Prof. Bhaskar Ramamurthi, Director, IIT Madras

Message from the Director

I am delighted to be a part of the third issue of Immerse. We can now para-phrase Ian Fleming and aver that with this issue, Immerse is neither happenstancenor co-incidence – it is here to stay. Glancing through some of the articles in thisissue, I am happy to note that they retain reader interest by combining an “xxxxmade simple” approach to science and engineering, along with glimpses of the peo-ple behind the research. The photographs and sketches bring the ideas to life. Theall-to-brief cameos about the authors add garnish to the articles. This issue, morethan ever, brings out the extent to which a vibrant research culture has embeddeditself at IIT Madras. I am grateful to GE for supporting this praiseworthy ventureof our students.

i| Immerse

From the Editors

Second row (L to R): Sanket, Nithin, and SwetamberFirst row (L to R): Kiranmayi, Raghavi, and Rohit

Greetings from Team Immerse!

Immerse is the IIT Madras Magazine on Research in Science and Engineering. Our en-deavour is not only to showcase some of the recent developments in research and innovationat IIT Madras, but also to communicate the science behind them in the simplest way possiblefor better understanding and appreciation.

This year marks the third year of Immerse. Our cover page this year is in celebration of2015 - UNESCO’s International Year of Light. Just like the seven colours of light, this year’sissue covers a spectrum of seven themes - communication, computing, economics, health,energy, defense and materials. In each of these features, besides the technical and emotionalfacets of research at IITM – from the faculty’s passion and expertise to the research scholars’enthusiasm and perseverance – we have tried to share with you our own fascination for theadmirable work going on.

We thank the Director and the Dean of Students for their constant encouragement. Wegratefully acknowledge our sponsor GE. We save our most special thanks for all the profes-sors and students who have been gracious and generous with their time and effort whileacquainting us with their work.

Above all, we will consider our job well done if this issue rekindles the curiosity of evenone reader, inspiring them to indulge themselves in the exciting world of science and engi-neering. Carl Sagan once said, “Somewhere, something incredible is waiting to be known”,to which we add – “in the following pages.” Immerse yourself!

ii| Immerse

IMMERSEIIT Madras Magazine on Research in Science and Engineering

www.t5eiitm.org/immerse

For Immerse

Editors Kiranmayi MalapakaNithin RamesanRaghavi KodatiRohit Parasnis

Sanket WaniSwetamber Das

Contributors Akshay GovindarajAnanth Sundararaman

Aparnna SureshArundhathi Krishnan

Aryendra SharmaAslamah Rahiman

Ayyappadas AMTejdeep Reddy

Isha BhallamudiNikhil MulintiRahul VadagaSachin Nayak

Shivani Guptasarma

Consulting Editor Nithyanand Rao

Photographs Vivekanandan N

www.shaastra.org

Design Amritha ElangovanSree Ram Sai

Vishal UpendranRaghavi Kodati

Typesetting Nithin RamesanRohit Parasnis

Sanket WaniSwetamber Das

For Shaastra

Sponsorship Raghul ManoshBhavik Rasyara

Shashanka S RaoMahesh Kurup

Except the images, which belong to the respective copyright holders as identified, all content is ©The Fi�th Estate, IIT Madras.This magazine was typeset entirely in LATEX.

Industrial Internet: Bringing Big Iron and Big Data forBetter Quality of Life

Vinay JammuTechnology Leader - Asset Performance Analytics, Global Research,

GE India Technology Centre, Whitefield, Bangalore-560066

Connectivity and computing power have been changing our world in a significant way over the past twodecades. E�ficiencies of consumer oriented industries such as retail, banking, transportation, and hospitalityhave been transformed. Today we carry the same computing power in our cell phone as the supercomputer ofthe ����s. The Consumer Internet has enabled �+ billion people to be connected to the internet to exchange in-formation and get improved services. This has been enabled by apps such as Uber, Amazon, iTunes, which con-nect the service provider directly to the consumer through digitisation, eliminating intermediary processesthat drive ine�ficiencies. New Consumer Internet giants such as Amazon, Apple and Google were born in thepast couple of decades and grew significantly by adding $��� billion in new revenues in just one decade.

The Industrial Internet will have bigger impact by improving e�ficiencies in industries such as power gen-eration and distribution, healthcare, transportation and manufacturing, to name a few. Improved e�ficien-cies in these industries means lower cost and better quality for healthcare services for all of us, more reliable,uninterrupted power, lower cost of travelling and reduced emissions and greenhouse gases. The IndustrialInternet connects big iron such as jet engines with big data to optimise the performance of these big industrialassets thereby saving crores of rupees for the owners, operations and users. For example, a �% improvementin India's power grid would amount to additional capacity of � gigawatt of power equal to � coal plants of ���megawatt capacity. It saves the cost of adding these plants and eliminates the greenhouse gas emissions thatwould have come from these plants.

To enable these e�ficiencies arising from the Industrial Internet, new technologies are needed. New sens-ing technologies that can provide measurement of critical parameters such as temperatures and pressure injet engines are needed. Low cost, reliable and secure communication technologies to transmit large amountsof data from mobile and remote assets operating in harsh environments would be a key enabler for the In-dustrial Internet. Finally, cloud enabled technologies to manage the data and perform advanced analytics tooptimise performance of these machines, extend the life of the critical parts, predict and prevent unscheduledoutages and failures, optimise parts inventory, perform the right maintenance at the right time with the leastamount of time and resources. To achieve the best e�ficiencies of scale, end-to-end solutions are needed thatcan improve e�ficiencies from fuel production, fuel transportation, electrical generation, transmission anddistribution and electricity use as an example for power production. This requires an ecosystem of interoper-able solutions that can be stitched together for di�ferent applications.

iv| GE

To support this broader vision, GE is working on two technologies that enable delivery of value that theIndustrial Internet promises. GE's Predix platform is the first cloud-based open Industrial Internet platformdesigned to host industrial data and industrial applications (https://www.geso�tware.com/predix). It is anopen platform that enables combining solutions from di�ferent companies to solve Industrial Internet e�fi-ciency challenges. The second technology GE is working on is the Digital Twin. Digital Twin is a digital cloneof a physical asset such as a jet engine or MRI machine that provides key insights and recommendations onimproving e�ficiencies of the assets. A Digital Twin can provide key insights into remaining useful life of partsin industrial equipment, prognose and prevent impending failures and provide recommendations to opti-mise its performance. To enable this Industrial Internet revolution, GE along with IBM, AT&T, Cisco, Intelhas founded the Industrial Internet Consortium (http://www.iiconsortium.org/) with the vision of setting thearchitectural framework and direction for the Industrial Internet.

The Industrial Internet will be the next technology wave to revolutionise e�ficiencies of industrial pro-cesses and provide better quality of life for all people. ⌅

v| Immerse

“It is an experience like no other experience I can describe, the best thing

that can happen to a scientist, realising that something that’s happened in

his or her mind exactly corresponds to something that happens in nature. It’s

startling every time it occurs. One is surprised that a construct of one’s mind

can actually be realised in the honest-to-goodness world out there. A great

shock, a great, great joy.”

- LEO KADANOFF (1937-2015)

India shares a large tract of mountainousborders with multiple countries, manningwhich is a nightmare for the army. Notonly is it di�ficult to build roads at high

altitudes, but the army also has to contend withlandslides and avalanches while traversing theseroads. Naturally, one thinks, “Why not fly ratherthan go from one place to another on the ground?”

Yes, we are talking about jetpacks - devicesworn on the back that allow a single user to fly bymeans of propulsion produced by rapidly expelledgases. Jetpacks are for real now but they are nothingclose to the kind you see in science fiction movies.They have been used only for ceremonial purposesand no armed force in the world currently uses it.We talk to Major Lakshyajeet Singh Chauhan, anengineer in the Indian Army, about his researchproject on developing a rocket backpack motor forthe Indian Army.

Maj. LS Chauhan details the hardships facedby the Indian Army in remote locations throughhis own experience of working in the army for �years. He says, “I was posted in the Siachen glacieras part of my infantry attachment. We used towitness sudden snowfalls and all the routes used tobe blocked for days at a stretch. It is impossible toland a helicopter at such altitudes. If someone wasill, there was no way you could take him out of therein such conditions.” He goes on to say that due toruthless weather, he had to spend a much longertime at the Siachen than he was designated to.

Jetpacks can be a real boon for the Indian Army.One can go to remote places without any hassle.Consider the same example of the Indian Armyguarding mountain borders. If a couple of peoplein a company can fly, they can build a ropewaybetween the cli�fs of two mountains, instead of

the entire company coming down and climbingup again. The time taken to go from one peak toanother is reduced. With this idea in mind, Maj. LSChauhan started his M.Tech. project at IIT Madras.He discussed it with his guide Prof. PA Ramakrishnafrom the Department of Aeronautical Engineering,who agreed to work with him on this and give himthe required guidance. Prof. Ramakrishna explains,“It is more like this... if you see monkeys climbingtrees to get at the fruit, you will observe that theywill first climb one tree and jump from one tree toanother. People working in such fields do the same.It is possible in this terrain too. It is basically thesame thing. One somehow manages to climb onepeak and uses that advantage to go to another. Thiswas the idea with which we started this work.”

Yes, we are talking about jetpacks -devices worn on the back that allow a

single user to fly by means of propulsionproduced by rapidly expelled gases.

Rocket backpacks have been around forsometime now but they have been only used forfun and recreation. They have a very low flighttime and can just about carry a person. This projectaims to build a portable one that can carry a soldierand his equipment. There was an attempt by BellHelicopters to make one such device for the USMilitary. They built a device called the Bell RocketBelt which was eventually rejected because it wasexpensive and there was no scope for carryingpayload.

Let us take apart the rocket used in a jetpack.In the simplest terms, a rocket has a combustionchamber in which fuel comes in contact with achemical called the oxidiser that helps the fuel burn.

Major Lakshyajeet Singh Chauhan was an M.Tech. student (2013-2015) in the Depart-

ment of Aerospace Engineering at IIT Madras. Before his Masters, he worked for 9 years

for the Indian Army and was posted in Siachen and Kargil, among other places. He is

currently an instructor in the Faculty of Aeronautical Engineering at the Military College

of Electronics and Mechanical Engineering, Secunderabad.

�| Aerospace Engineering

Jetpacks — A Boon for the Indian Army

Surprise Attack in the Mountains:A quick climb onto the mountains will help carry outa surprise attack on the enemy who is still strugglingto climb up. In general practice, the Indian Army usesmountaineering equipment to climb mountains forsurprise attacks.

Counter Insurgency Operations:A quick li�t-o�f with the rocket backpack while chasing amilitant will enable a soldier to take a shot at the enemyor guide his forces suitably.

NSG (National Security Guard) Operations:The NSG operation on the Taj hotel during ��/�� sawimmense involvement of helicopters. Jetpacks willenable NSG commandos to easily reach the roofs of theestablishments under attack.

Attack Operations:It is di�ficult to attack the enemy from the front, whenthey have laid down mines and put barbed wire. Rocketbackpacks can be used to jump over these obstacles andcharge on the enemy.

High Altitude Operations:The Indian Army has been deployed at Siachen - thehighest battlefield in the world. Jetpacks will help relocatefrom one post to another, get supplies and evacuatecasualties. During attacks, it will help in raiding enemyposts located at peaks and in faster deployment of troops.

Recovery Operations:Jetpacks can be utilised to recover fallen vehicles, attendto casualties, recover important documents, etc. They willalso be e�fective in reaching remote locations and assistarmed forces in withdrawal from covert operations.

What they (the Indian Army) have is the following: when they enter enemy territory, the last ��� metresare a minefield guarded by the entrenched enemy. They face huge casualties while running through this.It is a throw of dice - there’s a high probability that one might step on a mine and lose one’s leg or lose one’slife. With this rocket backpack motor, one can catapult oneself over the last ��� metres and then take onthe enemy.

�| Immerse

Prof. PA Ramakrishna obtained his PhD from the Indian Institute of Science, Bangalore

and joined IIT Madras in 2005. Currently, he is working as a Professor in the Department

of Aerospace Engineering. His research interests are in aerospace propulsion, especially,

solid and hybrid propellant combustion.

The burning of the fuel and the oxidiser controlsthe thrust. Controlling this thrust is crucial tocontrol the motion of the person wearing the rocketbackpack. Initially, for one to go up, the thrust has begreater than the pilot’s weight. And when the sameperson has to come down, he has to reduce his thrustso that it is less than his weight.

Maj. LS Chauhan compares hisexperiences of working in the army withthat of working on the research project.

The type of fuel and oxidiser involved is used toclassify the rocket. For example, in a liquid rocket,both the fuel and oxidiser are liquids while in a solidrocket, both of them are solids. The liquid rocketused in the Bell Rocket Belt had Hydrogen peroxideas its oxidiser which was hard to handle since it wasexplosive and ate through metal.

Maj. LS Chauhan and Prof. Ramakrishna werelooking for something more benign than that. Theywanted hybrid rockets to do the job for them. Suchrockets have a liquid oxidiser and a solid fuel. Theirprimary advantage is that you can control the flowrate of oxygen and get the thrust level you require.This feature also exists with liquid rockets but notwith solid rockets. But it is easier to control thrustin hybrid rockets than in liquid rockets as only oneout of the fuel or the oxidiser is a liquid. Moreover,hybrid rockets are known to be very safe. Prof.Ramakrishna’s lab has been working on them forover �� years now. They have conducted over ����experiments on them without any mishaps. It isimportant to take safety into consideration whenthere is a human being on the other side. They plan

to use water as the oxidiser in this system as it is notonly very safe but also readily available.

Maj. LS Chauhan compares his experiencesof working in the army with that of working onthe research project. He says that he was onlydoing a maintenance job in the army. A�ter joiningacademia, he had to put his mind back into thethinking mode. He describes his experience ofworking for �-� months on something new andinnovative as wonderful. The work was as strenuousas the army where he had to spend ��-�� hoursevery week. Yet, he says that he did not really mindworking long hours as the work was very interesting.He feels that failure is a part of research. “Eventhough I put in my best e�forts, something or theother kept going wrong. Sometimes, I thought I wason the wrong track. But, I kept going and eventuallyfinished the propulsion part of the project and evendesigned the combustion chamber for the rocketbackpack motor.” The di�ferent stages of the projectare illustrated through the flowchart below.

�| Aerospace Engineering

Ignition Continuous combustion

Experimental set-up

He also acknowledges the need for a goodmentor in a research project with the followingwords: “My professor helped me out a lot. Onlythe idea was mine. Most of the thinking and thedesign was done by him. I was only followinghis instructions.” One might ask the followingquestion: “Bell Helicopters was a large organisationdeveloping the rocket belt. How can a few studentsin the lab do it?”, to which Prof. Ramakrishna hasa ready reply. “There are no technical di�ficultiesassociated with the project. We do have all thetechnology to develop it. Only its implementationand engineering within the weight budget and safetyconsiderations is hard. All these things need to be

taken care of and the product needs to be broughtout.”

As a consequence of Maj. LS Chauhan’s work,the proposal of inclusion of the project under theArmy Technology Board has been forwarded by theMilitary College of EME, where Maj. LS Chauhan isposted currently, to IIT Madras. This collaborativeproject worth |�� lakhs is expected to start inJanuary ����. It aims to develop a rocket backpacksystem capable of carrying a soldier for at least oneminute at a cruise speed of �� km/hr. All thisneeds to be done within a budget of |�� lakhs perbackpack to ensure that the large requirement forsuch backpacks can be met at a reasonable cost.

�| Immerse

Despite the availability of advanced technologyand the required funding, the main challenge ofthe human body not being adapted to fly naturallyremains an obstacle in this endeavour. The jetpackmust accommodate for all factors of flight such assu�ficient li�t and to some extent, stabilisation. And,of course, the soldier operating the jetpack must betaught to fly it.

In conclusion, Maj. LS Chauhan wishes to givethe following message to the students – “I wish good

engineers and scientists would stay in India andwork for the country instead of going abroad justfor the sake of a higher pay. Being Indians, weshould not leave our country. The joy of getting afat paycheck is nothing compared to the satisfactionderived from serving the country, say by working inthe army. My dear Indians, please stay back, work forthe development of India and make it a strong anddeveloped country.”⌅

The images are from Maj. LS Chauhan’s M.Tech. thesis.

Meet the Author

Sachin Nayak is a final year B.Tech. student of the Department of Electrical Engineering

at IIT Madras. He loves microcontrollers, coding, swimming, running and, of course,

reading and writing. He has been part of several organisations in the institute like the

CFI Electronics Club, The Fifth Estate, Shaastra, etc. He plans to pursue graduate studies

in Computer Science in the near future.

�| Aerospace Engineering

Through the Looking Glass

By Ayyappadas A M

How digital photoelasticity research empowers the mobile revolution

Imagine a research group consisting ofmechanical engineering graduates andheaded by a world renowned expert inthe subject area of experimental stress

analysis. Could they be silently empowering themobile, IT revolution and technology convergence,by doing engineering research from their subjectdomains? This is the story about such a technologybridge, made possible through furthering theunderstanding of science and its applications. Theirstory o�ten goes unsung, although their work shinesout from the smiles you capture through Instagramor the high quality videos played from your BluRayDVD. We will see how diverse technologies suchas image processing, numerical computing, andstress analysis have come together to help the massproduction of optical devices.

Let there be lightThe history of science and technology is replete

with examples of cross-disciplinary applicationof scientific principles which have acceleratedtechnological developments. In this case thecommon thread happened to be light. In fact, it isinteresting to note that the phenomena which finallycame to draw the demarcation line between classicaland modern physics also links di�ferent disciplines.The digital photoelasticity research group, headedby Prof. K Ramesh from the Department of AppliedMechanics, IIT Madras, are the people behind thisimpressive feat. Their journey began through theIndo-European Union project under FP� initiativeknown as SimuGlass. They started o�f with a straightforward, though not necessarily simple, objective - tovalidate the results from a finite element simulation

�| Immerse

Dr. K Ramesh is currently a professor at the Department of Applied Mechanics, IIT

Madras. He was its Chairman during (2005-2009) and formerly a professor at the De-

partment of Mechanical Engineering, IIT Kanpur. In recognition of his significant con-

tributions in photoelastic coatings, the F. Zandman award was conferred on him by The

International Society of Experimental Mechanics in 2012. He is a Fellow of the Indian

National Academy of Engineering (2006).

tool developed for predicting the residual stressesproduced in optical lenses during the manufacturingprocess known as precision glass moulding. Thechallenges faced during their research made themrise to the occasion by unravelling importantinsights about the physics behind the phenomenaof residual stress in moulded lens. Their work hasthe potential to make the production of opticallenses much more e�ficient, adding to the mobilerevolution.

Prof. K Ramesh has been involved withresearch in digital photoelasticity for more thantwo decades. He has produced milestone papersin the subject area, and is recognised as anauthority in his field. Therefore, it is not surprisingthat he is a key collaborator in a high impactindustry project involving European and Indianinstitutes and industrial partners. The Europeanpartners of the SimuGlass project consisted of twoacademic research institutes - Fraunhofer Institutefor Production Technology IPT and Centre deRecherches de l’Industrie Belge de la Ceramiqueand two industry partners – Kaleido Technologyand EcoGlass, while the Indian partners includeCentral Glass and Ceramic Research Institute(CGCRI), Indian Institute of Technology Madras(IITM), Indian Institute of Technology Delhi (IITD)and Bharat Electronics Limited (industry partner).The Indian wing of the project was headed byDr. Dipayan Sanyal from CGCRI. The stated aimof the project was to enhance the understandingof the precision glass moulding process and toincrease the quality of the process by developing anintegrated Finite Element tool. It was also envisaged

that through the project “the state-of-the-art ofmanufacturing in India and parts of Europe, whichfollows the grinding and polishing route, will bereplaced with the advanced precision press formingroute, especially for manufacturing the advancedoptical elements.”

Their work has the potential to make theproduction of optical lenses much more

e�ficient, adding to the mobile revolution.

Precision Glass MouldingAs a collaborating institute, the role of IIT

Madras was initially limited to the measurement ofresidual stresses in optical lens using methods fromphotoelasticity and validation of the finite elementmethod tool for process optimisation. Tarkes DoraP and Vivek Ramakrishnan, PhD students workingwith Prof. K Ramesh at the Digital PhotoelasticityLab, Department of Applied Mechanics, have beeninvolved with the project from the time they joinedthe Indian Institute of Technology Madras. Tarkes,whose PhD work is almost exclusively based on theresearch done under the auspices of the project,had this to say: “you know, it was like solving along puzzle, by taking one challenge a�ter another indiverse fields, until a larger picture emerged.”

So what is precision glass moulding and whyis it relevant? Tarkes replies, “The conventionalmanufacturing process of optical lenses is amulti-step and highly time consuming process.And is that all? “No”, Tarkes continues, “Thedigital revolution has unleashed a new market forsuch precision glasses. In today’s world, everyone

�| Applied Mechanics

Steps involved in Precision Glass Moulding

with a mobile phone is an amateur photographer.The convergence has come to the level that oneor more optical devices have become standard inmost consumer electronics, and particularly in thecommunication devices. The demand for thesedevices has been exponential in the past decade.This has created a situation where dependency onconventional production process involving grindingand polishing has become virtually impossible.”

It was like solving a long puzzle, bytaking one challenge a�ter another indiverse fields, until a larger picture

emerged.

Precision glass moulding involves five steps as

shown in the figure. First the cold glass blankwith a defined geometry, called gob, is loadedinto the mould. The oxygen is removed from theworking area, followed by nitrogen filling. Thewhole system is then heated. The variables thata�fect the quality of the product - temperatureand the force applied come a�ter this stage. Theglass is heated to a point slightly above what isknown as glass transition temperature. Glasstransition temperature is the temperature at whichglass changes from a hard and relatively brittlestate into a molten or rubber-like state. The valueof this depends on the composition of the glass.Optical glasses have low transition temperatures,typically between ���°C and ���°C. The mouldquality degrades due to wear and tear if glasses withhigher transition temperature (���-���°C) are used.

�| Immerse

A�ter heating, the glass gob is pressed to the desiredshape. The whole assembly of lens and mould iscooled to room temperature. The final lens can bedirectly used for the desired application.

Residual stress is the internal stressdistribution locked into a material. These stressesare present even a�ter all external loading forceshave been removed. Residual stresses are developedin the lens during the cooling. In conventionalmethod, the glass lens is annealed, (a heat processthat involves slow cooling) intermediately. Thisminimises the residual stresses. There is also theissue of shape deviation in the lens. Repeated trialand error experimentation, in order to minimise theresidual stresses and shape deviation is prohibitivelyexpensive, given that a large cost is involved in themanufacturing of the moulding tools. Very fewcompanies such as Aixtooling GmbH produce theultra-precise high quality moulding tools. Thus,numerical computation with finite element methodhas become the obvious choice to get the optimisedshape of the mould and for selection of thermalcycles.

Manufacturing meets photoelasticity

As soon as the team at IIT Madras took aplunge into the problem, they encountered severalhurdles. It was one thing to do photoelasticmeasurements with plastic materials. Opticalglasses were, however, a totally di�ferent territory.The fringe widths to be measured were in micronsas compared to millimetres, and the requirement ofprecision was also higher.

Glass is a weakly birefringent material. Thefringe order usually observed is less than one.Measurement of birefringence in glass up to �nanometers is possible using an instrument calledautomatic polariscope (for e.g.: manufactured byM/s GlassStress Ltd., Estonia which uses phaseshi�ting technique). There were ‘low’ and ’high’photoelastic constant glasses, whose stresses are tobe measured. The accuracy of measurement su�ferswhen the range of values go from one extreme to the

other, while using the same process. They overcamethe precision related hurdles by devising twoexperimentation techniques. The team proposed anew experimental technique using Carrier FringeMethod for photoelastic calibration of glass forhigher values of photoelastic constants. Also forrelatively low photoelastic constant materials, anew experimental method involving phase shi�tingtechnique was devised.

The Finite Element Method based Pro-cess Optimisation

The accuracy of the final product is ofutmost importance as far as precision moulding isconcerned. The conventional method for obtaininghigh accuracy has been based on trial and error,and domain expertise of the process engineer. TheSimuGlass envisaged at changing this paradigm.And indeed, they found a breakthrough. A simplermethod to obtain the optimised design of the mouldwith submicron-form accuracy (shape deviation < �micron) was proposed by the IITD and IITM teamsin ����.

Bench-Top Precision Glass Moulding ApparatusSource: PGPL, Isfahan University of Technology

Finite Element scheme for simulation andanalysis of the moulding process involves severalother challenges. The thermo-mechanics of glassinvolves two aspects - viscoelasticity and structuralrelaxation. Viscoelasticity models are available incommercial so�tware packages like ABAQUS. Thestructural relaxation has to deal with the nonlinearthermal expansion. There was no standard method

��| Applied Mechanics

PhotoelasticityPhotoelasticity is an experimental method to determine the stress distribution in a material. It isadvantageous because of its ability to give a fairly accurate picture of stress distribution, even aroundabrupt discontinuities in materials. A typical photoelastic measurement setup consists of a light source, apolariscope, a transparent material which is to be analysed, and an image capturing device. First describedby the Scottish physicist David Brewster, the method has come a long way to become a very important toolin engineering and research.

available in commercial packages for solving thisproblem with su�ficient accuracy. Tarkes fromthe IIT Madras team in collaboration with IITDTeam developed the code for solving the structuralrelaxation problem.

At the beginning of the project the use of digitalphotoelasticity was only envisaged as a validationtool. But soon this was to change. The key issuewas that the physics of heat transfer involved inthe moulding process was not well understood.This is to say that the heat transfer mechanismbetween the mould and glass, the glass and thesurrounding N2 atmosphere should be known forsure, if the numerical computation results were tobe reliable. The experiments were conducted inthe facility at Fraunhofer Institute for ProductionTechnology, Germany. When the results from thefinite element model were compared, it was evidentthat the assumptions made about the heat transfermechanism for glass and N2 interaction was farfrom accurate. This prompted them to relook theproblem from the heat transfer point of view. Theydid a computational fluid dynamics simulationof the cooling stage of the moulding process.This revealed that the mechanism was not likelyconvection, as it was believed or assumed previously.

Tarkes analysed the temperature measurementdata and paved the way to account the actualheat transfer mechanism. They ended up usingdigital photoelasticity values as a significant inputparameter by which they are able to predict moreaccurate values to be used in the finite simulation toaccount this heat transfer mechanism.

The Bottom Line

In business everything comes down to a bottomline. I asked the team how their work fits into thelarger frame. And this is what they had to say: “theaccuracy of the simulation has been of the order of��-��%. The work done by our group has potentiallyimproved this to ��-��%. Most importantly, thescience of the process is now better understood. Thediscovery that the cooling of the lens in the precisionglass moulding does not happen through convectionis a breakthrough. An increment of accuracyby ��% is one significant step towards massiveimprovement in the mass production levels.” Inshort, thanks to such collaborations, all of usare able to get higher resolution cameras in ourmobiles. More importantly, this will go a long wayin improving the quality of optical lenses, which hasa direct impact on clinical fields. ⌅

Meet the Author

Ayyappadas A M is a PhD scholar working in Fluid Dynamics, at the Department of AppliedMechanics. Apart from the science-y stuff he does, by way of research and personalinterest, he is an avid reader of philosophy and history. Infrequent blogger, but mostlyharmless.

��| Immerse

Iremember my first visit to the Himalayas,standing thousands of feet above sealevel. It was a spectacular view – densegreen forests, massive rock cli�fs, chains

of silvered peaks and grasslands on the banksof the Mandakini river that ran deep and silent.Just imagining that the ground below me whilestanding there starts shaking at an accelerationapproximately equal to one-fourth of that of a freefalling object gives me goosebumps. Althoughqualitatively similar, it certainly is not a ride atan amusement park. This exact phenomenon hashappened in the region around Kedarnath templemultiple times, the most recent one being in ����.

The temple is located near the ChorabariGlacier, one of whose two noses is the source ofMandakini River. This terminates at Chorabari Lakewhich is approximately ��� metres long, ��� metreswide and ��-�� metres deep and is situated about �kilometres upstream from Kedarnath. On June ��,���� heavy rainfall together with the melting of thesnow from the glacier led to the complete drainingof the lake within a matter of minutes. Once the lakehad burst, the water carried along mud and debrisdown the valley and caused massive devastation tothe entire town downstream.

Heritage conservation doesn’t meanfreezing a building in time or creating a

museum.

However, there was no major damage tothe temple structure. It is immensely strong,made of thick, massive granites and high-grade’metamorphic gneissic’ rock slabs, pillars and bricks.Simply put, these are rocks that changed form dueto heat and pressure while buried deep below theearth’s surface, have banded appearance and aremade up of mineral grains, typically quartz minerals.They have a bright sheen, are rough to touch and veryhard. The walls pillars are about �.� metres thick. Theroof is an assemblage of multiple blocks with dressedstone on the exterior. This is one of the majorreasons that the temple withstood the earthquakes.

The survival of the temple can also be attributedto the presence of a man-made platform whichraises the temple super-structure that prevented thetemple from the direct flow of gushing floodwater.

Chorabari lake and Kedarnath town.Courtesy: Dr. Menon

India has one of the largest stocks of heritagestructures in the world out of which �� are formallyrecognised by the United Nations Educational,Scientific and Cultural Organisation (UNESCO).Of these sites, �� are cultural properites and� are natural properties. Formal systems thatrecognise conservation of heritage structures as amultidisciplinary engineering e�fort do not exist inIndia. Heritage conservation doesn’t mean freezinga building in time or creating a museum. Instead,it seeks to maintain and thereby increase the valueof buildings by keeping their original architecturalelements, favouring their restoration rather thanreplacement and, when restoration is impossible,recreating scale, period and character. Addressingthe task of understanding and protecting heritagestructures from natural hazards, ageing andweathering e�fects is a serious problem in Indiasupplemented by the lack of adequate quality andquantity of manpower.

��| Immerse

Dr. Arun Menon is an Assistant Professor at the Department of Civil Engineering at IIT

Madras, where he has been since 2010. He also currently serves as the Convener of

National Centre for Safety of Heritage Structures (NCSHS). His research interests center

on structural conservation of historical monuments which include seismic response, as-

sessment and retrofitting of masonry structures, historical seismicity and seismic hazard

analysis. He received his M.Tech. in Civil Engineering from IIT Madras. He received his

M.Sc. and PhD from ROSE School, University of Pavia, Italy.

With the intent of beginning a formal approachto address the safety of heritage structures, theNational Centre for Safety of Heritage Structures(NCSHS) was established at IIT Madras in July ����with Dr. Arun Menon, a professor in the StructuralEngineering Laboratory at the Department of CivilEngineering, designated as the Convener. Dr.Menon tells me that NCSHS is envisioned as along-term programme towards addressing thechallenge of ensuring structural safety of historicalmonuments and other heritage structures in India.The plan is to collaborate with implementingagencies such as Archaeological Survey of India(ASI) which would help in fundamental research andeducation.

Most historical structures have multi-leafwalls which are composed by one or more

external and internal leaves.

Dr. Menon’s primary research interest isthe seismic vulnerability assessment of buildingstructures. The Indian subcontinent has a highseismicity - the frequency of earthquakes in a region.The country has been divided into � seismic zones(Zone �, �, � and �) where Zone � and Zone � expecthighest and lowest levels of seismicity. Kedarnathtemple, situated on the Garhwal Himalayan rangenear the Mandakini river in Uttarakhand, standsin Zone �. Its neighbourhood has seen severalearthquakes in the recent past such as in ����(Uttarakashi) and ���� (Chamoli). Again, in majorityof the cases, no major damage to the templestructure was reported.

The temple is believed to have been built in the�th century A.D. in Nagara architectural style. TheShastras, the ancient texts on architecture, classifytemples into three di�ferent orders; the Nagara or‘northern’ style, the Dravida or ‘southern’ style, andthe Vesara or hybrid style which is seen in the Deccanbetween the other two. The Nagara style’s primaryfeature is a central tower (shikhara) whose highestpoint is directly over the temple’s primary deity.This is o�ten surrounded by smaller, subsidiarytowers (urushringa) and intermediate towers; thesenaturally draw the eye up to the highest point, like aseries of hills leading to a distant peak. Setting thetemple on a raised base (adhisthana) also shi�ts theeye upward, and promotes this vertical quality.

The type of damage in a masonry wall dependson the relative alignment between the direction ofthe ground shaking and the wall. If the groundshaking is perpendicular to the wall, the mostcommonly found cracks are vertical and if theground shaking is parallel to the wall, the cracks areusually diagonal.

Out–of–plane collapse mechanism.Courtesy: Dr. Menon

Both of these lead to formation of di�ferentfailure mechanisms – analysis of defect in design,quality and other parameters which led to the failure

��| Civil Engineering

of the process. In the former case, an out–of–planecollapse mechanism is observed in which case thewall leaves may be detached or the entire wall mayoverturn. In the latter case, an in-plane-shearmechanism is realised and the entire column snaps.

Masonry structures typically show two typesof mechanisms under earthquake ground shaking,namely global mechanisms and local mechanisms.Global mechanisms, which are typically in-planeshear response of structural walls, occur whenthe masonry structure is constituted of wallsand floors/roofs that are well connected to eachother. In-plane shear response in masonrywalls is characterised by diagonal (x-cracks) orhorizontal cracks. In most ancient masonryconstructions, connections between structuralelements and between roofs/floors and walls arepoor, and these lead to out-of-plane mechanisms,or local mechanisms. Out-of-plane mechanismsare characterised by out-of-plane bulging/pushout or overturning of walls, parapets and otherfree-standing elements.

The project started in ���� with the aim ofrestoring the damaged parts of the temple andpreventing damage against future events. Thereare a lot of challenges in structural assessments ofheritage structures. Prior to a physical inspection ofa historic structure as much information as possibleabout the structure must be gathered. Not only is itimportant to understand the actual structure of thehistoric building in question, but also the times inwhich it was built.

Multi-leaf walls.Courtesy: Dr. Menon

Most historical structures have multi-leaf wallswhich are composed by one or more external andinternal leaves. External leaves contain stoneworkor brickwork and internal leaves are usually made ofrubble masonry or very weak infill materials, such asearth or loose material.

Dr. Menon mentions the lack of materialcharacterisation of the inner and outer leaves ofthese walls. For example, the exterior and interiorof the wall at Kedarnath temple are gneiss stoneleaves and the infill is rubble stone, which is irregularin shape, size and structure. Since these are notappropriately characterised, it is di�ficult to carryout further investigation. Other challenges includeunavailability of proper geometrical/architecturaldrawings and poor understanding of ancientconstruction practices, especially the sequence ofconstruction: contact and connection betweenexternal and internal leaves.

Other challenges include unavailability ofproper geometrical/architectural drawings

and poor understanding of ancientconstruction practices.

Dr. Menon and his team have visited the templefour times between June ���� and June ���� forgeophysical studies, structural investigations andpreliminary structural health monitoring. Thereis a smile on his face as he mentions that theywere allowed to enter the temple premises duringthe nights a�ter all the pilgrims had le�t, througha small gate at the back. During their visits,the team used a lot of interesting equipment foranalysis. An endoscope was used to analyse thewalls of the temple and the inner structure wasfound to contain voids in the core masonry. Theendoscope is very similar to the one a doctor uses tocheck the interior of the body. Another instrument,an infrared camera, which takes pictures of theradiation conditions, was used. It is similar toa common camera that forms an image usingvisible light. Instead, infrared camera forms imagesusing infrared radiations. By recording surface

��| Immerse

Courtesy: Dr. Menon

temperatures under di�ferent exposure conditions,hidden voids and cavities were detected in thetemple structure.

Structural analysis of a structure is done byemploying various techniques. Di�ferent forces,deformations and accelerations are applied to thestructure and its components to assess their e�fectssince excess load may cause structural failure. Theseapplied forces are called loads. One type of load isa dead load which includes loads that are relativelyconstant over time, including the weight of thestructure itself, and immovable structures such aswalls and plasterboards. The technique of subjectingthe structure to dead load to assess its e�fect is calledGravity Load Analysis. When this was used on thetemple, it was concluded that the structure has veryhigh safety margin against gravity loads.

...the inner structure was found to bedeteriorating ...hidden voids and cavities

were detected in the temple structure.

Another load which the temple’s structure wassubjected to is the lateral load. These loads arelive loads (temporary or of short duration such asthe load due to wind) whose main component is ahorizontal force acting on the structure. Most lateral

loads vary in intensity depending on the structure’sgeographic location, structural materials, height andshape. Dr. Menon’s team considered hydrostaticpressure to complete the lateral load analysis andfound that the pressure has no damaging e�fecton the structure. And that the structure is safeagainst hydrostatic pressure. The temple wasalso subjected to gravity loading and a monotonicdisplacement-controlled lateral load pattern whichcontinuously increases through elastic and inelasticbehaviour until an ultimate condition is reached.This method of analysis is called Pushover Analysis.And again, it was concluded that the structure issafe against such loads. The pushover analysis is amethod to establish the capacity of the structure.Pushover-based seismic assessment (comparingdemands to capacity), showed that the structure wassafe against lateral loads.

The magnitude of acceleration mentioned inthe first paragraph is a measure of how hard theearth shakes at a given geographic point. It is knownas the Peak Ground Acceleration (PGA). A�ter all theanalysis, it was found that the temple structure issafe against earthquake ground motion below �.�g.It was also found that the timber sloped roof over theSabhaMandapa and inner mandapa (mandapas arehalls in the temple) is a poor quality construction and

��| Civil Engineering

is vulnerable to earthquake shaking.

The front gable wall (the triangular portionof a wall between the edges of intersecting roofs)has shown significant dislocation of stone blocks.To safeguard it from future earthquakes thereconstruction of the stone masonry supportingwalls of the truss and gable walls has been proposed.The proposal includes introduction of timber bandbeams with timber cross ties along the four sides ofthe SabhaMandapa above the stone wall to ensureintegral action of the entire structure in the eventof earthquake shaking. The gable walls would alsobe provided with band beams to ensure greaterout-of-plane resistance that is required underearthquake shaking.

Prof. V Kamakoti from the Department ofComputer Science and Engineering at IIT Madrashas worked on a wireless sensor network forstructural health monitoring. It is desired tomonitor the inclination/tilt of the temple structureover a period of time. The process is automated andperiodic. The sensors will be installed at Kedarnathtemple and will send periodic readings every fi�teen

minutes to a remote server located at IIT Madras.It is engineered to wake up upon an impact of�.���g and is designed to work under sub-zerotemperatures up to –��oC. Network connectivityseems to be an issue. BSNL is being used whichworks from �� am to � pm and there is no certaintythat the power during the winter season when thetemple is shut down for six months will remain on.

The work has been challenging for the teambecause of unpleasant weather at the site which isstrikingly di�ferent from the weather in Chennaiwhere the team stays for better part of the year. Alot of work needs to be done in the northern regionaround Zone �. There is a ��� km central seismicgap in the Himalayan front which also includesUttarakhand. A recent study (����), claims that thereis su�ficient energy stored in the ongoing tectonicprocess which could generate earthquakes of highmagnitudes as high as touching � on the Richterscale. But the time of the event cannot be predicted.It might occur tomorrow, or �� years later. The onlycertain thing is it is going to happen and we need tobe prepared. ⌅

Meet the Author

Sanket Wani is a final year student pursuing B.Tech. in Chemical Engineering at IITMadras. He can usually be found browsing popular science content on the internet. Oflate, he has developed an interest in science writing. He also takes a keen interest ineating at fancy restaurants and watching football.

��| Immerse

Much of chemistry deals withthe study of the interactionsbetween di�ferent types ofmatter. Such interactions are

easily imaginable between two or more fluids as themolecules of one fluid have the freedom to occupythe empty spaces between the molecules of the other.But what happens when the candidates involvedare solids? Due to their physical constraints, twosolids do not interact until you provide them amedium that facilitates such interactions. Moreo�ten than not, these mediums are liquids calledsolvents, the excess liquid phase in which one ormore solids (now called solutes) are dissolved toform a homogenous solution. With a global marketworth billions of dollars, solvents are an essentialpart of many sectors of the economy, includingmanufacturing, processing and transportation.

Despite being critical in addressing some ofthe most important problems in chemistry researchas well as other challenges that society is currentlyfacing, solvents also raise many environmental,health and safety issues. Firstly, most solventscome from finite sources, such as petrochemical orfresh-water resources. To add to this, the life-cycle ofa solvent, right from its manufacture to its disposal,requires a considerable amount of energy input.Also, most of the traditional non-aqueous solvents,i.e., those other than water, such as benzene andtoluene, tend to be toxic and evaporate easily(also called volatile). This makes them di�ficult tostore and transport, and eventually poses threats ofatmospheric pollution and accumulation in livingorganisms. Given these drawbacks, there are severalchallenges that need to be overcome for the short-and long-term usage of solvents.

In the pursuit of more environmentally friendlysolvents, scientists stumbled upon a class ofmolecules called ionic liquids. Ionic liquids arenon-volatile molecules that seem to be able todissolve everything, and thus have the ability toreplace the conventional manufacturing medium.Sounds magical, doesn’t it? But as the great sciencefiction writer, Arthur C. Clarke pointed out – “Magicis nothing but science that we don’t understandyet.” And this magic is what Prof. Sanjib Senapati’sgroup at the Computational Biophysics laboratory,Department of Biotechnology, are trying to unravel– figuring out what makes ionic liquids (or ‘greensolvents’ as they have come to be known in recenttimes) a chemical possibility and gives them thesespecial powers.

With a global market worth billions ofdollars, solvents are an essential part ofmany sectors of the economy, including

manufacturing, processing andtransportation.

Ionic liquids have a very unique chemistry.They are composed of ions like all salts, but, unlikeconventional salts, such as sodium chloride, whichhave ions of comparable sizes, in an ionic liquid, theions di�fer a great deal in their sizes. The comparablesizes of the ions in a conventional salt lead to strongelectrostatic interactions between them, allowingthem to pack together uniformly, giving rise to aregular lattice structure. This does not happen inan ionic liquid, and therefore, they remain liquids atroom temperature.

Prof. Sanjib Senapati received his M.Sc. degree in Physical Chemistry from Univer-

sity of Calcutta, Kolkata. He obtained his PhD from IIT Kanpur. He has worked as

a Post Doctoral Fellow at University of North Carolina, Chapel Hill and University of

California, San Diego, USA. In 2005, Prof. Senapati joined the Department of Biotech-

nology, IITM where he is currently a full Professor. Apart for ionic liquids, his other

research interests include identifying drug targets and designing novel drugs for HIV

and heart disease.

��| Immerse

Prof. Sanjib Senapati with his research group in the Department of Biotechnology at IIT Madras.Courtesy: Prof. Sanjib Senapati

So, to put it in simple words, ionic liquidsare salts that remain liquids at room temperature.Interestingly, since the ions involved in ionic liquidsare large, a part of them is polar, or in otherwords, charged, whereas the other is non-polar.By rule, polar solvents only dissolve polar solutes,whereas, non-polar solvents only dissolve non-polarsolutes. Now since ionic liquids have both kindsof components, they have the advantage of beingamphiphilic – they can dissolve just about anything,even cellulose.

MD simulations allow one to see exactlywhat is happening at the molecular level

in the course of a reaction.

Prof. Sanjib Senapati, a theoretical chemist bytraining, focused on the application of chemistryin the field of biology during his post-doctoraldays. Here, he got introduced to the fascinatingfield of ionic liquids and has since been studyingthem keenly. His lab, recently declared one ofthe best performing bioinformatics facility inIndia by DBT Biotechnology Information SystemNetwork, approaches research problems usingstate-of-the-art molecular dynamics simulations.

By using structural parameters (properties of amolecule such as lengths, angles and planes betweenthe atoms) to define the interactions betweenmolecules and computer codes to replicate theconditions of a lab experiment (such as temperature,pH, salt concentration and atmospheric pressure),they are able to simulate what goes on inside atest tube on a computer screen. The advantageof simulations over test tube is quite evident;MD simulations allow one to see exactly what ishappening at the molecular level in the course ofa reaction.

Debostuti, one of the graduate studentsworking on ionic liquids at Prof. Sanjib’s lab isamused at the naivety of our query when we ask her,“But what have ionic liquids got to do with biology?”She replies, “Well, there is much more to ionic liquidsthan meets the eye. A�ter ionic liquids became asuper hit in the manufacturing industry, the nextbig step for the industry was quite obvious – to trytheir hands on ionic liquids in enzymatic catalysis.”Catalysis is the speeding up of a chemical reaction inthe presence of an additional participant called thecatalyst. When bio-molecules act as such catalysts,the bio-molecules are termed as enzymes and thereaction is termed as enzymatic catalysis. She then

��| Biotechnology

continues, “But, to be able to carry out an enzymaticreaction in ionic liquids, it becomes important thatthe enzymes themselves are stable in ionic liquidsfirst. That is where the stability of bio-molecules inionic liquids came into the picture.”

Stability of bio-molecules such as DNA andproteins is one of the most critical issues thatresearch in bio-sciences faces every day. Thesebio-molecules are delicate entities, with very shorthalf-lives when stored at room temperature in water,their natural solvent. This is either due to hydrolysis(explained later) or the action of degrading enzymes,both of which cause their breakdown. Long-termpreservation of bio-molecules is carried out bystoring them at refrigerated conditions of -��oCto -��oC, temperatures which reduce the rate ofhydrolysis and render the degrading enzymesinactive. Common sights in all biology labs arerows of refrigerators which house many precioussamples. The failure of these refrigerators can leadto very significant losses – sometimes the wholecareer of a scientist is lost, sometimes irreplaceablesamples such as rare cancer tissues are lost andsometimes, arrays of samples for future experimentsare lost. The maintenance of samples at suchultra-low temperatures is therefore crucial and addssignificant costs and concerns to research. It isin fact ironic that this article was written whileguarding a -��oC freezer which threatened to go o�fduring the Chennai floods power outage.

Stability of bio-molecules such as DNAand proteins is one of the most critical

issues that research in bio-sciences facesevery day.

In such a scenario, the evolution of ionic liquidsto the latest generation of bio-compatible ionicliquids ushered in the promise of an economicaland hassle-free solution to this age-old problem.Research in this field was pioneered by Dr.Prabhakar Ranganathan, an alumnus of IIT Madras,and his group at the University of Monash, Australia.

They found that when DNA is stored in thesebio-compatible ionic liquids, the structural featuresof DNA were maintained even a�ter six months ofstorage at room temperature as opposed to a fewweeks in water!

DNA structureCourtesy: Dept. Biol. Penn State ©2004.

DNA is one of the three common bio-molecules(DNA, RNA and proteins) and functions as theinstruction manual for our body. These instructionsare, however, not written in the �� alphabetsof English, but in a code which has only fouralphabets – Adenine, Cytosine, Guanine andThymine – collectively known as nitrogenous bases.As can be seen in the figure, DNA has a veryinteresting structure. It is a double helix formedby two anti-parallel single strands made up ofthese nitrogenous bases. The backbone to thesenitrogenous bases is provided by alternating sugarand phosphate groups. These two single strandsare held together by hydrogen bonding, or inother words, the attraction between the positivelycharged hydrogen atom on one strand and anothernegatively charged atom such as nitrogen, oxygenor fluorine on the other. The two strands twistaround each other with an o�fset pairing, resultingin the formation of two kinds of grooves – majorand minor: these are structurally opposite toone another and run alternately along the entire

��| Immerse

Figure 1. Left: Water (in orange) surrounding DNA (in blue) in a 5 wt % ionic liquid (in green) solution. Right:Ionic liquids take the place of water in a 80 wt % ionic liquid solution.

Courtesy: Prof. Sanjib Senapati

length of the DNA. This double helix of the DNA isprovided structural support by a single layer of watermolecules, referred to as the ‘spine of hydration’,which remain hydrogen-bonded to the DNA in itsminor groove. Another way in which water lendsstructural support to the DNA is via the ’cone ofhydration’ – small clusters of water that surroundthe backbone of the DNA, the alternating sugar andphosphate groups.

. . .water is both a friend and a foe toDNA.

Interestingly, the same water molecules fail tostabilise the DNA when it comes to its covalentbonding. They disrupt these bonds, breaking theDNA down to smaller fragments which are againamenable to further break down right up to themonomeric components, the nitrogenous bases.This is called hydrolysis. In order to circumventhydrolysis, two strategies are attempted in DNAstorage – it is either stored in a dry state, i.e., it isdehydrated, or in water but frozen, so that the rateof hydrolysis diminishes rapidly. However, both themethods are not foolproof. With every dehydrationattempt or a freeze-thaw cycle, DNA, being made

up of extremely long, thin strands, tends to break,leading to structural damage.

From the study by Dr. Ranganathan sproutedmany unanswered questions that excited Prof.Sanjib and group. They figured out that MDsimulations could serve as highly valuable tools toexplore what ionic liquids are doing to the DNAto be able to have such a stabilising e�fect on it.They found that on introducing ionic liquids into asetup of DNA surrounded by water, the ionic liquidmolecules gradually start replacing water, and a�terabout ��� to ��� nanoseconds, there are very fewwater molecules le�t surrounding the DNA (Fig. �).

Now, this was both a cause for relief as wellas worry because water is both a friend and a foeto DNA. Interestingly, ionic liquids replace both thespine and the cone of hydration in the DNA. Thisshould actually lead to the entire DNA structurecollapsing; but with ionic liquids it does not, becauseof their ability to hydrogen-bond with the DNAthe same way that water does. The result is theformation of a spine and a cone of ionic liquidsthat now support the DNA structure instead of aspine and a cone of hydration (Fig. �). So, we havesomething that gets rid of most of the water and,

��| Biotechnology

Figure 2. (a) Spine of hydration in the minor group of DNA. Emergence of spine of ionic liquids in (b) 5 wt %and (c) 80 wt % ionic liquid solutions. Courtesy: Prof. Sanjib Senapati

therefore, reduces the rate of hydrolysis to almostnegligible while at the same time mimicking thesupporting role of water. It’s a win-win situation!

At this point, Debostuti acknowledges one ofthe drawbacks of using MD simulations for a studylike this – that a technique with a timescale ofnanoseconds cannot really predict whether the DNAwill remain stable for years to come or not. So, tosubstantiate the findings of their MD simulations,the group stored the DNA in a large number ofionic liquids at di�ferent concentrations and at roomtemperature. The DNA remained stable even a�ter ayear of storage at room temperature.

Despite such positive findings, the use of ionicliquids as storage molecules is a field that has notyet been embraced by the scientific community. Amajor reason for this is that while ionic liquids havebeen found to be fantastic hosts for DNA, theirtrack record with proteins has been patchy. Proteinstructure involves a great amount of diversity andcomplexity as compared to DNA structure. Giventhis, while researchers are now able to recommenddi�ferent ionic liquids for storing di�ferent kindsof proteins, they have still not reached a stagewhere they have been able to pinpoint a singleuniversal ionic liquid as being apt for proteinstorage. Nevertheless, the search is surely on.

More importantly, the field of study is stillin its infancy, where people have figured out thepros of the technique, but are still not sure aboutthe cons – such as any likely adverse e�fects dueto the long-term storage of bio-molecules in ionicliquids. Debostuti’s current job is screening out asmany ionic liquids from as many di�ferent classes aspossible for their potency in stable storage of DNA.The goal is to be able to make a confident statementone day: “Yes, all ionic liquids are good for long-termDNA storage.”

Ionic liquids get rid of most of the waterand, therefore, reduce the rate of DNA

hydrolysis while at the same timemimicking the supporting role of water in

the maintenance of DNA structure.

Meanwhile, the group is also exploring whathappens to DNA in an ionic liquid under conditionsof environmental stress such as high temperature,which is known to melt DNA. The results oftheir preliminary study are exciting – the meltingtemperature of DNA in ionic liquids is way higherthan its melting temperature in water, which meansthat DNA in ionic liquids is resistant to highertemperatures.

��| Immerse

Besides all this, Debostuti also sees applicationsof ionic liquids in many other fields, includingextraction and separation technologies and, ofcourse, the hot topic of drug delivery. However,the high viscosity of ionic liquids and their lack ofspecificity are proving to be challenges in the way,which scientists are trying to overcome in order tohelp these miracle solvents make their mark.

Debostuti is on the verge of her thesissubmission. She admits that when she was o�fered

the project initially, she was sceptical; DNA wasnever her ‘comfort-zone’, she was more of a ‘protein’person. But a�ter five years of exciting work onionic liquids and DNA, she seems to have changedher mind. She says in a very cheerful tone, “Today,if somebody asks me what I would like to domy post-doctoral research on, I would say DNAnanotechnology. Because, now, I can only think ofDNA and there is so much more to be done!” And wewish her all the best with that! ⌅

Meet the Author

Kiranmayi is a PhD student in the Department of Biotechnology. As part of her thesis,

she studies the involvement of DNA changes in the development of hypertension and

diabetes in Indian populations. A great admirer of the English language ever since she

can remember, she aspires to be a technical writer after she completes her PhD and tries

to find as much time as possible between her late nights at lab and her research seminars

to keep up with this passion.

��| Biotechnology

Algae on Fire

by Akshay Govindaraj

How complicated is the whole chemistry of burning fuels? What about that of algae burning? Furthermore,how do we tackle these di�ficulties to harness the enormous potentials in microalgae? We talk about how

extraction of fine chemicals and energy from algae is di�ferent from other conventional sources.

Algae are largely single-celled organismslacking roots, stems and leaves. Mostalgae, like plants, use photosynthesisto produce energy. Their simple

structure makes them highly energy e�ficient as well.They can vary in size from a few microns to largemacroscopic multi-cellular organisms which can beup to �� metres in length and can grow as fast as ��centimetres a day. Microalgae generally refer to allthose species of algae with sizes between �� and ���microns. Most of the estimated ���,���+ species ofalgae are expected to be unicellular and fall withinthis range.

But why do we care about these tiny organisms?For starters, their absence would take your breathaway, quite literally! It is a common misconceptionthat most of the oxygen we breathe comes fromforests. On the contrary, forests consume almost asmuch oxygen as they produce. It is estimated thatabout three-fourths of the oxygen we breathe comesfrom algae alone.

Perhaps more relevant to the state of theenvironment today is the fact that when nutrientsare available in plenty, algae populations can growvery rapidly, doubling in number every few hours.They can then be harvested quickly to produce usablebiofuel. Traditional crops would require ��� timesthe land to produce an equivalent amount of biofuelas microalgae. Part of the reason why algae havesurvived on this planet for so long is that they adaptvery quickly. This suggests that we might be able togrow algae in conditions where traditional crops failto grow.

It is estimated that three-fourths of theoxygen we breathe comes from algae

alone.

The idea of fuel sustainability is considered tobe the Holy Grail for global ecological health. Butwe are still quite far away from a state of sustainablegrowth, since we have thus far failed to find andutilise a source of energy that is renewable and can

��| Immerse

Dr. R Vinu obtained his PhD in Chemical Engineering from Indian Institute ofScience, Bangalore, in 2010. He has authored over 30 research papers, one bookchapter and filed a patent. He is the recipient of Young Faculty Recognition Awardfor excellence in teaching and research from IIT Madras in the year 2015.

act as a substitute to fossil fuels. Some arguethat usage of biofuels derived from algae will takeus closer to that goal, since algae take in carbondioxide from the atmosphere during the processof photosynthesis and e�fectively close the carboncycle. On the other hand, fossil fuels would bringadditional carbon into the atmosphere. At the veryleast, we can say that algae based fuels are greenerthan fossil fuels.

Other sources of energy like solar and windhave high capital costs, and since the demand forenergy is only ever going to increase, it is unlikelythat solar and wind will completely cater to them. Wedefinitely need to find some other sources of energy.Some argue that microalgae might just be the perfectsolution.

Use of algae as an ingredient in themanufacture of fine chemicals is a

futuristic thought.

In the conventional method of producingalgal fuels, algae are fed carbohydrates and theirsecretions are collected. The high fat content in thesesecretions results in a high calorific value. It hasbeen estimated that a total area the size of Francewould be enough to power the whole world’s energydemands. Algae can grow on land which is otherwiseunsuitable for agriculture or even in water bodies;therefore it might be precisely what mankind needsto avoid the energy crisis without compromisingmuch on our other dependencies on land.

But Dr. R Vinu from the Department ofChemical Engineering, IIT Madras, believes thatthere is a lot more about algae we haven’t exploredyet. He says, “Currently, uses of algae are mostlyfocused on generation of energy worldwide. Use

of algae as an ingredient in the manufacture offine chemicals is a futuristic thought.” Recently, ithas been observed that when algae are burnt in theabsence of air (a process known as pyrolysis), theresultant chemical composition, which di�fers fromspecies to species, sometimes contains compoundswhich are of high value. These compounds canbe used in pharmaceuticals, cosmetics and variousother industries. Dr. Vinu and his team havebeen analysing the process of pyrolysis with severalspecies of algae to try and understand the structureof the species and more importantly, find out if anyvaluable chemicals can be obtained using the sameprocess in a large scale.

In chemical engineering, there are two kinds ofproducts manufactured. Bulk chemicals, includingmost petroleum products, are those which aremanufactured in large quantities and usually havea continuous supply. Fine chemicals, such asingredients for synthetic drugs and cosmetics, aremuch more valuable and are usually manufacturedand marketed in relatively small quantities. Recentobservations suggest that pyrolysis of algae canbe used as a process to manufacture several finechemicals.

It has been known for a few years thatdirect algal pyrolysis of some algae species givesus an end product which is rich in a class oforganic compounds known as aromatics, which arecharacterised by pleasant smells. These compoundsare used extensively in the manufacture of variousplastics, detergents and drugs including aspirin.But only recently, Dr. Vinu and his team havediscovered that some species yield another classof organic compounds, called cycloalkanes. Thesecompounds have a plethora of uses in fields like

��| Chemical Engineering

Dr. Vinu’s group at IIT Madras.Courtesy: Dr.. Vinu

refrigeration, pharmaceuticals and the manufactureof other important chemicals. But there areseveral challenges that have to be addressed beforethe manufacture of these compounds can becommercialised.

Before any reaction or a set of reactionscan be conducted on a large scale, we must beable to mathematically simulate the whole processaccurately. This gives us an idea as to how thewhole process will react to any disturbance andalso provides us with a quantitative estimate on therisks associated with the reaction, which is criticalto prevent accidents. We depend on the field ofchemical kinetics to provide us with an accuratemathematical description of the entire process. Theset of all reactions expressed in the form of ordinarydi�ferential equations is known as a ‘mechanisticmodel’.

The way chemical kinetics is taught in schoolis o�ten misleading. It gives us the impressionthat all chemical reactions are highly predictable innature and it is just a matter of finding their reactionmechanism to describe the process, which is true, ina way. To find the reaction mechanism the wholeprocess has to be studied on a laboratory scale. Butin practice, some processes such as pyrolysis of algae

are just too complex to get a complete mechanisticmodel. The number of reactions, products andintermediate species are o�ten so high in numberthat it is close to impossible to find the exact reactionmechanism. In the large scale production of energyor manufacture of fine chemicals, an incompletepredictive mathematical model of the process cangive results which are completely aberrant.

The number of reactions, products andintermediate species are o�ten so high in

number that it is close to impossible to findthe exact reaction mechanism.

The team faced a similar problem during theinitial study of biomass pyrolysis. It was laterunderstood that the problem could be simplifiedusing predictive models which are mathematicallysimpler. Predictive models o�ten include only afew reactions which have high reaction rates orwhich give rise to more reactive species in theprocess while ignoring the other reactions. Youcan choose a predictive model whose complexity iscommensurate with the accuracy needed.

The search is on for more accurate predictivemodels for algal pyrolysis. But these predictive

��| Immerse

models are likely to be specific to some speciessince di�ferent species are di�ferent in their inherentchemical nature and result in di�ferent end productcompositions even under similar conditions.

Most species of algae are likely to give only asmall fraction of compounds which are economicallyvaluable amongst a deluge of by-products and sincevalue is only associated with pure chemicals, theymust be separated before they are ready to be used.But separation is o�ten a di�ficult task. What makesseparation di�ficult? According to thermodynamics,the process of mixing two pure components is o�tena spontaneous process, especially for compoundswhich are similar in structure. Since mixing isa spontaneous process, the exact opposite of it isnot. Additional energy of a higher grade mustbe provided to separate the components of themixture. The process of pyrolysis will o�ten result incompounds of similar nature which are even moredi�ficult to separate.

It has been observed, however, that algaehave high nitrogen content and when burnt in thepresence of air directly, release excessive amountof nitrogen oxides. Besides this, oil obtained

from pyrolysed algae will probably not be usable astransportation fuel since most conventional internalcombustion engines are only suitable for fuelswhich have a specific behaviour. The algal basedfuels can’t be modified to something which has achemical composition similar to the conventionaldiesel or gasoline, but in the future we might reduceour dependence on conventional sources of fueland build internal combustion engines which aredesigned for algae based fuels. Direct pyrolysis ofalgae may still be useful to generate energy in powerplants.

The whole idea of using algae to manufacturefine chemicals is still in its infancy and has a longway to go. So far, only a handful of algae specieshave been studied, but it is estimated that thereare more than seventy thousand other species, andperhaps many of them can be used to manufacturemore useful organic compounds. Today, most ofthe organic chemicals are obtained from sourceslike crude oil which are exhaustible. But given howquickly algae can grow, in the future we might onlydepend on algal sources to meet most, if not all, ofour demand for organic fine chemicals. ⌅

Meet the Author

Akshay Govindaraj is a student in the Department of Chemical Engineering atIIT Madras, whose interests are in certain areas of applied mathematics. Whileworking on this project he understood much more about how applied mathematicsis useful in chemical engineering. For comment or criticism, he can be reached atakshaygvdrj@gmail.com

��| Chemical Engineering

In ����, in a lecture entitled There is Plenty ofRoom at the Bottom, the renowned physicistRichard Feynman said “ . . .But I am notafraid to consider the final question as

to whether, ultimately, in the great future, we canarrange the atoms the way we want; the very atoms,all the way down!” What Feynman did in his lecturewas to explore the possibility of advanced syntheticchemistry by direct manipulation of individualatoms and molecules. The conceptual insight,though revolutionary, was not able to generateenough waves in the scientific community, at leastinitially. It was not until the late ����s that thetechnology could reach a stage where molecules andatoms could really be controlled and engineeredby direct engagement in their level. The scale ofinterest, as one can imagine, is exceedingly smallor nano as we know it now. For comparison,human hair is about ��,��� - ���,��� units wideon nanometer scale. The ‘Nano’ revolution usheredin synthesis of several brand new molecules andstructures of nano size with remarkable propertiesand hence diverse applicability. The field ofnanotechnology grew rapidly in the years thatfollowed and had already seen two Nobel Prizes bythe end of the next decade.

One such man-made nano structure which hasreceived considerable popularity since its synthesisin ����, is dendrimers and its assemblies. Theseare nanoscale molecules with beautifully symmetricand repetitively branched structure. A cursorysearch for ‘dendrimer’ on the Web of Science(Thomson Reuters) database produces more than��,��� results. Here at IIT Madras, the e�forts toprepare them and other light weight moleculesfor large assemblies to be used in wide variety of

applications is spearheaded by Dr. Edamana Prasadwith his group in the Department of Chemistry.

What Feynman did in his lecture was toexplore the possibility of advanced

synthetic chemistry by directmanipulation of individual atoms and

molecules.

“So, why have we chosen these molecules forour work?” says Dr. Prasad as we speak in hiso�fice next to the newly constructed Chemistrydepartment building. Let us consider dendrimers,for instance. These organic molecules closelyresemble bio-molecules such as protein in shape,size and weight. Proteins are known to self-assembleand generate unique hierarchic nanoscale structuresfor performing various functions in the human body.Taking this important clue from proteins, one can infact, in more or less similar fashion, generate higherorder complex structures with useful functionalityby properly customised aggregation of dendrimers.“But what useful functionality are we taking about?”was my follow-up question. Vivek, one of Dr.Prasad’s students, promptly replied - “What if I tellyou that such appropriately designed big assemblieshave self-healing ability and can assist in oil spillrecovery too? This is just a couple of their myriadusages.” I was intrigued.

In order to better understand the processof designing them, I stepped into Dr. Prasad’slaboratory and spoke with his PhD students Parthaand Madhu. The individual molecules undergo theprocess of self-assembly to create bigger assemblies.

Dr. Edamana Prasad is an Associate Professor in the Department of Chemistry at IIT

Madras. He worked at Photosciences and Photonics Laboratory, NIIST Thiruvananthapu-

ram (CSIR, Govt. of India) and obtained his PhD (Chemistry) in 2000. His research

interests include study of aggregation kinetics in dendrimers, finding the mechanism of

supramolecular self-assembly in dendrimers, and determining the excited state dynam-

ics in self-assembled systems. Dr. Prasad is also working as the Head of the Teaching

Learning Centre (TLC), at IIT Madras.

��| Chemistry

The protagonist of this story — dendrimers, whichconsist of ‘chemical shells’, organise themselvesvery beautifully in a symmetric pattern around thecore in a spherical form. Each shell is made upof molecules which are functionlised to create abranched structure around the core. This structureis known as dendrimers. The number of branchingevents from the core to the periphery is known as‘generations’. One can easily visualise them as a treewith many branches and sub-branches. The namehas been derived from Dendron which happens tobe the Greek work for ‘tree’.

As a direct consequence of fractal nature,their light emission properties show an

unprecedented enhancement.

Incidentally, branching in a tree remindsof fascinating objects called fractals – repeatingnever-ending patterns which appear self similaron various scales. So, do these dendrimers alsoshow fractal nature? Yes, they do. Dr. Prasadwith his former student Dr. Jasmine, reportedthat a popular class of dendrimers named PAMAMorganise themselves in an aqueous medium andshow fractal structures. The self-assembly ofPAMAM is achieved by electrostatic forces. Fractaldimension is a statistical index to estimate thefractal nature of an object. It essentially quantifiesthe change in fractal pattern with the change of scaleat which measurement is done and if this index is anon-integer (e.g., �.�), the corresponding structureis a fractal. For PAMAM dendrimer assembly, thisindex was shown to be a little above �.� and therefore,the fractal nature was confirmed. As a directconsequence of fractal nature, their light emissionproperties show an unprecedented enhancement.Fractals are perhaps the most preferred way ofgenerating captivating complexities in the naturalworld. It clearly works in the nano world too.

One is then led to think “What holds themtogether or to be more precise, what kind of forcesmediate this self assembly of small molecules?”

Partha reminds me of covalent bonds that weencounter in Chemistry ���. A covalent bold is astrong chemical bond which involves the sharingof electron pairs between atoms. But, for thesebig assemblies, we need the bonding to be ofthe non-covalent kind which essentially meansweak interactions. These forces operate beyondmolecular level and are responsible for spatialorganisation of complex molecular architecturethrough self-assembly. Therefore, we name them– supramolecular forces. There are many kinds ofthem, one, for example is hydrogen bonding.

Dendrimers and other such self assembliesmay have multiple supramolecular interactionsholding the individual parts together. Note thatthese interactions are weak in nature, thereforethe system is amenable to tuning during synthesis.We need freedom to break and make the bondseasily to control the process. An importantway in which some derivatives of dendrimersaggregate themselves is in the form of helicalstructures. Partha has recently explored themechanism in minute detail. He has found that suchself-assembled systems may exhibit well-definedalignment leading to chirality on the macroscopiclevel which means that the systems are not identicalto their mirror images. This is the origin of thehelical structure formed. This understanding of themechanistic aspect is involved, as Partha says, butcrucial to construct supramolecular systems as perthe requirement. Now that we are armed with allof this information, it is time to design some gelsystems which happen to be a natural consequenceof the aforementioned self-assembly.

We all have encountered gels in our daily lives.Butter, jam, shoe polish, hair gel, etc. are some ofthe gels we use regularly. A gel state is always easierto recognise than define, noted the British scientistDorothy Jordan Lloyd in ����. Typically, a liquidsystem made of two or more components turns intoa gel when one of the solute components formsa three dimensional crossed-linked network insidethe bulk gas or liquid. Formation of such a solidnetwork within the fluid restricts its flow resulting

��| Immerse

Research interests in the Dendrimers laboratory, Department of Chemistry at IIT Madras.

in a jelly-like substance. Now, if the cross-linking ofthe network component of a gel is supramolecularin nature, we get supramolecular gels. These gelsare formed, for example, as a result of self-assemblyof dendrimers in organic and aqueous solvent. Inrecent years, Dr. Prasad has been actively pursuingthe study of the formation and properties of these‘physical’ gels.

“The two kinds of supramolecular gels that weuse in our study are - hydrogels and organogels”,informs Dr. Prasad. As the names suggest, hydrogelshave water as their solvent while organogels areformed when an organic solvent (which has carbon)is a component of the gel. Dr. Rajamalli, a formerresearch scholar in Dr. Prasad’s lab has investigatedboth the gels in a series of publications. Organogelshave stimuli responsive character, which means thatan external or internal physical stimulus can promptsuch gels to tune their properties.

Dr. Rajamalli was able to design and synthesisean ‘instant’ organogel based on a class of dendronswith specific kind of linkages. The gel may be useddetect the presence of fluoride ions which has an

important role in biological systems. When the gelcomes into contact even with small concentration offluoride ions, a gel-solution transition occurs whichchanges the colour of the solution from deep yellowto bright red and hence the presence can be detectedby ‘naked eye’. A similar gel formation which wasinduced by metals, was also prepared to detect leadions as reported last year by Dr. Prasad with hispost-doctoral scholar Vidhya Lakshmi.

“The two kinds of supramolecular gelsthat we use in our study are - hydrogelsand organogels”, informs Dr. Prasad.

More recently, Madhu has prepared aninteresting three-component organogel. Thisorganogel consists of one-dimensional nano fibres.One of the components possesses cholesterol whichis a biocompatible molecule and is, in fact, knownfor its ability to form one-dimensional structuresand gels. Cholesterol e�fectively guides the processof supramolecular structure formation by stabilisingvarious hydrogen bonds and regulating positive

��| Chemistry

and negative charge transfer in the system.These charges originate from the other chemicalcomponents of the system. Therefore, in thepresence of an applied potential, the system exhibitselectric conductivity. Such a system has beensynthesised for the very first time here in this lab.Dr. Prasad’s student, Sitakant continues this work tounderstand the mechanism of conduction in detail.

Vidhya Laksmi and Madhu have alsosynthesised another promising organogel whichcan assist in recovery of spilled oil from water. Inmarine areas, accidental oil spill is a major concernas it has detrimental e�fects on the surroundingecosystem. Our friends in the lab observed that whena customised dendron-based solution or geletorcomes into contact with oil on sea water surface,a robust gel system is formed by readily absorbingthe oil floating atop. This gel is hydrophobic(water-hating) in nature and attains a wafer-likeform almost instantaneously and floats on the watersurface. These wafers of the gel can then removed,manually or mechanically. Oil is easily retrieved byheating the wafer gels. Surprisingly enough, geletorremains intact and can be re-used up to five or sixtimes with reasonable success. It is a highly e�ficientprocess of oil spill recovery.

This gel is hydrophobic in nature andattains a wafer-like form almost

instantaneously and floats on the watersurface.

Such a dendron based geletor has beensynthesised for the first time in this lab. Inaddition, the geletor is useful for its anti-wettingand self-cleaning properties and in the formation ofinvisible ink. It is indeed fascinating to note whathatred for water can do. But as we shall see, love forwater could also be equally rewarding. Hydrogelsconsist of hydrophilic (water-loving) structuresin them. These three-dimensional structures arecross-linked enabling the gel system to hold largeamounts of water.

Hydrogels are so�t, flexible and resemble livingtissues. Here, in the Dendrimers laboratory, thesehydrogels have been studied for their magnificentluminescence properties by Dr. Rajamalli, Supriya,and Sadeepan. Dr. Prasad’s research student,Prashant is attempting to use a hybrid hydrogelmedium to enhance the photoluminescenceproperties of lanthanide ions by a pheonomenenoncalled resonance energy transfer (RET). Versatileand stable light emission are highly desiredfor optoelectronic applications and in cellularbioimaging. However, as it happens, these are notthe only compelling features of hydrogels.

Dr. Prasad’s student, Vivek explains thathydrogels have this remarkable ability of self-healingwhich means two or more separate fragments of thegel can stick together spontaneously as fresh bondsare created in the process. This is something likethe flour dough we make to cook chapati. Two piecesof dough easily stick together if kept close enoughto create a larger piece. Hydrogels are so�t, flexibleand closely resemble living tissues and, therefore,have several biomedical applications. These gels arecurrently used in reconstructive tissue engineering,wound dressing, contact lenses, etc.

Vivek has recently synthesised a novelthree-component hydrogel with some commonlyavailable chemicals. Many of the known hydrogelshave this self-healing ability only in acidic mediumwhich limits their usage. The hydrogel synthesisedby Vivek maintains its self-healing property even ina medium which is neither acidic or alkaline, suchas water. Perhaps the most important use of thishydrogel is in the purification of water. When heavymetals ions (which are toxic in nature) and organicresidues present in water come into contact withthe gel, they get collected on the gel’s surface. Thegel can then be easily removed, leaving pure waterbehind. The gel also has robust mechanical strengthand high swelling capacity which are two highlysought-a�ter features of hydrogels from the pointof view of applications.

These supramolecular big assemblies that wehave seen so far are truly striking. A significant

��| Immerse

Dr. Edamana Prasad working with his students in the newly established LaserFlash Photolysis Laboratory at IIT Madras. The inset shows the laser beam at 532 nm.

application of supramolecular hydrogels is indrug delivery - an increasingly popular methodof targeted administration of medicine in thebody. It is one of the on-going works in theDendrimers lab and Dr. Prasad’s student Ramyawalks me through the details. “We are workingon a control release system”, she apprises me. Apopular alternative approach of drug delivery is byusing macromolecules such as higher generationdendrimers. In these dendrimers, as we have seenearlier, a series of chemical shells are attachedon many levels. Such an arrangement gives it aspherically symmetric three-dimensional shape,much like a flower. The structure is highly porousand consists of many empty pockets in which a drugis loaded. The drug get released slowly as and whenrequired. But, there are some serious issues thatneed to be resolved.

Higher generation dendrimers are di�ficult tosynthesise in a laboratory. In addition to that, someof these dendrimers are cytotoxic in nature, i.e.,they can kill the living cells. Ramya is attemptingto resolve these by using low generation dendrimersor dendrons. The structures of such dendrons areflat fibers and not a three-dimensional morphology.

Therefore, to hold the drug, she uses a gel systemcreated by an intelligent combination of water andan organic solvent. The gel thus formed is thesame hybrid hydrogel which we saw earlier. Inthe presence of the appropriate solvent and water,the dendrons self assembles in the form of longfibres, entangled with each other. The drug is loadedwith the solvent in the gelation process itself andresides inside those fibres. The initial results doindicate that it is an e�fective control mechanismfor the di�fusion of drugs. Ramya is now beginningto study the biological aspects of her findings, incollaboration with Dr. Vignesh Muthuvijayan fromthe Department of Biotechnology, IITM. She ishopeful that the stability of this system and its naturewould be favourable for drug delivery applications.

The initial results do indicate that it isan e�fective control mechanism for the

di�fusion of drugs.

Interestingly, there exists another kind ofdendrimers called Janus type dendrimers nameda�ter the Roman God with two faces, one lookingtowards the future and one at the past. Janus

��| Chemistry

dendrimers have an amphiphilic nature, i.e., theyconsist of both water-loving and water-hatingparts. Dr. Prasad’s student Prabakaran isinterested in synthesising them and in studyingtheir self-assembly properties. These dendrimerstend to form vesicle in a mixture of solvents, canform hydrogel, thermo-reversible organogels andalso exhibit liquid crystal behaviour. No wonder thatthese gels find potential applications in the fields ofdrug delivery, gene delivery and sensors.

Dendritic structures can be good host systemsfor metal nano particles and quantum dots. Quiteinterestingly, Dr. Prasad with his student, TufanGosh, has recently shown that some graphenequantum dots (GQDs) immersed in aqueoussolutions under certain conditions can be stableeven in the absence of dendritic support and emitbright and pure white light. These GQDs arezero dimensional and have tuneable luminescenceproperties. On investigation, it was found that itis an assembly of those GQDs that forms in thesolutions which generates this white light emissionwhen suitably excited. The process of aggregation ofGQDs is similar to that of dendrimers. The workhas convincingly demonstrated that pure whitelight emission can obtained by a well-designednanoscopic assembly of a single material, GQDs inthis case. Dr. Prasad’s research students, Kaviya andLasitha, in similar manner, use metal nano particlesand nano-scale assemblies as sensors and catalysts.

In future, Dr. Prasad’s laboratory envisagesto create more fundamental and applied researchbased on molecular assemblies. One of the majordevelopments in this direction was the recentestablishment of an ultrafast kinetic study facility(Laser Flash Photolysis) with the help of funding

from the Department of Science and Technology fora group project. This is the first of its kind facility atIIT Madras. The experimental set-up can be utilisedto analyse electron transfer kinetics which mostlyoccur from the electronically excited state of themolecules at nano-second time scale. Dr. Prasad andhis group are now heading to analyse the electrontransfer kinetics in molecular assemblies such asgels, which is an unexplored area in the frontierresearch.

Therefore, we discovered that big assemblieswith incredible features can be achieved byintelligent design of small molecules. The gelsthat we encountered find major applications in thedevelopment of smart materials. These are novelmaterials which tune their properties under theinfluence of external stimuli such as temperature,pressure, electric field, nature of the ambience etc.They are immensely useful in fabrication of sensors.The self-healing ability provides the material longlasting durability. On the other hand, biomedicalapplications of these gels are only limited by ourimagination. From drug and gene delivery to tissueengineering and biosensors, the list is long andrapidly expanding. Water purification and oil spillrecovery may be categorised as their non-trivialfields of usage. They also pave the way for newgeneration of optoelectronics. The developmentof organic light emitting devices, light harvestingsystems, photovoltaic cells, etc. can greatly benefitfrom their versatile light emitting properties. Thereis no doubt that these supramolecular big assembliesare playing an important and decisive role in shapingfuture – the great future that Feynman talked aboutin his lecture more than five decades ago. ⌅

Meet the Author

Swetamber Das is a PhD student in the Department of Physics at IIT Madras. He is in-

volved in various activities for popularisation of science. Working on this article exposed

him to the fascinating world of Chemistry. He feels grateful to Immerse and Dr. Edamana

Prasad for it. He is also exploring his newly found interest in the history and culture of

the Indian subcontinent. He is an Assistant Editor of Scholarpedia. For comment or

criticism, he can be reached at swetdas@gmail.com.

��| Immerse

The singer swings his arms around ashe delves into a raga alapana. Theviolinist listens carefully and plays theappropriate following phrases. The

percussion artists join once the composition starts,and the audience begins to keep track of the talaenthusiastically. Applause occurs sometimes in themiddle of a piece, when a particularly telling svaraphrase is sung or an interesting mrudangam patternplayed. As the three hour concert comes to a closeand the mangalam is sung, the curtains come downand listeners leave, content and filled with music.

Music is an art form, a source of entertainment,a means of communication, a way to celebrate, amethod of therapy, a source of joy. So what docomputers have to do with music? Can a machinerecognise ragas? Can Indian music be given anotation? Can a computer transliterate mrudangambeats? Can it separate a concert out into di�ferentsongs? Can it identify why certain songs make uscheerful, and others make us melancholic?

The first challenge faced by the team wasthe concept of a svara in Indian classical

music.

These are some of the questions that arebeing explored by Prof. Hema Murthy and herstudents in a project that is part of CompMusic,a worldwide Music Information Retrieval Projectthat is examining various traditional forms ofmusic. The music genres covered by the projectare Carnatic (South India), Hindustani (NorthIndia), Turkish-makam (Turkey), Arab-Andalusian(Maghreb, Northwest Africa), and Beijing Opera(China). The project deliberately focuses on certain

traditional forms of music that have not beendocumented the way some other systems of music,such as western classical, have. One of the primarygoals of the project is to showcase these systems ofmusic to the world.

Prof. Xavier Serra, coordinator of the project,a researcher in the field of sound and musiccomputing, and a musician himself, first heardCarnatic music when he was an expert speaker atthe “Winter School on Speech and Audio Processing”in ���� on “Audio Content Analysis and Retrieval”.In his own words, he had not heard anythinglike this before. He convinced Prof. Murthy tojoin the CompMusic project, overcoming her initialreservation that dissecting and analysing musicwould lead to losing the pleasure of simply listeningto it. Prof. Murthy immediately saw that it wasimportant for a musician to be part of the project,and TM Krishna, a popular Carnatic vocal musician,agreed to be a collaborator. In addition, a student ofhis, vocalist and engineer Vignesh Ishwar also joinedthe project o�ficially.

Tonic Determination

The first challenge faced by the team was theconcept of a svara in Indian classical music. Althoughloosely translated as a musical note, a svara is not somuch a note of a single fixed frequency as a rangeof sounds hovering around a certain frequency. Onemay vocalise a particular svara, but really be singinga combination of them. When say, one sings thesvara ‘ma’ in the raga Sankarabharanam, ‘ma’ is theonly svara pronounced, but the svaras ‘ga’ and ‘pa’ arealso touched upon.

In western music, pieces are composed to beperformed in a prescribed fixed scale.

Prof. Hema A Murthy is a Professor in the Department of Computer Science.

She obtained her PhD from the Department of Computer Science and Engi-

neering at IIT Madras in 1992. Her areas of research are speech processing,

speech synthesis and recognition, network traffic analysis and modelling, mu-

sic information retrieval, music processing, time series modelling, and pattern

recognition.

��| Immerse

Group Delay FunctionThe group delay function depends on the frequency of a signal and is the negative derivative of the phaseof the Fourier transform of the signal. The deviation of the group delay function from a fixed number is ameasure of the non-linearity of the phase as a function of frequency. The peaks in the group delay functionare inversely proportionate to the bandwidth of the group delay function and a peak signifies less inflection.

However, in Carnatic music the tonic is the referencenote established by the lead performer, relative towhich other notes in a melody are structured.

Musicians generally perform across threeoctaves, the lower, middle and upper octaves. Anoctave consists of seven svaras, ‘sa ri ga ma pa dha ni’.

The tonic is the note referred to as ‘sa’ in themiddle octave range of the performer. In order toidentify a raga, the first requirement is to determinethe tonic, because it gives a frame of reference.The same phrase of svaras may be identified ascompletely di�ferent ragas if the frame of referenceis di�ferent!

The basic unit of measurement of musicalintervals is a cent, which is a logarithmic measure.An octave is divided into ���� cents spanning ��semitones of ��� cents each. However, based on thetonic, the range of frequencies in an octave changes.For example, if the tonic is ��� Hz, then the higheroctave ‘sa’ is at ��� Hz, and the range of frequenciesin that octave is ��� Hz, and the ��� Hz are dividedinto ���� cents. But for someone whose tonic is��� Hz, the range of frequencies in an octave is ���Hz, and those ��� Hz are divided into ���� cents.Yet both these octaves are heard the same way by alistener. So tonic normalisation needs to be done.That is, the pitch histogram in the Hz scale needs tobe converted to a pitch histogram in the cent scale.

Group delay synthesised histogram

How does a listener determine what the tonicis? Usually one can immediately determine whichnote is the ‘sa’ (even if the note itself is not articulatedand only the lyrics of a song are sung). This is becausethe svaras ‘sa’ and ‘pa’ are ‘prakruti svaras’, or fixedsvaras, which are sung in a plain way compared to theother svaras. That is, the bandwidths of frequenciesfor these notes are sharper. Hence Prof. Murthyand the team used signal processing and machinelearning techniques to find out which are the sharpernotes. The pitch histogram of the music is processedusing what is known as a group delay function.

The group delay technique emphasises thesvaras ‘sa’ and ‘pa’ and this gives the tonic. Withthis method, the group was able to achieve about ��percent accuracy in identifying the tonic. To furtherfine-tune these methods, they segmented the soundof the tambura, an instrument that provides pitch tothe musician. They determined the tonic from thetambura alone, and this helped increase the accuracyof tonic identification to around �� percent.

Each raga has certain unique typicalmotifs that can be viewed as time

frequency trajectories.

Raga Verification

The next problem addressed was that ofmelodic processing. Can a computer listen to a songin a particular raga and determine what that ragais? First o�f, what is a raga? Loosely, it a collectionof svaras with some rules as to what combinationsthey can be sung in. But a raga is not merely thenotes that make the scale. The phrases (sequencesof notes) that are intrinsic to the raga, the variousgamakas (ornamentations) employed, the silences

��| Computer Science

between the notes, the inflections - all of these makeup the aesthetics of a raga.

How does a listener identify a raga? Somehow,within a few seconds of a musician singing a raga,a reasonably musically literate listener is able toidentify it. As the group realised, each raga hascertain unique typical motifs or signature phrases.A typical motif is a phrase that a particular ragais replete with, and that does not figure in anyother raga. A motif can be quantified by pitchcontours and viewed as a time frequency trajectory.The group realised that most of these motifs comefrom compositions set to tune in the raga, typicallythose composed by the ‘musical trinity’ – ShyamaSastri, Thyagaraja and Muthuswamy Dikshitar. Inparticular, they realised that it is the pallavi or thefirst segment of a song that typically contains therichest phrases of the raga. As an analogue, it is inthe initial few phrases of an alapana (a particularform of melodic improvisation typically performedbefore a composition) that the identity of the ragathat is to be performed is established. Prof. Murthylikens this to an artist drawing an outline of alandscape, or a portrait, before filling in the details,or a computational mathematician, who can gaugethe behaviour of a matrix by looking at its first feweigenvalues.

A mrudangam syllable has an onset,attack and decay.

The first task was thus to build a databaseof typical motifs of commonly performed ragas.The team used an algorithm called the longestcommon subsequence and a variant of it called therough longest common subsequence to identifythose phrases that are frequently repeated incompositions.

When we hear a raga, we first identify it witha smaller group of ragas, its cohorts. The team setabout the task of defining the cohorts for a numberof commonly performed ragas. They eventuallyrealised that what really happens when one listens

to a piece of music is raga verification rather thanraga identification. When we hear a raga, we do notactually compare it with every one of the hundreds ofragas we might know. First we think, ‘hey, it soundslike this song’. We identify it with some smallersubset – its cohorts – and then verify which of the ra-gas it is in the smaller group. The machine is trainedto do the same - the given time-frequency trajectoryis first identified with a small set of cohort ragas,each of which has some typical motifs derived fromcompositions and alapanas. These typical motifs arefed as queries and if they occur in the input raga, theraga is thus verified.

Time Frequency trajectories ofKalyani and Shankarabharanam

The ragas Sankarabharanam and Kalyani di�ferby a single svara. But anyone with a little knowledgeof Carnatic music can tell them apart. Interestingly,comparisons of their time frequency trajectories alsoshow how unlike each other they are. This is why thecomputer has to be trained to recognise raga motifsas time frequency trajectories rather than as merenotes. Carnatic music is primarily an oral traditionand notations only provide a rough framework.

Percussion

The next task that the team worked on involvedpercussion. Percussion is a complex part of Carnaticmusic. The raga is the melody, while the rhythmaspects are the laya and tala. Specifically, the talais the rhythmic structure in which a compositionis set. The mrudangam is the primary percussion

��| Immerse

instrument used in Carnatic music, while otherinstruments such as the kanjira, ghatam and mors-ing are also used. The mrudangam playing can varydepending on the lead artist, the mrudangam artisthimself, the song, the emotion conveyed by themusic and so on. The silences in between strokes areas important as the beats themselves. Sometimes theplaying is deliberately slightly o�f beat. This is called’syncopation’. Improvisation is a very important partof Carnatic music and musicians usually meet for thefirst time on stage with no prior rehearsals.

One of the purposes of analysing percussionis to put markers on a composition, to determinewhich part of the tala the song starts in, endsin and so on. TM Krishna pointed out that ‘mo-harras’ (certain predetermined patterns played) aremore or less fixed. But first, the beats of themrudangam have to be transcribed in some way.Prof. Murthy was actually approached by PadmaVibhushan Sangeeta Kalanidhi Dr. UmayalapuramSivaraman, a renowned senior mrudangam artist,who wanted the beats of his mrudangam to bedisplayed on a screen when he played. Every beatplayed on the mrudangam can be articulated orally assyllables, say for example, ‘ta ka dhi mi’. This processof the vocal articulation of percussion syllables, or‘sollus’ as they are known, is also called ‘konakkol’.

Stroke occurrences for instruments in differentpitches

The mrudangam stroke is viewed as an FM-AM(frequency and amplitude modulation) signal since

the sound of the mrudangam involves both pitch andvolume. The strokes played on the right hand side ofthe mrudangam are pitch strokes, while those playedon the le�t side are not. There is still however acertain coupling between them. In order to analysethe strokes and syllables of the mrudangam, Prof.Murthy relied on her work in speech recognition.Just as in speech, each syllable has a vowel withconsonants on either side. In a similar way, a mru-dangam syllable has an onset, attack and decay.Onset is the beginning of the syllable, which reachesits crescendo in the attack, and then it begins todecay. They used what is known as a hidden Markovmodel, in which attack, onset and decay are the threestates. Transitions between these can also be madestates in this model. Using this model, they wentabout the process of classifying strokes. The teamdevised features to process strokes of the mrudan-gam, kanjira and some other percussion instruments.

An important goal of music informationretrieval is music archival.

A�ter transcribing the strokes played by anartist, the machine uses the Markov model to locatethe moharra. The moharra in turn gives the numberof aksharas or beats in the tala and so the cycle lengthis determined. The team is now working on how tofind the point in the tala where the song begins.

Music Archival

An important goal of music informationretrieval is music archival. In this respect, one taskhandled by the group was concert segmentation,a major part of music archival. Most availablerecordings of Carnatic music are continuous butwe o�ten need to listen to only one particularsong or a raga. This requires that the concert besegmented into di�ferent segments. Prof. Murthyinitially suggested that they segment concerts intodi�ferent songs using applause. How can applausebe detected? Mapped as a time vs amplitude graph,it has the shape of an eye. A few in the audiencestart clapping, it reaches a crescendo and then

��| Computer Science

again becomes subdued. The team developed somedescriptors to detect applause, and some criteriawere fixed for the duration of an applause.

. . . the team was able to quantifyapplause. The strength of applausesindicates the highlights of a concert.

However, the assumption made here is thatapplauses occur only at the end of a certain piecein a concert. But this is simply not true in aCarnatic concert! For example, they found that theyhad a concert recording with, say, � items, whichwas segmented into, say, �� parts if applause wereused as a means of segmentation. This is becauseCarnatic music has certain kinds of improvisationthat take place at various points during a concert,and the audience may spontaneously clap if they finda particular part of the alapana or svaraprasthara (animprovisation technique in which svaras are sungextempore to a particular line of the composition)appealing.

Now, in every complete item of the concert,at some point or another, the vocal, violin andpercussion all occur. Moreover, they all appeartogether as an ensemble at some point. Also,a composition occurs in each item, and an itemalways ends with at least a small segment of thecomposition. The machine is thus trained to look forthose places where the vocal, violin and percussionare present together. It then goes backwards andforwards in order to identify the complete item andsegment it out. How is this ‘merging’ done from theright and le�t? Changes in the raga can be detectedand this helps determine when the performer hasmoved onto the next composition. The group wasable to successfully apply this technique to segmentabout �� continuous recordings of concerts intoseparate items. Web discussion groups such asrasikas.org that carry reviews and lists of itemsperformed in concerts made it possible to match thesong names to the segmented items.

In addition, the team was able to quantifyapplause. The strength of applauses indicates the

highlights of a concert. This algorithm can be used topick out the best or most popular parts of a concert.

The group has also been able to do some work onsinger identification. Drawing on experience fromProf. Murthy’s work in speaker recognition, theyhave been able to work on identifying musicians bythe timbre of their voices. This is possible by actuallytraining a machine learning algorithm to learn voicecharacteristics of di�ferent singers. This is alsoimportant in archival and concert categorisation.

Applause detection

Music Applications

The CompMusic team has contributedmetadata for items from Carnatic music concertsand albums to MusicBrainz, an open online musicencyclopedia and database which aims to serve asa repository of metadata for music from acrossthe world. The group is also considering workingon an application like SoundHound, in which onecan query a song by humming a part of it. TheCompMusic project has developed a web browsercalled Dunya which is freely available for use. Thework that has been done in di�ferent genres bydi�ferent groups can be tested on this browser. Forexample, one could think of applications like:

�. One can sing a snippet and the browserdetermines the tonic and range of frequencies,

�. If one is listening to a piece of musicand reproducing it, the browser tells howaccurately it has been copied – it can tell

��| Immerse

Dr. Hema Murthy and her students in the Department of Computer Science and Engineering at IIT Madras

whether what has been reproduced is in theright ‘sruti’ (pitch),

�. One can feed it a piece of music and it canchange the pitch and play the music alone(without the lyrics) at a di�ferent pitch.

All these applications are likely to be of greatuse to students of music. The browser is maintainedby the core CompMusic team. Each group developsthe algorithms and once they are robust, they areintegrated into the tool kit on the browser.

Initially, Prof. Murthy did not expect to findenough students interested in working on Carnaticmusic. To her surprise, she found that a largenumber of them – many with no background inCarnatic music at all – were enthusiastic about theproject. Shrey Dutta, for instance, went on to learnto play the veena, began to listen to Carnatic music,and did much of the motif recognition work. Hesays now that he only needs to look at the timefrequency trajectory of a raga to identify it, not evenrequiring the analysis that the machine does! JomKuriakose also came to Prof. Murthy with littlebackground in Carnatic music. He was, however,fascinated with percussion, and now works directly

with Umayalapuram Sivaraman on onset detection.Prof. Murthy is very happy that students have beenincredibly open about working in what is considereda niche and old fashioned genre of music.

All these applications are likely to be ofgreat use to students of music. Thebrowser is maintained by the core

CompMusic team.

Prof. Murthy stresses several times that anykind of machine learning should be implementedonly with proper context. A proper knowledge baseshould be the frame of reference for any machinelearning techniques. Machine learning involvesbig data, and as one pumps more informationin, it learns on the average. Signal processinggives results in the particular, but can make errors.Combining the two makes for an ideal recipe formusic information retrieval techniques.

Far from taking away the joy of listening tomusic, Prof. Murthy says analysing Carnatic musichas brought light to many things she had not realisedearlier. A raga can sound drastically di�ferent when

��| Computer Science

interpreted by di�ferent musicians. Moreover, theirstructures have changed over the centuries, andonly keep evolving with time. Why do some thingswork in music, and others fail? Why can onesing a particular gamaka in a particular raga butnot in another? Several things are intuitive aboutmusic. Can a machine understand these subtleties?Computers can be trained to play chess, to provemathematical theorems, to diagnose diseases, torecognise ragas and so much more. Can they also

be trained to think intelligently and creatively? Togive elegant proofs, and discern between good andmediocre music? If pointed in the right direction,can they even see things that humans may miss? Thisrequires a deep understanding of human cognitionand the human creative process. By working onvarious forms of music, and more generally in theunderstanding of art and creativity, CompMusic is asignificant contribution to the vast ocean of artificialintelligence.⌅

All the images have been taken from Prof. Murthy’s research papers.

Meet the Author

Arundhathi Krishnan is a PhD student in the Department of Mathematics and a Carnatic

musician. Her area of research is Functional Analysis and Operator Theory. She can be

reached at arundhathi.krishnan@gmail.com

��| Immerse

Centre for Innovation (CFI), IITMadras calls itself a place wherea student can ‘walk in with anidea, and walk out with a product’.

Known within the institute as a place wherestudents slog from midnight until daybreak, itencourages students to pursue innovative ideasand supports committed and ambitious studentteams to participate in prestigious internationalcompetitions.

Team Amogh is one such team that has madeIIT Madras’ first Autonomous Underwater Vehicle(AUV). This vehicle, codenamed AUV Amogh, iscapable of performing a set of predefined tasks onits own without human intervention.

But why work on an AUV when there are manyother interesting avenues needing attention? Oceanbodies cover over �� per cent of the earth’s surface,but they remain unexplored to a large extent.Although industries making Remotely OperatedVehicles (ROVs) have taken up many underwatermissions, there is still a need to develop AUVs,

especially because there are places deep below inwater where e�fective communication is an issue.Also, the research and development pertaining tounderwater projects is progressing at a much slowerpace as compared to land-based and air-basedprojects. Determined to push the envelope ofunderwater technology for exploration, the teamdecided to develop a fully autonomous underwatervehicle.

In their endeavour, the team has also provedtheir mettle in several competitions. Theyhad their first taste of success in ���� whenthey participated in the national competition –Student Autonomous Underwater Vehicle (SAVe)– organised by the National Institute of OceanTechnology. Outperforming all the other teams inevery aspect, the team secured the first position.This success spurred them to participate inan international competition called RoboSub,conducted by the Association for Unmanned VehicleSystems International (AUVSI). Competing againstseveral teams, they emerged as strong contenderson an international level.

Side View of AUV Amogh.

��| Immerse

Apart from these two competitions, they alsoparticipated in various innovation challenges andthe vehicle has also been selected as one of the topstudent innovations across the nation.

As it turns out, like most other success stories,Amogh too had humble beginnings. The team hadinitially set out to build nothing more than an ROVwhich was stable and manoeuvrable. In this phase,they designed the frame and the hulls, decided onthe material to be used, analysed the structure, andcame up with a waterproofing mechanism. Whilepassing their prototype through a series of testsand upgrades, they faced numerous challenges,the biggest one being trying to make the vehiclewaterproof. A�ter a detailed examination of variousoptions, they chose water-tight PVC pipes. Thesepipes were fixed tightly to both fore and a�t ends ofthe hull, and then sealed using an epoxy substance.

The Team’s ROV.

Improving their prototype at every stage, theyeventually built a self-powered ROV. The vehicle hadlithium polymer (LiPo) batteries on board to powerthe thrusters. These batteries were controlled usinga terminal connected to a wireless router, which wasin turn connected to the vehicle through an ethernetcable.

The team’s success in the ROV phase furtherfuelled their ambition to build an AUV. However, thiswas by no means a mere continuation of their taskuntil then. Each aspect of the AUV design demandedexpertise in a specific area. Three sub-teams wereformed to tackle this – the mechanical, electrical,and the so�tware teams. Although these teams werecarved out of the original team, they needed tofunction in coordination with each other.

Designing the AUV

Needless to say, the mechanical team wasentrusted with the indispensable task of designingand manufacturing the vehicle. More precisely, theyhad to design the following parts: pressure hulls,frame, and camera enclosure.

Since pressure hulls provide a watertightenclosure for the vehicle’s electronics, their designcould not be overlooked. A configuration of twocylindrical hulls was chosen because it helped reducethe resistance of the vehicle and provided enoughroom for the electronics to be mounted on. Thebottom hull was chosen to be heavy in order tocounter the buoyant forces that would push thevehicle out of water.

In order to achieve a higher speed per unit ofpower input, the hull was fitted with a nose havingan ellipsoidal shape as it o�fered the least drag onthe structure. Furthermore, simulations helpeddetermine the thickness that would enable the hullto withstand the pressure of water at a depth of�� metres. Being the backbone of any underwatervehicle, waterproofing was done meticulously. Acustomised cap, with grooves to accommodate tworubber o-rings, was permanently attached to the a�tend of the top hull. The cap was further coveredwith a flat disc, which consisted of � co-axial holes(not visible in the figure) to mechanically squeezethe o-rings and ensure watertightness. In order toprevent any chance of water entering the hull, thegap between the cap and the flat disc was sealedusing silicon grease.

The Top Hull.

��| Centre For Innovation

At this juncture, it is worth noting that anAUV is also required to perform certain slickmaneuvering tasks. For this purpose, the vehiclewas equipped with � thrusters to achieve control in �degrees of freedom. Two thrusters placed on eitherside of the frame facilitate surge (forward/backwardmotion) and yaw (tilting in its own plane) control.Two thrusters – fore and a�t – positioned axiallyupwards provide heave control.

The AUV’s Frame

Powering the Motion

Let us now move on to the team withoutwhich the vehicle would be rendered powerless– the electrical team. They were responsible forpower management, circuit design, and missioncontrol. The electrical module consists of a CentralProcessing Unit (CPU), a micro-controller, powersupply units, sensors, thrusters, and other essentialperipherals.

One of the seemingly insurmountablegoals the teams set for themselves was to

eschew ready-made components anddesign their own components instead.

The motherboard or the CPU, just as the one inyour desktop, does the main job of image processingand mission controlling, and provides a platform forall the components of the vehicle to communicatewith each other. The micro-controller controls the

motion of the vehicle by changing the rotation speedof the thrusters, on receiving commands from theCPU.

The primary control board, an interface forvarious sensors used in the vehicle, initially hadall the components soldered onto it by fitting thewire leads of the components into the holes on theboard – referred to as through-hole technology inelectrical hardware parlance. They re-designed thecircuit using Surface Mount Technology (SMT), inwhich the components are mounted directly ontothe control board. They were indigenously designed,except the surface of the board. This significantlyreduced the size of the board because by havingsmaller or no leads, SMT components were smallerthan their through-hole counterparts.

The bridge between the micro-controllerand the thruster is the motor driver. It’s acircuit that draws power from the batteries, anddrives the motor at the speed demanded by themicro-controller. By now, it is natural for the readerto assume that all such high-tech components, suchas the micro-controller or the motor driver, werepurchased from electronics stores in the market.However, this is not true. One of the seeminglyinsurmountable goals the teams set for themselveswas to eschew ready-made components and designtheir own components instead.

They intended to pursue this slowly, replacingthe ready-made circuits with the ones they designed.This helped them build components tailored to theirneeds.

Among the most essential of these componentswere the thrusters which consumed about �� percent of the total power. As a result, they demandedhigh capacity batteries to run the vehicle. Therefore,four lithium polymer batteries which together lastedfor a minimum of �� minutes were chosen forthe entire vehicle. The � higher voltage batteriespowered the thrusters, and the remaining � lowvoltage ones supported all the other peripherals.

However, this isn’t all that there is to anAUV even when looking at it solely from an

��| Immerse

A Labelled Diagram of AUV Amogh

electrical engineering viewpoint. For anything tobe autonomous, sensors are essential. Amogh uses� sensors – pressure sensor, inertial measurementunit (IMU), current sensor, voltage sensor and leakdetection sensor – and a pair of cameras. Thepressure sensor is used to determine the depthof the vehicle below sea level. The IMU measuresthe orientation of the vehicle in degrees. Currentsensors were used to measure the current flowingthrough each device since a high surge in currentmight permanently damage the device. Since anexcessive discharge of lithium polymer batteriesleads to catastrophic failures, voltage sensors wereincorporated to regularly monitor the voltage acrossthe batteries. In order to prevent any damage dueto water leaking into the hull, a circuit was built toidentify the intrusion of water. Two circular probeswere mounted near the end cap of the hull. Sincewater conducts electricity even with slight impurity,the voltage across these probes gets amplified inits presence. This voltage signal is sent to themicrocontroller to trigger a shutdown of the system.

Steering the Ship

Last but far from least, the so�tware teamwas responsible for image processing, missioncontrolling, and designing a simulator. Thesignificance of their role can be best explainedusing the following example of a competition theyparticipated in.

As per the problem statement in the RoboSubcompetition, the vehicle was supposed to touch �buoys. The vehicle was guided towards the buoy bya plank placed on the floor of the water body. Thebuoy had to be traced by the front camera before theAUV reached the end of the plank. Once the vehicletapped the buoy, it would bounce back and traversetowards the other buoy, as before.

In order to achieve this, the vehicle leveraged� cameras, placed in the front and the bottom ofthe vehicle. However, there was a challenge: theimages weren’t clear enough for a spotted object tobe detected. They had to be corrected to remove the

��| Centre For Innovation

blue tinge, a characteristic of underwater images,and brightened in order to improve visibility. Anyimage had to be preprocessed to ensure high chancesof the corresponding object getting traced on thecamera. How did they accomplish this?

Note that a camera treats images as beingcomposed of a large number of tiny coloured squarescalled pixels. Each pixel is a combination of threecolours – red, green and blue. So if a bluish tinge hasto be made negligible, the red and green channelscan be boosted in intensity.

In another such issue, water, because of itshigh refractive index, deviates light from its originalpath, leading to reduced visibility. The resultingdark images are corrected by a method called gammacorrection. In this method, the RGB colour space isconverted to another colour space called HSV (Hue,Saturation, Value). The dullness of the images can be

rectified by increasing the saturation of the image.This leads to bright objects becoming brighter andthe dark ones becoming darker.

Furthermore, the orientation of the plank withrespect to the frame was determined to correct thepath of the vehicle. As the vehicle got closer to thebuoy, the area occupied by the buoy in the imagegot larger. Once it reached a certain threshold, thevehicle was programmed to move further, hit thebuoy and come back to take another course.

The mission-controlling part of the so�twaredetermined the power needed to be given to eachof the thrusters to move along a particular course.Because the LiPo batteries had limited enduranceand needed a significant amount of time to getcharged, the team also designed a simulator to solvethe challenges of mission controlling.

Team Amogh with their AUV.

��| Immerse

AUVs have a plethora of applications. They areused mainly in detecting leaks in oil pipelines deep inthe ocean. They are even used in detecting corrosionin a ship’s hull, ballast tanks, piles of a dock, and oiltanks. A technique called non-destructive analysis,where theories pertaining to ultrasonic sound areused, can detect the thickness of the corrosive layer.These frequencies, in the order of a few MHz,penetrate the corroded layer before being reflectedby the underlying metal.

Team Amogh’s project is representative ofhow student teams work together in groups to

participate in competitions, taking charge ofdi�ferent lines of work to perfect every singlecomponent involved in the design of the vehicle.The team now plans to upgrade its present designby using brushless thrusters, and slowly transformit into a modular design. To be at par with thepresent day technology, they have also decidedto use acoustic sensors to determine the vehicle’slocation precisely. A startup, named PlanysTechnologies, has also emerged out of the project.Currently incubated in IIT Madras, Planys plans todeliver customised autonomous vehicles specific todi�ferent underwater applications. ⌅

All images are courtesy of Team Amogh, CFI

Meet the Author

Rahul Vadaga is a 4th year Dual Degree (B.Tech.-M.Tech.) student in the Department

of Electrical Engineering at IIT Madras. Fascinated by the idea of ‘building things on

one’s own’, he joined the Centre for Innovation (CFI). After a year-long thrilling ride at

CFI, he decided to write about one of its notable and successful endeavours. He feels

grateful to Immerse and Team Amogh for presenting him with an opportunity to do so.

Of late, he has been exploring the area of Artificial Intelligence in order to understand

its immense possibilities for the future. For comments or criticism, he can be reached at

rahul.vadaga@gmail.com

��| Centre For Innovation

As the torrential rains lashed at the hillslopes of Uttarakhand, life came toa complete standstill in one of thebiggest tourist centers of the country.

Hundreds of lives were lost and property wasdamaged. Fear and panic spread as survivorsunsuccessfully tried to contact their loved ones. Insuch situations, rescue and relief operations havebeen extremely di�ficult. Due to lack of authenticinformation and communication breakdown, thedays following a disaster have never been devoid ofpanic and confusion.

This system makes use of the existingtechnologies available to provide an

immediate and temporarycommunication means.

To address this vital aspect of disastermanagement, namely, the establishment of apost-disaster communication system, the Japanesegovernment, in collaboration with various technicalinstitutes in India and Japan, has setup theDISANET (Information Network for Naturaldisaster Mitigation and Recovery) program.The main aim of this program is to develop acomplete model that covers the various aspects ofdisaster management which include monitoringand modelling of weather and seismic activity,developing a robust communication network andexecution of e�fective relief in a post-disastersituation.

The entire program was split into four majordivisions, each of which were taken up by specificresearch groups from India and Japan. Among

the four divisions, the development of sustainablecommunication architecture was undertaken byProf. Devendra Jalihal and Prof. David Koilpillaifrom the Department of Electrical Engineeringat IIT Madras along with Keio University, Japan.They have developed a very innovative and e�fectivemeans of communication that can function as goodas a cell-phone network even when all the existingcommunication systems fall prey to a disaster.

It has all the important features requiredfor post-disaster communication: less time forinstallation, greater accessibility, e�fective outreachand broadcast of authentic information. It does notrequire any custom built equipment which makes itreadily available at any location within short time.Mobile phones are the most common means ofcommunication and are heavily depended upon.Therefore this system makes use of the existingtechnologies available in mobile phones to providean immediate and temporary communicationmeans.

The physical setup includes an LTE (Long TermEvolution, commonly known as �G) transmitteror antenna and other related equipment that arehoused in a hoisted helium balloon for ensuring alarge coverage area. The coverage is enhanced byusing an FM broadcast system, which can conveyinformation to victims regarding relief supplies,precautions and other rescue operation details. Thisinformation is broadcasted by an authentic sourcesuch as the district collector, over a certain frequencythat the victims can tune to and get informed. Thelow bit-rate digital data also called RDS can be usedto broadcast centralised relief information in textformat than can be read from the mobile.

Dr. Devendra Jalihal is a Professor in the Department of Electrical Engineering at IIT

Madras. He recieved his B.Tech. from IIT Kharagpur in 1983 and then completed his

Masters in Engineering at McMaster University at Hamilton, Canada. In 1994 he joined

the Department of Electrical Engineering at IIT Madras. Prof. Jalihal enjoys teaching

the fundamentals of Electrical Engineering such as signals and systems and communi-

cation theory in undergraduate courses. His research interests include Statistical Signal

Processing, Detection and Estimation Theory and Digital Communication.

��| Electrical Engineering

Generally, RDS is used by FM radio channels todisplay text such as the name of the song, channeland sometimes even the song lyrics on the screenof the audio setup. A similar text containing thehelpline numbers, details of the whereabouts ofrelief materials, etc. can be broadcasted throughRDS. Since FM-RDS is a feature available on a largenumber of modern GSM handsets, it o�fers thegreatest outreach to the victims.

As mentioned above, FM broadcast is used toconvey information from the authority at the controlcentre to the victims. Similarly the victims also cansend text messages, images and even short videosgiving details of their location and condition to acertain number over the network. To facilitate this,WiFi and LTE (�G) technologies are used. Since WiFihas a short range, multiple antennae are set up in thesurrounding areas and the main antenna is placed

in the basket of the helium balloon. LTE has a largerange of �� to �� km and therefore eliminates theneed for any intermediate towers. Hence it savestime and is the best means of communication whenthere is no possibility of setting up towers. Both WiFiand LTE technologies provide high bit-rate and canbe used for streaming videos and images. For thebenefit of those having mobile phones without anyWiFi or �G technology, there is also a GSM setup inthe basket of the helium balloon that functions asa temporary tower. This acts as any other ordinarycommunication tower to transmit voice messagesand calling. The details sent by the victims viathe above mentioned modes, are gathered by centreauthorities and provided to rescue workers to carryout the relief operations. Also, rescue troops can usethis communication network to be in constant touchwith each other and the main control unit.

Circuitry inside the GSM Base StationCourtesy: Nithyanand Rao

��| Immerse

The DISANET TeamCredits: Prof. Devendra Jalihal

During any disaster, the main causeof communication breakdown is an increasein communication tra�fic over the usualcommunication networks. “Due to people tryingto contact their loved ones in the a�fected areas,there is a large and sudden increase in the number ofcalls being made to a particular subscriber, leadingto congestion and eventual breakdown”, says ProfJalihal. Therefore the DISANET communicationsystem has introduced the ‘I am Alive’ feature toaddress this problem. A victim in the a�fected areasends a text message or an image to the call centerwhich then updates his/her mobile number alongwith the message, date and time, on the internet as asearchable entity thus being accessible to everyone.In this way, the well-being of a victim is conveyedto a large number of people at a time, thus avoidingexcess tra�fic.

During the floods in Uttarakhand, the

casualties’ details and information were notavailable even a�ter four or five days following thedisaster.

The well-being of a victim is conveyed toa large number of people at a time, thus

avoiding excess tra�fic.

This led to uncertainties in the whereaboutsand well-being of victims. To overcome thislimitation, the DISANET communication systemmakes use of the ‘person finder’ feature developedby Google. It uses various attributes of a personto confirm his identity. The rescue operators takeimages or videos of the victims and send them to themain operation center. This data is presented to theworld in standard formats known as PFIF (PersonFinder Information Format). It consists of victimdetails displayed on dashboards. Initially, only the

��| Electrical Engineering

picture and a few details gathered by the rescueoperators are available, but with time, people whoknow the victim can add and update other detailsthus making the information complete. This waydata can be refined and augmented with time.

This completes the entire framework of theDISANET communication system along withits features for providing e�fective and robustcommunication during disasters. The system wassuccessfully tested on a small scale at IIT Madrasin July ����. Prof. Jalihal mentions that there

are ongoing talks with the Chennai Police andRailways for implementing some of the features ofthis technology with a few modifications in heavilycrowded areas during festive seasons in the city. Itcannot be emphasised enough that malfunctioningof communication systems during crisis situationsamplify the di�ficulties faced by the victims as wellas the rescue troops. Hopefully, the establishmentof such an e�ficient communication network willreduce the confusion during such times and inthe process make the rescue work easier and moree�fective. ⌅

Meet the Author

Tejdeep Reddy is a third year undergraduate student, pursuing his B.Tech. in Naval Ar-

chitecture and Ocean Engineering at IIT Madras. He is also actively involved in the

development of an Autonomous Underwater Vehicle at the Center for Innovation, IIT

Madras. To know more about this remarkable vehicle, go to page 44 and start savouring

its story!

��| Immerse

Adisobedient mass of cells – looselycalled a cancer, or tumour – sits inthe midst of healthy tissue. Evadingthe body’s immune system, and

drawing sustenance from the blood vessels that itmanages to recruit around itself, the rogue masscontinues to grow - as cells within it divide, anddivide again. As it works hard at performing thisdeadly exercise, the tumour cannot help but warmup and give o�f some heat. This heat radiatesoutwards and is ordinarily lost to surroundingspaces. Dr. Kavitha Arunachalam and her group atthe Department of Engineering Design have beenworking on ways to detect this naturally-emittedheat reliably, using microwave radiometry. Theyalso use externally-supplied heat to help destroythe growing mass, using an approach known ashyperthermia.

A cancer generally does not announce itsarrival in a hurry, and it cannot be madeto leave without a sacrifice of some of the

body’s healthy tissue.

When a microbe (a bacterium or a fungus)infects a human body, we manage to attack it withchemicals, such as antibiotics, which take advantageof the bacterium’s vulnerabilities; vulnerabilitiesthat are not shared by our own cells. In contrast, acancerous cell is like a healthy cell in almost everyway, except that it somehow manages to divideuninhibitedly. A cancer generally does not announceits arrival in a hurry, and it cannot be made to leave

without a sacrifice of some of the body’s healthytissue.

In the specific context of breast cancer, whicha�fects more than a million women every year,another unfortunate fact needs to be faced. Today,the most reliable method of detecting breast canceris X-ray mammography; a technique which involvessending high-energy radiation through breasttissue. X-rays have enough energy to damage DNA,and create mutations. X-rays, therefore, can actuallypotentially cause cancer even as they are used todetect its presence. Thus, there is a dire need foralternative imaging methods.

Dr. Kavitha’s group works with low-energy,low-frequency, non-ionising electromagneticradiation. The group works with microwaves (whichhave already given us tools such as radar, radiotelescopes, GPS, mobile phones and microwaveovens). Microwave frequencies are lower than thoseof red and infrared light, and go down all the way tothe frequency ranges of radio waves. Dr. Kavitha’sgroup believes that microwaves have potential formedical imaging and treatment which is only justbeginning to be explored.

Microwaves, unlike X-rays, cause no mutationsin DNA. Also, the heat that a tumour generatesincludes a component of microwave radiation. Thismakes microwaves ideal for use in both treatmentand imaging, firstly because one does not needto introduce any external radiation to performimaging, and secondly because much less harm isdone when one does need to send some microwaveradiation into cancerous tissue, to destroy it.

Dr. Kavitha Arunachalam is an Assistant Professor in the Department of Engineering

Design at IIT Madras. She works on microwave antenna design, non-destructive test-

ing of materials, and development of instrumentation for biomedical applications. Dr.

Arunachalam obtained her B.E. in Electronics and Communication Engineering from the

College of Engineering, Guindy, Anna University, and her PhD from the non-destructive

evaluation laboratory at Michigan State University. Following postdoctoral research at

the hyperthermia research laboratory, Duke University Medical Centre, she joined IIT

Madras in 2010.

��| Immerse

(from left) Dr. Kavitha, Geetha, Rachana and Vidyalakshmi.

A microwave radiometer transmits nothing to theobject that it images. It merely receives andmeasures the radiation generated by the object,making it completely safe for use with tissues.When used for cancer detection, the radiometersimply detects the heat that a tumour generates atmicrowave frequencies, and uses this information tofind the tumour.

The heat radiated by a tumour, and by healthyparts of the breast, carries information abouttemperatures. This allows a radiometer-baseddevice to create a three-dimensional temperaturemap of the breast. Going inwards, as thetemperature rises, one encounters a series ofisothermal (equal temperature) contours centredaround a hotspot – the cancerous mass. How wellthe device is able to locate a tumour depends onits ability to measure di�ferences in temperature.The tumour’s size at this stage is roughly five to tenmillimetres across – making it large enough to beresolved from surrounding tissue through the use ofmicrometre wavelengths.

The temperature of an object tells us howenergetically its atoms and molecules are movingaround. Every object which is at any temperature

above absolute zero (at which atomic motion ceases)radiates some energy. The object might make upfor what it loses this way by absorbing radiationthat falls on it, in order to maintain a steadytemperature. This is a fundamental consequenceof the restlessness of the charged particles insideit; the energy of that motion is converted to theenergy of the radiation emitted. The nature of theradiation emitted by such an object depends on itstemperature. This is how astronomers estimate thetemperatures of stars. In the spectrum of lightreceived from a star, a particular frequency hasthe maximum representation in terms of energy.The higher this peak frequency, the higher thetemperature of its source. So blue stars are hotterthan red ones.

The heat radiated by a tumour, and byhealthy parts of the breast, carriesinformation about temperatures.

If, instead of receiving and analysing the entirespectrum, we were to build a device focusing on aselect group of frequencies – e.g., the microwaveregion of the spectrum – then the amount of energythat such a device would get from the radiating

��| Engineering Design

object would be directly related to its temperature.Of course, this is an oversimplification and onlyroughly true; but it holds for low frequencies.Microwave is low enough in frequency for this tohold true, making it a reasonable choice for thenarrow band of frequencies that one chooses todetect in the case of cancers.

Another factor that determines this choice ishow deep the technique allows us to look. Thetumour emits heat at all frequencies, but as thesewaves make their way to the antenna at the surfaceof the breast, their energy gets absorbed by layersof muscle, fat and glandular tissue. It is possible toform a picture of breast tissue by measuring infraredemissions as well, but this picture would go lessthan a centimetre deep. This is because infraredemission, having much higher-energy, would dieout much faster with distance. Microwaves, on theother hand, can bring information from as deep asthree to four centimetres into the breast. As Dr.Kavitha explains, decent resolution and very goodpenetration are what make the microwave frequencyrange a good choice for such applications.

It is possible to form a picture of breasttissue by measuring infrared emissions

as well, but this picture would go lessthan a centimetre deep.

Note, however, that while these waves reach uswith more of their initial energy intact than thatof infrared, this says nothing about how much ofthat energy was there to begin with. VidyalakshmiMR, the research scholar who designed and builtthe device circuitry, gives me an idea of just howweak the signal that they’re trying to catch is. AsI enter the lab with her, she sits down with asheet of paper and proceeds to explain with extremee�ficiency everything I can understand about howthe device works, despite her dismay at my lack ofknowledge of anything but the very basics of fieldtheory. I watch in awe as she lists out all the mobile,bluetooth and WiFi signals that are always zipping

across everywhere around us, and shows me, in themiddle of that chaos of frequency bands, the tinysignal that their device works with. It’s a signal,she tells me, as weak as what one would receive onthe Earth’s surface from a satellite in orbit. A goodquality call on a mobile network would generally usea signal about a hundred thousand times strongerthan that, and such mobile signals would mostlikely be abundantly available to interfere with thedetection device, anywhere that it might be used.

The problem, then, is not only to sort out thefrequencies but to detect and deal with such a weaksignal in the first place. It is a signal in picowatts,a trillion times weaker than the milliwatt scale atwhich most power sensors operate. Measuring abroader range of frequencies, which would haveincreased the detected power, is made impossible bythe flanking communication bands. What’s more,every measurement device has random internalvariations, called noise, that are usually too small tomake much di�ference to the signal but can distortit beyond recognition if the signal itself is equallysmall.

To be able to detect the power without addingany noise, while also making sure that it was a signalonly from the cells and not from the environment,the lab had to meet the challenge of designing a verygood front-end for the instrument. A front-end isa component of every communication system thatdirectly takes the signal collected by the antenna,processes it and passes it on. The front-endnormally starts with what is called a band pass filter,which passes on the required frequency and getsrid of the rest. Following this, the output fromthe filter is amplified so that later stages in thecircuit can work with a better signal. So at first,Vidyalakshmi tried placing the filter and amplifiersin this configuration. It didn’t work; what the filterreceived from the antenna was too weak for it towork with. The front-end that she finally developednow has three stages of amplifiers to get the strengthup to a decent level before it reaches the filteringstage. At the end of each stage, there are isolatorsthat act like valves or one-way gates, so that nothing

��| Immerse

happening in any part of the circuit feeds back to thestage before it, to a�fect it.

I ask her if she was apprehensive aboutchoosing such a weak signal in such a crowded zoneof the spectrum. She is surprised by the question; shedid not ever doubt that it could be done. The deviceis now ready, complete with casing. With a heatedwater bath and a thermometer, she has been testingit to see how well it can measure temperatures. Sofar, the results have been rewarding.

The tabletop radiometer, needless to say, ismore portable than any X-ray system used forrecording mammograms. It does not involveshielding, isotope handling, or specially builtunits, and is almost a hundred times cheaper tomanufacture. It runs on two AA batteries.

Measured properties of PVAL solutions can becompared with known tissue behaviour

Courtesy: Dr. Arunachalam

Along with the practical applications ofmicrowave radiometry for imaging come a wholehost of complexities. Rachana S Akki, also a researchscholar in Dr. Kavitha’s group, is working to identifyand model the factors that a�fect the quality ofthe scan and develop protocols that will ensure itsreliability. Some factors, she explains, are beyondour control – the size and depth of the tumour,for instance, and the balance of fat and glands inthe breast tissue. Because the amounts of powernormally emitted from di�ferent patients’ breastsare di�ferent, the device must be able to tell whether

a change in the measurement is because of thepresence of a cancer, or merely due to diversity inthe composition of the breast tissue.

The elimination of subjectivity, I realise frommy conversations with Rachana and Dr. Kavitha, isthe ideal that medical imaging of all sorts constantlytries to attain. A scan that can be carried outany number of times, by di�ferent people withdi�ferent skill levels, and still look the same, is ascan that can be trusted. An ultrasound probe, forinstance, with all its constant movement, o�fers nohope of obtaining such a scan reproducibly, evenwhen the person operating it is skilled. In bothX-ray mammography and radiometry, the chancesof such disturbances are reduced by keeping thedevice steady and compressing the breast betweentwo plates. This compression is necessary with X-rayso that large volumes of tissue do not absorb toomuch radiation. It also gives a better quality scan,since the healthy tissue is compressed more easily,bringing the tumour closer to the surface. But X-raymammograms are painful, with the breast beingcompressed to half its size. Rachana emphasises thatthe microwave radiometer, on the other hand, canwork with much, much lower levels of compression;about a quarter or even a fi�th.

A scan that can be carried out anynumber of times, by di�ferent people with

di�ferent skill levels, and still look thesame, is a scan that can be trusted.

To model the breast, Rachana preparedsolutions of a chemical called polyvinyl alcohol(PVAL) in water, changing the amount she addedeach time, to get gels with di�ferent properties.A�ter making a wide range of gels, she studiedtheir sti�fness and other mechanical properties.Combining available data about how fatty, glandularand mixed breast tissues behave under compressionwith the results of her experiments on thesePVAL ‘phantoms’, she was able to deduce whichPVAL solution would mimic which kind oftissue. This allowed her to design and control

��| Engineering Design

computer simulations of the breast, and to extractinformation, for instance, about the power emittedby hotspots in fatty or glandular breasts, whencompressed to di�ferent extents. One way in whichthis is relevant is that a more glandular compositionleads to more discomfort under compression,and this has to be taken into consideration whiledeciding on imaging procedures. These studies alsogave Vidyalakshmi an estimate of the power levelsthat the front-end must be designed to take, as input.

Developing protocols for imaging by studyingall these factors will someday allow the device to bemade suitable for use in clinics in urban as well asrural areas where, unlike in the lab, the environmentis not controlled. That is why it is importantto carefully look into which factors influence themeasurement, their relative significance, how muchthey may vary, and the corresponding e�fects ofsuch variations on the results. Researchers can thenoptimise the parameters that can be controlled, togive reliable results while minimising cost, pain anddiscomfort. “If a defined protocol is there, thenthe person who is handling the device will have achecklist,” says Dr. Kavitha. “That will give us greaterconfidence that a hotspot detected is from the tissue,and not from the influencing environment.”

As a supplement to chemotherapy,microwave hyperthermia acts by

improving the delivery of the drug totumour cells.

The microwave radiometer is intended tobecome an alternative screening tool, one that canavoid unnecessary exposure to ionising X-radiationduring regular screenings. If a cancerous growthis suspected, however, X-ray mammography wouldstill remain the golden standard. Apart frompreliminary scans, the microwave radiometer couldalso be used for intermediate screenings to checka patient’s response to treatment – currently doneusing infrared thermograms – and for follow-upsthat check for recurrence, which may currently tendto combine ultrasound and X-ray examinations.

Looking into how microwave can be used alongwith these techniques could help increase the scandepth while reducing the risk of exposure to harmfulradiation.

It turned out that most of the pieces ofequipment that I saw in Dr. Kavitha’s lab hadbeen built in-house; these include, amongst otherthings, a large variety of antennae. The antennathat Rachana has made for the radiometer is a smallcircular, nearly flat one, like a stethoscope disc. Anantenna can cover either a wide angle up to a shortdistance, or a long distance in a specific direction– and this applies to both microwave transmittingand microwave receiving antennae. In this case, theantenna needs to collect radiation from as large aregion as possible, and the source of this radiationis not too far away. So a wide angle antenna is bestto use. In other applications, directionality becomesimportant. For microwaves can be used to not onlydetect cancers but treat them as well, and in thise�fort – with the location of the tumour becomingknown – an irradiating antenna must transmit in avery specific direction.

Another of the PhD students, GeethaChakaravarthi, is working on one such application,and the lab shelves are dotted with antennae andother devices she has prepared to investigate the useof microwaves to kill cancerous cells. This technique,called hyperthermia, works by heating up cancerouscells by focussing microwave radiation upon them.Hyperthermia supplements chemotherapy andradiotherapy as well, by di�fering mechanisms.

Prototype of the microwave radiometer

��| Immerse

Applicator placement on a healthy volunteer during preclinical pilot studyCourtesy: Dr. Arunachalam

Radiotherapy, where cancer cells are killed withX-rays, works better in those parts of a tissuewhich are rich in oxygen. Thus, tumour cells aremore sensitive to radiotherapy if they are locatednear a major blood vessel. Microwave therapy,on the other hand, works better where the bloodsupply is poor, because the blood is not able toe�fectively transport away the extra heat. It can thusreach areas of the tumour where radiotherapy fails.Together, radiotherapy and hyperthermia can defeatthe tumour more e�fectively.

. . . it is very important to get rid of airbubbles in the water if the microwave is

to reach deep tumours.

As a supplement to chemotherapy, microwavehyperthermia acts by improving the delivery ofthe drug to tumour cells. In chemotherapy, theblood carries a drug to destroy the tumour, theidea being that the higher metabolic activity ofthe cancer cells will lead them to take up moreof the drug than healthy cells do. Heating thetumour with microwaves e�fectively forces the bodyto raise circulation in the heated region in an e�fortto regulate its temperature, and more circulationmeans better drug delivery. It is crucial that theheating be highly localised, so that there is as littledamage as possible to healthy cells.

While it is known that microwave therapy canimprove radiotherapy and chemotherapy results,e�fective devices have not yet been developed forthis. The ongoing e�fort in Dr. Kavitha’s lab isto develop patient-friendly devices well-suited totreating both large and small tumours with verysite-specific application of therapeutic microwaves.The ‘patch applicator’ which Geetha has developedis the smallest one currently available and itsgently-concave surface, unlike the rigid ones of mostof its predecessors, allows it to rest comfortably onthe patient’s skin.

To get microwave to the tumour, a transmittingantenna sends radiation through a water bag placedon the skin (without the water bag, the antennawould burn the skin). Computer simulations aswell as clinical measurements done by the grouphave shown that it is very important to get ridof air bubbles in the water if the microwave isto reach deep tumours. But ‘degassing‘ systemsfor getting bubble-free water are expensive, eitherincorporating specialised bubble traps or filters, ordegassing the entire liquid before use. And since lesspower reaches the target cells if bubbles are present,any ine�ficiency in removing bubbles makes higherdoses of radiation necessary to achieve the samee�fect.

Geetha’s work has led to a new degassingsystem that is much cheaper than existing ones. In

��| Engineering Design

this setup, a pump circulates water through pipeswhile an electronic feedback system maintains thevolume and temperature of water in the applicator.Equipped with sensors, the degassing system uses

A cost effective Inline degassing systemCourtesy: Dr. Arunachalam

its control over flow rates to remove bubbleswith a vacuum chamber. Because it can degas

a circulating fluid e�ficiently and economicallywithout disrupting the ongoing process, thisinvention has the potential to revolutionise anumber of other medical procedures – dialysis beingone case in point.

Both medical imaging and therapy, as far ascancer is concerned, are fields dominated todayby toxic chemicals and high-energy radiation.Technologies in both fields are far from ideal, but thegoals are clear. Imaging needs to be as non-invasiveas possible, with higher and higher degrees ofaccuracy. Therapy needs to be as specific as possible,with little or no e�fect on healthy parts of the patient’sbody. In the battle against cells that can look justlike their healthy sisters, and invite widespreaddestruction with every e�fort to kill them – butwhich cannot conceal their own heat footprint –there is little doubt that the future will see enormouscontributions from microwave research. ⌅

Meet the Author

Shivani Guptasarma grew up in Chandigarh, where she attended school at the Sacred

Heart Senior Secondary School for girls and developed interests in all areas that are

currently classified under the STEM subjects. She joined IITM in 2014 for a B.Tech.

in Engineering Design and an M.Tech. in Biomedical Design. She is excited about her

courses because they allow her to continue to study biology, maths, electrical and me-

chanical engineering, with prospects of someday developing insights and tools to help

human beings.

��| Immerse

Inside Healthcare PolicyBy Isha Ravi Bhallamudi

Image credit: Prashanth NS via Wikimedia Commons

Much like research in Science, Humanities and Social Sciences

research is multifaceted and can be approached from a variety of

perspectives, methodologies and tools. To o�er a glimpse into re-

search in the social sciences, this piece explores the field of Health

Economics and Public Policy through an interview with Prof. VR

Muraleedharan from the Department of Humanities and Social

Sciences at IITM.

��| Humanities and Social Sciences

In a career spanning three decades, Prof. VRMuraleedharan (or Prof. VRM) has madeenormous contributions to research onpublic health policy. In order to learn more

about what policy research entails and understandthe complexities involved in the field of healthcarepolicy, I found myself stepping into Prof. VRM’so�fice one sunny evening for an interview about hiswork and extensive research experience in this field;in particular, his several recent and very excitingprojects.

Policy Research: The Big Picture

Policy refers to a broad set of decisions, plansand actions undertaken to achieve specific goalsin a region or nation. In the field of healthcare,formulating e�fective policies is crucial as goodpolicies can have profound e�fects on the state ofhealth in a particular area. For example, a policysubsidising contraceptives in HIV-prone areas couldin the long run lead to drastically reduced rates ofdisease. To ensure that there is a high level of qualityand access to healthcare, it is necessary to have wellthought out, e�fective healthcare policies. Comingto what policy research is: it involves examinationof the design and process of policy making andimplementation, evaluation of policy outcomes,and also an analysis of factors that constrain thee�fectiveness of policy, including figuring out exactlyhow and why particular policies worked or may workunder certain circumstances. “Normally, a policy isviewed as a black box: the interest lies in the inputsor elements that form the policy, and the output oroutcomes of the policy”, Prof. VRM points out. “Butworking on policy research means being interested

in what goes on inside the black box, understandingthe pathways and dynamics that make particularpolicies work. So, for e�fective policy analysis, youhave to open the box and correlate the two.”

To ensure that there is a high level ofquality and access to healthcare, it isnecessary to have well thought out,

e�fective healthcare policies.

Is there any one broad theme that reflectsthe essence of policy-oriented research taking placeacross the projects that Prof. VRM has been involvedin? “Our focus across several projects has centredon one single question, one I am very fond of”,Prof. VRM says thoughtfully. “The first part of thisquestion is the realisation that every rupee spenton one person is a rupee denied to another – assomeone working in the development field, this isa daily chant. But the second part is a criticaleconomic question from the public point of view:is that rupee well spent, given that someone elseis being denied it?” This really is a very interestingand tough question. To illustrate: the governmentbudget this year for Tamil Nadu for health is about|���� crores – and this money was spent witha view to certain benefits and their distribution.“But what we would like to know”, Prof. VRMsays, “is how it was distributed across di�ferentsocioeconomic spectra, the particular benefits ofpublic spending on healthcare, how equitably thebenefits of government spending are distributed,how it can be improved with a sense of equity andhow much of the pie do the poor get in terms ofbenefits.”

In a research career spanning over three decades, Prof. VR Muraleedharanhas held multitude of significant roles in academia, research, and the policysphere including as Member of the Mission Steering Group of the National RuralHealth Mission, Govt. of India; Senior Researcher for national and internationalbodies such as DFID, and full Professor since 2000 in the Humanities and SocialSciences department, IITM. He is also an IITM alumnus, having completed hisPhD here in 1988.

��| Immerse

A Deeper Look at Policy OrientedHealthcare Research

Across Di�ferent Time Periods and Regions

A large part of Prof. VRM’s work has focusedon studying the costs of, access to and coverage ofhealthcare in Tamil Nadu, especially by comparinghealthcare interventions and health indicators inTN to those in other regions/states of India andcountries, and using the research findings to cra�tconstructive healthcare policy. For example, onesuch project, carried out between ���� and ����,is titled ‘Good Health at Low Cost, �� Years on: WhatMakes a Successful Health System?’. It carried forwarda research project on comparing healthcare acrossa particular set of countries which was carried outin ����. �� years later, it seeks to analyse how andwhy each of these and other countries accomplishedsubstantial improvements in health or access toservices or innovative health policies relative toeconomically comparable regions or countries . InIndia, only the state of Tamil Nadu was studiedbecause the scale of diversity in India makes itdi�ficult to generalise such a study for an entirecountry based on a few states.

We have several programs in India,targeted for particular diseases for

example, and the same programs in everystate. But some do better than the others.

How does one explain that?

This project necessitated analysing the past ��years of healthcare in Tamil Nadu and was carriedout by Prof. VRM and Prof. Umakant Dash fromthe HSS department. This was done using their pastextensive experience and research in the field as wellas more than �� interviews with higher level o�ficialswho could explain policy changes, had worked atthe district level before and were closely involvedin the implementation of various programmes.Speaking to higher level o�ficials who had worked in

di�ferent states was the best way to get a comparativeperspective relative to other Indian states and findout how TN made use of certain financing measuresand central government program features to achievea higher level of healthcare. The project found thatTamil Nadu had not spent lavishly on healthcareuntil ���� and even a�ter ����, when the NationalRural Health Mission was instituted. “We wantedto look at the ��-year period before that to see howplaces that spent relatively little on health managedto bring out better health outcomes”, Prof. VRMexplains. “We have several programs in India,targeted for particular diseases for example, and thesame programs in every state. But some do betterthan the others. How does one explain that?”

One way is to construct the story behind theseevents: all �-� central secretaries interviewed duringthis project gave the impression that “Tamil Nadu isgood at seizing money fast when there’s a big poolof money for allocation.” But the other question tobe asked is what percentage of the allocated moneyis spent e�fectively, that could have positive impacton health outcomes. Many states don’t spend a highpercentage, as e�fectively, but instead underspendthe allocated funds and attribute the relatively pooroutcomes to a lack of capacity. “This is a veryinteresting point. Tamil Nadu spends relativelymore e�fectively than others. If ��% of the moneyis spent e�fectively, and say ��% goes through otherhands (meaning, down the drains), that ��% is stillquite well spent. But it’s the other way round in otherstates, as we found a�ter distilling our observationsand interview responses over several years”, Prof.VRM explains. But this leads to a third, evenmore interesting point. “If I have spent ��% of theallocated money well, and you have spent ��% ofyour allocated money well, this ��% di�ference inspending, cumulated over �� years, makes a hugedi�ference in terms of outcomes.” This di�ference,if repeated consistently, and aided by other factorssuch as an e�ficient bureaucracy, a diligent work ethic,supported by other systemic factors such as goodroads and transportation, and media, add up to acumulative di�ference that counts for a lot. This

��| Humanities and Social Sciences

o�fers one explanation for the relatively positivehealth outcomes in the state.

Consortia for Comparative Research

Comparing di�ferent health systems to arriveat better policy practices for a particular regioncan be carried out in several ways. In recentyears, Prof. VRM has been involved with atleast two research consortia that seek to do justthis. The first, the Consortia for Research onEquitable Health Systems (or CREHS), carried outcomparative research between ���� and ���� in sixcountries – India, Nigeria, South Africa, Thailand,Kenya, and Tanzania – to generate knowledgeon how to strengthen health system policies andinterventions in ways that would “preferentiallybenefit the poorest”, such as by examining theimpact of mobile health units on access to care.The second consortium, which evolved from CREHS,is called RESYST (Resilient and Responsive HealthSystems), and aims to enhance the resilience andresponsiveness of health systems to promote healthand health equity and reduce poverty.

Working in a consortium necessarily meansthat a lot of time is spent in sharing and discussingeach stage of the research process across countriesand teams. This involves regular meetings, timespent to structure the content and regularity ofthe meetings, arriving at common questions,developing a methodology and research instruments(the questionnaires) together, and interpreting andsharing the findings. Comparative studies take upa lot of time because all the teams involved have toreach a consensus on common questions that aremeaningful in each country and are comparableacross them as well, and they must establish clearlywhat each country gets out of the exercise. Eachstep of the research process for one team must bein tandem with the steps taken by the other teams,and it can be di�ficult to maintain an equitablerhythm and balance while carrying out the workover a long period of time. There are internal checksand balances and timelines to ensure that the workproceeds relatively smoothly and sub-groups that

keep moving back and forth before arriving at asatisfactory conclusion.

How does international comparative researchhelp cra�t good policy at country or state level? Priorexperience shows that learning from the experiencesof other countries as well as our own historyhelps construct e�ficient and well-constructedinfrastructure and delivery structures. “The impactof research on policy here is not a linear, director clear relationship because it is di�ficult topredict exactly where, when and how researchinfluences policy”, Prof. VRM remarks, “butwe have very interesting ways of capturing thisrelationship.” One aspect of this is the engagementbetween researchers and policymakers, which buildsgradually and takes o�f over time. “For example, eachof the hundred odd meetings and talks I have hadwith the government the past year is evidence of myphysical and mental engagement – and its impact isdi�ferent from handing in policy reports that nobodyhas time to read anyway (even if they want to). It isimportant to find ways to engage as researchers withpolicy makers in your own way and style”, says Prof.VRM.

The impact of research on policy here isnot a linear, direct or clear relationshipbecause it is di�ficult to predict exactly

where, when and how research influencespolicy.

Insights from Grounded, Participatory Policy Re-search

However, policy oriented research work hasits own complexities. One way to illustrate themis through Prof. VRM’s ongoing project on UniversalHealth Coverage (UHC), which seeks to pilot thisconcept in Tamil Nadu for the state government intwo districts, one of which is Krishnagiri. In thisdistrict the research is being carried out specificallyin the block of Shoolagiri. The project has beenongoing for around six months.

��| Immerse

The Thottapattu sub-centre, or lowest unit in the healthcare delivery system (serving 5,000 - 6,000 people), inTamil Nadu. (From right to left) Prof. Umakant Dash, the head nurse who manages the entire sub centre, and

Prof. VRM (2013).

For policy to work out e�fectively inpractice, the research must also

incorporate the psyche, ecology, terrain,geography and multitude of other factors

surrounding the region.

Piloting the UHC involves a large numberof household surveys, facility surveys, groupdiscussions and focus group discussions acrossvillages and intense discussions with fieldfunctionaries. This is with the objective of collectingground level knowledge of illnesses, learning theexpectations of the villagers, and keeping track ofthe facilities that are currently functional on ground.“For this, the complete mapping has been done tofirst assess what is present on the ground. Next, it isimportant to find out how much people are spendingout of pockets for healthcare (this was capturedthrough a large survey of ���� households) andrecording health seeking behaviour of the villagers ofthe last one year, including for deliveries, prenatal,postnatal care, immunisation, access to public orprivate facilities, out of pocket expenditures for

various illnesses, and so on”, explains Prof. VRM.Further, state level consultations take place ondeveloping an Essential Health Package or EHP,including its contents and ways to guarantee itsdistribution through a publicly financed system.“Coming up with the EHP involves an intricate set ofnegotiations which include consultations with statelevel o�ficials, field functionaries and people living inthe villages. Thus, the bottom-up views are collectedalong with the expectations residents have from theEHP”, says Prof. VRM. This material, which reflectspeople’s voices and their needs, is used to then reflecton what is doable and what is expected. It helpsnegotiate di�ferent meanings and consequences ofthe EHP and proceed forward to arrive at a packagethat combines the needs and expectations of all inan equitable way.

“For policy to work out e�fectively in practice,research must also incorporate the psyche, ecology,terrain, geography and multitude of other factorssurrounding the region”, Prof. VRM emphasises.For example, one peculiarity particular to Shoolagiriis that people speak three languages, with di�ferentlanguages used for di�ferent activities. Thus,taking this into account, particular areas and

��| Humanities and Social Sciences

policy recommendations have to be treated withsensitivity: for example, as Prof. VRM argues,“you cannot place someone from Tirunalveli as aVillage Health Nurse (VHN) into Shoolagiri; shewould speak not just Tamil but a di�ferent dialect ofTamil.” Thus, even just the process of recruitmentin public systems is one that is fraught full ofproblems. Such issues may arise at di�ferent partsof the research work or policy implementation andmust be anticipated (or at least, mechanisms forswi�t redressal conceived) in order to ensure smoothfunctioning of policy.

Research Narrative: Coming Full Circle

My last question is one that perhaps shouldhave been the first. How did Prof. VRM find himselfin this field? “A�ter completing an MA in Economicsfrom BITS Pilani followed by a two year break”, Prof.VRM says, “I found myself engaged in two researchprojects across Maharashtra and later as a researchassistant on a project on assessing PHCs in Orissain ����. Around this time, aided by two wonderfulresearch guides, I travelled all across five districtsof Orissa and developed an interest in healthcare.Studying healthcare systems in Orissa in that timewas very tough, and that project, the field work andthe travelling stimulated my interest in the field. SoI really owe a lot to Orissa!” I am surprised to findout that Prof. VRM is an alumnus of IITM, havingcompleted a PhD here on the history of healthcare in

South India under a renowned economic historian,Prof. S Ambirajn, who taught at HSS IITM from���� till ����. Prof. VRM’s thesis, and subsequentresearch, was based on archival work. A�ter a oneyear sabbatical at Harvard in the ��s, Prof. VRM’sfocus shi�ted more towards more recent policy, andthis has shaped his current research interests andwork. In a similar vein, he also enjoys guidingresearch scholars from diverse disciplines, thoughthe general rule is that scholars whatever be theirdisciplinary background, should have an interest inpublic health policy issues.

Speaking about how all facets of policy researchtend to come together, Prof. VRM says, “Rightnow, we are working towards using our work inRESYST to help inform the UHC project; especiallyfor increased exposure.” In fact, the UHC projectitself has also tied into yet project funded by USAid,where � research institutions in India are trying topilot UHC in eight Indian states (and this is just onecomponent out of three, of this project). “By now,it’s di�ficult to say what one thing I am working onby way of a funded subject, because as you can seeall these projects are an organic evolution, and areconnected in essence.” And as Prof. VRM points out,“You know, despite doubts about how meaningfulyour work is, you continue your research and keeppushing the frontiers of policy studies through yourengagement with policy makers . . .and, naturally,there is no end to this process.” ⌅

Meet the Author

Isha Bhallamudi is a fourth year Integrated M.A. student of the Humanities and So-

cial Sciences majoring in Development Studies. She has been involved in writing about

research and innovation through Immerse and T5E. Her research interests lie in policy

research and its cross connections with health, poverty and gender. Isha can be reached

at b.isha.ravi@gmail.com for comment, criticism or discussion!

��| Immerse

Born an Entrepreneur

by Ananth Sundararaman

Dr. Muhammad Yunus has said multiple times that “All humans are born entrepreneurs.” Therevolutionary idea of Grameen bank has brought millions of people, employed in small enterprises such as

farmers, out of poverty. Dr. Arun Kumar and Dr. Suresh Babu at IIT Madras are trying to synthesiseavailable evidence on the links between microfinance and poverty.

It was ���� when Bangladesh was facingone of the worst famines in history.A�ter quitting his job as a deputy chiefof the General Economics division in

the government’s planning commission, Dr.Muhammad Yunus was serving as the head of theeconomics department at Chittagong University fora little over a year then. It was during this periodthat the inspiration for Grameen Bank came to Dr.Muhammad Yunus in the form of a trip to the villageof Jobra in Bangladesh.

What began by helping a woman strugglingto make ends meet as a weaver of bamboo stoolssoon became a model to assist families avoid thehigh interest rates under predatory lending. Whileallotting small loans of $�� to a group of �� familiesas start-up money, little did he realise that he waslaying the foundation for the now familiar GrameenBank and what would go on to be hailed as a ‘miracle

cure’ for global poverty – microfinance. He went onto win the Nobel Peace Prize in ���� for his e�fortsin creating one of the world’s most high-profile andgenerously funded development interventions.

“Microfinance to a layman is basically acommunity initiative to induce very small savingsfor the people by pooling their savings and lendingwithin the group”, says Dr. Arun Kumar, a professorin the Department of Management Studies (DoMS).Dr. Arun Kumar, along with Dr. Suresh Babu fromthe Department of Humanities and Social Sciences,has spent a considerable amount of time in the pasttwo years visiting rural villages across various statesin India. The primary aim of microfinance is to helppoor people having minimal experience in dealingwith banks achieve a sustainable source of cash flowon a monthly basis. Aimed at people employed insmall enterprises such as petty shops, tailoring andeven farmers especially in rural areas, this concept

��| Management Studies

Dr. Arun Kumar G is currently engaged as a professor in the Department of Man-agement Studies, IIT Madras, and is involved in teaching Corporate Governance,Financial Accounting and Mergers & Acquisitions. His research interests lie pri-marily in Development Finance and Joint Ventures & Alliances. He has co-authoredtwo textbooks on Management Accounting and can be reached at garun@iitm.ac.in

has been a breakthrough in unlocking immenseopportunities for them. “It has even gainedtraction in urban areas, and also involves a gendercomponent”, informs Dr. Suresh Babu about theactive participation of women and their role in selfhelp groups.

Microfinance is basically a communityinitiative to induce very small savingsfor the people by pooling their savings

and lending within the group.

The concept of microfinance is to findcollaborative ways to meet the needs of a group,primarily through creating and exchanging cashwithin the group. Consider a group of �� women,and let us say each of them contribute |��� towardspooling their savings in the group every month. Thistotal amount of |���� can be given as a loan to oneof the members to conduct her business under thecondition that in the second month while all theother members contribute |��� each, she repays aninstallment of her loan by contributing |��� withinterest on top of her monthly contribution. In thesecond month, the group now has a little over |����which can be given as a loan to some other memberin the group. By doing this on a rotating basis, thegroup continuously increase their savings and alsokeep a check on each other so that they do not fail torepay the amount.

This model has been a great success incombating the fact that a large part of the ruralpopulation is not a part of the formal financialsystem. By eliminating the middleman, borrowing

from a money lender at high interest rates and beingtrapped in a vicious circle of debt for generationsis no longer their only option. By being a part ofthis kind of a group an individual can manage toraise a sizeable amount as loan without collateraland at a reasonable rate of interest. The fact that theborrowers paid Muhammad Yunus back in full andon time spurred him to start travelling from villageto village, o�fering more tiny loans and cutting outthe middlemen.

Dr. Yunus was determined to prove that lendingto the poor was not an ’impossible proposition’, andGrameen bank adopted its signature innovation:making borrowers take out loans in groups of five,with each borrower guaranteeing the others’ debts.Thus, in place of foreclosure (banks selling theproperty of the loanee to recover the debt they areowed) and a low credit rating that usually definesborrowing from a bank – Grameen depends on anincentive at least as powerful for poor villagers:the threat of being shamed before neighbors andrelatives.

Dr. Arun Kumar and Dr. Suresh Babu haveundertaken research on this financial system overthe past year, in collaboration with the Departmentfor International Development (DFID), UK andHand-in-Hand, a non-governmental organisation(NGO) headquartered in UK. Hand-in-Hand was infact started in Tamil Nadu by Dr. Kalpana Shankar,the wife of a district collector in Coimbatore withthe help of Dr. Percy Barnevik, a Swedish businessexecutive and philanthropist.

Interestingly, Dr. Kalpana earned her PhD in

��| Immerse

A microfinance meeting in progress in Kerala; Image Publicly available via Wikimedia Commons

nuclear physics and had little training in financeand economic development. It is today presentin over ten countries across the world includingAfghanistan and Lesotho. Their mission is towork for the economic and social empowerment ofthe poorest and most marginalised population bysupporting the development of businesses and jobs.They receive funding from a number of di�ferentsources including individuals, corporations,bilateral and multilateral institutions, and trusts andfoundations. As of ����, Hand-in-Hand India hadcreated �.� million jobs, �,��� Citizens’ Centres andcovered over � million households under its variousprogrammes.

The terms of the project were to synthesiseavailable evidence on the links between

microfinance and poverty.

“There were two factors that led to the both ofus joining forces to work on this area”, says Dr.Suresh Babu. “He (Arun) works in finance while I aminterested in development processes. Microfinanceprovided an overlap in terms of finance and

development and we discovered that there is somepotential research that can be conducted together inthis area.” They had individually worked on previousassignments for Hand-in-Hand and decided tocollaborate on research when a new assignment onmicrofinance was o�fered.

The terms of the project were to synthesiseavailable evidence on the links betweenmicrofinance and poverty. “We were assessing if theinvestments into the NGO were yielding the resultsthey were expected to and if there is a social returnon those kinds of investments”, says Dr. Arun. TheNGO assisted them with their field based research invisiting di�ferent villages from Pali in Rajasthan toCuddalore in Tamil Nadu where they had a presence.They looked into various cases of self-help groupsand communities which received support from theNGO to create a system of lending and borrowingamong the community.

“We came across di�ferent scales of operationduring our field visits. One of the most successfulcases is ‘Kudumbashri’ in Kerala.” A female-oriented,community-based, poverty reduction project thatwas initiated by the Government of Kerala,

��| Management Studies

‘Kudumbashri’ receives regular aid and assistancein conducting their activities and increasing theirreach from the government. Their strategy is oneof forming women’s collectives in di�ferent villagesand provide skill upgradation and training sessions.Small savings generated at the families are pooled atvarious levels as thri�t and used to attract credit frombanks to support micro-enterprises for sustainableeconomic development. One of the key reasons forits success is cited as the ability of any woman tobecome involved with the organisation irrespectiveof whether she is below or above the poverty line. Byconducting thorough background checks, the needfor a voter ID card or a valid identity proof becomesunnecessary and thus paves the path for the wideimpact that Kudumbashri has managed to create.There have also been individual instances of successstories that have been detailed in the research.

While the ultimate goal of this systemis to generate employment and sustain localentrepreneurial businesses, this may not always bethe case. By visiting di�ferent villages across thegeography of the country and covering a wide rangeof specific industries from the ��,��� villages wherethe NGO operates, they concluded that while themodel leads to a tangible benefit for borrowers, theremay be specific situations when the system may not

work. For instance, money borrowed to pay tuitionfees for children would lead to a case where themoney borrowed does not have an immediate returnon investment. This may lead to a situation wherethe model may break down as the resources pooledin by di�ferent members help only a few individualswhile the rest would have to wait for a long timeto see the amount be repaid. Both Dr. Arun andDr. Suresh have spent a lot of time in analysingdi�ferent types of SHGs and have compiled a reportdetailing the complete spectrum of the impact ofmicrofinance in helping eradicate poverty in India.

Moving forward, the goal is to conduct researchto understand more about the consumption ofthese loans by the people taking them. The aimof microcredit is, a�ter all, to teach the financiallydisadvantaged the basic financial principles theyneed to sustain the growth that is initiated by SHGs.While it’s not a one stop tool for the eradication ofpoverty, it is definitely an important precondition toeconomic development. In the words of Dr. Yunus,“All human beings are born entrepreneurs. Someget a chance to unleash that capacity. Some nevergot the chance, never knew that he or she has thatcapacity.” ⌅

Meet the Author

Ananth Sundararaman is a third year undergraduate student in the Department of Civil

Engineering at IIT Madras. He is also pursuing a Minor in Economics. He spends his

time between his hobby of quizzing and watching Liverpool play, come the weekend. A

believer in the underlying philosophy of Asterix, his biggest fear is the sky falling on his

head.

��| Immerse

Perhaps one of the most importantaspects of society is communication.It’s what allows diverse populationsfrom across the world to cooperate and

socialise, and ideas and opinions to spread aroundthe globe and make us a truly global community.

The Allied e�forts to crack Enigma,largely aided by the work of Alan Turing,

marked the beginning of the era ofcomputers - an era that would see

cryptography mutating into awell-defined field of study.

From writing on papyrus scrolls to sendingmessages instantly around the world, methods ofcommunication have evolved rapidly over the years.And yet, a concern that has never ceased to existis that of privacy. The need to keep importantmessages safe from prying eyes and ears resulted inthe field of study now known as cryptography.

The word conjures to mind images of hackerslocked in battle with cryptographers, attemptingto ferret out secrets of international import. Butcryptography has far older origins than one mightthink. The oldest use of codes can be traced toEgypt in around ���� BCE, where non-standardhieroglyphs were found carved into stone. Sincethen, codes and ciphers grew progressivelycomplicated, from the Caesar Cipher employed byJulius Caesar (simply shi�t every letter of the alphabetto the le�t or right by a fixed number of letters) tothe supposedly unbreakable Enigma employed bythe Germans in WWII. The Allied e�forts to crackEnigma, largely aided by the work of Alan Turing,marked the beginning of the era of computers - anera that would see cryptography mutating into awell-defined field of study.

It is in this field that Dr. Santanu Sarkar,from the Department of Mathematics, works. Hisresearch considers encryption, as he says, “from theattacker’s point of view.” It concerns the use of

mathematical constructs called lattices in attemptsto break the RSA cryptosystem, one of the mostcommon encryption methods in use today.

Before explaining his research, Dr. Sarkaroutlines the history of modern cryptography.“The foundations of modern cryptography werelaid in ����, when Whitfield Di�fie and MartinHellman published a revolutionary paper.” Thispaper outlined a new concept – the public-keyencryption system. Till that point, cryptosystemsused symmetric-key encryption. This meantthat both the receiver and sender of informationused a single, shared key (a term for a largenumber used in the encryption process) toencrypt plaintext (unencrypted information) and todecipher ciphertext (encrypted information). Thisnecessitates the use of a secure channel for the keyto be shared between the sender and the receiver.But the necessity of a secure channel in order toset up a secret key was a practically insurmountablechicken-egg problem in the real world.

Di�fie and Hellman’s paper, on the other hand,posited that a shared key was not necessary. Instead,they proposed that two keys be used - a publickey and a private key. The public key would beavailable to anyone who wanted to communicatesecurely with a system, while the private key wouldbe known only to the system itself. Encryption wouldbe carried out with the public key and decryptionwith the private key. It was also mandatory thatthe private key not be deducible from the publickey, as that would compromise the system’s security.Since the private key didn’t need to be shared withanyone via a potentially insecure channel, public-keyencryption was clearly a better choice.

Although Di�fie and Hellman were able to provethe feasibility of such an encryption system, theyweren’t able to come up with a viable systemthemselves. That was accomplished two yearslater, by Ron Rivest, Adi Shamir and LeonardAdleman, researchers at MIT. That eponymouscryptosystem ‘RSA algorithm’ has since become oneof the strongest known encryption standards in theworld.

��| Immerse

Dr. Santanu Sarkar is currently an Assistant Professor in the Department of Mathematics

at IIT Madras. He was previously a guest researcher at the National Institute of Standards

and Technology. He received his PhD from the Indian Statistical Institute, Kolkata. His

main research interests include cryptology and number theory.

It is now used mainly as part of hybrid encryptionmethods: data is encrypted using a symmetric-keysystem, and the shared key is then encrypted usingRSA. This is largely because of the RSA algorithm’scomputational ine�ficiency – encrypting the dataitself using RSA would take a very long time.

At a basic level, the RSA algorithm is based on thepremise that the product of two large prime numbersis very hard to factorise. Put into mathematicalterms, consider � prime numbers, P and Q, andtheir product, N . Then, the integer solutions to theequation p(x, y) = Nxy are the factors of N . Trivial(irrelevant) solutions to this equation include (x =

1, y = N) and (x = N, y = 1). The importantsolutions, though, are (x = P, y = Q) and (x =

Q, y = P ). While a computer can solve this equationrelatively fast when N is small, larger values of Nresult in runtimes that render decryption attemptsinfeasible. For example, a ����-bit value of N (i.e,N ⇡ 21024) will take approximately 10211 years tofactorise. To put things in perspective, the age ofthe universe is around 1010 years! In cryptographicterms, here N is the public key and P and Q are theprivate keys.

At this point, Dr. Sarkar mentions a caveat.“RSA encryption can’t be broken by conventionalcomputers. However, there is an algorithm, calledShor’s algorithm, that can be used to break RSAencryption using quantum computers.” But sincequantum computers are still in nascent stages ofdevelopment, the RSA algorithm is still consideredto be a bastion of cryptography.

The mathematical framework above describesthe first and simplest RSA variant. There have been

others proposed since then, and it is on one ofthese variants that Dr. Sarkar works. This variantconsiders the equation ed = 1 + k(N + s). eand N are known (and hence, are public keys) andd, k and s are unknown (and hence, private keys).This can be expressed as a polynomial p(x, y, z) =

ex�1�y(N+z). As before, obtaining the non-trivialsolutions of this polynomial is equivalent to breakingthis particular variant of the RSA encryption.

At a basic level, the RSA algorithm isbased on the premise that the product oftwo large prime numbers is very hard to

factorise.

Decrypting any variant of the RSA algorithmwas considered infeasible until Don Coppersmith,an American mathematician, established a theoremrelated to the factoring of numbers. When usedin conjunction with mathematical constructs calledlattices, it was found that the RSA algorithm couldbe broken in polynomial time (This means that therunning time is a polynomial function of the inputsize. It’s largely used to denote programs whoserunning times don’t blow up too fast). Fortunatelyfor cryptographers around the world, the guaranteeof success for such an attack was attached to certainconditions.

The encryption could be broken in a feasibletime scale only if d, one of the private keys, was lessthan N0.292 - which implies that for a secure RSAdesign, d would have to be greater than N0.292. ButDr. Sarkar prefers to think of it as an upper boundfor the system to be vulnerable, rather than a lower

��| Mathematics

A schematic explaining how public-key encryption works.Image source: Wikimedia Commons

bound for security. “I always look at the problemfrom the attacker’s point of view”, he says with asmile. “Hence, I think of values of d for which thesystem is insecure.” This bound was proven by twocryptographers, Dan Boneh and Glenn Durfee in����. For example, if N was a ����-digit number,the concerned RSA system would be secure as longas d was a ���-or-more digit number. However, themathematical community conjectured that for theRSA system to be truly secure, d would have to begreater than N0.5. Using the example from before, dwould have to have more than ��� digits.

I always look at the problem from theattacker’s point of view.

Any increase in the bounds on d would havetwo consequences. First, it would expose any RSAsystems that used values of d below the new boundas insecure. Secondly, an increase in the value of din any RSA system results in a significant increasein the time taken to decrypt it using Coppersmith’stheorem. Hence, improving the lower bounds on d

contributes greatly to improving the security of RSAsystems used across the world.

Dr. Sarkar goes on to explain that he worked

on a further variant of this RSA scheme. “N doesn’thave to be only a product of two prime numbers.It can instead be of the form N = P rQ, where ris another integer. I worked on proving bounds ond when r = 2.” Dr. Sarkar’s work improved thebounds on d from d < N0.22 to d < N0.395. Talkingabout the implications of his work, he says, “Theresults published will prompt RSA system designersto revise their designs. Since there is a larger rangeof d over which RSA can be broken, systems will haveto be designed with the new bounds in mind.”

A lattice, rendered in Sage.Image by the author

I mention to Dr. Sarkar that his work seemshighly theoretical. He’s quick to point out that it does

��| Immerse

involve some simulations – he runs attacks on RSAvariants using an open-source so�tware called Sage.This serves to validate his results. “My work, and infact all work since Boneh and Durfee’s paper, involvesome implicit mathematical assumptions that noone’s formally proved. I need to run actual attacks inorder to validate the bounds I derive.” But, he admitswith a rueful grin, “It can get tedious at times. Youjust have to keep trying the problem from di�ferentangles. I also like what I do.”

When I ask him how he decided to venture

into cryptography, he points to his alma mater, ISIKolkata, as his inspiration. “ISI is well known forcryptography. Once I started working in this field,I saw problems of this type, and they interested me.I still work with colleagues there, as well as withcollaborators in China.”

Dr. Sarkar is currently attempting to improvethe bounds described above even further. He’s alsoworking on problems related to another encryptionalgorithm, RC�, primarily employed in wirelessnetworks. ⌅

Meet the Author

Nithin Ramesan is a B.Tech. student of Electrical Engineering. He likes to quiz,

write and read Calvin & Hobbes. For bouquets or brickbats, he can be contacted at

nithinseyon@gmail.com.

��| Mathematics

Amarathon runner has persistent painin his knee that leaves him unableto walk. He sees an orthopaedicspecialist at a hospital and undergoes

a knee replacement surgery, which takes all of oneday. He can now perform any activity he did beforeand wins that marathon he was training for. A happyending - the story advertised by every bone andjoint specialty hospital. What they don’t talk about,though, is that it takes several weeks to design andmanufacture the implant that’s tailored to replacehis knee precisely. Moreover, these implants havean average lifetime of only about �� years, and willhave to be replaced by another surgery. That it takeshim several weeks of physiotherapy to regain his fullrange of motion and even then he shall experiencechronic pain is an issue that is conveniently ignored.

They envision a cost-e�fective instrumentintegrated with a diagnostic tool that

designs and manufactures the part to bereplaced.

Total Joint Replacement is seen as the biggestsuccess story of orthopaedic surgery. It has helpedhundreds of thousands of people regain or maintaintheir functional independence and live fuller, moreactive lives. However, the surgery is prohibitivelyexpensive, rendering it out of reach for an estimated�� million people who su�fer hip, knee and shoulderjoint failure every year. Dr. Soundarapandianand his team from the Department of MechanicalEngineering are determined to change this.

They envision a cost-e�fective instrumentintegrated with a diagnostic tool that designs andmanufactures the body part to be replaced. A patientwith joint failure walking into a hospital will gothrough the rigmarole of diagnosis to surgery ina few hours. As a first step towards this dream,they are working on eliminating the shortcomingsof existing bone implants by synthesising amagnesium implant with a calcium-phosphate (CaP)coating by a laser-based coating technique.

The high loads that orthopaedic implantsneed to support restrict the selection of feasiblematerials. Today, stainless steel, cobalt, chromiumand titanium alloys have been successfully usedto fabricate implants because of their strength,comparatively low sti�fness, light weight and relativeinertness. However, these implants release toxicmetallic ions. Moreover, analysis of metal ormetal alloy devices provides convincing evidencethat implant failure is because of a mismatch ofthe mechanical and the chemical properties ofthe implants with the bone at the bone-implantinterface. This mismatch leads to the formation of afibrous layer of tissue at the interface, giving rise tosmall gaps which cause movement at the interface.Ultimately, this causes a failure of the implant andrequires subsequent surgeries to replace the looseimplant. One approach to alleviate this problem hasbeen the use of CaP coatings applied to the implantsurface. This enables researchers to considermaterials with attractive properties that have earlierbeen rejected for their lack of biocompatibility.

The CaP mineral hydroxylapatite (HAP),Ca5(PO4)3OH , has attracted considerableattention because of its close resemblance tothe chemical and mineral components of teethand bone. As a result, HAP is biocompatiblewith bone. Instead of forming a fibrous tissuelayer at the implant-bone interface like normalbiomedical alloys, implants with HAP coating havebeen shown to form a thin layer bonding with thebone and even promoting bone growth. Plasmaspraying is the most popular and the only Food andDrug Administration (FDA) approved method forapplying CaP coatings to implant surfaces. Thisprocess involves the high-velocity spraying of moltenHAP powder onto an implant surface. Upon impactwith the substrate, the material rapidly cools andforms a dense coating with a morphology consistingof layers of HAP impact splats. Coatings synthesisedby this method form a dense, adherent layer of CaPon metal substrates.

��| Mechanical Engineering

Dr. Soundarapandian did his Ph.D. in Mechanical Engineering (2010) at Southern

Methodist University (SMU), Dallas, USA followed by Postdoctoral research at University

of North Texas (UNT), Denton, USA. Currently, he is an Assistant Professor of Mechan-

ical Engineering at IIT Madras. His research focuses on synthesis and characterisation

of structural and bio-materials, LASE, computational modelling, manufacturing automa-

tion, fabrication of next-gen bio-implants, and laser applications in medical industry.

While plasma spraying is a well-understoodprocess, the control of variables is quite complicated.The extremely high temperatures (10, 000oC to12, 000oC) used in the plasma spray process canvastly a�fect the properties of the final coatingand result in potentially serious problems such asthe coating of complex implant devices containinginternal cavities. More serious is the potential forthe formation of amorphous CaP phases with a Ca/Pratio between �.�� and �.� in the film rather thanstoichiometric HAP which has a Ca/P ratio of �.��.There is also concern over alteration of the coatingstructure. In addition, spraying plasma to coatwithin the pores of porous metal materials provesdi�ficult because it is a line-of-sight process.

A detailed schematic of the deposition process

During his PhD training, Dr. Soundarapandianidentified Magnesium (Mg) as a suitable alternativebecause its mechanical properties are closer to that

of natural bone. However, due to the corrosionof Mg in physiological environments, it cannot beused directly. The solution proposed by the teamis deploying HAP coatings on Mg implant surfacesusing a laser-guided manufacturing technique.This exploits the bio-compatible and bone-bondingproperties of the ceramic, while using the superiormechanical properties of Mg implants.

Additive manufacturing caught the professor’seye during his Master’s in mathematical modellingat Blekinge Institute of Technology, Sweden wherehe developed the modelling technique to predictthe right manufacturing process given the requiredgeometry and material. However, typically, additivemanufacturing isn’t intended to accommodatematerials with dissimilar properties. Bones are acomposite of both organic and inorganic materials.This implies that exactly mimicking a bone wouldrequire materials with disparate properties andexisting additive manufacturing techniques werethus inappropriate for the task at hand. Further,bones have a porous geometry which must alsobe mimicked by the implant in addition to beingcompatible with the bone environment, a propertycalled osteo-integration.

This exploits the bio-compatible andbone-bonding properties of the ceramic,

while using the superior mechanicalproperties of Mg implants.

To address all these concerns, the researchgroup has developed a new instrument that canaccommodate metals, ceramics and polymers.The novel technique involves a commonly used�D printing method called Fused Deposition

��| Immerse

The different components of the 3D printing lab.

Modelling. The process basically involves a hotair gun controlled by a robot arm that zig-zagsback and forth depositing layers of powderedHAP mixed with a polymer mixture that promotesbinding to the surface of the metal. This surfaceis air-dried and then bombarded with a laser at apre-characterised energy density that minimisesthe corrosion rate for a given combination ofmaterials. Lasers are very precise and powerful andthe process is a non-contact process and hence idealfor bio-implants.

The next step in the process of designing anovel implant is a set of rigorous tests. Thefirst step to ensure bio-compatibility is a set ofin vitro tests. In this step, you immerse theimplant in artificially developed bio-fluids thatmimic the physiological conditions and study themover several days. The implants manufacturedby the research group passed this stage and theynoticed an interesting e�fect. Not only was itbio-compatible, meaning it wasn’t harmful to thebody, but it also turned out to promote bonegrowth. This led to another set of tests to studycell behaviour, particularly adhesion to the implant.

The experiments performed in association with acollaborator showed increased adhesion of bonecells (osteoblasts) to the implant surface. Theysubsequently worked on tweaking several factorssuch as surface chemistry and topology to enhanceadhesion. Surface chemistry was altered by usingseveral polymers proven to increase cell adhesion.The surface roughness was also altered to studye�fects on cell adhesion.

Not only was it bio-compatible, meaningit wasn’t harmful to the body, but it also

turned out to promote bone growth.

Currently, the team is busy ensuring that themethod, the instrument and the process can beused for wildly di�ferent materials including severalbodily derived materials. One such bodily derivedmaterial being considered is Fibrin, a fibrous proteininvolved in blood clotting. It’s a tough, resilientmaterial that has properties very similar to thatof cartilage. They have already extracted fibrinand are currently working on fabricating implantsfrom it. As with any biomedical device, ensuring

��| Mechanical Engineering

reproducibility continues to be a major challengeand they’re working on verifying and validating theirmethods for the same. The next step will be touse these implants in animal models for in vivostudies. Dr. Soundarapandian is currently looking

for collaborators to carry out some of the biologicaltests. He says that the next stage will take � to �years a�ter which he would be allowed preliminaryhuman trials in his long haul to see his dream cometo fruition. ⌅

All Images are courtesy of Dr. Soundarapandian.

Meet the Author

Aparnna Suresh is a final year student of Biotechnology at IIT Madras. While not holed

up in the lab dreaming about creating a Jurassic Park, she enjoys quizzing, reading and

swimming. Her long term goal is to pursue a career in the academia and her research

interests include synthetic and systems biology, and biological computing. She can be

reached at aparnnaa.suresh@gmail.com.

��| Immerse

� millimetres is just the right size to dosomething big, claims Apple. Theirunusually sleek iMac which is just � mmthick, is as awe-inspiring to a material

scientist as to a gadget geek, because joining theultra-thin monitor panels is a challenging materialengineering problem. Which is why, Apple overruledconventional metal welding processes and used arelatively new approach called friction-stir welding,creating a product enclosure that was too thin andseamless to take apart.

Since ancient times, humans haveknown that thermal and mechanical

processes can be used to morph materialsto our needs

Friction-stir welding is a process where pressureand friction-generated heat are used to joinmaterials. Since ancient times, humans have knownthat thermal and mechanical processes can be usedto morph materials to our needs – liquid watercan be solidified on cooling, carbon can turn intodiamond at high pressures and temperatures, andsheets of metals can be joined when their edgesare melted by hot flame. Friction-based processesare similar, albeit they use friction to generate heatunder intense pressure (imagine a crushing ���kilos weight supported on an area equivalent to twofinger tips). The study of these processes is one ofthe main focuses of the Materials Joining groupat IITM’s Metallurgical and Materials Engineeringdepartment. The group uses friction to achieve avariety of things – from improving the strengthof metals, to coating their surfaces and welding or

compositing very di�ferent materials.

Prof. Ranjit Bauri, one of the faculty membersin the group, describes his work as ‘surfaceengineering’. He uses a method called friction-stirprocessing to enhance hardness and wear resistanceof metal surfaces. The apparatus is as basic as shownin the image below:

Schematic of friction stir processing setup.Courtesy: Dr. Ranjit Bauri

The vertical tool, which is under intensedownward pressure, rotates and translates alongthe metal substrate. This produces frictional heatand local mixing at the interface, causing what’scalled a plastic deformation of the material. Plasticdeformation is almost like flow in solid state. Flowin solid state may seem counter-intuitive but thisphenomenon of plasticity is universal in our dailylife. Whenever we pound a bar of iron, the bar getspermanently deformed, literally ‘flowing’ into itsnew shape without melting. Or even when we ironour clothes which are typically made of polymers, thecreases flow out.

Materials Joining Laboratory at IIT Madras comprises of Prof. GandhamPhanikumar, Dr. GD Janaki Ram, Dr. Ranjit Bauri, and Dr. Srinavasa Rao Bak-shi from the Department of Metallurgical and Material Science Engineering. Theirresearch interests span surface engineering, microstructure analysis, additive man-ufacturing, welding and welding simulation, study of composites and alloys. Housedin a large Central Workshop bay, they use a wide array of testing and analytical toolsto investigate and improve material behaviour.

��| Immerse

Metals, which are the consideration of thestudy here, get such plasticity from their grainyinternal microstructure. One way to understandmicrostructure is to look at metals as not onebig solid slab, but as made of many microscopicinterlocking polygons. Each of these polygons is a‘grain’ and shares boundary with other neighbouringgrains in three dimensions. Sizes and boundariesof these grains strongly a�fect almost all industriallyuseful properties like strength, ductility, hardness,corrosion resistance and wear-resistance of thematerial. Hence, understanding and improvingmicrostructure of metals has been the holy grail ofthe metallurgical sciences.

Grain structure of aluminium obtained using.Electron Backscattering Diffraction,

Courtesy: Dr. Ranjit Bauri

Friction-stir processing (FSP) is one such processthat helps refine the grain size, says Prof. Ranjit.When this method was discovered in late nineties,the group here was the first to apply the processto make surface composites. Composites are animmensely useful class of materials as they give usnew properties – say, a mix of strength of one metaland low weight of the other. Aluminium (Al) and

nickel (Ni) fit this bill, as nickel gives higher hardnessto a widely used, low density metal like aluminium.

Now, the easiest and crudest way would beto melt the metals and mix them. But meltingthem to make an alloy results in formation ofbrittle intermetallic compounds like Al3Ni. Theseunwanted compounds arise because energy suppliedto melt the metals also makes chemical reactionsbetween them thermodynamically feasible. One wayto overcome this trouble, as this group discovered,is to make the composite using a solid-state processlike friction stirring. The second metal can beintroduced into the ‘stir zone’ in a variety of waysand embedded into the other metal’s surface by themovement of the vertical rod tool.

Composites are an immensely usefulclass of materials as they give us new

properties – say, a mix of strength of onemetal and low weight of the other.

The end product in case of aluminium and nickelis a metal-metal composite that’s three times harderthan aluminium on the surface. This means it canresist wear more e�fectively. “The beauty of theprocess though”, as Prof. Ranjit puts it, “is that itdoesn’t decrease the ductility, which is the abilityof aluminium to be drawn into wires, too much;we are able to retain �� percent of aluminium’sductility.” This is a big deal because there iso�ten a trade-o�f between strength and ductilityin conventional strengthening processes. Prof.Janaki Ram, another faculty member in the group,has achieved similar results with metal-ceramiccomposites despite ceramics being a completelydi�ferent class of materials from metals.

Making composites though is just oneof the multitude uses of friction-relatedprocesses. Prof. Janaki Ram is also keen onapplications of friction-related processes in additivemanufacturing. Additive manufacturing or �Dprinting is a computer operated layer-by-layermanufacturing of an object. This is unlike in a

��| Metallurgical and Materials Engineering

normal setting, where di�ferent parts of an objectare casted first and then welded together. In ajournal paper in ����, this group was the firstto propose that friction surfacing, a process verysimilar to friction-stir processing, could be usedfor layer-by-layer manufacturing of metal objects.The only di�ference, in fact, between friction stirprocessing and friction surfacing is that the rod toolused in the latter is consumable. While traversingthe substrate, the rod tool itself significantlyso�tens at the interface because of high temperatureand pressure. This leads to establishment ofmetallurgical bonds between metal atoms of the rodand the substrate, causing some material to come o�fthe rod and deposit on the surface of the substrateas one layer in a step-by-step layer addition process.

Six-layer cylindrical deposit consisting of a fullyenclosed internal cavity and its X-ray radiograph,

Courtesy: Dr. GD Janaki Ram

Friction surfacing’s biggest advantage is in thefact that it is a solid state process. This meansthat it is suitable for use with dissimilar materials,which would say, be incompatible with each otherin melt state. There are many ideas-in-waiting forproducts using dissimilar materials – like a turbinewith the input end made of a material optimised forheat resistance, and the output end optimised forstrength. Or a bottle with a magnetic bottom to beheld in place by a magnetic holder.

Close-up of Friction Stir Welding.Image source: TWI, via Flickr

The possibilities are unlimited.

One of the more established uses of friction isthe friction-stir welding process used in buildingproducts like Apple’s iMac, NASA’s rovers or moretraditionally, aerospace components produced bythe likes of Boeing and Airbus. Invented in ����sby The Welding Institute in UK, friction-stir weldingwas competitively patented until recently, closing o�favenues for external research, says Prof. GandhamPhanikumar, another faculty member in this group.Once the patent expired, research opened up andtechniques like friction processing and surfacingwere proposed as a modifications of the originalwelding process. While these techniques are yetto be undertaken on a large commercial scale inIndia, organisations like Naval Research Board arelooking at possibilities of using friction welding andsurfacing for in situ repairs or application of coatingsfor marine vehicles.

...this group was the first to propose thatfriction surfacing ...could be used for

layer-by-layer manufacturing of metalobjects.

Reading through the PhD thesis of H KhalidRafi, who worked in this group and graduated in����, one gets the idea of immense potential offriction-based processes to serve as an alternativeto the conventional techniques. In fact, one of hispapers on friction welding aluminium alloys has

��| Immerse

already been cited over fi�ty-five times in five years.While there is a still long way to go, the group hopes

to continue drawing more insights into friction andits applications in material processing.⌅

Meet the Author

Raghavi Kodati is a senior undergraduate student in the Chemical Engineering de-partment, whose research interests are in microfluidics and materials. While work-ing on this article, she got fascinated by the history of material joining processes –from their use in iron pillars in ancient India to today’s aluminium-lithium SpaceXrockets. Excited about science writing, she has written for three issues of Immerse.

��| Metallurgical and Materials Engineering

Water – one of the basicnecessities for life, holdssecrets that never cease toastonish researchers. Its

liquid form is denser than the solid form. It expandsboth when heated to 4oC and cooled to 4oC. It isa universal solvent, dissolving a large variety ofsubstances. In addition to all peculiar qualities ofwater, researchers have now understood that waterat freezing temperatures and high pressures canstore certain gases too. Dr. Jitendra Sangwai, aProfessor at the Department of Ocean Engineering,IIT Madras has done research that reveals thisunique aspect of water.

Water at very high pressures (�� to �� bar) andchilling temperatures (0oC to 10oC) turns into a newform of ice-like crystalline structure. This crystallinestructure has small gaps in it making room for gasstorage. This crystalline structure together with gasis known as a clathrate hydrate or a gas hydrate.A gas hydrate then, is just a cage made of water.Most low molecular weight gases such as hydrogen,oxygen, nitrogen, methane (natural gas), carbondioxide, and hydrogen sulphide can be trappedinside the cage, but each gas needs di�ferent pressureand temperature conditions for this to happen. Eachhydrate derives its name from the gas present in thecage. So if methane is present in the cage, thenit’s called methane hydrate. With this interestingproperty of water, harmful gases can be stored oruseful gases present in the existing hydrates can beextracted.

The property of water transforming into cagesand trapping gases was discovered by an Englishchemist, Joseph Priestley in the ��th century andby another English chemist, Humphry Davy inthe ��th century independently. But attentionwas drawn towards it in the ��th century whenEG Hammerschmidt, an engineer working in aTexas-based natural gas company found that thesehydrates were blocking natural gas pipelines inwinters.

Clathrate hydrates were looked upon as obstaclesto the flow of natural gas in pipelines until aRussian petroleum engineer turned professor, Yuri FMakogon discovered in ���� that clathrate hydratescan be used as source of energy by extracting thenatural gas that has accumulated inside them. By����, he had estimated the amount of natural gas inhydrates present worldwide and paved the way forpresent researchers to find ways to extract it andunderstand the behavior of hydrates.

Molecular structure of gas hydrates. Courtesy: MIDAS(Managing Impacts of Deep Sea Resource Exploitaiton)

The oceans are a conducive environment for theformation of methane gas hydrates. At the depthswhere hydrates are found, pressure is very highdue to the sheer height of water above, and thetemperature is low as the sun’s rays can’t penetratesuch depths. Vast reservoirs of methane hydratesare found in marine sediments, at depths greaterthan ��� meters, close to continental margins andin onshore permafrost – soil, rock or sedimentthat is frozen for more than two consecutive years.Availability close to continental margins meansreduced extraction and production costs, withouthaving to spend on deepwater drilling which is acostly and risky a�fair. This is a boon for India,since it has a very long continental margin. In Indiaalone, the natural gas present in methane hydratesis estimated to be about ���� trillion cubic meters -���� times the currently known natural gas reservesfrom other sources.

��| Ocean Engineering

Dr. Jitendra Sangwai is an Associate Professor in the Department of Ocean En-gineering at IIT Madras. He received his PhD in Chemical Engineering from IITKanpur and is the founder of Gas Hydrate and Flow Assurance Laboratory at IITMadras. He holds eight patents in the field of gas hydrate, enhanced oil recoveryand flow assurance. His research interest lies mainly in the field of gas hydrates,enhanced oil recovery, rheology of drilling fluids, flow assurance, and polymer andnanotechnology applications for upstream oil and gas engineering.

“Natural gas hydrates o�fer a realistic solutioncompared to other polluting fossil fuels,” saysProf. Sangwai. Methane, which is present in thecage-like structures of methane hydrates, can beexchanged with the greenhouse gas CO2 producedby burning the methane. This is kind of a zero carbonenergy scheme. This helps in CO2 sequestration –capturing CO2 and burying it back in the earth aspart of the hydrate. This method of fixing CO2 hastwo advantages. One is cleaning up theCO2 that hasbeen emitted and the other is that it is unlikely thatthe CO2 stored in the form of hydrates will comeback to the Earth’s surface.

A High Pressure reactor used for gas hydratestudies at Gas Hydrates Flow and Assurance Lab

While drilling, CO2 is injected into the well forrecovering the oil trapped in the tiny pores of rocks.The injected CO2 pumps out the trapped oil and, inthe process, can also be consigned to the cage of thehydrates. This idea led researchers to the concept

of flue gas separation. The drilling environmentis a dirty place, emanating many dangerous fluegases like oxides of carbon, nitrogen and sulphur.Gas hydrates can save us here. If the pressure andtemperature at which flue gases form hydrates areknown, then by sending in water at that precisetemperature and pressure, one can trap the gases inthe cage-like structures of the hydrates.

In India alone, the natural gas presentin methane hydrates is estimated to beabout ���� trillion cubic meters - ����times the currently known natural gas

reserves from other sources.

Gas hydrates o�fer an alternative to the highexpenses of transporting and storing LiquefiedNatural Gas (LNG). LNG infrastructure o�ten addsto the cost of importing as it needs a special typeof floating tanker and heavy refrigeration facilitieswhereas gas hydrates need very minimal storagespace – � cubic meter of methane hydrate can store��� cubic meters of methane. Transporting hydratesis quite simple – it can be done using existingpipelines, with the hydrates in the form of slurries.

Another potential application of gas hydrates isemploying them in desalination. Here, seawater istaken in a chamber and CO2 gas is passed it atvery low temperature and high pressure resultingin the formation of CO2 hydrates. By taking theformed CO2 hydrates into another chamber anddissociating them by increasing the temperatureand decreasing the pressure, pure water that is freefrom salts can be obtained. This happens becausesalts cannot form hydrates. Desalination using

��| Immerse

Dr. Jitendra Sangwai and his research team at IIT Madras

this technique is considerably cheaper than otherconventional techniques.

The problem is that little is known aboutthe stability of the structure of the cages whichform at di�ferent combinations of pressures andtemperatures. The environment where hydrates arefound is very harsh. It is not possible to obtain ���%pure methane hydrates. Seawater already containsmany dissolved salts that inhibit the formation ofthe gas hydrates by taking away the required water.E�fects of these salts on gas-hydrate formation arenot completely understood. Further, the impactof di�ferent types of porous medium in the oceanlike silica gel, silica sand, activated carbon on theformation of hydrates are to be studied.

Apart from understanding what promotes theformation of hydrates, the study of substances whichdissociate these hydrates are quite important. Gashydrates are notorious for flow path blockage ofpipelines in the oil and gas industry, which is howthey were initially noticed. “Both promoting anddisassociating of hydrates are to be mastered,” says

Dr. Sangwai.

Dr. Sangwai and his team have been workingwith a variety of additives to improve the stabilityof gas hydrates at di�ferent temperatures andpressures. Rather than conducting experimentspiecewise, which is expensive and time-consuming,they are developing a model to predict theirbehavior.

Dr. Sangwai’s aim now is to developnew kind of additives that will reducethe cost of extracting methane from

methane hydrates.

Additives are of two types. They can eitherpromote or inhibit the gas hydrates formation.Inhibitors help in freeing the pipelines, andpromoters help in trapping gases in their cages.Additives can be classified on how they workwith hydrates. Certain additives that tweakthe temperature and pressure conditions atwhich hydrates form or dissociate are known as

��| Ocean Engineering

thermodynamic additives. Other additives that donot a�fect the temperature and pressure conditionsbut still a�fect the formation of these hydrates andare known as kinematic additives. New hybridadditives are emerging which serve the purpose ofboth thermodynamic and kinematic additives.

Safety is paramount while extracting methanehydrates. During the drilling operation of hydrates,if huge amounts of methane gas are releasedsuddenly from the drilling site, an explosion underwater is highly probable. To prevent this, a cheaper -highly volatile methanol is used as drilling fluid. Thisacts as a thermodynamic inhibitor. Prof. Sangwaiand his team are looking for options to replace thehighly volatile methanol with polyethylene glycol.

If gas hydrates have so many advantages, thenwhy aren’t we seeing natural gas extracted frommethane hydrates in the mainstream industry?The main problem is that the transportation oflarge amounts of water from the recovery site tothe extraction site is a costly a�fair. Moreover,

formation of gas hydrates is a very slow processwhich forms a bottleneck in the supply chain.The extra transportation costs coupled with thekinetically slow process has proven to be one ofthe deterrents in adoption of this technique bythe industry, which is why LNG still dictates thenatural gas industry. Dr. Sangwai states cheerfully,“This is where we come into the picture.” Gashydrates can be made competitive by minimising thetransportation between recovery site and extractionsite. Dr. Sangwai’s aim now is to develop new kindof additives that will reduce the cost of extractingmethane from methane hydrates.

When asked about how he decided to work inthe field of gas hydrates, Dr. Sangwai says “Righta�ter my PhD, I wanted to work in the area whichwill be the future of oil and gas industry. I believegas hydrates will be our future energy sources.” Withcommercial production already started in Japan,though at a slow pace and small scale, we can expectthe production of natural gas from gas hydrates soonto begin in India. ⌅

Meet the Author

Nikhil Mulinti is a final year Dual Degree (B.Tech. - M.Tech.) student in the De-partment of Ocean Engineering at IIT Madras. He is fond of science and his fasci-nations range from the cosmology to anthropology. He is currently working on thebubbly flow technology, a trending research area in marine hydrodynamics. For anycomments or criticism, the easiest way is to drop a mail on nikhil.mulinti@gmail.com.

��| Immerse

Suppose one routine day you wake upto find another living being who is anidentical copy of your mirror image.That is, this mysterious creature

behaves and moves exactly the way you do but forhis body being a laterally inverted version of yours.However, if you both bump into each other, you getkilled instantaneously! Bizarre, isn’t it?

Courtesy: CERN

While we currently know nothing that evenremotely hints at the existence of such alien beings,we are aware of a similar phenomenon in the realmof particles, owing to the Nobel Prize winninginferences of the renowned physicist Paul Diracand several other studies that were founded uponhis seminal work. Interestingly, Dirac combinedEinstein’s special relativity and the theory ofQuantum Physics into an equation that yielded twosolutions – one associated with positive energy andanother with ‘negative energy’. He conjecturedthat for every class of charged particles, thereexists a class of ‘antiparticles’ – particles with thesame mass but opposite charge. For example, theantiparticles corresponding to electrons (a class

of matter particles) are particles called positrons.Equivalently, for every class of electrically neutralmatter particles, there exists a class of electricallyneutral antiparticles having the same mass. Thus,for every entity of matter we are familiar with, thereexists a corresponding ‘antimatter’ entity. Severalexperiments that ensued from Dirac’s postulatesproved the existence of such antiparticles. It wasalso learnt that when a particle collides with itsantiparticle, both of them get annihilated andrelease energy in the process, regardless of whetherthey are charged or neutral. Thanks to yearsof toil of physicists all over the world, today wecan understand and appreciate antimatter and itsrelationship with matter much better.

But there is one thing we still do notcomprehend: If matter and antimatter are exactlyequal and opposite, why does the universe containmuch more matter than antimatter? Manyexplanations have been proposed so far, but noneof them is fully convincing.

If matter and antimatter are exactlyequal and opposite, why does the

universe contain much more matterthan antimatter?

Nevertheless, we do know that certain conditionscalled Sakharov conditions must be satisfied for thereto be an imbalance between matter and antimatter.One such fascinating condition is the existence ofa phenomenon called Charge Parity Violation (CPViolation) during the first few seconds following theBig Bang. It turns out that the most enduring theoryof particle physics, known as ‘The Standard Model’,provides some explanation for CP Violation.

Dr. Jim Libby is an experimental particle physicist working at IITM since 2009. Dr. Libby

received his undergraduate and postgraduate degrees from the University of Oxford. His

PhD work was with the DELPHI experiment at the Large Electron-Positron Collider at

CERN. Since completing his PhD in 1999, he has worked with accelerator experiments

at CERN, Stanford, Cornell and KEK (Japan). He also participates in studies related to

the Indian-based Neutrino Observatory (INO).

��| Immerse

However, whether the Standard Model describesCP Violation correctly or not is not known withcertainty. This is currently an area of active researchand Dr. Libby has been engrossed in it for the pastseven years at IIT Madras. Before we move on to Dr.Libby’s specific interests within this area, let us seewhat the Standard Model itself has to o�fer.

Developed throughout the latter half of the ��th

century, this exhaustive theory seeks to explain thecharacteristics of the vast multitude of subatomicparticles and their complex interactions with thehelp of only three kinds of particles – six quarks, sixleptons and force carrier particles.

Courtesy: DESY

Quarks are constituents of the familiar protonsand neutrons that make up most of the matterthat we see around us. They do not have anindependent existence; they exist only with otherquarks in composite particles (particles composedof other particles). Scientists are currently aware ofthree pairs of quarks: the up-down, the charm-strangeand the top-bottom pairs. Among these, up, charmand top quarks carry positive electric charge whereasthe other three carry negative electric charge. On alighter note, the characteristics of these quarks areas weird as their names.

Leptons are solitary matter particles and thushave an independent existence. The six kinds ofleptons are electrons, muons, tau particles and threekinds of neutrinos. All leptons except neutrinos carry

electric charge.

The third category of particles in the StandardModel is that of force carrier particles which giverise to three fundamental forces: the strong force,the weak force and the electromagnetic force. Thestrong force holds quarks together, the weak forcecauses the decay of massive quarks and leptons intoheavier quarks and leptons, and the electromagneticforce causes electrically charged particles to repel orattract each other.

Every conservation law is associatedwith a symmetry inherent in nature.

These forces in turn give rise to three kindsof interactions: the strong, the weak and theelectromagnetic interactions. These interactionsinclude particle decays and annihilations andcan be represented pretty much like chemicalreactions characterised by reactants and products.Furthermore, their most vital properties aredescribed by quantum mechanics. Two suchproperties are the familiar electric charge (denotedby C) and parity (denoted by P ).

Before asking what ‘parity’ means, it would beworth recalling two laws of Physics that you studiedin high school: the law of conservation of linearmomentum and the law of conservation of angularmomentum. If you wrack your brain long enough,and if you have the genius of the great Germanmathematician Emmy Noether, you might gain oneof the most precious insights into physics: Everyconservation law is associated with a symmetryinherent in nature.

For example, linear momentum is conservedbecause of spatial symmetry – there is no particularlocation in space that is preferred to other locations.In the case of angular momentum, there is nopreferred direction in space. Likewise, the initialassumption that nature was unbiased and thustreated matter and antimatter identically, or thatthere was a symmetry between antimatter andmatter, reinforced the classical hypothesis that the

��| Physics

total parity of a system is always conserved in aparticle interaction. In essence, this means that aninteraction and its mirror image can be representedby the same particle equation.

However, as is the fate of most scientific theoriesthat for a long time enjoy an unquestionablepresence as truths even in the most critical of minds,the theory of parity conservation was proved tobe incorrect. Physicists learnt through reluctantlyperformed experiments that parity is not conservedin weak interactions.

Usually, when a promising hypothesis is refutedby experimental evidence, scientists understandablytry to modify it or extend it to a more general caserather than discarding it altogether. That is exactlywhat happened in this case too – physicists tried tofind another quantity Q such that the combinationQP of this quantity and parity would remainsymmetric even in weak interactions. The renownedphysicist Lev Landau proposed that in this case Qis nothing but C, the charge. Parity conservationthus came to be replaced by CP Symmetry. Thismeant that a process in which all the particles areexchanged with their antiparticles was assumed tobe equivalent to its mirror image.

How should the Standard Modelaccount for CP Violation?

Although CP Symmetry succeeded to an extentin explaining weak decays, history was destined torepeat – this extended notion of symmetry too wasfound to be violated in decays of particles called neu-tral kaons. It was observed that even the combinationof charge-parity was not conserved in these decayswhich again happened to be weak interactions. Thisphenomenon came to be known as CP Violation. Tocomprehend this more clearly, consider the decayB� ! DK�. We now perform a CP operationon this decay, i.e,. we laterally invert the decay inthe �-dimensional space (or convert it into its mirrorimage) and then invert the signs of the charges of theparticles involved. We thus arrive at the decayB+ !

DK+, which is a charge-conjugated version of themirror image of the original decay. Experimentshave shown that the rates at which these two decaysoccur di�fer by a remarkable amount. In other words,the total amount of CP on the reactant’s side of thecombination of the two decays B± ! DK± doesn’tequal its total amount on the products’ side. Thisis how charge-parity conservation (or CP symmetry)is violated in this decay. The overall significance ofthis symmetry violation for particle physics can begleaned from the fact that its discoverers, Croninand Fitch, were awarded the Nobel Prize in ����.

At this juncture, the following question arises:How should the Standard Model account for CPViolation? The answer lies in the properties of theCabibbo-Kobayashi-Maskawa matrix (CKM matrix), asquare array of numbers that is central to theStandard Model. In order to get an idea about thismatrix, we must first note that there are weak decaysin which negatively charged quarks (bottom, downand strange) get converted into positively chargedones (top, up and charm). This gives rise to � possibledecays. The CKM matrix, having � rows and �columns, contains information on the strengths ofeach of these decays. CP Violation is incorporatedinto the Standard Model by allowing this matrix tohave complex number entries.

However, there is a crucial constraint: for theCKM matrix to make physical sense, the StandardModel requires it to have a mathematical propertycalled unitarity. This property can be expressed as aset of equations which the matrix must satisfy.

“These equations involve �� variables!”, saysPrashanth, a student of Dr. Libby, as he laughs atthe sheer complexity of the whole thing. Fortunately,the unravelling of a few relationships between thevariables reduced their number by an alarmingdi�ference - from eighteen to just four! These �parameters are comprised of � Euler angles and �phase variable. The unitarity of the CKM matrix nowreduces to fewer constraints, of which one states thatthe Euler angles should be the angles of a triangle,i.e., they must sum up to ��� degrees. The triangle

��| Immerse

formed by them is called the unitarity triangle. Ascan be seen from the figure below, its angles aredenoted by↵,� and�. Although the phase variable isthe one responsible for CP Violation, the area of theunitarity triangle indicates the degree of violation.This triangle is unique in terms of its lengths andangles. It therefore has the potential to indicatehow accurately the Standard Model describes thissymmetry violation.

The Unitarity Triangle (An Illustrative Image)

Many physicists thus shi�ted their focus to thedetermination of the values of ↵, � and �, theEuler angles or the angles of the unitarity triangle.They have been able to determine ↵ and � witha reasonable accuracy by studying the interferencebetween various decays involving particles calledmesons, but � remains elusive. You may naturallyask, “Why can’t they determine � solely from theirknowledge of the other two angles simply becausethe three angles form a triangle?” Note that sucha method of determining � would rest on theunitarity assumption, the constraint on the CKMmatrix that in fact gave rise to the three Eulerangles. In order to test this assumption, it isnecessary to determine � from other measurements.Moreover, “the current world average precision on �

is significantly worse than that of the other anglesof the unitarity triangle”, says a paper recentlyauthored by Dr. Libby. So the next question is: howmust one go about enhancing this precision?

This is where Dr. Libby’s research enters thepicture. His aim so far has been to improve ourknowledge of � by studying how certain observablequantities violate CP symmetry in B± ! DK±

decays. These quantities are called CP-violating ob-servables. Dr. Libby’s focus has been on a special

class of these observables – observables associatedwith particle states called CP eigenstates. Thesestates are the products of certain kinds of B± !DK± decays. Two kinds of CP eigenstates havebeen of greater interest: CP-even eigenstates and CP-odd eigenstates. Another way of saying that a CPeigenstate is CP-even is: ⌘ = +1. Similarly, ⌘ = �1

implies that the state is CP-odd.

However, this does not mean that no e�fortswere made in this direction previously in order toknow more about the Euler angle. Four kinds ofB± ! DK± decays had already been studied.Essentially, each of these is a decay of a D meson to aunique CP eigenstate. Furthermore, in each decay,only a fraction of D mesons decays to the desiredeigenstate. This fraction is known as the branchingratio (BR) of the decay. With this background inmind, we can represent the end results of these fourdecays through the table below:

State ⌘ Branching Ratio (%)

⇡+⇡� +1 �.��K+K� +1 �.��K

S

⇡0 �1 �.��K

S

⌘ �1 �.��

As can be seen from this table, only a very minutefraction of the D mesons decays to a given CPeigenstate. Measurements on these states are thuslimited because not many samples are available forstudy. So the challenge facing Dr. Libby and histeam of students was to identify more easily availableeigenstates to which these mesons decay. And theydid. They used the following critical observationmade in a separate study: the branching ratio forthe decay D0 ! ⇡+⇡�⇡0 is 1.43%, a fractionsignificantly greater than the ones listed in the tableabove. This implied that the state ⇡+⇡�⇡0 wascertainly more useful than the four states studiedearlier. But it was not known whether it was aCP-even or a CP-odd eigenstate. As it turns out, theanswer to this question facilitated a more precisedetermination of �.

The data pertaining to the CP observables

��| Physics

associated with the decays described above as wellas several other decays has been collected by agiant particle detector called CLEO-c. “CLEO is thegranddaddy of flavour physics, with a history ofachievement dating back over �� years”, says Dr.Libby. This machine of monumental importancewas designed way back in ���� to collide electronswith their antiparticles called positrons at an energyof approximately �� GeV, an amount su�ficientto accelerate ��,���,���,��� electrons througha potential di�ference of one volt. Of particularinterest is the Cornell Electron-positron StorageRing (CESR), the part of CLEO where these collisionstake place. It has a circumference of ��� metres andis located �� metres below the ground level! Sinceits initial construction, CLEO has been upgradedseveral times for various purposes. Its final version,CLEO-c, has been tailored to the study of charmquarks such as D mesons. CLEO-c has so farcollected � million pairs of D mesons.

Dr. Libby and his team thus set out to analyse thedata gathered by CESR that contains informationon the CP-content of the previously mentioned Ddecay. You may have correctly guessed that thedata concerned was extremely vast – so vast thateven a�ter the physicists concerned analysed all thedata using the methods they had planned to employ,they felt compelled to use an altogether di�ferentclass of methods just to validate their analysis.This class of methods, known as the Monte Carlomethods, involves generating random numbers andperforming repeated simulations on the acquireddata using these random numbers as inputs for thesimulations.

“CLEO is the granddaddy of flavourphysics, with history of achievement

dating back over �� years. ”

CESR Quadrapole Magnet.Courtesy: CLASSE, Cornell University

��| Immerse

The next part of the analysis was to determinewhether the state ⇡+⇡�⇡0 was CP-even or CP-odd.Results indicated that the state was in fact inbetween these two extremes; a quantity called CPfraction (denoted by F+) that was determined forD ! ⇡+⇡�⇡0 revealed that it was almost a pureCP-even eigenstate (with the “purity” being close to��.�%). This key observation was examined furtherin order to understand its various implications. Asfar as � was concerned, Dr. Libby and his teamshowed that further investigation of the decay modeD ! ⇡+⇡�⇡0 could enhance the precision on �.What is more, they were also able to propose an exactanalytical method for the determination of this Eulerangle. It is just a matter of time before this methodis implemented in the near future. So it is safe to say

that in the worldwide e�forts to unravel the mostintriguing mysteries of matter and antimatter, Dr.Libby and his team, and hence IIT Madras, havetaken a step forward.

However, the story is not over yet. “Theformalism needs to be adjusted to incorporate F+

to account for the small CP-odd component in thefinal state”, as per a recent paper of Dr. Libby’s.Another massive particle collider named BeijingSpectrometer III could supply the data necessaryfor this purpose. Analysing this data would thencontribute to our knowledge of � and push us closerto answering the larger questions involving matterand antimatter. Undoubtedly, this means that thereis a long way to go and that innumerable excitingdiscoveries are in store for us. ⌅

Meet the Author

Rohit Parasnis is a final year Dual Degree student pursuing his B.Tech. in Electrical

Engineering and M.Tech. in Biomedical Engineering at IIT Madras. One of his long-term

goals is to make science more interesting and more accessible to all. Some of his past

endeavours include generation of video content for familiarising school students with

experimental science, translation of scientific and social scientific Wikipedia articles into

regional languages and performing a managerial role for the National Service Scheme

(NSS) at IIT Madras. He can be reached at rohityparasnis@gmail.com.

���| Physics

Could you briefly tell us about the history of NPTEL?NPTEL was started in ���� with the initial aim ofcreating course material for college students andmaking it available to everyone on the internet. Thefirst phase of the project extended from ���� to ����,when we created ��� archived courses; ��� videocourses and ��� textual courses. The second phase,which is almost nearing its end now, had a targetof ��� courses, but we were successful at raisingit to nearly ���. The Ministry of Education hasnow sanctioned the third phase, which focuses onMassive Open Online Courses, or MOOCs.

The online courses and the subsequentE-Certification programme is di�ferent from thearchived course material on NPTEL in a way thatthey follow the pace of a regular classroom insteadof the enrolee’s pace.

The long term aim is to increase thecontribution of other institutes towards

these courses and to provide a widerrange of courses.

Regarding the MOOCs, when did they start and howhas it progressed since then?We started with � course in March ����, � inSeptember and by the time we reached January ����we had started �� courses. Then in July ���� waswhere we saw the real expansion in courses, westarted �� courses by then. In January ����, wewill reach �� and by the time we enter March, weshould be at �� courses. We are expecting to reacha steady state where a certain set of courses will beo�fered in the odd semester and a certain set in theeven semester. The long term aim is to increasethe contribution of other institutes towards thesecourses and to provide a wider range of courses.

How does the e-certification programme via onlinecourses work?Mainly, we o�fer � formats of online courses; ��hours courses of � weeks duration, �� hours coursesof � weeks duration and ��-�� hour courses, which

Prof. Andrew Thangaraj, Coordinator,NPTEL

extend over nearly a semester. All the courses beingo�fered in a phase will have the same starting dateand students who are interested will have to enrollfor it beforehand. Every week, parts of the coursecontent and related assignments will be uploadedon the portal. The assignments will have a duedate and the students have to submit them to getthem evaluated, which makes up a part of their finalscores. There is also a discussion forum for everycourse, where students can get their doubts clarifiedby the teaching assistants in charge.

At the end of the course, the student will have toattend an o�fline proctored exam in their designatedcentres. Currently, we have at least one regularexamination centre in all the states. If there isa substantial number of students from a regionwhich does not have a centre, new centres are setup for their convenience. Overall, we have nearly�� examination centres all over India. The finalcertificates will be prepared with marks scored in theassignments and the final exam.

What is the role played by the Teaching Assistants(TA) at NPTEL?Students, or rather teaching assistants play a veryvital role in the smooth functioning of our courses.Many MS, PhD and even final year B. Tech. and DualDegree students are a part of this programme as TAs.Depending on the length and degree of di�ficulty ofthe course we have the number of TAs ranging from

���| NPTEL

The work that goes on behind the scenes for NPTEL.

� for certain courses to even �� for some.

They learn about the portal and upload materialon it. They are a part of the discussion forumto answer the various questions raised about thecourses. They also help in the correction of the exampapers. The TAs have also helped out students withtheir preparation for exams such GATE. They havemapped out the answers to the questions asked inGATE in the last � years over � disciplines. Overallwe have had excellent interaction with the TAs.

How is NPTEL di�ferent from other MOOC plat-forms like MIT Open Courseware and Coursera?The largest faction of our viewership consists ofIndians, apart from people from USA, Pakistan,Africa, etc. We prepare our courses so as to suitthe requirements of the Indian educational systemmore and to make our audience more comfortablewith the learning process. Also, we conduct o�flineproctored exams at the end of our online courses.The certificates that students receive in this way hasa greater value attached to them since they serve asproof of their own achievement.

Could you briefly tell us about the outreach pro-grammes and workshops carried out by NPTEL?We have held so many workshops in recent weeksthat we are starting to lose count! We have startedthis e�fort of creating local chapters in colleges. As a

part of this programme NPTEL has a representativein each of these colleges. He provides the necessaryinformation regarding the feedback on the courses,the work that must be carried out, how to improvethe course and also on the various courses requiredby the students.

We have coordinated workshops all over thecountry. We have gone to Kerala, Andhra Pradesh,Karnataka to name a few places. We have evengone to tribal areas in various states such as Orissa.At these places we go and describe what NPTELis, what a local chapter is, what online courses areand how they might benefit from them. We haveorganized more than ��� local chapters till nowand get requests for more each day, with requestsfor local chapters even coming from countries likeEcuador.

The certificates that students receive inthis way has a greater value attached to

them since they serve as proof of theirown achievement.

As the coordinating institute for NPTEL, what rolehas IIT Madras been playing in its functioning anddevelopment?IIT Madras has been one of the primary institutes

���| Immerse

in laying the foundation of NPTEL. It has providedfinancial support for this programme right from itslaunch in ����. It has also helped in the properdistribution and e�ficient use of this money. Manycomponents of the courses have been initiated atIIT Madras. Online courses, transcription andeven the subtitling of these courses have seen theirbeginnings at IITM.

The classroom where NPTEL lecturesare recorded at IIT Madras

What is the budget of NPTEL and how is it dis-tributed in the phases?The budget that we were provided for Phase � was|�� crores for five years. This amount was meantfor the creation of ��� courses. Since there was a lotof money le�t even a�ter the creation of ��� courses,we used the remainder for setting up of studios andfor hiring more sta�f members in the � institutes

involved in NPTEL. This amount which was meantfor � years in fact lasted for � years and was used verye�ficiently among the institutes.

As a part of the next phase which begins in����, the Project Advisory Board has sanctioned |��crores. This money is meant for the next � years andis to be used for the creation of more online courses.

In the future we want to be a body whichkeeps o�fering courses online that studentscan take from anywhere and use for credit

or even for employment.

What is your future vision for NPTEL?We are looking towards creating a virtual technicaluniversity. We have laid out a clear plan for thispurpose and are working towards it. We are alsolooking to initiate a credit based curriculum for alluniversities in India and this will hopefully becomea way through which students can earn creditswherever they are. In the future we want to be a bodywhich keeps o�fering courses online that studentscan take from anywhere and use for credit or evenfor employment. ⌅

Aryendra and Aslamah are second year un-dergraduate students at IIT Madras. They arecorrespondents for The Fifth Estate.

���| NPTEL

Cover Images CreditsFront Cover

Design: Vishal Upendran

ContentsDesign: E Amritha & Sree Ram Sai

Computer Science and EngineeringTM Krishna performing in a concertDesign: E AmrithaSource: S Hariharan

Aerospace EngineeringA jetpackDesign: Vishal UpendranSource: Wikimedia Commons

Applied MechanicsA set-up for precision glass mouldingDesign: Rohit Parasnis & Kiranmayi MalapakaSource: IPT Fraunhofer Griechenland

BiotechnologyA water color painting of DNADesign: E AmrithaSource: Caitie Magraw Art

Chemical EngineeringA type of algaeDesign: Swetamber & KiranmayiSource: Wikimedia Commons

ChemistryStructure of dendrimersDesign: E AmrithaSource: “Dendrimers Market”, NANOTECH-MAG Issue �� (����)

Civil EngineeringKedarnath templeDesign: E AmrithaSource: Debdutta Purkayastha via Blogger

Center For InnovationA concept image of an AUVDesign: Sree Ram Sai

Source: CC-BY-SA Erik Scott

Electrical EngineeringTesting the DISANET system at IIT MadrasDesign: Sree Ram SaiSource: Wikimedia Commons

Engineering DesignA breast cancer cellDesign: Sree Ram SaiSource: Wikimedia Commons

Humanities and Social SciencesA primary health care center in KarnatakaDesign: Raghavi & SwetamberSource: Wikimedia Commons

Management StudiesDesign: Kiranmayi & SwetamberSource: Shutterstock

MathematicsDesign: E AmrithaSource: YouTube

Mechanical EngineeringOrthopaedics: On-bone settingDesign: Vishal UpendranSource: Wikimedia Commons

Metallurgical EngineeringFriction weldingDesign: E AmrithaSource: TWI via Flickr

Ocean EngineeringA gas hydrate block embedded in the sediment of hy-drate ridge, o�f Oregon, USADesign: Vishal UpendranSource: Wikimedia Commons CC-BY-SA �.�

PhysicsCollision of matter and antimatterDesign: Vishal UpendranSource: Newscom

���| Cover Image Credits

���| Immerse

Thanks for Reading

Readers of Immerse will include faculty members and students at other IITs, IIS-ERs and NITs, where we will send copies, just like we did last time. Apart from theonline audience, of course.

If you are a student at IIT Madras and would like your research project to be fea-tured, we would be delighted to hear from you. Email us at immerse.iitm@gmail.com.

If you found this issue exciting and would like to contribute next time, be it as aneditor, writer, photographer or graphic designer, please get in touch with us at thesame email address. We will let you know how you can contribute.

In any case, if you have anything to say, be it praise or criticism, we welcome yourviews. Let us know by filling this feedback form goo.gl/BCSdkf, also accessible viathis QR code.

t�eiitm.org/immerse

���| Immerse