dmf simulation techniques - schaeffler group · the dual mass flywheel (dmf) ... a minimum by using...

5555LLuuKK SSYYMMPPOOSSIIUUMM 22000066

DMF simulationtechniques –Finding the needlein the haystack

Alexander FidlinRoland Seebacher

With the development time for new vehicle mod-els continuously falling (Figure 1) and thedemand by the automotive industry for cost sav-ings during the development phase increasing,the supplier chain must also rethink its productdevelopment.

In concrete terms, this means that fewer andfewer experimental vehicles are being built andmade available to the supplier industry. In thedevelopment and optimization of torsional vibra-tion dampers, the practical opportunities forvehicle testing are continuously shrinking,which has resulted in simulation technologygaining in importance.

Rising comfort demands (Figure 2) alongsideever stronger engines pose ever more difficultchallenges for torsional vibration damperdevelopment. In order to transmit increasingengine torques within the same packagespace, the characteristic curves must be madesteeper. Since higher engine torques also

result in greater irregularities, the dampermust do ever more to minimise vibrations inthe drivetrain.

The boundaries of what is physically feasiblecan, in principle, only be reached with the use ofsimulation technology and then expandedthrough the application of innovative technicalideas.

Although simulation technology already playsa significant role in the development and opti-mization of torsional vibration dampers, itsuse to date has been mainly problem-oriented.With the design concept already chosen, thereis often very little freedom to fully implementthe improvements identified.

The goal is thus, to integrate simulation technol-ogy into both the pre-development stage andinto the very early phase of product develop-ment, using simulation to optimize the productsfrom the beginning.

The DMF as an element in improvingvibration comfortDuring the 1980s, LuK made a decisive contri-bution in the area of driving comfort. The dualmass flywheel (DMF) made it possible to inso-late the torsional vibrations of the engine fromthe rest of the drivetrain (Figure 3). Annoyinggearbox rattling noises were eliminated andbody boom considerably reduced. It becamepossible to drive at very low engine speeds,therefore saving fuel.

Driving along with a situation dependent andthus varying engine speed is the most importantand dominant vehicle operating point. Never-theless, there are many other operating pointswhich must also be considered. Firstly, theengine must be started and later stopped atthe end of the journey and perhaps also attraffic lights. The drive itself begins with thevehicle launch. Changes in the accelerator pedalposition as well as gear changes cause loadchanges in the drivetrain, or the vehicle coastswithout load. These are only a few of the addi-tional operating points in which there is a high

5566 LLuuKK SSYYMMPPOOSSIIUUMM 22000066

44 DDMMFF ssiimmuullaattiioonn tteecchhnniiqquueess

Introduction

Figure 2 Expectations of increasing comfort

Figure 1 Changes in the development process

demand for comfort.Figure 4 shows theengine speed andengine speed irregu-larity curves for themain driving states.

In order to meet ever-increasing comfortdemands, LuK devel-oped a new methodfor optimizing the torsional vibrationdamper. The primarymetric is to opti-mize the DMF, giving appropriateconsideration to alloperating points andall relevant vehicleparameters. To takeadvantage of the full potential of theavailable parameters,


DDMMFF ssiimmuullaattiioonn tteecchhnniiqquueess 44

Figure 3 The effect of the DMF on comfort

Figure 4 Important vehicle operating points

optimization must be implemented as early aspossible in the concept phase. The method ofoptimization will be covered in greater detailbelow.

In achieving this metric, additional requirementsfor the simulation technology arise, which aredescribed below.

Simulation technologyprerequisitesRelevant operating pointsFirst, all potentially problematic operatingpoints must be defined. The overview in Figure5 lists the most important drivetrain prob-lems.

These problems can be largely divided into 3groups.

• Acoustic problems (gear rattle and bodyboom)

• Driveability problems (surging and judder)

• Durability (component strength)

Metrics, subjective evaluation and sensitivity graphsSuitable metrics must be found in order to objec-tively evaluate the design quality. This meansdetermining which physical values must bemeasured and establishing the extent to whichthese measured values correlate with subjectiveevaluation.

Figure 6 is a schematic representation of thedetermination of the metrics for the operatingstate “full load in drive.” Possible problemshere are gear rattle and body boom. To evaluatethe quality of the drive, the vehicle is accel-erated with maximum engine torque. Speedsare measured with high resolution at the trans-mission input, the differential input (for rear-wheel-drive vehicles) and at the wheel.Depending on the metric, the signals are differ-entiated or integrated over time and theiramplitudes graphed as a function of enginespeed. To evaluate the gear rattle, the averageamplitudes of the angular fluctuations of thetransmission signal are taken as the metric andevaluated over the critical engine speed range.Body boom correlates well enough with theover-accelerations at the wheel and at the dif-ferential input. In the range below 1500 rpm,the maximum acceleration of the wheel servesas the metric. Especially in rear-wheel or all-wheel-drive vehicles, drivetrain resonance canbring about boom noises. As the metric for this,the peak acceleration is evaluated at the differ-ential input.

Ideally these measurements are performedwith different DMF designs and damper sys-tems. In this way, the largest possible range ofmeasured metrics is captured. The various sys-tems are evaluated subjectively when themeasurements are taken. When the subjectiveevaluations are plotted against the respectivemetrics, the result is known as the sensitivitydiagram (Figure 7).

The use of systems that exceed the comfortachievable with a DMF alone is optimal. Forexample, torsional vibrations can be reduced toa minimum by using an add-on clutch slip con-troller. In this way, it can be seen, for example,



Figure 5 Drivetrain problems

Operating point Problem

Idle Gear rattle

Drive Gear rattle, boom

Coast Gear rattle, boom

Engine stop Gear rattle,clatter

Tip-in and back-out Surging

Vehicle launch Judder, rattle, surging

Engine start Durability, comfort

Sub-idle speed reso-nance drive

Durability

to what extent a noise can be reduced by min-imising torsional vibration. The above men-tioned add-on slip controller was specificallydeveloped to produce the most comprehensiveand continuous sensitivity diagram possible.This add-on slip controller is integrated directlyin the release system, similar to a clutch-by-wire system.

Figure 7 shows the subjective evaluations plot-ted against the metric using the example of bodyboom at low engine speed. A rating of 10 corre-sponds to noiseless driving, a rating of 1 to intol-erable body boom. For most customers, the tar-get rating is 8, but a rating of 6 is the absoluteminimum.

With such measurements, on many vehicles ofdifferent types, LuK expands its know-how on acontinuous basis.



Figure 6 Full load in drive mode: Analysis of the metrics for rattle and boom

Figure 7 Sensitivity diagram for boom at low enginespeeds

EfficientmodelsLimitations due tocomputational capaci-ty, demand the use ofefficient simulationmodels with adequateaccuracy.

A model with 9 rotat-ing inertias is used to represent the vehicle in the drivemode (Figure 8).When possible, onlythe relevant naturalmodes and drivetrainparameters are con-sidered.

In contrast, all power-transmitting ele-ments and excitingobjects are modelledin greater detail. Thetorque fluctuationsof the combustion



Figure 8 Full load in drive mode: analytical vibration model, measurement and simulation

Figure 9 Engine start: analytical vibration model, measurement. simulation and metrics

engine are represented as a function of thecrankshaft angle and the static torque. Thetransmission behaviour of the DMF is consid-ered with all of the complex and nonlinearproperties of the arc spring. The stiffness,damping and lash of the drivetrain are deter-mined experimentally with special measure-ments.

Such a model is able to describe the operatingpoint drive (Figure 9) with sufficient accuracy.

Figure 9 shows a comparable analytical vibrationmodel, measurement-simulation comparisonand metrics for the engine start operating point,which can also be very well described.

As was the case with the drive mode, the enginestart as a whole is considered. In addition to theDMF parameters, the engine, engine manage-ment system, starter motor and accessory driveare considered. The metrics here are the startingtime and the maximum torque in the DMF. Whilst the starting time metric is a comfort vari-able, the maximum damper torque metric is adurability variable. Similar procedures havebeen developed for all of the other operatingpoints.

Consideration of the complete vehicleA total system consideration is another impor-tant feature in performing simulation calcula-tions. Not only LuK products such as the DMFand the complete clutch system are consid-ered, but also the entire drivetrain, the engineand the engine management system.

This makes it possible to better assess the fullpotential and limitations of the various LuKproducts so that other areas of the drivetraincan also be included in the problem solvingprocess when needed.

The total system consideration also makes itpossible to detect potential problems with inter-actions between the torsional vibration damperand other components in the vehicle. In particu-lar, the engine management system, with idleregulator, cylinder balancing controller, anti-jerkcontroller, etc., is very often in complex interac-tion with the DMF. Here too it is very desirable tobe able to affect an improvement early in thedevelopment phase.



Figure 10 Total system consideration



Figure 11 Engine start: Interaction of the DMF with the starter and engine management system

Figure 12 Idle: interaction between the DMF and engine management system

Figure 11 shows the effect of engine manage-ment and starter speed on the engine start.The 3 graphs show the speed signals of theengine and transmission. First is a phase inwhich the unfired engine is propelled bythe starter motor alone. After a few cycles, the engine is started. Then the engine speedruns through the DMF resonance, which fre-quently causes problems during vehicle tun-ing.

A drop of just 30 rpm in the starter speed hasnoticeable negative effect on the engine start-up. (Figure 11, top right graph). The situationbecomes critical if the starter disengages toosoon, e.g. before the first combustion event. The consequence is unacceptable resonancevibrations, which in an extreme case couldcause a resonance hang-up (Figure 11, bottomright graph).

Interactions between the complex torquetransmission behaviour of the DMF and theengine management frequently have a nega-tive effect on driving comfort in idle mode aswell. A lash is intentionally introduced inorder to achieve perfect isolation of thefirering frequency between the primary andsecondary masses of the DMF. Annoying gearrattle being completely eliminated (Figure 12,left graph). This lash and the combinedeffects of the engine speed and regulator sys-tem can also excite low-frequency oscilla-tions, causing unpleasant vibrations (Figure12, right graphs).

For both examples, it is clearly necessary toinclude engine related parameters early in thedesign. This naturally requires that the cus-tomer be included early in the developmentprocess. The accuracy of the vehicle data hasa decisive effect on the quality of the simula-tion results and thus on the quality of theproduct.

Automated calculation sequenceThousands of simulations must be performedfor the calculation, depending on the numberof influencing parameters and operatingpoints to be calculated. This requires a simu-

lation environment with the greatest possibledegree of automation. Appropriate damperand vehicle models are generated from ProEdrawings and vehicle data. The calculationplan is created using the influencing para-meters. The simulations in the time domain provide raw data for analysis. From the metrics thus determined, an optimal torsion-al vibration damper is then designed. LuKuses programs developed in-house to per-form the simulations and determine the met-rics.

Integration in theearly developmentprocessUp to now, vehicle tunings have occurred very close to the start of production, i.e. thedesign concept is already fixed and thereremains little freedom in which to makeimprovements. However, to take advantage of the full potential of the available para-meters, it is necessary to include the damper optimization very early in the devel-opment process. It is much easier to imple-ment suggestions for optimization in thisphase.



Figure 13 Opportunities for change throughout the devel-opment process

Optimization procedureThe process that has been developed is orient-ed around three essential requirements thatare included by definition in product optimiza-tion.

• The function of the LuK products depends onmany design parameters. Specifically for thedual mass flywheel, there are normallybetween 18 and 24 parameters (dependingon the complexity).

• The products are not optimized in relation toany individual metric, rather they must func-tion well in all customer-relevant driving situ-ations.

• The products are mass-produced. Thereforethe production tolerances must be taken intoconsideration when developing an optimalproduct.

These requirements essentially determine theoptimization strategy. Figure 14 shows the partsthat are contained in a simple DMF.

Naturally some of these components and theirgeometry can be reduced to a limited number ofparameters in the evaluation of the function and

strength of the DMF. However, since a descrip-tion of, for example, a coil spring must considernot just its characteristic curve, but also the wiregeometry to calculate the stress, the number ofparameters becomes relatively large. It becomeslarger the more complex the specific structure is(simple DMF, DMF with inner damper, DMF withinner damper and torsion-damped clutch disc,etc.).

This gives the above mentioned number ofparameters, which determines the dimensionof the parameter spectrum in which the opti-mization is carried out. This dimension raisesexponentially the number of points that wouldbe necessary to scan the spectrum evenly. Figure 15 illustrates this fact. If we were to sub-divide each parameter axis in the two-dimen-sional case into two sections, we would have 2² = 4 regions (squares). In the three-dimen-

sional case, there are2³ = 8 regions(cubes). In a realistictwenty-dimensionalcase, there would be 220 = 1,048,576regions.

Unfortunately, it isnormally completelyinsufficient to varythe parameters onjust two levels. If wesubdivide each axisinto several sections,this produces anastronomical numberof possible parametercombinations, whichcould not be man-aged even with themost modern simula-tion technology.



Figure 15 Example: parameter spectrum divided into two-and three-dimensional cases

Figure 14 Components of a DMF

The solution is found in the systematic applica-tion of statistical methods, which were originallydeveloped for test planning. These methodsmake it possible to limit the number of parame-ter combinations and reduce them to a size thatcan more easily be handled. One factor in favourof this method is that the variation limits of indi-vidual parameters must be related to the toler-ance variations in series production. This idea isillustrated in Figure 16.

Assuming that the curve in this diagram repre-sents the optimization target, the best resultwould then be achievable at the narrow peakon the right in the diagram. If, however, weplace the width of this peak in relation to pro-duction tolerances (red rectangle), it can beseen that a clearly worse result is obtainedwithin the tolerances than if the wide plateauin the middle of the diagram were selected asthe optimization target (this variance is indi-cated by the blue rectangle).

In other words, it must be ensured that no partin mass production is worse than a certainlimit.

Another and entirely more important aspect isthe diverse number of driving situations men-tioned above that can be influenced by theDMF. This is illustrated schematically in Figure 17 in the form of a radar graph. Only themost important driving situations are listed.Each of these terms includes several, some-times very different phenomena, which shouldbe kept separate from one another when evalu-ating a vehicle. Under engine start, for exam-ple, starting time is distinguished from com-fort, and under stop, rattling is distinguishedfrom clatter.

All of these initially subjectively evaluated phe-nomena are represented using the sensitivitydiagrams mentioned above.

In every product development or product opti-mization process, the question naturally arisesas to how the various contradictory require-ments of a system can be combined. In otherwords, how can we find an acceptable trade-off?The usual procedure is to combine individualmetrics using a weighted average into one over-all metric.

Z is the overall metric, z1, z2, ... , zN has the indi-vidual metrics or driving situations (total of N),k1, k2, ... , kN shows the weighting of the driving situations. This method also has abasic disadvantage. If a parameter combina-tion is found that results in very poor ratingsin certain driving situations, but looks good in all others, it could be viewed as optimum by the automated optimizer. However, thisis contrary to real world requirements. A product must not perform below a certain limitin any driving situation. This is called theacceptance boundary. In addition to comfort-related metrics, there are also durabilitytargets (e.g. stress). These boundaries mustbe met in all cases.



Figure 17 Diversity of driving situations and individualevaluations

Figure 16 Relation between narrow peaks and wideplateaus during optimization for massproduction

In order to take this into account, LuK does notuse the arithmetic mean, but the geometricmean:

g1, g2, ... , gN shows the respective acceptanceboundaries for the relevant metrics. This formensures that as soon as one metric falls belowthe acceptance boundary, the entire parametercombination is also automatically rated unac-ceptable.

The acceptance boundary for comfort-relevantmetrics generally corresponds to a subjectiverating of 6. For durability related metrics, it cor-responds to the minimum permissible safety fac-tor (e.g. 1.2). Every effort is made, however, toachieve a subjective rating of 8 for the comfort-relevant metrics.

The weightings k1, k2, ... , kN make it possible toprioritize the targets specifically for each cus-tomer and/or vehicle model. Naturally, the driv-ing characteristics and the resulting comfortdemands are different for a limousine than for asports car. When establishing the weightings,intensive discussions between OEM and suppli-er are essential.

In summary, we can say that the developedmethod enables a comprehensive and holisticanalysis and optimization of a product within abroad field of variations. The proceduresdescribed and the conclusions to be drawn fromthem are illustrated in the next section by way ofan example.

Optimization example using a DMFThe described procedure is illustrated in theoptimization of a conventional DMF. The selected object of the optimization was a com-pact vehicle with a modern 4-cylinder dieselengine with rear wheel drive. This selectionposes particularly stringent requirements ofthe torsional damper and is therefore represen-tative for the conditions in which such systems

will have to prove themselves in the nearfuture.

Before the optimization can be carried out, a fewsteps must be taken. First, a preliminary study ismade of various pre-selected concepts-, where-by a few concepts – featuring modular design –are compared with one another. For example, thefollowing components are available for selec-tion:

• A one- or two-stage arc spring

• The DMF can be equipped with a one- or two-stage inner damper or have no innerdamper

• The clutch disc can be rigid or have a torsiondamper

Once a basic concept has been decided upon,the necessary parameters and their variationlimits can be established. The first considerationhere is the available installation space.

Every optimization starts with a basic designdrawing. The quality of the basic drawing ishighly dependent on the experience of thedesigner. However, by using the optimizationmethod, the influence of the designer's experi-ence is greatly reduced. The quality of the basicdrawing does not determine the optimizationresult, only the time taken until the optimum isreached.

In our case, a DMF with a two-stage arc springwas selected for the optimization (see Figure14). Its function and durability can be



Figure 18 Optimization results for DMF with two-stage arcspring

described with 16 parameters. The functionwas evaluated in seven driving situations. Nor-mal driving was described with two differentphenomena: drive-rattle and drive-boom. This takes the special importance of these phenomena into account. Around 5600 simula-tion calculations are performed in the opti-mization. The optimization results are shownin Figure 18.

The blue coloured area shows the range of varia-tion of the evaluations made during the opti-mization. There is no DMF that would correspondto the inside or outside edge of this range. A DMFassigned a rating of 10 in coast, for example, wasonly given a 7 in stop. The boundaries show,however, what is possible in general with theconcept under study in the various driving situa-tions.

The red line indicates the optimization result.This DMF is actually acceptable in nearly all driv-ing situations. Only for booming in drive modedid it receive a rating just under 6, and so it has

delivered the best possible rating for the con-cept under study.

Since the optimal evaluation still lies within theunacceptable range, the introduction of addi-tional components in the DMF must be consid-ered. The next obvious step is to add an innerdamper to the two-stage arc spring. This designis shown in Figure 19.

This raises the number of parameters to 22 andresults in a “softer” torsional characteristiccurve, with reduced hysteresis, which benefitsthe overall vibration isolation (see Figure 20).

The optimization result for this concept is shownin Figure 21. Figure 22 shows the comparisonbetween these two concepts.



Figure 20 Comparison of spring characteristics of DMF withand without inner damper

Figure 19 Components of a DMF with inner damper

Figure 21 Optimization results for DMF with two-stage arcspring and inner damper

Figure 22 Comparison of DMF with and without innerdamper

With considerable additional expense, we areable here to raise the evaluation of the drive-boom at its optimal point to a rating of 6.5. Thisrelates to the nominal design, in which massproduction tolerances are entirely ignored. How-ever, using the previously described simulationtool, we can also undertake a tolerance study.Here, all parameters are varied within produc-tion tolerances. The variation of individualparameters was determined from actual produc-tion data. The result (Figure 23) shows a cleardispersion of the expected evaluations, with arange up to the equivalent of an entire rating.

It can be seen, however, that under mass pro-duction conditions, a rating above 6 can beensured in all driving situations. This methodalso makes it possible to estimate the probabili-ties of individual evaluations. The correspondingfrequency distributions are shown in Figure 24.

The analysis of the most important influencingfactors also makes the conflicting goalsbetween start and boom in the drive modeapparent. In order to improve starting behav-iour, for example, the secondary mass had to bereduced, and the basic friction increased. Bothchanges worsen boom in the drive mode, how-ever (Figure 25).

As can be seen from Figure 21, further improve-ment with regard to booming was not possiblewith the concept under study. Alternative tech-niques with new influencing factors are requiredto realise significant further improvements.

Solution throughinnovationThe problem described in the last section canonly be efficiently solved by changing the struc-ture of the torsion damper. If we stay with theconventional spring-mass systems, we must notonly consider the described conflict betweenstarting and booming, but also the perpetualconflict between the transmission of enginetorque and the quality of vibration isolation. If the available installation space remains



Figure 23 Results of the tolerance study Figure 25 The most important influencing factors and theireffects in relation to boom in drive mode andstart behaviour

Figure 24 Distributions of the ratings for start and drivemode boom

unchanged, increasing torques require everstiffer spring rates, which degrade the vibrationisolation.

One alternative comes from an idea known formany decades – but as yet never successfullyapplied in the automotive industry – of a cen-trifugal pendulum-type absorber. The functionalprinciple of a secondary side centrifugal pendu-lum-type absorber is illustrated in Figure 26.

The centrifugal pendulum-type absorber workslike a damper, whose effective stiffness is gener-ated by centrifugal force. The corresponding cor-recting torque is

.

This means that the natural frequency of thevibration absorber is proportional to the speed.This makes it possible to combat the selectedorders of excitation (e.g. the main excitation)very efficiently with the right tuning of the vibra-tion absorber. Figure 27 shows the transfer func-tion of an ideal centrifugal pendulum-typeabsorber tuned to the second engine order.

The effect of the centrifugal pendulum-typeabsorber is of course limited by its mass and thevibration angle. For that reason, the centrifugalpendulum-type absorber can only be used inaddition to a spring-mass damper in the drive-train (see Figure 26). The basic isolation stillcomes from a spring-mass system. The remain-ing vibrations are counteracted by absorption atthe engine firing frequency.

One design realization of the DMF with a second-ary-side centrifugal pendulum-type absorber isshown in Figure 28.

The space in a conventional DMF where the innerdamper would be located is instead used toaccommodate the centrifugal pendulum-typeabsorber.

The optimization method already described is anideal tool for designing such a system. With theuse of the centrifugal pendulum-type absorberto reduce torsional irregularities in drive, it



Figure 26 Functional arrangement of the centrifugal pendulum-type absorber

Figure 27 Theoretical transfer function of a centrifugal pen-dulum-type absorber

Figure 28 Design realization of a DMF with secondary-sidecentrifugal pendulum-type absorber

becomes possible to tune the spring compo-nents differently and to significantly improve thenoise level in drive (both rattling and booming).The optimization result is shown in Figure 29.

Due to the separation of functions, the vibrationisolation in drive could be improved by approxi-mately one rating. This improvement in the sub-jective evaluation means in practical terms halv-

ing the vibration amplitude in the lower speedrange, as illustrated in Figure 30.

This comparison affirms on one hand to theeffectiveness of the new system, but on theother hand points to possible further develop-ments. A certain worsening of the isolation in thehigher speed range is not perceived subjectively



Figure 29 Comparison of optimization results of a conventionalDMF with inner damper and a DMF with secondary-side centrifugal pendulum-type absorber

Figure 30 Comparison of test results between a conven-tional DMF with inner damper and a DMF withcentrifugal pendulum-type absorber

Figure 31 DMF with cover-side centrifugal pendulum-type absorber

as critical in the tested vehicle, but must be over-come so that the new system can be used in thebroadest possible spectrum of applications. Thisworsening has two causes. Firstly, the spring-mass system now becomes stiffer without theinner damper (see Figure 20) and secondly,because the friction in the suspension of thecentrifugal pendulum-type absorber itselfincreases with speed. These reduce the degreeof isolation.

In order to overcome at least the first problem,LuK has developed the concept of the cover-sidecentrifugal pendulum-type absorber (Figure 31).This new concept makes it possible to recoverthe space needed for the inner damper. In ourexample, the drive-boom evaluation is clearlyone rating better. The expected rating of this sys-tem is given in Figure 32.

SummaryWe have presented a new technique to supportthe development of torsional vibration dampers.The aim of this method is to integrate simulationtechnology into the early concept phase of prod-uct development. Here, the proposed conceptsare tested using simulation technology, and theinfluencing parameters on dynamic behaviouroptimized with respect to the entire drivetrain.

What is particularly new about this method is thefact that all operating points and production tol-erances are considered in optimizing the tor-sional vibration damper. This provides for con-tinuous quality and process assurance betweendesign, testing and production. Another impor-tant factor in the assurance of product quality isthe availability of very detailed vehicle andengine data. This requires the close integrationof the customer, from the very beginning, withinthe product development process. Potentialproblems, such as interactions of the dual massflywheel and the engine management systemcan be recognised early on and solved in theirentirety.

Since all operating points as well as a great vari-ety of parameters and variations must be consid-ered, a large number of simulation calculationsare required, necessitating a high degree ofautomation. The models and the resulting simu-lation results must achieve a high level of accu-racy. Metrics that best reflect subjective evalua-tions are defined to evaluate the simulationresults.

This method was developed for the optimizationof dual mass flywheels, but is in principle alsoapplicable to all torsional vibration dampers, aswell as to the optimization of the entire clutchsystem. One example can be found in the article“The Clutch and the Release System” [1]. Thesimulation tools developed can also be used fordesign quality control.

In addition to product optimization and qualityassurance, the systematic application of thismethod reveals conflicting design targets andtests the limitations of existing concepts.Prompting the search for new and innovativeideas that can be comprehensively tested,evaluated and developed virtually, right up tothe prototype phase using the methodsdescribed.

Literature[1] Zink, M,; Hausner, M.; Welter, R.; Shead, R.:

The Clutch and the Release System – Anengaging topic 8. LuK Kolloquium 2006



Figure 32 Comparison between centrifugal pendulum-typeabsorbers, inner and outer with additional ID

dmf simulation techniques - schaeffler group · the dual mass flywheel (dmf) ... a minimum by using...

Documents