sample project review.pdf

8/14/2019 Sample Project Review.pdf

1/27

Batman

(Each presentation has been given an

unique name so as to hide the identity of

the presentation author from the Reviewer,but known to the Instructor)


2/27

Finite Element Validation of Low

Impact Response on a Lab-Scale

Space Frame Structure

Jagadeep Thota, Mohamed Trabia & Brendan OTooleDepartment of Mechanical EngineeringHoward R. Hughes College of Engineering

University of Nevada, Las Vegas

Las Vegas, NV, USA.

ASME 2012 Verification & Validation SymposiumMay 2nd4th, 2012, Las Vegas, NV, USA


3/27

Background Space frame structures have been commonly used in

vehicles to enhance their structural strength while reducing

the overall weight.

When a vehicle, with an internal space frame structure, is

subjected to an impact load, the individual frames and

joints of the space frame play a critical role in mitigating

the generated shocks.

In order to effectively design the space frame structure, it

is important to predict the propagation of these shocks

through the space frame members.

While performance of space frame structures under staticloads is well-understood, research on space frame

structures subjected to impact loading is minimal


4/27

Literature Review Gaul and Lenz (1997) showed that nonlinear shock transfer

performance of joints has substantial influence on the dynamics

of the structure as they induce large amount of damping.

Sandia National Laboratory (2001) conducted FE studies for

investigating energy dissipation due to micro-slip in the bolted

joints.

Song et al. (2002, 2004) developed a beam element, which can

simulate the non-linear behavior of bolted joints on a vibrating

frame.

Ibrahim and Pettit (2005) suggested that friction in bolted joints

is a main sources of energy dissipation in mechanical structures.

Thota et al. (2011) conducted computational studies on a military

vehicle space frame subjected to high impact load.


5/27

Objective

Develop a lab-scale space frame structure

having bolted joints.

Conduct a low impact experiment on the space

frame structure and measuring the resulting

acceleration (shock) response.

Propose a FE method that can predict the shock

response measured in the experiment.


6/27

Lab-Scale Space Frame To consider the shocks within a 3-D structure, a lab-scale space frame

is designed.

The overall length of the cube shaped structure is 482.6 mm. The frame members are hollow, having square cross-section with wall

thickness being 3.175 mm.


7/27

Joint Design The joint halves are C-shaped sections, which are bolted together through

the frame members.

The joint has two orthogonal branches of 114.3 mm, and the combinedwidth is 50.8 mm.

The angle joint houses the ends of the longer frame members while the

shorter frame member ends are enclosed by the joint halves.

The angle joint legs are 100 mm long, and width is 50.8 mm.

The wall thickness of the joints (including angle) is 6.35 mm.

The length of frame members outside

the joints are:

Horizontal members are 254 mm

long, Vertical members are 381 mm long.


8/27

Joint Halves

To eliminate noise in the acceleration (shock)

signal, the faces of the opposing joint halves are3.175 mm apart.

This arrangement ensures a more homogenous

contact between the joint and the frame members.


9/27

Space Frame Sections

D

W

tD

W

t

D

W

t

Frame member Angle Joint Joint half

Al l dimensions are in mm

Type D W t

Frame member 38.1 38.1 3.2

Angle joint 88.9 88.9 6.4

Joint 50.8 25.4 6.4


10/27

Material All components of the lab-scale space frame, except the bolts, is made

of Aluminum 6061 alloy.

The structure approximately weighs 11.4 kg.

Density

(kg/m3)

Youngs

Modulus

(GPa)

Poissons

Ratio

Yield

Strength

(MPa)

Tangent

Modulus

(MPa)

2700 68.9 0.33 276 562

Strain

Stress

Yield

Point Tangent

Modulus

Elastic

Modulus

Failure

Point

MAT_PLASTIC_KINEMATIC


11/27

Bolt Tightening Grade 8 bolts were used in the space frame structure .

The bolts are tightened to reduce the noise in the output signals that can result

from loose connections. Applying the same tightening torque on all bolts ensures the repeatability of

the results.

The bolts are tightened to a preload of 10.8 kN and a torque of 12.5 Nm.

These values are computed from the standard design equations:

= 0.9 = 10.8

= 0.21 = 12.5

Sp= Proof stress of the bolt material = 586 MPa

At= Tensile stress area of bolt = 2.1e-5 m2

dp= Pitch diameter of the bolt threads = 5.525e-3 m


12/27

Impact Experiment Low velocity, non-destructive, impact experiment was carried out.

The structure is placed on an aluminum support during the experiment.

An upper frame member is impacted at the mid-member location with a forcehammer.

Acceleration is recorded, through an accelerometer, in the middle of the

opposite frame member location.

Fast Fourier Transform (FFT) of the resulting acceleration signal is conducted.

FFT is used to determine the natural frequencies of the space frame.


13/27

Finite Element Model All components of the lab-scale structure are modeled as beam elements.

Common elements between the different components of the structure are merged to

obtain contact.

The angle joint and the bolts are not structurally modeled, but their masses are

accounted for by adding mass-elements at each corner of the cube.

Preprocessor: Altair-HyperMesh (v 9.0)

Solver: LS-DYNA (v 971)


14/27

Finite Element Model

(3-Drepresentation

is for

illustration

purpose only)


15/27

Finite Element Model The joint is modeled as two parts:

The first part (blue) comprises of

combined cross-sections of the frameand joint.

The second part (red) is the cross-

section of the joint.

The length of the beam elements are

maintained at 3.2 mm, resulting in a

total of 1,832 elements.

The loading condition and output

similar to the experiment are

mimicked.

Total simulation run time: 8 ms CPU (3 GHz Intel Xeon processor

with 2 GB RAM) time: Approximately

6 minutes.


16/27

Results The unfiltered force signal obtained from the experiment is used to define the

impact curve for FE analysis.

The acceleration signals from experiment and simulation are filtered usingButterworth low-pass filter with cutoff frequency of 10,000 Hz.

The sampling rate for the experiment and FE analysis is 1 mega-

sample/second.

Typical Force

Signal


17/27

Results: Acceleration Signal The predicted acceleration signal captures the first peak of the experiment.

Most of the subsequent acceleration peaks for the FE model are smaller than the

experimental result. The frequency of the predicted signal matches well with the experimental signal.


18/27

Results: FFT The first predicted natural frequency is very close to the experimental one.

The rest of the natural frequencies of the cube, including the predominant one, 1500 Hz,are predicted by the FE model.

There is an additional frequency, 810 Hz, predicted by the FE model:

This may be due to the absence of some structural components such as angle joints and bolts,and holes in the FE model which might have suppressed this additional frequency.

The acceleration amplitude of this frequency is small.

Overall, this a very good match for a space frame structure such as the cube comprising ofa total of 48 bolts, 36 structural components, and 8 joint locations.

Experimental Computational


19/27

Conclusions

Shock propagation though a space frame with bolted

joints is not well-understood. An approach using finite element analysis for

predicting shock transmission within such structures

is proposed.

The proposed approach is verified using a lab-scale

space frame structure.

Comparing experimental and finite element results

lead to the following observations: The initial peak of the acceleration signals match closely.

The FE model is able to predict all the experimental

natural frequencies.


20/27

Future Work

Incorporate bolts in the FE model. Explore ways for mitigating shocks by optimizing

the joint design variables and include shockabsorbing material.

Expand current research to model shocks resultingfrom high velocity impact.


21/27

Acknowledgments

We would like to thank Mr. Ami Frydman, U.S.

Army Research Laboratory, for interacting withthe authors.

We are grateful to Dr. Douglas Templeton, U.S.Army TACOM-TARDEC, for being helpful in

developing the ideas of this research. This work was funded through a cooperative

agreement with the U.S. Army ResearchLaboratory under contract DAAD19-03-2-0007.


22/27

Thank you

&

Questions?


23/27

Robins Comments

(type your name in place of Robin!)

6 comments on what is good in this


24/27

6 comments on what is good in this

presentation

The model looks awesome

This is a bad comment!

The comment is too abstract

No explanation is given for why the reviewer thinks the

model is awesome

Good comments:

The quality of mesh is quite good as all the elements

comprise of the same element size and even the

smallest of the components is finely meshed. The description regarding why the joint halves should

not touching each other is informative and useful for

other researchers.

8 comments on what the presentation is


25/27

8 comments on what the presentation is

lacking OR not clear

I dont like the figures

Bad comment

Too abstract: all figures? particular figure? Why?

Sounds personal than technical (avoid using I, we, this

person, this student.)

Good comments:

Not clear on why static material properties were used for a

dynamic analysis.

Not mentioned anywhere what type of beam elements wereused, i.e., will the space frame members acting as beams take

into effect the shear in the beams.

The natural frequency values, and the axis values on the FFT

plots on slide 18 are not clear and too small to read.


26/27

Do not include this slide when submitting

The deadline for the review comments is

14thMay 2013(by 5:00 pm).

Emailme back the entire presentation with

your review comments at the end of the

presentation. 8 %of your project grade will be based on

how good/bad you have reviewed the

assigned project. The review comments for a project will not

have an effect on the grade of that project.


27/27

Do not include this slide when submitting

Project grade breakdown:

Structural model: 10%

Heat transfer model: 8%

Modal analysis model: 4%

Presentation (including the results shown in thepresentation): 15%

Review comments: 8%

Total Project grade: 45% Total HW grade: 50%

Inclass:5%

sample project review.pdf

Documents