LS-DYNA TUTORIAL 10: Three Point Bending of Sandwich Structure with EPS Foam as Core

This is a long and heavy tutorial, so buckle up everyone!

Topics covered in this video, includes but not limited to:
- How to make a simple sandwich structure
- How to set up three point bending test simulation
- How to model failure with ADD_EROSION
- How to model styrofoam with MAT063 CRUSHABLE_FOAM
- How to prevent negative volume error in solid elements
- How to combat hourglassing in soft materials
- How to 'glue' parts with tienodes contact
- How to use apply velocity/displacement with BOUNDARY PRESCRIBED MOTION RIGID
- How to extract the forces with BNDOUT option

As mentioned, I did the modelling of
EPS foam for my final year project. Here the link to the published paper:
(do check it out and hope you can cite if it is helping you in your own research!) 😁😁😁

Here is the link to the stress-strain file:

Here is a sample of 3 point bending of foam:

[alparslansolak]

Thank you Ameen Topa, this tutorial will be useful for engineers, can you make the same video for a composite sandwich structure which includes face sheet(carbon fiber, fiberglass) and honeycomb(Nomex, aluminum) ? Kind regards

[vladimirlubimov]

Good afternoon, Ameen. Thank you very much for the tutorial video, thanks to them I have become much better at modeling. I have two questions for you:
1) Could you make a video in which you showed how to use the basic tools of LS-DYNA (for example, editing geometry, building a grid on a 3D model, measuring distances between points, planes, and more).
2) Please advise which material model can be used to model a polymer material (sealant, mastic), for example, such as is used for vibration isolation of the car body or for sealing the seams of the cooling system of the car.
I am very grateful and grateful to you in advance. You're doing a wonderful job!!

[mulukenmeaza]

Thank you vey much. I am enjoying so much your videos. I am in situation that I have to apply load on the specific area using rigid body impactor. My plate is refined mesh and the impactor is course mesh. I choose the impactor because I can simply change the mesh of my plate and I don't have to select the node set where i apply the load manually. So, I have tried to apply force on the plate using rectangular box so that I can have the force on specific area on the plate (rectangular area). My assumption is to make the impactor to be rigid and apply force on the impactor (10KN divided by number of nodes of the impactor). I expected the rigid body to transfer the load to the plate. It works; however, I don't get the same result when I apply the forces directly on the plate on the same area by manually selecting nodes using node sets. What I feel is that the impactor doesn't transfer all the load to the plate. Can you give a tip what contacts I have to use when the impactor touches the plate they go along together and transfer the whole load?

[sriharikurukuri]

Appreciated your great work and sharing your modeling knowledge.

[vaibhavnmete]

Great work man. I appreciate your effort in providing knowledge to us. Similar work I have done, but I am not getting, what is "Tensile Stress Cut-off"?

[jacovanniekerk]

Great tutorial, thank you! I have been trying to simulate compression of different densities of EPS blocks that are impacted by a steel bar (50J kinetic energy). I'm particularly interested in ensuring model stability at the point where the foam bottoms out. I've used your 2014 paper as reference for the basic model setup, and stress strain data from compression experiments on my material, but cannot achieve realistic behaviour. I've tried adding *CONTACT_INTERIOR and changing the contacts from *AUTOMATIC_SURFACE_TO_SURFACE to NODES_TO_SURFACE. I have not yet tried adding *MAT_ADD_EROSION as I've set ERODE = 1 in *CONTROL_TIMESTEP. I still get negative volumes and weird deformations at high strain (>95%). I've tried HEX elements (ELFORM 1) and TET elements (ELFORM 10 and 13). Any suggestions on what to look at?

[satisfyingyounes]

Thank you for your tutorial. Can you please add a tutorial that cover composite materials and their fiber orientation.
Regards,

[김창호기계공학과]

Is the s-s graph of the data for LCID a compression test, a tensile test, or a bending test? What graph does it insert?

Also, can I put the EPS foam, the shock absorbing foam you provided on YouTube, into the LCID?

[AntonioGarofano-xk]

Thank you for your informative and detailed videos, this one was also very useful to me. I am modelling aircraft seats to be integrated into a fuselage to analyse a drop test with dummies. Some papers I found suggest using MAT57 card for seat foam.
Which approach do you think is best suited to my case? Mat57 or Mat63 like the one in the video?

Also, what is the main role of the contact_interior? Is it also important to set it in my case?

[abdulkadirgenc]

Thank you for this helpful video package.
Can you please explain why we defined node set on the core but segment set on the upper and lower plates prior to connecting them? Why not segment set on both or node set on both? I would appreciate if you explain.
Thank you in advance.

[anaychk]

Thanks for the tutorial. Much appreciated your work.

[emregungor]

very successful and comprehensive work, thank you.

[vladimirlubimov]

Good afternoon, Ameen! In dynamic implicit analysis, compression of a foam cube works perfectly and according to the results, the graph of the deformation rate in time looks perfect and converges with the calculated one for the entire test interval (compression). But with explicit modeling of the same process, the graph of the dependence of the strain rate on time looks quite strange and after a certain deformation begins to "jump" very much, which affects, respectively, the results of stress determination (since several curves are laid for different speeds). I tried to change the contact and do it without contact, applying speed to the upper nodes of the cube, the effect is the same. Obviously there is some kind of error in the calculation. I use tetra elements. Can you tell me in which direction you can move to eliminate such fluctuations in the deformation rate?

[haidermraih]

Thanks, Ameen. In Tutorial 10, should we fix the Sandwich beam in the Y axes?
I think It can move when we apply the load.

[chanduiithyd]

Hi. In my analysis i dropped crushable foam on fixed rigid block. I have defined nodes to surface contact. still its not detecting contact and foam block is passing through rigid base. What might be the reason?

[thedomevlog]

How you get the load curve ID for the crushable foam

[김창호기계공학과]

I performed sensitivity analysis on the unloading parameters HU and Shape Factor using only one model that compresses a single element. However, in the case of Shape Factor, it was confirmed that the unloading behavior changes well as the value is changed (Shape Factor is performed from 1 to 30), but Hu Factor is set to (1e-3, 1e-2, 1e-1, 1). We confirmed that all numbers were the same after making the change. In the case of the original HU factor, does the Hu curve not change significantly when changed?

[sandeepprajapati]

Hello Sir, I want to do crash analysis of helmet in LS DYNA, so it has basically three parts, visor, outer shell, impact absorbing material and then comforting foam, which are made of ABS foam, polystyrene, polyurethane. I have to impact the helmet at different velocities ranging from 5 to 10 m/s on an anvil, and then Charphy izod test and then to calculate the Head Injury Criterion, can you please guide me how to do it, it will be very helpful.

[alvarogarciamendez]

Hello sir, thank you for this helpful video. How do you get LS DYNA program manager? I just have free access to PrePost. Thanks!!

[Akash]

Dear Topa, using the specified contact types (also the one at the end of the video), my impactor (punch) is going THROUGH the foam. Why this might happen? Thanks!