LS-DYNA TUTORIAL 12: Static and Dynamic Axial Tube Crush

[@albertsanjuan4057]

Hello Sir.. Thank you so much for all the uploaded videos. It is helping me a lot since I'm studying it on my own. I hope you can upload as well a simulation of a bolt mounted part in a structure subjected to impact. Please continue to be an inspiration to us, beginners, in LSPrepost. Cheers!

[@AmeenTopa]

Thanks for your kind words. Please share my channel among your friends. Help my channel grow (more subscribers) so I can make more videos! Regarding your question, you can check my tutorials on ball impact on a plate and springback analysis. These two will help you with impact and bolt 🔩 prestress.

[@ChanghoKim-y7l]

when quasi static analysis, you said you don't need to apply strain rate parameter . but for the accurate value is it better to apply strain rate parameter?

[@AmeenTopa]

for quasi static analysis, whether you apply the strain rate parameter or not, it won't make any difference.

[@nathanprimana9580]

hi thank you for the informative video! may I ask if it is possible to make multiple honeycombs to become a honeycomb structure inside of LS prepost in itself, since for my final thesis I am supposed to make a honeycomb structure, thus I imported from the honeycomb creator in Designmodeller, but it turned out to be individual shells that makes it difficult later in the material classification, thank you in advanced.

[@AmeenTopa]

@@nathanprimana9580 you can use the same method in this video to create a single cell. Then, use the Transform -> Translate tool to make copies of that cell. It takes some practice but if you are good at geometry, then it should not be a problem.

[@aklilualemneh9885]

Thank you for a thing what you have done, they are so interesting and i have to see each of them. by using ls dyna software i try to do some analysis regarding to human spine, have you an idea specially the contact and materials

[@AmeenTopa]

Full human body models are available online. Try to look for THUMS model provided by Toyota company. From these models, you can see what contact and what material models are used for the spine.

[@aklilualemneh9885]

Thanks so much, but how i download or see the models deeply , thank you

[@tjj4801]

I’m studying a lot about lsdyna with your video. Thank you. But I have a question. I wanted to match the results of lsdyna with the previous paper, so I tried to find some data but failed. Also, it is interpreted quasi-static and the strain rate is not considered in the physical properties, so I think the results of lsdyna should match the graph pressed at 2mm/min speed in that paper Details done to match the load data in the paper with the lsdyna results are as follows 1. The mesh size of the shell was attempted at 2 mm, 1.5 mm, 1 mm, and 0.5 mm, respectively, with a maximum load of 30 kN, which was approximately the same as the paper value, and the crumpled shape of the shell was approximately the same. However, the shape of the load data curves was somewhat different. 2. I entered the SIGY and EPSF values in MAT18. The over-transforming elements were removed from the simulation, but this result was almost similar to the method.1 3. Finally, adjusted the end time and the movement of the ridge using the gravitational conditions and the speed at which 2mm/min is pressed. And set it to a dynamic implicit newmark, not static. The shell is lightweight, so there is no effect of inertia, and it‘s a natural result, but it’s the same as number one. The simulation time is very long and inefficient I was so curious why the paper and simulation load data are different, so I tried many things, but in the paper, the physical properties of the shell are marked only with aluminum, and I don‘t know what kind of aluminum it is specifically. So I think the reason why the simulation results and the paper are different is the material properties. What do you think?

[@AmeenTopa]

@@tjj4801 You are welcome. I am glad that my videos were helpful. Regarding the validation, as long as you get the similar peak force and MCF, then that is already good enough. As for the long run time, you may apply mass scaling as well as time scaling. Time scaling means increasing the speed so the termination time is smaller. Mass scaling means increasing the mass of the system in order to get a bigger timestep size. Applying too much of either mass or time scaling might increase the inertial effects. The general rule is to look for kinetic energy. If it is less than 5% of the total energy, then we can accept the simulation as quasi static.

[@tjj4801]

@@AmeenTopa Thank you. This means that the internal energy and kinetic energy ratio can increase the density by 5%, 95%. Thank you for the good tip.

[@AmeenTopa]

@@tjj4801 Kinetic energy will increase if mass scaling or time scaling is done. Internal energy will not be effect. This can be observed from kinetic energy equation: KE = 0.5 m v^2 K.E. is related to mass and velocity.

[@tjj4801]

@@AmeenTopa I understand, thank you so much, I am learning a lot thanks to you.

[@AmeenTopa]

@@tjj4801 You are welcome. If you are required to publish your work in a journal and I can be one of the authors, then I am open to help you further with the validation process by checking your model (k file).

[@ChanghoKim-y7l]

I want to ask you something. I did the compression analysis, and the max of the value is too different when the rcforce dt is 0.05 and 0.00005 in ascii. How should I proceed in this case? The data is calculated relatively smoothly when it is 0.05 and when it is 0.00005, there seems to be a lot of data that stands out.

[@AmeenTopa]

It is hard to tell. The data with 0.00005 has oscillations and noise, which is normal. Even in real experiment there will be noise. You can apply some filters after plotting the curves to remove/reduce noise.

[@peymanrahimbakhsh9674]

Hello Ameen Topa, thank you for your helping and your answers I have another question how can I simulate a quasi-static simulation in LS Dyna? I hope you have a great time

[@AmeenTopa]

You can refer to my video on Compression Test to see how to do this

[@peymanrahimbakhsh9674]

@@AmeenTopa thanks

[@bellal1329]

hi great video! just wondering, what's the default unit in ls- prepost for velocity? in the reference paper, it says -10mm/min. but i assume this is not the unit here?

[@AmeenTopa]

You can search "LS-DYNA Unit Consistency". There is a good explanation on the dynasupport website.

[@bellal1329]

thank you for the response! but for example the tube geometry is 38mm*38mm, how can you know that the default unit is mm, so that you just input 18.4 and -18.4? Besides that, for the implicit method, why we don't need to define velocity? Will the velocity not affect the deformation? THANKS!@@AmeenTopa

[@AmeenTopa]

@@bellal1329 In LS-DYNA, you need to follow a set of unit. For example, we have the standard international units (Newton, meters, seconds, Kg). Also, we have another unit system (kg, mm, millisec, kN). So, it is your task to make sure that your input follows one of the unit systems. For more, read here: www.dynasupport.com/howtos/general/consistent-units For the implicit simulation, it is a quasi static test, therefore, we do not define INITIAL VELOCITY. We use PRESCRIBED MOTION to define the displacement or velocity of the top plate.

[@bellal1329]

@@AmeenTopa Thank you so much for the clarification! one more question, in which video did you talk about using material properties edited in the textbox?

[@AmeenTopa]

@@bellal1329 yes, you can edit the keyword files with a text editor like Notepad or Wordpad.

[@윤재광-x4p]

Thank you for the great video! Could you please share the file for this example??

[@AmeenTopa]

Sorry. This video is from four years ago and I don't have the file anymore.

[@윤재광-x4p]

@@AmeenTopa I'm sorry to hear that... Could you recommend an example related to MAT_JOHNSON_HOLMQUIST_CERAMICS?

[@gggisbatman]

Hello thank you very much for the tutorial it helped with school project. Can you tell me if it is possible to remesh a box shell along y axis (length)?

[@AmeenTopa]

Hi there. You are welcome. Please share my video if you find it useful. Regarding your questions, yes it is possible. Check out my playlist: MESH IT WITH PREPOST. You will find a lot of useful information there.

[@alparslansolak6115]

You light my way Ameen Topa, Thank you :)

[@AmeenTopa]

Thanks for your kind words. Please share my channel among your friends. Help my channel grow (more subscribers) so I can make more videos!

[@ChanghoKim-y7l]

When I thought about it, wouldn't the buckling mode come out well and give an appropriate critical load result compared to the theoretical formula? (fixed-fixed Euler equation)

[@AmeenTopa]

@@ChanghoKim-y7l Euler equation is for static buckling. In this video, the load is quasi static and dynamic, so the peak load will be different from the theoretical value of Euler's formula.

[@ChanghoKim-y7l]

@@AmeenTopa When I looked at your paper, I found that the initial buckling was formed at about 30000N. Is it unreasonable to derive a critical buckling load by applying the corresponding peak force value, even if it is quasi-static?>

[@ChanghoKim-y7l]

@@AmeenTopa Then, if you want to find the value by applying the theoretical value, is there any other way? It's just for validation.

[@AmeenTopa]

@@ChanghoKim-y7l may I know which paper you are referring to? Theoretically, the peak force from quasi static analysis would be lower than the value from Euler's formula.

[@ChanghoKim-y7l]

@@AmeenTopa a study on the axial crush analysis of a rectangular tube with experimental comparison In the equation, (w+h) x 2 x t x sigma_yield

[@yugeshwaransubramaniam476]

Hye. I have one question regarding the ls-dyna axis. I’ve chose different axis for the simulation. So, may know how to fix the axis permanently. Thanks in advance.

[@AmeenTopa]

Sorry I don't understand exactly what you are referring to. Are you saying you want to use a user defined axis?

[@edeplessnesss]

Hi Ameen, i want to ask a question. If we would use composite material instead of aluminum, which material model would suitable instead of power law plasticity material model?

[@AmeenTopa]

Two good candidates are MAT022 composite damage and MAT054 enhanced composite damage. These two are widely used therefore it will be easy to find the input parameters from literature.

[@edeplessnesss]

@@AmeenTopa i tried to Mat54 model for aramid epoxy composites by implicit solver, but i got to convergence error. problem did not converge.

[@aminmohammadigolafshani2015]

Hello Mr. Topa, I have a question. how can we assign our square tube thickness? I wanna simulate the crash for square tubes with different widths and thicknesses. How can I do that?

[@AmeenTopa]

SECTION SHELL is where you define the thickness of the part.

[@aimrinnuqman4298]

Hello Mr Topa. I have a question here. Some research paper mentioning that they use Belytschko-Tsay shell element with one centre integration point and five through-thickness integration point. I understand that through-thickness number can be change, but what does it mean by 'one centre integration point' ?. Currently I am trying to simulate dynamic axial tube crashing

[@AmeenTopa]

You can check the presentation below: www.dynamore.de/de/download/papers/2013-ls-dyna-forum/documents/review-of-shell-element-formulations-in-ls-dyna-properties-limits-advantages-disadvantages

[@haidermraih2559]

Thanks, Ameen. You assigned 47 mm as the displacement. I want to ask how we can manage the load rate. I need the plate to move with 0.05 mm/sec, for example, all the way to 47 mm. How can I do it using the load or displacement, not velocity? And is there any relationship to the termination time of the model?

[@haidermraih2559]

I think when we defined the curve, we said when the time equals 1, the displacement equals 47 mm. This means that the termination time governs the loading rate. For instance, if the termination time is 2 sec, the load rate is 47 mm/2 sec, which is 23.5 mm/sec.

[@AmeenTopa]

@@haidermraih2559 In DEFINE CURVE, we put can put displacement vs time. So, for example, if I put 0,0 and 10, 50 then it means at time =0, the displacement is zero. And at time = 10 sec, the displacement is 50 mm. The load rate or velocity is the differentiation of this curve. So, it will be (50-0)/(10-10) = 5 mm/sec. Remember physics: velocity is the differentiation of displacement.

[@haidermraih2559]

@AmeenTopa understand, would that govern the termination time of the model? Let assume that I have 50 mm thick block, and I need to apply the load with 0.5 mm/ sec, which results in 10^-3 s^-1 as a strain rate. What would be the termination time of the model? Assume that the displacement curve is 0, 0 1, 50

[@haidermraih2559]

I think it can be managed from the load curve itself, and it is not related to termination time.

[@AmeenTopa]

@@haidermraih2559 Yes, it is related only to the curve used, as I explained in my previous comment. However, take note that if you change the load rate, you might want to change the termination time too. For example, if you made the load rate 0.5 x the original speed, then you need double the termination time in order to reach the same displacement.

[@bhavanrana1123]

Can you do compression and bending test on Railway sleeper tutorial please

[@AmeenTopa]

You can check my tensile test tutorial and bending tutorial. That should be enough for you to get the basics.

[@mayal-rawi6906]

Thanks a lot brother it really helps

[@AmeenTopa]

You are welcome. Please share my videos so more people can benefit. 😬

[@vladimirlubimov2318]

Hello, Ameen! Please tell me, I use MAT 83, the properties of which depend on the deformation rate. But for some reason, the implict analysis does not take into account the speed.. it always counts according to one curve, the very first, regardless of what loading speed. Is it possible to fix it or is it the features of implicit analysis? Thank you very much.

[@AmeenTopa]

Hi there, it's you again! :D Make sure you are using implicit dynamic analysis not implicit statics.

[@vladimirlubimov2318]

@@AmeenTopa I apologize for the annoyance....) I always enjoy watching your videos and it's only thanks to this that I started doing my mini projects! I am very grateful to you for your advice. That is, the implicit analysis should take into account the influence of speed on the behavior of the speed-dependent material, but you need to specify an implicit dynamic analysis?

[@AmeenTopa]

@@vladimirlubimov2318 It's fine. You can always ask here if you have any questions. *CONTOL IMPLICIT DYNAMICS. You can read more about this keyword in the user manual vol. 1.

[@vladimirlubimov2318]

@@AmeenTopa Hello, can you provide email contacts by which you can be contacted and discuss the issue of assistance in person?

[@AmeenTopa]

@@vladimirlubimov2318 check the following link: sites.google.com/view/topacademyofficial/about-topacademy

[@mtaha7476]

sir, even tough ı have selected nodes before the analysis and adding nodfor and nodout option ın database, my nodfor analysis file is empty(1kb). Do you have any idea why it happened? The error: "File has no information to be processed!" I can see other results just ı cant see nodfor.

[@AmeenTopa]

I think you need to define DATABASE HISTORY NODE SET. Please check the manual for confirmation.

[@lalinlamCambodia]

Hi Ameen, I have some question related to mesh. After i export the mesh (hex8) from hypermesh to prepost, the element in prepost is solid with 10 node format. I’m not sure why it is not a 8 node solid? And can we change it to 8 nodes solid or anything wrong with my export from hypermesh? Thanks.

[@AmeenTopa]

Hi Lam, I am not sure about this problem. But in general, Element formulation type can be changed in SECTION SOLID ELFORM

[@sharanagoudapatil4038]

Can you say? How can I import experimental stress-strain values to the Model?

[@AmeenTopa]

How to import the stress strain curve? I have covered that in a previous video. You can check blast simulation video titled "LS-DYNA Tutorial: Explosive and Blastwall"

[@peymanrahimbakhsh9674]

Hello Ameen Topa, I have question about this simulation, what if I want to fill this tube with foam? , what parameters change and what parameters I should add to the keywords? thank you for your helps

[@AmeenTopa]

Dear Peyman, You need to include solid elements for the foam part. Watch my tutorial on three point bending where I used EPS foam as the core: kzbin.info/www/bejne/aoezgol5oJp5aaM For more information about the foam modelling, you can refer to published paper here: www.hindawi.com/journals/mse/2014/292647/ www.jstage.jst.go.jp/article/mej/2/5/2_14-00510/_article/-char/ja/ Sincerely,

[@liuyi7742]

Hello, Ameen. Thanks a lot for sharing. I have a question. How to delete the element when it is failed? For example in the contact with adhesive. When the structure bears a large force, the adhesive layer will peel off at the connection. How to achieve this in the program?

[@AmeenTopa]

Check my tutorial: LS-DYNA TUTORIAL 14: Delamination Test and Cohesive Elements

[@Prisonbreak910]

Hello, I have a question for you. If i use contact suface to suface for implicit solve so my result is true ?

[@AmeenTopa]

It is better to use AUTOMATIC SURFACE TO SURFACE MORTAR for implicit analysis as it improves the stability. If you didn't use the MORTAR option, then you might not reach convergence in the analysis.

[@Prisonbreak910]

@@AmeenTopa so CONTACT SUFACE TO SUFACE only use for explicit analysis ?

[@AmeenTopa]

@@Prisonbreak910 For more about contacts, you can check this webinar: kzbin.info/www/bejne/oGjPdnybjrGBf7s

[@vladimirlubimov2318]

Good afternoon, Ameen! Can you tell me if there are any general recommendations for choosing a time step for static implicit analysis. Well, let's say 1 to 100 of the total time, for example, if you need 1 second, then the step is 0.01 or 1;1000? I noticed that this significantly affects the result..to calculate, say, the same load of one object on another under its own weight under the influence of gravity? and which scheme is optimal for typical similar tasks, when one body loads the second under its own weight and gravitational load? thank you in advance!!

[@AmeenTopa]

If you are using static implicit analysis, I recommend to use IAUTO = 1 to invoke the variable time step. Since you are using static implicit, then I assume it is a quasi static load. Therefore, you can put the termination time = 1 and initial time step as 0.01. The load should be applied and starts from 0 and reaches its maximum value when t = 1 (which is termination time).

[@vladimirlubimov2318]

@@AmeenTopa I understand everything. thank you very much for the answer. And what happens if, at completion time 1, I set the initial step to 0.1 or 0.001, instead of 0.01? how will this affect the result theoretically?

[@nj5587]

Hello Ameen, I am working on a similar model but my rigid plate seems to be moving axially into and through the surface of aluminium rather than crushing it, what can be the cause of this?

[@AmeenTopa]

Hi N J, you may want to check again your Contacts. Maybe it was not defined properly.

[@anaychk590]

Excellent tutorial. Thanks very much.

[@AmeenTopa]

My pleasure! please support my channel grow (more subscribers) so I can post more free tutorials!!!

[@김창호기계공학과]

Why don't you set up a control_contact card that reflects the shell thickness?

[@AmeenTopa]

In the contact types that I used, the shell thickness is calculated by default settings.

[@김창호기계공학과]

@@AmeenTopa thank you

[@mulukenmeaza4844]

Very informative video, thanks! Can you explan me how do you verify the mesh element order( linear or quadratic)? Is it possible to modify? Thanks!

[@AmeenTopa]

Element order is selected from SECTION keyword.

[@mulukenmeaza4844]

@@AmeenTopa Thank you very much for prompt reply. Have you define the element order in the tutorial? Am I missing something here? Thanks!

[@AmeenTopa]

In SECTION SHELL. ELFORM is the element formulation. There are many options under there.

[@mulukenmeaza4844]

@@AmeenTopa Thank you very much once again. I have the following questions: What are the option I have to change if I want to use force instead of displacement? What would be the default analysis if we don't add or specify either implicit or explicit key options? Lastly, I would appreciate if you could me tips for bolt pretention, how to use bonded contact (bolt and nut) and frictional contact (bolt and clamp). Thank you in advance, appreciate your help.

[@AmeenTopa]

​@@mulukenmeaza4844 You are welcome. To use force instead of displacement, you can use LOAD NODE SET. Check out my tutorial: Compression Cylinder. If we don't add any keywords, the default is explicit time integration. For bonding contact, check out my tutorials: (a) 3 Point Bending of Sandwich Structure (b) Delamination Test and Cohesive Elements

[@carpi20]

Hi Amen thank you for this awesome tutorial. I have two question , is it normal to get huge result files in second analysis. For hexagonal shape, explicit analysis whole results are 22gb. I didn't get that much bigger result before. ( Termination=10, d3plot : 0.001). And my second question is ; in my analysis, the swelling in the box which is a result of folding is much less than yours. I got similar graphs but in my opinion it is not matching with experimental results like the page you refered. I checked all but there is only one different than yours: My ELFORM is not same with yours. My elform is different because of my pre post version (v4.6.17). What can I do for getting results similar to experimentals. Sorry for that long questions, but you are my savior :) Thank you again for this great videos! Yours sincerely.

[@AmeenTopa]

Hi there, first of all. We can discuss through email for more convenient communication. Put your email address here and I will follow up. First, yes, the results will be big even if your k file is small. With those settings, you will have 10000 frames of results which is too much. Secondly, you can use different formulation. Just download the latest LS-Prepost. Different ELFORM will give different results. Matching experimental work or results of others is not an easy task to do. But it can be done. I need to know more details to be able to give a useful answer.

[@srihari74]

Thanks for sharing this video. Very informative

[@AmeenTopa]

Glad it was helpful! Please support my channel grow (= more subscribers) so I can post tutorial videos more frequently hehe :D

[@sharanagoudapatil4038]

Excellent!!!! Thank you so much brother it helped me a lot!!!

[@AmeenTopa]

You are welcome. Please share my videos so more people can learn! How to import the stress strain curve? I have covered that in a previous video. You can check blast simulation video titled "LS-DYNA Tutorial: Explosive and Blastwall"

[@peymanrahimbakhsh9674]

Hello Ameen, I have a question, how can I calculate peak crushing force in this tutorial? thank you for these excellent videos

[@AmeenTopa]

Hello there, as you saw in the video, we plot of the force vs time graph. From there, you can easily see the maximum force.

[@danielhambissa5357]

Why don't you make some videos on types of constraints on ls dyna with their dominstration?...is there is any book you can recommend me for this?

[@AmeenTopa]

Thanks for the suggestion. But there are some reading materials and examples online. That's why I don't have plans to make any. www.dynasupport.com/tutorial/ls-dyna-users-guide/constraints-and-spotwelds

[@peymanrahimbakhsh9674]

Hello, what parameters peak crushing force depends on? thank you

[@AmeenTopa]

Here there, it depends on many factor such the the cross section type, the thickness of the tube, and the material properties...etc...

[@AmeenTopa]

@Vedant Deshmukh You will know the answer if you watch the video.

[@Ankit-df8rw]

Sir thanks for such an incredibly good videos. But i have encountered a problem....the solver says i need double precision version for the implicit method. Kindly help!!

[@AmeenTopa]

Yes you need to change the solver from single precision to double precision. I don't know what system and what version you are using. So, you need to contact the technician who installed the software to do that.

[@Ankit-df8rw]

@@AmeenTopa actually its a cracked version R11 and do i have to download a double precision solver??

[@AmeenTopa]

@@Ankit-df8rw if it is a crack version then I don't think I can help you because I never used that.

[@Ankit-df8rw]

@@AmeenTopa 1 more thing sir i wanted to ask..is it possible to perform implicit analysis in single precision version of LS-DYNA??

[@AmeenTopa]

Yes possible for simple analysis.

[@lalinlamCambodia]

Hi Ameen, I have followed your model but I am not sure why i cannot get the rcforc result as yours. My rcforc is zero (it’s like there is no force). Any ideas on this. I tried to do it twice but still cannot get the result as yours.

[@AmeenTopa]

Dear Lam, are you putting correct value for DT in ASCII OPTIONS?

[@AmeenTopa]

Also, the tube is crushed or not in your simulation?

[@lalinlamCambodia]

@@AmeenTopa Hi, the tube is crushed and the deformation is look like yours. I got the same result in implicit. But I cannot get the result for explicit of both square and hexagon tubes. The DT that I put is 0.001 as yours. I tried to checked many times whether which part that I set it wrong. Is it possible if you can help me check the k.file. My email address is [e-mail]. Thanks for your help and time.

[@김동현-f9t6w]

Hello Ameen Topa Thank you very much for your videos I'm leaving a comment because there's a difficult part while watching the video. When I run Animate, my last number is 1 instead of 101, so it doesn't run. Do you happen to know the cause? Thanks

[@AmeenTopa]

Did you put the termination time? That's weird. Everyone else tried the tutorial and they didn't get the problem. There might be some steps that you missed in my video.

[@김동현-f9t6w]

@@AmeenTopa At 13:47, I don't see the numbers on my screen, and does "unknown" appear to be the cause of the error?

[@AmeenTopa]

@@김동현-f9t6w There is something wrong. I am not sure what it is. I don't have enough information to give you a useful answer.

[@김동현-f9t6w]

@@AmeenTopa Thank you for your answer.

[@edeplessnesss]

Ameen you are wonderful! If possible, can you prepare a video about delamination of composites by the cohesive zone modelling ?

[@AmeenTopa]

Yes, I have that in my to-do list....has to wait for its turn haha.

[@AmeenTopa]

Check out my latest videos.

[@pradeeppal787]

giving error when running on MPP. CONTROL_SOLVER commnad 0.85 plz help

[@AmeenTopa]

It seems you have a problem with the solver. Try to use double precision smp solver.

[@PiloteBoeing747]

Hi Ameen, thank you for your tutorial. I did a simulation with a beam which is crashing into a sheet of metal (the opposite of you)... I have this error : "Discrete nodal inertia tensor must be positive definite. Possible zeros on diagonal". Do you know what it could be? Thank you :)

[@AmeenTopa]

I am not sure as I have not ever faced similar problem before. It might has something to do with the beam elements. Make sure you have defined their cross section properly.

[@lalinlamCambodia]

Hi Ameen, I have some questions which may not related to this video. In case that the element is Lagrangian or Eulerian elements, from where we set Lagrangian or Eulerian to the element?

[@AmeenTopa]

This is controlled from ELFORM in SECTION keyword. In this video, I used Lagrangian elements.

[@lalinlamCambodia]

Ameen Topa thank you so much.

[@lalinlamCambodia]

Hi Ameen, I have some more questions. 1. in my case, the load is a pressure-time history on the rigid plate, do I need to define the segment set on the rigid plate and then apply load on the segm set, am I right or wrong? 2. For contact, we need to create every possible contacts which may occur during the crushing right?

[@AmeenTopa]

@@lalinlamCambodia If it is a RIGID plate, then just use LOAD BODY RIGID

[@ChanghoKim-y7l]

why didn't use compression material property

[@AmeenTopa]

@@ChanghoKim-y7l for metals, the tensile and compressive properties are the same except for high strains.

[@ChanghoKim-y7l]

@@AmeenTopa i got it.. thank you then i want to steel compression test(quasi -static) for using lsdyna. then i can use johnson_cook or MAT_PIECEWISE_LINEAR_PLASTICITY right? but i don't know what is the best mat card for the steel compression simulation can you give me some advice?

[@AmeenTopa]

@@ChanghoKim-y7l both are fine. You can just use MAT024. MAT015 is used for high strain rate simulations with damage involved.

[@goktugyilmaz1504]

Dear Sir, I am very excited to see your tutorial. Thank you very much for your explanations. I really appreciate your effort! But, I want to ask a question. My purpose is to do a quasi-static simulation of the same profile. However, I really do not understand how you simulated 100mm/min press velocity. You did not enter any value for the velocity of the rigid body in quasi-static analysis. But how could we make sure that we are doing quasi static analysis with 100mm/min press velocity. In the article, there are different speeds for the press.

[@AmeenTopa]

Dear Yilmaz, the quasi static analysis is done using implicit solver. When using implicit, physical time is irrelevant.

[@karlsruhestolz69]

Just a question, I am trying to do the same but with an aluminium honeycomb model and I got this message " *** Error 70021 (OTH+21) Memory is set 598383462 words short memory size 900000000 Increase the memory size by one of the following where #### is the number of words requested: 1) On the command line set - memory=#### 2) In the input file define memory with *KEYWORD i.e., *KEYWORD #### or *KEYWORD memory=#### What can I do?

[@AmeenTopa]

This means you need to allocate more memory to solve the simulation. You can put MEMORY = 1000m like what I do my most of my videos.

[@karlsruhestolz69]

@@AmeenTopa thanks, but where is this option?

[@AmeenTopa]

You can when I open the LS DYNA manager, before the I run the file. There is a space in front of MEMORY.

[@AmeenTopa]

Check the video at: 13:30

[@ashutoshmaurya915]

Hello Sir, I learned very much from your video...Thanks for uploading the video. I need a little help in modelling the Collapse behavior of the building in LSdyna. Please reply As soon as Possible.

[@AmeenTopa]

You are welcome. What is your problem with the simulations?

[@ashutoshmaurya915]

@@AmeenTopa Thanks A lot for your Reply. I was trying the simulation with the mat erosion, but it simply erode the material which crosses the defined failure limit, But I want the actual collapse behaviour of the building. Could you please tell me a way, how to proceed for this, if possible please give me alternate way for contact to you, my email id is - [e-mail]

[@AmeenTopa]

​@@ashutoshmaurya915 It could be a problem with the failure criteria.

[@dacdungtruong6986]

Dear Dr. Topa, Thank you for your video, it is very informative and helpful. Could you kindly make a video on ice-structure crushing, like a ice block hitting a rigid plate? Thank you in advance.

[@AmeenTopa]

Dear Dac Dung, You are welcome. The simulation of ice crushing is no different than this tutorial or my first tutorial. The only difference is the material properties which can be found in the literature. Here, my first tutorial: kzbin.info/www/bejne/aZLChWp-nZZsnbM Here, some ice-simulation related papers: www.dynalook.com/conferences/8th-european-ls-dyna-conference/session-3/Session3_Paper3.pdf www.dynalook.com/conferences/13th-international-ls-dyna-conference/constitutive-modeling/simulation-of-compressive-2018cone-shaped2019-ice-specimen-experiments-using-ls-dyna-r ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20060022142.pdf www.sciencedirect.com/science/article/pii/S0020768310004154 www.researchgate.net/publication/237465173_A_High_Strain_Rate_Model_with_Failure_for_Ice_in_LS-DYNA

[@dacdungtruong6986]

@@AmeenTopa Thank you very much for your kind support. I understand that there are several materials available in LS-Dyna such as MAT_13, MAT_63 and MAT_153. However, those still have some limitations in defining ice behavior more realistically. Therefore, using MAT_41-50 (user-defined material) may be more preferred. I would be very grateful if you could share a simple user subroutine (e.g. elastic plastic material), and how to apply this as well. Thank you in advance.

[@AmeenTopa]

Dear Dac Dung, I have no tutorials for UMAT subroutine. I have no deep knowledge in User Defined material and I don't feel comfortable putting a tutorial on something that is not in my field expertise. Anyway, for modelling ice with user defined material, you can check this paper: core.ac.uk/display/29409769 More on user defined subroutine: www.dynamore.de/de/frontpage-de/dokumente/eyecatcher/research-and-development/simulation-2-1.pdf www.researchgate.net/publication/327623424_How_To_-_user_defined_material_models_with_LS-Dyna_on_Windows kzbin.info/www/bejne/gKLafWqpbaulfZY

[@dacdungtruong6986]

@@AmeenTopa Thank you so much for the kind reply and good references. These sources are really helpful to me. Again thank you for your effort and supports.

[@AmeenTopa]

You are welcome. Sorry I couldn't help more. UMAT is not my specialty. But if you have any other doubts, you can ask. I will answer if I know. :D

[@chaimongkolpadungtree2878]

excuse me sir,i got some warning and i dont know how to fix it .I got Warning 40571 (SOL+571) Initialization - node will not be moved penetrating node near or at shell mid-surface, Warning 30131 (INI+131) total number of massless nodes = 1,and Warning 30128 (INI+128) Massless nodes I design this model in soldiwrk and mesh in hyperwork then import to ls-dyna.These is different from this video.I dont know that involve about my warning.I hope you will reply my comment.Thank you very much

[@AmeenTopa]

Hi, a common mistake when importing geometry from solidwork or other CAD software is that you make the geometry as solid parts, but when you mesh it, you mesh it as shell. Massless nodes can be removed by: Element Tools, Node Edit, Delete Nodes, Delete unreferenced Element Nodes.

[@chaimongkolpadungtree2878]

@@AmeenTopa According to your suggestion,My model can run but impactor runs through crashbox which doesn't appear collapse of crashbox. Is it possible this warnings involve with my result?

[@chaimongkolpadungtree2878]

@@AmeenTopa Warning 40571 (SOL+571) Initialization - node will not be moved penetrating node near or at shell mid-surface single surface interface # = 2 penetrating node # = 5534 segment # = 2917 penetration = 9.6500E-01 remaining distance to shell midsurface = 0.0000E+00 s,t = 1.0000E+00 1.0000E+00 and i have already seen this warning that is involve to my animation.

[@AmeenTopa]

yes, it is highly possible that this penetration is due to the meshing. That's why, if possible, I prefer to create my parts from scratch using LS-Prepost

[@chaimongkolpadungtree2878]

@@AmeenTopa What is unit of force in dynamic case?

[@dequan-g2z]

Excellent video! Sir, I got a problem : why is it that my initial speed setting is in the negative Z direction, but the rigid plate is still moving in the positive Z axis? I adopted the key word *INITIAL_VELOCITY_GENERATION😂

[@AmeenTopa]

@@dequan-g2z it could be that it is moving correctly. It's just that the target is so stiff so the plate bounces back.

[@dequan-g2z]

@@AmeenTopa But when I used*BOUNDARY_PRESCRIBED_MOTION_RIGID，the plate can move correctly.

[@AmeenTopa]

@@dequan-g2z BOUNDARY PRESCRIBED MOTION is like a compression test. No matter how strong the material is, it will get crushed. INITIAL VELOCITY GENERATION is like a drop test, if the target material is strong, the impactor will bounce back. To prove this to you, in your model with the initial velocity, check the stresses in the target. You will find there are stresses. This means that the impactor has already hit the target. Because if it was moving upward, it won't hit the target and there will be no stresses.

[@dequan-g2z]

@@AmeenTopa Thank you so much, sir! Which means the keyword *BOUNDARY_PRESCRIBED_MOTION is more suitable to simulate the compression test in the universal testing machine?

[@AmeenTopa]

@@dequan-g2z Yes, correct. I have a tutorial on tensile test and compression test. Check those.

[@zurameen]

Link to download LS-DYNA: getintopc.com/softwares/simulators/ls-dyna-free-download/

[@azimdaniel222]

Hello Ameen, can I get your email? I have a lot of questions regarding this ls-dyna software. I have been asked to use this software for my final year project. Thankyou in advance :)

[@AmeenTopa]

Hi there. Put your email here and I will follow up in my free time.

[@azimdaniel222]

@@AmeenTopa [e-mail] , thankyou :)

[@azimdaniel222]

@@AmeenTopa i have sent to your email

[@Kiran58399]

Can I use the same procedure in the video for crash box(in shape of cone/square with some thickness)if not,what are the changes to be made in the procedure sir!!???

[@AmeenTopa]

Yes you can.

[@Kiran58399]

@@AmeenTopa ok sir thank you.suppose if that cube section should to be made of composite..is it enough to change the material cards and follow the same procedure!?or any changes to be made sir!!?

[@AmeenTopa]

@@Kiran58399 yes only changing the material properties. To control the no of Laminates and their direction, you need to modify the parameters in SECTION SHELL.

[@Kiran58399]

@@AmeenTopa thank you so much sir, I'll let you know if it works.