LS-DYNA TUTORIAL 15: Shaft Torsion with Mass scaling, Ductile and Brittle Failure

[@yonca2538]

A very informative video, thank you! But why didn't you choose to do an implicit analysis?

[@AmeenTopa]

Very good questions. With Implicit analysis, everything needs to be fixed and have no loose ends. I could use it, but up until the point of fracture only. Once an element is deleted, then it would be difficult to reach convergence (as you can see in the vibration I mentioned in the energy graph)

[@sriharikurukuri5458]

Excellent work! Thanks for sharing...

[@AmeenTopa]

Thank you! I'm glad it's helpful! Please help my channel grow ( = more subscribers) so I can post tutorials more frequently! Feel free to share my channel hehe :D

[@anaychk590]

Thanks for sharing your knowledge.

[@AmeenTopa]

Glad it was helpful!

[@swapnilmane5759]

can a similar setup be used for a compression test of the granite specimen, where instead of twist, displacement is given? if not, what other changes will be required?

[@AmeenTopa]

yes, I have a tutorial con compression test. You can check it out.

[@abrahammorenoreyna2758]

Great job!! Thanks!

[@AmeenTopa]

You are welcome. Please support my channel grow (= more subscribers) so I can post videos more frequently! :)

[@amindoxiti1201]

Thank you, please make more video tutorials for civil engineers.

[@AmeenTopa]

You are welcome. Yeah, I have few more tutorialsl ideas in my for Civil Engineers.

[@amindoxiti1201]

@@AmeenTopa Good news, thank you so much.

[@roma1qwerty]

Thanks for the video. I checked D1-D5 values in the sources listed below the video, but they are different there. Maybe I don't know where to look? I would appreciate if you answered my question.

[@AmeenTopa]

You are welcome. You need to search in the literature for these values. Take note that my tutorials are only for learning purposes and values might be fictitious.

[@roma1qwerty]

Thank you for the answer!@@AmeenTopa

[@roma1qwerty]

I've done everything and have a problem. The Rotating end is rotating but only first row of elements of Cylinder is affected.

[@AmeenTopa]

Most probably you are rotating too fast. Check out the rotational velocity.

[@roma1qwerty]

@@AmeenTopa Thank you so much. You were right, your advice helped.

[@hamzagungor3029]

Hi sir, in this video, why didnt you make any contact between middle part and end parts ? is making one section meaning they are connected each other ?

[@AmeenTopa]

There are connected to each other because their nodes are shared at the connection zones. And the nodes are shared because of the way I created this model. Initially, it was one single part then I used MoveCopy to change the part ID. Using this tool will not disconnect the mesh. It only changes the ID of the part.

[@hamzagungor3029]

@@AmeenTopa you are great. Please increase number of videos 🙏🙏🙏

[@AmeenTopa]

@@hamzagungor3029 Finish all of my videos on the playlist LS-DYNA Basic Tutorials and I believe you would be able to do the simulations you want to do.

[@김경민-e5o]

Thanks for your tutorial sir. I am now trying to use johnson-cook model in tensile test, and I have one question regarding it. How can I define the 'PC' in the material card, which is the tensile failure stress?

[@AmeenTopa]

You set PC to a very huge negative number because you want the failure (erosion) of the elements to occur in the specimen due to tension.

[@김경민-e5o]

@@AmeenTopa Thank you very much sir! I will try as you advised me

[@AmeenTopa]

@@김경민-e5o You are very welcome.

[@김경민-e5o]

@@AmeenTopa I have another question. I've done tensile test simulation, but element is not deleted during the test. Is there any way to make element deletion occur? And if you don't mind, can I send my .k file to you ? Thanks in advance sir!

[@AmeenTopa]

@@김경민-e5o it could be that 1- the damage criteria is not defined properly. 2- the cutoff pressure value is not big enough 3- the deformation is still low. With regard to sharing your model, there will be some consultancy fees. If that's fine with you, please contact my team at [e-mail]

[@caesolutions9625]

Very helpful tutorial Ameen. I appreciate the time you give. Why did you define the C2/P parameter in JC material model? Since parameter VP = 0, isn't this redundant? Once more, thank you and keep up! Theo

[@AmeenTopa]

Hi there, It has been a mistake as I was more focusing on inserting the damage parameters in this tutorial. But yes, you are correct. C2/P parameters are only effective when VP = 1. And here is where I knew that: www.dynalook.com/european-conf-2007/optional-strain-rate-forms-for-the-johnson-cook.pdf Anyway, since the load rate is low in this tutorial, I would expect that it won't make much difference. You can try out with VP = 1 and let me know if you get interesting results :D

[@shubham22kar]

hi, nice tutorial & very informative, but I am getting the rupture at rotating end, not in the middle for ductile. The rotating end gets seperated from shaft, Please suggest solution for that. Thank you

[@AmeenTopa]

Decrease the speed of rotation, that should work. If the above still didn't work, then you may want to use a refined mesh.

[@tilaksah6577]

Hello Ameen sir, the tutorial is very informative. I have a doubt in my model using mass scaling. As in ur video u just increased the density, how can we extract the data of how much mass is added in the whole/part model and how can we verify that the added mass is within the limit and do not effect our force results. Thank you!

[@AmeenTopa]

Hi there, Regarding this method of mass scaling: 1-You can measure the total mass of the system before and after increasing the density 2-Regarding whether it will effect the force or not, as a rule of them you measure the energy, and check if the kinetic energy is within 10% of the total energy of the system as I mentioned in the video. Of course, to check this, you can run the model without mass scaling and compare the results.

[@tilaksah6577]

Thank you for your reply sir! I will do n check that. Well after Mass scaling kinetic energy is below 5% but without mass scaling the duration of analysis shows more than 3600hrs. Is there any other way to check without mass scaling?

[@AmeenTopa]

If that's the case, then I think the results are still acceptable. You can convert to implicit analysis if you want to check without mass scaling. But you might get difficulties to reach convergence if your model is complex.

[@tilaksah6577]

Thank you sir! yes, implicit analysis will be difficult for my model, as of now my results are pretty much acceptable. I will try improving it.

[@akshaychavan1537]

where are you getting the data related to materials, values like D1, D2, D3 and etc.?

[@AmeenTopa]

Please read the description of my videos. It is literally listed there!

[@rosariapalumbo5789]

hi ameen thanks for your tutorial, you talked about another way to do mass scaling by setting dt2ms, but how can i do it?

[@AmeenTopa]

As far as I know, there is no way to do. For mass scaling, you can check my tutorial on car collision.

[@nj5587]

what are the numerical effects of mass scaling on the solution? will it have effect on strains?

[@AmeenTopa]

Yes it will effect. Mass scaling means increasing the mass. When you increase the mass, you will increase the kinetic energy (1/2 m * v^2) as well as the forces ( m * a ). However, if the amount of mass of scaling is not much, then the results won't be different so much. But there will be a different.

[@jacobmeza9743]

Hi Ameen Topa! It was an interesting video! I like the way that you induce the rotational movement. However, I was wondering If there was an alternative to applying a torsional moment to a component made of solid elements. I know they only allow translational movement.

[@AmeenTopa]

Hi there. You can try LOAD NODE POINT with DOF 4 or 8. Look it up to see how it works.

[@jacobmeza9743]

@@AmeenTopa It is a bit late but you were correct. I use LOAD NODE POINT with CNRB and then I released the rotation about the axis of interest for applying a torsional moment to the shaft.

[@AmeenTopa]

Glad it worked.

[@isurudias978]

thanks for the video, how do we calculate D1 - D5 values ?

[@AmeenTopa]

These are the damage parameters. You can refer to the papers in the video caption for further reading.

[@keoagilekerileng663]

Very powerful

[@AmeenTopa]

Thank you. Please share my videos if you find them useful! :D

[@diegoarmandogomezmarquez879]

Hello Ameen, nice video. I have a doubt: in a thermo-mechanical model, how do we modify the thermal properties, such as specific heat, to reflect the increase in the mass as a consequence of changes in density?

[@AmeenTopa]

For thermal simulations, you need to define thermal material models. The most basic one is MAT_Thermal _Isotropic in which you need to define the thermal conductivity and heat capacity of the material.

[@diegoarmandogomezmarquez879]

@@AmeenTopa yeah, but, should we modify the properties because there is more mass now?

[@AmeenTopa]

@@diegoarmandogomezmarquez879 How does the mass increase suddenly? If you heat up your specimen, it will expand, then melt down. The mass will be the same but only the volume will change.

[@AmeenTopa]

@@diegoarmandogomezmarquez879 The density of the material will change. But you do not need to worry about this. Because you define specific heat capacity which is a quantity that is per unit mass. If you read about heat capacity, you will find the following: Specific heat capacity is an intensive property of a substance, an intrinsic characteristic that does not depend on the size or shape of the amount in consideration.

[@diegoarmandogomezmarquez879]

@@AmeenTopa That was my doubt; I thought we needed to scale the thermal properties. Thanks.

[@nik2903]

Thanks alot

[@AmeenTopa]

Most welcome

[@malikrabnawaz427]

Thanks for the Video … I am modelling Special type of Concrete called as strain-hardening cementitious material (SHCC) under Torsion. I am using mater model O72R3 Damage material model. When i changed a certain Parameter below a certain value my System didnot converge. I have changed the time step size and also i have chnaged the displacement and Energy limits in implicit solution. But i am not able to solve.. I am stuck for atleast a week . Please can you guid me in solving my Problem. If i have your email, i can send you my K file. I would be very grateful to you. Kind regards Rab nawaz

[@AmeenTopa]

Hi Malik, based on my experience, MAT072R3 gives better results using Explicit analysis. Due to some reason, while modelling with implicit, the results seems to diverge from experimental data. If by "system cannot converge", you mean in implicit analysis. This is due to the failure of the concrete. When failure occur, there will be huge deformation which cannot be modeled with Implicit solver due to the instabilities. Drop your email here and I will follow up. Sincerley, Ameen

[@malikrabnawaz427]

@@AmeenTopa Thanks for a quick reponse. [e-mail] I have to use implicit solver because I have a specific problem for my Thesis. I would really appreciate if you take a look at the k.file. Kind regards Rab Nawaz

[@AmeenTopa]

Hi Malik, I've sent an email. Please check you inbox.

[@krishnajagadeesh6307]

Hi sir

[@AmeenTopa]

Hello