# Rigid Body settings in case of contact shell-solid **URL:** \ **Category:** General Questions\ **Created:** [February 28, 2024, 12:06pm UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570 "2024-02-28T12:06:51Z")\ **Posts on this page:** 11\ **Page:** 1
**Author:** ![pkropotin](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/pkropotin/32/2951_2.png) [@pkropotin](https://prepomax.discourse.group/u/pkropotin)\ **Post date:** [February 28, 2024, 12:06pm UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/1 "2024-02-28T12:06:51Z")
Hello everybody! I’m calculating the composite rope that goes around the pulley (rope image in the end). Now I use an equivalent beam instead of a rope, so as not to increase the number of elements while setting up of the calculation scheme. The pulley is designed as a shell. On the first step, I bent the beam around the pulley. An the second step I will need to rotate the pulley at a constant speed. To rotate the pulley I need to define it as a rigid body. But now the program gives a calculation error if I define the pulley as a rigid body. Please tell me what is my mistake? If instead of a rigid body I fix all the pulley nodes, then everything works well. In case of rigid body: \*ERROR: increment size smaller than minimum best solution and residuals are in the frd file ![Beam_Pulley](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/5/5b2ba951bd5a66baa6db24eb1954d719365b6dd5.png) ![image](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/b/b6279a4cf69f17040ad3af2c1416315de12cd7a8.jpeg) PreProMax without Rigid: > **[test5\_work\_UBeam.pmx](https://drive.google.com/file/d/1AW_gtiIYmzDsV2v3ByL1PHdbz7yPLsqc/view?usp=drive_link)** > > Google Drive file. PreProMax with Rigid: > **[test5\_Rigid\_UBeam.pmx](https://drive.google.com/file/d/1i6puzQ36fdK3tanB-XskuzPiTUWZI3R6/view?usp=drive_link)** > > Google Drive file. ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [February 28, 2024, 12:28pm UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/2 "2024-02-28T12:28:07Z")
Rigid body constraints with shells and Nlgeom can cause convergence issues in CalculiX: [Rigid-body constraint convergance problems - CalculiX (official versions are on www.calculix.de, the official GitHub repository is at https://github.com/Dhondtguido/CalculiX).](https://calculix.discourse.group/t/rigid-body-constraint-convergance-problems/369) So it’s better to avoid that and use workarounds (like you did here or in different ways). ---
**Author:** ![pkropotin](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/pkropotin/32/2951_2.png) [@pkropotin](https://prepomax.discourse.group/u/pkropotin)\ **Post date:** [February 28, 2024, 12:48pm UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/3 "2024-02-28T12:48:30Z")
Great! Now it’s clear. So I will have to move the rope with zero friction instead of the pulley rotation. In that case I want to set a constant speed for the surface of the rope that situated on the top of the pulley (it’s fixed in the first step). Could you help me, how it’s better to set up? What type of dynamic analyses do you recommended for that? ---
**Author:** ![pkropotin](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/pkropotin/32/2951_2.png) [@pkropotin](https://prepomax.discourse.group/u/pkropotin)\ **Post date:** [February 28, 2024, 1:03pm UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/4 "2024-02-28T13:03:44Z")
Or maybe it’s better to design a pulley as solid body instead of shell. And set up the body as rigid. Will this then lead to a significant increase of elements? ---
**Author:** ![pkropotin](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/pkropotin/32/2951_2.png) [@pkropotin](https://prepomax.discourse.group/u/pkropotin)\ **Post date:** [February 28, 2024, 1:13pm UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/5 "2024-02-28T13:13:56Z")
I think I found some way here for the rope movement instead of the pulley rotation: [![](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/6/6aba663f6388ca9b083c9199e9fd463b48f7a5b1.jpeg "PrePoMax & CalculiX - Wear analysis") ](https://www.youtube.com/watch?v=0tHW6A3IRjA) But I have to use a rigid body option to connect a reference point to the surface of the rope. After that the reference point can be moved by Amplitudes settings. ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [February 28, 2024, 1:49pm UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/6 "2024-02-28T13:49:26Z")
> [@pkropotin](#): > > Will this then lead to a significant increase of elements? Not if you use a single layer of hex elements since shells in CalculiX are extruded to such a layer anyway. This might be the best approach. > [@pkropotin](#): > > But I have to use a rigid body option to connect a reference point to the surface of the rope. After that the reference point can be moved by Amplitudes settings. Yes, at the cost of the surface being rigid so it’s usually done with rigid tools. ---
**Author:** ![fatmac](https://avatars.discourse-cdn.com/v4/letter/f/b782af/32.png) [@fatmac](https://prepomax.discourse.group/u/fatmac)\ **Post date:** [February 28, 2024, 4:08pm UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/7 "2024-02-28T16:08:43Z")
I have tried for some time to model ropes - around sheaves and in other areas but have not really got very far. One of my issues has been to arrive at a body that has decent tensile stiffness along its major axis way and little or none in the other axes. I have even tried some sort of chain links with multiple short trusses but with all have not got very far. The best approach I have got is to use a belt -wide but thin to keep its bending stiffness down. To test it it ran it round a sheave and had it pull against a spring - see video. On the contact issue u mentioned with contact between solids and shells I have had no luck here- i get models to run in static analysis but as soon as I try non-linear the same model dose not run. For non linear studies I no longer even try -i just use solid to solid. If I have a cad model with a surface that I want in a contact study I simply extrude it into a solid of the relevant thickness before importing it into my meshing or FEA software. I will be keen to hear how u go with modelling the rope sheave problem -good luck! [BeltOnSheave.zip](https://prepomax.discourse.group/uploads/short-url/Ad6dxg8xJ5wfGOj9vN74WRQ9Jth.zip) (7.6 MB) ---
**Author:** ![fatmac](https://avatars.discourse-cdn.com/v4/letter/f/b782af/32.png) [@fatmac](https://prepomax.discourse.group/u/fatmac)\ **Post date:** [February 28, 2024, 7:31pm UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/8 "2024-02-28T19:31:24Z")
as soon as I sent this I realised I had made an error! Contact is non-linear! What I had in mind was node-surface connections - I get them to work with surfaces in static but non linear analyses. That aside I still have run into issues with shell to solid contact and for non-linear avoid it. ---
**Author:** ![pkropotin](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/pkropotin/32/2951_2.png) [@pkropotin](https://prepomax.discourse.group/u/pkropotin)\ **Post date:** [February 29, 2024, 4:04am UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/10 "2024-02-29T04:04:50Z")
Absolutely! I would be very grateful if you could share the pmx files of this model. If you don’t feel comfortable posting this publicly, here’s my email: [pkropotin@gmail.com](mailto:pkropotin@gmail.com) In turn, I promise to send you a model of the composite belt when I finish it (PreProMax and FreeCAD files). Thank you! ---
**Author:** ![pkropotin](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/pkropotin/32/2951_2.png) [@pkropotin](https://prepomax.discourse.group/u/pkropotin)\ **Post date:** [February 29, 2024, 8:31am UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/11 "2024-02-29T08:31:44Z")
Thank you for help! I’m a beginner in PreProMax and it’s very helpful! I have changed the shell pulley to solid body pulley. So, I can use RigidBody in the model now, but it looks like there are some errors in PreProMax. During the calculation it changed the reference point units from mm to meters (1000 multiplication). I correct it, but it appears again. Have you ever face with such errors? ![image](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/3/396664cf3ddf6789c7f59a00d40093b08f5d55b4.png) ![image](https://global.discourse-cdn.com/free1/uploads/prepomax/original/2X/0/01dbf32fc00a499694d6a7691a6b53c578f879b0.png) ---
**Author:** ![FEAnalyst](https://yyz2.discourse-cdn.com/free1/user_avatar/prepomax.discourse.group/feanalyst/32/688_2.png) [@FEAnalyst](https://prepomax.discourse.group/u/FEAnalyst)\ **Post date:** [February 29, 2024, 8:34am UTC](https://prepomax.discourse.group/t/rigid-body-settings-in-case-of-contact-shell-solid/1570/12 "2024-02-29T08:34:06Z")
Try redefining the reference point, maybe using a different method. It should work.