Contact problem - deformation starts before contact

Hi,

If you are using NLSTAT as analysis type then small displacement analysis is used for solving unless you use LGDISP which include large displacements in nonlinear analysis.

Can you share the .out file using the File transfer link available in my signature?

Also, I think you need to consider the thickness of master and slave which are not displayed in HV. 

To confirm can you check the displacement of the node where enforeced displacement is applied?

Thanks Prakash.

.out file has been shared.

I still don`t understand why I am having contact before the parts touch each other. I tried different property and adjust options but none seemed working.

Hi,

I did not see any warning or error message in .out file 

Are you using solid elements or shell elements?

If it is shell elements thickness of shell elements are not visible in HV and what you are looking is the mid surfaces of master and slave.

I am using shell elements and sorry, I was wrong. The simulation comes to a completion but without any contact. The shell thickness is way smaller than the gap between the components (about 9-10 times) so this shouldn`t be a problem. The attached picture shows the initial gap between the components and the thicknesses. As soon as the cylinder moves through the other body, deformation starts in this body as if there is initial contact between them.

<?xml version="1.0" encoding="UTF-8"?>

Hi,

Can you share the model file please?

File sent.

Thanks

Hi,

So I see the normals of the elements are not facing each other. I think this is what causing the weird behavior of contact. 

Can you reverse the normal of the elements so that the both the cylindrical and curved surface normals face each other and try again?

I did but still have the same problem. 

I did but still have the same problem. 

OKAY, I will check the rest of the model setup and will update to you soon... /emoticons/default_smile.png' srcset='/emoticons/smile@2x.png 2x' title=':)' width='20'> 

Hi @ibozsoy1

Looks like it is a drop test which you are doing, is my understanding correct?

If that is the case better run a explicit analysis instead,

here Iam attaching model files and result file from RADIOSS explicit.

Unable to find an attachment - read this blog

Well, this is a crush test (not crash) and the inertial forces are negligible. This is just a simplified version of my model, which has several other parts in contact. I guess I can still use RADIOSS explicit but I need to know how to do mass scaling (since the loading rate is small (small time step), simulation will take a lot of time to complete). Can you provide any info about mass scaling?

Thanks

Hi,

You can control the timestep in RADIOSS using /DT/NODA/CST card in the engine file. The user has to specify Tmin, which is the target time step. And Tsca, which is 0.9.

Please refer the video for timestep control in Radioss Explicit: https://altair-2.wistia.com/medias/o0bfml9ah5

Thank you very much. I will work on it and let you know if I have more questions.