Hello everyone,
I modelled a three-point bending problem and the result I got shows a lot of fluctuation in the force output. Is there a way to improve the output and produce a smoother graph (without using filters).
Hi @syed2468,
How did you retrieve this force data? For example: Contact force, section force, beam force etc.
Why is that high frequency signal bothersome to you?
@icebreakersours Becuase I don't know how to explain the oscillation. In the experimental testing the force output is a lot more stable and smooth. Is this high frequency a normal observation when outputting contact force using RCFORC?
On a side note, does either of you have any experience working with *MAT_240_Cohesive_Mixed_Mode? I'm trying to determine what strain rates I should test the material properties at to properly define the material model.
Thanks
@syed2468 ha yeah noise like this is actually very common when plotting rcforc. Think about it, you have a simulation that is trying to determine the force between two meshes that are constantly trying to cross eachother and the solver math is pulling back any nodes that start to penetrate the other mesh. In a basic sense, that's why you see these oscillations in the contact force for a lot of simulations.
Now, you can refine your contact to maybe cut this down or you can using section force which I always recommend if possible. Also, you can 100% filter that as long as it's resonable and you report it in your study.
Any chance one of the objects in the contact is a deformable solid?
@syed2468 Such high-frequency "noise" is usually seen in explicit analyses - which is what I am assuming this analysis is. If this frequency content has little to do with the frequency content in the forcing function (loading/boundary condition) or the structural normal modes, then you can easily ignore it.
Also, another way to look at this "noise" is to think of it as an artifact of not checking for convergence (which is what you do using an implicit scheme).