Dunlop

Did you upload this to your answer ? I only see the visualisation of the pressure and flow at the end. sorry I missed that

I think I am working out what could be wrong, it seems like the mean in Tim Laska's code is not correct. I guess the idea of this is to calculate the middle of the elements and then calculate everything there. but it seems to be averaging the x with y in each coordinate and not x1 with x 2 .. maybe I am getting there. If it truly is a bug I will comment Tim's code

this is great! I am still struggling a bit in decoding Tim Laska's code to calculate the forces, but almost there I think

The next thing I will do is to go for higher res, but first I need to get the different steps working on something small

To answer your question the mesh is not very nice, but I wanted to start with small poor resolution images before I go to high res stuff. This means that everything will be rough and poorly represented

Hey thanks for this, I think I am getting the code by Tim Laska working now. the structure of his code is a bit different to mine.

Should we move this to chat?

I copied the image from MMA and it changed the format upon being uploaded. The code should run now, but with different dimensions

I see, the image changed its size upon being uploaded, I had a strange format which came from a CT scan so think something went wrong there. will check

I try again, could be there was a problem with copy paste from MMA to SE will check.

@Alex Trounev Thanks, I must have missed that, could be that it is already there in the other question . Could be that I should delete the question as it gets close to a duplicate then..

Will do, looking forward to testing out MMA 13.0

This is great! I think I understand now.Thank you so much!. I am really keen to use this more in my teaching and research.

Ah, think I have worked it out now. I was getting completely confused by the Voigt notation and forgetting that we are dealing with tensor multiplication. If I understand right, each of the three x three matrices is one of the "submatrices" of the full elasticity tensor.

i.e., {{-c15,-c14,-c13},{}{}}.nabla w = {sigma xx, tau xz, tau xy}

this is quite helpful. Does this mean say the first term in the top line i.e.{{-c15,-c14,-c13},{}{}}.nabla w gives sigma xx, tau xz and tau xy ?

Ah, I just saw below in the solid mechanics text the following

the gradient of w is a list of (gamma xz, gamma yz and epsilon zz)

the first term in the top row consists of a 3x3 matrix multiplied by the gradient of w

Perhaps the way for me to understand it is as follows:

Maybe things are becoming clearer (especially with the new solid mechanics tutorial)

Each of these terms consists of different 3x3 matrices (which are all sub-matrices of the elasticity matrix), multiplied by the gradients in the displacements (u,v,w). My question is how are these sub-matrices created?

Here I see that the top row would be the equivalent of the nabla sigma term (in the x-coordinate), which is the sum of terms in w, v and u.

Take the image of the output of the SolidMechanics PDE Component

Perhaps I am just thinking the wrong way.

The thing that I don't quite get is how one goes about creating this operator. I see now in 13.0 with the new solid mechanics package it is already there, but say if I wanted to create my own version of an operator, how do I go about doing this

when one runs it gives an unevaluated form of the equilibrium equations.

In fact what I am trying to understand is how to write the stress operator form of the -nabla sigma epsilon.

Sorry for the delay had to look after the kids.