Get max stress values at the nodes instead of averaged

Hi,

I am currently working on developing a Python script model in which an rst file will be read through dpf reader (dpf.Model(rst_file)) and get the nodal stress values. But, the problem here is that the nodal stress values should not be the averaged ones but the maximum among the element contributions. I mean let's say node 1 is connected to elements 1,2,3 and 4. So, what's happening now is when I convert the elementalNodal stress values to nodal internally it is averaging nodal stress values at node 1 from all the elements (1,2,3 and 4). But what I need is the maximum nodal stress value at node 1 among all the element's contributions (max. of stress at node 1 among elements 1,2,3 and 4). I am able to do this using a for loop and iterating through all the nodes and elements but it consumes a lot of time for bigger models. So, is there any efficient way of doing this?

Find more posts tagged with

Python

Structures

DPF

Accepted answers

All comments

Chris Harrold

@Ramdane can you have a look at this?

Ramdane

Hi @Samukham , writing a C++ operator will allow you to have a faster solution. we can help you on this.

For scripting version, are you using this from Mechanical or pyDPF? the syntax will be a little bit different between Iron python and CPython.

Thanks,

Ramdane

Samukham

Hi @Ramdane , thanks for your response and for offering help. We're trying to implement this from pyDPF, through which we read the .rst file from the Python environment and the above-said post-processing task has to be performed.

Ramdane

@Samukham , OK let me get back to you then with a script example that we can evaluate. If the perfs are not good than we can write a C++ operator.

Samukham

@Ramdane Sounds good

Ramdane

hi @Samukham , sory to get back to you so late.

@Lea PARADIS , please could you help here writing the python script?

Lea PARADIS

Hi @Samukham,

Here is a link to a pull request on pydpf-core repository that sets up an example on how get maximum stress values at the nodes instead of averaged: https://github.com/ansys/pydpf-core/pull/1043

I added information about performances when averaging, it looks correct.