Why is there a discrepancy between the mold flow tensor and the rebuilt global Tensor of Ansy s

Murugan
Murugan Member Posts: 2
Name Dropper First Comment
edited October 2023 in Structures

Hello everyone,

I have a question regarding element orientation from injection mold data mapped on the ansys mesh.
I simulated with this data and I got some euler angles as you can see here below. [87.893,-5.402,89.805]

Then, I tried to reconstruct the global tensor coordinate system from the local eigen value and euler angle based coordinate system through python.
My eigen values which in this case are : array([0.84282816, 0.13237687, 0.02478798])
A11 and A22 (the first 2 are the same values that I get in ansys and Python calculation as well)
So the reconstructed global coordinate Tensor looks like this

Whereas my Tensor from Mold flow xml data looks like this below: (Note: only the yellow marked values are different and thus cause trouble ). I am just confused how the reconstructed global Tensor matrix (that too in this case just for one node) can look different than my mold flow input just by one single entry (just the same entry highlighted twice as it occurs twice in the tensor because of symmetry) ?

Because of this , I am not actually able to get the same euler angles , when I try to calculate directly from the mold flow data, as the eigen vectors are different in both the case (see below). Hence, I kindly request you to involve the developers here as well , as that would be of great help to quickly resolve my issue. This is actually turning out to be a concerning issue with my customer. I request you for your kind understanding.
It is causing the topic of uncertainty quantification to play a major role.

Due to this discrepancy, my euler angles are completely different

Looking forward for your kind reply.

Best Answer

  • Javier Vique
    Javier Vique Member, Employee Posts: 61
    First Answer First Anniversary Name Dropper 5 Likes
    Answer ✓

    Hi @Chris Harrold ,
    This user reached out to us by support and I told him that PyAnsys questions must be handled by Dev Forum. Anyway, this was related to a mistake in his script, I verified on my side that Mechanical works as expected and he told me that, when using Matlab, he was getting results as expected too.
    @Murugan Could you please confirm?
    Thanks and regards,
    Javier.

Answers

  • Chris Harrold
    Chris Harrold Member, Administrator, Employee Posts: 100
    First Anniversary First Answer Ansys Employee Solution Developer Community of Practice Member
    admin

    @AKD-Scripting-Team or perhaps @Maxime Rey could get to the right person for this one?

  • Murugan
    Murugan Member Posts: 2
    Name Dropper First Comment

    Hello @Chris Harrold ... There was nothing wrong in my script . There is an issue with the python library. And I found that with matlab I was able to calculate it right.

    Now coming to the main issue, I require help from ansys developers to get the open source PyDPF Composite library working on my side. Javier gave me the script. But I am not able to make it run on my side. And the ansys support says I need to get the help from you in order to make it run and ansys support cannot help me with this issue. So can you give me a contact with whom I can have a teams meeting, so that he can help me make the script run with the ansys.

    Thanks in advance

  • Chris Harrold
    Chris Harrold Member, Administrator, Employee Posts: 100
    First Anniversary First Answer Ansys Employee Solution Developer Community of Practice Member
    admin

    @Ramdane is best for DPF stuff or will know who to send you to at least. (Sorry for the delayed reply - was out last week.)