Stress output in PSHELL (CQUAD8 - from op2)

57 views
Skip to first unread message

David Lessard

unread,
Dec 20, 2022, 4:07:04 PM12/20/22
to pyNastran Discuss
Hello everyone,

When I look into the op2 object, I can find each element (CQUAD8) and then, when I browse through, I was previoulsy able to extract the von mises Element-Nodal stress for each of the elements (like shown in the image below, I took the last item in the list, and it corresponded to the von mises stresses I found in NX).

cquad4 datas.png

So I thought I could only take the 4th item in the list to find the T_XY (shear stress), but the value doesn't correspond to the value I have in NX. 

When I am using NX, I use the native coordinate system to display my results. Could it be that my op2 file output result in a different coordinate system? Or is there another way (to specify in the header) to extract the shear stress in the native coordinate system?
Also, could it be that the CQUAD8 element has different type of structure to retrieve stress, I couldn't find it? (Once again, I was able to find values for Von Mises, but oxx, oyy and txy doesn't appear where it is suppose to be).

Thanks a lot for your help!
David

steve

unread,
Dec 21, 2022, 3:10:24 AM12/21/22
to pyNastran Discuss
David,

I think you're just mixed up on your indices.  Python is 0-based, so:
[0=fiber_distance, 1=oxx, 2=oyy, 3=txy, 4=angle, 5=omax, 6=omin, 7=von_mises]

The op2 could be in a different coordinate system, but that's rare.

Steve

David Lessard

unread,
Dec 21, 2022, 1:30:34 PM12/21/22
to pyNastran Discuss
No, I checked for this mistake and it's not the problem in my case. Here I am showing all datas associated with my 1st CQUAD element (units are kPa shown in Python). I highlighted the values associated with the 1 node of my 1st CQUAD8 element. 

Element 1 - All datas.PNG

The last item of the list is 1.193e+03, which represents 1.19 MPa for the Von Mises stress in my simulation. And when I look at my simulation, I have exactly that result (Node #1 = 1.19 MPa, units used in NX). So all is good.
Element 1 - Node 1PNG.PNGElement 1 - Node 1 - Von Mises.PNG

When I look at the o_xx = -5.939e+2, which represents -0,594 MPa (which is the 2nd term), it does not correspond to what I find in NX (using the native coordinate system in NX). In NX, I find 0.82 MPa.
Element 1 - Node 1 - oxx.PNG

I find it weird that everything is fine when I look at von mises, but nothing works when I look at oxx, oyy, txy.

Also, I should add that for my 1st frequency tested, the oxx, oyy and txy seemed to be matching (NX and Python yield the same result), but as I move foward in my random vibration analysis, the values start to diverge from one another. For the first 4 frequency, they are matching. But as I move further and further into higher frequencies, the values start to be a bit different, then a lot different until they are completetly differents at some point.

Thanks for your help, I really appreciate it
David

Steven Doyle

unread,
Dec 21, 2022, 1:47:48 PM12/21/22
to pynastra...@googlegroups.com
Per the NX/Simcenter DMAP 2019.2
image.png
That's what defines the order and the code and output vector seem consistent with the definition.

I suspect you are trying to get the results for the material coordinate system instead of the element coordinate system.  can transform shells from element->material cid using:

from pyNastran.op2.data_in_material_coord import data_in_material_coord
bdf.read_bdf(bdf_filename)
op2.read_op2(op2_filename)
op2_new = data_in_material_coord(bdf, op2)

If that's not the case, I'd need to see an example.

Best,
Steve

--
You received this message because you are subscribed to the Google Groups "pyNastran Discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pynastran-disc...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/pynastran-discuss/061c4c52-b6db-4d72-b63e-7b72de436950n%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages