After exporting Stress - Element-Nodal, Unaveraged,Von-Mises for different nodes and elements to ASCII file or EXCEL I found that the co-ordinates of grid points are wrong. But I need those co-ordinates.
Here is a simple example to investigate (see attachment).
I did select a feature face of my structure where all co-ordinates are:
x = 40 mm
y = 0 - 10 mm
z = 0 - 10 mm
Every first grid point of new element has wrong co-ordinates. In this special case it's always grid point 166 which is the GP in the middle.
I can find it in Excel AND and also in CSV file. See below:
It's interesting, but the values vary between CSV and EXCEL. Additionally, repeating leads to different values too.
I need the correct co-ordinates of the grid points in that primary NX excel sheet for additional post-processings in secondary excel sheets.
Is it a known error?
Is it only here in my case? - Am I the only one?
I could re-produce it in Windows 7, Windows 10 on different hardware.
Our NX version is 220.127.116.11.
What shall I do?
Who can help? - Best wishes, Michael
PS: I put my model and given EXCEL and CSV file in zip file
I'm able to duplicate the problem using your data in Simcenter 12.0.2.
When I run Simcenter 3D 2019.1 or 2019.2, I get the correct values out for all occurrences of grid 166 coordinates. It appears that the problem has been fixed in 2019.1.
Hi JimB, thank you for answering and testing.
Unfortunately, I can't wait until NX 2019-1 is part of our company's software pool. I don't know when it will be installed here. Mostly we don't get the latest version of software suites because of all those "teething problems".
The error also is happening in case of:
- grid point force - element-nodal, unaveraged,
- grid point moment - element-nodal, unaveraged,
- stress - element-nodal, unaveraged
- strain - element-nodal, unaveraged
- elastic strain - element-nodal, unaveraged
- and maybe other (in Thermal: flux???)
EDIT: Additionally, if I can't rely on the co-ordinates, can I rely on the values themselves???
Is NX 2019-1 in enumeration the second or the third version beyond NX12?
Maybe the error also appears in NX11 or NX1847 or NX1872?
Is it imaginable to force that correction in NX12 in addition?
As I assume NX12 is still in maintenance update state, right? - Is this error still to be corrected in such a case? - Do I have to create an Incident Request or so?
Can you help? - Best wishes, Michael
Sorry, I should have clarified better. I ran with your data in Simcenter 12.0.2 and encountered the same behavior as you reported - coordinates for node 166 were output incorrectly. I then ran in Simcenter 3D 2019.1 (and also 2019.2) and did not encounter the issue, so concluded that, if there was a bug, it had been fixed. I did not do in-depth research to verify that a PR had indeed been submitted and was marked as closed.
I did just open the .op2 file in Simcenter 11.0.1 and verified that nodal coordinate output to Excel was correct there as well, so the issue appears to be limited to 12.0.2 at least or 12.0.0-12.0.2 at most. You would need to contact GTAC and open an IR to go more in-depth as to which specific release(s) are affected and if a MP is possible for a particular release.
In enumeration, release order is:
End of maintenance dates are documented here:
thank you for your contribution. In the meantime I did found a new component which is not exported correctly:
Beam results, as they are unaveraged element-nodal results by definition.
Additionally, I was not able to get correct annotation for that nodal results. It looks like the annotations have there anchor point of leader at zero and not at nodal position. And they are those very nodes with wrong co-ordinates in excel. This is really weird.
I will make an Incident request. I feel very uncomfortable with those errors. All of that results are really important for me, I need it for daily work!
Best wishes, Michael
Last info before waiting for solution:
I could have managed to create an INCIDENT REQUEST: IR 9559456
Best wishes, Michael