Error when trying to load cubit file (Problem with Acis version)

jsilva.mit · October 26, 2021, 7:10pm

Thank you for the help with this. I can confirm that I was able to read your .cub file using Coreform Cubit 2021.5. There are also two new behaviors that I did not found before when I saved my model using version 2021.5:

I am able to save the same model that you sent (using save as…) in the .cub5 format without any crash or problems.
I am able to save the model in .cub and then read it again without any problems.

This might be an indication that the version 2021.5 might be causing the problem when saving my model, and that this problem was solved in version 2021.10 that you used.

Let me know what you find using version 2021.5.

Thanks also for the advice regarding using python and the command line.

Thanks,
Josimar

jsilva.mit · October 27, 2021, 11:04am

Hi @karl,

Just wanted to give you a quick update on this:

I opened the .cub file that you sent me by email, imported additional surfaces and then perform additional webcuts. When I try to save the model using Coreform Cubit 2021.5 I get the same error message as before:

selected filter: Cubit files (*.cub)
Cubit>open “./modeV3.cub”
Geometry engine set to: ACIS Version 28.0.2.0
ERROR: ACIS API error number 73004
ACIS API message = missing logical in restore file
ERROR: did not read anything from the input file ‘C:\Users\joa790\AppData\Local\Temp\CBT6D48.tmp’
You may be trying to read a file that is from an Acis version
newer than you are using.
ERROR: Command Failed.

I will send you the journal file that I created above along with the additional surfaces so that you can try it if you want.
I used Coreform Cubit version 2020.2 and confirmed that the error above still occurs. In this case I used the original journal file that I sent you by email.

Do you have any other ideas of how I could fix this problem? I know you mentioned that version 2021.10 will be released in the next few days, but would the current development version already have the bug fixed?

Let me know if I can do to help.

Thank you,
Josimar

karl · October 27, 2021, 12:24pm

You can try the latest development release at Coreform Product Downloads under “Latest Development Builds.”

Karl

jsilva.mit · October 27, 2021, 1:15pm

Hi @karl,

Thank you for the suggestion. Quick question: did you manage to make it work with the 2021.5 version in your end? I know it worked with 2021.10, just curious whether you obtained similar results for the 2021.5.

Thanks,
Josimar

karl · October 27, 2021, 1:19pm

Hi Josimar,

I’m working on getting 2021.5 installed on my Fedora 33 box this morning. I’ll let you know the results.

Karl

jsilva.mit · October 27, 2021, 1:23pm

Hi @karl,

Thanks, very much appreciated. Just a quick note that the errors that I get with the 2021.5 version occurs both on windows and linux (ubuntu) — I have tried in both.

I will install the latest development version ( Coreform-Cubit-2021.10-rc+20708_39151fb8-Lin64.rpm) and will try again.

Let me know what you find.

Thanks,
Josimar

karl · October 27, 2021, 2:46pm

Hi Josimar,

I just ran 2021.5 on Fedora 33 and save the .cub5 file without a crash. I was able to open that file up and get the same model as before. Is your disk low on space?

Karl

jsilva.mit · October 27, 2021, 3:15pm

Hi @karl,

My disk is not low on space, I double checked that.

Another update:

I used Cubit 2021.5 on Windows and rerun the journal file that I sent you yesterday and I still get the error message above. I will send you the exact journal file along with all the input surfaces so that you can try it out if you want.
I am waiting on the result for Cubit 2021.10 development version.

Let me know if you have any other ideas.

Thanks,
Josimar

jsilva.mit · October 27, 2021, 3:28pm

Hi @karl,

I sent the journal files and the input surfaces to support@coreform.com. Note that in the zip file there is a folder called CoreformCubit which contains a .cub file. This is the output from the journal file. If you try to open this file using cubit version 2021.5 you would get the error messages that I refer above.

Let me know if you have any questions.

Best,
Josimar

jsilva.mit · October 27, 2021, 4:43pm

Hi @karl,

Just wanted to report the results from testing with the development version ( Coreform-Cubit-2021.10-rc+20708_39151fb8-Lin64.rpm )

Unfortunately, the error still persists. There is an improvement that I am able to save the file as .cub5 without Cubit closing and crashing.

The error occurs when I try to load the geometry back to my projects. What I do is: 1) Close Coreform Cubit, then open it again and try to load the previously saved .cub or, in this case, the .cub5 file. The exact error message that I get is below.

I sent to support@coreform.com the exact journal file that I used to create the .cub file as well as all the input surfaces. To reproduce this error, perform the following steps:

Install docker container on Ubuntu 18.05
Download and build the Coreform development version ( Coreform-Cubit-2021.10-rc+20708_39151fb8-Lin64.rpm )
Inside the docker, run the script 1_MakeModelStratigraphy_v2.py which will create the .cub file
Using save as, save the project as .cub5 version as well
Close cubit, and then try to open the .cub file or the .cub5. You should then see the above error message.

Let me know if you need any additional information from my end.

Thanks,
Josimar

jsilva.mit · October 28, 2021, 6:02pm

Hi @karl,

Just wanted to follow up on this. Let me know if you need anything from my end.

Thanks,
Josimar

karl · October 28, 2021, 6:07pm

Hi Josimar,

I will try running your exact script. I probably won’t install a docker image. My windows box is running low on disk space. It is frustrating, because I see the problem when I look at your .cub5 file. But I can’t generate a .cub5 file with the problem.

Karl

jsilva.mit · November 3, 2021, 5:27pm

Dear @karl,

Just wanted to check if you have any updates on this. If I understand correctly, you can’t reproduce the error using the exact scripts that I sent you?

Let me know if there is anything that I can do to help.

Thanks,
Josimar

karl · November 4, 2021, 1:36pm

Hi Josimar,

I can not replicate your problem while running your scripts. I see the problem when I look at the cub files you generate, but I cannot create my own. I will spend a little more time today looking at this.

Karl

jsilva.mit · November 8, 2021, 4:13pm

Hi @karl,

I am still working on this problem. I would like to know your thoughts regarding the following ideas:

I noticed that after saving the cubit session as .cub5 before imprint all, I can then open the session normally. However, if I save it after imprint all I cannot open the session (see the error message above, in the previous discussion). Would it be possible that the imprint process is corrupting the Acis geometry and causing the error that I observe?
Instead of imprint all, would you recommend me using imprint only for specific surfaces that I know are at the interface between volumes? Would imprint tolerant help?

Thanks,
Josimar

karl · November 8, 2021, 4:42pm

Hi Josimar,

I think what is happening is that the imprint creates many new very complex curves and surfaces. The size of the ACIS file increases with the number of entities. For some reason on your system, you fail to write the large ACIS file into the cub file.

What is the difference in file size between an exported ACIS file before and after the imprint?

Karl

jsilva.mit · November 8, 2021, 4:52pm

Hi @karl,

Thanks for the reply. The difference in size is around ~3 to 4 Gb. Before imprint typical size of the model geometry is ~4-5 Gb. After imprint the size decreases to 1 - 2 Gb. These sizes correspond to the .cub5 and .cub formats.

Note that if I save the model in ACIS format (.sat) after imprint I have a file with size of ~5 Gb.

Is there a away to test your hypothesis that my system might be preventing me to write the ACIS file after the imprint?

Would it help if I send you my environmental variables?

Let me know what I can do to help.

Thanks,
Josimar

karl · November 8, 2021, 4:56pm

Hi Josimar,

You previously did an export acis after imprint. Can you compare that file with an export acis before doing the imprint?

Karl

jsilva.mit · November 8, 2021, 7:01pm

Hi @karl,

I regenerate the model and here are the results.

The exact command lines that I run were:

## Exporting Acis geometry before imprint
cubit.cmd("export acis './CoreformCubit/ModelGeometry_BeforeImprint_{}.sat' ascii overwrite".format(modelNumber))

cubit.cmd("imprint all")

## Exporting Acis geometry after imprint
cubit.cmd("export acis './CoreformCubit/ModelGeometry_AfterImprint_{}.sat' ascii overwrite".format(modelNumber))

Here are the results:

Before imprinting, the .sat file has 3.8 Gb
After imprinting, the .sat file has 5.9 Gb

What do you think about these results?

Thanks,
Josimar

karl · November 9, 2021, 3:44pm

Hi Josimar,

That is what I expected. The file is significantly larger after imprinting. I have asked my IT people if they have any ideas about what could be causing this on your environment. Hopefully, they have some directions to try.

Karl