I homogenized the unitcell using TexGen4SC for a plate/shell model (Kirchhof-love), and copied filename.sc.k, filename.sc and filename.sc.opt from cmdHUB storage, then imported the homogenized properties from the cdmHUB using Abaqus-SC, and analyzed the structure using Abaqus GUI.
When I wanted to dehomogenize it using Abaqus-SC, I receive an error
"OdbError: The path 'XXX/temp/filename.odb' has an existing ddb associated with it."
Some of your steps are not very clear to me. I don't understand the meaning of "imported the homogenized properties from the cdmHUB using Abaqus-SC ". After your homogenization using TexGen4SC, you should get homogenized plate stiffness matrix in .k file. Then, you should use the homogenized plate stiffness matrix to define shell elements in Abaqus to do your plate structural analysis. The above two steps (homogenization and plate structural analysis) do not need Abaqus-SC GUI. At last, you want to do dehomogenization, you can use Abaqus-SC GUI as shown in your attached figure. Based on my understanding, you should have .sc file and .ech file in your working directroy, and you import .sc file, which is the input file for SwiftComp. Then, you need to input the structural responses (e11, e22, 2e12, k11, k22, 2k12) based on your structural analysis. Please first to make sure you have done the above steps.
Yes, up to the structure analysis is exactly correct as you said.
Yes, in the Abaqus-SC GUI manual under the subsection Dehomogenization it is mentioned that.sc, .sc.k and .sc.opt are required to be in the current work directory.
I input the structure responses, now dehomogenization completes successfully; however, the following error prompts after dehomogenization (please see the attached picture).
A 2D Shell macroscopic model was chosen in importing the homogenized properties and in dehomogenization steps.
You can check if .sn file is generated in the working directory and make sure it is not empty. If .sn file is ok, then the problem is mainly about the Abaqus-SC GUI. I think my colleagues will answer you questions.
Any ideas about the above mentioned error in Abaqus-SC?
Xin mentioned that the .sn file might be empty but it was not, so another possibility is that there might be some problems with the Abaqus-SC GUI.
Does anybody has an example in which a model is homogenized in TexGen4SC, then the homogenized stiffness matrix is used in Abaqus to analyze the structure, and finally dehomogenized using Abaqus-SC?
I want to find out what caused the aforementioned error.
It seems that the information in the .sc.sn is not complete. For C3D8 element type, each element requires 8 nodal values for each element. If the element number provided in the sc file is not correct, this error message will prompt out.
It seems that the information in the .sc.sn is not complete. For C3D8 element type, each element requires 8 nodal values for each element. If the element number provided in the sc file is not correct, this error message will prompt out.
I used your .sc file and run a dehomogenization analysis using the GUI, and there was no problem. I checked your .sc.sn file and found no problem there. Maybe you can download the newest GUI and have another try.
I downloaded the updated Abaqus-SC GUI and went through the process again, I received the same error which is attached here again.
Besides of the above mentioned problem, I modeled a very simple geometry to test Abaqus-SC by following the provided tutorial, but I get an error stating that sc.k file should be in the directory when I want to dehomogenize it in Abaqus-SC GUI. I assume that means that homogenization process was not complete.
Another thing I noticed was "You are running SwiftComp in Parallel version!; Problem control parameters: analysis elem_type trans_flag tem_flag" while dehomogenizing the model and it would freeze and do nothing unless you close the whole software.
Are there any manuals or examples for TexGen4SC and importing .sc file in Abaqus-SC GUI?
The files in cmdHUB are not satisfactory for my purpose and cannot obtain any results.
I homogenized the model (3D Solid) using TexGen4SC in the cloud, saved the .sc, .sc.k and .sc.opt files, and imported the .sc file into Abaqus-SC GUI (Analysis type: Elastic, Macroscopic model 2D).
I analyze my model but in dehomogenization step I've received several errors, and this step have not been carried out successfully in Abaqus-SC GUI up to this point.
1. Have the .sc from the TexGen4SC ready, copy it to your working directory (in your PC). You also need to have SwiftComp.exe and License file in this directory.
2. Open the command window, navigate to the working directory, type command: swiftcomp < scfile.sc> H 2D and press 'Enter'.
This step is running the homogenization analysis and to obtain sc.k, .sc.opt files in your working directory before the dehomogenization.
3. Open Abaqus SwiftComp GUI, set the working directory of the Abaqus GUI to your working directory.
Reza Sourki @ on
I homogenized the unitcell using TexGen4SC for a plate/shell model (Kirchhof-love), and copied filename.sc.k, filename.sc and filename.sc.opt from cmdHUB storage, then imported the homogenized properties from the cdmHUB using Abaqus-SC, and analyzed the structure using Abaqus GUI.
When I wanted to dehomogenize it using Abaqus-SC, I receive an error
"OdbError: The path 'XXX/temp/filename.odb' has an existing ddb associated with it."
and in the command line interface:
"Macroscopic displacements
(0.0, 0.0, 0.0)
Macroscopic roatations
((1.0, 0.0, 0.0), (0.0, 1.0, 0.0), (0.0, 0.0, 1.0))
shell macroscopic strain and curvatures :
(0.0, 0.0, 0.0, 0.0, 0.0, 0.0)"
Any thoughts?
Thanks in advance.
Xin Liu @ on
Some of your steps are not very clear to me. I don't understand the meaning of "imported the homogenized properties from the cdmHUB using Abaqus-SC ". After your homogenization using TexGen4SC, you should get homogenized plate stiffness matrix in .k file. Then, you should use the homogenized plate stiffness matrix to define shell elements in Abaqus to do your plate structural analysis. The above two steps (homogenization and plate structural analysis) do not need Abaqus-SC GUI. At last, you want to do dehomogenization, you can use Abaqus-SC GUI as shown in your attached figure. Based on my understanding, you should have .sc file and .ech file in your working directroy, and you import .sc file, which is the input file for SwiftComp. Then, you need to input the structural responses (e11, e22, 2e12, k11, k22, 2k12) based on your structural analysis. Please first to make sure you have done the above steps.
Reza Sourki @ on — Edited @ on
Hi Xin,
Yes, up to the structure analysis is exactly correct as you said.
Yes, in the Abaqus-SC GUI manual under the subsection Dehomogenization it is mentioned that.sc, .sc.k and .sc.opt are required to be in the current work directory.
I input the structure responses, now dehomogenization completes successfully; however, the following error prompts after dehomogenization (please see the attached picture).
A 2D Shell macroscopic model was chosen in importing the homogenized properties and in dehomogenization steps.
Xin Liu @ on
You can check if .sn file is generated in the working directory and make sure it is not empty. If .sn file is ok, then the problem is mainly about the Abaqus-SC GUI. I think my colleagues will answer you questions.
Reza Sourki @ on
Hey Xin,
No the .sn file is not empty.
unitcellplate.sc.sn
6 MBClick to download
Reza Sourki @ on
Hi all,
Any ideas about the above mentioned error in Abaqus-SC?
Xin mentioned that the .sn file might be empty but it was not, so another possibility is that there might be some problems with the Abaqus-SC GUI.
Does anybody has an example in which a model is homogenized in TexGen4SC, then the homogenized stiffness matrix is used in Abaqus to analyze the structure, and finally dehomogenized using Abaqus-SC?
I want to find out what caused the aforementioned error.
Bo Peng @ on
It seems that the information in the .sc.sn is not complete. For C3D8 element type, each element requires 8 nodal values for each element. If the element number provided in the sc file is not correct, this error message will prompt out.
Can you also provide your .sc file please?
Bo Peng @ on
It seems that the information in the .sc.sn is not complete. For C3D8 element type, each element requires 8 nodal values for each element. If the element number provided in the sc file is not correct, this error message will prompt out.
Can you also provide your .sc file please?
Reza Sourki @ on
Hi Bo,
Thank you for your reply.
The .sc file is the one I saved from the cloud after using TexGen4SC. Please find it in the attached file.
unitcellsplate-1.sc
534 KBClick to download
Bo Peng @ on
I used your .sc file and run a dehomogenization analysis using the GUI, and there was no problem. I checked your .sc.sn file and found no problem there. Maybe you can download the newest GUI and have another try.
Reza Sourki @ on — Edited @ on
Hi Bo,
That's good to know thank you!
I downloaded it almost a month ago and was not able to get results, so, there might be a problem with the GUI or even the license?
I will download it again, but I believe I have to get a new license from Professor Yu as well.
Thanks Bo.
Bo Peng @ on
Actually I don't think there is any problem related to the GUI and license in your case. But please have another try.
Reza Sourki @ on — Edited @ on
Hi,
I downloaded the updated Abaqus-SC GUI and went through the process again, I received the same error which is attached here again.
Besides of the above mentioned problem, I modeled a very simple geometry to test Abaqus-SC by following the provided tutorial, but I get an error stating that sc.k file should be in the directory when I want to dehomogenize it in Abaqus-SC GUI. I assume that means that homogenization process was not complete.
Bo Peng @ on
You need to have the .sc, .sc.k, .sc.opt files in your working directory before the dehomogenization.
To have these required files, you can use rerun the homogenization in your working directory.
To rerun the homogenization step, you can use a command line :
swiftcomp scfile.sc H 2D
in your command window.
To open the command window, type : cmd in your windows search bar, and click the command window, then navigate to the working directory.
Good luck.
Reza Sourki @ on — Edited @ on
Hi Bo,
I tried that but it states it cannot READ scfile.sc!
Bo Peng @ on
scfile.sc is your swiftcomp file name.
swiftcomp <Your scfile name.sc> H 2D
Reza Sourki @ on
Another thing I noticed was "You are running SwiftComp in Parallel version!; Problem control parameters: analysis elem_type trans_flag tem_flag" while dehomogenizing the model and it would freeze and do nothing unless you close the whole software.
Are there any manuals or examples for TexGen4SC and importing .sc file in Abaqus-SC GUI?
The files in cmdHUB are not satisfactory for my purpose and cannot obtain any results.
Xin Liu @ on
You can find all texgen4sc related materials (introduction, manual, video) at the link "https://cdmhub.org/tools/texgen4sc".
Reza Sourki @ on — Edited @ on
>I've watched the videos and have read the files.
I homogenized the model (3D Solid) using TexGen4SC in the cloud, saved the .sc, .sc.k and .sc.opt files, and imported the .sc file into Abaqus-SC GUI (Analysis type: Elastic, Macroscopic model 2D).
I analyze my model but in dehomogenization step I've received several errors, and this step have not been carried out successfully in Abaqus-SC GUI up to this point.
Any ideas or examples?
Bo Peng @ on
1. Have the .sc from the TexGen4SC ready, copy it to your working directory (in your PC). You also need to have SwiftComp.exe and License file in this directory.
2. Open the command window, navigate to the working directory, type command: swiftcomp < scfile.sc> H 2D and press 'Enter'.
This step is running the homogenization analysis and to obtain sc.k, .sc.opt files in your working directory before the dehomogenization.
3. Open Abaqus SwiftComp GUI, set the working directory of the Abaqus GUI to your working directory.
4. In the GUI, run the dehomogenization analysis. You can follow the Abaqus-SwiftComp_GUI_v1.2.2_Manual.pdf : https://cdmhub.org/resources/1134/supportingdocs
I also have the results files attached as your reference.
plateunitcellHomo.zip
7 MBClick to download