Issue: Example anl_blade_cs_rcas fails with FileNotFoundError: cs\main_cs_set1.sg

Hi, everyone! I am a new user to VABS/iVABS. I’m trying to run the iVABS example below on Windows, but the analysis stops during the cross-section (cs) analysis step because a required .sg file cannot be found.

Versions / packages

  • Examples: ivabs-examples-0.11.build20240623.1.zip

  • iVABS: ivabs-0.11.build20240624.1-windows-portable

  • VABS: VABS4.1ReleasePC08-01-2024.zip

Steps to reproduce

(ivabs) C:\Users\whhong\ivabs\examples\anl_blade_cs_rcas> ivabs analyze main.yml

Expected behavior

The example runs successfully and completes the cross-section analysis.

Actual behavior

It fails with:

FileNotFoundError: [Errno 2] No such file or directory: 'cs\\main_cs_set1.sg'

Log excerpt:

(ivabs) C:\Users\whhong\ivabs\examples\anl_blade_cs_rcas>ivabs analyze main.yml
raw args = ('C:\\Users\\whhong\\ivabs\\scripts\\msgd\\__main__.py', 'ivabs', 'analyze', 'main.yml')
list args = ('analyze', 'main.yml')
parsed args = Namespace(subparser_name='analyze', loglevelcmd='info', loglevelfile='info', logfile='eval.log', inputfile='main.yml', paramfile='', resultfile='')
variant = ivabs
[12:31:19] INFO     reading main input main.yml...                                                              io.py:70
...
           INFO     [step: cs analysis] running cs analysis (h)...                                             sg.py:149
cs:   0%|                                                                                        | 0/9 [00:00<?, ?it/s]
           CRITICAL Traceback (most recent call last):
...
                    FileNotFoundError: [Errno 2] No such file or directory: 'cs\\main_cs_set1.sg'
...

Additional info / what I tried

I suspect this happens because prevabs did not generate the .sg file.
I tried running prevabs manually inside the cs folder:

prevabs -i main_cs_set1.xml -h -v

However, the command produces no output in the terminal, and it also does not generate the corresponding .sg file.

Question

Is cs/main_cs_set1.sg supposed to be included in the example package, or is it generated by a preprocessing step that I’m missing?
If it should be generated, could you please point me to the correct command/workflow to create it (and how to troubleshoot prevabs producing no output)?

I solve this problem by updating the prevabs file from this link: cdmHUB - The Composites HUB - Resources: PreVABS: Supporting Docs

@su.tian.22 is this the most recent version of prevabs? I thought it is not released with iVABS.

The issue could be due to the incompatible DLLs between prevabs and vabs.

we should fix the most recent one if the most recent one has issues.

Thank you very much for your timely reply! I think the iVABS version I’m using is the latest one available on GitHub.

@su.tian.22 Hi! I’m new to PreVABS and VABS, so I have a few questions—some of them may be quite basic. Recently, I’m using PreVABS v1.5 to generate .sg files and then running VABS v4.1 for the structural analysis. I noticed that the latest PreVABS release on GitHub is v1.6, and I’m wondering:

  1. Does using different PreVABS versions (e.g., 1.5 vs 1.6) lead to differences in VABS results?

In addition, I’m having geometry-generation issues when running PreVABS on airfoils from the IEA 15MW wind turbine blade, mainly in two aspects:

  1. For the FFA-W3-xxx series of sections, the trailing edge (TE) sometimes appears to be an open curve (not closed).

    • Does PreVABS provide any built-in handling for an unclosed TE (e.g., automatic TE closure, tolerance-based stitching, or point reordering)?
    • If not, what is the recommended preprocessing workflow (e.g., enforcing closure within a tolerance, adding a TE point, trimming/merging endpoints)?
  2. For some airfoils, I see a self-intersection / crossing issue near the TE on the pressure side (PS): in the TE region, there are points with y > 0 on the PS, which causes PreVABS to fail and no results are produced.

    • How should this geometry problem be handled?
    • Is there a recommended method to repair the airfoil coordinates (e.g., re-sorting by arc length, splitting into suction/pressure sides and enforcing monotonicity, smoothing, removing duplicate points, or rebuilding the TE region)?

Any guidance, best practices, or example scripts/tools for cleaning these airfoil coordinate files would be greatly appreciated. Thanks!

PreVABS may not be suitable for fine control of a cross-section right now. If you do not consider doing parametric and automated cross-sectional building and analysis, you can try manually building the cross-section in a tool like Abaqus/CAE and generate the VABS input file. You can check out this example Build Cross-section in Abaqus and Export to VABS — SGIO. If you prefer other tools than Abaqus, please let us know. We can update sgio to handle the data format you need.