Open ThomasChauve opened 3 years ago
Hi Thomas,
Check this issue opened by myself some time ago: https://github.com/dicengine/dice/issues/185
As far as I know, you can't make an incremental correlation directly from the graphical user interface. Just create a custom input and params XML file (I attach some examples) and then call dice.exe from the command line as follows:
dice.exe -i input.xml -v -t
PS. in the input.xml file change the paths (as well as the subset and step sizes) according to your particular case. Also, in my case, I performed the DIC in a specific region of interest (defined in line 5 of the file), delete or comment this line accordingly. You can set the strain window in line 13 of params.xml file. Let me know if you have any problem
Cheers, Marco
Marco is right in what he comment. Thanks Marco. Maybe we should enable this from the GUI since it seems like more people are trying to use it. I'll change this issue label to enhancement, which would be to have the incremental formulation available from the GUI.
Thanks for you quick answer. I will use the command line then. Best Thomas
Hello everyone.
I am using v2.0-beta15 with ParaView 5.01 on windows 10.
I am trying to calculate local strain during tensile test of very ductile material. Using GUI everything I got that deformation near neck region is near 200% (VSG Strain), but in the middle of the neck I see black band because of the very sharp deformation in this region.
I guess that incremental DIC analysis would be helpful in this case. I add
<Parameter name="use_incremental_formulation" type="bool" value="true" />
in params.xml and started the analysis via PowerShell using the command
& "C:\Program Files (x86)\Digital Image Correlation Engine\dice.exe" -i input.xml -v -t.
Finally I got very strange result. VSG strain was very small ~0.03. And I do not see any deformation, neck etc.
What did I do wrong?
PS And what does it mean [unused] in PowerShell? For example: User specified correlation Parameters: interpolation_method = 2 [unused] sssig_threshold = 20 [unused] optimization_method = 1 [unused] initialization_method = 6 [unused]
Regarding the params having [unused] denoted for each of them, this is an artifact of the Teuchos::ParameterList object. Those parameters are indeed being used, but because the debugging print statement for the parameter list is called before any of the parameters are accessed, they are all flagged as having not yet been used.
Regarding why you are getting strange results for the middle of the neck it's hard to say what is going on. There are lots of things that get complicated for large deformation, cracking, necking, or stress concentrations. There are experimental things that could be causing this (the speckle pattern cracking or coming off, etc.) There are numerical issues: for example the shape functions used in the default version of DICe really only capture linear affine motion so if you have anything more complex, it will likely fail. There are also frame rate issues that could come into play. The bottom line is that even with the incremental formulation using DIC for large deformation or failure is still an immature area of research.
If you have some representative images you could share, that might enable use to see what's happening and give you some advice. Perhaps there is something else going on that could be easily fixed.
Thank you for your fast response.
I've attached small set of images in my last message (Experiment.zip).
It was very common test of AISI 304 flat sample. Experiment was ~20 minutes long and I took ~600 photos every 2 secs.
I've uploaded all the images to google drive. Feel free to download what is necessary to you.
All photos from the experiment (~600 images, 1.2 gb): https://drive.google.com/file/d/1dWoH2swHg_Fnm2NgLC56-1cX8GToplYt/view?usp=sharing
Short set of photos for DIC testing (21 images, every 30s photo, 42 mb): https://drive.google.com/file/d/1iXwWlPxF7wWfkXHjk_5m26DGM5iaEStL/view?usp=sharing
My settings and results files (183 mb): https://drive.google.com/file/d/1gnXh_JexFPv6E_6ZT3uRtk6WLbh-EcIb/view?usp=sharing
Hello
Thanks for this nice software.
I was wondering if it is possible to do incremental DIC analysis with your software ? I mean in the set of images (IMG_1, IMG_2, IMG_3) pefomed 1 DIC between IMG_1 and IMG_2 and a seconde between IMG_2 and IMG_3. Is there a convenient to do this ?
I am using v2.0-beta16. Best Thomas