simtk.org

Hello, I'm having problems scaling the upper extremity model. When the scaling process finishes, it just shows the ulna and the humerus in a very strange position and the experimental markers are shown very far away from the scaled model.

Thank you for your help,

Nicolas Barrera Bateman

I am having similar problems scaling a full body model that includes upper extremities. Have you found any solutions since you last posted?

Hei,
i think i have the same problem. My ulna is translational dislocated to the humerus.

did you find a solution to your problem yet?

thanks,
Michael

Hi Michael,

Did the model geometry appear correctly before you do the scaling? I've seen cases where the geometry files were using a frame different from that assumed by the model and the bones ended up looking funny. If you have a custom set of geometry files, they need to be added to the geometry path in the GUI (Edit->Preferences->GeometryPath) to visualize correctly. This's particularly important with scaling since the scaled model doesn't live on disk so it's geometry search path is exactly that specified in the GUI.

Hope this helps,
-Ayman

Hei Ayman,

thanks for your fast reply.
I think i solved the problem in a more intuitive way. After the scaling i saved the scaled model, closed it and reopened it again. After reopening it, everything looks fine.
So i quess your suggestion might be right.

i've just got one more question:
after applying the scaling tool, i get the following error message (attached).
Am i right when i assume it just refers the graphical issues? because, further, after scaling the model the resulting model is not visualised correctly. not only the elbow joint is dislocated but also many bones (hand, trunk) are not visualised at all. the reappear all after reloading the model...!? or does the message imply something else which need to be taken care of?

Thanks for your fast help,
Michael

Hi, Michael. The error message occurred when a muscle activation level was being accessed for display purposes (i.e., to decide how to color the muscle path), but looked to me to be due to some other calculation that got done at the same time.

Error detected by Simbody method lapackInverse(Mat<>): Matrix is singular so can't be inverted (Lapack getrf info=3).

That may mean that the model does not have reasonable mass properties, or if there were some other numerical problem leading to improper 0, NaN, or Infinity values.

Regards,
Sherm

I have the same errors. How I can do more.
Thank you for help.

Hi,

To isolate if this is an issue with the ScaleTool or the Model, can you save the scaled model to a file and load it from the file in the GUI? If you get the same error then it's an issue with the model resulting from scaling, otherwise it's a scaleTool issue.

Please let us know how that goes,
-Ayman