ERROR: The area of segment 'ThoracicAorta' is negative.

[Karen-Helene Støverud]

Dear SV team,
I am building a 1D model of the aortic arch and subclavian artery based on the 1D simulation guide since I only have information about length and radius of each vessel segment.
However, the simulation crashes with the following "ERROR: The area of segment 'ThoracicAorta' is negative".
I have tried to increase number of elements and reduce timestep, but without success.
How to solve this problem?
Is it in the boundary condition or is the geometry a problem?
Please see attached file (changed format to txt to be able to upload).

All the best,
Karen

[Martin Pfaller]

Hi Karen,

I had a look at your model. I didn't find anything obviously wrong. I made some modifications and got it to run for 0.1s (attached). The things I changed were: add segment SubclavianA, change material parameters, divide R's by 10, multiply C's by 10.

When I look at the output, it seems that the segment ThoracicAorta is experiencing backflow, i.e. blood is entering through the descending aorta. I'm not sure why this is happening, but maybe this can give us more clues.

Did you generate this model through SimVascular? If so, can you send me the centerline. I'd like to see if something went wrong during model generation. I also attached one of our aortic models for comparison.

Best
Martin

[Karen-Helene Støverud]

Hi Martin,
Thanks a lot!
I need the model to be more stable such that i can run it for 10 s and is it ok to change the parameters that much?
Will I still get a realistic result? (I may have been a factor ten of because of unitconversion)

The goal of this model is to include the arteries in the arm -branches from the subclavian. And in the end I also need information about the veins.

I did not use any image data for this model so it is not based on centerlines, I simply made a drawing on paper of the arch and used diameters and length I found in a paper and used the coordinates in x,y plane for the nodes. Then I tried to connect the nodes via joints. Therefore, you are probably right that there are problems with the geometry.

Thanks for the file with your arch this was very helpful, but I think it would be easier to expand the model based on centerlines. Do you have a model like that available, or should I try to get my own dataset?
I can probably get MR data for the arteries from the aortic arch, but for the study I am working on I only have images from the arm and need a more general model for the aortic arch.

This was probably too much information, but hopefully it makes sense.

Best
Karen

[Martin Pfaller]

Hi Karen,

I made those modifications to generate some preliminary results that can point us to problems in the model. After the problems have been fixed, you'd go back to the parameters you want in your model.

I'm happy to hear that you're creating your 1D model from scratch. In many cases, the 1D results are already sufficient and you don't even need to build a full 3D model. So experimenting with 1D models can save you a lot of computation and model-creation time. Unfortunately, we haven't yet built a pipeline that starts from a 1D drawing of the vessels. I think this is something we should include in SimVascular in the future.

I sent you a private message about datasets.

Best
Martin

[Karen-Helene Støverud]

Hi Martin,
Ok, so when the geometry is ok, then I will not have problems with changing parameters

We plan to stick with 1D for this study and now I think that I might start with the aortic arch you sent me and try to continue to build the subclavian from this. It should be possible as I have the nodes and I can create new ones and view the results in Paraview.

I am so happy for your help!!

best
Karen