applied prestress on shell element

[achitsaz]

Hello,
I spent three months working on a project involving the mechanical analysis of arterial walls in aneurysm regions using FEBio. I obtained genuine geometry of aneurysms and parent arteries from patient images, but I ran into a problem when I couldn't extract the precise domain for the arterial wall from the images. As a result, I simply drew the surface of the arterial wall from the segmented images. Therefore, I believe that using shell element instead of solid mesh for my analysis would be best. Additionally, the geometry I obtained isn't a free-loaded geometry, so I decided to use prestrain elastic material for my cases. I have attached the feb file for your review. (all parameters and dimension in this file in the CGS unit system)

I have two primary questions about prestrain material in FEBio:

1. How can I determine the prestain gradient tensor F0 for each case? I only have the pressure inside the artery and the loaded geometry, so how can I obtain F0 with this information? I tried using an iterative algorithm to optimize F0 based on the residual displacement and strain by applying pressure gradually. For a few cases, it works. However, the algorithm crashes during the process for most cases because FEBio cannot solve the negative Jacobian error. How can I get rid of the Negative Jacobian error? ( the file I have attached for the a case is have a same problem)

2. I read about the two-step method in the user manual, but I am not sure how to use it. Could you please guide me on how to reach the prestrained geometry?

I tested and checked everything, but I couldn't understand how to fix this problem. Your help is very appreciated.

Thank you.

https://drive.google.com/file/d/1xQtlayB4rbK73bv-I-W00m27OtivPtpN/view?usp=sharing

[ateshian]

Hi,

I am not sufficiently knowledgeable about the restrain gradient method (you can look up the published paper on this subject) to answer your first question, but I can answer the second one: If you know the amount of pressure that produced the deformed geometry, you can apply that pressure in reverse (e.g., outside-in) to recover an approximation to the pre-strained geometry. Once you get that deformed mesh of the pre-strained geometry, you export the mesh (from the plot file, set at the final time point, select File->Save as... and select the *.feb format), then import it into a new model where you reapply the properties and boundary conditions that you want.

This procedure is only approximate, because of material and geometric nonlinearities. In other words, if you take your new model, prescribe the pressure the normal way (i.e., inside-out), you will not recover the original geometry exactly, but hopefully the difference will be small.

Best,

Gerard