Recurring Meeting of Cleveland Clinic - University of Utah

Date: March 26, 2014

Time: 3:00 PM EST

Means: Conference call

Attendees:

1. Ahmet Erdemir (Cleveland Clinic)
2. Jason Halloran (Cleveland Clinic)
3. Jeff Weiss (University of Utah)
4. Steve Maas (University of Utah)
5. Ben Ellis (University of Utah)

Agenda:

1. Progress on in situ strain test problem.

   • http://wiki.simtk.org/openknee/Febio/InSituStrain

2. Update on local coordinate systems and connector elements.

   • http://wiki.simtk.org/openknee/Febio/RigidBodyMotions

3. Update on set definitions in FEBio input and output.

   • http://wiki.simtk.org/openknee/Febio/SetDefinitions

4. FEBio solution repeatability.
5. Decide action items for next meeting.
6. Other.

Immediate Action Items:

• Cleveland Clinic (Ahmet)
  • Run test problems to evaluate solution paths of FEBio using single and multiple threads.
• University of Utah (Ben)
  • Provide documentation on the effect of Pardiso on FEBio solution paths when using multiple threads.
  • Run test problems including contact to evaluate solution paths of FEBio using single and multiple threads.
• University of Utah (Steve)
  • Solve in situ strain problem for cardiovascular test case for potential CMBBE 2014 abstract submission.

Notes:

1. Progress on in situ strain test problem.
   • Steve was able to solve a three-dimensional ligament problem illustrating the general purpose utility of in situ strain feature. He utilized data from previous work on the medial collateral ligament conducted using Nike3D. He didn't have access to original in situ strain data; therefore utilized previous model results. Nike3D and new FEBio implementation resulted in similar strains; quantitatively stresses predicted by FEBio were higher. This may be a result of different material model implementations between Nike3D and FEBio.
   • If another problem can be solved by CMBBE 2014 deadline, an abstract submission will be possible. A simplified version of the cardiovascular problem may serve for this purpose. Jeff and Steve will discuss this opportunity.
   • Ahmet wondered if changing on boundary conditions can be implemented appropriately during and after in situ strain prescription. For example, during in situ prescription the ligament ends can be fixed; in following step one end may be set free to allow settling of the tissue in equilibrium. Steve already illustrated this with a simple artery example, where the opening angle was modeled: tied nodes during in situ strain application, relaxed to simulate cutting of the artery. Jeff noted that in situ prescription can be conducted along with various displacement and load conditions at tissue boundaries.
   • The team discussed the need for a format to specify in situ strains. In situ strains are implemented at the element level and can be prescribed in a spatially varying manner. This can be done within the FEBio file or maybe by providing element or location-specific in situ strains in a certain format (in Preview for example). A workflow for Open Knee(s) project will be discussed.
2. Update on local coordinate systems and connector elements.
   • No updates on this issue; the University of Utah team has been focusing on the in situ strain feature.
3. Update on set definitions in FEBio input and output.
   • No updates on this issue; the University of Utah team has been focusing on the in situ strain feature.
4. FEBio solution repeatability.
   • Ahmet relayed the problem of repeatability in solution paths of FEBio experienced by the Stanford University team (illustrated by different solution file sizes). This issue has been discussed years back as it was observed when Pardiso (underlying parallel linear solver for FEBio) was used with multiple threads. Ben will talk to Dave Rawlins to pull out previous data on this issue. Based on his anecdotal experience, Ben also noted that this issue may happen with contact problems even when solved on a single thread. Both Cleveland Clinic and University of Utah teams will conduct some simulations to troubleshoot under different simulation cases and computing conditions.
5. Decide action items for next meeting.
   • See Immediate Action Items above.
6. Other.
   • Jeff noted a request by an investigator interested in an "Open Spine" project. He will introduce this investigator to Ahmet and Ben by e-mail for prospective discussions.