Recurring Meeting of Cleveland Clinic - University of Utah

Date: October 23, 2013

Time: 1:00 PM EST

Means: Conference call

Attendees:

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

Agenda:

1. Discuss FEBio feature specifications development task progress.

   • http://wiki.simtk.org/openknee/Specifications/FebioFeatures

2. Discuss infrastructure/simulation software information updates.

   • http://wiki.simtk.org/openknee/Infrastructure/SimulationSoftware

3. Decide action items for next meeting.
4. Other

Immediate Action Items:

See notes for details.

• All

  • Explore moving November 27, 2013 meeting.

• University of Utah (Steve & Ben)

  • Start a wiki page to show progress on how to implement and test the given test problem for in situ strain (priority 1). A brief outline of algorithm and links to relevant paper will be included.

• Cleveland Clinic (Ahmet & Jason)

  • Provide a summary of how connector elements for joint coordinate system are used in finite element analysis, as a reference for the University of Utah team to incorporate such capabilities to the existing FEBio code.

Notes:

1. Discuss FEBio feature specifications development task progress.
   • Jason and Ahmet updated the team with the expected FEBio features based on their priority.
   • Features categorized based on conditions they are used in: constitutive modeling, pre and post processing, rigid body kinematics or surrogate modeling.
     • In situ strain.
       • Test problem to implement the feature was discussed. The University of Utah team already has an interested in FEBio implementation. Once developed for the given test problem, in situ strain can be prescribed in the joint model.
       • Implementation of this feature provides a good publication opportunity, potentially Steve as lead.
       • If a target strain is applied to an element, the element will no longer be in equilibrium. Without an applied force on the free end, the mesh will retract (until zero strain). When the end is fixed, the in situ strain will not be exact, it will be lower.
       • Various approaches will be worked on to reach the desired in situ strain application: apply close enough prescribed strain, or iteratively get to the desired strain, or apply a desired force, or finding a reference configuration that corresponds to the in situ strain configuration. As such, multiple algorithms can be developed to dictate the prescribed strain conditions. Initial consideration will be given to the first two approaches.
       • A pre-analysis step will be needed to equilibrate the applied strain fields. The pre-analysis step should accommodate implicit static and implicit dynamic simulations.
       • Ahmet will create a task to start a wiki page for in situ strain feature development progress. The task will be assigned to the University of Utah team.
     • Element, node and surface set definitions.
       • Sets will be useful to prescribe various properties, e.g., material properties, and loading and boundary conditions. With surface/node/element sets it will be easier to keep track of information especially when using scripts to extract information from result files.
     • Connector elements for joint coordinate system.
       • Connector elements can be useful to apply and interpret loading and boundary conditions, particularly for rigid bodies. This capacity provides easy adoption of a standardized clinical coordinate frame for a joint. The connector elements can be used to directly output relevant joint kinematics and apply physiological loads without the need to figure out the transformations between the given joint bodies.
       • For connector elements (and for pre-/post-processing) local coordinate systems may need to be defined for the femur or tibia. Kinematics and loads can utilize these local coordinate frames. Such a feature will need to transform the kinematics in a given coordinate system into the global coordinate system, at which FEBio calculates the positions.
       • Currently, in FEBio, the rigid body kinematics are defined in terms of the center of mass coordinate system which is automatically calculated by the program. The user can also position the center of mass anywhere as desired. The coordinate axes are still parallel to the global system. This feature will provide an ability to define coordinate axes and origin that is different than the global coordinate system.
       • An ability to define the connector elements is desirable not only for Open Knee(s) but also for finite element analysis of joints in general.
     • Penetration based contact.
       • This feature has already been implemented.
       • A test problem will be created to evaluate the existing capability.
     • Spring element and wrapping.
       • This feature aims for simplified modeling of ligaments with wrapping around bones.
       • Spring elements can be connected together to implement this. Such discrete elements (springs) will need to interact (like a contact) with the surfaces underneath (contact algorithm between a line element and a surface).
     • Shell elements of cartilage.
       • Shell elements will simplify the joint model, with the potential to predicting deformations in the contact zone as well as contact pressures.
       • Jeff noted that shell elements are not necessarily designed for prediction of transverse loading. There may be a need to find appropriate element formulation.
2. Discuss infrastructure/simulation software information updates.
   • Ben updated the relevant wiki page to provide information on simulation software. Ahmet reviewed the page.
3. Decide action items for next meeting.
   • See immediate action items above.
   • Ahmet will create tasks to manage progress in FEBio feature implementation.
4. Other

   • Jeff suggested using GoToMeeting for desktop sharing in future meetings.