OpenSim only for gait simulations?

[Emil Walczak]

Hello everyone!

At the beggining an introduction to my model:

I'm simulating the jaw movement. My model consist of a jaw, two articular discs and crane.
Crane is defined as a ground body, discs joints are defined as free joints and for jaw I am using a custom joint.
I've five elastic foundation contacts between the bodies, attached the discs to the jaw with 14 ligaments and introduced 22 muscles. My object is to obtain muscle activations for desire jaw trajectorie.

My calculation algorithm:

1. Scaling - my model was obtained bby 3D scanning and all mass properties were calculated with MSC Marc so I decided not to use scaling tool for my model.

2. Inverse Kinematics - it is not possible to experimentally mesure the movement of discs so I can't introduce virtual markers for discs movement. As a result the discs aren't moving and I've simplified the model only to jaw and crane.

3. I've made RRA but there are no external loads in my model so only usefull part of RRA tool is smoothing the jaw movement trajectories obtained with IK.

4. CMC tool, I think that I have well understood the purpose of reserve actuators.

However I'm performing the same simulation that was previously performed on LIFEmod. There all the forces were sufficient for performing the movement so I don think that I should introduce any changes of muscle forces.
I'm also getting very big muscles forces which amount to 1000 N when they should have about 1- 3 N.
I've impression that I may be introducing the actuators in the wrong way and these big forces are a result of that.

And now questions:

1. Is OpenSim a software dedicated for other simulations then gait movement? I'm working with your software for about 6 months and all the tutorials in Users Guide concern gait also all the papers that have been written about OpenSim application into skeletal modelling concern gait (Thelen 2003,2006). I've found a paper on your site describing the RRA function and it is also written I think only for gait simulations. In theoreticall part of RRA and CMC describtion there is also only gait model.

2. Do you have any kind of models of complicated joints? In my simulation there is a conctact during all time of the simulation, is it correct for OpenSim? Were these kinds of simulations considered by you?

[Andrea Gargamelli]

Hi Emil,
With this software you can do several simulation:running, jump, gait, anly arm movements and so on.

This is my Bechelor degree thesis where I simulated a forward squat jump. I hope that you can find interesting information.

Greeting
Andrea Gargamelli

[Ayman Habib]

Hi Emil,

Although the examples posted use gait analysis (primarily because our lab uses OpenSim for gait research) but there's nothing built into the software that is gait specific. There're many example models that are not gait related (e.g. the webpage below contains links to upper extremity, human and animal models as well as many other educational models)
http://simtk-confluence.stanford.edu:80 ... tal+Models
Also simtk.org has more OpenSim models contributed by users.

Keep in mind that creating a model that is simple enough to analyze yet physically accurate to simulate the activity you want to study may not always be a simple task and could easily be a full master or Ph.D. thesis, particularly if you collect data to validate your model/simulations.

Please post specific questions you have about your model and the user community can point you to specific examples that will help you out.

All the best,
-Ayman

[Emil Walczak]

Hello Ayman and thanks a lot for fast reply!


Generally this model is a subject of my Master degree.

As I have written I am simulating jaw movement with two temporomandibular joints (TMJ).
However, the main purpose of my work is to verify if the OpenSim software is capable to replace LIFEmodelere since the second one is no longer developped.

It is actually funny that you wrote about Phd becouse I am working with a model which was actually used during thesis and that is why I am sure that the model should work with my properties.

Now introduction into model. There are many models of jaw with two TMJ, one of them was even simulated in OpenSim https://simtk.org/home/jaw but the main difference is, that in this model there are no articular discs and they are replaced with contraints.

Also in my model there are no standard joints like ball-socket joint since I am modelling the joint.

I've wrote my model with C++ without any bigger problems, they began when I started using calculations tools.

Since there are no markers for the discs (we can place them on the condyles but not on the discs), the discs are not moving during IK.

That is why i decided to perform the first part of simulation IK ->RRA ->CMC with simplified model and use complete model only for forword dynamics.

So now we have model only with skull (defined as a ground) and Jaw. We have also 22 ThelenMuscles.

I have actuated 6 DOF of the jaw with residual actuators with optimal force equal 1N and excitations of range <-300,300>. But I am not obtaining the good results. There is almost no muscle activation but the residuals are active, I dont understand that since the residuals should help the model to perform the movement if the muscles cannot generate the sufficient force.

I've tried to run the simulations without residuals but OpenSim gave me the typical error about lack of the force.