simtk.org

Hello everybody,

I have found on a recent paper that the Flen (max normalized lengthening force) for young(~20 y.o.) people should be around 1.4 instead of the regular Thelen2003 value which is 1.8 and is more for older people (>40 y.o.). What do you think about that ?

Cheers

I think that muscle parameters in OpenSim are based generally for gait models and if you want to work with muscles much different than those used during walking you will need to adapt the muscle model properly. And btw. on which version of OpenSim you are working? My default Fl value is equal 1.4.

Hi Emil,

Thanks for your reply. I am working on Opensim 3.2. But I am using a model that has the Thelen2004 muscle model implemented with. Maybe you are using the newer Millard2012 model for muscle I guess ? Could you send me a copy of your model so that I can look at the values you have for the muscles ?

Thanks

Nope, I'm working also with Thelen, and I have been working a lot with on the influence of muscle properties on muscle behaviour. This model is well described in this paper:

http://www.google.fr/url?sa=t&rct=j&q=& ... 1959,d.d24

I can tell you that default OpenSim Thelen muscle properties values are dedicated for gait movement and in other cases you need to introduce different muscle properties.

I've attached two PDF's for you, frist one with Zajac work on which Thelen based his model and second one with the method for calculations of the passive force shape factor,I've also included a graph(shape factor file) that presents passive force with default passive force shape factor (4) and the one that I have calculated for my jaw movement simulation (25). You can see that with default value the passive force in muscle is almost 25 larger.

Maximum contraction velocity I've obtained from some publications dedicated for masticatory muscles.

Extrapolation treshold is used in normalized force - velocity factor calculations in Thelen muscle model, I've marked it in "treshold" file.

Both those factors have large influence on muscle behaviour and stability of the simulation.

Hello Emil,

Thank you very much for all these informations. However it seems that the pdf where you explain the method for the passive factor calculation is not in the attachement or am I mistaken ? Would you mind sending it again ?

I mean this file "second one with the method for calculations of the passive force shape factor,I've also included a graph(shape factor file) that presents passive force with default passive force shape factor (4) and the one that I have calculated for my jaw movement simulation (25). You can see that with default value the passive force in muscle is almost 25 larger."

Thanks a lot.

And concerning the complete description of the Thelen2003... it seems that the Flen is 1.8 (see page 3) in the paper and not 1.4. What do you think ?

Frederic

Hello,

I think that this paper is from 2013 and that may be the reason, also force-velocity factor is different in the paper and gausian force - length factor also. In my simulation I'm using the 1,4 value. Generially in Zajac publication you can find out that maximum normalized muscle force value should be smaller then 1,8. I've put in attachement also .png file with equation that I'm using to calculate the passive force shape factor.

Hey thanks again for all the information.

However I still don't get it how you made to chose the value for kpe. How did you know you could choose 25, because you would need to have already some value for the passive force to compare with and that is not really possible, is it ? In my case at least I will not be able to have passive force value to compare in order to choose a kpe value, so I don't know how to calculate it. Why was it choosen 4 anyway to begin with ?

Cheers

Frederic

Hello,

yes I have the passive forces that occurs in muscles when they have maximal length, I think that default value is equal 4 since the muscle is scaled for gait model, you can look in this Millard publication when he presents his model based on experimental results and you can observe that the passive force curve corresponds to this from Thelen muscle model.

Ok so from what I understood,

It has been chosen that at 1.5 * optimal fiber length, the passive force should be half of the max isometric force and that's why you need a k of 4. I think because we do not have any data on the individual passive force for each muscles used in the lower limb model, they decided to make a general behaviour of the passive force. It's a pity I can't find data for lower limbs muscles because it would have been interesting to adapt the model. In case anyone got any data on this, I'll be very interested.

Thanks,

Frederic