Muscle Specific Strength


I have been using AnyBody from the last few months only and I am still somewhat new. In the last few days I have been trying to understand the muscle strength data of the software and compare the results with real data. Specifically, I have been doing some simulations with the shoulder model to compare MVC values from AnyBody (at MMA = 1) and from literature. My problem has been that some of the cases (cases being different directions of force) match quite well, while some are quite far.

I am using AnyBody version 6.0.2 with AMMR 1.6.2

My question is: what kind of strength percentile does the shoulder model in AnyBody represent?

In the last days, I spent quite some time trying to understand the 3 element model used in AnyBody. While, I did not find the detailed description (I mean equations) describing the behaviour of the 3E muscle model in AnyBody, I tried to understand the different parameters used for the modelling. The parameter I am interested in is F0 (ideal muscle strength; strength at neutral fibre length).

In the 3E muscle model, F0 = Strength ScalePCSAMuscle Stress

Correct me if I am wrong, but, if I don’t use any scaling, Strength Scale = 1 for all the shoulder muscles. PCSA data has been obtained predominantly from cadaver studies (VU and MAYO, and a few others). While Muscle stress = 90 N/cm^2 for all the shoulder muscles. Unfortunately, I could not find the source for the value of muscle stress = 90N/cm^2. I tried to search the forum and came across a couple of threads which have dealt with this topic before:

Finally, I could find the literature reference for the muscle stress value, but, it has still left me confused as to what strength percentile does 90N/cm^2 represent. Are we talking about a young adult or an old adult? While, the anthropometric data of the standard human template is “around” 50th percentile European male, what about the strength aspects?

I am sorry for re-posting about topics already discussed. But, I am still not clear about this topic.

I would greatly appreciate if you could provide me with some literature references on this topic as well.


Hi Dave,

The data represents cadaver data so a specific subject, it may not represent a specific precentile in this respect.

If you measure the elbow flexion strength in the model it is around 40Nm for 90 deg angle, if compared with for example this paper

This is on the low side since the paper says around 60 Nm for young males.

If you want the model to fit a specific subject then it needs to be scaled to fit, this subject.

Best regards

Hi Søren

Thank you for your reply. Indeed, in my data as well, the MVC results have been lower in the simulations (but, also higher sometimes).

You wrote about scaling the model to fit the subject. I know that there are various scaling laws which also affect the strength of the model. Could you guide me some more about how to match, specifically, the model’s strength data with my subject? What kind of measurements do I need from the subject and how should I implement it in the model (perhaps, using the define arm strength index)?

From the experimental data I saw, I do realize that human strength is very variable. It’s not as simple as saying that A is stronger than B. The subjects tend to have different directions in which they are stronger. Which brings me back to my question again. When I try to match the strength of the model to the strength of my subject, how could I take into account this variability due to directions as well, besides, the absolute value itself?

Thanks and regards

Dear Dave,

Sorry for late answer.

This is a kind of hard question for us to answer.

There are some hardware devices which can measure muscle strengths, for instance, Biodex isokinetic strength test system.

I would rather to test the maximum joint torques of some human joints such as elbow flexion.
Then you can run AnyBody simulation for simple elbow flexion motion with appying that torque on the elbow joint.
And you can see how much muscle activation will be.

Then we can adjust the arm strength index until the maximum activation of the muscles will not exceed 1.0.

I tried to simply the case but there may be more works than I expect.

Best regards,

1 Like