Problem Parameters of Muscle

Main Forum
1

> Hi All
>
> I have a problem about anybody in the parameters of muscle and tendon.
>
> This link is attached to a common parameter of muscle.
> http://www.yousendit.com/download/MnFnclVNcklRR2VGa1E9PQ
>
> But I do not know the parameters in muscle Anybody in general is the belief in
the existence?
>
> If I want to change the parameters of the muscles of the National believes
Asia Possible?
>
> How can I know the value of those parameters? Determine how they are?
>
> I hope you will recommend.
>
> Thank You
>
> BestRegards
> Wongsakorn
> Cmerc Lab Thammasat University

2

Hi Wongsakorn,

The muscle parameters in the models come from cadaver studies carried out by various research groups all over the world. If you look inside the models you should be able to find comments listing the Journal paper references where the data is coming from.

We only have the data that is already in the models and getting anatomical data is generally the biggest problem of musculoskeletal modeling. If we had more data we would have included it in the models.

So I cannot offer you much help on this one except that if you ned muscle data for different populations then you must either locate them in the literature or perform your own cadaver studies.

Best regards,
John

3

Hi Johnr
[LEFT]My research study about knee that I do not know What value should be entered into the parameter tendon in Anybody software.

Now in my lab have tensile machine the properties of the material. The value that I expected from the experiment is stress-strain curve. I will put a tendon from a medical test, but I do not know process the converter stress-strain curve from an experiment into the Parameter muscle Anybody?

If you do not know how parameter of tendon or muscle to find.

Please introduce the literature review to me. On topics related to how to find the parameter values of muscle and tendon

[LEFT]If I can solve this problem. I think it will make my research much smoother.[/LEFT]

Thank you
Best regards
Wongsakorn
[/LEFT]

4

Hi Wongsakorn,

I presume we are talking about the three element muscle model, AnyMuscleModel3E, which contains a serial-elastic element that represents the tendon elasticity. Have you studied the tutorial on muscle models here: http://www.anybodytech.com/index.php?id=684 ?

This muscle model has a number of settings, whose standard values ensure a reasonable relationship between muscle strength and tendon stiffness. However, you can also modify these parameters and obtain other properties. Please refer to the tutorial for a description of the different parameters.

If you want to look further into the mathematics, of the muscle model, then you can find some references here:

Anderson, F. C. and M. G. Pandy (1993). Storage and utilization of elastic energy during jumping. Journal of Biomechanics 26(12), 1413–1427.
Bahler, A. (1968). Modeling of mammalian skeletal muscle. IEEE Transaction of Bio-Medical Engineering,
BME 15(4), 249–257.
Daxner, T. (1997, August). Simulation von beuge- und streckbewegungen des menschlichen knies mit dads. Ph.D. dissertation, Institut fur Mechabik der Technischen Universitat Wien.
Duda, G. N., D. Brand, S. Freitag, W. Liese, and E. Schneider (1996). Variability of femoral muscle attachments. Journal of Biomechanics 29(9), 1185–1190.
Gfohler, M., T. Angeli, T. Eberharter, and P. Lugner (1999). Dynamic simulation of cycling powered by lower extremity muscles activiated by functional electrical simulation. In XIIth Biomechanics Seminar, H¨ogfors C, Editor, pp. 101–122. vol. 12 (ISSN 1100-2247). Goteborg.
Hatze, H. (1977). A myocybernetic control model of skeletal muscle. Biological Cybernetics 25, 103–119.
Hill, A. (1938). The heat of shortening and the dynamics constants of muscle. Proc. Roy. Soc. B. 126, 136–195.
Huxley, A. (1957). Muscle structure and theories of contraction. Progr. Biophys 7, 255–318.
Kaplan, M. L. (2000, May). Efficient optimal control of large-scale biomechanicla systems. Ph.D. dissertation, Department of Mechanical Engineering, Standford University.
Kaufman, K. R., K.-N. Ai, and E. Y. S. Chao (1989). Incorporation of muscle architecture into the muscle lengthtension relationship. Journal of Biomechanics 22(8/9), 943–948.
Klute, G. K. (1999, October). Artificial muscles: Actuators for biorobotic systems. Ph.D. dissertation, University of Washington, Department og Bioengineering.
Neptune, R. and M. Hull (1998). Evaluation of performance criteria for simulation of submaximal steady state cycling using a forward dynamicmodel. Journal of Biomechanical Engineering 120, 334–341.
Otten, E. (1985). Morphometrics and force-length relations of skeletal muscles. Free University of Amsterdam.
Pandy, M. G., F. E. Zajac, E. Sim, and W. S. Levine (1990). An optimal control model for maximum-height
human jumping. Journal of Biomechanics 23(12), 1185–1198.
Prilutsky, B.I. and Gregor, R.J. (2000). Analysis of Muscle Coordination Strategies in Cycling. IEEE Transactions on Rehabilitation Engineering, 8, 362–370.
Purslow, P. P. (1989). Strain-induced reorientation of an intermuscular connective tissue network: implications for passive muscle elasticity. Journal of Biomechanics 22(1), 23–31.
Rasmussen, J., Damsgaard, M. and Voigt, M. (2001). Muscle recruitment by the min/max criterion - a comparative numerical study. Journal of Biomechanics em 34(3), 409–415.
Winters, J. (1985). Generalized analysis and design of antagonistic muslce models: effect of nonlinear muscle joint properties on the control of fundamental movements. Ph.D. dissertation, University of California Berkeley.
Winters, J. M. and S. L.-Y. Woo (1990). Multiple Muscle systems. ISBN 0-387-9730-9: Springer verlag.
Zajac, F. E. (1989). Muscle and tendon properties, models, scaling, and application to biomechanics and motor control. Critical Reviews in Biomedical Engineering 17(4), 359–411.

5

If I am interested in the topic of one ligament.
[LEFT]What number should I enter into L0, eps1, F1, a0, a1.[/LEFT]

AnyLigamentModelPol TestLigaModel1= {
L0 = 0.0214;
eps1 = 0.2;
F1 = 100;
// a0 = 1;
// a1 = 1;
// LinRegionOnOff = On;
};

Thank you
Best regards
Wongsakorn

6

Oh, I see. Your questions mentioned tendons several times, so I thought you were interested in those.

Anyway, there’s also a tutorial about ligaments available:
http://www.anybodytech.com/index.php?id=682

It explains in detail the mathematical assumptions and the physical influence of the different parameters. The model will give you a function for the dependency between length and force in the ligament, so it is quite accurately the same as the raw data coming out of the lab test of a ligament specimen.

Best regards,
John

7

Hi John

I'm sorry that I bother you again. :smiley:
I'm attachments power point that talk about test ligament problem.
:stuck_out_tongue:

Thank You
Best regards
Wongsakorn

8

Hi Wongsakorn,

Looking at your graph, it occurs to me that it is somewhat different from the basic behavior that we are trying to mimic with the ligament model in AnyBody. AnyBody’s ligament model presumes that there are three parts to the curve in the following order:

  1. A slack part. If the ligament is completely slack, then it is presumed that the stress is zero.
  2. A nonlinear part. This is where the collegen fibers of the ligament are rotating to become more aligned with the strain direction, and hence the stiffness is progressive in this region.
  3. A linear-elastic part. This is where the fibers are mostly aligned with the strain direction and the ligament works like an elastic.

Looking at your graph I can see that you have
a. No slack. You begin the measurement when the ligament is taut.
b. A nonlinear part with decreasing stiffness.
c. A part with negative stiffness.
d. A nonlinear part with increasing stiffness.

So your behavior is more complex than what was intended with the ligament model in AnyBody. You can still model your behavior by simply defining the ligament as an AnyKinSPLine or AnyKinPLine and then attach an AnyForce to the line. The AnyForce can be any explicit function of the line length that you type in, so you can simply work out your own polynomial representation of your curve and type it in. Notice that you have to change it from strain/stress to length/force.

However, before you do so, please consider what part of the curve you have measured is relevant to model. The part of the curve with negative stiffness indicates that significant damage has occurred to the tissue somewhere between parts b and c. So everything after part b may be a damaged ligament that may not be relevant for a model of a normal activity.

9

If I cut section B and C in graph. How I can read and convert values from data in graph (i.e. young's modulus, poison ratio ultimate stress from calculate from graph) into Anybody's ligament model in 3 parameter (L0, eps1, F1).I'm not sure that these values associated with ligament ?
Please answer me.

Thank You
Best regards
Wongsakorn

10

Hi Wongsakorn,

I’d like to help you, but you have to help me a little first. Please create and send an Excel file with the following information:

  1. Slack length of the ligament.
  2. A table of data from the tensile test. Please convert the stresses to Pascal (instead of PSI) and please include only the part of the graph you wish the model to contain.

Thanks,
John

11

Now I have not tested ligament.But I will test soon.This specimen ligament is limited.[LEFT]Now if possible I want to know graph. What is axis X and axis Y. Characteristics curve that I should have from the experimental for convert into 3 Anybody's parameter.
[LEFT]If I test to find stress-strain curve. It enough for you to help me?

Thank You
BestRegards
Wongsakorn

[/LEFT]

[/LEFT]

12

Yes, if you send actual data for the stress-strain curve, then I can help you set up the model if you cannot work it out on your own based on the explanation in the tutorial.

Best regards,
John

13

Hi John
Stress relaxation from testing ligament have influence to parameter of ligament ? If not influence to parameter of ligament I will not test in this section. Because it can be difficult.

Best regards
w. phakdeepinit

14

I attach the data. Please help me setup tendon parameter of anybody.

Thank you

Best regards
w. phakdeepinit

15

Thanks, those are good data. I can see that you load the tendons until failure, and you are probably mostly interested in the non-failing part of the curve, i.e. where the force(displacement) curve has a positive slope. Is that a correct assumption on my part?

I am asking because the modeling might be more complicated if you also need to include the descending part of the curve.

Best regards,
John

PS. Replies may be slow the next few weeks. Most of the guys in the research group and company are on holiday.

16

I interested in the non-failing part of the curve only.

Best regards,
Wongsakorn

17

Hi Wongsakorn,

I am enclosing a model that has two different concepts of ligaments that mimic your data. I have added comments to the code that is hopefully self-explanatory.

In the first approach I defined one of AnyBody’s ligaments with the polynomial ligament model just as in the tutorial. I went to your spreadsheet and looked at where the nonlinear part of the ligament elasticity stops and the linear part sets in. Then I recorded the strain and force correponding to that point and used them in the ligament model af F0 and eps0. I also set a0 = 0 to get a continuous transfer of force from the slacked to the tensed regions. This very easily created an approximation of your data.

For the second approach I am enclosing a spreadsheet (zipped) that I pasted some of your tensile data into and used for calculations. I did the following:
[ol]
[li]I copied only the data that corresponds to the ascending part of your tensile curve.[/li][li]Then I converted the data to SI units, i.e. from millimeters to meters.[/li][li]I added some artificial data to ensure that the data starts from zero rather than from the slack length of 0.025 m. The artificial data also ensures that the curve is belo zero for the entire slck part of the ligament length.[/li][li]I created a scatter plot of force versus displacement.[/li][li]I clicked on the graph and added a trend curve (third order polynomial) and asked to have the equation written on the plot.[/li][/ol]Then I went ahead and coded the function directly into AnyBody with a max function to ensure I cut off the slack part of the function.

This should illustrate quite well what you have to do.

Best regards,
John

18

Thank you for your suggestion.
File that you attached (Ligament.any) I can not run.
File that you attached (Fit curve in excel).What effect if I will fit curve polynomial 4 5 or 6 order.

Best regards,
Wongsakorn

19

Hi Wongsakorn,

The model probably fails to run because you are using an older version. But it is impossible for me to know unless you give me some information about the type of error you are getting.

A higher order polynomial will give you a closer fit to the data, but it also increases the possibility that the curve might oscillate. I usually prefer lower order functions.

Best regards,
John

20

Where I will load anybody software.
I use version 3.0.4.

Best regards,
Wongsakorn