# Not congruent definition of the knee rotation axis with the Klein Horsman measurement

For building a Femur COSYS that I can reconstruct in a FEA-Model I plan to use the mechanical proximal-distal Axis and also the cylinder axis of the femur epicondyle as descript by Klein Horsman.

By doing so I found a bug in the AnyBody Code.

Picture 1.

The y-Component is in the Horsman definition positive but in the AnyBody-Model negative. The black line in the knee joint should have the same axis as the cylinder.

Changing sign of the first RotMat in the “AnyRefNode CylCenter” about the x-Axis from 2.46 to -2.46 the problem is solved.

ARel = RotMat(…Sign*-2.46pi/180,x) * RotMat(…Sign-21.70*pi/180,y)RotMat(Cipi,x) ;

The same is true for the knee joint. As defined in the Horsman thesis the Knee Joint axis is:

Rotation axis X Y Z
Knee -0.53 -0.11 0.84

But with the AnyBody Code the ARel of The KneeJoint is:
ARel = {{0.8457278, 0, 0.5336145}, {0.06928131, 0.9915357, -0.1098042}, {-0.5290979, 0.129834, 0.8385694}}

The sign is not Correct of the ARel(1,3) entry.

ARel = RotMat(…Sign*-7.46pi/180,x) * RotMat(…Sign-32.25pi/180,y)RotMat(…SignAxisRotpi/180,x);

If I change the Sign of -7.46 to 7.46 the bug is solved.

Then the ARel Axes of the KneeJoint is:
ARel = {{0.8457206, 0.003505665, -0.5336145}, {-0.07339077, 0.99124, -0.1098042}, {0.5285551, 0.1320261, 0.8385694}}

This would be the same as the axis of Klein Horsman.

Hi rohneern,

Thank you very much for making us aware of this, we will look into this and include it in the next release of the repository.

Best regards
Søren