Boundary conditions for rigid bodies 2

[NataCalc]

Hello,

I have transformation matrices generated via alignment of rigid bodies in the loop. These matrices consist of rotation and translation. The alignment was done using ICP algorithm ''PointtoPoint'. Bodies which were aligned were not identical in vertices and faces / point clouds. When I apply generated transformation matrices to two rigid bodies and visualise results in the loop, I see the movement which mimic my needs.

For FEA I use open source software, Febio. When I do apply same transformation matrices, displacement and rotation, to rigid bodies, they do not show positive trend in movement. My system unit is mm. In displacement vector there are values of 50 and 70 mm.

What can be wrong?

Cheers,
Nataliya

 

[FEA way]

So you perform alignment of point clouds in some non-FEA software (probably MatLab) and then you want to apply obtained transformation matrices as boundary conditions (prescribed displacement) in Finite Element Analysis, right ? What is the purpose of this FEA ? I’m afraid that the syntax of generated transformation matrix may not be compatible with FEA software. From what I’ve read Febio has some unusual way of treating rigid body rotations. Check its documentation and analyze how it works in this software.

 

[GregLocock]

Sounds like you may be having problems with the order of rotations.

https://en.wikipedia.org/wiki/Euler_angles

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[NataCalc]

Yes, I perform alignment of point cloud using pcregistericp of Matlab. This function does not give sequence of rotation ((( but allow to align not equal in vertices and faces bodies. The purpose of FEA is to mimic 4DCT movement of the wrist and calculate stress-strain distribution in soft tissue connecting rigid bodies.

The syntax of FEBio in rotation: (1) rotation unit is radian (2) convert rotation matrix to rotation axis and angle. By multiplying rotation axis x angle will give magnitude of rotation in Rx, Ry and Rz directions used in FEBio.

My rotation matrix generated from alignment is
ans =

0.9980 -0.0462 -0.0429
0.0459 0.9989 -0.0088
0.0432 0.0068 0.9990

Calculate magnitude of rotation in Rx, Ry and Rz directions, respectively, is
ans =

-0.0078 0.0431 -0.0461


Can someone check the math of this vector?

Regards,
Nataliya