I'm having a frustrating issue at the moment. I set up a simulation (at the moment with basically no particles), and then import an ascii stl mesh (produced in meshlab) with the following line:
fix MT all mesh/surface file MT.stl type 1 rotate axis -3.832567 0.000000 2.832567 angle 17.365650
When I look at the output, the mesh has rotated, but it has also become larger. Am I doing something wrong here, or is this a bug?
tar attached containing full input script and original stl mesh.
[edit] sorry, should have said, this is compiled with make openmpi, and I'm running v2.3.2.
| Attachment | Size |
|---|---|
 meshtest.tar_.gz | 3.12 KB |
pfalkingham | Thu, 04/11/2013 - 20:39
Follow up
Ok, so it turned out I was normalizing my axis vector in a weird way, and doing it properly results in the mesh behaving as expected.
(the above fix should read
fix MT all mesh/surface file MT.stl type 1 rotate axis -8.041962e-01 0.000000e+00 5.943639e-01 angle 17.365650 move 0.139673 -0.084280 0.071767)However, it's still work asking if it's intentional that non-unit axis vectors scale the mesh?
ckloss | Fri, 04/19/2013 - 15:07
Hi Pete,
Hi Pete,
thanks for the note. The axis was unintentionally not normalized before calculating the roatation quaternion, I'll fix that in the next release
thanks a lot,
Christoph