After a check in the documentation, the "anisotropie" parameter of GeomPlate_BuildPlateSurface means : "If the Boolean Anisotropie is true, the algorithm's performance is better in cases where the ratio of the length U and the length V indicate a great difference between the two. In other words, when the surface is, for example, extremly long".
Following the previous mail of Arnaud, I would say that Anisotropie, in the algorithm, is the physical capacity of the surface to be bent along U or V. In GeomPlate, we assign 1 to the Anisotropy. Tests showed that 1 was the best value in spite of the real computation result which is (Length of the Surface along V / Length of the Surface along U). If you want, you can insert it in GeomPlate_BuilsPlateSurface in ComputeAnisotropie() and test it.
About the next step, approximation, if the result is insufficient, try to reduce the tolerance.
Be carefull about your constraints (G1 and G0), they don't have to be opposed.
It is true that I should have specified where, in the documentation, you could find the answer.
GeomPlate_BuildPlateSurface is the base computation tool of the BrepOffsetAPI_MakeFilling class. In the BRepOffsetAPI_MakeFilling documentation, more details are given on the parameters (which are almost the same than GeomPlate_BuildPlateSurface). On the on-line documentation, check at:
Thu, 06/15/2000 - 04:00
Stefan,
After a check in the documentation, the "anisotropie" parameter of GeomPlate_BuildPlateSurface means : "If the Boolean Anisotropie is true, the algorithm's performance is better in cases where the ratio of the length U and the length V indicate a great difference between the two. In other words, when the surface is, for example, extremly long".
Regards,
Arnaud.
Thu, 06/15/2000 - 10:41
Thanks for your answer.
> After a check in the documentation, the
I couldn't find information about the "Aniostropie"-parameter in the "documentation". I c checked the online documentation and that's all I got: http://www.opencascade.org/doc/ModelingAlgorithms/cursur.htm#P_35_15
bye, Stefan
Thu, 06/15/2000 - 10:53
Hello Stefan,
Following the previous mail of Arnaud, I would say that Anisotropie, in the algorithm, is the physical capacity of the surface to be bent along U or V. In GeomPlate, we assign 1 to the Anisotropy. Tests showed that 1 was the best value in spite of the real computation result which is (Length of the Surface along V / Length of the Surface along U). If you want, you can insert it in GeomPlate_BuilsPlateSurface in ComputeAnisotropie() and test it.
About the next step, approximation, if the result is insufficient, try to reduce the tolerance.
Be carefull about your constraints (G1 and G0), they don't have to be opposed.
Thu, 06/15/2000 - 10:57
Dear Stefan,
> Thanks for your answer.
A pleasure.
> I couldn't find information about the
> "Aniostropie"-parameter in the
> "documentation". I c checked the
> online documentation and that's all I got:
> http://www.opencascade.org/doc/ModelingAlgorithms/cursur.htm#P_35_15
It is true that I should have specified where, in the documentation, you could find the answer.
GeomPlate_BuildPlateSurface is the base computation tool of the BrepOffsetAPI_MakeFilling class. In the BRepOffsetAPI_MakeFilling documentation, more details are given on the parameters (which are almost the same than GeomPlate_BuildPlateSurface). On the on-line documentation, check at:
http://www.opencascade.org/doc/ModelingAlgorithms/offset.htm#P_42_7
Regards,
Arnaud.
Wed, 06/27/2012 - 22:13
I know it's an old thread, but for the sake of completeness
The parameter Anisotropie has no effect, it is used only at one point.
See also Line 6533 at GeomPlate_BuildPlateSurface.cxx:
-----------------------------------
if (myAnisotropie)
{
//Temporary
return 1.0;
}
else return 1.0;
-----------------------------------
Regards
Alex