Message: Re: HepRep and boolean issue Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

None Re: HepRep and boolean issue 

Forum: Visualization
Re: Question HepRep and boolean issue (Dennis Haggerty)
Date: Oct 28, 13:36
From: John Allison <John Allison>

This is not specific to HepRep. It could happen for any vis driver that uses polyhedra to represent the Geant4 solids. (RayTracer is immune to this because it uses Geant4 tracking to create the image.)

Generating the polyhedron of a Boolean solid is a tricky business. The polyhedra of each component are generated, then each facet of the first is intersected with each facet of the second to generate the resultant polyhedron. This is computationally very tricky and subject to rounding errors that occasionally cause unphysical results. In your case, the difference between 8.486635*mm and (Side+Ring/2.0)*mm (they may well be slightly different, again due to rounding errors) is enough to trigger this.

The situation arises particularly if facets coincide, or very nearly coincide.

We are seeking effort to improve the robustness of the BooleanProcessor but meanwhile our recommended workaround is to avoid coincident facets by deliberately shifting one of the components by a small amount, say 1.*um, that will not affect the physics of your simulation.

Sorry not to be more helpful.

 Add Message Add Message
to: "Re: HepRep and boolean issue"

 Subscribe Subscribe

This site runs SLAC HyperNews version 1.11-slac-98, derived from the original HyperNews


[ Geant 4 Home | Geant 4 HyperNews | Search | Request New Forum | Feedback ]