Message: Re: general questions regarding G4LogicalBorderSurface Not Logged In (login)
 Next-in-Thread Next-in-Thread
 Next-in-Forum Next-in-Forum

Feedback Re: general questions regarding G4LogicalBorderSurface 

Keywords: optical G4LogicalBorderSurface volume G4LogicalSkinSurface
Forum: Processes Involving Optical Photons
Re: Question Re: general questions regarding G4LogicalBorderSurface (Onno Kortmann)
Date: 10 Jan, 2007
From: John Apostolakis <John Apostolakis>

I believe that you have misunderstood, or are finding strange behaviour.

It is certainly not the intention that "solids [need] to overlap[] by a certain amount" in order to ensure that volumes touch.

However it is possible that the rotations in transformation could introduce numerical inaccuracies. For the relevant volumes, can you check the rotation matrices -- how did you create these ?

The problem that I envision, magnified many times would look like:
and it would be due to roundoff errors.

Can you give some more information regarding the problem that you observe ? Your description is very terse. What are the relevant volumes ? Boxes, cylinders, other ? How are the parts positioned that are expected to touch ?

I do have a drastic suggestion that may enable you to dramatically reduce the number of problems that you see. It is to chance the value of the constants kCarTolerance in the file geomdefs.hh (located in global/HEPGeometry/include) to a larger value. It is hard to provide a reasonable range for this, but for setups with typical dimensions down to 1mm, I would not make the tolerances larger than about a micron, and I would start a couple of orders of magnitude below that.

I note that the current value is very small:
 static const G4double kCarTolerance = 1E-9*mm;

In case you are dealing with cylindrical or spherical shells, then you will also need to change kRadTolerance and kAngTolerance in a similar manner.

I would expect that such changes will enable solids to respond that a point is on their surface if it is within the revised tolerance.

If this does significantly improve the behaviour of your application, please do let us know.

Best regards, John Apostolakis

Inline Depth:
 1 1
 All All
Outline Depth:
 1 1
 2 2
 All All
Add message: (add)

 Add Message Add Message
to: "Re: general questions regarding G4LogicalBorderSurface"

 Subscribe Subscribe

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