Rather obscure division by zero in triangleFuncsintersectAxesBundle

gschaider · August 12, 2008, 08:42

Version: OF-1.5

Description: While reading in a STL into snappyHex it crashes with a division by zero in line 142 of meshTools/triSurface/triangleFuncs/triangleFuncs.C

The cause seems to be these parameters for intersectAxesBundle
V0

-31.71428 0 -15.10714) V10

-1.285715 8.99165e-16 -1.142858) V20

0 0 -1.678573) i0:0

For these parameters the local variable det = 8.9e-16*1.67 > SMALL (not by much ;) ) but u1=8.99e-16<SMALL and the routine then tries to divide by u2=0

I know it is a one in a million chance that this might happen, but it happened. I propose the following patch that tightens the restriction on det (and fixes this):

triangleFuncs.patch

Bernhard

mattijs · August 12, 2008, 19:01

Thanks - I put it in. Does the mag(u1)<SMALL need some non-dimensionalisation as well?

gschaider · August 13, 2008, 08:08

It might be good for consistency but is in my experience not necessary.

Bernhard

August 12, 2008, 08:42	Version: OF-1.5 Description	#1
gschaider Assistant Moderator Bernhard Gschaider Join Date: Mar 2009 Posts: 4,225 Rep Power: 51	Version: OF-1.5 Description: While reading in a STL into snappyHex it crashes with a division by zero in line 142 of meshTools/triSurface/triangleFuncs/triangleFuncs.C The cause seems to be these parameters for intersectAxesBundle V0-31.71428 0 -15.10714) V10-1.285715 8.99165e-16 -1.142858) V200 0 -1.678573) i0:0 For these parameters the local variable det = 8.9e-161.67 > SMALL (not by much ;) ) but u1=8.99e-16<SMALL and the routine then tries to divide by u2=0 I know it is a one in a million chance that this might happen, but it happened. I propose the following patch that tightens the restriction on det (and fixes this): triangleFuncs.patch* Bernhard __________________ Note: I don't use "Friend"-feature on this forum out of principle. Ah. And by the way: I'm not on Facebook either. So don't be offended if I don't accept your invitation/friend request

August 12, 2008, 19:01	Thanks - I put it in. Does the	#2
mattijs Senior Member Mattijs Janssens Join Date: Mar 2009 Posts: 1,419 Rep Power: 26	Thanks - I put it in. Does the mag(u1)<SMALL need some non-dimensionalisation as well?

August 13, 2008, 08:08	It might be good for consisten	#3
gschaider Assistant Moderator Bernhard Gschaider Join Date: Mar 2009 Posts: 4,225 Rep Power: 51	It might be good for consistency but is in my experience not necessary. Bernhard __________________ Note: I don't use "Friend"-feature on this forum out of principle. Ah. And by the way: I'm not on Facebook either. So don't be offended if I don't accept your invitation/friend request

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
About flow division	sunnysun	OpenFOAM Running, Solving & CFD	5	March 2, 2009 13:34
Division by zero	vitke	OpenFOAM Running, Solving & CFD	5	September 1, 2008 06:35
Division by zero in Xoodles	hannes	OpenFOAM Bugs	3	August 4, 2008 12:04
ScalarField division	maka	OpenFOAM Pre-Processing	2	August 27, 2007 06:10