Hexpress: failed feature line recognition

timfranke · January 7, 2014, 08:37

Hello,

I came across a phenomena and I wonder if someone could tell a work around.
Given a cube where one edge is not completely sharp since after
half of the edge length the sharp edge runs over into a radius
that increases from R=0 to R=max at the corner point.
After importing as stl eleven edges a shown a feature lines but the
single edge decribed above is absent, although I would expect
to see a feature edge that covers half of the cube edge.
From a feature angle point of view the featre edge should be there.

Regards,
Tim

Kaskade · January 7, 2014, 10:28

A picture would help understand the problem.

But I don't know if a feature edge can just stop, without connecting to other edges.

Do you have to use STLs?

timfranke · January 7, 2014, 11:52

Just imagine a cube where one edge radius develops from being 0 (sharp)
to a finite value (round). In Hexpress the whole edge is not marked as
a feature line although the first part is perfectly sharp.
If I read in stl the only way to detect feature lines is to evaluate the angle
of surface normals between neigbouring cells. If that angle exceeds a
given threshold then the common edge is a feature line.

Obviously this is treated differently in Hexpress leading to the described deficit.

Importing the geometry as Parasolid leads to the same failure.

Regards,
Tim

Kaskade · January 7, 2014, 13:44

How about splitting the cube at the point where the edge starts getting rounded?

January 7, 2014, 08:37	Hexpress: failed feature line recognition	#1
timfranke New Member Tim Franke Join Date: Mar 2009 Posts: 8 Rep Power: 17	Hello, I came across a phenomena and I wonder if someone could tell a work around. Given a cube where one edge is not completely sharp since after half of the edge length the sharp edge runs over into a radius that increases from R=0 to R=max at the corner point. After importing as stl eleven edges a shown a feature lines but the single edge decribed above is absent, although I would expect to see a feature edge that covers half of the cube edge. From a feature angle point of view the featre edge should be there. Regards, Tim

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Foam::error::printStack(Foam::Ostream&) with simpleFoam -parallel	U.Golling	OpenFOAM Running, Solving & CFD	52	September 23, 2023 04:35
whats the cause of error?	immortality	OpenFOAM Running, Solving & CFD	13	March 24, 2021 08:15
simpleFoam parallel	AndrewMortimer	OpenFOAM Running, Solving & CFD	12	August 7, 2015 19:45
Problems of Duns Codes!	Martin J	Main CFD Forum	8	August 15, 2003 00:19
error while compiling the USER Sub routine	CFD user	CFX	3	November 25, 2002 16:16

January 7, 2014, 10:28		#2
Kaskade Senior Member Onno Join Date: Jan 2012 Location: Germany Posts: 120 Rep Power: 15	A picture would help understand the problem. But I don't know if a feature edge can just stop, without connecting to other edges. Do you have to use STLs?

January 7, 2014, 11:52		#3
timfranke New Member Tim Franke Join Date: Mar 2009 Posts: 8 Rep Power: 17	Just imagine a cube where one edge radius develops from being 0 (sharp) to a finite value (round). In Hexpress the whole edge is not marked as a feature line although the first part is perfectly sharp. If I read in stl the only way to detect feature lines is to evaluate the angle of surface normals between neigbouring cells. If that angle exceeds a given threshold then the common edge is a feature line. Obviously this is treated differently in Hexpress leading to the described deficit. Importing the geometry as Parasolid leads to the same failure. Regards, Tim

January 7, 2014, 13:44		#4
Kaskade Senior Member Onno Join Date: Jan 2012 Location: Germany Posts: 120 Rep Power: 15	How about splitting the cube at the point where the edge starts getting rounded?