tangent building

From: Yong Zheng (yzh@pris.EEAP.CWRU.Edu)
Date: Mon May 12 1997 - 22:21:09 EEST

Hi, Mr. Jager:

We have been working on tangent laser cutting for RP. For cutting
trajectory generation, we adopt a similar approach -- approximating
the layer edge-surface with ruled surface. we do the segmentation
in both the horizontal and vertical directions by analyizing the CAD
model, thus adaptive slicing (using layers of variable thicknesses)
in the vertical direction and piecewise ruled surface along the layer
edge are used.

Am I correct that you and Ray are evaluating the error by comparing the
curvatures or normals between the approximating ruled surface and the
original layer edge-surface? As pointed out by one of the responders,
surface normal and curvature are a local properties at a point. How are you
using them to evaluate the goodness of the ruled surface on the entire
layer edge surface?

Branching in topology in the building direction in one of our major
problems. If we don't have the luxury of continuous variable layer
thickness and want to approximate a layer with branching, what will
be the optimized approximation? I cann't see a definite solution to
this problem.

A couple of papers about our work were/will be published (one in SFF 96
in Autin, another in European Rp&M Conference 97). I appreciate it if you
would point me to your paper (you mentioned) or any other related



This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:39:34 EEST