On Mon, 24 Feb 1997 Nkin@aol.com wrote:
> Ray Hope wrote
> Problem 1.
> Joins between two or more surfaces, and vertices can cause the error
> approximation to give incorrect results. Previously published work (by
> others) has tried to solve this problem by slicing the part so that the
> surface joins coincide with layer joins. However this can only work if the
> surface joins are in the same plane as the layers. In many cases where a
> part is defined by two intersecting surfaces, the intersection curve is not
> in the layer plane. So what do we do?
> Of course your slices could have any edge shape and could be assembled to
> create any requred form. Today, however, I assume the "adaptive slices" you
> contemplate are limited to cross sections which have edges which define
> "compound curves"- which could be visualized as having been cut by a laser
> beam (a straight edge, at a varying angle from vertical).
Yes, we are using a five-axis waterjet to cut layers.
> If this is correct, it may be helpful to look at a relatively simple example
> before getting too involved with software details. Consider modeling or
> fabricating a threaded shaft (simple "triangular threads"), with layers
> sliced perpendicular to the shaft . Such a threaded shaft is just one
> example to illustrate the general need to be able to create sharply defined
> details (in addition to flat planes) at ANY elevation.
Good example of the problem.
> To minimize "stairstepping" and enable such general "Z" capability, how can
> one avoid either: 1.) using fine layer thicknesses (perhaps varied
> according to the needs of the particular geometry) , or 2.) resorting to even
> more complex edge shapes?
Lets consider a threaded bolt. In the threaded section the best results
will be achieved with fine layer thicknesses, but for the head of the bolt,
and the unthreaded section, much thicker layers can be used with no loss
of detail. My problem is mainly in finding a robust procedure to select
the best layer thickness in the threaded section, to keep the error within a
specified tolerance. In this case, using the curvature to estimate error does
not work as the triangular threads have zero curvature. So what do we use
in this case as the criteria to select the layer thickness?
This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:39:22 EEST