Join our MCAD Central community forums, the largest resource for MCAD (Mechanical Computer-Aided Design) professionals, including files, forums, jobs, articles, calendar, and more.
I don't have ISDX, so do not know, but I'll offer a guess. There's no difference in the surfaces except as may be the result of having additional controls and options for creation. If it were me andI wanted to know I would create two surfaces using the same boundaries and constraints, export them and examine them in Rhino. I would expect them to be, more or less, the same; e.g. degree 3 surfaces with similar U, V spans, etc. If you want to create the surfaces and post them, I'll take a look.
- - - or - - -
Do you simply want to know what ISDX offers over and above Pro/Surface and are asking the wrong question?
Now I'm really curious (and really do need to get a copy of
SE so I can check it out for myself).
When I think "parametric" it means the surface conform to
changes in boundaries, maintaining defined constraints. Not
so with ISDX?
When I think "conceptual"; polygon mesh (as with
Pro/Concept?) or subdivision surface modelers come to mind.
I feel certain you don't mean that?
My assumptions re ISDX are that it provides extensions to
Pro/Surface functionality; interactive curvature graphs,
additional curve (and surface) creation and manipulation
functions, etc. Not true?
An "expert" told me that ISDX doesn't do anything that proSurface can't do. Example: Curve-on-surface (ISDX) can be done with with regular (point-to-point) curve if you change attribute from "free" to constrain it to a "surface".
From my experience, ISDX does some great surfaces on more bulbous shapes while ProSurface does better on more slab(ish) shapes. ISDX has better curve shape control. (Can you move an internal point on a regular curve by .001" in the Y direction in ProE?) And ISDX does make reverse-engineering easier by easily building curve on IGES files.
Since ISDX came from CDRS, I'd suspect the NURBS equations would be a bit different than ProSurface... But you never quite know how these bought-n-sold "deals" affect the parent software. To me, the bigger question is, "Do you see any difference in the tooling (molds) from one to another?"
ISDX surfaces are a bit easier tobuild than bnd blend. The real advantage iscreating curves.Much more control as mr koontz has said which results in much more control over the surfacecurvature. I would agree that ISDX ismore suited to conceptual modeling because its quicker, but there isjust as much if notmore parametric control available. The only time I use bnd blend over style feature is patching imported data. For some reason I haven't had much luck getting style features and import features to work well together.
ISDX surfaces ARE more controllable because they are created from freeform curves. they can be controlled parametrically by building datum curve skeletons as a feature prior to creating the Style surface. I read recently that to be truly accurate to a freeform surfaceideal NURBS surface must be to the order of 50. In reality most CAD packages only construct surfaces to the order of 3. This has generally always been acceptable although there have been reports of unacceptable gaps in some surfaceboundy edges. I think if your surfaces were created from identical boundaries using the same boundary conditions then you will find they are the same if you analyse in another CAD package.
PJW, it would be interesting to read the article you're refering to.
It's my understanding that successively higher orders (degrees) are necessary to constrain to higher orders of continuity (G1, G2, G3, ...).
I believe it is true that Pro/E will only create surfaces to degree 3 yet it will constrain G2. To constrain opposing boundaries G2 requires degree 5 and six CV's unless an additional knot is introduced by thesystem.
Accuracy of boundary coincidence is controlled by knot spans, e.g. where there are rapid changes in boundary curvature knot (and hence U or V) spans should be decreased to maintain edge coincidence.
It's interesting to point out where G1, G2 and G3 curvature continuity was derived from. It is analogous to position, velocity and acceleration over a boundary edge. If two surfaces share the same position then but not necessarily tangent then they are said the be positional. This is is referred to as G1 positional. If the two surfaces share the same curvature at the joint then they are said to be tangent or G1 (first derivative of position S or dS/dT). Automobile exterior designers will realise that if this type of bounday is used on a car, there will be evidence of a 'ripple' at the joint which would probably look like a flaw.
G2 curvature continuity is what most high end surfacing gurus will aim for and ensures that the acceleration over the joint (second order derivative of position d^2S/dT^2) remains the same. Higher orders only lead to the surfaces having identical derivatives to the Nth order furtheraway the seam. In other words the seam is 'invisible' to the human eye.
Just a bit of info really for anyone who wondered where the concept came from.
Thanks, it's always worth mention. I believe, tho', that you've shifted the scale. G0 is positional, G1 tangent, etc., each influencing a successive Control Vertex (or 'vertice'?) in the curve. Aninconspicuous boundary willbe G3 (curvature + one order of acceleration or rate). It's interesting to note that this ~can~ beaccomplished (the visual effect satisfactorily approximated)using degree 2 (conic)curves and G1 continuity. We just can't get the system to do it automaticallyor maintain it parametrically and it's often too time consuming or impractical to do it manually. It's easier to do using using G2 curvature continuity as a basis but curvature continuity, in itself, doesn't guarantee avisually pleasing transition acrossa boundary.
Oh, no prob! I have to dig out my notes whenever I wanna talk about this stuff, what with all the G's, C's, degrees, orders, cubics, quadraphonics, quintics, ...
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
