Solidworks, Catia and Pro-E

Does anybody have experience with Catia or Pro-E, how well they can be
used in conjunction with Solidworks, which would be better to "upgrade"
to from Solidworks, relative cost of each? My application is aircraft
sheet metal, where all bends have substantial bend radius, and few bends
are actually straight.

We once subcontracted some modeling to a Catia company, and I was deeply
unimpressed with the results.

Thanks,
--
J Kimmel

www.metalinnovations.com

"Cuius testiculos habes, habeas cardia et cerebellum." - When you have
their full attention in your grip, their hearts and minds will follow.

 

They probably only want low-end stuff <G>.

 

UG NX not taken in consideration?

 

In originally selecting Solidworks, I rejected Solid Edge. I didn't
know what "UG NX" was before your post, so I visited the web site and I
still don't know--the web site is a marketing fog bank.

I'll consider any comparable software. Solidworks used to be the "best
bang for the buck", now it's starting to look like "you get what you pay
for".

Any comments appreciated.

--
J Kimmel

www.metalinnovations.com

"Cuius testiculos habes, habeas cardia et cerebellum." - When you have
their full attention in your grip, their hearts and minds will follow.

 

John,

With aircraft surface skins, you're dealing mostly with sheetmetal "forming"
as opposed conventional bending. It's possible that both UG and Catia have
modules for this. Both of these programs originate in the aircraft sector,
and this remains there primary focus.

I don't know that either Solidworks, or Pro-E have an add in for this.

Regards

Mark

 

"Forming" requires non-linear analysis, if you are talking of
compound surfaces from flat stock, unlike conventional sheet
metal bending (which is limited to developable shapes).

Such surfaces may be easy to *model* (as formed) in
many systems ...... but non-linear stuff is in the realm
of FEA (see LS-Dyna as an example) and very specialized
software packages.

They may have been too.

 

Cliff,

Catia was developed by Dassault, originally, to model and CNC machine parts
for their fighter jets. (Mirage and others)

UG didn't really originate in aircraft, but alot of the advanced development
was done when they were owned by McDonald Douglas.

Regards

Mark

 

Hi John,
Solidworks will not form curved bends.

Catia will flatten curved bends. Most curved bends in aerospace are very
mild in curvature so Catia ignores the shrinkage/stretching in the flange.
if the contour will cause wrinkling most designers incorporate cutouts or
flute the flange at the inter rivet positions. e.g. at the leading edge of
wing ribs.

Catias sheet metal module is streets ahead of Solidworks, but so is the
price, the module is an expensive addon not included in the base package.

If we pattern curved bends we do it manually in Solidworks, ie flatten a
straight flange with the same bend and flange criteria and then offset that
distance from the curved spline to get the curved flat pattern.

In most instances sheet metal curved flanges are tried out on the tooling
first and any problems with wrinkling back engineered into the design.

I do hope you will put in a request to Solidworks, there are a lot of
smaller aerospace based companies would love to have this feature in
Solidworks.

 

We do much different stuff (hand-held computers), but have talked to Catia,
NX, and Pro/E about the problems we have with SolidWorks surfaces. The Pro/E
people haven't tried very hard to get our (small, five seats) business. The
Catia and NX people made an effort to show how their product could generate
our type of parts. We liked what we saw with Catia best and had them build
one of our most troublesome parts. They couldn't do it any better than
SolidWorks could in the two days on-site that we scheduled. They took it
home and are working on it. Still haven't gotten results back. I'm impressed
by their professionalism. The NX sales people started out pretty lame, but
got better when we pointed out how the Catia people were kicking their
butts.

I don't know how well any of the packages work with SolidWorks.

Jerry Steiger
Tripod Data Systems
"take the garbage out, dear"

 

ALL of the early CAD/CAM systems grew out of the need for CAM
programming capability.
Once you can plot a toolpath you can plot geomerty as well, if

United Computing.

Much came after EDS/GM bought them. Started (in part) by Dr.
Hanratty's software development long ago, as such things go these
days.

Sadly, much was based on UniAPT instead of APT-IV (IMHO).

http://www.plmworld.org/museum/

BTW, Please consider snipping a bit & bottom posting <G>.

HTH

 

http://www.eta.com/Products/Dynaform/
http://www.arup.com/dyna/applications/forming/forming.htm
Etc.

 

IF you keep SW uptodate you can probably easily exchange
ParaSolid data with UG.

 

If you were dissatisfied with your contractor's CATIA work, it was most
likely the contractor and not CATIA that was to blame.

I noticed some strangeness when importing CATIA sheetmetal parts to SW
and Pro/E. CATIA has a way of skewing sheetmetal surfaces that should
be parallel. These surfaces are off generally by .000005° or less,
but it is enough to "disqualify" a part from being treated as a sheet
metal part in Pro/E and SW. This error originates in CATIA, and is not
just a translation error.

I've used SW, Pro/E, and SW extensively. Currently using SW with
access to Pro/E Wildfire for legacy data.

If you have to deal with a lot of customer imported data that has no
features, buy UG, it will be worth it. UG has amazing abilities to
manipulate non-parameterized suface and solid geometry. The trick will
be finding a user savvy enough to do it.

 

More likely a failure to A) test date translation with a few simple
parts first to get it right and B) failure to provide clear & proper
instructions as to what was desired.

Those two are usually problems as vendors fear looking dumb and
clients often don't grasp it all in the first place (nor do they want
to work much with vendors that ask lots of questions),

IIRC IBM Mainframes (and any software that can run on them) have
options as to the number of decimal places and what they are optimized
for. Often this is accounting or transaction processing where .01
matters but .00001 does not. So they may be able to count trillions of
pennies really fast but not do really well on floating point
calculations (needed here).
But they can also be configured the other way IIRC.

I'd look to the CATIA install options chosen as the possible problem
first.

Not really amazing, except when compared to constrained parametric
modelers (if that's what you are used to). That places "natural"
limits on the users .... & what's easy to model.

 

Not a "dig", per se.

Not really.

Based on Dr. Hanratty's AD-2000 IIRC.

Some of that was (and probably still is) proprietary. See notes
about "foreign surfaces" and their evaluation.

I know ... adding the public capability to output ASCII
ISO-4343 (similar, anyway) CLDATA (version 9.0 IIRC)
was my "suggestion".

UniAPT is far more limited than the later APT-IV. Which
also ran on the DEC VMS platform.

There were many versions of APT long before the IBM 360
(circa 1967 IIRC). IIRC The first versions of APT (from MIT)
were circa 1952 and later such folks as CAM-I.

Perhaps you are only thinking of APT/360? DAC ARLEM?

It was needed. IIRC The GM C4 folks started adding that in ~1992.

Crossposted to alt.machines.cnc <g>.