PDA

View Full Version : Solid Edge, or Solid works?



Dr Claw
06-14-2004, 02:53 PM
Hey, who uses these? what are your thoughts on them? We have 1200 seat of SE at our school and 10 SW licesnse exclusively for formula.

having used both for the past 5 years (Since my sophmore year in high school), what do you like about them? even though solid works currently pays my bills, i personaly am a bigger SE fan. I cant wait for SE v16 to come out because it has a whole suite for doing spaceframes, which was it's major fault for FSAE (it miters tubes and everything apparantly.)

what other design software do you use for formula btw?

Charlie
06-14-2004, 05:03 PM
We use Solidedge. Its good IMO because it is very easy to learn, and fairly powerful. I am not too familiar with Solidworks, but it seems to be used more in industry than SE.

If you've used both, why not tell us what's better? http://fsae.com/groupee_common/emoticons/icon_wink.gif And hows does Solidworks pay your bills, I want in on that. http://fsae.com/groupee_common/emoticons/icon_smile.gif

Denny Trimble
06-14-2004, 06:40 PM
Our school uses SolidWorks, and it's great for us. Easy to pick-up, and powerful enough to work for our 2000-part full vehicle assembly, if you set it up right. We use configurations a lot (left/right suspension assemblies, fastener models driven by design tables, etc). Integrated FEA / CFD / Dynamics simulations help too.

We export our wireframes (spaceframe) or surfaces (monocoque) to ANSYS or Algor for beam & composite shell element FEA. Machined parts & small weldments are analyzed in Cosmos/Works (integrated).

We also have a few design tools setup that use Excel to move SW assemblies around, record dimensions, and generate plots. I'm sure you can see how this is useful http://fsae.com/groupee_common/emoticons/icon_smile.gif

I haven't used SE, so I don't know how it compares. Will this new version do beam element FEA on spaceframes?

Dr Claw
06-14-2004, 07:31 PM
it pays my bills because TRW pays me to do crazy stuff in it http://fsae.com/groupee_common/emoticons/icon_smile.gif. going through the design lifecycle sucks ass IMHO using SW...drafting things is just a pain (compared to SE. both are a dream compared to UG and CATIA though..) and 1/2 of the "advanced modeling" features in SW dont even work. they wont work for another two releases either and that is just how SW ishttp://fsae.com/groupee_common/emoticons/icon_razz.gif

I wouldnt trust cosmos FEA as far as i could throw it though...i believe it is only good (say 95% accruate) for displacement. your other readings are 70-80% reliable or so.

as for FEA portability though, SE works with both design space, and algor. like i said, the only thing that held SE back was it's non-easy space frame setup though, there are simple ways arround that...just no one at my school was willing to do it as such. in short, it couldnt do the 3D points like you can in SW, you would have to intersect two surfaces to get your composite intersection line in 3 dimentions.

in the new version though, i was told that it "assigns cross sections for a length of tube, and then miters your joints for you" or some paraphrased version of that. i dont know how redily 'known' this info is, but that is waht was handed my way from a very reputable source.

For prismatic parts, and general assemblies, they are equals. for sheetmetal parts, and industrial design (organic type shapes) SE hands down, every day. i'd also pick SE drafting over SW, because things are a lot more customizable, and the things actually work. The only good thing about SW though, is it has Photo Works. its always fun to add the bling, to your thing.

It seems that SE's only problem (outside of photoworks) is the lack of market share..and its a shame i cant get another job using it.

Denny Trimble
06-14-2004, 07:48 PM
Do you have any examples (model files?) of this inaccuracy in Cosmos? Anything at all?

Dr Claw
06-14-2004, 09:09 PM
not really, just TRW guidelines. any stress, strain, or anything else really, we need to send to our FEA guys. Besides, SW is more of a designer's program and COSMOS is intended to give a designer a clue as to how his part will react in certain cases. he then gets an oppertunity to fix the blatantly obvious ones and must leave the finding of the local minima to the FEA experts.

i do remember one time thouhg, me, and LTU's driveline guy tried putting a torque on a stubshaft. it did something that can be described as "spiraled out". it was like you were spinning pottery or something, and the splined surface flayed outwards in a cone type shape. it was crazy. who really knows if we were doing it right, but we fixed holes, and applied torque over the temp axis; that to me though, is a correct execution of FEA.

Greg H
06-15-2004, 12:50 AM
I've always been a fan of autocad and I like SW because of the ability to copy and paste sketches directly from autocad. Is this possible in other programs?

Also, we were also trying FEA on our splined axle and we got the cone thing. The really disturbing part, however, is when you animate it. I'm just going to say do it for yourself and leave it at that.

Denny Trimble
06-15-2004, 01:24 AM
Yeah, any time you model a bar in torsion it will do that. It's because it's exaggerating the displacements to some insane amount by default, just so it's large enough for you to see.

I remember one of our guys freaked out because he ran FEA for the first time, and he could see his part bending on the screen. "No way that's gonna work on the car, man!". Turns out, the numbers are what matter, and the visualization helps you digest it.

Our physical testing on anti-roll-bars this year was within 5% of the FEA results. That's on a thin-walled part that a solid mesh isn't supposed to accurately model. I'm confident it's good enough for us.

Frank
06-15-2004, 03:20 AM
I have used both a fair bit

both are about even, SE has better surface modelling capabilities

both seem fairly stable when using links.... try using "INVENTOR" now that sucks

the most important difference I can remember, is that SE supports "disjointed bodies" in part files, while SW does not

In my mind, this makes SE a heap easier to use (think chassis model)

You guys spinning out about COSMOS.. do you realize that you cant add more than one "face constraint" without making "artificial stiffness"? ...... This is the most common mistake, which I keep seeing

Frank

Denny Trimble
06-15-2004, 09:49 AM
As of the 2003 version, SolidWorks does support disjoint bodies in a single part file.

Yes, it's easy to overconstrain a part in any FEA program. There is some skill involved in doing any analysis correctly.

I keep hearing people knock on Cosmos. What's the magic program you're all using that a monkey can use and get better results?

Rob Davies
06-15-2004, 10:10 AM
ANSYS is what we use and its pretty easy to get the hang of.

Apparantly ANSYS has been coded so that it can deal directly with tubular structures really easily rather than cosmos cutting them up into thousands of tetrahedrals or whatever

Not too sure myself just what I heard on the grapevine...Perhaps someone else will be more able to explain

Rob

Denny Trimble
06-15-2004, 11:42 AM
Yeah, ANSYS and ALGOR (and Cosmos Geostar) can do linear elements (beam / truss), that make spaceframe analysis practical. Previous teams here tried to do a shell mesh on a tube frame... bad idea. 30hr run times, with 30% crash rates. Compare that to a beam element analysis that takes 30mins to set up and 2 mins to run.

gug
06-15-2004, 07:02 PM
ansys is easy to get the hang of?

arrrrgh! im never going to be able to do fea! http://fsae.com/groupee_common/emoticons/icon_frown.gif

i normally wind up hitting my head against the keybord when trying to use ansys...

Dr Claw
06-15-2004, 10:36 PM
<BLOCKQUOTE class="ip-ubbcode-quote"><font size="-1">quote:</font><HR> ansys is easy to get the hang of?

arrrrgh! im never going to be able to do fea!

i normally wind up hitting my head against the keybord when trying to use ansys...
<HR></BLOCKQUOTE>

i've always held that it takes a special breed to do actual FEA. that is not knocking a person that does FEA once or twice in his/her life for FSAE, but i mean to do it your whole life. Some people just arent cut out for it, and the ones that are....are both deadly strange and weird. we have a FEA guy at my work that works 20 hours a day, sleeps at work (showers at our gym..) and has police tape arround his office barring cleaning ladies from...well, cleaning and vaccuming and stuff. strange bird.

not saying that that means you cannot, beyond a shadow of a doubt, be weird, but just my own personal thoughts on lifelong FEA guys http://fsae.com/groupee_common/emoticons/icon_razz.gif

Frank
06-16-2004, 02:02 AM
you would have to be nuts to 3D mesh a spaceframe.. linear elements.. the only way

http://www.uq.edu.au/fsae/strand7.jpg

Rob Davies
06-16-2004, 02:13 AM
ansys is easy to get the hang of?

arrrrgh! im never going to be able to do fea!

----------

Make a search for University of Alberta ANSYS tutorials.. They are what got me going

Rob