pro/mechanica vs ansys workbench 10

[gpaladin]

Hi

SO far I have worked in ansys WB10 and tried to work in pro/mechanica.

In pro/mech I couldn't get any results. The solver would brake and leave me with some cute number like 30432.

I would be pleased to hear your opinions.

 

[mrfrog]

We had an ANSYS and Pro/Mechanica rep come in and show us both products. They said that they liked the ANSYS better, although it was twice as expensive.


Pro/Mechanica:


For the most part, it was a simple analysis package for a linear CAD system. It used a nodal analysis to do the calculation. For large deflections, or when the nodes don't line up like they did originally, the calculations are wrong. Also, the calcuations take a long time to generate. Another bad aspect is that in order to do the analysis correctly the design had to change. For instance, Pro/E will lump together parallel surfaces at the same elevation. If you had two parallel surfaces with different loads, you had to shave off .001" off of one of the surfaces in order to do the calculation, but now your model is messed up.


ANSYS:


The environment is really nice for this. ANSYS and Pro/E talk back and forth, so that changes made in one will update real-time to the other. It will also generate analysis reports automatically. They revamped the analysis, so it'll spit out the correct answer. Consequently, this nodal analysis thing is a problem for Pro/Mech and COSMOS.


Again, the issue is how much do you need. The rep left us with this. If you are doing a simple analysis, or if you are using it every once in a while, Pro/Mech is the way to go. If you were good having a rough estimate, Pro/Mech works. However, anything beyond that would require ANSYS. The issue is the cost. Pro/Mech, the base package (no thermo) was half that of ANSYS, which had thermo and vibration as well.

 

[Remmerv]

Hi mfrog,<?:namespace prefix = o ns = "urn:schemas-microsoft-comfficeffice" />


I must say that the MECHANICA REP did a really bad job if he left you with this opinion.
MECHANICA is a easy to use but powerful package which is integrated in Pro/ENGINEER like no other package. All MECHANICA model data is saved in the Pro/E .prt and .asm files so with your model you have always access to your data. All Pro/E parameters are available for optimization in MECHANICA and you can use Behavior Modeling to perform very complex optimization studies.
Since PTC decided to split MECHANICA horizontally in Basic and Advanced STRUCTURE and THERMAL (previously separate full Structure and full thermal), you may need the advanced version to do some more advanced stuff like large deflections. For large deflections you need to use the Large Deformations option in your static analysis. Non-linear contact analyses are already possible within the basic package.
For large steel constructions like cranesMECHANICA has the very useful midplane compression in order to create shell elements. I've used this more than once to perform static and prestress-modal (to tension the cables modeled as beams) analyseson complete container and off-shore cranes.


You can also easily split a Pro/E surface into MECHANICA Surface Regions to partially load or constrain a surface. You can alsouse volume regions to isolate a piece of a part.For accuracy and speed to produce a correct answer we have absolutely no problems in taking on Ansys when linear material behavior is applicable. Only when needed we transfer the model to MSC.MARC to do the non-linear stuff we cannot perform with MECHANICA but even thenwe like to use theMECHANICA results to check the MSC.MARC analyses (at least the first linear part of it) for the correct convergence/mesh.


Your impression shows that the software is only as good as the guy who is using (or in this case "showing" it) and from our experience (users and REP of the total PTC software line) I can tell you that your MECHANICA-REP visitor should best be looking for another job.I'm not sad whena potential customer chooses for Ansys based on facts and specific requirements (we like to build longterm relations with happy customers), I'm only disappointed when this happens on incomplete information.

 

[gpaladin]

Those are some interesting points of view.
I know of MECHANICA potentials but I am unable to use it like it should be used. Therefore I use ansys.
Since my company invested in mechanica I would be wery happy if I could use it appropriately.
The large problem I encounterd is inability to repair the model in order to get some results. It seems that when mechanica is ran as integrated part of pro/e that i cannot look where the problem has occured. Or I just do not know how to do it.
In any case it seems that in order to make some assembly analisys I need to learn more about MECHANICA.

 

[Miko]

Hi all and thank you very much for basic info about this powerfull softwares. At this time I am learning the basics of FEA and any tips or opinions are welcome... Thanks

 

[Remmerv]

If you install Pro/E with the Structural and Thermal Simulation option be sure to install the HELP files also. PTC has put a lot of information in the HELP files, including illustrations. As a PTC customer with a maintenance contract you can also access the PTC Knowledgebase on their website which also includes a lot of HOW TO's, suggested techniques and background information on the MECHANICA P-elements.Also a very nice way to learn the package is to take a PTCU (PTC University) subscribtion for one or 2 years (2 years is approx only 30% more expensive than 1 year!).The creation subscriptiongives you access to all PTC creationsoftware (Pro/E and almost all its modules, including Pro/MECHANICA!). This web-based classhas the same content as the class room course but you can do it in your own pace. I do not know the USD price but based on our Euro price (about 730 euro) you should be looking at 600-700 USD for the two year subscription which is a absolute bargain for the stuff you get access to!

 

[slashct]

Check out this thread as well "Is Pro/Mechanica a good Analysis Tool..


[url]http://www.mcadcentral.com/proe/forum/forum_posts.asp?TID=29 740&KW=slashct[/url]


I have used Mechanica from when PTC first purchased it, we were looking at the software before PTC purchased the company. As well I have used Cosmos, I found that Cosmos is good for a few things, it a has come along way, and the reason we went to mechanica was for additional functionality (years ago).
Any FEA software is as good as the person driving it, each has limitations and you have to know how to set up the problem. Everyone has their personal preference, I have worked in a large company that had Ansys, Ideas, Mechanica,and Patran. Each group used something different, but the results were all close, it depends on the accuracy you need. If you need to get inside the ball park Mechanica is a good choice, if your doing precision optics and other high accuracyitems I would look at Patran or ansys. Mind you thesemuch more intensive and require quite a bit of training, also the cost is quite a bit more.
On a second note, we had a part that someone spent a week in Solid Works and Cosmos trying to run an analyze, cosmos would get stuck after 7 hours or so. Imodeled the part in ProE and than ran the analysis on it, took me some time to model the part, Mechanica did the analysis in just over 2 minutes. We have been pretty happy with Mechanica, and when it requires additional we have a license of Patran.

 

[Natewebb]

Workbench and mechanica are both good if you know how to use them. The mechanica rep mrfrog described obviously has never used mechanica,and neverattempted to learn how to use it. Nothing in his description is accurate.


If he was a car salesman, he would be telling you the honda civic runs on diesel and gets 12 mpg downhill. Thats about how much he knows about what he sells.

 

[MechTeeth]

Resurrecting an old thread here.


I've been using Workbench for a few years (now on 12.1) and am playing with Advanced Mechanica aswe also own that. I went through the online 1 week training course and tested out as "proficient" but obviously don't have much experience (with Mechanica).


I took an assembly and ran it in ansys. As is typical in ANSYS, it was easy to set up and meshed and ran on the first try yielding satisfactory results (Assembly with contact).


In workbench, no matter my efforts, I couldn't even get it to mesh (same assembly, very similar initial settings). Fail.


On to the part. I took one of the parts, which - - again, runs easily in ANSYS and about 3 minutes to set up the study. It took me about 1 hour to get it set up in Mechanica such that it would mesh and run. Run time in ANSYS = 7 seconds. Run time in Mechanica = 58 seconds. Similar numbers of elements. Similar setup except that I let ANSYS apply weak springs automatically while I had to fix an edge in XY in Mechanica (I couldn't find any sort of weak springs option)


What am I missing here? I know Mechanica can't be this bad.... can it? I think I'd rather use ANSYS even with ProE.

 

[moriarty]

Perhaps you should put up a simple default model and
challenge the punters to do better than ANSYS - accuracy
speed etc.

Perhaps the sweetest part of Mechanica is its ability to
monitor its own convergence. How does ANSYS do this?

I would have thought Mechanica would also use a lot fewer
elements but similar DOF to achieve similar accuracy - as
its been said elsewhere its elements are as big as Texas.