Discussion Forum

Hi

I do not agree, I find it much better from the methodology than the earler version, it's much more physics oriented, and less application oriented, what fits me well.
But I do agree that v4.0 is not a finished product (as for most software: never trust a X.0 release for client work, wait at least to X.1, but it remains good for training) and they have changed the approach sufficiently that you need some time to get used to it.
Having to start fully from scratch I agree is probably quite painful, I was lucky to attend the last automn Comsol conference and after two presentations on "how to" understand the new approach, I got rather entousiastic about it. And by now it feals quite natural, even I'm reluctant to return on 3.5.

But again, it all depends on your needs and approaches, and if you have had the opportunity to follow some of the COMSOL training courses, for me these have all been very interesting, and have been a good kick-start into each new filed I have approached.

Comsol covers almost all physics, it took me 4 years+ at the university to get really familiar with physics, plus a couple of decades industrial developments, so do not expect to understand everything overnight, but with training one still become a master, even of COMSOL, at least that is my conviction and experience.

Have fun Comsoling
Ivar

It is so difficult to find anything, I haven't dig much so no idea about its solver's ability and other things..but at first glance i didn't like it..may be I will continue with 3.5a or switch to gPROMS..

Hi

there is nothing against continuing in 3.5, I use it still for all my customer work, but I'm preparing myself on V4.0 to switch when I feel its OK enough (and that I control it well enough ;). But it really helps to follow the workshops organised by COMSOL, not all are payed courses, several are free.
And there are a few video demos worth to follow, see the main www site
I'm currently sitting in on here in Neuchatel, we are some 30 persons from all around the region gathering, new and "older" users, to listenand get the "tricks" how to handle the new version.

Have fun Comsoling
Ivar

Hi,
Today I defined one reactor model in comsol 4.0, which involve momentum, mass transfer, heat transfer, and reaction; the same model I defined in the previous version, do you know how much more time I took to define it in 4 compared to 3.5a..!! I took almost 2 hours more compared to 3.5a..this is ridiculous...
Mayur

Hi

Well in my opinion the conclusion is: continue in 3.5, at least until 4 is better documented and up
Ivar

yea your are right..i have to continue with 3.5..I was expecting much better from comsol as I am enjoying working in 3.5.

and also have you observed in 4.0, we can't see model equations in subdomain and boundary settings!!, I really don't like this..any way will wait for proper documentations..thanks for your responses..

Mayur

Hi

Yes I agree it's not there yet and it's frustrating (its' on the planning list for next release I have been promished), but on the other side, how many tousands hours do you expect it to be to rewrite all COMSOL like that ? and they are still less than 190 persons, luckily, because each of them wants to have their salary, and who is paying that ? It's us the users, so less they are the cheaper it is for us; but then it takes them somewhat longer to make such large steps, as between 3.5 and 4, than if they where 500.
On the other side, I'm convinced they had no choice they had to reorganise the modeliing tree approach to keep in line with the general development of more physics, so I'm just glad it's now, and I'll leave them as I do for all my CAD, FEM, and even OS soft to go up to X.1, then only I might complain.
In the mean time we should encourrage them with hints how to improve, beacuse it's now we can influence it without any heavy heritage to take into account

Think of it, if it would be you who had to rewrite everything, how would you go about it ?
I find they are doing great, could be better: yes, but nobody is perfect, or ?

Have fun Comsoling
Ivar

I've been using COMSOL since ver 1 (when it was called FEMLAB) and have been very impressed with each release and especially with COMSOL's support. However, I must admit that the transition from 3.5 to 4.0 to be a major change in COMSOL structure, that has resulted in a very non intuitive interface. While I'm sure the interface needed changing so as to allow greater flexibility, it is now very cumbersome because one has to hunt for options/inputs that you now (being a 3.5 user) must be somewhere, and were quite obviously placed in earlier versions.

Had I initially tried COMSOL with this release, I'm not sure if I would have stuck with it. Of course, I will stick with it because I know it's a great piece of software and COMSOL will continue to work to make it better, but this does not bode well for new users.

I'm glad to see they are directly supporting clusters now! That's huge.

BTW, it appears you can get to the equations (some what) by checking "show equation view" under options, preferences, model builder.

Hi,

I agree with your initial reaction.

My first impression was that the new interface was written by and for financial analysts as a posed to Engineers.

COMSOL 3.X generated a great following and lots of momentum. COMSOL 4.X is a clear fork in the road.
My hope is that COMSOL will come out with a version that has all of the under the hood improvements in 4.X but allows the user the choice on which interface to use, i.e., 3.X or 4.X. It is not clear that this is possible, but if it is, COMSOL would make many many users happy.

It is important to consider that many past text books, teaching materials and COMSOL documentation were written consistent with the intuitive, established approach of the 3.X products. It makes good business sense to ensure a smooth, consistent path forward as new products are released.

Hi

I have full understanding for some of your arguments, when you have spent weeks/months/years to learn a tool, it's not easy to accept such a changes that COMSOL is doing wit 3.5 to 4.
I'm only hooked up from late 3.3, and I was skeptical to these changes last year at the beginning of the conference, but the 2-3 presentations made me understand the idea behind V4, and now I like it much better than 3.5 w.r.t methodology of applying "physics" and not application modes.
On the other side I'm still struggling with "where is now this feature in V4".
Today with 4.0a I feel everything is there and even more than with 3.5. But I still expect to spend a few months before I will fully catch up, as my boss cannot pay me that, it's on my free time (and my bosses at home are not really pleased either, that I understand and have to accept too ;)

Once they have managed to get the code clean I expect that the COMSOL developers will improve the doc, but they are really rewriting ALL physics and then probably too post more presentations showing the methodology that remains very similar to 3.5 I find.

Today I'm only missing ways in V4 to shorten the documentation of my models, I'm still loosing too much time to document the model and dump the requried graphics for my reports (but V3.5 was neither not a reference for speed writing model reports). As well as for scripts for housekeeping and PA/QA check tasks these I must rewrite mostly from scratch

Anyhow, V3.5a is still there and will probably remain for quite some time

I'm still having fun Comsoling
Ivar

also i don't like fluid dynamics in Comsol compared to Fluent, and thus i can't rely on Comsol for my chemical reaction and mass transport part. anyway i am going to switch to fluent, no more comsoling from my side..good luck friends..

Hi

I believe you are too quick there, the CFD code is hardly out, but if you have another code with easy access and it does the job, eveyone is free. One should think global (when necessary): what does your client use, which other physics do you need.
For me my clients use several other tools, that is a pain, for me, as they do not use COMSOL, but I do so many multiphysics, that I have no equivalent to COMSOL, and I believe the others will soon come after us with COMSOL.

Good luck
Ivar