Need a book? Engineering books recommendations...

Return to index: [Subject] [Thread] [Date] [Author]

Re: Structural Analysis Software for New One-Man Firm

[Subject Prev][Subject Next][Thread Prev][Thread Next]
Bill:

You need to get a bigger and faster computer.   I work with RISA quite a bit and I've been to their three day training.  On my machine, my models run almost instantaneously compared to when I was using my IBM AT.    I get pretty much instant response from their on-line consultants and have had great assistance from Josh Plum.   If you don't want to do multiple cases, then plan your output requests more carefully.   Sometimes you get to much.   Yes, in the beginning the graphic interface to the old DOS system was the pits - but I was using SAP2000 then.   What I'm into now is some stuff that has to be run in NASTRAN and I just sub that out.  

None of my runs on the big models (which I really don't do real often)  take over 15 to 20 seconds.   I would say that is a big improvement when I would leave the IBM on all night and come back the next morning with the results.   But all in all it's still better than previously  having to go down to CDC in San Jose with my deck of punched cards and play with multiple runs which took about 20 minutes per cycle.

Obviously that's some history that I'm talking about and was just hoping that you would stick to history.



Neil Moore, PE, SE, SECB
neil moore and associates
consulting structural engineers


established 1972





On 1/19/2011 10:00 AM, Bill Polhemus wrote:
On 1/19/2011 10:34 AM, Keith Erick Fix wrote:

One more thing: I really did hear those of ya’ll voting for Visual Analysis, and I thought about that REALLY hard, but the “sales” guy pretty much ruined that alternative for me. There’s a certain back-woods feel to that bunch that makes me think of the kind of guys I meet in Arkansas that practice engineering from a home office. I’m sure the IES is really professional, well-staffed, and technically competent, but I just can’t shake the feeling that I couldn’t depend on them if I got into trouble with a model. Obviously, if you recommended the software, then your mileage varies from mine.

 
I can see where you'd get that impression, because they ARE a small firm and their self-marketing (i.e. their public relations) leaves much to be desired. I've chided them about this several times, but they seem to take the approach that they don't care if they "stay small," because significant growth just isn't in their business plan. That's why you don't see their advertising as much, and why they don't "partner" with other products (such as Autodesk's Revit - although they do have a nice Revit interface that they wrote in-house).

Not to change your mind, but just for the record...

I have never dealt with software vendors/publishers who were more responsive than the folks at IES. There were two instances that come to mind in my experience with VA, the first a bug (caused the program to crash) and the second a feature I thought should be there that was missing, and in each case they responded to me within a short period of time (about six hours in the first case, and about three days in the second) with a patch.

I have no affiliation with IES whatsoever, aside from giving them money in the past to buy their products, and being a beta-tester from time to time. I love 'em, love their products, love their approach to dealing with customers' needs.

And insofar as "all software having flaws," I simply found the particular one I pointed out with Risa to be intolerable. The day-to-day modeling that most of us do (particularly the small-fry) may well never come up against the limitation I described enough to matter, but I had a very large model, about 2,000 members with several loads and load combinations including seismic. I was performing an earthquake design as well (and I would put in here, that I found Risa 3D's approach to seismic analysis input VERY elegant, so much so that I described in detail to the folks at IES, advising them they should try a similar approach).

But in the end, it was taking HOURS per run, after which I would focus in on the members I wanted detailed info from, then I'd have to run it again to get that info. It was ludicrous, in my humble opinion, indicative of Risa's roots in the memory-limited world of DOS where the program started.

IES began to develop their software in the early days of Windows NT, and focused on the Windows environment and took advantage of the memory management of Windows. Most of the other, older software (including Risa and STAAD), early on had a "bolted-on" interface with the same old DOS-era algorithms running behind it, and it took some time before they were able to correct it. STAAD took a VERY long time (it got better after Bentley purchased them), and in the case of Risa, some limitations appear to remain.