Structural Analyis Software Recommendation 2

[VAStrEngr]

I would appreciate if anyone who currently uses any type of structural analysis software could provide some reviews and/or suggestions for a reliable software package that I can pitch to my company.

We are a multi-disciplinary engineering firm (50 employees) and our structural needs range from the design of small buildings and frames to large storage tanks and pipelines. Any software package has to be versatile enough to handle a large variety of designs and include all the major building materials (wood, concrete, and steel).

We currently use a very dated version of RAM Advanse but given the pace that technology has progressed, I would like to explore options for whatelse might be out there.

I have heard of the Risa software and some of the other big names but would like to hear from people who use the software before looking into it too far.

Thank you.

 

[jhardy1]

OK, at the risk of high-jacking the OP's thread, can I put a different slant on the original question:

Of the various structural analysis / design programs that people have advocated, is anyone using them in conjunction with a high-end 3D plant design system (such as Intergraph SmartPlant 3D) for mechanical / process driven design, such as oil & gas plants, mining projects, etc?

How do you go about keeping the engineering analysis / design model "synchronised" with the 3D plant model, as the plant layout "evolves"? Do you import / export via a neutral file format such as DXF (shudder!), CIM-Steel, SDNF, etc, or do any of them support "live" interaction with the multi-discipline plant-wide 3D model and database?

How does the process support multiple iterations of "round trips" for design iteration. (Most of the packages I have seen are sort of OK for member size changes as long as the frame "topology" doesn't change, but how does it cope when the mechanical engineers make significant geometry changes to the structure to accommodate design changes for their plant and equipment?)

Any comments welcome - I fear I may have stirred up a hornet's nest here!

 

[271828]

Honestly, I am envy at our colleagues in auto or aerospace industry who can utilize general purpose FE like Abaqus or Nastran.

LOL, it's quite silly to compare Abaqus and other high end FEA programs to programs that sell for only a few thousand dollars!

Besides, when would most structural engineers use 99% of the features that Abaqus has, but RISA doesn't? Almost never, and the higher end programs require much, much longer learning curves in my experience. A halfway experience guy can be up and running with RISA or SAP2000 in just a few hours.

To the OP: Probably RISA-3D. SAP2000 if you think you'll want to do much beyond simple small displacement second-order analyses.

 

[271828]

Also it has a very robust manual to help learn the software.

Do I sense sarcasm there? Last time I used Abaqus, the manual was badly needed!

On a somewhat related note, I tried to learn Ansys a few months ago and started going through the tutorials. After 3 hours, I still couldn't get a basic beam in there with a point load on it. Then I got distracted and did something else, LOL.

 

[dik]

I like SAP and Risa... more frequently using the latter for run of the mill designs. The high end FEM programs are OK for automotive and aerospace stuff, I suspect, because with automotive, if you save a dollar in the production and you make a million of them, then you have directly added $1M to profit. In aerospace stuff, you are looking to minimise weight.

Most of the stuff I do is 'meatball'... the cost savings is in the ease of connection or fabrication. If I spend an additional half-hour to analyse a connection to save a pound... unless I have a 'kazillion' of them, there is no advantage.

Dik

 

[JAE]

RISA 3D.

Agree with Ron on the older STAAD versions - very buggy. Haven't used it recently so it may be much better today.

 

[OCI]

I would reccomend Scia Engineer. It's very popular in europe and is getting more popular here in the US. It offers most of the higher end features of SAP and Robot with a much better interface and report generation capabilites.

 

[irizarry]

I would say, if it's multi-disciplinary and versatile what you need, SAP2000 is the answer. You have almost the full versatility of highend FEA programs, but can use it for pretty much everything without it taking three hours to analyze a simple model.

 

[BAretired]

I don't know anything about structural software, so it is not likely that I am going to provide you the answers you seek.

Is there a software package which can accept 3D co-ordinates, 3D loads and come up with a complete and accurate analysis of the structure, including unsupported length (buckling length) of individual members? I do not believe that there is. If I am wrong, please let me know.

BA

 

[IDS]

Is there a software package which can accept 3D co-ordinates, 3D loads and come up with a complete and accurate analysis of the structure, including unsupported length (buckling length) of individual members?

If you do a linear analysis you need to do your own assessment of buckling loads, which you can either do with a separate buckling mode analysis, or include geometric non-linearity effects in the analysis.

If you do a complete non-linear analysis, including allowance for material and geometric non-linearity, and with proper conservative material stiffness values, there is no reason why any 3D analysis package should not come up with the right answers, providing the package allows sufficient control so you know what it is doing. Whether that is the best way to do it is another matter.



Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[BAretired]

Thanks, Doug. I think I understand most of what you said. Why would it not be the best way to do it? Does it still require engineering judgment, no matter what the software spits out?

BA

 

[IDS]

Yes, you still need engineering judgement in setting the material properties, and assessing the output. Probably more judgement needed than when using tried and true simplified methods.

Also you would need to do separate analyses for serviceability and ultimate checks, and maybe another one for collapse as well, so the question is, is it worth the effort?

Re-reading the question, if you were asking if there is a package where you can just plug in the geometry and the loads, and let the software set materials properties etc, and rely on the design that is spat out at the end, then I'd agree the answer is no (as far as I know).

Doug Jenkins
Interactive Design Services

http://newtonexcelbach.wordpress.com/

 

[VAStrEngr]

Any specific feedback on Robot? The partner-in-charge (not a structural engineer) is interested in that because of the tie in with Revit. I checked the website but first-hand reviews are much more helpful.

 

[DWHA]

I have only ever used RISA, but I will say that it is fairly easy to pick up. And, as stated before, the support is great. If you have a question about you model, they have an answer.

 

[OCI]

I have used Robot. It has some strong features but the interface is very cumbersome. Autodesk bought it a few years ago and so far has done almost nothing to the software as far as improvements. Like I said above, if you like robot I would take a look at Scia Engineer.

 

[JoshPlumSE]

It's funny, I've heard a lot of speculation about how well Robot and Revit will link together now that AutoDesk has aquired Robot. But, I haven't really heard any buzz from engineers on the topic.

If you're interested in the link between a structural analysis package and Revit, then I would suggest starting another thread about that specific topic.... or, looking for a forum dedicated to Revit or BIM.

 

[DHKpeWI]

For those of you who like RISA and use Direct Analysis:

Have you worked through the "benchmark" problems in the commentary of Appendix 7 using RISA? Please see page 16.1-435. According to the information on RISA's web site, many additional joints must be added between the member ends to adequately model the P-Delta (big and little) effects. However, when the compressive load is small compared to the Euler buckling load, the number of additional loads is required is minimal.

How have you handled this in your designs?

 

[bbookz]

Risa Floor/3d is user friendly, but it is not without flaws. Risa 3d will not compute live load reduction for column loads even though this is done in Risa Floor. You'll have to trick the program to calculate unbraced lengths for beams supporting steel joists. Also, you'll have to trick the program to size composite beams correctly for deflection when they support post composite applied dead loads. Concrete design is limited in its usefulness. When I used Ram prior to our purchasing Risa I did not notice as many design bugs, but the Ram interface was archaic and visually checking a model for errors was more difficult. That being said Ram did have slightly better output of design information for Reports.

 

[JoshPlumSE]

DHKpeWI -

Have you tried to work through those Benchmark problems in any other program (STAAD, RAM, SAP). Last time I checked, they ALL require that you add intermediate joints to the member in order to fully match the "theoretically correct" equations.

Now, the RISA documentation makes it pretty clear exactly what you have to do in order to capture this P-little delta effect. Not sure if the other guys address the issue as thoroughly in their documentation. Maybe they do, maybe they don't.

Disclaimer: I am actually a RISA employee. I also sit in on most of the AISC task committee meetings for Chapter C and Appendix 7. So,

 

[Jhonel]

Whaaaat?...No GTSTRUDL??... ;-)

 

[SpecialEddie]

RISA 3-D is hands down the best. If RISA were to disappear off the face of the earth, I would quit my job and become an architect.