Plate Stress contour plot

[dozer]

I've just started using STAAD.Pro 2004 and I was surpised to find that you can't do a stress contour plot using the nodal stresses. Am I missing something?

In case you're wondering why I care. The guys that built the model I'm looking at used 12" square plate elements for a wall with external vertical stiffeners. If you look at the color of an element you think that the stress in a wall element adjacent to a stiffener is around 12 ksi, but when you look at the nodal stresses you see that it is about 18 ksi at the edge of the element. BTW, I calculated what I thought it would be before I looked at the model and got . . . drum roll . . . 18 ksi.

The range that STAAD shows in the legend uses the element centroid stress. Yes there's a toggle that says "Index based on center stress" but it doesn't seem to do jack. Anyway, so the maximum value in the range (if it wasn't for some other areas of the model) would say 12 ksi.

Other than the fact that smaller elements should be used in this model (IMHO) is there really no way in STAAD to view nodal stresses graphically? BTW, what's the deal with the solver? This model has 13381 nodes and it took forever to run. Forever is 40 minutes, btw. There was a lot of output, but still, that's ridiculous.

And since I'm complaining (I really don't like STAAD), there is a huge output file. I viewed it inside of STAAD and it was taking forever to load a 31 meg file. I just hit close because I got tired of waiting and I opened it with Textpad. It opened instantly. What a piece of crap STAAD is.

Well, now that I'm way off the subject, don't forget my original question about stress plots.

Thanks

 

[bylar]

As you may figure out from the anser to other post, you may not get an answer here. If you have a legal copy, try support from STAAD

 

[dozer]

Yes, we do have a legal copy but in a classic case of holding a grudge, many years ago I had what was clearly a bug but they maintained it wasn't the program's problem, it was a modeling problem. I've been P.O.'d ever since, so I thought I'd give the engineering community a shot at this.

Also, I want to make people aware of the pitfalls of FEA. I think the way STAAD does the range of values for stress is very dangerous and misleading. For instance, I was looking at a gusset that was part of an overall structure. According to the legend, the maximum stress was 10.1 ksi, but when you examine the nodal stresses the max was 23 ksi!!

Yes, this was at a boundary so there are some accuracy issues and yes, as I stated before I think the elements are too big, but come on, tell me graphically what the highest stresses are and let me decide if they are real are not.

Oh, by the way, when I said if you turn off "Index based on center stress" it doesn't do jack, that's not right. It does worse than jack. It shows graphically stresses adjacent to the 23 ksi node that I mentioned that you would read from the color contour to be around 10 ksi.

Bylar, I understand your comment about not getting an answer. At the risk of incurring the wrath of STAAD users, I don't think the average STAAD user has much experience with FEA other than frames. My fear is most people just accept what the program is spitting out without really questioning it. Hopefully, people will start casting a more critical eye at the results.

 

[zer]

dozer let me agree with you. I got my hands dirty with STAAD.Pro 2004 numerous times and the interface is just trash (ever tried printing a figure from the screen?) and inconsistent. This is reflected in the design code parameters as well. Let's not even get into the connection design and the STAAD.etc module.
The only attractive feature about it is that due to the input/output file the whole process of checking/reviewing is more transparent. I mean I find it okay to carry out a design of a not-so-important structure and wrap the input/output file nicely in a report, but doing anything else with it is suicide. And I mean not just being doubtful about the results, but the process as well.

Any other opinions on the matter would be appreciated.

 

[bylar]

zer
What is a chemical eng doing with STAAD

 

[zer]

I misclicked the field when In the sign up process

 

[dozer]

Zer,

To print the screen, click the printer icon with a Q on it. No idea what the Q means. I never have found this on the pull-downs. Common sense would dictate it would be File > Print, but nooo, that doesn't work.

Until I figured this out, I was taking a picture and then putting it in a report and then printing a report.

 

[zer]

Thanks for the tip dozer, even though that was just an example to give an idea on how crap the interface is.

 

[DTJ]

Under Plate Results click on the "Results along a line" tab. This will allow you to cut a plane anywhere along the structure and get the results at that specific point rather than the average across the plate. You can also get global moment results under the plate contours tab.