*I am running P-Delta Analysis on a reinforced concrete tank
PDELTA ANALYSIS CONVERGE 2 SMALLDELTA
*The bottom of the tank is supported on soil and I am using foundation support as follows:
SUPPORTS2578 TO 3393 PLATE MAT DIRECT Y SUBGRADE 31 PRINT COMPRESSION
*The analysis gives my many warnings about instability which I am assuming though to Compression only support definition (see above line).
*so I went ahead and changed the support definition to get rid of compression only as follows:
SUPPORTS2578 TO 3393 PLATE MAT DIRECT Y SUBGRADE 31 PRINT
*I then ran the P-Delta analysis and all the warnings went away.
*Question is: Compression only supports do not work with P-Delta analysis?
*Using the normal "PERFORM ANALYSIS" does not give me any warning with the following:
Thank You in advance.
Pdelta analysis should work fine with compression only but just ensure that you are not analyzing component cases ( like LL alone, wind alone or seismic alone ) with compression only springs as supports as that can cause the solution to diverge. You should first have the supports without compression attribute, specify the component primary cases, define PERFORM ANALYSIS and CHANGE, redefine the supports with compression attribute, specify the REPEAT LOADs to combine the primary loads, add the PDELTA ANALYSIS command at the end
Instead of using the CONVERGE option, use something like PDELTA 15 ANALYSIS SMALLDELTA
I did exactly that in the following order
Primary load cases are defined next followed by:
PERFORM ANALYSISCHANGESUPPORTS2578 TO 3393 PLATE MAT DIRECT Y SUBGRADE 31 PRINT COMPRESSION
Repeat loads were defined next followed by
*I checked the reactions and finding lateral reactions and moments which should not be.
*I use CONVERGE to make sure and verify that load cases are converging. Is that problematic?
Thanks
Then check for the instability warnings and check for the displacements. You may be having abnormally high displacements due to instabilities leading to high reactions/moments.
Converge option is not related to instabilities. We do however recommend that you use more iterations, which would ensure convergence anyway, rather than forcing the software to converge within a certain number of iterations as it may not be enough.
1) Regarding instabilities and lateral reactions:
I think in my case having lateral reactions & some moment reactions is normal as the R.C. tank will experience skin friction and the surrounding soil will contribute to its lateral stability as well.
2) As far as convergence, If I did not specify the "converge" command, will STAAD tell in the output notes if any load cases failed to converge after performing the specified iterations? My thought that STAAD will only tell this info if the converge command is used. Am I misunderstanding something?
STAAD.Pro will tell you whether the solution converged or not and in how many iterations. The information is printed in the analysis output file. You do not need to include the converge option for that.