Moment Curvature Analysis - failed

[mastan.kanjula]

Hi,

While doing moment curvature analysis, for some of the axial loads, the analysis is not performed. Can you please check the attched file and advise .

thanks

Mastan

[dave_geeves]

Hi Mastan

As you hadn't attached a zipped up data file I had to build your section from scratch based upon the pdfs you attached, so I had to make a few assumptions (such as the design code you were using etc .....) so I hope they are correct.

I don't think there is a failure here, it is just that in the results printout it appears that some of the cases have just not been analysed.  To resolve this, instead of just clicking on "Analyse", which analyses just the selected axial load tab, click on "Analyse all" before opening the results viewer.  This worked for me and I attach the resultant pdf results file.

Hopefully this should solve your issue and if so please mark this post as a solution so that others can benefit.  Thanks

Kind regards

Dave Geeves

[mastan.kanjula]

Thank you Dave for your explanation. However please note that I have used the option of 'Analyse All', even then the following meassage is displyed "Curvature -Ductility unknown. Analysis failed".

I am not able to attach SAM file with this message, now I have attached a zip file; can you please check and advise.

[dave_geeves]

Hi Mastan,

Thanks for attaching your zipped data file - I downloaded and open it no problems.  I experienced the same problem as you when I tried to analyse the moment curvature so I looked a bit deeper as the analysis certainly fails for some of the load tabs.  It seems rather strange as the section I created from scratch was almost identical to yours ( the reinforcement may have been in slightly different positions but only very slightly.

I believe the analysis is carried out using a strain compatibility approach and the discontinuities in the curve are determined with an iterative approach which in some cases can fail to converge.  I think that this may be what is happening here.  There are a number of parameters that can be changed to assist the convergence, such a convergence tolerances, starting planes etc so I made some investigations into this.

I first of all set the iteration parameters back to the default values - and surprise - all the moment curvature cases were than solved and gave results.  This is simply done by opening up the bending and shear analysis form and from this form opening the "Control Iteration" form.  There is a button at the bottom to reset to default values.  If you "OK" these forms and go back to the moment curvature analysis it then seems to work.

I cant give you a full explanation on why this works but sometimes it has to be done trial and error style with all the parameters.  It seems that the program can occasionally internally adjust some of these parameters for some reason and going back to the defaults sorts the problem out.

I hope this helps and if this has now answered your query please make  note on this post.  Thanks

Kind regards

Dave Geeves

[mastan.kanjula]

Thank you Dave for quick reply. That file worked. But If I change the reinforcement arrangement and loads, then same issue. Can you please check the attached file also.

Thanks

Mastan

[dave_geeves]

Hi Mastan,

I'm afraid you have beaten me with this one.  I have tried all the tricks I know but with little difference. so I don't know what else to suggest.  It is strange that the axial cases that are failing are not close to any limit and that the cases with both larger and smaller values work fine.  It is a shame that the program does not give any further information on the reason for the failure, which may have shed some light on it.  I will continue to think about this but hopefully the developers will be able to respond at some time.

Kind regard

Dave

[mastan.kanjula]

Thank you Dave for your kind help. will wait for the developers advise.

Regards

Mastan

[mastan.kanjula]

Hi Developer,

Can you please advise on the below issue at the earliest.

Thanks

Mastan

[mastan.kanjula]

Hi Developer,

Can you please advise on the below issue at the earliest.

Thanks

Mastan

[karoly_tarcsi]

Hi Mastan,

Thanks for you post.

Section analysis in ASBD may fail to converge in the presence of large axial force in some circumstances. Calculation of first yield moment is very sensitive to the initial iteration parameters.

I examined your model and I introduced some enhancements to to make it converge for all your axial force cases. See Coulmn 5.1x2.0 - Output_enhanced.pdf attached.

Unfortunately usual techniques like relaxing the tolerance to the balancing forces or converting arcs in the cross section to polygons did not offer any solve the issue.

The only way I managed to make all the axial cases converge is by reducing the start yield/stress to 0.00155 and 310MPa for the rebar material property. See Rebar_matprop.jpg and Coulmn 5.1x2.0 - Output_modified_start_yield.pdf.

By doing so, both the ultimate moment and the first yield moment will be slightly underestimated. See Workaround_vs_enhancement.pdf.

Regards,

Karoly Tarcsi

AUTODESK

Software Development

[mastan.kanjula]

Hi Karoly,

Thanks for your response. Can you please send attachments mentioned in the reply.

Also please note that we are analysing different types of cross sections with varying reinforcement arrangement depending on the structural design requirements. Can you please advice any other solution without reducing start yield stresses.

Regards

Mastan

[karoly_tarcsi]

Hi Mastan,

I do apologize, I am not sure what happened with the attachment. Please find the four files zipped and attached.

I have not found an other way to force your section to converge for those axial cases.

Unlike in simple section analysis there is no facility to control the iteration/convergence for the  moment-curvature analysis in the current version.

Are you having difficulties with other sections too? 

Regards,

Karoly Tarcsi

AUTODESK

Software Development