Lateral Buckling on Tapered Section

[Dirgs]

Hi,

I'm getting diferent analysis results when defining a steel frame with haunched beams.

I have for bar 76, half span bar with an end bracket to column. For the other case, bar 82 with constant section and bar 81 with bracket to column. Forces are equal on both cases and member types accordingly for a lenght lz of 1.70m.

For each bar the calculation point is at 1.60m with same internal forces. The diference seems to be on the mechanical properties of the section, as warping constant and depth from load to shear center are the inputs that change, causing diferente Mcr for same beam elements.

In this case, how do you recomend to appropriate modelling for member verification of beam elements as such?

And how RSA is analysing the variable haunched section?

Thanks,

Dirgs

[Artur.Kosakowski]

The difference is caused by the fact that in first case you analyze a super bar which is built of section with the constant inertia at part of its length and the variable on the other and in the second case only the bar with variable inertia. For the first one the stiffness of the whole super bar is averaged (according to the lengths of the segments of different properties).

If you create a new member wilt list of bars 81 and 82 and assign the same label type as for bar 76 the results of the code checking should be the same.

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Dirgs]

Good morning Artur,

I maybe missing something. If the critical zone is the end length of the beam with positive flexural moment, and both member types use same lz length for lateral buckling, the type shouldn't influence the ratio of the bar. 

As you can see in following pictures, with same member type for bar 76 and 81, I get the exact same results as before. So, I still think that the problem is in how the mechanical properties are calculated. Thus, my questions remain. 

Can you please further explore how properties of  bar sections with brackets are defined in order to understand the correctly procedure to the analysis in ultimate limit states. Perhaps a small example to help understand how RSA internally calculates these data.

Regards,

Dirgs

[Artur.Kosakowski · Accepted solution]

Please check http://forums.autodesk.com/t5/Autodesk-Robot-Structural/Critical-moment/m-p/3335273 (message 2)

In the formula F2 for Mcr Robot uses the averaged values. If a member has got different sections along its length the average (weighted) value is assumed (the weight is the length of each of the segments with a different cross section). This is what causes the difference in the verification between the bar with the variable section only (81) and with variable part and the constant one (76).

If you find your post answered press the Accept as Solution button please. This will help other users to find solutions much faster. Thank you.

[Dirgs]

Thanks Artur,

Besides the possibility of enter directly Mcr for member verification, consider also for the wishlist  to output properties used (Iw and It) along with C factors and Zg on detailed results. 

Regards,

Dirgs