RAM ELEMENTS - MODELLING SAME STEEL FRAME (SAME INPUTS EXCEPT CONNECTIVITY AT NODE) WITH DIFFERENT OUTCOMES

[PT99]

I AM MODELLING STEEL FRAME (AISC ASD), RUNNING ONE MODEL WITH SAME SECTION DIVIDED AT NODE, AND ANOTHER MODEL WITH ONE LONG SECTION, ALL OTHER INPUTS ARE THE SAME. STRESS RATIO'S WAY DIFFERENT, BUT RAM ELEMENTS IS SUPPOSED TO CONSIDER 2 CONNECTED BEAMS SAME AS ONE BEAM, RESULTS SHOULD BE INDENTICAL.

 

[driftLimiter]

Did you verify the program understands the unbraced lengths properly? Beam flexure is significantly impacted by the unbraced length and many programs don't get this correct without user override.

 

[Celt83]

Unless you manually assigned the unbraced lengths it is taking them as node to node.

 

[jerseyshore]

From RAM website:

The engineer doesn’t type anything in the spreadsheet cell (either directly or with the tool) and thus the value remains the default “0”: RE uses the physical length of the member as Lb and Cb will be automatically calculated based on the physical length of the member.

 

[PT99]

I believe I did this, with bracing at 6 ft in both models. Can I send you RAM file.

 

[Celt83]

There are multiple locations to set bracing some apply to axial load and others apply to flexure top or bottom. Recommend reading through their manual on how to set unbraced lengths.

 

[PT99]

This is 2 beam with 3 columns. (Middle beam twice span on end beams). Moment and Shear are exactly the same in both models. Connections are shear only, I am modelling with hinge.

For "Assign Flange Bracing Points" (Bending spreadsheet), I have set all Lu Pos as 6 ft and all Lu Neg as 0. Purlins will brace top flange only.

I dont see what's wrong

 

[zurch1818]

I'd recommend looking at the comprehensive report and seeing what is different with the capacities between the two members (since the loads are the same). RAM is nice that it gives you some breakdown on where the capacities are coming from. My hunch says LTB being the problem (like others said).

I've also had torsion issues and also issues with RAM not treating combined bending on pipes correctly. I still want to know where the weak axis is on a pipe.

 

[zurch1818]

RAM will also give you a percentage of where the controlling part is and the resulting internal forces/moments, which can be useful in troubleshooting.

 

[jerseyshore]

If the "analysis" is identical, then it has to be related to something in the "design" inputs. That's sections (and orientation), materials, bracing etc.

I prefer to break members up at nodes as opposed to long members, but bracing is important to input correctly and most likely the cause here.

As zurch said just look at the comprehensive output and compare.