RISA PDelta and Buckling

[RonRoberts]

RISA's Pdelta analyis kicks out after the deflection reaches 1000 times the initial deflection (RISA Error 2011).

If this trigger (Error 2011) was eliminated in RISA, would the Pdelta analysis of a column converage on the Euler Load (at least roughly converge)?

The attached model should buckle at about 100 kips, but kicks out at a little over 22 kips.

 

[JoshPlumSE]

Ron -

Euler Buckling Load (Pe) = pi^2 *EI / (KL)^2

In your case EI = 0.35 * Igross
This is because you have not un-checked the box that says "consider cracked sections" on the concrete tab of the global parameters.

Pe = pi^2 * 1,814,400 / [(59*12)^2] = 35.7 kips

**Note: ACI doesn't usually let you get above 75% of the theoretical value = 26.7 kips before your amplification becomes infinite.**

Therefore, the point at which the P-Delta method starts to diverge is at about 65% (i.e. 23/35 = 0.65) of the theoretical buckling value.

That's pretty typical. I don't usually see this method as being capable of predict P-Delta amplification above about 70% of the theoretical elastic buckling value.

 

[RonRoberts]

Thanks Josh.

If you change the "material" to General, but use the properties from concrete will that let you look at a purely elastic buckling case?

 

[JoshPlumSE]

Using a general material is a way to avoid the cracking that is normally assumed with any beam defined as concrete. However, you will probably run into the same sort of convergence issue as you approach 75% of the theortical elastic value.

But, personally, I think it is easier to do by un-checking the "used cracked sections" box on the global parameters. '

Alternatively, if I want to avoid cracking on only one or two members, then I can modify the Icracked value which is part of each members concrete design parameters. If I set this value to 1.0, then no cracking will be assumed for that member.

I hope that helps!

Josh

 

[RonRoberts]

Thanks!

Found a similiar thread in "Structural Engineering Other Topics" addressing similar questions with a rather active discussion going on. You might find it interesting.