PCA Column: what to do with nonorthogonal beams?

[hippo11]

Please help...Thanks...

PCA Column uses the moment magnification technique from the ACI code.

If I have a beam framing in at a skewed angle, how would I account for that non-orthogonal angle in my slenderness calcs? Would it be appropriate to account for the slenderness contribution of this skewed beam by assuming the orthogonal component of the length of that beam as the “span”, i.e., span= length*cos(angle) or length *sin(angle)?

 

[JAE]

I think its a bit more complicated than using an orthogonal distance replacement.

The formulation in the ACI code deals with the effect of the stiffness that the beam provides at the column joint to resist orthogonal bending in the column. So when you put a unit deflection on the column (in an orthogonal direction) this displacement is a pure rotation, ...say about the Y axis. If the beam is at an angle to the column, some degrees off the Y axis, then the resulting rotational displacement puts a combination of bending AND torsion into the column. So the beam resists the column bending with flexural and torsional stiffnesses....some combination of the two.

At 0 degrees offset, the beam uses 100% flexural and 0% torsional. At 90% offset, the beam uses 0% flexural and 100% torsional. In between - well - I don't know how it transitions between the two - not sure if it is linear or not.

One idea would be to always underestimate the resistance that the beam provides...this could be done by using a weak beam in flexure only and then ignoring the torsional effect totally. The orthogonal component of the span would be shorter, resulting in a stiffer beam so I wouldn't do that...use the same span, just reduce the stiffness by some amount - again, I'm not sure how to determine this reduction - possibly a linear reduction with an added safety factor to further reduce the beam stiffness.

Just some off the cuff thoughts. Anyone else have any ideas or knowledge about this?

 

[JAE]

...also - be careful as PCACOL is fairly dated and the moment magnification procedure may not properly correspond with the applicable building code and version of ACI 318.

 

[hippo11]

Maybe I could multiply the EI, not the L, by cos(angle).

 

[JAE]

....oops - sorry - the second to last sentence in the second paragraph above should end with "...resulting rotational displacement puts a combination of bending AND torsion into the BEAM."