Different Response Modification Factors for Orthogonal Lateral Resisting Systems

[EngineerRam]

Hello,

When calculating the seismic demands on structures as per ASCE 7 for a regularly shaped building with different lateral systems in the orthogonal directions, the computation is done based on the primary lateral system in each of the directions. My question pertains to the shear going to the system perpendicular to the direction of the applied seismic load. For instance, if I have a SMF system and a SSW system, the shear that goes to the SSW system due to accidental and inherent torsion would be based off of the SMF system, with a higher response modification factor. Reviewing ASCE 7, I have not seen anything that requires the shear going to the system perpendicular to the direction of the applied load to be scaled up in accordance with its lower response modification factor, but to me it makes sense to do this. Has anyone ever had this question and gotten a decent answer?

Thanks

 

[driftLimiter]

Interesting question! I can't find a direct answer in ASCE either. But if we think about the R factor conceptually I think we can safely answer the question.

The R factor is meant to account for the ductility of the given system. Unless the building is extremely (and I mean EXTREME) prone to torsion, One would expect that the predominant load for any direction under consideration is the DIRECT shear load. This would be the load that generates the ductile displacements.

The amount of load due to torsion on the perpendicular system should be significantly less than the Direct load of that system for it's own direction.

This to me implies that the direct shear load would cause yielding in the system of that direction, and no yielding in the orthogonal system. Thus the eccentric moment and resulting perpendicular shears would be based on the R factor of the system in the direction of the direct shear.

If for some reason the loads to due eccentricity from one direction are greater than the direct shear on that same line are greater, then you might consider scaling the forces based on the R factor?

 

[HTURKAK]

I think the code is clear for this case .
Copy and paste of Section 12.2.2 ;

Combinations of Framing Systems in Different Directions. Different seismic force-resisting systems are permitted to be used to resist seismic forces along each of the two orthogonal axes of the structure. Where different systems are used, the respective R, Cd, and Ω0 coefficients shall apply to each system, including the structural system limitations contained in Table 12.2-1.


That is, different seismic force resisting systems may be used in the two orthogonal directions of a building. If there is no interaction with shared structural elements , the R, Cd , and Ω0 in Table 12.2-1 will be applied independently to the two orthogonal directions..


You are expected to perform the analysis with R= 1.0 then the design effects shall be calculated using relevant R values in Table 12.2-1 independently to the two orthogonal directions..

My opinion...










Don't underestimate a nail. A nail saves a horseshoe, a horseshoe saves a horse, a horse saves a commander, a commander saves an army, an army saves a whole country.. GENGHIS KHAN

 

[driftLimiter]

You are expected to perform the analysis with R= 1.0 then the design effects shall be calculated using relevant R values in Table 12.2-1 independently to the two orthogonal directions..

The base shear is computed in each direction then distributed to each level the resulting inherent and accidental torsion for each level is based on the seismic design story shear which is based on the base shear in that direction.

My opinion is that if one scaled the shears resulting from torsion it could lead to a non-conservative result. The R factor in the transverse direction could be higher than the direction being considered. This would essentially imply that the designer is expecting yielding of the transverse SFRS due to torsional loading which as I stated above is not a typical situation.

 

[milkshakelake]

I think it depends on how much of an effect the torsion has. If it's very significant, I wouldn't scale up the R values. I'd just use the most conservative R value, similar to a mixed lateral system in one direction, or a building with curved/diagonal lateral system.

A lot of this comes from practicality for me. How would you even scale up the shear in one direction? Use a load combination with a higher factor, calculated with something like R2/R1 with an adjusted Tcal? You'd have a hundred load combinations in different directions, assuming you're doing the 45 degrees and eccentricity thing.

 

[driftLimiter]

Sorry I can't let this one go....

No matter how many different ways I think it about, my conclusion is the same.

The torsion due to eccentricity is generated as a function of the Story Force.
The story force is based on Cs which is based on the R factor for that direction.

There is no need to scale the perpendicular forces resisting the torsion.
It is not reflective of the assumed seismic behavior, and it can be unconservative.

At the end of the day some frame or shear wall line will have earthquake forces that are classified under the seismic load combination of each Direction. ELX and ELY for example.
The R factor for each of those directions is contained within those load combinations.