When to Consider Seismic Loads on Columns in Shear Wall-Dominant Designs?

[MAB70]

Hello everyone,

I’m working on a building where shear walls are the primary lateral force-resisting system for seismic loads. My goal is to let the shear walls handle most, if not all, of the lateral load, potentially reducing seismic design requirements on the columns. I’m aware that Eurocode 8 provides a specific 15% threshold for secondary seismic elements, where columns might be considered non-seismic if they carry less than 15% of the lateral force.

However, in American codes (ACI 318, ASCE 7, and IBC), I haven’t found a similar explicit rule. ASCE 7 requires load distribution based on stiffness, which implies columns should be designed for the portion of load they attract, but I’m curious how practitioners approach this. Here are a few questions:

1. Have you used a stiffness-based analysis to demonstrate that columns attract minimal seismic load and could be excluded from seismic detailing? If so, how did you document this approach to align with ACI or ASCE requirements?
2. Is pinning column ends T&B a viable strategy to limit their participation in lateral resistance and channel more load into the shear walls? Are there any practical challenges or code implications with this approach?
3. What’s your interpretation of treating columns as “secondary seismic elements” in American practice? Do you apply similar thresholds (e.g., Eurocode’s 15%) to justify this, or do you have another method of ensuring code compliance?

I’d love to hear your perspectives on handling this design issue, particularly any code references or practical advice you can share.

Thanks!

 

[KootK]

Looong ago I was taught to just assume that everything goes to the shear walls so long as they are six times as stiff as the sum of all of the columns. I don't do this any longer.

In North America, I believe that the dominant practice is to:

- Design the designated lateral system for all of the lateral loads.

- Check the "non-structural" things for compatibility with the estimated seismic drift. This is particularly important for punching shear at slab to column connection.

As things shift towards performance based design in some markets, we start returning to analyses that somewhat consider all of the structure as participating laterally to some degree.

 

[MAB70]

@Kootk, thanks as always for your insights.

On the topic of shear walls as the main lateral system—do you typically run two model configurations? Specifically, one with the columns engaged to capture the loads for their design, and another setup where the columns are pinned to direct 100% of the lateral forces to the shear walls for their design?

Also, you raised a good point regarding punching shear. Assuming we need the slab to accommodate the ultimate drift (Elastic Drift x Cd), do you typically factor in unbalanced moments derived from Seismic Force x Cd? Or do you account for overstrength to ensure the slab remains elastic for punching shear considerations?

 

[meamin]

if it's concrete, aci chapter 18 has requirements for non-lateral components. Applies for sure to seismic design category D and above.

I believe the aisc seismic manual has some requirements for non SFRS system component as well.

 

[MAB70]

Thanks Meamin, Its SDC B .. the code seems to be silent on them and doesnt go into much detail unlike SDC D and above where a whole section and requirements were given.

 

[KootK]

On the topic of shear walls as the main lateral system—do you typically run two model configurations?

Yessir.

Assuming we need the slab to accommodate the ultimate drift (Elastic Drift x Cd), do you typically factor in unbalanced moments derived from Seismic Force x Cd? Or do you account for overstrength to ensure the slab remains elastic for punching shear considerations?

Neither. I'll actually treat it as an imposed deformation problem using Cd rather than as an applied force problem. Firms often have column design spreadsheets that make enough simplifying assumptions that it becomes possible to tackle the columns independently from the whole building model.

 

[MAB70]

Thanks Kootk, Just on these responses:

1) Would you check your drifts based on the one that has both walls and columns participating in resisting lateral loads or you take it from the model where you pinned the columns and having the walls get the total force?

2) On Punching shear, i know ACI 421.2 has a simplified method close to what you have mentioned using the elastic drift and utilizing the slab as being simply supported at mid spans and using 50% effective inertia but it seemed for me too empirical with no reasoning behind the method.