Seismic Design - Penthouses 1

Check if it meets 2-stage analysis requirements?

Assuming the penthouse is not an extension of the building frame and you meet the requirements to use Chapter 13 for the design of the penthouse, I think you design using Fp equations in ASCE 7. I would consider z/h = 1 as it is mounted to the roof of the building structure. Ap and Rp would be per the table in Chapter 13.

Apologies for not being clear, I am concerned about running equivalent lateral force on the building structure as a whole, not for design of the penthouse.

What "h" would you use for the overall structure.

S&T

I think you should use structure height. Part of designing the penthouse per Chapter 13 is the understanding that the element is not large enough to significantly impact the building seismic design. I would just consider the penthouse mass at the roof level. If the weight exceeds 25% of structure weight, I think there are other requirements.

See Section 15.3 of ASCE 7 and commentary for that section.

Table 12.6-1 shows the Permitted Analytical Procedures. In this case, the penthouse implies mass irregularity . Pls look Table 12.3-2 Vertical Structural Irregularities .

2. Weight (Mass) Irregularity: Weight (mass) irregularity is defined to exist where the effective mass of
any story is more than 150% of the effective mass of an adjacent story. A roof that is lighter than the
floor below need not be considered ..

So, if you want to use ELF, height will be taken the height of the roof not the penthouse.


That implies the penthouse wt is too less then the storey wt. If you want to use ELF, you would add the wt of penthouse to the wt of roof level for the global analysis of the structure and use Chapter 13 ( Fp) for the penthouse storey analysis.

Thanks jdengineer and HTURKAK, I ran ELF by combining the weight of the penthouse to the roof with the height equal to the roof height, similar to how you two seem described doing it as well.

Jdengineer said:

I would just consider the penthouse mass at the roof level.

Click to expand...

HTURKAK said:

If you want to use ELF, you would add the wt of penthouse to the wt of roof level for the global analysis of the structure and use Chapter 13 ( Fp) for the penthouse storey analysis.

Click to expand...

I have yet to convince my colleague of my train of thinking; they are stuck on the ASCE definition of Structural Height being:

ASCE said:

Structural Height - The vertical distance from base to the highest level of seismic force resisting system of the structure.

Click to expand...

S&T

The penthouse isn't part of "the structure" if it is not designed as part of "the structure".
Either design it as part of the structure per chapter 12 or design it separately per chapter 13.

To answer your question, "h" would be the height of your main building lateral system. In this case since your penthouse is not framed by an extension of your moment frames, "h" is the height to the main roof. This is assuming that the penthouse qualifies as an appendage and can be treated as a nonstructural component. Side bar: I personally feel that there is somewhat of a disconnect between the logic of using a 25% weight cutoff for ASCE 7 Section 13.1 but a 10% weight cutoff for ASCE 7 Section 12.2.3.1, so I tend to tread carefully once I'm beyond 10%.

I don't believe you'll find anything more definitive in the code, but pictured below is an excerpt from the steel example in the FEMA P-1052 training and instructional materials. You may also have luck digging through some of the references for the approximate period equations used in ASCE 7.

Out of curiosity, have you modeled the building with the penthouse and compared the actual period to the approximate values?

Deker said:

Out of curiosity, have you modeled the building with the penthouse and compared the actual period to the approximate values?

Click to expand...

Capped period is 0.868 seconds.

Actual period (without penthouse):


With the penthouse:


S&T

Makes sense. My understanding is that the code equations are empirical and capture real-world stiffness effects from cladding, partitions, gravity columns, etc. that are not typically included in a lateral system frame model. So you would expect your bare-frame model to have a longer period.