Seismic Design of Small Mezzanine 1

[AThor]

I just wanted to get some feedback on several aspects of a project I have been working on.

I am designing a small (40'x25', 4' tall) mezzanine. It will be used for storage, so design LL is 250 psf, for heavy warehouse storage. The design is a composite deck, W beams, HSS columns, 12 bays. The spans are fairly small, 7-9 feet, so that the columns can be attached to the slab on grade below, without cutting holes for footings.

Most of my questions revolve around seismic design. I am in seismic design category D. I will just go ahead and list some of the assumptions I am making, and am curious to get thoughts on the accuracy and/or validity of the assumptions:

- I am looking at ASCE 15.5, non-building structures similar to buildings. It looks like I can select a seismic force resisting system from Table 15.4-1, but I get directed back to chapter 12 for design procedure.
- The concept we came up with was: tie the mezzanine into an existing, stable wall for lateral resistance in one direction. For the other direction, have the beams span between columns, with moment connections at the column base and beam connections.
- I see this system as an ordinary moment frame. Being in design category D, referencing Table 15.4-1, I have to use R and omega0 both =1 to avoid AISC 341, which gives me a base shear of 1.5*seismic weight. This seems high.
- In general, what is the difference between selecting a system with R and omega both =1, or both =3? Wouldn't both yield the same design forces?
- I'm thinking whatever I select, the high live load might govern member selection anyways.

I'm just looking for thoughts on if I am interpreting ASCE seismic design correctly.

 

[JoshPlumSE]

Why not look at AISC 341 and use R = 3.5? The rules for OMF shouldn't be too terrible. Worst case is using the R=1 forces for the moment connection design. But, you will save on base plate reactions and such.

 

[AThor]

Thank you for the reply. My initial instinct was to go for the most conservative choice, since I am fairly inexperienced with seismic design, AISC 341, connections, etc. I'm still trying to understand all the nomenclature and how it all fits together.

It seems like I can use a steel OMF with R=2.5 and omega = 2, from Table 15.4-1, if I follow AISC 341, which it seems like just boils down to using FR or PR connections for beam to column. My adviser mentioned to use an all around fillet weld for this connection. Would this type of weld satisfy the 341 OMF requirements?

I'm also a bit unsure how to use the overstrength factor. Do I apply overstrength load combinations only when looking at force in the connections?

 

[msquared48]

Better be sure you can use an OMF in your seismic zone.

Mike McCann, PE, SE (WA)

 

[AThor]

If I'm understanding correctly, an OMF is permitted in design category D for a non-building structure similar to buildings, if I use R = 2.5, and follow AISC 341 requirements, and have a maximum height of 100 ft (this mezzanine is only 5 feet tall.)

 

[msquared48]

OK.

Mike McCann, PE, SE (WA)

 

[KootK]

If you're attaching to a solid wall (block/concrete) on one side, I'd be very tempted to call that your lateral stabilizing element in both directions (in and out of the plane of the wall). I suspect that will be much cheaper than going with moment frames. It may also be a better reflection of stiffness in the system compared to moment frames.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[JGard1985]

Agree with reducing cost per KootK. Depending on headroom/storage requirements under mezzanine, have you considered braced bays in lieu of moment frames. You have a good application with squat bays and good bracing angles

Jeff
Pipe Stress Analysis Engineer

http://www.xceed-eng.com

 

[AThor]

They want to keep the underside open for storage which is why I'm avoiding bracing options. That is true that the wall connection will probably end up taking a lot of the lateral force in both directions. The idea was to fabricate each of 4 frames (4 columns, 3 beams, baseplates each) in the shop and ship these frames pre-made. Then in the field, they will just need to be bolted to the floor, and have the deck attached and poured. This would save on field welding costs.

I'm hesitant to use bolted connections for the beam-column joints. At the very least, I could determine seismic forces as if it was an OMF, then the wall connection would take a portion of that load, reducing the demand on the baseplate connections, and adding some conservatism to the weld connections.

Thanks for the replies.

 

[Shu Jiang PEng SE]

The seismic load calculated per “steel ordinary moment frames with unlimited height” is larger than that per “steel ordinary moment frames with permitted height increase”. The reason for the specification is that: the extra requirements of configuration and connections (AISC 341) for “steel ordinary moment frames with permitted height increase” will guarantee the structure having certain level ductility, which reduces the seismic load the structure experiences. No any extra requirements are imposed (AISC360) for “steel ordinary moment frames with unlimited height” to guarantee the structure ductility.

Use full moment connection between beams and columns for “steel ordinary moment frames with unlimited height” is not enough.

 

[AThor]

I think that is narrowing in on my confusion. What exactly defines connections that satisfy AISC 341 requirements, versus connections that meet AISC 360? For example, AISC 341 says beam column connections are permitted to be fully restrained or partially restrained moment connections in ordinary moment frames. It then goes on to list several conditions that qualify a connection as FR or PR. The steel construction manual also has discussions on FR and PR connections in chapters 11 and 12. Are the definitions of FR and PR connections different in AISC 341 compared to the SCM? Are certain types of welds prohibited? Or is it all about the design method, and any weld type can be used if designed properly? i.e., can you use a fillet weld but not a groove weld, or visa versa?

 

[DETstru]

Since your column spacing is so small and the mezzanine is so short, why not use special cantilevered columns?
No special connections other than making a fixed base at the column.

 

[Shu Jiang PEng SE]

Take FR moment connection as example: some requirements per AISC 341 are: "The required shear strength, Vu or Va, as appropriate, of the connection shall be based on the load combinations in the applicable building code that include the amplified seismic load. In determining the amplified seismic load the effect of horizontal forces including overstrength, Emh, shall be taken as:
Emh = 2[1.1RyMp]/Lcf (E1-1)" "FR moment connections shall be designed for a required flexural strength and a required shear strength equal to the maximum moment and corresponding shear that can be transferred to the connection by the system, including the effects of material overstrength and strain hardening." --Section 9.1-29 AISC 341 Ordinary moment frame (OMF).


FR moment connection shall be designed per the required beam flexural strength AND loads due to structural self weight and other environmental effects.

 

[Toby43]

This topic is of interest to myself also and look forward to others thoughts. A few issues that I see would need considering--
1) would the wall be able to provide adequate capacity for the out-of-plane loading from the mezzanine. What sort of connection would you provide? Even if wall could be justified to carry the load per length of wall at mezzanine elevation, I imagine it would be difficult to maintain integrity of connection at this junction if wall becomes highly damaged due to out-of-plane flexural action.
2) If mezzanine is been attached to 1 wall only, then out-of-plane twist would occur in wall when motion is parallel to wall
3) If mezzanine is provided with its own stiff lateral system out-of-plane of wall, would incompatibility with surrounding occur i.e. maybe flexible roof above or differential in wall flexibility at mezzanine termination?

Toby

 

[AThor]

Toby,

The "wall" isn't really a wall, it is more like a big block of concrete. There are two adjacent floor areas, one with a floor elevation ~5 feet above the other. The mezzanine is going in the lower area, the the top of the mezzanine will match the floor of the upper area. So, the "wall" it will tie into is really the side of the slab on grade of the upper floor area, which is also resting on a thick block of concrete acting as a 5 foot tall wall.

One tie in connection we came up with is a steel angle, one leg bolted into the wall, and the mezzanine deck spot welded onto the other leg. One reason for this thread is that I'd like to get a good idea of the seismic load so I can see if the bolts will hold it. I guess none of the frame seismic coefficients would apply for loading in the direction tied into the wall. Any ideas for estimating the seismic force acting to pull out that wall connection? Bolt shear would come from the seismic load in the perpendicular direction, where the seismic frame is being used. Although it is a bit unclear how much seismic force in that direction would be taken by the wall connection versus the frames.

@Shu, thanks for the explanation. I will have to spend some time working through that to build up an understanding.

 

[KootK]

So, the "wall" it will tie into is really the side of the slab on grade of the upper floor area, which is also resting on a thick block of concrete acting as a 5 foot tall wall.

With this being the case, I don't think that it's appropriate to be using moment frames in either direction except perhaps as a redundancy measure if your client is willing to pay for it. The wall/blob/SOG will wind up being your lateral resisting element in both directions baring a very unconventional connection detail to the wall. And that's mostly good news. Cheaper construction. I'd recommend designing both the in plane and out of plane connections to the wall as R=1. As far as lateral systems go, this should be about as stiff and non-ductile as it gets.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.