masonry wall design question 1

[tcaoz]

I am designing a building with concrete roof and masonry walls. When I designed the roof I assume the roof is supported by three walls and calculation shows that the middle wall has a moment beside reaction. My question is when I design the middle wall do I need to consider the moment for out-of-plane force? thanks

 

[bnickeson]

It depends on how you detail the connection of the concrete roof to the masonry. Since you have moment in the wall, I'm assuming you left this connection fixed in your model. If so, in order to get a true moment transferring connection, you would have to rigidly connect the roof structure to the wall using fully developed bent bars which are lapped with the concrete reinforcing (negative moment) and also lapped with the masonry wall reinforcing. In this case, you would then consider the moment for out of plane bending in your wall.

If you detail the connection to act more as a pin, in which case there wouldn't be the reinforcing I mentioned above, you would not have to design the wall for the moment forces. But I would warn you that whichever detailing method you decide to go with, it is best to model it using those assumptions as well, as it will affect the moments in your conrete roof slab.

Also, don't forget there is a 5 psf minimum distributed load on interior walls, regardless of your design assumptions. Hope this helps.

 

[tcaoz]

bnickeson:

Thank you for the reply. I am a little confused. If calculation shows that there is moment on wall, can this moment be ignored by using certain connection detailing? Probably I should use two simple span beams instead of using continuous beam with 2 spans if I want to eliminate the moment. please advise, thanks.

 

[bnickeson]

How are you modeling the concrete roof and wall system? Do you just have a continuous beam over a central support or do you have the walls modeled as well (as stick figures)? The reason I ask is you can potentially get several different answers depending on the way you have it modeled.

Is your concrete roof system a slab-on-metal-deck style roof or is it going to be a formed roof? This question will help out on the detailing issues a little bit.

As for your questions: if your model or calculations show moment on a wall, that moment has to be designed to as long as your detailing reflects the modeling conditions you have. See my first reply for help on that. On the modeling end of things, the amount of moment transferred into the wall depends on the stiffness of the wall itself. If you have an extremely stiff wall, say a 12" solid grouted CMU wall, there will be a larger portion of moment transferred into that wall than say a 6" CMU with grouted cells at 48" o.c.

If your concrete roof is a monolithic pour, I would not model it as two simple span beams. You would end up with much more positive moment in the roof than need be and not any negative moment over the interior wall. If you want to be conservative, I would design the wall for the out of plane moments that you are getting. That way you can rest assured that you are designing for the worst case moment that would get into that wall, regardless of the connection detailing.

 

[kslee1000]

You made mistake in the analysis if you do not wish the wall to take moment. Your slab is continuous over 3 simple supports, but remember to release the rotation restrain on top of the walls if using computer analysis, thus there is moment in the slab over the middle wall, but none of the walls will have moment.

 

[tcaoz]

Thanks bnickeson and kslee1000. When I did the calculation I modeled the roof as a continuous beam with 3 supports and the roof will be a monolithic pour concrete roof.

per seismic design requirements (ASCE-7) roof has to be anchored to wall. to my understand, if there has the anchorage, there will have the rotation restrain, am I right? please help. another question is, are the moment inside roof and moment on the wall related, just like shear and reaction are related?

please bear with me if they are stupid questions.

 

[kslee1000]

It's matter of reinforcing details. The dowels from wall to slab should be kept minimal, but still adequate to avoid complete seperation of the elements at the joint, it's mainly a shear phenomenon, not necessary to resist moment. Remember, it only takes slight rotation to release the majority of bending moment. However, I would add some add'l top & bottom bars in the edges of the end slab (only) in anticipation of some moment (M=0 when you assume pinned connection), and only continuing the wall bars, that are obsolutely necessary for the strength of the wall, into the slab with minimum development length.

Look into record drawings of other similar projects in the office, you should be able to find good details for use.