Earthquake Design 2

[GWSystems]

I am designing a RC square shaped tunnel. This structure will set on grade and have a conical pile of rocks piled on top of it. I have questions pertaining to EQ design of this structure maybe someone can help:

The code I am using gives the "minimum required analysis procedures" that I am allowed to use. Are they referring to the minimum as being the most conservative? Or are they referring to being the least accurate? Which method is more conservative and or least accurate? Response Spectrum Analysis, Equivalent Lateral Force? Response History Analysis?

Since I am familiar with structural dynamics and computer modeling, I'd rather use RSA or RHA, but is this OK if the code say that the "minimum allowable precudure" is equivalent lateral force?

Lastly, does anyone have suggestion on how to account for the load that the pile of rocks will induce on the tunnel when an earthquake strikes?

Thanks. Please help if you can!!!

 

[AggieYank]

GWSystems. If the code you're using says to use equivalent lateral force, i would definitely use that, since it is the minimum. However, you can also do a RSA or RHA analysis, and from the combined results, do your design.

How the pile of rocks acts depends. I would assume they're "fixed" to the top/roof, basically from friction. Determine the weight of the rocks, its fairly simple from there. It seems like it would get messy if they aren't fixed somehow to the top, unless its a place where the rocks won't be able to move off the top (high walls on each side or some similar situation).

Are the rocks going to just sit on top of the tunnel? Or are they going to be attached somehow?

 

[Qshake]

Typically the minimum required analysis procedures will result in the more conservative force based analysis and will also result in a design that ensures no loss of life though the structure may have extensive damage.

A more refined (RSA or Time History) analysis will be more accurate for most structures but I'm not certain with a tunnel and a pile of rocks. Using this analysis it is more likely that you can design the structure so that it meets a owner specified critieria of being servicable following a large earthquake.

Another consideration is that particular to underground structures. Most will not see a large force from inertia like above ground structures do. Most concern is over the differential ground motion and the usual lack of movement joints in a tunnel to go along with the movement.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[GWSystems]

Aggie - The tunnel is box shaped (similiar to a culvert) and the conical pile of rocks is loosly piled on top. There is nothing to fix it laterally other than friction.

If I chose to do a RSA or THA, what would be the best way to include the pile of rocks in determining the mass matrix for the eigensolution? Or similiarly, for the equivalent lateral force procedure, what mass should I use in coming up with the equivalent laterl loads?

Thanks for the help!!

 

[Qshake]

You may simply not include the rocks for the lateral force. Friction has never been an acceptable "positive anchorage" for any analysis involving seismic forces.

Lastly, I'm not sure that the person or entity that derived the design criteria for this project is correct in asking for the lateral force method. There is a great deal of soil structure interaction that must be accounted for which will not be in the crude lateral force method.

I suggest a discussion with a geotechnical engineer or geophysicist on the matter and check out ASCE's Journal of Geotechnical Engineering for related information.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[AggieYank]

Qshake. When you say "Friction has never been an acceptable "positive anchorage" for any analysis involving seismic forces.", I agree. However, I believe the idea is that you can't engineer something to be held down by friction for seismic resistance. In the case of an analysis though, the worst case would be the rocks not sliding off the top, essentially acting like their attached to the top of the tunnel. I would definitely use the full weight of the rocks in my seismic analysis.

 

[Qshake]

Aggieyank,

You've mentioned it in your reply..."essentially acting like their attached"...they are not, so why include them. They could slide across the top in or out of phase with the ground motion. I would include them for vertical dead load though. This will keep the compressive force in the walls of the tunnel and help in the beam-column action of the walls.

Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[AggieYank]

Qshake. I would include them as if they were fixed because as you said, they might slide accross the top in phase with the motion of the top, or, they may not slide at all.

This seems similiar to the analysis of a suspended pool. In the case of a pool, I believe you are required to use a multiplier of 2.2, or something along those lines for the seismic lateral force of the water. There is a thread on this site about it.

 

[Qshake]

First let me just say that I think this is a good discussion and I don't wish to detract from that. Learned people will benefit from this discussion.

I would not be opposed to keeping the load in the lateral analysis given the right conditions.

But an enclosed pool offers a boundary condition that is at least consistent the a positive load path. that is the inertia of the water acting against the sides of the pool or in the case of a tank, the water mass acting on the wall of the tank.

In the case of rocks, it is just friction. However, I do agree that not all of the rocks may fall away from the tunnel.



Regards,
Qshake

Eng-Tips Forums:Real Solutions for Real Problems Really Quick.

 

[VoyageofDiscovery]

I am absolutely in awe of the abilities of refine computer analysis, however you should have such a soil-structural system verified through lab tests or similar journal research. Not to discount this type of analysis, however I get really wary of computer models when there are huge repercussions.

My two cents.