seismic forces on combined foundation 1

[Deckard451]

The following is a Thread I posted in "foundation engineering" - I got two repiles effectively agreeing with my concern. Any comments from Seismic Experts here?

I have two large stoves approx 30m tall, 7.7m diameter, 10m apart weighing about 1700tonnes each. Seismic forces will govern strength design of an (unfortunately) combined foundation.

Obtaining seismic forces is pretty straightforward but the question is ----

In the event of an earthquake, can the two stoves go "out of phase" of each other or will they always have forces acting on each of them in the same direction.

If they oscillate like a tuning fork then both can move in towards each other and bend the hell out of my foundation. Looking in the orthogonal direction, if one goes one way and the second goes the other way then the foundation suffers a huge torque or twisting action which will double my reinforcement.

Is it possible for the earthquake to induce oscillations in the stoves such that they move in different directions and effectively “fight” each other using my foundation.

I am inclined to think it is possible (mainly because I’m not sure and feel the need to design for it). However we are talking about a fairly sizable increase in build costs and there are some people trying to argue that it can’t happen and that earthquake forces from the two vibrating structures will always act in the same direction.

This alternative arguement is that it is the footing that moves under the stoves and the stoves are mearly trying to "catch up with" the footing. If this is so then I can see that the forces cannot oppose. But my arguement against this is pictures from Kobe which clearly showed that close buildings smashed into each other indicating opposing oscillations.

The two stoves are essentially the same.

Any views would be appreciated.

 

[JAE]

Just looked through the new IBC2000 and they included a section that describes "Building Separations" and it includes this:

"Adjacent buildings on the same property shall be separated by at least dMT where

dMT = square root(dM1^2 + dM2^2)

dM1 and dM2 are the displacements of the adjacent buildings."

Apparently, the IBC feels that there is the possibility that two similar structures can sway together and apart simultaneously, but they temper it a bit with the square root of the sum of squares of the deflections.

I would approach it the same way as you never get two perfectly matched structures (in terms of period, T) so they could definitely develop an opposite sway condition.

 

[cyak]

In Greece when we have two structures, identical or not that are sharing the same foundation (generally combined) we are analyzing and designing the foundation with the possibility of non-synchronous earthquake excitation. Note that this procedure is applied even when the two structures are seperated by a seismic joint, i.e. they are very close.

In general however we use spectrum analysis with SRSS or CQC so the actions on the foundation follow the SRSS rule that JAE gave you... The load combinations are many more than in the single building case.

If you are using statically equivalent loading then the situation will be more conservative...

I do not know how your combined foundation looks but in most cases I have come across I do not get that much of torsion in the foundation and the non-synchronous loading generally causes positive bending moments to increase...

 

[Deckard451]

Thanks for your helpful comments above - I am now finalising design and preparing drawings for construction, confident that the additional foundation costs incurred by this assumption are justifiable.

 

[emceditor]

I believe that building codes "learn" and get refined by natural occurances. A building designed to according to the building code will be always designed for just to a minimum acceptable level according to the standard of the code councils' knowledge.

As the code does not address the loading for this scenario yet, my personal opinion is that unless the building happens to be over an active fault, ground acceleration will be always in the same orientation. But in any case, not knowing anything about these stoves I would think that there is a chance were one stove will be operating while the other is not. In that case the operating weight will be quite different and with that the response to earthquake motion different too. Simplified, I would design this for one stove under full operating weight and the other under the least weight with opposing seismic forces, and naturally also for both stoves under the full operating weight with the seismic forces oriented in the same direction. I would think both cases will induce different stresses on the footing, while the second case will maximize over-turning and sliding of the overall system.

 

[LouRobinson]

There is always the possibility that the two stoves are not quite the same or that the ground under the foundation varies somewhat. In addition, the foundation will experience more than just simple translation: it will experience all six components of ground motion. The stoves will therefore NOT be subject to the same base acceleration at the same time. Even the simple S-waves that are traditionally used for constructing response spectra, will arrive at different parts of the foundation at different times. It is therefore appropriate to design for the out-of-phase response of the two stoves, as intended. However, it is not necessarily essential to design for the combined actions, if you could assure that the foundation would respond in a ductile fashion.

 

[MaxJasper]

The designer of this tower also found an innovative method to save in foundation materials:

http://medweb.rcm.upr.edu/anes/Pizza.JPG