Overturning and Sliding Resistance

[strucguy]

This is the exerpt from IBC-2009.
"1807.2.3 Safety Factor. Retaining walls shall be designed to resist the lateral action of soil to produce sliding and over-turning with a minimum safety factor of 1.5 in each case. The load combinations of Section 1605 shall not apply to this requirement. Instead design shall be based on 0.7 times nominal earthquake loads, 1.0 times other nominal loads, and investigation with one or more of the variable loads set to zero. The safety factor against sliding shall be taken as the available soil resistance at the base of the retaining wall foundation divided by the net lateral force applied to the retaining wall.
Exception: Where earthquake loads are included, the minimum safety factor for retaining wall sliding and overturning shall be 1.1 "

The above change seems to address all the misinterpretations with regard to what load combinations need to used when checking wall over-turning and sliding as per IBC 2006 and earlier codes. But, my concern is with regard to reduced 1.1 factor of safety when earthquake loads are considered with a reduced load factor of 0.7. Something doesn't sound right here. Am I missing something ??

 

[WillisV]

I don't think so - 0.7E is a service level earthquake just like 1.0H is a service level soil pressure - they are just saying its OK under the reduce the factor of safety from 1.5 to 1.1 when checking seismic loads.

 

[strucguy]

WillisV

Thank you for your prompt response. But, what could possibly be the rationale for this reduction in Factor of Safety?

strucguy

 

[WillisV]

My guess is that typically retaining walls have to resist soil pressures constantly and do so with a great deal of reliability with limited rotations etc. - so a 1.5F.S. is most likely warranted - seismic events on the other hand are infrequent and basically as long as the wall doesn't completely fail in the seismic event it would be considered adequate - so a reduced factor of safety is considered adequate.

 

[JoshPlumSE]

The seismic portion of this new provision is, in my opinion, a very radical INCREASE in the design requirements. Anyone else have any thoughts on this?

To my knowledge, there isn't much of a history of retaining walls failing in earthquakes. In fact, I don't think that even CalTrans designs their retaining walls for seismic.... And, they usually love buried treasure (i.e. really expensive foundations). My assumption was that the wall tends to move with the earthquake motion. Therefore, you don't have inertial forces developing that you would need to design for.

When I was studying for the California SE exam, I spent some time researching the calculation of sesimic forces on retaining walls.

It was fairly difficult to find any design examples where seismic forces were considered for Retaining walls. The only provisions that I found was the "Mononabe - Okabe" method. But, this method is pretty conservative. Anyone know of any advancement in the last 5 to 7 years that would use a more reasonable method?

 

[msquared48]

Also, according to a geotech I frequently work with, when a geotech gives you the lateral load for the wall, the sliding factors, etc., the 1.5 FS is already built into their numbers, so, why should it be applied twice?

Mike McCann
MMC Engineering

 

[WillisV]

Mike, that is a big generalization and not true from many of the geotechs I work with. It is certainly prudent to ask them though to make sure that they are giving you the ultimate failure values of the sliding coefficient etc.

 

[msquared48]

Perhaps so...

Nevertheless, I find it directly stated in many soils reports that the FS is 1.5. I still never rely on this FS in my own calcs ... never ... but it is still there - an obvious overdesign.

The thought was just brought up in passing here - seemed possibly relevant to the code change to me.

Mike McCann
MMC Engineering

 

[InDepth]

You may want to read the NEHRP commentary. Some of the load factors were chosen based on an average....so the actual EQ force could be higher or lower....depending on the resonance between ground shaking and the structure/soil mass. For retaining structures a lot depends on the soil behavior of the wedge that is accelerating...which is dependent on the behavior of the wedge under dynamic loads. Global stability failure plane checks (FS) seem to be rarely checked.

I would avoid the comment that retaining walls don't fail under earthquakes. Check out older ATC and EERI reports with retaining wall failures documented in Chile, Japan, peru, and venezuela....even some walls failing in the US. Even underground structures built by MWD failed by offset or had walls blow out.

Behavior of basement retaining walls is not well understood. Some claim soil arching between floor slabs relieves soil pressure on the walls. Soil properties / stiffness can also play a role. Global arching around basement structures and damping may also play a role.

Liquefaction beneath retaining structures is also an issue.

Geotechs should always provide the factor of safety used in their parameters. They should even indicate if the FS was applied to the final quantity, or to the phi factor (USACE approach). Have the geotech document this is his report.

Monobe-Okabe seems like the only hard written equations for seismic, but is limited by its assumptions. If you want earth pressures on a wall that accounts for SSI, a FEM model with dynamic input would be useful. It would also help to check global stability FS.

 

[JoshPlumSE]

I'm not surprised that there are SOME failures of retaining walls.... Afterall, there are very, very few retaining walls which have actually been designed to resist seismic forces at all.

Even so, I betcha we could take a trip within 10 miles of the epicenter of the Northridge earthquake and inspect 100 walls that didn't have any damage to them even though they weren't designed for ANY seismic loads.

I'm just not all that convinced.... Maybe I should contact the code committee (or the SEAOC seismology committee) to see what the justification behind this change in wording was.

My only real point was always that this requirement is a dramatic increase in the code requirements for retaining walls and EQ's. Especially, if the buidling department is going to start asking us engineers to build "a FEM model with dynamic input". Who has time to do a soil FEM model for a crappy little retaining wall?

If the retaining wall failure would result in the failure of an adjacent building, I think the requirement is reasonable. But, what percentage of retaining walls built fall into this category?