IBC 2009 1807.2.1 Retaining Wall Keyway

Never noticed that before, but I'm not sure I agree with it anyway. As you know, particularly for tall walls, this can really make a big difference in design. I don't normally do that, and I am not sure how you accurately account for it without being ridiculously conservative for tall walls.

Why would active pressure not be applied to the keyway?

EIT

RF:

Exactly, as all loads, to include lateral, must be traced to the foundation, of which, in this case, the keyway is an integral part.

Sorry for the excessive punctuation...

Mike McCann
MMC Engineering

There is no active earth pressure if one side is not excavated. A keyway is surrounded by soil on both sides. If you look at NAVFAC DM 7.2, attached here, they suggest passive resistance be applied to the keyway and not active pressures. Think of a trench filled with concrete. The opposing lateral earth pressures cancel each other. Now make a small excavation on one side of that trench and you created active earth pressure on the cut side and passive on the embedded portion.

So are we saying walls design in 2005 with are underdesigned because active pressure was not imposed?

I do not see where FixedEarth's attached NAVFAC Figure 15 says not to take active pressure to the bottom of the key. The figure does not show or define P_H. Additionally, the figures on the previous NAVFAC page do not show any keyways below the footing. Therefore, the shown P_H's extend only to the bottom of the non-keyed footings.

The attached design example #2 picture from Hugh Brooks' Basics of Retaining Wall Design, 9th edition, shows the active pressure going only to the bottom of the footing, not to the bottom of the keyway. However, Figure 8.1 in Brooks' design mauual shows pictorially and states "For sliding force, extend (active pressure diagram) to bottom of key."

Brooks also states, "When using a key to increase sliding resistance, both IBC '09 and the 2010 California Building Code have added Section 1807.2.1 which reads: 'When a keyway is extended below the wall base with the intent to engage passive pressure and enhance sliding stability, lateral soil pressure on both sides of the keyway shall be considered in the sliding analysis.' This means the wall lateral force must be calculated to the bottom of the key rather than the bottom of the footing. This can significantly increase lateral force and require design modifications. This requirement does not affect overturning which is still calculated about the base of the footing."

www.PeirceEngineering.com

I've always wondered about this. It takes several (sometimes many) feet for your passive pressure to overcome your additional driving active pressure on the keyway (because your starting points for your pressure is at different elevations). The effect is that the first couple of feet make your sliding worse. Good luck if your geotech recommend you ignore the first 2 or 3 feet of passive pressure.

Does anyone really think that adding a small keyway to a retaining wall will make the problem worse?

Lets say you have a stable retaining wall without a keyway and decide to add a 2ft deep keyway for giggles, do you really think the addition will make your wall start to slide?

I personally have a hard time accepting that. If you find that easy to accept, please help me understand. I agree that the forces need to balance, but maybe our view of the forces is simplistic.

PEinc- Thanks for your input. Here is my issue with the 2009 IBC on keyways. There is no distinction on two variables:

One- If active earth pressure should be extended below the bottom of the footing (when you have a keyway), does that active earth pressure start at zero at the base of the footing, or do you continue it from the lateral earth pressure value at the base of the footing. For a 10 ft tall wall with a 3 ft deep keyway that would make a huge difference.

Two-It should matter where the keyway is placed. Obviously if the keyway is below the toe the active earth pressure will not walk laterally to engage the keyway. Similarly if keyway is below the stem. The case ca be made for a keyway below the heel to be designed for active if it is flush with the stem. If the keyway is offset from the stem, the earth pressure line of application is broken, so we have to start the active at zero value. Logically, there has to be an offset distance that will eliminate this active earth pressure completely.

If you look to other countries, a relieving slab is incorporated into the stem to zero out the active earth pressure just below this slab. Same thing for the keyway.

Now we open the gates to problems with older designs. Retaining walls designed with IBC 2003 and 2006 with a keyway will now fail! How come- they are perfectly straight in the real world.

The backfill soils (active earth pressure) does not recognize property line, author's conclusions and changes in building code. It follows stresses and strains. Our designs should do the same. We should follow the soil loading.

I know few structural engineers who stopped using keyways because it is very conservative under the 2009 IBC. I have talked to others who aren't even aware because they just use sofwtare and check for o.k. output.

We should have in my opinion 4 cases: Keyway under the Toe, under Stem, between toe and stem and under L-shaped wall(flush with the stem).

FixedEarth, can the active pressure really be zero at the top of a key if there is also vertical surcharge (bearing pressure) pushing down on the footing? I say no. If you haven't finished designing the wall, you don't know the bearing pressure. Therefore, it seems that applying active pressure all the way to the bottom of the key is about as reasonable close as can be modeled in the design process without doing iterations. I suspect that's IBC's rationale.

It is actually pretty interesting because we generally use wedge theory (ranking / coulomb) to establish lateral earth pressures. Unless the key is located at the heel of the footing the lateral earth pressure from the active wedge is actually pushing on the earth between the key and the vertical face of the wedge. I would think this soil would have some "shear" resistance which is not accounted for. The passive wedge is interesting as well as it is in a similar situation.

EIT

I agree with bpstruct & Gumpmaster. RF- keep it coming-you may join our camp.

PEInc- Yes I though about the bearing pressure and you have a point. Technically, we can say the trapezoidal bearing pressure on the base slab has 0.3 to 0.4 lateral stress component. But when you have a keyway, these pressure are opposite and for the most part cancel each other. If you had a keyway and the passive side was exposed, then yes, you can apply the small active earth pressure due to the toe or heel bearing pressure.

Anyway, I will finish a small paper and submit to IBC in few weeks, I will attach a copy of it to this for at that time.

In my last sentence, I was trying to say, I am putting together a paper for ICC and will post it here when it is available. Would be nice to hear from our international members on retaining wall keyway design.

Attached is a sketch of what I was trying to explain earlier.

This now has me thinking about overturning...If checking about the toe you may just us the weight of the key but real for OT to occur there would be some sort of passive resistance from the assumed active soil pressure side.

EIT
www.HowToEngineer.com

Your sketch looks about how I would draw it. However, a sliding analysis based on the sketch would require adding the two bearing pressure surcharges (acting in opposite directions under the toe and heel) to the sliding analysis and calculating the shear strength of the soil behind the keyway, under the heel. The analysis would probably give a less conservative wall design than what IBC is saying to do for the sliding analysis. Code writers and many engineers do not want a less conservative design - even if it is more correct.

Two-It should matter where the keyway is placed. Obviously if the keyway is below the toe the active earth pressure will not walk laterally to engage the keyway. Similarly if keyway is below the stem. The case ca be made for a keyway below the heel to be designed for active if it is flush with the stem. If the keyway is offset from the stem, the earth pressure line of application is broken, so we have to start the active at zero value. Logically, there has to be an offset distance that will eliminate this active earth pressure completely.

Click to expand...

How far do you think this offset should be? A few feet for most small-scale retaining walls? I'm going to make a habit of putting the keyway at the edge of the retaining wall toe to eliminate this issue.

Based on the 2006 ICC code change proposals and the documentation (available on the ICCsafe website) the IBC provision was inserted based upon design requirements for flood walls and similar use on saturated soils. The provision does not REQUIRE the application of active pressures on the back of keys, it requires a designer to "consider" the need to apply loads on the key. ("...lateral soil pressures on both sides of the keyway shall be considered in the sliding analysis."

I am currently rewriting our design guide on cantilever retaining walls, and we will discuss the specific applications that require these loads to be applied (specifically saturated soils where they act as a fluid), and where they do not (everything else.)

Abusementpark-

It turns out on the extreme end of things, if your keyway depth is more than the heel offset, you are safe to avoid any active earth pressure on the back side of the keyway. See my diagram. Let us hope that the ICC agrees to this for the 2015 IBC version.

With a large horizontal offset between the key and the back edge of the foting, the active pressures could be more like that those of a properly separated pair of terraced retaining walls. There probably would not be a full, height active pressure, but rather one large active presure diagram extending fron FG behind the wall to the bottom of the heel and also a second, smaller, active earth pressure behind the key. Movement is needed to develope passive pressure. This movement should also cause active pressure behind the key. The important point to consider is offsetting the key, as FixedEarth has shown, far enough away from the back of the footing so that you don't develop one large active pressure that extends from FG to the bottom of the key.

The USACE Engineering Manual for retaining walls shows the driving earth pressure extend to the bottom of the key. Even though there is soil on both sides of the keyway, the lateral pressures do not balance since one side of the retaining wall is heavily surcharged vs the other side. If the keyway were placed at the end of the heel of wall, the driving earth pressure would extend down the face of the concrete, just as it does along the thickness of the base slab (which may also have soil on both sides).

It is reasonable to assume some reduction in the driving pressure as the keyway moves away from the end of the heel - I use the same concept as shown in the sketch by RFreund, exccept that I assume that opposing lateral pressures due to vertical surcharge from the slab above are balanced (approximation). The code only requires that the lateral pressures be "considered" since this is a complex issue with pressures varying with keyway position.

The design concept is based on a simple "free-body" diagram of the forces on each side of the wall - unfortunately, the exact values of all forces is not so "simple".

It might be a little late but I have been wondering about this very thing, then I stumble across CalTrans Bridge Design Specifications 5.6.3, it states "For walls with footing keys of depth, Dk , which is greater than the distance, Bk , from the back face of the footing key to the back face or heel of the wall footing, the lateral earth pressure loading shall extend to the level of the bottom of the footing key." Which goes along with what FixedEarth has said.

Here is a link to Section 5 of the Cal Trans Specs.