Coefficient of Friction for Slab on Grade

[pricklyPete]

Does any know, at a slab on grade, how much a vapor barrier above or below the sand cushion affects the coefficient of (sliding) friction between the bottom of the slab and the soil. I would assume it would be detrimental.

To make a long story short, I have a block foundation wall which has the S.O.G. at 100-0 on the interior side and the grade is down at about 91-0 on the exterior. Therefore I am retaining a pretty good amount of soil, which means that I have to deal with a calculated 1000 plf horizontal reaction at the top of the wall. This reaction has to be resisted by the slab. If I have a 4" S.O.G. weighing 50 psf and I use a coefficient of sliding friction of 0.40 I would need 50 feet of slab per linear foot to resist this load. I could probably do this if the contractor doesn't discontinue the mesh at the control joints. Any thoughts on how I might be better able to deal with that reaction going into the slab? (Note: I am treating the wall as fixed at the base and pinned at the top, so I have already attempted to reduce the top reaction in that way. Also, the length of the wall is 70' so I can't try to span horizontally between the side walls. Finally, cantilevering the wall up is not an option either for various reasons)

 

[PSlem]

Put a grade beam in about 12' from the wall and treat it as a shear key.

Engineering is the art of not constructing...of doing that well with one dollar what any bungler can do well with two after a fashion.

 

[pricklyPete]

PSlem,

Am I understanding you correctly to say that the grade beam would be parallel with the wall whose top I am trying to brace?

The main concern I have about doing that is that the shear key will place a horizontal load on the soil in front of the key. I don't know how much, but some of this horizontal load will go right back into the wall. Maybe as a geotech, you can help me to understand at what angle the load (the load that is applied to the soil by the face of the shear key) will go down and out into the soil towards the wall that the shear key is holding back. If 50% of that load goes right back into the top of the wall reaction, how much am I really gaining?

 

[flamby]

Your case is not a simple one. If slip-cirle comes into play, your calculations based on coefficient of friction of 0.4 may be optimistic. If you place the shear key deep enough, you will get larger area of slip circle and it improves the stability. You need to study the soil-properties, you are dealing with and any help from a geotechnical person will be very useful.

 

[PSlem]

Figure the resistance from the shear key will act towards the wall and up to the surface at 45*+phi/2 and the active zone of the wall is at 45*-phi/2 from the horz. So for a 9' wall and a phi of 30* the active zone is 4.5' wide at the top and the passive zone from a 1-/2'shear key is 3'. You want to leave about 5' between them for safety so 12.5' back. For a phi of 30* the Kp is about 10 times Ka so there should be plenty of resistance when you take moments about the base.

Engineering is the art of not constructing...of doing that well with one dollar what any bungler can do well with two after a fashion.

 

[rollofox]

By the way, a sand cushion under a slab may be a liability. See;

http://www.buildingscience.com/resources/foundations/sand_layer_under_slab.htm

 

[Focht3]

An interesting hypothesis - too bad it's generally incorrect...

It's pretty clear that the author doesn't have a working knowledge of soil mechanics in general, or of soil suction, the ideal gas law and how relative humidity works. I'd love to hear what Bob Lytton has to say about that 'expert'!

Free water doesn't collect in the "sand cushion" unless there's a leak present. Relative humidity controls the presence - or absence - of moisture in the "sand cushion."

The thin layer of sand acts as a part of the vapor barrier by breaking the capillary rise of moisture. It is not present to protect the polyethylene vapor barrier, reduce concrete cracking, slab curling, or protect the soil from high water-cement ratios.



Please see FAQ731-376 for great suggestions on how to make the best use of Eng-Tips Fora. See faq158-922 for recommendations regarding the question, "How Do You Evaluate Fill Settlement Beneath Structures?"

 

[3dboy]

You have to be very careful in how you detail your slab to wall connection, slab to grade beam or "deadman" connection to make this work. Also .4 is pretty loose on the coefficient of friction we always use .35 for concrete to soil and a vapor barrier is going to knock that down to less than .1 also be careful saying the base of your wall is "fixed" the definition of a fixed base is that there is no rotation. This is something which your computer model does very well, but just isn't very realistic in the real world of soil-concrete interation. If you look into how much rotation you have when you model it as pin-pinned you will find that the rotation is very miniscule and you will really have a hard time detailing the wall to not move that much. Remember the rule of thumb that deflection criteria are usually more stringent than strength criteria. Finally your "deadmen" (or gradebeam) have to be far enough back that the lateral load works its way downward (at probably a 2:1 slope) to below they footing of the wall you are restraining. Finally, no way you are going to get mesh to transfer that tensile load. my money says you gotta reinforce it fairly heavily. If you are trying to make the base of the wall fixed, why not just make it a retaining wall? It is only 9 feet tall.

 

[PSlem]

Just to add another thought, if you use the slab to retain the wall, you will have to prop the wall to support it while backfilling and placing the slab.

Engineering is the art of not constructing...of doing that well with one dollar what any bungler can do well with two after a fashion.

 

[pricklyPete]

Thanks 3dboy and PSlem,

yes Pslem, you're right. We will need to shore the top of the wall until the slab and grade beams have achieved strength.

3d boy, there are two reasons that we chose not to design the wall as a retaining wall. #1) the wall is a block wall and it probably would have required a 16" block with #8 in each core to achieve the base moment in the wall of about 13,000 lb-ft/ft. With the new IBC lap lengths, these bars would have been very long and not easy for the masons to work with. Secondly, there was some concern about the deflection at the top of the wall if it were designed as a cantilever. If we went with a floating slab on grade, I would think the top of the cantilevered wall retaining the earth would have a noticeble amount of deflection in relation to the edge of the slab.

If I had the opportunity to do it over, I would have involved the geo-foam people. Although the cost of the geo-foam may not have been less than the cost of the extra concrete I had to use to get this to work, I think the cost of having to shore the wall might have tipped the scale in favor of the geo-foam. Plus it would have been a lot cleaner and harder to screw up.

 

[3dboy]

yeah you couldn't do that with masonry and those reinforcement requirements. With it being masonry, I definitely don't think you can consider your base as fixed either.

 

[J1D]

pricklyPete,
The picture is not quite clear to me. As you described I have a block foundation wall which has the S.O.G. at 100-0 on the interior side and the grade is down at about 91-0 on the exterior. Therefore I am retaining a pretty good amount of soil, which means that I have to deal with a calculated 1000 plf horizontal reaction at the top of the wall. This reaction has to be resisted by the slab. Why 1000 plf horiz reaction at the top of the wall? How can the slab (S.O.G. at 100-0?) resist the reaction?
Thanks in advance for clarification.

 

[pricklyPete]

J1D,

I was considering the wall to be pinned-pinned. So it was pinned at the footing (at or below 92-0) and pinned at the top at 100-0 (i.e. that SOG). It is retaining earth, but not necessarly a retaining wall in the sense that it has a fixed base with no support at the top. In this case it is supported at the top. Does that clear it up?

 

[J1D]

pricklyPete,
Looking at a wall section, when you consider pinned-pinned support. The top pin is a physical horizontal support (by bracing, ties, etc), then you have the 1000psf horiz reaction at the top. The reaction is to be reisisted by the S.O.G. (as you mentioned) because the wall is tied to the S.O.G. at the top. Am I right so far?

If this is true, the slab must tie to the foundation (wall, grade beam, or whatever) on the other sides. In that case you may not need the slab-grade friction to hold slab.

 

[pricklyPete]

J1D,

I think that you are interpreting the situation correctly.

I generally agree with your suggestion. However, there are a couple of issues that come up.
1) There are going to be contraction joints (if not construction joints) that run parallel with the "retaining wall" and the wall on the opposite side of the building (which happens to be 90' away for this building). Our office standard is that the slab reinforcing stops at both the control and contraction joist. Therefore, there is a break in the continuity of the slab on grade which is acting as a tension member across the 90' length of building connecting wall to wall.

2) If the first point I brought up could be dealt with and the slab were able to be used as a tie across the building, then, in order to deal with 1000 plf, we would need to provide reinforcing bars (as opposed to mesh) in the slab for the entire floor slab of the building in the direction perpendicular to the retaining wall and the wall on the opposite side of the building. I would also think that you would want a slab thicker than 4" too in order to get decent cover on the rebar. So that would be another cost item.

But this second point in an of itself is probably not prohibitive. I think it is more the combination of both of the above issues. But it might be an option worth exploring again.

 

[J1D]

pricklyPete,

For control and construction joints (rather than expansion joints) rabar can go across (continuous or overlap splice)the joints. Even with cracks the wire mesh across a control joint can still transfer tension force.

If using the slab as a horizontal tie is not a solution (because of your office std, cost, etc). How about using tie bars (wall to wall) below the slab? I like ties better than the friction.

BTW, wire mesh may have trouble being secured the position.

Regards,