Subgrade modulus for multiple layers

[rainandcm]

Typically the recommended subgrade modulus values are for one type of soils, say compacted gravel, stiff clay, etc. I happen to have a case that I will have 2 feet of compacted gravel sitting on medium-stiff clay. What is the best way to estimate a good subgrade modulus for this case? Maybe weighed average for soil within 5B range?

Thank you.

 

[oldestguy]

If this is for a rigid pavement, likely the shallowest is used. For AASHTO Ottawa Road test design method it doesn't make a much difference in required thicknesses if the number varies by a lot.

 

[Ron]

If for any pavement section, layering is common. Each has a different strength and influence. Treat them separately, whether you are using AASHTO methods or elastic layer analysis.

 

[rainandcm]

Thanks everyone. This was actually for a concrete pad (16'x40').

Ron said "treat them separately". I am not sure what this means. My structure guy would only ask for one value for his analysis.

 

[Ron]

If you have distinct layers, give your structural consultant a value for the layer immediately beneath the slab. The lower layer would be considered subgrade and would not be counted on to mitigate loading, but would be a "responsive" support layer only.

You have mentioned a "subgrade modulus", which is different than a "modulus of subgrade reaction". A subgrade modulus would be given in psi; while a modulus of subgrade reaction (K), has units odd units of psi per inch of deflection.

If you are required to give the K value, then keep in mind that ultimately, it takes a relatively large variation in the K value to greatly affect the thickness of the overlying slab.

 

[Mccoy]

rainandcm,
if I understand well you need a Winkler Modulus, or maybe a spring stiffnes, for structural design.

You have first to assign an active depth. You hint at a 5B value but that may be too large. Opinions are not all the same and there are good reasons for differing opinions.

I would follow the suggestions of the GCR 12-917-21 NIST publication, available on line, which are specific for thsi kind of problem.
Relations 2.18b to 2.18d give the effective depth, zp, for different impedances (translational, rocking-torsional).
Relation 2.18a is an harmonic mean, weighted to the layer's thickness, although this is not written there.

The procedure would be the following:

1)Assign a zp depth by the 2.18 relationships
2)Assign elastic moduli, E or G, to each layer
3)Calculate the harmonic average of E or G by formula 2.18a (just substitute G or E to Vs)
4)Apply a formula where the Winkler modulus or the spring stiffness is correlated to E or G, using the above average (like the Vesic formula, or other similar ones).

Pls note that the publications also carries the Pais and Kausel formulas to calculate the spring stiffnes from G.

http://www.mccoy.it