100 pci subgrade stiffness

[abusementpark]

Maybe this trend is localized to the area in which I practice. But EVERYTIME a geotechnical engineer recommends a stiffness value for designing a soil supported mat or pavement, the design value is 100 pci. It never fails.

Are these guys even doing calculations? Why would the value never change site to site? It this is a safe rule-of-thumb and none of them want to stick their neck out and give a higher value?

Please help me understand.

 

[FixedEarth]

If we have soft to firm clay, the 100 pci value is a default. However, if we start getting into very stiff clays it is near 150 pci and in granular soils, the range of vertical subgrade modulus would be in the 250 to 350+ pci range.

Look up at Bowles Foundation Analysis and Design book for a range of values. On your next job, email the soils engineer that Bowles table and ask for an addendum letter.

 

[Ron]

abp..if I remember correctly, you practice in a coastal plains area. Given that, most of what you deal with are relatively clean fine sands, silts, organic silts, and some clays. All can be mixed together, obviously, with the properties of the mixture not necessarily corresponding to the properties of any of the constituents!

If the material is clean enough to be used as subbase for a pavement or compacted subgrade for a slab on grade, there is no reason to use a k value of 100. To answer your question...no, they are not calculating anything. They're using a very conservative standard that was established long ago by someone in your area and it gets repeated by the local geotechs without question. If they did a few plate load tests, they would likely find a k of 200 or more.

All that being said, the modulus of subgrade reaction doesn't greatly affect the thickness of the slab in reasonable increments. A rule of thumb you can use for k vs. CBR is about 20:1, so for a CBR of 10, the k value would be close to 200. There is a better correlation in Huang's pavement design book, but that's reasonably close.

 

[abusementpark]

We see mostly soft to stiff clays (or silty clays) where I practice. But usually for the softer clays in the greater New Orleans area, we almost always have to use deep foundations.

The other thing I wonder about is how much the value should be predicated on what type of base/drainage course you use. Some geotechs don't even comment on the effect of the different options. For instance, sometimes we decide to put 8" of crushed limestone as a base instead of our typical drainage layer. Should that make a difference as well?

 

[Mccoy]

Some geotechs don't even comment on the effect of the different options. For instance, sometimes we decide to put 8" of crushed limestone as a base instead of our typical drainage layer. Should that make a difference as well?

Maybe it should, in relation to foundation width, to the stiffness ratio with the original soil and to the aspect to be examined.

If we get back to the definition of subgrade modulus, that is loading pressure/settlement, everything which is going to influence settlements will necesarily influence the winkler modulus. the foundation width is a primary concern.

So your 8" would perhaps make a difference beneath a (very) narrow strip, whereas the effect would be null beneath a wide slab.

Such a limited thickness of crushed limestone looks like having a different purpose from that of limiting settlements and improving modulus subgrade, it probably gives some uniformity to the base and improves the phi_cv or phi_constant volume to be used in the sliding failure analysis.

 

[msucog]

abusementpark, is anyone actually paying for a "real" k value? or is it lumped in to a lowball rfp that was bid against 5 other geotechs with a bunch of generic language, a few (too few) borings and no lab testing? not trying to sound pesimistic but too often the rfp wants the world for $3000 on a 30 acre site. if you want "real" recommendations, may i suggest selecting real geotechs that are pricing the work and not trying to underprice the next guy? it is always frustrating when i'm underbid by those that are just trying to underprice the next guy and not doing the client any real service. real geotechs looking out for their clients' interests are not the cheapo firms...ever. that's not to say the clients should pay massive amounts for geotech work...but it has gotten a bit ridiculous with the lowball bid strategies.

 

[ATSE]

abp,
Seems like geotech should be providing two k values, one for short term load, such as traffic loads, and one for long term creep in soft clays.
For a large mat-slab load with a deep lower zone, the k value in the middle should be different than at the edge.
After considering creep, an equivalent k value can be relatively low and change (increase) the flexural demand on a mat-slab.
Even with expensive geotech reports and direct requests, I only ever get a single k value - usually 150 kcf (87 pci).

 

[BigH]

Further to McCoy's note, the 8 inches of crushed limestone may not be anything other than giving a working platform as it is difficult to work on soft to firm clay directly - and no, for any footing having any width, it should be neglected . . .

 

[fattdad]

the whole notion of "long-term" subgrade modulus values irritates me. It's a structural engineers construction of a geotechnical problem. We already have ways of addressing geotechnical performance related to areal loading (i.e., industrial warehouse slabs).

I worked as a geotechnical engineer for an engineering firm that specialized in industrial warehouses. I ran a CBR on every project for both heavy-duty pavement design (AASHTO method) and for correlation to subgrade modulus. Why guess when the data can be had for a few hundred bucks?

Here's the way I see the problem: An industrial subgrade will either be compacted structural fill or prepared native subgrade. The CBR shows the soaked penetration resistance when the on-site soils are compacted to 95 percent relative compaction. So if the on-site soils can be reasonably characterized by the CBR result the correlation to subgrade modulus will be relevant. I also use the ACI subgrade bump depending on the subbase thickness.

f-d

¡papá gordo ain’t no madre flaca!