If you would be checking out what benefit you get from differing thickness of base courses using the rigid pavement design criteria developed by the Ottawa Road tests back in the 50's, which is still used by many state highway agencies, you would see differing thickness have very little effect as to life of the pavement for differing traffic counts. About all the base appears to be good for is preventing pumping of the subgrade through the joints and for providing a surface upon which to place the concrete. Thus 12 inches is not much better than 3 inches.

http://www.fhwa.dot.gov/infrastructure/50aasho.cfm

Thanks for your reply OG. I do not have that specific Ottawa Road Test information, but I am aware that for rigid pavement design, the thickness of the base course does not affect significantly the thickness of the pavement (this is perhaps most of the load is taken by the pavement and high stresses do not get transmitted to the subgrade). In my case, with a weak subgrade (high compaction grade cannot be achieved for these soils), the recommended thickness of granular base under floor slabs takes into consideration the low compacted subgrade. For instance, the bottom 50mm~100mm of the base is acting as a "sacrificial" layer to "rectify" low compacted subgrade. This will facilitate compaction operations and ensure high compaction grades for the surface underneath the floor slab. This has been my approach for subgrade preparation for slab-on-grades in locations where the subgrade is weak. I was just wondering if there is any guidance for subgrade preparation for slab-on-grades. Thanks again OG.

Take a look at the figures provided in this reference.

http://onlinepubs.trb.org/onlinepubs/circulars/ec1...

thanks OG, I will take a look at that document. It looks interesting...

Thanks OG... good publication, and a well deserved star. I think Okiryu has it right by calling you OG... has a certain ring of antiquity about it.... <G>

Dik

Okiryu....the road test to which OG refers is the original AASHO (American Association of State Highway Officials, which is now AASHTO, the American Association of State Highway and Transportation Officials) Road Test done in 1959. It has been the primary design and performance impetus for essentially all of the AASHTO pavement design criteria since. Subsequent research has validated the road test many times over and some adjustments to design criteria have evolved over the years.

The graded aggregate subbase you propose will have little affect on the slab thickness. As an example, if you assume the aggregate gives you a modulus of subgrade reaction (k) of 225 you can determine the slab thickness. If you have a relatively clean sand subbase, you would assume a k value of 150 to 175, maybe even 200. The slab thickness will not change appreciably.

The graded aggregate can be used to enhance the properties of a low CBR subbase or subgrade, but should be mixed in to do so. If you use the aggregate as a discrete layer below the slab, it will serve as a good capillary break for drainage or water migration mitigation.

Thanks Ron for the input. Yes, I am aware that the subbase layer underneath the slab has little effect on the slab thickness. When I use subbase granular material, I normally considered an "effective subgrade reaction" due to the contribution of the granular subbase (the "k-value" of the subgrade is increased due to the presence of the granular subbase. USACE has some charts for calculating this "effective subgrade reaction").

Also, this is a different question, but it is related to my original question: what do you think about placing clean gravel (crushed gravel like concrete aggregate, I think it is similar to ASTM C33 no. 6) underneath the slab, say 4" to 8" thick? I understand that it cannot be compacted, but it can be vibrated to interlock the gravel particles. I saw this clean gravel material installed in several projects underneath slab-on-grades and I think that it provides good support for the slab (if it is adequately vibrated).

Requesting further clarification:
Caltrans Empirical method for flexible pavement design(based on Hveem method, I believe) uses GE (gravel equivalent) and GF (gravel factor), in addition to TI and R-value (not CBR).
Table 633.1 attached. There is a "credit" for thicker aggregate base (reducing HMA). How is that different than AASHTO? Learn me.

ATSE - HMA is flexible pavement, not a concrete slab.

cvg - we can agree on that.
The papers included in EC118 included both rigid pavement (plain concrete) and flexible pavement (asphalt) - lots of material there - I didn't read it all.
So OG and Ron's comment apply solely to concrete slabs for pavement.

OKiryu: On your observation abut using clean gravel under pavement slabs. Whether compacted or not, I would be interested in the long term performance of the slabs under traffic, specifically is there a reduction in joint faulting (like driving down a set of stairs). City of Milwaukee, at least back some years, used such a clean base course, whereas, the WDOT did not, but useing rather a well graded base course. This difference was clearly shown on one street where the City paved one side and the WDOT the other. The City's side had practically no faulting whereas DOT side was a rough ride (down the stairs). With no fines to be pumped out of the joint when saturated, the city's road section performed better. In my experience as to performance of rigid pavements, faulting at the joints is one of the main problems that tends to raise the cost of maintenance of these pavements before they break up or have other major problems. Grinding the slabs extends life, but doesn't stop the faulting. This faulting isn't a problem with the continuous reinforced pavements, in my view. That's a whole different story.

OG, as always thanks for share your experiences. Your history indicates that clean gravels are able to adequately support loads. Also, may I ask what happened after that? Did the WDOT start to use clean base courses? Thanks again, OG.

As to WDOT base courses, they did a few two layer base courses just after I left, but with new people in the design management, I didn't see that later. However, I was not involved much with state highway pavements any more, so who knows. The comforting thing about those later days however was the fact that the geotech report procedures and general routine established was carried forth by the new guys.

WinPAS pavement design software is a great way to walk though the AASHTO 1993 design procedure for PCC or asphalt. You can see the effects on the pavement section of changing subgrade material properties and thicknesses (and all the other inputs) really quickly. I highly recommend that software.

Thanks Terratek, I will take a look at that software. For your reference, USACE has also a good software for pavement design called PCASE. It is available freely in internet.