Is anything less than 95% standard proctor acceptable? 2

[RioVerdeDirt]

For example, using a sandy/gravel fill with a maximum dry density of 129.2 pcf the contractor was only able to achieve 91% compaction. This house pad is between 2-4' and will support a slab on grade post-tensioned foundation. Any thoughts on if the compaction is adequate and will not settle over time?

 

[oldestguy]

Assuming testing had no errors, and errors are common, what would you do if it was natural ground and at the same density? Also, where did that magic number of 95 come from? Why not 100? Why not 90? Anyhow, I suspect the classification would be "dense". The foundation appears to be one that can bridge uneven support. You probably would build on the natural ground situation without question.

 

[cvg]

review / redo the proctor, 95% should be easy unless the gravel size is large, you might need to do a rock correction

 

[Ron]

I suspect your contractor either has a wrong Proctor as cvg noted or he is not putting appropriate effort into the compaction. I have seen many residential contractors who do not have a clue how to properly compact soil and have no idea of the moisture-density relationship. A sandy gravel should be relatively easy to get 95% of the standard Proctor. We would typically specify 95% of the modified Proctor for such material.

 

[RioVerdeDirt]

Thank you for your replies.

Had another test done by another geotech company. Turns out the first company was using a dry density number that was too high. The pad was above 95% in 2 locations, and above 98% in the other 2 locations.

 

[BigH]

Was the original (or even the new) testing company using a "real" proctor value of the material as on site or just a value that is typically used in the area??

 

[oldestguy]

Gosh. I didn't get any bites. Was hoping to stir up some conversation.

 

[cvg]

well, practically speaking - if I am going to build on engineered fill, than I want it to be well compacted and I want to verify that. I would never build a structure on natural ground without any idea what the in-situ density was. Perhaps with topsoil stripped and then proof rolled, I might build on "natural ground". Assuming a well graded mixture of sand and gravel with adequate water, than 95% of modified proctor is not difficult to obtain and will improve the long term performance of the structure, than why not require it? Why skip steps on the foundation? Certainly there are other areas to save time and money besides that.

 

[RioVerdeDirt]

BigH,
The original testing company tested the material before it was delivered then used that value to compare. The new testing company tested proctor value on site.

 

[BigH]

RioVerdeDirt - well, there you go. Our Contractor is doing proctors, etc on the borrow pits but the specs (and rightly so) specify the proctor and other tests on the material as delivered. OldestGuy - there was an article some time ago in Ground Engineering about the "95% Fixation". Why not 100%, why not 90% - depends on standard or modified Proctors, engineer's experience and bias and an "historical" (or is that hysterical) fixation on 95%. I'm not sure anyone would be able to reasonably tell me the behavioural difference between 93% vs 95% relative compaction (presume, say, standard). If one states, well, we want to minimize settlements . . . and there is really a difference between 93 and 95, then would there be settlement problems if I achieve 95 and 98?

Most times when slabs are designed, it is "always" with a layer of crushed or gravel beneath. I've even seen one recently (to my consternation) that the site had a granular base, then the poly vapour barrier, then a 50 mm lean concrete mud mat, then placing the BRC and a 100 mm slab. I can't say I've seen a mud mat between the slab and granular before . . .

 

[BigHarvey]

oldestguy, I tried to help by putting a star but it didn't work !

 

[GeoPaveTraffic]

Ok oldestguy, it is late on Friday and I'm ready for a beer so I'll bite.

To me there is a difference between natural soil at a given density and fill at a given density. Addmittedly that difference is much less with a well graded sand and gravel such as being discussed here, but there is still a difference.

Part of the difference is that natural deposits are assumed to varry uniformally between borings. Now we all know that is not always the case, but it is still almost always one of our base assumptions. Fills on the other hand can and do vary considerably in short distance, unless the contractor does a good job of control. That is where density testing comes in, consistently good tests generally mean that the contractor has good control.

As for 95% modified versus other compaction requirements, nothing special in my mind. The higher you spec the better job the contractor has to do to meet the requirement. Most of my work is done in silts and clays, we typically use 95% modified under structures and slopes and 85 to 90 for landscape areas. For base rock we typically use 98% standard (our base is usually limestone and we found that the material tended to breakdown some under modified energy).



Mike Lambert

 

[BigH]

There's a nice paper in ASCE Special Publication #068, "Unstaurated Soil Engineering Practice", by Iraj Noorany where he talks about structural fills in California - usually with expansive clays . . . they typically compact wet of optimum by a 3 to 5% and only to 90% Modified Proctor.

 

[oldestguy]

Ok a few bites.

Then, I can't recall ever checking out an excessive settlement job that can be blamed on say 88% or some very far off test result.. Usually it is more in the category of no compaction or excessive lift thickness leaving loose zones. Building on cinders or, a sideways slip into loose stuff.

So, at 95% we know it works. Some scientific reasoning there.

I do know of an excessive swelling job or two when 95 percent was actually achieved on an expansive clay. The learning curve was expensive then.

In a few jobs with residual clay soils for K-Mart parking lots I have asked for an unconfined compression of at least a given figure, say 2.0 T/sf, with a minimum percent of Modfied Proctor at 83% (because acceptable undisturbed ground was 83%.) Mr. Freezing ground then takes over regardless the next winter.

 

[darthsoilsguy2]

my theory is that people defaulted with 95% standard, then they had a job with a problem. The problem was not the spec, but the materials/contractor/testing not working out.... and so they brought the numbers up.... 98%.... 100%... 95% modified..
when i wrote earthwork specs, i wondered what kind of feedback i would get from contractors if i specified 103% standard compaction for the stone.

one item to point out for the non-soil-testing types. There is a precision statement to the proctor testing. You can certainly get 2 different numbers on the same soil and both are valid tests. I'm always skeptical when i hear about soils being double-checked by others. Fill soils are rarely homogeneous and technician locations are dubious.

 

[oldestguy]

I wonder if Mr.Proctor tested his "new" lab method against the sheep treading on fresh earth.

Also, testing can have some built in errors. A funny example was a time I was on a proposed landfill site and an opposing party hired a nearby very experienced professor to counter what I might find. We had a test pit dug. I mentioned the soil there was "structured" (a soil scientist term). He hadn't heard of that term. Probably also never heard of "peds", as in pedology. He had along his lab tech to do some field density tests. His main device was a quart jar fitted with a small cone and filled with white sand. That goes to tell what sort of soils engineering education comes from his office and the degree of precision his students were taught..

I looked forward to having him on the other side in court some day, but it didn't happen.

 

[aeoliantexan]

I think there is a significant difference between, say, 90%, 95% and 98% of maximum Standard Proctor density. For your example, relative density would probably be a more consistent predictor of strength and compressibliy than Proctor, because cohesionless soils tend to jump out from under the Proctor hammer and give erratic Proctor results. But let's say that 95% Proctor is equal to 70% relative density. The difference between 0% and 100% relative density is probably only about 25 pcf. At 90% of maximum Standard Proctor density, the relative density would be roughly 50%, generally considered too low for reliable foundation support.

For clayey fills, the big difference appears when the completed fill is subjected to water infiltration. The lower the percent compaction, the more voids available to fill with water. Assume a silty clay with a maximum Standard Proctor density of 105 pcf and optimum water content of 20%. Compacted at optimum,the consistency will probably be very stiff, with an unconfined compressive strength between 4 and 8 kips per square foot. At 98% compaction, it can absorb enough water to have a water content of about 23%, 3% above optimum. It would probably be stiff (2 to 4 ksf). At 95% compaction, the potential saturated water content is about optimum + 5%, and it would be firm (1 to 2 ksf). At 90% compaction, the potential saturated water content would be about optimum + 8%, and the soil would be soft, unsuitable for footing support and likely to compress significantly under its own weight.

If you can keep water out of the fill, The degree of compaction is much less critical than if the fill may get wet in the future.

 

[Ron]

I agree with aoliantexan about the significant differences in the compaction from 90 to 98 percent of the standard Proctor. The relationship is one of void ratio, which is not linear when related to compaction.

While relative density might perhaps be a better predictor of performance, it is difficult to find a lab with the equipment to properly run the test and its repeatability with various soil types is questionable. The moisture-density relationship (Proctor) coupled with appropriate in-place density testing is a valid and repeatable test series that can be readily applied and interpreted. One problem is that most contractors don't know much about it and do not use it to advantage.

But the question of what percent is appropriate remains. For repetitive loading such as pavements, 95% of the standard Proctor is woefully deficient, particularly in sandy soils or clayey/silty sands above optimum moisture. One must also keep in mind that compaction does not equal stability. You can achieve compaction in soils that have little or no load-induced stability. Don't mix up the two!

Back to the void ratio. Can the void ratio change in soils that were only compacted to 90% of the standard Proctor? Of course, but the change has to be induced by something. The most common "somethings" are changes in groundwater levels (if the level rises into the 90% compacted soil and then recedes, the void ratio will generally change, resulting in compaction greater than 90%....which results in settlement. The quantification of the settlement is site specific and depends on several variables) and vibration. The vibration can come from adjacent traffic, equipment within the building or even thunder. The goal of selecting an appropriate compaction level to specify is to mitigate such future settlement.

Also keep in mind that the two Proctor methods, standard and modified, were developed based on available equipment to cause compaction. In the early days of the use of moisture-density relationships, the equipment was not particularly heavy and certainly not sophisticated. Vibratory compaction was not used. For those considerations the "standard" Proctor was correlatably useful, particularly in clayey sands, clays and silts. As equipment became more sophisticated and variations in material blends were used, it was recognized that one could easily achieve over 100% compaction in some soils with the better equipment. That led to the development of the modified Proctor to better exemplify the compactive effort that could be put into the soil with better equipment.