Soil Cement Compaction

[geoguy65]

I am providing regulatory oversight for construction of a soil-cement dam. The aggregate is similar to ABC with around 10% fines, 40 to 55% finer than No.4, and non-plastic. The fill is being constructed as soil cement (relative compaction and moisture content) as opposed to an RCC mix. My problems/questions:

1) For several nuclear and sand cone densities, the degree of saturation for the compacted fill exceeds 100% - often > 105%, and occasionally as ridiculous at 180%! I have not been able to come up with a logical explanation.

2) The difference between the sand cone and corresponding nuclear densities (wet) range from 2 pcf to as high as 13 pcf, and more typically in the range of 5 to 6 pcf. The typical range is higher than I would expect to see. Again, I cannot find a reasonable explanation.

We've checked the "usual suspects" - sand cone calibration, sand calibration, sp. gr. of the soil cement components, math errors, etc. Nothing stands out.

Any thoughts on what might be happening? Has anyone had similar experience? Any and all thoughts/ideas/comments are welcome. Thanks in advance.

 

[rockiologist]

Have you tried taking drive cylinders to see a field density? This may not be possible depending on the max aggregate size. Was there any proctor taken and was it checked for errors? Did you try doing a test strip with the nuclear gauge to see if you can come up consistent dendisties? Are you able to proof roll the area? I hope this helps

 

[cvg]

is the soil cement being produced from an approved stockpile? If not, is the soil uniform or could there be deviations in the properties from one lift to the next? Is there tight control on the water content in the mix? Is it being mixed in a batch plant? Do you have good correlation between contractors QC and your QA tests? eg: do his nukes match up with yours?

 

[msucog]

run sand cones and stay away from the nukes.

 

[muuddfun]

How are you doing the moisture contents for the sand cones? You may want to dry one back in the lab oven set low to dry it slowly and see if there may be some chemical reason that you are burning off more than the water when you do your sand cones in the field, or that may be being picked up by the nuc and giving the highly variable readings. Material that has a high hydrogen content other than water can be read by the nuc as extra water and will throw off the readings. Try two nucs as the same location, and then dig a sand cone at the spot under where you positioned the nuc and get a calibration between all three, if it is within about 3 or 4 pounds of each other then that is the best that can be expected with these methods.

 

[msucog]

if you got aggregate, that could make the test more variable if you happen to have small samplings (say you pick up a rock). also, the geology of the rock may fool the nuke due to the mineral makeup of the rock. also, you might check to see that the pulverizer is thoroughly mixing the material so that aggregate is dispersed in to the mix.

 

[geocrete]

If you are calculating the degree of saturation, or you are plotting the density tests on a proctor graph and they plot to the right of the zero air voids curve - indication 100%+ saturation then the specific gravity you are using may be too low. The specific gravity of soils is typically 2.65 and cement is usually 3.15. You may want to have the specific gravity tested.