Technical Papers and Articles

»Nuclear Gauge Vs. Sand Cone for Compaction Test

The following is taken from an October 25, 2001 memorandum from Dr. Tom Petry, consultant for Geotechnical Engineering Consultation & Research.

Re: Problems with the use of nuclear densometers on treated ground

According to our conversation, the contractors are having problems getting the lime treated subgrade dense enough, as tested using nuclear densometers, and winding up with the materials on the dry side of optimum.

It is imperative that the moisture content be kept as high as feasible and practical to aid in the chemical reactions of the lime with the soil and during the building of pozzolans. Moisture levels at the optimum for treated soil plus 5% are good targets for all of the process of treatment until it is time for compaction. For compaction the moisture level should be at the optimum for compaction of the treated soil. After compaction the subgrade should be kept damp at the level of at least the optimum throughout the curing period.

When using a nuclear densometer, one is using the resistance of alpha particles to go through the soil to measure the soil’s bulk denseness or moist unit weight. The level of moisture in the soil is determined by the device sending high-speed neutrons through the soil and measuring the number of slow-speed neutrons that get through. High-speed neutrons are slowed down to the slow-speed to be counted when they go through or near a Hydrogen atom. This means that things in the soil that have Hydrogen atoms in them, other than water, affect the reading to make it appear that there is more water than there actual is in the soil. When Lime is added, of course, many Hydroxyl ions are added that contain a great number of Hydrogen atoms. If the nuclear densometer is calibrated for the natural soil without Lime present, then each reading of a lime treated soil would appear to be higher in moisture and lower in dry unit weight than it really is. This occurs since the meter takes the moist bulk unit weight and divides it by 1 + the water content found to determine the dry unit weight. All of this occurs automatically in the device so no one really knows what the intermediate results are for sure.

This problem can be overcome. The nuclear densometer can be calibrated to read the actual moisture content and dry unit weight before going to the field or when first in the field. This is done by compacting a known weight of treated soil at a known moisture content into a device that allows for the meter to read it. A more likely scenario is to use another kind of device to determine the moist denseness and to find the moisture content using an oven to dry it out, for field compacted treated soils. The two devices I have seen used successfully are a drive cylinder to take a specimen back to a laboratory and a sand cone to measure the volume of the hole out of which a sample is taken. Since both of these methods are known to be less accurate than use of the nuclear densometer, the sampling and testing must be done very carefully. The most practical method seems to be the sand cone method, since most companies have such devices.

To “calibrate” the nuclear device using a sand cone method, one needs to either compact treated moist soil under highly controlled conditions or use both on the same treated subgrade during construction until calibration can be determined. The subgrade is “read” using the nuclear device, then sampled by digging a hole, saving the soil in a sealed container, and using a well calibrated sand cone and sand to find the volume of the hole. The soil is sealed in a container to maintain field moisture levels until it is taken to the laboratory, where its moist and oven dry weights are determined. When a sufficient number of these tests are done, one can determine what the nuclear device is really reading as denseness and moisture levels. It is imperative that all of the sampling and testing done for this calibration be done as carefully as possible.

In summary I would offer the following. Since the nuclear densometer is measuring, without calibration for treated soils, moisture contents higher than reality, the subgrade is thought to be too wet and not dense enough. This forces the contractor to work drier to bring the moisture levels down and to bring the dry unit weights up. This is contrary to the best methodologies to achieve proper treatment results in a lime treated subgrade. In fact, it may well starve the chemical reaction of the water needed to develop the reactions and pozzolan cements desired. The nuclear device being used needs to be calibrated to determine the real moisture level and dry unit weight achieved in the treated soil subgrade. The most likely way is to do side-by-side determinations of water content and dry unit weight using the nuclear device and some other, carefully done, method such as with a sand cone.