Telling someone to “go pound sand” is not exactly a term of endearment, but it is essentially what is required at any significant earthwork project performed in preparation for the construction of a structure and its supporting foundation system.

By Daniel P. Duffy

Unless they are built on an exposed, stable bedrock surface or are resting on a supporting infrastructure of driven piles, most foundations require that the underlying soil be compacted in place to provide the necessary strength to ensure that the proposed structure remains stable. In many instances, the compacted soil itself may be the structure, as when an earthworks job constructs a berm of structural fill. In addition to achieving desirable strength characteristics, soil can be compacted to reduce its vertical hydraulic conductivity and serve as a barrier to groundwater migration or contain water stored in a pond or lagoon. If left to itself, poorly compacted soil or soil that was never compacted in the first place can significantly settle at different rates over different areas of the site (a process called differential settlement). This can cause damage and even structural failure to occur in the building, foundation, embankment, or pavement resting on the soil. To avoid this potentially catastrophic failure, a stable soil base created by compaction is required.

Soil Compaction Mechanics
Soil compaction involves the controlled application of force to a given depth and area of soil so that the density of the soil increases as its volume decreases. The depth of compaction is especially critical since different types of soils effectively transmit the compaction force to a unique limited depth. As such, loose soil is usually spread out over the compacted area in layers of controlled and predetermined thickness called lifts. The machinery used to apply this force can utilize only the dead weight (static load) of the machine itself, or cyclical dynamic force that is generated by a rotary rammer driven by a reciprocating motor, or (more typically) a combination of the two. The applied loads can be static (continuous and constant applied pressure from a dead weight), kneading (the physical mixing of the soil like bread dough with objects that penetrate to some depth into the soil), vibratory in nature (consisting of a series of rapid but relatively small load applications that literally shakes the soil), or consist of impacts (fewer but heavier and sharper applied loads).

The characteristics of the soil itself also affect the compaction project, setting boundaries on the extent and effect of the compaction process. The first of these is the soil’s moisture content. Each type of soil has a typical void volume. This is the amount of space that is not taken up by solid particles. This void volume can be filled with either air or water. The presence of water can greatly aid the compaction effort by acting as a lubricant that allows the soil particles to slide past each other as they move into a more compact, denser formation. By preventing clay particles from bonding and sticking together, water also allows for a more effective compaction of cohesive soils.

However, there can always be too much of a good thing and water in soil is no exception. After a certain moisture content level has been achieved (measured as the percentage of total soil weight consisting of water), the potential for densification (measured as dry unit weight, the weight of soil once the water has been completely removed by heating) is actually reduced. Compaction normally drives out air from the soil and reduces the void space that is not occupied by water. Excess water over air in the voids prevents further reduction of void space as the densified soil approaches a saturated condition and there is no more air to expel. In cohesionless soils with large particle sizes (gravel, broken rock, etc.), the soil’s moisture can freely drain away, but this is not usually an issue. But in most other soils, water is retained due to their relatively low permeability. The effects of water content result in a bell shaped curve, called the “moisture-density curve” that shows increasing density as water content increases up to a mid range point. Then the curve decreases, showing falling density as water content continues to increase. Between these two conditions at the mid range is the soil’s optimum water content. This corresponds to the soil’s maximum dry density for the particular compaction effort being used.

There is another graph that can be plotted to represent the soil’s moisture and density relationships. This curve compares various saturated conditions (the voids are completely filled with water and no air is present) at differing moisture contents to their resultant dry densities. At these various values, the volume of water in the soil is equal to the soil’s void volume. As such, these values represent the soil’s theoretical maximum compaction and the resultant curve is called the “zero air voids curve.” The zero air voids curve runs from the upper left of the graph (at zero moisture content) to the lower right, at some high moisture content. The theoretical 0% moisture value on the zero air voids curve represents the weight of pure solid soil, with no voids whatsoever. By definition, the other extreme end of the curve (100% water content) would be pure water without any soil, resulting in a dry unit weight of 0 pounds per cubic foot. Most zero air voids curves don’t go beyond the 30% to 40% water content range. Since it represents the theoretical maximum soil density, this curve lies above the moisture-density curve.

The other physical characteristic of the soil that affects its compaction is its particle size gradation. Gradation is a measurement of the distribution of the sizes of the particles that make up the soil’s solid content. Particle sizes are measured by the grain’s diameters in millimeters, and their corresponding screen sizes in the sieves are used to separate out the soil particles. Soils can be uniformly graded, gap graded, or well graded. Most of the particles in uniformly graded soils have the same size. Gap graded soils tend to have two or more particle sizes that predominate. Well graded soils have a smooth and even distribution of particle sizes that range from large gravel to small clay particles. Generally, well graded soils will compact more than poorly graded or uniformly graded soils since the well graded soils have smaller soil particles that can more easily fill the void spaces between the larger particles.

Testing and Measuring Soil Density
Standard and Modified Proctor Tests—There are two types of tests and four different types of field measurements of soil density. The tests are called the Standard Proctor and Modified Proctor compacting tests. These are the tests that result in the moisture-density curves described above. Each represents a different compaction effort and results in a unique moisture density curve that is offset from the other on the density-moisture graph. The Standard Proctor involves the compaction of 1/30 cubic feet (57.6 cubic inches) of soil in three separate lifts. Each lift is subject to 25 blows from a hammer weighing 5.5 pounds that is dropped from a height of 12 inches. This results in an applied compaction force of 12,400 foot pounds. The Modified Proctor requires a greater compaction effort. The same amount of soil is compacted in three lifts as in the Standard Proctor test, but the applied force is delivered by a 25 blows from a hammer weighing 10 pounds that is dropped from a height of 18 inches. The resultant compaction force is equivalent to 56,200 foot pounds (about 4.5 times greater than the Standard Proctor). Each time either test is run, the soils moisture content is varied to achieve a different final dry density. The results are plotted to create the moisture density curve for that soil for the applicable compaction force. For most applications, soil has to be compacted to either 95% of the optimum density determined by the Standard Proctor test, or at least 90% of the Modified Proctor test. In both cases, the matching water content should be wet of optimum, which usually translates into 0% to 5% wetter than the optimum moisture content.

Moisture Content—The field tests used to directly measure the density of in-place moisture and density typically involve the removal of a sample of the soil and the backfilling of the excavated space with a material of known volume. The soil sample is initially weighed and then weighed after heat has been used to evaporate the moisture contained in the soil. The difference between the two weights gives the water content of the soil.

Volume Test—Meanwhile, the hole that the soil has been taken from is filled with sand or a blown up balloon to measure the volume of the now empty space. Backfilling with sand is referred to as the sand cone method. A jar of sand with an upside-down funnel attached is set over the hole and a valve is opened, releasing the sand. The markings on the jar’s side allow for the measurement of the amount of sand that has left the container. It allows for large samples and is highly accurate. However, it is a complicated procedure, and is relatively slow to complete. The balloon densitometer sets a balloon attached to a hand pump and a pressure gauge inside the hole. The pump is used to expand the volume of the balloon until it completely fills the hole. The volume of the balloon and the soil sample can be determined from the air pressure needed to maintain the balloon’s inflation.

Indirect Tests—Indirect methods include the Shelby Tube and the nuclear gauge tests. These both allow for smaller samples and eliminate the need to directly measure the void left by the soil sample. The Shelby Tube is a thin-walled sampling tube driven into the soil to a depth equal to 10 to 20 diameters. The volume of soil in the tube can be measured to a high degree of accuracy since a thin-walled sampler greatly reduces the two main sources of disturbance, displacement, and friction. The nuclear gauge test does away with the soil sample completely. It measures the in-place moisture and density of the soil with radiation from a gamma ray source. Since dense soil absorbs more radiation than loose soil, the soil’s density can be measured by the amount of gamma rays that are radiated back to the detector located in the bottom of the unit. The gamma ray source can be set to various depths or set at the surface, though radiation backscatter from a surface source can be less accurate than a deep source. A small diameter hole is dug to allow the placement of a deep gamma ray source, though the soil volume is not measured. After the test is completed, the hole is backfilled with a low permeability material, such as bentonite powder.

Compacting Different Kinds of Soil
Each soil type has typical optimum moisture content. This can vary from a high of 24% to 12% in clays to a low of 7% large aggregates. The optimum moisture content of clays varies from 12% to 15%. The values for silts range from 11% to 15%, while sands have optimum moistures from 10% to 15%. In general, the range of possible optimum moistures is proportional in mixed soils such as clayey sands, clayey silt, sand-silt-clay, etc. Furthermore, the values given above are for inorganic soils. Organic soils tend to have much higher optimum moisture contents. However, organic soils tend to make poor foundation soils and are difficult to properly compact so are typically removed and their excavation backfilled with stable soils.

There are two broad categories of organic soils: cohesive soils and aggregates. Granular soils range in particle size from fine sands (0.003 inch) to coarse sand (0.08 inch) to fine gravel (0.08 inch) to medium gravel (1.0 inch) and cobbles (3.0 inches). Sand and gravel are noted for their free draining properties resulting from their high hydraulic conductivities. As such, sand densities are not as dependent on moisture content as other soils. The best method for compacting sands is vibration applied with a vibratory smooth drum or padded drum. The lift thickness subject to compaction should be 10 inches to 12 inches If the vibrations can be set to approximate the natural frequency of the soil, then maximum density can be obtained.

Smaller than aggregates, cohesive soils have a much lower hydraulic conductivity. Clays are the smallest particles with sizes ranging up to only 0.002 inch. Silts are somewhat larger with particle diameters up to 0.003 inch. Their cohesiveness results from the slight molecular bond between the soil particles. Cohesive soils are far more dependent on moisture content to determine their optimum densities. For compacting silts and clays, a type of roller (either pneumatic tire or padded sheepsfoot) is preferred. Pneumatic tires tend to work better on non-plastic silts, while the sheepsfoot works best on plastic clays.

Measuring Productivity, Operating Costs and Payment
We now go from the theory of soil compaction to the practice of properly measuring soil compaction out in the field. Production rates for soil compaction are measured in cubic yards per hour (CY/h). The kinds of cubic yards being measured depend on their location and physical state. Cubic yards of soil in their natural state (in the ground prior to excavation and use) are referred to as Bank Cubic Yards (BCY). Soil that has been excavated and is either being hauled by a truck to the job site or is sitting in a stockpile waiting for use is measured by Loose Cubic Yards (LCY). After loose soil has been compacted in-place, the resulting volume is referred to as Compacted Cubic Yards (CCY).

The excavated loose state is an intermediate stage useful for projecting hauling requirements, transportation costs, as well as stockpile volumes and their footprint areas. An apples-to-apples comparison requires a ratio between the initial bank state and the final compacted state. The act of compacting soil by definition reduces its volume. The ratio of final compacted cubic yards to initial bank cubic yards is the “shrinkage factor” (SF), which is almost always less than 1.0:

Shrinkage Factor, SF = CCY / BCY

Similarly, the excavation of soil disturbs its natural structure and increases its void volumes so that the soil experiences an overall increase in volume. The ratio of intermediate loose cubic yards to initial bank cubic yards is the bulkage factor (BF, not to be confused with the soil’s “swell factor,” which measures the increase in soil volume resulting from increased moisture content). The value of the bulkage factor varies from soil to soil depending on their physical characteristic (moisture content, initial void volume, cohesive or aggregate soil types, etc.), which is always greater than 1.0:

Bulkage Factor, BF = LCY / BCY

Finally, there is the relationship between the loose soil and the compacted in-place soil. Unlike swell factors, which are a result of the soils inherent physical characteristics, the ratio of spread loose soil volume to subsequent compacted soil fill volume is usually a function of the compacted fill’s construction specification (though the soil’s inherent physical characteristics place boundaries on the amount of compaction that can be accomplished). For example, if a structural fill specification calls for the soil to be placed in loose lifts of 8 inches thick and subsequently compacted to lift thickness of 6 inches, the resultant “compaction factor” (CF) would be 75% (or 6 inches / 8 inches). Compaction factors are always less than 1.0:

Compaction Factor, CF = CCY / LCY

Each factor and volume measurement has a different role to play in the evaluation of soil compaction productivity and the costs of soil compaction. Excavation costs are based on bank cubic yards; hauling costs are based on loose cubic yards; and compaction costs are based on compacted in-place cubic yards. Soil needs are derived from the application of the shrinkage factor of the borrow soil, to the design volumes of the required field compaction.

Payment, as opposed to costs, is usually made for the amount of the finished product—the total volume of compacted cubic yards. Payment, even payment to a hauling contractor, is almost never made on the basis of loose cubic yards or truckloads delivered to the site. This is because it is almost impossible to judge the amount of soil coming into a job site per truck load or to physically measure individual truck deliveries to see if each is actually at load capacity. The truck loads capacity can also vary greatly depending on the truck model and whether the truck has been filled to struck or heaped capacity. Sometimes, the weights of the loads being carried by incoming trucks are determined by running them over a truck scale located at or near the job site. However, since the unit weight of the soil can vary greatly depending on its moisture content (which can change during excavation, stockpiling, and even during hauling), payment based on weight is not recommended. For these reasons, payment for soil compaction work should be based on actual in-place compacted cubic yards as determined by volume measurements made using before and after ground surveys of the soil compaction site.

Measurement of soil compaction field productivity depends on the dimensions and operating speed of the soil compaction equipment, the resultant thickness of the soil lift being compacted, and the number of passes required to achieve the specified in-place density. These factors are combined in the following formula:

P = (W × S × L × 16.3) / N
where
P = compaction productivity (CCY per hour)
W = the compacted width, usually the width of the contact drum or twice the width of the compactor’s wheel (feet)
S = the compactor’s operating speed (miles per hour)
L = the compacted lift thickness (inches)
16.3 = a conversion factor that takes the value of the first three multiples, feet × inches × miles per hour, and translates it into compacted cubic yards per hour (5,280 feet per mile × 1 foot/12 inches × 1 cubic yard/27 cubic feet, resulting in a value of 16.3 CY / (mile × inch × feet).
N = Number of passes made with the compacting equipment needed to achieve the required in-place density, as previously determined by field testing on a Boutwell pad (each) (http://www.gradingandexcavation.com/mw_0606_boutwell.html).                    

Daniel P. Duffy, P.E., is employed by URS Corp. in Akron, OH.

GEC - March/April 2008

Home | Search | Subscribe | About | News | Advertise | Register | Services | Industry Events
Keep Informed | Contact Us | Current Issue | Back Issues | ForesterPress | StormCon

© FORESTER MEDIA, INC.
P.O. Box 3100 • Santa Barbara, CA 93130 • 805-682-1300