Field Compaction Control

The specification names a dry unit weight, and the field test proves you hit it.

The plans call for 12,000 cubic yards of embankment, compacted to 95 percent of a Proctor maximum of 120.0 lb/ft³. Before it's built, someone has to answer three questions: how an inspector with a shovel and a jug of sand will prove each lift is dense enough, how many cubic yards to dig from the borrow pit, and how much water to add on the way. They sound like three problems, but they're one. Dig the soil, haul it, wet it, and roll it, and the grains never change. Only the space between them does. Keep count of the grains, and every answer falls out.

Grains per cubic foot

The Proctor Test found the most grains a set effort can pack into a cubic foot of a soil, γD max\gamma_{D\,max}. A field specification doesn't ask the contractor to hit that number exactly. It asks for a percentage of it. The Handbook's Phase Relationships section calls the ratio relative compaction:

RC=γD fieldγD max×100\text{RC} = \frac{\gamma_{D\,\text{field}}}{\gamma_{D\,max}} \times 100

Both are dry unit weights, and that's the point. γD=Ws/V\gamma_D = W_s / V counts the grains alone, so RC compares the grains a roller packed into a cubic foot of fill with the most the lab could pack into a cubic foot of the same soil. The moist unit weight won't do. It counts the water too, and the water in a fill changes with the weather. When a field test reports a moist unit weight, take the water out first with γD=γ/(1+ω)\gamma_D = \gamma / (1 + \omega).

A spec of 95 percent sets a floor: the minimum acceptable field dry unit weight is 0.95 γD max0.95\,\gamma_{D\,max}. For the soil in the Proctor Lesson, with γD max=111.9\gamma_{D\,max} = 111.9 lb/ft³, that's 106.3 lb/ft³, and any lift at or above it passes. A heavy roller can even outdo the lab's hammer, so a relative compaction a little over 100 percent is unusual but not an error.

Just above it, the Handbook prints relative density, DrD_r, which also uses a field and a maximum dry unit weight. Don't trade one for the other. Relative density places a sand between its loosest and densest possible states, with its own tests for both. Relative compaction is the plain ratio to a compaction-test maximum, and it's what a spec means by "95 percent of the Proctor maximum."

Checkpoint
1 of 2

A spec calls for at least 95% of a standard Proctor maximum of 115.0 lb/ft³. A field test on a lift reads a moist unit weight of 122.0 lb/ft³ at a water content of 14%. What is the lift's relative compaction, and does it pass?

A.
106.1%, so it passes
B.
91.2%, so it fails
C.
93.1%, so it fails
D.
98.0%, so it passes

Measuring a hole with sand

To find a field dry unit weight, you need the weight of some grains and the volume they came from. The weight is easy. Dig a hole in the lift, keep every crumb, weigh it, and dry a scoop to get its water content. The volume is the hard part, because a hole dug with a spoon has no shape a ruler can measure.

So fill it with something whose unit weight you already know. That's the sand-cone test. The sand is clean, dry, and uniform, and it's been calibrated, so that poured the same way it always settles at the same unit weight. It sits in a jug above a valve and a funnel-shaped cone. Set the cone over the hole on a base plate, open the valve, and sand runs down until the hole and the cone are both full. Weigh the jug before and after, and the difference is the sand that left.

Soil dug from the hole Base plate Cone Valve Jug of calibrated sand Sand in the cone Sand in the hole Compacted lift
A sand-cone test on a compacted lift

Not all of that sand is in the hole. Some of it stopped in the cone, and that amount is calibrated too, by running the test on a flat surface with no hole. Subtract it, and the rest measures the hole. Divide by the sand's unit weight, not the soil's, because sand is what filled the hole:

Vhole=Wsand used−WconeγsandV_{\text{hole}} = \frac{W_{\text{sand used}} - W_{\text{cone}}}{\gamma_{\text{sand}}}

From there it's the Proctor reduction again: γ=W/V\gamma = W/V for the soil that came out, then γD=γ/(1+ω)\gamma_D = \gamma / (1 + \omega), then RC. The Handbook has no sand-cone test, so carry the procedure yourself. In SI the scale reads masses and the sand comes with a density, so V=m/ρV = m / \rho, and W=mgW = mg, from the Handbook's Concept of Weight, turns the soil's density into a unit weight.

Worked Example A sand cone on a lift
1 of 3

A sand-cone test on a compacted lift gives these results. The soil's standard Proctor maximum dry unit weight is 117.0 lb/ft³. Find the field dry unit weight and the relative compaction.

Sand-cone test data
MeasurementValue
Jug and sand, before18.40 lb
Jug and sand, after7.75 lb
Sand that fills the cone3.30 lb
Unit weight of the sand98.0 lb/ft³
Soil dug from the hole9.45 lb
Water content of the soil12.0%

A lift described on paper

Sometimes a question skips the field test and describes the lift by its phase properties: GsG_s and the void ratio, or GsG_s, the water content, and the degree of saturation. You built the dry unit weight for that in Unit Weights of Soil. With Vs=1V_s = 1, the grains weigh GsγWG_s\gamma_W and the box holds 1+e1 + e:

γD=GsγW1+e\gamma_D = \frac{G_s\gamma_W}{1 + e}

There's no water in it, since dry counts the grains alone. When the question gives ω\omega and SS instead of ee, get the void ratio first from Se=ωGsSe = \omega G_s, which the Handbook's Phase Relationships print as S=ωGs/eS = \omega G_s / e. Then RC is the same ratio as always.

Checkpoint
1 of 2

A compacted lift has Gs=2.68G_s = 2.68 and e=0.52e = 0.52. The soil's standard Proctor maximum dry unit weight is 113.0 lb/ft³. What is the lift's relative compaction?

%

Digging enough borrow

Now run the job from the other end. The plans fix the fill's volume, and the spec fixes its dry unit weight. The soil comes from a borrow pit, where it sits in the ground at its own dry unit weight, usually a looser one. How many cubic yards do you dig?

Count the grains. The fill needs Ws=γD fillVfillW_s = \gamma_{D\,\text{fill}}V_{\text{fill}}. Digging, hauling, wetting, and rolling move air and water around, but no grains are made or lost, so the pit has to supply exactly that WsW_s at its own dry unit weight:

γD borrowVborrow=γD fillVfill⇒Vborrow=Vfill γD fillγD borrow\gamma_{D\,\text{borrow}}V_{\text{borrow}} = \gamma_{D\,\text{fill}}V_{\text{fill}} \quad\Rightarrow\quad V_{\text{borrow}} = V_{\text{fill}}\,\frac{\gamma_{D\,\text{fill}}}{\gamma_{D\,\text{borrow}}}

This is the shrinkage factor from Cut, Fill, Shrinkage, and Mass Haul at work. Borrow is bank soil, and that Lesson defined SF=Vcompacted/Vbank=γD,bank/γD,compacted\text{SF} = V_{\text{compacted}}/V_{\text{bank}} = \gamma_{D,\text{bank}}/\gamma_{D,\text{compacted}}, so Vborrow=Vfill/SFV_{\text{borrow}} = V_{\text{fill}}/\text{SF}. When the fill is denser than the pit, as it usually is, you dig more yards than you place.

A borrow soil is often described by a moist unit weight and water content, because that's what a test pit measures. Take the water out before it goes in the ratio. Moist soil weighs more per cubic yard than its grains do, so a moist value on the bottom says you need fewer yards than you really do.

Worked Example Borrow for the embankment
1 of 2

The embankment from the opening needs 12,000 yd³ of fill, compacted to 95% of a standard Proctor maximum of 120.0 lb/ft³. The borrow soil weighs 118.0 lb/ft³ moist in the pit, at a water content of 10%. How many cubic yards of borrow must be dug?

The same count in void ratios

Some questions give void ratios instead of unit weights. The box with Vs=1V_s = 1 makes this quick. The grains take one unit of volume, and the voids take ee, so the soil occupies 1+e1 + e for every unit of solids. The solids are the same in the pit and in the fill, so the two volumes compare as their boxes do:

VborrowVfill=1+eborrow1+efill\frac{V_{\text{borrow}}}{V_{\text{fill}}} = \frac{1 + e_{\text{borrow}}}{1 + e_{\text{fill}}}
Borrow pit Compacted fill Vs​ = 1 Vv​ = e = 0.80 Vs​ = 1 Vv​ = e = 0.50 V = 1.80 V = 1.50 γD​ = 93.6 lb/ft3​ γD​ = 112.3 lb/ft3​ Same soil, Gs​ = 2.70, same weight of solids Ws​
One unit of solids in the borrow pit and in the fill
The grains are identical, so the pit takes 1.80/1.50 = 1.20 times the volume of the fill. The dry unit weights are in the same ratio the other way up: 112.3/93.6 = 1.20.

It's the dry-unit-weight ratio in another form. Divide one GsγW/(1+e)G_s\gamma_W / (1 + e) by the other, and the GsγWG_s\gamma_W cancels, leaving the two 1+e1 + e's. The 1 is the grains' own room, and dropping it compares the voids as if the grains took no space.

Checkpoint

A soil has a void ratio of 0.85 in the borrow pit and 0.55 once compacted in the fill. How many cubic meters of borrow does each cubic meter of fill take?

A.
1.19
B.
1.55
C.
0.84
D.
1.29

Water rides with the grains

The embankment's borrow arrives at 10% water, and the spec wants the fill rolled near its optimum, say 14%. How much does the water truck add? Water content is a fraction of the grains' weight, ω=Ww/Ws\omega = W_w / W_s in the Handbook's Phase Relationships, so the water travels as a set fraction of the grains. Raising it from ωborrow\omega_{\text{borrow}} to ωtarget\omega_{\text{target}} takes

ΔWw=Ws (ωtarget−ωborrow)\Delta W_w = W_s\,(\omega_{\text{target}} - \omega_{\text{borrow}})

The multiplier is WsW_s, the grains alone. Multiplying the moist weight instead would count the water already in the soil as if it were more grains to wet.

Questions usually ask per cubic yard, or cubic meter, of compacted fill, so the grains to count are the ones in a unit of fill: γD fill\gamma_{D\,\text{fill}} times 27 ft³ for a cubic yard, or times 1 m³. Per unit of borrow, you'd use the pit's dry unit weight, which counts fewer grains. To turn the water's weight into a volume, divide by γW\gamma_W. The Handbook's conversion factors give 7.481 gallons per cubic foot, and its Commonly Used Equivalents give 1,000 kg for a cubic meter of water, which is 1,000 liters.

Worked Example Water for the embankment
1 of 2

The embankment's fill is compacted to a dry unit weight of 114.0 lb/ft³. The borrow arrives at a water content of 10%, and the spec wants 14%. How much water must be added per cubic yard of compacted fill, by weight and in gallons?

Recap

  • Relative compaction is γD field/γD max×100\gamma_{D\,\text{field}} / \gamma_{D\,max} \times 100, from the Handbook's Phase Relationships: the grains in a cubic foot of fill over the lab's best. Use dry unit weights only, so dry a moist reading with γ/(1+ω)\gamma / (1 + \omega) first. A 95% spec sets the floor at 0.95 γD max0.95\,\gamma_{D\,max}. It isn't relative density.
  • A sand cone measures the hole with calibrated sand: the sand that left the jug, minus the sand that fills the cone, over the sand's unit weight. Then γ=W/V\gamma = W/V for the soil that came out, and dry it.
  • A lift described by its phase properties has γD=GsγW/(1+e)\gamma_D = G_s\gamma_W / (1 + e), with e=ωGs/Se = \omega G_s / S when the question gives ω\omega and SS.
  • Earthwork conserves WsW_s, so Vborrow=Vfill γD fill/γD borrow=Vfill/SFV_{\text{borrow}} = V_{\text{fill}}\,\gamma_{D\,\text{fill}} / \gamma_{D\,\text{borrow}} = V_{\text{fill}} / \text{SF}, or (1+eborrow)/(1+efill)(1 + e_{\text{borrow}})/(1 + e_{\text{fill}}) per unit of fill.
  • Water to add is Ws(ωtarget−ωborrow)W_s(\omega_{\text{target}} - \omega_{\text{borrow}}), with Ws=γD fillW_s = \gamma_{D\,\text{fill}} times 27 ft³ per cubic yard of compacted fill, or times 1 m³ per cubic meter.