Every layer of embankment, sub-grade, GSB and WMM on a highway is accepted on one ratio: the density reached in the field over the density the same material reached in the laboratory. Either half of that ratio can be wrong. A sound layer can fail on paper, and a loose one can pass.
This article explains OMC and MDD, the Proctor test and the modified Proctor test that produce them, the field dry density test by sand replacement, and the mistakes that spoil the result. The methods are the Indian Standard ones, the IS 2720 series, which is what MoRTH cites.
What OMC and MDD are
Maximum dry density (MDD) is the highest dry density a soil reaches when it is compacted with a fixed amount of energy, and optimum moisture content (OMC) is the water content at which that maximum occurs. Neither belongs to the soil alone. Change the energy and both move: more effort gives a higher MDD at a lower OMC.
That is why a density specification always names the test. "97 per cent" means nothing until it is 97 per cent of the maximum by a stated part of IS 2720.
Light versus heavy compaction: the Proctor test and the modified Proctor test
| Parameter | Light compaction | Heavy compaction |
|---|---|---|
| Standard | IS 2720 Part 7 | IS 2720 Part 8 |
| Name used on site | Proctor or standard Proctor test | Modified Proctor test |
| Rammer mass | 2.6 kg | 4.9 kg |
| Height of drop | 310 mm | 450 mm |
| Number of layers | 3 | 5 |
| Blows per layer, 1,000 cm³ mould | 25 | 25 |
| Energy per unit volume | About 590 kJ/m³ | About 2,700 kJ/m³ |
The 1,000 cm³ mould is 100 mm in diameter and takes soil passing the 19 mm sieve. A 2,250 cm³ mould, 150 mm in diameter, is used for soil with coarse material up to 37.5 mm and gets more blows a layer. The energy row is worked out from the rows above it, not quoted from the standards.
The heavy test puts about four and a half times as much energy into the same volume of soil, and for a given soil returns a higher MDD at a lower OMC. MoRTH's earthwork tables and its 98 per cent requirement for GSB and WMM are all written against Part 8. Some other specifications are written against Part 7, so check which one the contract names before two reports are compared.
Reading the compaction curve
The laboratory compacts the soil at five or more moisture contents. For each, bulk density is converted to dry density.
Dry density = bulk density ÷ (1 + w ÷ 100)
Bulk 2.10 g/cm³ at w = 10.5%: 2.10 ÷ 1.105 = 1.90 g/cm³
1 g/cm³ = 9.81 kN/m³, so 1.90 g/cm³ = 18.6 kN/m³
Plot dry density against moisture and the points rise, peak and fall. On the dry side there is too little water for the particles to slide past one another. On the wet side water occupies space that soil should. The top of a smooth curve through the points is the MDD, reported to the nearest 0.01 g/cm³, and the moisture content beneath it is the OMC.
The shape matters as much as the peak. A sharply peaked curve punishes a moisture error; a flat one forgives it. The MoRTH moisture band for earthwork, from 1 per cent above to 2 per cent below OMC, is a band on this curve, and outside it the roller will not reach density however many passes it makes. For each fill material the contractor submits the MDD, the OMC and the graph at least seven working days before compaction begins.
Field dry density test by the sand replacement method
IS 2720 Part 28 finds the in-place density by replacing the soil dug from a hole with sand of known bulk density. It comes in two sizes. The small pouring cylinder is for fine and medium grained soils in layers up to 150 mm. The large pouring cylinder is for soils containing stones and for layers over 150 mm and up to 250 mm, which covers a full MoRTH earthwork layer.
- 01
Calibrate the sand
Use clean, uniformly graded natural sand passing the 1.00 mm sieve and retained on the 600 micron sieve. Find the mass that fills the cone of the pouring cylinder. Then find the bulk density of the sand by running it into a calibrating container of known volume, of the same depth as the hole to be dug.
- 02
Prepare the surface
Expose and level a flat area, about 450 mm square for the small cylinder and about 600 mm square for the large one, and seat the metal tray with its central hole.
- 03
Dig the hole
Excavate through the hole in the tray to the full depth of the layer: up to 150 mm with the small cylinder, up to 250 mm with the large. Collect everything that comes out and weigh it to the nearest gram.
- 04
Pour the sand
Place the cylinder, filled to the same starting mass used in calibration, over the hole. Open the shutter and let the sand run until it stops by itself, with nothing vibrating nearby. Close the shutter and weigh the cylinder again.
- 05
Find the moisture content
Take a representative sample of the excavated soil for moisture content by IS 2720 Part 2, or dry and weigh the whole of it.
- 06
Calculate
Sand in the hole is the mass run out less the mass in the cone. Divide by the bulk density of the sand for the volume of the hole. The wet mass of soil over that volume is its bulk density, and dry density follows from the moisture content.
The degree of compaction formula
Degree of compaction, also called relative compaction, is the field dry density as a percentage of the laboratory MDD of the same material.
Sand run out of cylinder = 20,000 g - 9,560 g = 10,440 g
Less sand in the cone = 10,440 g - 1,450 g = 8,990 g
Volume of hole = 8,990 g ÷ 1.45 g/cm³ = 6,200 cm³
Bulk density of soil = 12,958 g ÷ 6,200 cm³ = 2.09 g/cm³
Dry density at w = 10% = 2.09 ÷ 1.10 = 1.90 g/cm³
Degree of compaction = 1.90 ÷ 1.96 × 100 = 96.9%
96.9 per cent: accepted as embankment at 95, rejected as sub-grade at 97
MoRTH tests in sets of ten: one set for each 3,000 m² of embankment layer and each 2,000 m² of sub-grade or earthen shoulder. The tables behind the 95 and 97 are in embankment construction as per MoRTH Clause 305.
Core cutter and nuclear gauge: when each is allowed
The core cutter method, IS 2720 Part 29, drives a steel cylinder of 100 mm internal diameter and 130 mm length into the layer, digs it out, trims the ends and weighs it. It is quick and needs no sand. The standard limits it to fine-grained soils free from aggregations, meaning soils of which at least 90 per cent passes the 4.75 mm sieve, and says the cutter should not be used in stony soils. It also samples only the top 130 mm, which on a 250 mm layer is the better-compacted half.
MoRTH names Part 28, sand replacement, as the method for compaction control. A core cutter is therefore something to agree with the Engineer for clayey, stone-free fill, not a default.
A nuclear moisture and density gauge reads both in minutes. Clause 305.3.6 lets the Engineer permit one, on an agreed procedure, provided the gauge is calibrated to give results identical to the sand replacement test. Clause 903.2.2 then doubles the number of tests where a non-destructive method is used. The gauge carries radioactive sources, so custody, training and licensing have to be in place before it reaches site.
The errors that produce a wrong result
- Uncalibrated sand. Bulk density changes with grading, with dampness and from batch to batch. The standard wants each batch calibrated, and the calibration made or at least checked during each day's work. Sand recovered from a hole and reused without drying and re-sieving is the usual offender.
- A hole shallower than the layer. The top of a lift is usually denser than the bottom, so a 120 mm hole in a 250 mm layer reports the good half. The hole goes to the depth of the layer, and the calibrating container has to match that depth.
- Oversize stone. The laboratory MDD is found on the fraction passing the 19 mm sieve. IS 2720 Part 8 accepts that removing up to 5 per cent of stone changes little, warns that excluding a large proportion has a major effect on MDD and OMC, and recommends the larger mould for gravelly soils. A field hole that happens to hold one large stone reads high.
- Moisture measured badly. Dry density is bulk density divided by one plus the moisture content, so an error of one percentage point in moisture moves the result by about one per cent. A sample left open in the sun before weighing is enough to turn 97 into 96.
- A disturbed hole. Levering the tool against the side enlarges the hole and lowers the density recorded. Vibration from a roller working alongside packs extra sand into the hole and does the same.
- The wrong MDD. A borrow area changes with depth and distance. Dividing today's field density by a maximum found on last month's soil is an error no care at the hole can correct.
What to agree before the first test
A field density result is only as good as the laboratory curve under it and the hole it came from. Before the first layer, settle with the consultant which compaction test the contract names, which pouring cylinder the layer thickness calls for, how often the sand is calibrated and what triggers a fresh MDD. On our earthwork, levels, layer thickness and density are checked by whoever the contract nominates, and the work is set out to be tested. Which machine gets a soil to density is covered in which roller for which layer, and the scope we take on is on the earthwork and excavation page.