Field Density Test Report Generator for Nuclear Gauge and Sand Cone

Heading to the field? Install this site from your browser menu and this tool works with no signal. See how

How to use this tool

Fill the Proctor table first. Every test needs something to be compared against, and until a maximum dry density exists the tool can report densities but cannot report compaction. Give each Proctor a short ID like P1, because that ID is what the test rows point at. On a job with a base course and a subgrade you will have two, sometimes three, and mixing them up is the fastest way to certify a lot that should have failed.

Then set the specification. Required compaction drives pass and fail. The moisture tolerance is optional: leave it blank and moisture is reported but not enforced, which is what you want on a granular base where the specification only names a density. Fill it in on a cohesive subgrade where being wet of optimum is the actual risk.

Now add your tests. The moisture toggle matters. Most gauges display moisture as a mass of water per unit volume in the same units as the density, and that is the default. Some display moisture content as a percentage of dry mass instead. Pick the one your gauge is showing, because the two are not interchangeable and choosing wrong will move every compaction figure on the report. If you are doing sand cone work, switch the method at the top and the test table changes to masses instead.

Press Save Proctor references when you are done. On a contract that runs a season you enter the same maximum dry density dozens of times, and this removes that.

What the tool is calculating

A nuclear gauge does not measure compaction. It measures wet density and it measures water. Everything else is arithmetic, and the arithmetic is where reports go wrong.

If your gauge gives moisture as a mass per volume, dry density is simply wet density minus that moisture, and moisture content as a percentage is the moisture divided by the dry density. If your gauge gives moisture as a percentage instead, dry density is wet density divided by one plus that percentage expressed as a decimal. Both routes reach the same number when the inputs describe the same soil, and the tool was checked to confirm that they do.

Percent compaction is the field dry density divided by the maximum dry density from the Proctor, times one hundred. Moisture deviation is the field moisture content minus the optimum, in percentage points, carrying its sign so you can see whether the lift is wet or dry. A test passes when compaction meets the specification and, if you have set a tolerance, moisture sits inside it.

Sand cone runs differently. The volume of the hole comes from the mass of sand that filled it divided by the calibrated bulk density of that sand, after subtracting the sand left sitting in the cone. Wet density is the wet mass of the soil you removed divided by that volume, moisture comes from the oven dry mass, and dry density follows from those two.

Why one failing test matters more than a good average

There is a habit on site of looking at a run of tests, seeing that they average comfortably above the requirement, and treating the lot as accepted. It is an understandable instinct and it is usually wrong.

Most specifications judge tests individually. The requirement reads that each test shall achieve some percentage, not that the mean of the tests shall achieve it. Under that wording a lot with four tests at 99 and one at 91 has a failing test in it, and the average of 97.4 is irrelevant. The tool flags exactly this case, because it is the single most common way a lot gets accepted that should not have been.

The engineering reason sits underneath the contract wording. Compaction is not a quantity you average across an area the way you might average concrete strength across a pour. A soft spot does not get stronger because the ground fifty metres away is dense. It stays soft, it holds water, and it shows up two winters later as a depression in the wheel path with a pattern of cracking around it. The failing test is telling you where a specific piece of ground is, and averaging is precisely the operation that hides that.

This is also why the tool reports standard deviation alongside the mean. A lot averaging 98 with a deviation of 1 point is uniformly good work. A lot averaging 98 with a deviation of 4 points has something varying across it, usually lift thickness or moisture rather than roller passes, and the next lot will produce a failure somewhere even though this one squeaked through. The spread tells you about the process. The mean only tells you about the paperwork.

A worked check you can repeat on paper

Press Load example lot. It sets two Proctors, P1 at 2100 kg per cubic metre with an optimum of 7.0 percent, and P2 at 1850 with an optimum of 12.0 percent, against a specification of 95 percent compaction and moisture within 2 points of optimum.

Take test one. Wet density reads 2050 and moisture reads 150, both in kilograms per cubic metre. Dry density is 2050 minus 150, so 1900. Moisture content is 150 divided by 1900, which is 7.89 percent, shown as 7.9. Compaction is 1900 divided by 2100, which is 0.90476, so 90.5 percent. That is below 95, so the test fails on density even though its moisture sits only 0.9 points above optimum and is perfectly acceptable.

Test two reads 2200 and 140. Dry density is 2060, moisture content is 6.8 percent, and compaction is 2060 divided by 2100, or 98.1 percent. It passes on both counts, with moisture 0.2 points dry of optimum.

Test five is the instructive one. Wet density 1980, moisture 300, against the subgrade Proctor. Dry density is 1680 and moisture content is 300 divided by 1680, which is 17.9 percent. Optimum for that material is 12.0, so the lift is 5.9 points wet, well outside the tolerance. Compaction comes out at 1680 divided by 1850, or 90.8 percent. Both checks fail, and they fail for the same underlying reason: you cannot compact a cohesive soil that far wet of optimum no matter how many passes you put on it. Drying it is the fix, not more rolling.

Across the five tests the lot averages 94.38 percent with a lowest test of 90.5 and a standard deviation of 3.51 points. Three pass and two fail. Both the average sitting below the requirement and the spread being that wide point the same way, which is that this lot is not ready.

Every figure above was checked against an independent implementation before publishing. The two versions were run against each other on thirty four thousand generated cases spanning both test methods, both moisture entry modes, missing Proctor references, blank rows and deliberately malformed entries, and they agreed on all of them.

Mistakes that catch people out

Reading moisture in the wrong mode. Mass of water per unit volume and moisture content as a percentage are different numbers on the same screen. Entering one as the other silently shifts every dry density and therefore every compaction figure on the report. If your compaction results look strangely uniform or strangely low across a whole lot, check this first.

Comparing against the wrong Proctor. Base course and subgrade have different maximum dry densities, often by two hundred units or more. Pointing a subgrade test at the base course Proctor will usually produce a failing result on perfectly good ground, and the reverse will pass ground that should have been rejected. This is why the tool makes you select the reference per test rather than setting one for the whole report.

Backscatter where the specification wants direct transmission. A backscatter reading averages the top few inches and is influenced by surface texture. It is not comparable to a direct transmission reading taken at lift depth, and it tends to read lower. Record the probe depth you actually used, because a reviewer cannot tell from the number alone.

Swapping wet and dry mass on a sand cone. The tool rejects any test where the dry mass exceeds the wet mass, because that describes soil that gained water in the oven. It is always a transposition, and without the check it would quietly produce a negative moisture content and a dry density higher than the wet.

Retesting until it passes and reporting only the pass. A failing test followed by rework and a passing retest is a complete and defensible record. A failing test that disappears is not, and the gap in the test numbering is usually noticed. Put both on the report.

Forgetting the calibration date. Gauge calibration is normally annual and a report from an out of date gauge can be challenged in full. It takes one field on this form and it is the first thing a reviewer looks for.

Related Calculators and Reference

• Proctor Compaction Calculator, which produces the maximum dry density and optimum moisture that this report compares every test against.
• Field Density Test method reference, covering how the gauge and the sand cone actually work and where each one is appropriate.
• Proctor Compaction Test method reference, for the laboratory procedure behind the numbers in your Proctor table.
• Relative Density Calculator, for clean granular soils where relative density rather than percent compaction is the acceptance measure.
• Daily Work Report Generator, to record the testing alongside the crew, equipment and quantities for the same day.
• Site Inspection Checklist Builder, for the wider pass and fail record that sits around the density testing.
• ASTM International, publisher of D6938 for nuclear gauge testing and D1556 for the sand cone method.