A dry density measured on site cannot, on its own, show whether a layer is compacted well enough; it needs a laboratory reference for the same material. An Evd value, by contrast, is compared with the acceptance value in the specification, not with a laboratory reference. Below we explain the Proctor, water content and CBR tests we run in our soil laboratory , the classification tests that go with them, and how to bring us a sample.

Why a field result needs a laboratory reference

A dry density that is adequate for a clay fill may be well short of what a sandy gravel can reach. The in-place dry density is therefore divided by the laboratory maximum dry density of the same material and expressed as the degree of compaction (%). The required degree of compaction and water content range come from the project specification on each job. Field and laboratory work then run in this order:

Before earthworks start, a representative sample from the borrow pit, excavation or stockpile is tested for classification, Proctor and, where needed, CBR. If the material meets the specification, its Proctor result becomes the reference for site control.
Each layer is tested with the sand cone method (ASTM D1556/D1556M-24, TS 13872:2019+T1:2026) for in-place dry density and water content, giving the degree of compaction.
Where specified, Evd (dynamic deformation modulus) is measured with an LWD to ASTM E2835-21(2026). LWD measures stiffness and does not replace a density criterion.

The Proctor test: maximum dry density and optimum water content

We carry out Proctor compaction to TS EN 13286-2:2013. The material is prepared at several water contents and each portion is compacted in a standard mould with a defined rammer and number of blows. The peak of the dry density–water content curve gives the maximum dry density, and the water content at the peak is the optimum water content.

The standard has two compaction efforts, Proctor and modified Proctor. The modified test uses more energy and gives a higher maximum dry density and a lower optimum water content on the same material. The specification states which applies, and site results must be compared with a reference from the same method.

A common mistake in Cyprus is using one Proctor result from the start of a job throughout while the borrow pit material changes. Marly, clayey and limestone-derived materials vary over short distances, and an outdated reference gives a misleading degree of compaction.

Water content: why too wet and too dry both fail

We determine water content to TS EN ISO 17892-1:2014+A1:2022 by oven-drying the sample to constant mass, and use the same method on material from sand cone holes.

Too dry: friction between particles stays high, the roller cannot pack them together and the target density is not reached. Clay fills compacted dry of optimum may later swell or settle on wetting.
Too wet: water fills the voids and part of the compactive effort goes into the pore water. The layer pumps and ruts under the roller, and the dry density stays low.

In a Cyprus summer, heat and wind dry a freshly spread layer quickly; water sprayed only on the surface leaves the bottom of the layer dry, so it has to be mixed in. In winter, clay fills can end up well above optimum after rain. Most specifications therefore also set a water content range relative to optimum.

The CBR test: soaked, unsoaked and swelling

We carry out the California Bearing Ratio (CBR) test to TS EN 13286-47:2021. The specimen is compacted in a CBR mould at the water content and density the specification requires, usually based on the Proctor result. A standard plunger is pushed in at a constant rate, and the force at set penetrations, as a percentage of a standard reference force, gives the CBR.

Unsoaked CBR: tested straight after compaction; it shows the bearing ratio of the material as placed.
Soaked CBR: the specimen is immersed under surcharge weights representing the overlying layers, then tested. The soaking period is set by the specification, often four days, and represents the weaker wet-season condition.
Swelling: the rise in specimen height during soaking, reported as a percentage. For the expansive clays common in Cyprus it can matter as much as the CBR itself.

CBR is the subgrade strength input in pavement design and a suitability criterion for sub-base materials. The minimum CBR and allowable swell come from the project specification.

Classification tests that go with Proctor and CBR

Particle size distribution by sieving to TS EN ISO 17892-4:2016, showing the proportions of gravel, sand and fines.
Atterberg limits to TS EN ISO 17892-12:2018+A2:2022: liquid limit, plastic limit and plasticity index, which describe how the fines respond to water.
Soil classification to TS EN ISO 14688-2:2018 from those results, with linear shrinkage to BS 1377-2:2022 added for clays where useful.

A high-plasticity clay can give an acceptable maximum dry density and still show a low soaked CBR and high swell. Classification picks out this kind of material before work starts.

How much sample, how to bag it and how long it takes

The quantity depends on the tests and on the largest particle size. The figures below are approximate; we confirm the exact amount when we quote.

Tests	Clay, silt and sand	Gravelly and granular material
Water content only	About 1 kg	2–3 kg
Sieve analysis and Atterberg limits	3–5 kg	10–15 kg
Proctor	15–20 kg	30–40 kg
Proctor, CBR and classification together	30–40 kg	50–70 kg

Use strong, heavy-duty bags tied securely. For natural water content, use an airtight bag or container with the air pressed out, keep it out of the sun and deliver it the same day.
Sample a stockpile from several points rather than the dried surface crust, and do not pick out the coarse particles.
Do not mix materials from different sources, depths or layers.
Label the outside of the bag and put a second label inside in a plastic sleeve: sample number, source or location, depth, sampling date, a short description and the tests required.

Turnaround depends on the test; we confirm the programme with the quote. Water content needs oven drying, so results are usually ready within 6 to 48 hours. A Proctor test takes a few working days because several points are compacted and dried separately. Soaked CBR adds the soaking period and is usually the longest item. Bringing the sample before spreading starts means site control does not begin without a reference.

Frequently asked questions

Is one Proctor test per material enough?

For material from a single source whose properties do not change, it may be. A new test is needed when the borrow pit moves to a different level, when material of a different colour or texture arrives, or when site results scatter unexpectedly.

The degree of compaction came out above 100%. Is that good?

It usually means the Proctor reference does not represent the material on site; less often, compaction on site has exceeded the energy of the laboratory method. We normally recommend a new Proctor test on material from that area.

Should I ask for soaked or unsoaked CBR?

That depends on the project specification and the designer. Subgrade and sub-base assessment usually calls for soaked CBR and swell, while unsoaked CBR shows the condition during construction. Send us the specification and we will agree the test programme with you.

For Proctor, CBR and classification tests, see the scope of our soil laboratory testing service and contact us to arrange sample delivery or sampling on site.