Micro-Deval Test: Procedure, Percent Loss and Limits

The Micro-Deval test is a durability check for coarse aggregate. A sample of stone is soaked, then tumbled in a steel jar with water and thousands of small steel balls. Whatever grinds down finer than a 1.18 mm sieve counts as loss. Highway agencies use it to judge whether a quarry's rock will hold up in asphalt, concrete and base courses, and to watch for changes in a source over time.

Crushed stone calculator →

The short version

What it measures: resistance to abrasion and grinding in the presence of water.
The standards: AASHTO T 327 and ASTM D6928.
The result: percent loss by mass. Lower is better.
Why water matters: many rocks are weaker wet than dry, and the test is built to show that.

Where the test came from

The method traces back to French work on aggregate wear and was developed further in Canada, where the Ontario Ministry of Transportation supplied a reference stone for labs to calibrate against. British Columbia, for example, replaced its older degradation test with Micro-Deval in 2005, noting that the older test depended too much on the technician. Many U.S. states now list it alongside or in place of the Los Angeles abrasion test.

Equipment

A jar mill that turns at 100 rpm, give or take 5.
Stainless steel jars of about 5 liters with a sealed lid.
An abrasive charge of 5,000 grams of magnetic stainless steel balls, each about 9.5 mm across.
Sieves from 19.0 mm down to 4.75 mm for building the sample, plus a 1.18 mm sieve for the loss.
An oven held at 230°F (110°C) and a scale that reads to 1 gram.

How the test is run

A plain-language outline. Follow the governing standard for the details.

Wash, dry and sieve the stone, then rebuild a 1,500 gram sample to the set grading. For the standard grading that is 375 g of 19.0 to 16.0 mm, 375 g of 16.0 to 12.5 mm and 750 g of 12.5 to 9.5 mm. Smaller gradings exist for finer stone.
Weigh it. Record the dry mass as A.
Soak the sample in 2.0 liters of tap water at about 68°F (20°C). The 2003 ASTM text required at least an hour. AASHTO-based state versions use an overnight soak of 15 hours or more.
Load the jar with the soaked stone, the same water and the 5,000 g steel charge, then seal it.
Run the mill. The standard grading turns for 2 hours, roughly 12,000 revolutions. The finer gradings run for 105 or 95 minutes.
Wash out the fines. Pour everything over a 4.75 mm sieve stacked on a 1.18 mm sieve, hose it until the water runs clear, pull out the balls with a magnet and throw away what passes 1.18 mm.
Dry and weigh what stayed on both sieves. That mass is B.

The math

Micro-Deval loss (%) = (A − B) ÷ A × 100, reported to one decimal.

Example (made-up numbers): a 1,500 g sample comes out of the oven at 1,272 g. 1,500 − 1,272 = 228 g lost. 228 ÷ 1,500 = 0.152, so the loss is 15.2 percent.

Precision depends on the loss level. The multi-laboratory table in the 2003 ASTM version lists a coefficient of variation of about 10 percent of the mean at a 5 percent loss, falling to around 5 percent of the mean at a 21 percent loss. In plain terms, two good tests on a soft stone will agree more closely, as a share of the result, than two tests on a very hard one.

Typical limits

There is no single national number. Each agency sets its own based on local rock and experience. As a reference point, the 2003 ASTM version carried a guidance table drawn from Ontario practice:

Use	Maximum loss (%)
Subbase made of granular material	30
Granular base	25
Asphalt base course and secondary road surface	21
Asphalt surface course	17 to 18
Open-graded base, structural concrete	17
Concrete pavement	13

Your state DOT may use different values, so treat the table as a sense of scale rather than a spec.

Keeping results honest

The balls wear down with use. Labs regularly run them over a 9 mm screen and throw out any that fall through. They also test a calibration aggregate, one with a known loss somewhere in the 15 to 25 percent range, every so often and chart the results. The Ontario reference stone averaged about 19.1 percent loss in a multi-lab study, and individual results were expected to land between 17.5 and 20.7 percent.

Micro-Deval vs LA abrasion

Dry vs wet. LA abrasion runs dry. Micro-Deval runs with water, which matters for rocks with clay seams or high absorption.
Impact vs grinding. The LA drum drops large steel spheres onto the stone. Micro-Deval's small balls mostly rub and grind.
Repeatability. Micro-Deval was adopted in part because it gives tighter results between labs.

Neither test alone proves a rock is good. Agencies often pair them with soundness, absorption and freeze-thaw checks, and with service history from roads already built with that stone.

Why it matters for driveways and base

Stone that breaks down under traffic and weather turns into fines. In a gravel driveway or road base, that means a surface that ruts, holds water and gets muddy. In asphalt, soft aggregate can lose its texture and become slick, the problem described in our page on polished aggregate. Hard rocks such as the basalt in a trap rock driveway are popular because they tend to resist that wear. When you are ordering stone, the crushed stone calculator converts your area and depth into tons.

Questions people ask

What does the Micro-Deval test measure?

It measures how well coarse aggregate resists wearing down when it is rubbed and ground in water. The result is the share of the starting mass, as a percent, that ends up finer than the 1.18 mm (No. 16) sieve. Lower loss means tougher, more durable stone.

What are the main numbers in the test?

A 1,500 gram sample, 2.0 liters of water and 5,000 grams of 9.5 mm steel balls go into a steel jar that turns at 100 rpm. For the standard 19.0 to 9.5 mm grading it runs for 2 hours, or about 12,000 revolutions.

How is Micro-Deval different from the LA abrasion test?

The Los Angeles test tumbles dry stone with large steel spheres in a big drum, so impact plays a big part. Micro-Deval uses small balls and water, so it leans toward grinding and wear in wet conditions. Some rocks that hold up dry break down faster wet, which Micro-Deval is designed to catch.

What is a passing Micro-Deval loss?

Limits are set by each agency and depend on the use. A table published with the 2003 ASTM version, based on Ontario practice, listed maximum losses of 17 to 18 percent for asphalt surface course, 21 percent for asphalt base and 25 percent for granular base.

How long is the sample soaked?

The 2003 ASTM text called for at least one hour of soaking. State methods built on AASHTO T 327, such as Colorado's and Montana's, soak the sample for 15 hours or more before it goes in the machine.

Why do labs run a reference aggregate?

The steel balls wear and machines vary. Running a known calibration stone at regular intervals and checking the balls on a 9 mm screen keeps results consistent from week to week and lab to lab.

As an Amazon Associate, AsphaltTons.com earns from qualifying purchases. Product links below are sponsored.