Marshall Stability Test: Procedure, Criteria and Results

The Marshall stability test is a lab test that squeezes a small, hot puck of compacted asphalt from the sides until it gives way. The peak load it carries is its stability, and how far it squashes on the way to that peak is its flow. Together with density and void numbers, those two readings are how the Marshall mix design method picks the right amount of binder for a paving mix. If you have seen “Marshall” on a mix design sheet or a job mix formula and wondered what the numbers mean, this guide walks through the procedure, the usual pass/fail criteria and what the results tell a buyer about the asphalt on a quote.

The short answer

What it measures: the largest load a 4-inch-wide, 2.5-inch-tall compacted sample can carry (stability, in pounds or newtons) and how much it deforms at that load (flow, in hundredths of an inch).
Test conditions: the sample soaks in a 140 °F (60 °C) water bath for 30 to 40 minutes, then is loaded at 2 inches per minute until the load peaks.
Typical criteria: the Asphalt Institute’s long-standing table calls for at least 1,500 pounds of stability and a flow of 8 to 16 for heavy-traffic mixes compacted with 75 hammer blows per side.
What it is for: choosing the binder content during mix design and checking that plant mix still behaves like the design.

Where the test came from

Bruce G. Marshall worked out the basic test at the Mississippi Highway Department around 1939. When World War II brought bigger, heavier military aircraft, the Army Corps of Engineers needed a simple way to design asphalt for airfields. Its Waterways Experiment Station picked Marshall’s test, added a flow measurement to help catch mixes with too much binder and kept refining the method into the 1950s. Pavement Interactive’s reference desk credits the Army’s worldwide use of the method, plus the fact that the gear is cheap, compact and quick to run, for making Marshall the mix design method likely used most around the globe. It also notes that about 38 U.S. states still use it in some capacity, even though many agencies now design with Superpave.

How a Marshall specimen is made

The lab first mixes trial batches at several binder contents. A common plan is five binder contents in steps of 0.5 percent by weight of mix, with three specimens at each, for 15 in all. At least two of the steps sit above the expected best binder content and two below it, so the curves have a clear peak or turning point.

Each batch is heated to compaction temperature and packed into a mold with a Marshall hammer:

Specimen size: a cylinder 4 inches (102 mm) across and about 2.5 inches (64 mm) tall.
Hammer: a 10-pound sliding weight dropped 18 inches onto a flat foot just under 4 inches wide.
Blows: usually 35, 50 or 75 on each face, depending on how much traffic the pavement will carry. More blows stand in for heavier traffic.

Once the specimens cool, the lab measures their height and bulk specific gravity (Gmb) before any loading, because those density numbers feed the void analysis. Our guide to air voids in asphalt explains Gmb, Gmm and the void formulas in detail.

Running the stability and flow test, step by step

Warm it to test temperature. Specimens go into a water bath held at 140 °F (60 °C), plus or minus about 2 °F, for 30 to 40 minutes. 140 °F is a hot-pavement temperature, when asphalt is near its softest.
Set it in the breaking head. The specimen is pulled from the bath, patted surface-dry and centered in the lower half of a two-piece breaking head. The upper half goes on top, so the load presses on the curved sides of the cylinder.
Load it at a steady rate. The press closes at a constant 2 inches (50.8 mm) per minute. Minnesota DOT’s lab manual allows no more than 30 seconds from leaving the bath to reaching peak load, so the sample does not cool.
Read the peak. The moment the load stops climbing and starts to fall, the operator records the maximum load as stability and reads the flow gauge for the deformation at that point.
Correct for height. A specimen that is not right at 2.5 inches tall gets its stability multiplied by a correction factor from a standard table, and the three results at each binder content are averaged.

ASTM D6927 allows two ways to run it: a traditional loading frame with a proving ring and dial gauges, or an automated setup with a load cell and an electronic displacement sensor that draws the load curve. AASHTO T 245 is the matching highway-agency method.

What stability and flow mean

Stability is the peak resistance the sample develops, reported in pounds or newtons. A higher number means a stiffer, stronger mix at hot-weather temperature, within reason. It rises and falls with the stone type and gradation and with the grade and amount of binder.

Flow is the total deformation from first contact to peak load, recorded in units of 0.01 inch (0.25 mm). Minnesota DOT’s manual gives the example of a sample that squashes 0.12 inch, which is called a flow of 12. ASTM spells out the logic: a flow reading over the top limit at the chosen binder content flags a mix that is too plastic and unstable, and a reading under the bottom limit flags a brittle one.

Neither number is a direct forecast of how many years a road will last. They are standard index values that let a lab compare trial mixes and check them against an agency’s limits.

Typical Marshall design criteria

Each agency sets its own limits, but many trace back to an Asphalt Institute table from 1979, reproduced by Pavement Interactive. Traffic is measured in ESALs, or equivalent 18,000-pound single-axle loads over the design life:

Light traffic (under 10,000 ESALs): 35 blows per side, minimum stability 500 lb (2,224 N), flow 8 to 20.
Medium traffic (10,000 to 1,000,000 ESALs): 50 blows, minimum stability 750 lb (3,336 N), flow 8 to 18.
Heavy traffic (over 1,000,000 ESALs): 75 blows, minimum stability 1,500 lb (6,672 N), flow 8 to 16.
Air voids, all levels: 3 to 5 percent.

The same source lists minimum voids in the mineral aggregate (VMA) by the nominal maximum stone size, for example 13 percent for a 1-inch (25 mm) mix, 14 percent for 3/4 inch (19 mm), 15 percent for 1/2 inch (12.5 mm) and 16 percent for 3/8 inch (9.5 mm). Always check the current spec from the agency or owner on your project; those numbers win over any textbook table.

How the optimum binder content is picked

After testing, the lab plots six curves against binder content: density, stability, flow, air voids, VMA and voids filled with asphalt (VFA). Some patterns are normal. Density climbs, peaks and falls, usually peaking at a higher binder content than stability does. Air voids drop steadily as binder goes up, VFA rises and VMA dips to a low point and then climbs again. Some recycled mixes show stability falling the whole way with no peak.

In the common procedure described by Pavement Interactive, the designer reads off the binder content that gives the spec’s middle air void value, typically 4 percent. That is the optimum. Stability, flow, VMA and VFA at that binder content are then checked against the limits. If any one fails, the mix goes back for redesign, often with a different gradation or stone blend. The finished recipe ends up on the job mix formula the plant runs from.

Limits of the Marshall test

Mix type: ASTM says the stability and flow results apply to dense-graded mixes with a maximum stone size up to 1 inch (25 mm). Big-stone base mixes and open-graded mixes need other methods.
Impact compaction: the sample is packed by a falling hammer, not by kneading like rollers and traffic. Superpave, which many agencies now use instead, compacts samples in a gyratory compactor.
One temperature: the test only looks at the mix when it is hot. Cracking in cold weather depends largely on the binder grade; see what PG 64-22 means on a quote.

What it means for your project and tonnage

For a driveway or parking lot, you will rarely see the test run. It matters because the supplier’s approved mix design, Marshall or Superpave, fixes the binder content and the compacted density your tonnage is based on. To turn a specific gravity into weight, multiply Gmb by 62.4, the weight in pounds of a cubic foot of water. Take a round, made-up example: a Marshall specimen with a Gmb of 2.360 comes to about 147 pounds per cubic foot. A square yard of pavement 1 inch thick holds 0.75 cubic feet of mix, which puts it near 110 pounds per square yard per inch, right at the common planning figure.

A denser or lighter mix moves that number, and your tons move with it. Ask the supplier for the mix’s design density if you want a tight estimate, then plug it into the asphalt tonnage calculator. Our guides to asphalt density in pounds per cubic foot and how density changes asphalt coverage walk through the adjustment.

Questions people ask

What is the Marshall stability test used for?

It is used in the Marshall mix design method to pick the best binder content for a dense-graded asphalt mix, and to check that plant-produced mix still behaves like the approved design.

What is a good Marshall stability value?

It depends on the agency and traffic level. A widely used Asphalt Institute table sets minimums of 500 lb for light traffic, 750 lb for medium and 1,500 lb for heavy-traffic mixes. Your project spec has the final say.

What is Marshall flow?

Flow is how much the sample deforms between first loading and peak load, read in units of 0.01 inch. A sample that squashes 0.12 inch has a flow of 12. Typical limits run from 8 up to 16, 18 or 20, depending on traffic.

What temperature is the Marshall stability test run at?

At 140 degrees F (60 degrees C). Specimens soak in a water bath at that temperature for 30 to 40 minutes and are then loaded at 2 inches per minute until the load peaks.

How many blows are used to compact Marshall samples?

Usually 35, 50 or 75 blows of a 10-pound hammer dropped 18 inches on each face of the sample. More blows are used for mixes meant for heavier traffic.

Is the Marshall method still used?

Yes. Pavement Interactive says roughly 38 U.S. states still rely on it in some form, and it remains common around the world, though many U.S. agencies now design mixes with Superpave.

Related guides

Open the asphalt tonnage calculator →

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