Every successful pavement starts with a mix design that has been tested before the trucks roll. The same principle applies to a shopping trip that ends with exactly what you came for: the outcome improves when you set intentions first. For hot mix asphalt, the Marshall stability and flow test on bitumen is the most widely used way to check those intentions in the lab, and it answers one question: will this mix hold together under traffic?
The test has served highway agencies since the 1940s, when engineer Bruce Marshall developed it for the Mississippi Highway Department. A cylindrical specimen about 4 inches in diameter and 2.5 inches tall is compacted in the lab, heated to the temperature a pavement reaches on a hot summer day, and crushed in a loading machine. The result is two numbers, stability and flow, that predict how the mix will behave on the road.
What the Marshall Stability Test Measures
Stability is the maximum load the specimen carries before it fails, reported in pounds or kilonewtons. The test applies that load at 140 degrees F (60 degrees C), the temperature a dark asphalt surface can reach in summer, when the binder softens and the mix is weakest. Flow is the total deformation the specimen undergoes at failure, measured in hundredths of an inch (0.25 mm units).
Together the two numbers describe how the mix behaves: stability is strength, and flow is flexibility. Reading them as a pair matters more than either number alone, the way a market forecast gains meaning only with context; the 2016 housing forecast that taught home builders to plan around shifting demand is a reminder that a single data point is a signal, not a verdict.
What the numbers tell you
High stability with low flow means a stiff, brittle mix that can crack under repeated loading. Low stability with high flow means a soft mix that ruts in hot weather. The design target sits between those extremes, matched to the traffic the pavement will carry.
Units and reporting
Report stability to the nearest pound or 0.1 kN and flow to the nearest 0.01 inch (0.25 mm). Test reports also record specimen height, compaction blows, and test temperature so results can be compared run to run.
Preparing Specimens for the Marshall Test
Specimen preparation follows a fixed sequence: blend aggregates to the job mix formula, add the binder at a controlled temperature, mix until the aggregate is uniformly coated, and compact the hot mix into a mold. The standard specimen is 4 inches in diameter and about 2.5 inches tall.
Compaction uses a Marshall hammer, a 10-pound weight dropped from 18 inches, striking each side of the specimen a set number of times. The blow count represents traffic: 35 blows per side for light traffic, 50 for medium, and 75 for heavy.
Compaction levels by traffic class
| Traffic class | Blows per side | Typical application |
|---|---|---|
| Light (less than 10,000 ESALs) | 35 | Residential streets, parking lots |
| Medium (10,000 to 1 million ESALs) | 50 | Collector roads, subdivisions |
| Heavy (more than 1 million ESALs) | 75 | Highways, truck routes |
ESAL stands for equivalent single axle loads, the standard way agencies express how much heavy traffic a pavement will see. The job mix formula fixes the aggregate gradation and binder content, and the lab verifies that the compacted specimen meets density and void requirements before the stability run.
Lab technicians and test repeatability
Operator skill drives repeatability more than any piece of equipment. A technician who heats the mix a few degrees too far, or compacts a specimen unevenly, produces a result that does not represent the mix; the same judgment about people that applies to the art of hiring right on a construction crew applies when you staff a quality control lab.
Control specimens
Run a control specimen from a known mix alongside each set of samples. If the control result drifts outside its normal range, the equipment or procedure needs attention before the batch results are trusted.
Running the Test: Conditioning, Loading, and Recording
Before loading, the compacted specimen goes into a water bath at 140 degrees F (60 degrees C) for 30 to 40 minutes so the entire sample reaches test temperature. It is then placed in a breaking head, two curved steel pieces that match the specimen’s shape, and loaded in a compression machine at 2 inches per minute (50 mm per minute).
The testing sequence
- Measure the specimen’s height and weight; record both.
- Condition it in the 140 degree F water bath for 30 to 40 minutes.
- Seat it in the breaking head and apply the load at 2 inches per minute.
- Record the maximum load (stability) and the flow meter reading at failure.
- Remove the specimen, check the failure pattern, and note any unusual cracking.
The flow meter is attached to the loading frame and measures the specimen’s deformation from the start of loading to the point of maximum load. A flow value of 12 means 0.12 inches (3 mm) of deformation, and values typically fall between 8 and 18 for well-designed mixes.
Why the procedure must be exact
Every variable is specified for a reason: bath temperature simulates the hottest pavement, loading rate mimics traffic speed, and the breaking head ensures the load is applied evenly. The patience with process that Norm Abram taught a generation of builders, and that shows up in his insistence on building quality homes one careful step at a time, is exactly what the Marshall run demands.
Interpreting Stability and Flow Values
Agencies publish acceptance criteria that tie stability and flow to traffic. The Asphalt Institute’s classic Marshall criteria are a common starting point.
Typical Marshall acceptance criteria
| Traffic | Minimum stability | Flow range | Air voids |
|---|---|---|---|
| Light | 500 lb (2.2 kN) | 8 to 18 | 3 to 5% |
| Medium | 750 lb (3.3 kN) | 8 to 16 | 3 to 5% |
| Heavy | 1,000 lb (4.4 kN) | 8 to 14 | 3 to 5% |
Air voids are measured separately on the compacted specimen and typically must fall between 3 and 5 percent. Too few voids means the mix can flush binder under traffic; too many means water can enter and strip the asphalt from the aggregate.
Finding the optimum binder content
Mix design runs a series of specimens at different binder contents, typically 4.5 to 6.5 percent, and plots stability, flow, and voids against binder content. The optimum binder content is selected near the peak of the stability curve while voids sit in the target band; that point balances strength, durability, and cost. Once the value is fixed, the plant has a target to hit, and the paving crew has a plan; the site planning discipline that the 2017 solar eclipse taught construction professionals, where preparation decided the outcome, applies to the paving window too.
Common failure patterns
A specimen that crumbles during the test usually means low binder content or poor aggregate coating. A specimen that flattens without breaking indicates excess binder or too much flow. Either pattern sends the design back to the lab rather than to the road.
Marshall Testing in the Quality Control Program
The design test is one moment in a longer quality program. During production, technicians take samples from the plant and run Marshall specimens at a set frequency, commonly one set per shift or per 500 tons, to confirm that the delivered mix matches the design. Field density tests on the finished mat check that the roller achieved the target compaction.
Sampling frequency and records
Keep every test record: date, time, plant silo, binder content, aggregate source, stability, flow, and voids. When a pavement distresses years later, those records are the first evidence engineers examine.
Recycled mixes need the same verification
Sustainability has pushed recycled asphalt pavement (RAP) into most commercial mixes, and RAP changes the binder’s behavior because aged binder is stiffer. Every recycled blend needs a fresh Marshall verification; the fruit flies that taught us about sustainable living showed that a small, controlled experiment can predict how a larger system behaves, and a 4-inch specimen plays that role for a lane-mile of pavement.
Following ASTM D6927 and Choosing a Test Method
ASTM D6927 standardizes the Marshall procedure: specimen dimensions, compaction effort, conditioning temperature and time, loading rate, and reporting format. Following the ASTM D6927 standard procedure for hot mix asphalt design keeps results comparable between labs and between seasons, so an agency can trust a stability number from any qualified laboratory.
Marshall versus Superpave
The Superpave method, developed in the 1990s, replaces the Marshall hammer with a gyratory compactor that kneads the specimen from all directions, which better simulates field compaction. Many agencies use Superpave for mix design and Marshall testing for quality control, so both skills remain in demand.
When to re-test
Re-run the Marshall series whenever conditions change. The test is fast and cheap compared with a pavement that fails early.
- Binder source or grade changes.
- Aggregate source or gradation moves outside the job mix formula.
- Plant production temperature moves outside the approved range.
- Field cores show density or voids outside the target band.