What is the International Roughness Index (IRI)?

The International Roughness Index (IRI) is a roughness scale that a quarter-car model computes from a measured road profile. IRI is the short form of International Roughness Index. The World Bank defined the scale from earlier US research (NCHRP). The International Road Roughness Experiment, held in Brasilia in 1982, led to it.

The IRI was designed to be relevant, transportable and stable with time. It can be measured with a wide range of equipment, because it needs only a profile. In the US, the FHWA made it the standard for the roughness data in the Highway Performance Monitoring System (HPMS). ASTM E1926 and AASHTO R 43 give the computation.

How the IRI is calculated

The input is a longitudinal profile, the height of the pavement along one wheel track. The IRI analysis simulates a quarter-car model on that profile. The model is a tire spring, an axle mass, a suspension spring and damper, and a body mass. It is called the Golden Car. The simulated speed is 80 km/h (49.7 mph).

The Golden Car has four set parameters. Divide each by the sprung mass (the body mass):

The simulation gives the speed at which the suspension compresses and extends. The IRI is the average of the absolute value of that speed over the length of the profile. This average is called the average rectified slope:

IRI = (1 / L) x integral of |axle velocity - body velocity| dt

L is the length of the profile, and the axle and body velocities are the vertical speeds of the unsprung and sprung masses. The result is a slope, so the units are m/km, mm/m or in/mi. The conversion is 1 m/km = 63.36 in/mi. A smooth pavement gives a low IRI.

The HPMS rules add three details. The profile points are no more than 6 in apart. Wavelengths longer than 300 ft are filtered out. The IRI is reported for sections of 0.1 mile. The FHWA states that the IRI is most affected by wavelengths of about 2.4 m and 15.4 m.

IRI, MRI and HRI

Strictly, the IRI describes one wheel track, often the right one. Some agencies measure both wheel tracks and average the two IRI values. That average is the Mean Ride Index (MRI). The Half-car Ride Index (HRI) runs the same analysis on the average of the two profiles. The HRI is lower than the MRI. For asphalt roads, Sayers found that the HRI is about 0.8 of the IRI. Check which index a report uses before you compare two values.

How the IRI is measured

The IRI is computed from a profile, so the measurement is a profile measurement. The FHWA lists these devices:

• High-speed inertial profilers. A vehicle carries an accelerometer, laser sensors and a distance measuring device. They are accurate and repeatable. They collect data at highway speeds without a lane closure. They cannot collect accurate data at low speed, below about 15 to 20 mph.
• Lightweight inertial profilers. They use the same technology in a small vehicle. Agencies use them on new construction, including concrete that cannot yet carry traffic.
• Walking profilers and rod and level. They give accurate profiles, but they are slow. Agencies use walking profilers to calibrate inertial profilers.
• Response-type instruments. A vehicle with a roadmeter measures the response. These readings are not stable with time and depend on the vehicle. They need calibration to the IRI scale.

A straightedge and a profilograph do not give an IRI. A straightedge shows local deviations only. A profilograph gives the Profile Index (PI), which is a different statistic. Its “in/mi” has little in common with the IRI in/mi. A point laser can read texture on tined concrete as roughness, so the FHWA recommends a line laser on textured concrete.

AASHTO and ASTM publish the supporting standards. AASHTO M 328 is the profiler specification, R 56 certifies profiling systems, and R 57 covers their operation. ASTM E950 covers the measurement of a profile with an inertial profiler. For construction acceptance, agencies typically certify the equipment each year and certify the operator.

IRI ratings for highways

The US federal rules rate highway pavement by IRI. The rule applies to the National Highway System. States report the IRI in inches per mile, for sections of 0.1 mile, and the IRI is computed to AASHTO R43-13.

The IRI rating is one input. For an asphalt section, the overall rating also uses the cracking and rutting ratings. For a jointed concrete section, it uses cracking and faulting. A section is Good only if all of its ratings are Good.

IRI in road construction

Agencies use the IRI to accept new pavement. The FHWA gives typical acceptable values for new highway pavements: 52 to 66 in/mi for asphalt and 57 to 72 in/mi for concrete. The value depends on the agency and the type of facility.

A 2014 survey found IRI-based smoothness metrics in 39 states for new asphalt and in 23 states for new concrete. Many states pay the contractor an incentive for a low IRI and apply a penalty for a high one. Limits that require correction before acceptance typically range from 60 to 150 in/mi. The FHWA notes that corrections are usually for localized roughness only. Most states also check local roughness with a straightedge. A typical limit is a deviation of 3.2 to 6.4 mm over a 3 to 4.9 m straightedge.

IRI and other measures

• PSI. The Present Serviceability Index is a 0 to 5 scale from the AASHO road test of the late 1950s. It predicts a panel rating of ride quality. Some agencies still correlate roughness data with PSI, but the FHWA notes that the PSI is falling out of use. Sayers noted in 1989 that no standard PSI roughness scale exists. Paterson derived an approximate IRI-to-PSI conversion from several sources.
• PCI. The Pavement Condition Index rates visible distress from 0 to 100. ASTM D5340 states that the PCI is no direct measure of roughness. A pavement needs a profile measurement to get an IRI.
• Boeing Bump Index. The airfield method that the next section describes.

The IRI on airfields

The IRI is a highway measure. FAA AC 150/5380-9 states that the highway industry defines roughness by the ride quality of a passenger. For an airplane, the AC defines roughness by fatigue of airplane parts. It also counts effects on safe operation, such as cockpit vibration and high g-forces. The AC adds that passenger discomfort lasts only a few seconds. It is not the basis of airfield roughness.

The AC groups airfield roughness into two kinds. A single event bump is an isolated change in elevation over 100 m or less. Profile roughness is a series of deviations that make an airplane respond in ways that add fatigue, reduce braking or impair cockpit operations.

The AC does not set IRI limits. It uses the Boeing Bump method. A virtual straightedge joins two points of the runway profile, up to 120 m apart. The method compares the bump height with the bump length against three zones:

The FAA ProFAA software calculates the Boeing Bump Index (BBI). A BBI below 1.0 is in the acceptable zone. The FAA standard survey interval is 0.25 m. Use of the AC is mandatory for projects that AIP or PFC funds pay for.

AC 150/5380-9 also warns about inertial profilers that apply a highpass filter. The filter can change the BBI, because the BBI uses longer disturbances than the IRI does. The AC does not recommend such profilers for BBI work. AC 150/5320-6G says to evaluate profiles with ProVAL or ProFAA.

EU guidance does not use the IRI either. GM1 ADR.OPS.C.010(b)(2) gives acceptable, tolerable and excessive heights for a runway surface irregularity of a given length. Above the acceptable limit, the guidance plans maintenance action. Between the tolerable and excessive limits, correction is mandatory. Above the excessive limit, the area of the runway warrants closure.

GM1 ADR-DSN.B.090(a) calls a 3 m straight-edge check, with no deviation greater than 3 mm, an achievable standard for the finished wearing course.

A TarmacView pavement survey rates visible distress to ASTM D5340. It does not compute an IRI.