This page was printed from

Long-term pavement testing by FHWA

Features | April 1, 2014 | By: ,

Photo 1: The testing equipment at the FHWA includes the Accelerated Loading Facility (ALF).
Photo 1: The testing equipment at the FHWA includes the Accelerated Loading Facility (ALF).

Contamination by subgrade soil causes deterioration in pavements from progressive loss of effective aggregate base thickness.


A major deterioration factor in pavements is the progressive loss of effective aggregate base thickness because of contamination by subgrade soil. In the service life of a pavement, subgrade soil fines migrate upward into the unbound base aggregate and some of the aggregate can push down into the subgrade soil.

In 1993, FHWA researchers wisely protected the pavement base aggregate layer under the ALF at the TFHRC in McLean, Va., using a separation/stabilization geotextile layer. The ALF (see Photo 1) conducts accelerated pavement testing and applies years of heavy highway truck traffic in a repetitive fashion in a relatively short period of time.

In this facility, different pavement setups were tested during the last 20 years, which imposed the equivalent of more than 8 million ESALs or 18,000-lb. axle loads.

When each pavement was rebuilt, the surface layers were removed and only the upper portion of the aggregate base was reconditioned. The lower part of the unbound aggregate base was left undisturbed. It was particularly critical for this test facility to maintain the same effective lower base section from test to test so there could be a credible comparison between the different upper pavement structures and treatments FHWA evaluated.

Exhumed geotextile

During the summer of 2013, a portion of the full pavement section beneath the ALF was exhumed, all the way down to the subgrade soil. TFHRC staff members performed the full-depth exhumation to see whether the lower part of the unbound base still had its original effective strength or had been compromised because of subgrade soil contamination. To their surprise, they discovered that the aggregate base had been protected and preserved by a separation/stabilization geotextile.

Photo: Test pit excavation down to the geotextile layer.
Photo 2: Test pit excavation down to the geotextile layer.

The area of the ALF pavement where the most rutting occurred was exhumed, excavating a 3-square-yard test pit through all pavement layers. The next generation of researchers found an AASHTO M288 Class 2 woven geotextile that was originally placed in 1993, between the subgrade soil and the unbound aggregate base, which was an approximately 24-inch layer of compacted Virginia DOT aggregate 21A to 95% of AASHTO T-180. The subgrade was a saprolitic soil with approximately 50% passing the 200 seive, a likely candidate for inducing upward fines migration to contaminate the overlying aggregate layer. In this case, because of the geotextile layer, the aggregate appeared to be as clean and intact as originally installed and there was no rutting at the interface between the base and the subgrade, as shown in Photo 2. This exhumation at the TFHRC presented a great opportunity to examine both the residual integrity of the unbound aggregate base and the survivability of the geotextile.

Figure 1: Base aggregate gradation curves—Grain-size curves for the exhumed aggregate superimposed between the limits for VDOT 21A. Source: FHWA
Figure 1: Base aggregate gradation curves—Grain-size curves for the exhumed aggregate superimposed between the limits for VDOT 21A. Source: FHWA


The aggregate was sampled and grain-size tested to compare to the original specification for Virginia 21A. The results of the grain-size curve for the exhumed aggregate are compared to the original 21A specifications in Figure 1.

Next, the exhumed geotextile was tested so its properties could be compared to the original properties for the AASHTO M288 Class 2 separation/stabilization geotextile, which was Amoco style 2006. Only a couple square yards of geotextile were exhumed and available for testing from the limited test pit. There were a few small punctures and tears in the geotextile, possibly because of the exhumation effort, but this damage had no significant impact on the performance of the geotextile.

The results of the testing of the geotextile and the original geotextile properties are compared in Figure 2.

Figure 2: Woven geotextile property comparison: Test results of exhumed geotextile vs. original properties (testing done by TRI).
Property ASTM Test Method New Amoco Style 2006 in 1993 (MD x CMD) Exhumed Amoco Style 2006 in 2013 Units
Mass/Unit Area D-5261 5.4 7.7 (not washed) oz/yd2
Grab Strength D-4632 300×300 288×402 lbs
Elongation D-4632 15×15 24×22 %
Puncture D-4833 120 185 lbs
Trapezodial Tear D-4533 120×120 91×103 lbs
UV Resistance D-4355 70 92 % Ret @ 500 hrs
Apparent Opening Size D-4751 40 60 US Sieve
Permittivity D-4491 0.02 0.05 s-1
Flow Rate D-4491 2.0 3.7 GPM/ft2

The results of this testing confirmed the benefit of using a separation/stabilization geotextile beneath pavements. The geotextile maintained its integrity, with properties after 8 million ESALs and 20 years being similar to the original geotextile properties. The aggregate base maintained its original gradation without losing its effectiveness because of the contamination/intrusion of subgrade soil fines. Photo 3 shows the distinct interface maintained between the aggregate base and the clayey soil below.

Photo 3: Deeper spot excavation reveals the clayey subgrade.
Photo 3: Deeper spot excavation reveals the clayey subgrade.


A study done in the 1980s by Jorenby and Hicks, proved that as little as 15% fines (<200 sieve) contamination of the unbound aggregate base can lower the effective strength of the unbound aggregate base by as much as 50%. Beyond affecting the strength, the aggregate permeability is dramatically compromised with the intrusion of subgrade soil fines. According to AASHTO pavement design, lower aggregate permeability must be compensated for by assigning a lower effective strength to the more poorly draining aggregate layer(s).

One reason separation/stabilization geotextiles are not used under every road with unbound aggregate bases is that most pavement designers do not understand the magnitude of the base contamination problem. Engineers design and build the pavements thinking they will last the design life of 20 to 30 years.

However, when the pavement needs significant added structure or full-depth reconstruction before reaching the end of its design life, there is rarely a forensic evaluation performed. Such an evaluation would—in most cases—prove that the pavement was failing because of loss of effective strength and permeability of the aggregate base because of subgrade soil contamination.

As a comparison to this FHWA exhumation section, no geotextile layer was used in the pavement section shown in Photo 4. As the base became progressively contaminated by subgrade soil upward intrusion, additional pavement structural overlays were required until the result was a full depth asphalt concrete pavement—not the practical way an engineer would design or build a pavement.

Photo 4: In this photo (not from the FHWA site in this article), the subgrade soil has completely contaminated the once-clean aggregate base layer in the pavement with no geotextile.
Photo 4: In this photo (not from the FHWA site in this article), the subgrade soil has completely contaminated the once-clean aggregate base layer in the pavement with no geotextile.


A significant finding with this exhumation is that the geotextile did the job of surviving both the construction stresses and the traffic loading by the ALF.

The 20 years of accelerated traffic loading here may have represented many more years of in-service performance by this geotextile under a typical moderate duty pavement. Also, the durable polypropylene geotextile showed little to no change in properties after 20 years.

The reason this material is called a separation/stabilization geotextile is that along with providing the separation function, this geotextile has been widely proven to add structural strength to the pavement structure by increasing the effective strength of the subgrade soil by up to 80% and by stabilizing/strengthening the aggregate through lateral constraint as the bottom of the aggregate seats into the geotextile. The survivability recommendations of ASHTO M288 National Guideline Geotextile Specification allows DOTs the advantage of selecting either woven or nonwoven geotextiles for separation/stabilization applications.

A geogrid may also stabilize the subgrade soil and the aggregate, but a nonwoven geotextile should be used beneath the geogrid to ensure positive separation.

Perhaps the most important point is that the typical cost of an installed AASHTO M288 separation/stabilization geotextile is less than the cost of 1in. of aggregate base, and any pavement with an unbound aggregate base is going to lose significantly more than 1in. of effective aggregate base thickness because of contamination and/or aggregate loss. Even if base contamination and subsequent pavement failure is limited to local areas of unforeseen weak subgrade or areas of water accumulation, the minimal cost of the geotextile is far less than premature reconstruction required in these areas.

Research has shown that more than 4in. of effective base is typically lost because of contamination and this is exacerbated by the presence of water in the road section (Jorenby and Hicks).

DOTs have tended toward using a tight, poor-draining aggregate base to help limit the upward intrusion of subgrade soil fines. However, that tight aggregate will not drain quickly and will still enable a slurry of subgrade fines to move up into the aggregate base under the pumping action of traffic loading. More progressive DOTs use a separation/stabilization geotextile that will allow the use of a stronger, more open, free-draining aggregate without the fear of contamination.

There are many good points to the new Mechanistic-Empirical Pavement Design Guides (MEPDG), such as the improved definition of the strength of each pavement layer.

However, there are inherent weaknesses to this new design method, such as the lack of a way to address the progressive loss of effective aggregate base thickness over the life of the pavement. The MEPDG can assign water content to layers but has no practical way to compensate for the negative effects of variations in the hydraulic condition of the unbound aggregate layer(s), because of progressive contamination and its inability to drain rapidly.

These factors were addressed in the AASHTO Pavement Design and should be treated with more consideration in future versions of the MEPDG. Lacking a way to compensate for progressive loss of effective aggregate strength and permeability in the MEPDG, it is strongly recommended that all pavements incorporating an unbound aggregate base place an inexpensive separation/stabilization geotextile between the base and the subgrade soil.

If the design indefinitely preserves the full unbound pavement “foundation,” there should never be a need to perform a full-depth reconstruction of the road. Instead, only a much less expensive, safer, and less traffic disruptive surficial treatment such as an overlay will be needed—and much less frequently!

Closing comment

There is a great need for more base contamination verification, and all DOTs are encouraged to measure the thickness and extent of contamination in any cores or in small or large area reconstruction or digouts of pavements.

This is the subject of a Research Needs Statement by the Transportation Research Board (TRB) and researchers are encouraged to investigate this under-appreciated pavement malady. DOTs, please report any data on measured base aggregate contamination to the Geosynthetic Materials Association (GMA): Lucie Passus,, 651 225 6956.See another exhumed geotextile article—“Geotextiles in unpaved roads: A 35-year case history”—in the June 2008 issue of Geosynthetics, pp. 26–33.

Download a copy of Geotextile Specification for Highway Applications AASHTO Designation: M 288-06 (PDF).

Mark Marienfeld, P.E., is an independent consultant with more than 30 years of experience in transportation infrastructure and geosynthetics.

Fred Chuck, P.E., is director of Mirafi corporate training and industry affairs with TenCate Americas.

Terms used in this article

  • FHWA Federal Highway Administration
  • ALF Accelerated Loading Facility
  • TFHRC Turner-Fairbank Highway Research Center
  • ESALs Equivalent Single-Axle Loads
  • AASHTO American Association of State Highway and Transportation Officials
  • AASHTO M288 National guideline specification for geotextiles in transportation applications
  • DOT Department of Transportation (U.S. states—e.g., Virginia DOT
  • TRB Transportation Research Board
  • GMA Geosynthetic Materials Association
  • MEPDG Mechanistic-Empirical Pavement Design Guide

Photos by Fred Chuck and FHWA

Share this Story

Leave a Reply

Your email address will not be published. Required fields are marked *

Comments are moderated and will show up after being approved.