Geotextiles in unpaved roads: A 35-year case history

By William M. Hawkins

Abstract

This article presents current information on geotextiles installed experimentally in an unpaved road 35 years ago.

In 1972, geotextiles were largely untested, and the site was set up as an accelerated field test to determine the comparative performances of several fabrics for use as a geotextile. But because the site was still accessible 35 years later, it offered an opportunity to review the ultimate potential lifetime of geotextiles in unpaved roads.

When the fabrics were exhumed in 2007, we learned that they had survived and continued in service despite 2 factors that had worked against them:

1. The lack of adequate cover (in some cases, less than 6in. of stone) had adversely affected the fabrics.
2. And, by current standards, an inadequate polymer stabilization package used when these fabrics were produced.

The unusual opportunity to look at geotextiles this old in situ and the fact that some of the fabrics survived and continued to perform under the adverse circumstances offers important information. With current stabilizers, design, and installation procedures, today’s geotextiles perform even better and longer.

Background

The purpose of geotextile separation is to prevent 2 simultaneous mechanisms that tend to occur in a roadway cross section over time (Figure 4). The first is that the stone base tends to penetrate into the subgrade soil, thereby compromising its load-bearing capacity. The second is that fine-grained subgrade soil tends to intrude into the voids of the stone base, thereby compromising the stone base’s drainage capacity. In both cases, when the base intermixes with fine-grained particles from the subgrade soil, the stone base (or the lower portion of it) is no longer effective for load bearing or drainage. The situation is heightened in areas of freeze/thaw and wet/dry cycling. Environmental changes such as frost, infiltration, drainage, and increased loading all adversely challenge the paved or unpaved road base.

It should be noted that many unpaved roads eventually become paved (usually with asphalt) and, if the stone base is protected from the beginning against subgrade soil contamination, the paved road design can be done with confidence.

In 1972, nonwoven fabrics were being used in Europe in road support applications on soft soils and at construction sites. The results appeared to be positive. Recognizing this, DuPont, an established nonwoven fabric producer, developed a program to produce a geotextile for use in similar applications. As part of that program, several existing fabrics were installed in unpaved road test sections for performance evaluations.

The purpose of the performance evaluations was to determine which fabrics would best perform the required functions of reinforcement, stabilization, and/or separation. Several materials were installed and evaluated in different geographic locations; the site near Smyrna, Del., is still functioning and is one of the oldest known existing accessible geotextile separation applications. In June 2007, this particular site was visited and samples were exhumed. Reviewed in this article is information about the initial installation and its current conditions, the field performance, and current status for these test sections. Through evaluation of physical, mechanical and chemical properties, the separation performance, survivability and durability properties of the geotextiles were evaluated and compared.

The Smyrna site and the 1972 design approach

The source of most of the historical information in this section is the original test evaluation report by Crane and Hutchins (1974) and discussions with its co-author, Dick Hutchins (2007).

The Smyrna, Del., test section that was created used a farm road built over a sandy clay soil with a load-bearing capacity of CBR (California Bearing Ratio) ≅ 1.0 when wet, and a CBR ≅ 6.0 when dry.

Unlike the other test sections constructed by DuPont at the same time, the Smyrna site was completely controlled by the designers. During the initial testing, the Smyrna road was not repaired.

The test focused on providing useful information on the performance of potential geotextiles used beneath stone base courses, in order to sell them into the road construction industry. A number of different commercially available fabrics were available and used at the site. The Smyrna test used a 1000ft (310m) section of road that was purposely underdesigned. The idea was to encourage or accelerate failure so the test geotextiles could be evaluated quickly. Using 40-kN wheel loads above the low-load-bearing soil normally calls for a 15-in. (38-cm) gravel base. However, only 6in. (15cm) of gravel base (40% of design) was actually used, according to Hutchins in 2007.

The tests were run in 2 stages:

1. a dry run, in which the loaded vehicle transversed while the road was dry and then samples
2. a wet test, in which loaded vehicles were run after a heavy rain and then samples were excavated and removed.

The site is in the area of a fill. There is slope of about 0.5% from north to south. As can be seen from Figure 1 and Figure 6, the road acts like a dike across the farmer’s field. Standing water in areas of this field indicates low points and poor drainage.

The site’s climatic conditions can be generally characterized by noting that it is in FHWA Region 1 and FHWA climatic zone I-A. This means that the site is located in an area with high potential for moisture.

Normal road construction techniques were used for installation of the geotextiles. Heavy construction equipment was used to make grade. Laborers spread the geotextiles by unrolling the materials on the subgrade in advance of a dozer spreading base material. In addition, a control section was installed where no fabric was placed under the 6-in. (15-cm) gravel base.

A general description of the road would be a “private road through a farmer’s field.” The number of passes on this road is low. However, during planting and harvesting season, the loads are heavy and frequent. From historical records, the CBR of the site before construction was 2 and the field CBR was approximately 8 (dry). The subgrade soil was a silty sand (SM) with 12% passing the #200 sieve, and the modified base was a well-graded gravel (GW) with thickness of 4-8in. Specifics about the geotextiles and site soils are in Tables 1 and 2.

The dry run (142 passes of loaded vehicles) produced no noticeable difference between the sections where fabric was used and the control section.

After a heavy rain, the wet test was carried out. In the control section (without fabric), complete failure occurred after 20 passes. At the other end of the longevity spectrum, in the GT-A (see Table 1) section, after 120 passes only soft spots were observed. From these initial tests, the GT-A fabric was determined to be the best candidate for these types of geotextile separation and drainage applications. It maintained sheet integrity and a conclusion was that this product provided the best results of all materials used at the Smyrna road project. (After the wet test, all candidates were excavated and evaluated.) It was concluded that, for heavy-duty construction stresses such as this, fabrics should be at least equivalent to the GT-A, at 3.5oz/yd² (136g/m²), and covered with at least 6in. of base material, or significant loss of properties will occur.

Exhumed after 35 years

In June 2007, 35 years after installation, George Koerner of GSI and the author returned to the Smyrna site to determine the status of the road and the condition of the geotextiles. After the various test plots were located, photographs were taken to characterize the general area conditions as well as the specific plots.

Exhumation of the samples followed. Pick and shovel were required to break up the hard crust of the unpaved road surface, which was well compacted because the exhumation was done in the most critical area–the tire tracks. After probing for the location of the geotextile elevation, which was 4-8in. (10-20cm) below the ground surface, careful removal of the fill by hand proceeded over an area of approximately 1m². The fabric was brushed, more photos were taken, and then the samples were removed and stored in plastic bags. Figures 2 and 3 illustrate how the geotextile samples and soil were collected, in addition to monitoring of field soil conditions.

Technical evaluation

General observations

Photographs confirm that even though the geotextile was installed 35 years ago and the project was under-designed, some of the geotextiles endured to effectively perform the primary function as a permeable separator. In fact, it was obvious where the geotextile was used because there was no significant rutting at those locations. It was equally obvious where no geotextile was used, as lateral spreading of the embankment was noted and rutting was clearly evident.

As observed from Table 3, there are 2 test sections—1 (GT-A) and 5 (GT-D)— that had minimum cover (6in. or more) and were still performing well. The others were not exhumed or were in “bad condition.” Test Section 7 (GT-E) had inadequate cover (only 3in.) and showed significant physical damage.

Evaluation

As shown in Table 3, 8 different geotextiles (plus a control section with no geotextile) were used at the site. Table 1 shows the results of index and performance testing of 6 of the fabrics used at this site, prior to installation. There were only 2 soils (subgrade and base) used for this project, and their characteristics are given in Table 2.

The geotextile samples were brought to the lab to compare their current physical characteristics with those of 1972. Unfortunately, only GT-A at 3.5oz/yd² (136g/m²) and GT-D (4.0oz/yd²) could be tested because the other geotextiles were significantly damaged. Grab tensile results show, on average, a 37% strength retention and a 52% elongation retention compared with historical production data for the GT-A and GT-D products. Trapezoid tear strength retention was approximately 50% and puncture strength 93% on average. Note that current testing was very limited. A summary of results for the 2 geotextiles can be seen in Table 4.

Analysis of the magnified polypropylene filaments showed some degradation. For photomicrograph analysis of the geotextile polymer, it was necessary to remove as much soil and other interference as possible. Repeated attempts to clean the soil from the geotextiles were ineffective, which is why mass per unit area and thickness results are not reported. As can be seen from the photos, polymeric deterioration was readily observed in all samples examined. This deterioration was not only observed in the outer layer of the surface, but some was also apparent in the core of the fibers.

It should be pointed out that the stabilizer package used in 1972 was quite different from, and much less effective than, today’s stabilizer. Currently, for example, the GT-A and GT-D geotextiles use the latest in hindered amine light stabilizer packages (HALS). HALS packages act as free radical scavengers no matter what type of free radical develops.

Nevertheless, one of the goals of the 2007 study was to determine whether the same amounts of antioxidants and ultraviolet stabilizers are present today as when the material was produced. In pursuing this goal, it became clear that a review of the heat flow (melting) curve and a review of the thermo oxidative time and temperature as compared to the 1972 stabilizer package would be of interest.

Differential scanning calorimetry (DSC) was performed on the aged polypropylene samples and compared to that of un-aged samples. The oxidative induction temperature of GT-A went from 228° to 212°C in 35 years (Table 5). However, the oxidative induction times of GT-A and GT-B are near 1 minute. This indicates that there is a small amount of the original package currently left in these materials.

Summary and conclusions

This report is unusual in that it documents the use of a geotextile type of fabric and its performance over a 35-year period.

The initial purpose of the test, 35 years ago, was to determine if and which fabrics would perform effectively as a geotextile in a separation application under an unpaved road. (Testing longterm durability was not part of the initial purpose.) The 1972 tests showed that GT-A could perform that function very effectively, even though it was not specifically designed for that use and was installed with inadequate “safety factors”—too little base cover for the extreme loads it was subjected to in wet and saturated conditions. The loads used in the initial testing would normally require a minimum of 2.5x the base used.

As it turns out, the fabric GT-A has performed the separation function for 35 years and is still working. Analyses of the fabric after 35 years for survivability and durability indicate the stabilizers used then are not nearly as effective as those used today. The fabrics suffered significant mechanical damage as a result of overloading but were still performing. Indications of inadequate protection of the polymers by stabilizers are not surprising because that need has been noted in other situations, and that is why the stabilizers used today last much longer and are more effective.

However, site inspection and samples indicate that, if at least 6in. of gravel remains over the geotextile, thermally spunbonded nonwoven geotextiles are still performing the function as originally intended 35 years ago, even though the site was grossly under-designed. Unquestionably, good performance is predicated on adequate soil burial. All geotextiles suffer survivability problems with a gravel thickness less than 6in.

Bill Hawkins, now retired, was a longtime employee at Fiberweb (and, previously, DuPont) and remains one of the true pioneers in the geosynthetics manufacturing business.

References

Cedergren, H. R. (1989), “Seepage, Drainage and Flow Nets,” J. Wiley and Sons, New York, N.Y.

Crane, J. P. and R. D. Hutchins (1974), Typar Road Reinforcement. Report TR434930, Project 704-236, Notebooks T-3420 and T3320. Textile Fibers Department, E. I. DuPont De Nemours & Co. Inc.

Hutchins, R. (2007), personal communications.

Koerner, G. R. (1997), “Data Base Development for Determination
of Long Term Benefit/Cost of Geotextile Separators,” Geosynthetics-1997, NAGS Conference Proceedings, Long Beach, Calif.,
pp. 701-713.

Koerner, G. R. (2000), “Geotextile Separation Study,” Geotechnical Fabrics Report (GFR), Vol. 18, No. 5, Roseville, Minn., pp. 14-21.