104 ■ Transportation Research Record 1819 Paper No. LVR8-1129 CSIR Transportek, P.O. Box 395, Pretoria, 0001, South Africa. ments in South Africa are designed using Technical Recommendations for Highways (5). This is essentially a catalog design system based on experience and mechanistic empirical analyses (6–8). The catalog extends to very low traffic categories (design lives of less than 3,000 standard axles per lane). Materials are classified primarily by their CBR strength with grading, Atterberg limits, and field compaction requirements also specified. Tables 1 and 2 are a summary of the prop- erties for grading. Summaries are also given for Atterberg limits (Table 3), bearing strength and swell (Table 4), and field compaction (Table 5). In TRH 4 (5), bases for low-volume roads are either G4 or G5, whereas subbases are of at least G7 quality. Support layers need to be at least 150 mm of G9 over material with a minimum CBR of 3%. Other catalogs with base material quality as low as G6 have also been developed for local use (9, 10). The importance of the CBR in the design of low-volume roads is thus clearly illustrated. CBR TEST Following the development of the moisture–density–compaction effort relationships by Proctor in the 1930s for the construction of earth dams (11), a standard test method for compaction was devel- oped. The Proctor test is currently designated as ASTM D698 (12) and adopted by AASHO as AASHO T99 (13). The California Department of Transportation extended the test by determining the load required for a standard plunger to penetrate the soil compacted to AASHO T99 effort at a standard rate and comparing this with the load required to penetrate a standard material (14). This measurement was defined as the CBR and is the basis of various pavement design methods. Limi- tations of the test have been identified, ranging from the fact that the results cannot be directly related to fundamental properties governing the strength of soils (15) to the variations among test methods. The compaction test was originally developed for subgrade materials [using only that fraction passing the No. 4 sieve (4.75 mm)]. The ASTM and AASHTO compaction tests, however, each provide a variety of methods depending on the particle size distribution, the size of the mold compacted, and the mass of the rammer used. Although the current AASHTO test method (T193) for determining the CBR provides for compaction using both the 2.49- and 4.54-kg rammers, this is a recent addition to the method, which previously only provided for the use of the light rammer. The ASTM method (ASTMD1883) for the bearing ratio has provided for the use of both the 2.49- and 4.54-kg rammers for some time. The CBR test method is defined for use in South Africa in the Technical Methods for Highways (TMH) suite of manuals No. 1 (16 ). However, minor modifications to the test method have occurred and the test now differs from the ASTM (12), AASHTO The relationships among strength, moisture, and density in pavement subgrades and layerworks are well known, but they have particular sig- nificance in low-volume roads. In these roads, the specified density is fre- quently not achieved (quality assurance testing tends to be reduced), and moisture fluctuations are more severe with greater consequences. Traditional studies of the material strength for these roads in southern Africa are based almost entirely on the California bearing ratio (CBR), a test with inherent problems. The test is routinely carried out to iden- tify whether the materials under consideration have the required soaked strengths (typically CBRs of 80% or 45% for bases of different stan- dards) at the design compaction density. Studies of the CBR at different moisture contents and densities should be carried out to identify the implications of variations in these properties on the behavior of pave- ment materials. A simple technique to be carried out during conven- tional laboratory testing was developed. Aspects pertaining to this type of study were evaluated, and the findings were related to low-volume road behavior. Research carried out on the relationships between materials and per- formance of low-volume roads has indicated that the environment probably has a larger effect than do either the material properties or the traffic (1–3). The environmental feature most dominant is the presence of moisture and water in the pavement. Traditionally, the California bearing ratio (CBR) has been used for characterizing pavement materials in South Africa (4). The selection of natural materials for selected layers, subbase, and base relies on the materi- als’ having an appropriate CBR value, and the pavement design thickness depends on the CBR of the in situ subgrade or fill. Because the CBR is such an important component of pavement design in South Africa (and in many other countries), a clear understanding of the impact of moisture and density variations on the result should be obtained during the conventional laboratory testing. This paper discusses some basic principles of the CBR test relevant to low-volume roads in South Africa and southern Africa and assesses the use of this test in conjunction with the standard compaction characteristics tests during the design of low-volume roads. LOW-VOLUME ROAD DESIGN IN SOUTHERN AFRICA The design of low-volume pavement structures in South Africa, and in southern Africa in general, is based mainly on the CBR strengths of the in situ and locally available materials. Conventional pave- Strength and Behavior of Materials for Low-Volume Roads as Affected by Moisture and Density P. Paige-Green