STABlLlSATlON OF EARTH ROADS WITH WATER-BASED POLYMER EMULSIONS

~olwnerisation, determined whether the emulsion would form Laboratory results show that by optimising the polymer type and stabiliser system, emulsions can be produced which, when diluted with water and mixed into sands or soils with a high clay content, can produce thick aggregates with high load bearing and water holdout characteristics. Unconfined compressive strength (UCS) and water uptake results on cores before and after soaking in water are given for a wide range of soil types and levels of polymer. The minimum requirement of 0,75 MPa for a C4 pavement is exceeded with only 0,25% of emulsion on soil. Practical results with surface applications only and incorporation to depth on several roads, a parking lot and the entrance to a sugar mill are reported. Introduction In southern Africa, finance for maintenance and construction of roads is scarce. The CSIR (Jones, 1996) have indicated that for gravel roads in townships a surface only treatment costing less than E l m 2 and remaining completely effective for one, but preferably two, years is required. For roads carrying heavier loads, the traditional method of importing aggregate is becoming prohibitive because of the scarcity of suitable material and the high cost of transporting aggregates over long distances. Polymer (plastic) emulsions are used extensively to form thin layers of less than 3 mm, in paints and waterproofing screeds on walls and roofs. Their excellent resistance to embrittlement, UV light and acid rain, and their good adhesive properties, made them suitable for blending at low levels into bitumen, tar and cement to upgrade these materials. This paper describes how polymer emulsions were modified to give thick consolidated layers with high load bearing strengths and good resistance to wear by water and traffic. Also discussed are the best techniques of mixing, drying and compacting the soils to give optimum results in laboratory tests and the performances of several polymer treated roads after extended periods of trafficking. Procedures Experiment I In 1983 the Transvaal Provincial Administration (Zadzick, 1983) carried out a practical road trial in which a diluted polymer emulsion was applied at the low rate of 0,06 L/m2 to the surface of a recently completed gravel road. The origin of this emulsion was from earlier laboratory work which showed that the type and level of stabiliser that was used during ihiik load bear& slabs, or thin surface skins with underlying loose sand, when poured onto the surface of sand and allowed to dry (Bishop, 1978). Those polymer emulsions stabilised with the lowest levels of high surface tension stabilisers produced the strongest sand aggregates. For the emulsion used in Experiment 1 the polymerisation was carried out with the lowest level of polyvinyl alcohol stabiliser necessary to prevent the polyvinyl acetate emulsion from coagulating on storage. The minimum film forming temperature (mft) of the polymer was reduced to 12°C with an external plasticiser to ensure film formation in most climatic conditions in southern Africa. This emulsion at 58% solids content was diluted 50 times with water in a spray tanker and then applied evenly over the surface of a road hear Pretoria, at the ;ate of 0,06 L in 3 L/m2 of water. The gravel of the road contained clay, with a plastic index (PI) of approximately 10, and the application was made in May, at the start of the dry winter months. The condition of the road was compared regularly with two adjacent sections which had received no aggregating agent. The results after four months of trafficking are given in Table 1. Experiment 2 To improve the resistance to water of the PVAc homopolymer a number of colloid stabilised copolymer emulsions were formulated with the following monomer combinations: vinyl acetate-acrylic, styrene-acrylic, veova-acrylic and acrylicacrylic. These emulsions were compared to cement, hydrated lime and a bitumen emulsion as soil bonding agents in a clay containing soil (PI=16). The unconfined compressive strength (UCS) was the preferred Civil Engineering test, as it compares load bearing performances in both the dry state and after submersion in water for four hours. These results are also applicable for other load bearing applications such as earth bricks. For an aggregate to conform to the requirements for the base of a C4 pavement it must achieve a UCS of 0,75 MPa. For the true potential of the polymer emulsions to be realised in laboratory tests it was found that the following mixing and drying procedures had to be closely followed. After determining the optimum moisture content (OMC) of the particular soil, the polymer emulsion at 1% active on aggregate, was pre-blended with 33% more'water than was needed to achieve OMC. This blend was then mixed thoroughly with a fresh sample of soil. The damp mixture was then spread out in a layer h50 mm thick and left in a shaded area (h23"C and 55% relative humidity) for 48 hours. It was Proc S Afr Sug Technol Ass (1998) 72 Stabilisation cf earth roads with water-basedpolymer emulsions RT Bishop, BA McAlpin & D Jones then compacted into the moulds. The moulds were split and the free standing cores were dried at ambient temperatures for 24 hours, and then at 60°C in a forced draft oven for 72 hours. Finally, the dry soil cores were weighed before one was crushed dry and the other duplicate was submerged in water for four hours. After removing and dabbing off the excess water the core was re-weighed and crushed immediately. The extent to which water had penetrated into the core was visually assessed. The results from other experiments using cores bound with 4% cement, hydrated lime and a bitumen emulsion, dried in the ways recommended, are included for comparison. The results are given in Table 2. Experiment 3 Using the same test procedure as in Experiment 2, decreasing levels of the best styrene-acrylic copolymer (B) in Table 2 were mixed into a high clay containing soil (PI=31) and a sand (PI=2) to determine the levels at which it ceased to have any visual bonding effects. The criteria for assessment after the four hour soak were the ability of the cores to (i) be removed and have a UCS measurement conducted on them, (ii) prevent the ingress of water, (iii) remain dimensionally stable without swelling or collapsing and (iv) be sufficiently bound so as to prevent clay particles from being permanently suspended in the supernatant liquid after physically stirring the collapsed cores for 20 revolutions with a palette-knife. Cores bound with 4% cement and hydrated lime were included for comparison. The results are given in Table 3. Experiment 4 In 1989, a road was built in a red sand (PI=3) using the styrene-acrylic copolymer (A) at Sodwana Bay, KwaZuluNatal. The area was 100 X 10 m and the copolymer was incorporated to 150 mm. To ensure drying in a reasonable time in the very humid environment, dry cement at 1% (assuming that 1 m2 X 0,15 m of sand weighs 270 kg) was premixed into the 150 mm layer. The styrene-acrylic emulsion with a solids content of 50% was diluted 1: 1 with water and was applied by a water tanker evenly over the test area at a rate of 2,4 ~ / m ~ . A grader thoroughly mixed the 150 mm layer before the road was shaped and compacted with a pneumatic roller. The dry copolymer content on sand was 0,22%. After four months of trafficking the load bearing strengths were measured in the 0 to 150 mm and the 150 to 300 mm layers (Table 4). The road was also visually assessed on an annual basis (Brotherton, 1997). Table 1. Condition of road four months after treatment in Experiment 1. Table 2. Unconfined Compressive Strengths and water uptakes in Experiment 2 (means of duplicates) at 1% dry polymer on soil. Treatment