- 15 - ROADMIX'S EXPERIENCE IN BITUMEN-RUBBER M J WATSON It and L K DAVIDSON '''' " Pr.Eng. C.Eng. B.Sc. (Hans) FICE Managing Director : ROADMIX SA (PTY) LTD **B.SC. MSAChemI Sci.Nat. MIMT Research & Development Chemist : ROADMIX (PTY) LTD 1. INTRODUCTION An important aspect of the rehabilitation of the N4 freeway between Pretoria and Witbank, undertaken by Fowler Construction in 1979, was the inj ection of some 600 000 linear metres of cracks with a bitumen- rubber compound. The cracks were of the classical pattern generated from a cement stabilised crusher. run base reflecting through a thin asphalt overlay. Rubber buffings, purchased from a retreading factory, were added to heated bitumen at the rate of approximately 20 per cent of the total mass. Special equipment was developed to apply the material to the cracks and good penetration was achieved. The properties of the bi tlliden-rubber compound were considerably better than those of plain bitumen and led us to the conclusion that it would be worthwhile investigating and developing further bitumen-rubber products for use in spray and hot rolled asphalt applications. We knew that a certain amount of work in the field of bitumen-rubber had been carried out in both America and Australia, and decided to send representatives to these countries to learn about their products and technology. Initially contact was made with Mr C McDonald, a Consulting Engineer in Arizona who, at the time when he was Roads Engineer for the city of Phoenix, had developed bitumen-rubber products in the mid 1960's. Roa&nix's experience in bitumen-rubber
- 16 - Contact was also made with Mr G Morris who at the time whe~ he was employed by the Arizona Deflartment of Transportation, was instrumental with Mr McDonald for the introduction of the use of bitumen-,-ubber products in rehabilitating roads in the state of Arizona. Meetings Were also held with Engineers of the Arizona and Texas Departments of Transportation, and with Research Engineers of the Universities of Texas and Arizona. At this time, 1981, the Americans were using bitumen rubbers in crack sealing and spray applications, but not in hot rolled asphalt and our representatives were able to see applications of these On road surfaces which had been re-sealed up to 15 years previously. With the spray applications, the following problems were noted: (a) (b) (c) Bleeding or flushing. Poor transverse distribution of binder (tramlining). Loss of aggregate under heavy traffic conditions. This latter problem was such that the maximum aggregate particle size was restricted at that time to 9,5 rom to prevent damage to windscreens. Rates of application 6f the binder varied from a minimum of 2 litres per square metre to a maximum of approximately 3,5 Ii tres per square metre. The percentage of rubber added to the bitumen was in excess of 20 per cent of the total binder by mass. It was observed that where application rates of less than two litres per square metre were used cracks reflected through the membrane. With the application - of the crack sealing compound the main problem observed was the variation in the degree of digestion of
- 17 - the rubber into the bitumen. This was caused principally by their use of pre-blended and pre-packed materials. Our conclusions from this visit were that the concept of adding crumbed rubber to bitumen could produce materials of good quality for use in road maintenance, but that the American technology as existed at that time would have to be modified to overcome the abovementioned problems, and adapt the materials to be sui table for South African conditions. 2. RESEARCH AND DEVELOPMENT Following this visit, we decided to institute a research programme to develop sui table products. This programme was sufficiently advanced by March 1983 to enable us to commence field operations. In the early years of our research, we discovered that the commercially available crumb rubber was variable both in its constitution and shape. This we found resulted in a variation in the reaction of the crumb rubber with the bitumen which in turn resulted in varying physical properties of the compound. To overcome the problem, we decided that it was necessary to purchase the rubber from selected sources in order to obtain a constant quality material. In South Africa there are, broadly speaking, three types of tyre produced; heavy duty tyres; public service vehicle tyres and light duty vehicle tyres.. These tyres and their various parts, are made from different rubber formulations with varying natural and synthetic rubber contents. It was this variance which we considered created the problem of varying properties of the compounds. We therefore selected our sources of rubber on the basis of the type of tyre which each supplier processed. Roadmix1s experience in bitumen-rubber
r, 18 Our researches further led us to the conclusion that the shape of the rubber particles had a significant effect on the reaction of the rubber with the bitumen, which also affected the physical properties of the compounds. We therefore developed a milling process through which the selected rubber was comminuted to what we considered to be the most desirable amorphous shape. The rubber grading which also has significant effects on the compound properties was likewise controlled. By introducing these two measures, the bitumen-rubber compounds which we produced in the laboratory had consistent physical properties. It is important to realise that in the original American concept of bitumen-rubber blends the rubber was not dissolved in the bitumen, but instead reacted with the oils in the bitumen so that an interlocking matrix of rubber was formed. We considered this principle to be vital to the ultimate quality of the compounds and persevered with this approach instead of adding ingredients to dissolve the rubber, although the latter principle would have facilitated the application of the compounds. Our research also illustrated that vulcanised rubber compounds were better than normal rubbers as they tended not to dissolve. At this stage of our research we concluded that it was necessary for the bitumen rubber compounds to have different properties for the various applications. We considered the following three applications: crack seal, spray and hot rolled asphalt, and decided that the properties for each should be: (a) Spray Application i I I I (i) Good adhesion to aggregate. I
- 19 - (ii) Low temperature susceptibility both at high and low temperatures~ (iii) Balanced elasticity/plasticity in order to obtain spontaneous self healing and receptivity to aggregate under normal chipping conditions. (iv) High tenacity (toughness). (b) Crack Sealing Application (i) High elasticity. (ii) Rejection of aggregate. (iii) Good penetration. (iv) Good adhesion. (v) Low temperature susceptibility both at high and low temperatures. (c) Hot Rolled Asphalt Application In this instance we investigated both preblending bitumen and rubber, and also using rubber in a dry crumb form when mixing the asphalt, the latter being the system adopted. In both cases, it was felt that the binder would require the same properties, viz: (i) (ii) (iii) (iv) High elasticity. Low temperature susceptibility both at high and low temperatures. Good adhesion. Good cohesion. However, in the case of the hot rolled asphalt, the properties of the final mix in addition to the binder were also investigated.
- 20 - r The above properties were achieved by varying the type, grading and content of the rubber additive in each of the compounds. Having produced materials in the laboratory which had the desired physical properties, we then had to overcome the problems of manufacture and application. We found that commercially available equipment would not" adequately blend or apply the spray product, and consequently, were obliged to design and construct our own blending and spraying equipment. An important factor in producing a compound with constant physical properties was noted to be the period of digestion of the rubber into the bitumen. combination of of three hours time and o at 175 C. This varies with temperature, but the optimum temperature was found to be in the order We also observed that when the compound was held at the digestion or application temperature for excessive periods, then the physical properties of the compounds were impaired. This created further difficulties with the spraying operation. This led us to the conclusion that our compounds cannot be pre-manufactured and stored, but must be manufactured at the site of application. This applies of course to the spray and crack sealing products only. 3. EQUIPMENT AND FIELD APPLICATION Our manufacturing and applying equipment was designed around this premise and our total operation is an on-site manufacture and apply process. With regard to the hot rolled asphalt, the rubber additive again has a different grading and constitution to that for the other products. Here we are able to add the rubber into both batch plants and continuous mixing plants. The asphalt has distinctly
,- 21 - enhanced elastic properties compared to conventional premix especially when the binder content is increased to 11 per cent. Again, with asphalt the digestion process is of the order of three hours and we find that this can be successfully achieved by storing the asphalt in insulated trucks for this period before paving. In the case of asphalt the time period at which it can be held at high temperatures is not so critical. The mix temperatures can in fact be maintained at this level for a period of up to 6 or 7 hours without detriment. Paving and rolling techniques are different to those for conventional asphalt, particularly as the mixing and digestion temperature is considerably higher than normal rolling temperatures. The asphalt once paved must be allowed to cool to a certain extent before compaction. Because of the good compactability of the material, the Use of steel wheeled rollers only is necessary. Depending upon the temperature of rolling, compaction to zero air voids can be achieved, particularly with high binder contents. The risk of subsequent flushing under heavy traffic conditions is negligible. It is also interesting to note that this material can be made to remain lalive 1 after cooling down to normal air temperature, and can therefore be used as a cold asphalt for patching. 'fhe self healing property of the binder is important in this regard. In 1983 we again sent delegates to America to investigate further developments which may have happened in that country. The problems of bleeding and aggregate stripping were still found to exist as well as poor distribution of the binder in certain instances.
- 22 - r il 'I t ~ At this stage there was no adequate national specification being applied in America. This resulted in problems associated with variation of quality and application of the bitumen-rubber compounds giving variable end results. In a subsequent visit to Australia, we discovered that similar problems had been experienced, and to overcome these, lower rubber contents in the binder were used, e.g. approximately 13 per cent compared to our 20 per cent. 4. SPECIFICATION AS with all new technologies, there is a certain amount of risk involved in their usage. It is difficult to draw up compatible specifications with limited knowledge of bitumen rubber technology and performance. It is our opinion that a specification relating to the physical properties of the, materials is more pertinent than a recipe specification. We have therefore scheduled the following requirements for the spray grade materials which relate to tests to be carried out on samples of the applied material, simply and speedily. The bitumen-rubber shall comply with the requirements listed. property Requirement Elasticity Plasticity 25 0 C Temperature Susceptibili ty Brittleness OOC Temperature Susceptibility Flow at 60 0 C Elastic Recovery After Flow No Fracture Flow 70% Min. 40% Max. 25% Min. 20 rom Min. 80 rom Max. Adhesion Self Healing Aggregate Bonds Firmly Seals Together
- 23-5. CONCLUSION It is our firm belief that bitumen-rubber compounds can be used very effectively for the rehabilitation of roads and for surfacing new works. They 'have been developed as an alternative to overlay and reconstruction, and used properly in this context, can be extremely cost effective.
- 24 - APPENDIX DESIGN PROCEDURE For spray applications pre coating of the aggregate is essential for the successful application of the seal. Also as normal seal design criteria do not apply, we developed our own modified design techniques. These are necessary to obviate flushing which could otherwise occur with the much higher rates of application of binder applied to prevent crack reflection. The size of aggregate must be selected with consideration to climate, traffic density and binder application rate. Aggregate Type As for normal seals. Aggregate Precoating As use. for normal seals. The precoated aggregate should be dry before Aggregate Sizes For Stress Absorbing Membrane (SAM) initial seals 9,5 rnrn, 13,2 rnrn or 19 rnrn aggregate may be used depending on traffic requirements and the existing surface. For SAM reseals, as in the case for normal seals it is desirable that the existing surface should be prepared before sealing. This would entail removing surface irregularities by a slurry seal or thin overlay or enrichment of a lean surface by a dilute emulsion spray. Then again depending on requirement, 9,5 mm, 13,2 rom or 19 rom aggregate may be used.
- 25 - If such remedial surfacing is not done, the choice of aggregate size may be limited by the surface texture, and allowance for this may be required in calculation of spray rates for bitumen-rubber application. For Stress Absorbing Membrane Interlayers (SAMI) 9,5 mm or 13,2 mm aggregate may be used; the larger aggregate enabling use 'of a higher rate of application,of bitumen-rubber binder which is desired for eliminating reflection cracks. In general where cracking is prevalent it is necessary to use sufficient bitumen-rubber binder to prevent transmission of cracks into the new surfacing. Aggregate smaller than 9,5 mm should not be used in bitumen-rubber work because it would not permit the desired high rate of application of binder and this could permit cracks to reappear over long-term. For the same reason, an unduly flaky aggregate of lower than normal ALD should not be considered. The choice between a SAM or SAMI surfacing is dictated by the type and severity of existing cracks. A SAM seal will normally prevent reflection of fatigue cracl<s whereas a SAMI will prevent reflection of all cracks. Aggregate Rate of Application - SAM To obtain an even single layer of aggregate may require up to 10 per cent more aggregate than for normal sealing, because with the use of bitumen-rubber the rolling does not push the aggregate to the flattest dimension as readily. Allowance for Aggregate Loss on site ALD mm Below 5,0 5,0-6,0 6,0-7,5 7,5-9,5 Over 9,5 Allowance for Loss 10% 8% 6% 4% 2%
r - 26 - Bitumen 80/100 Bitumen shall be used. Rubber Specially comminuted tyre crumb complying to Specification shall be used. It shall all pass a 1,18 mm screen. Binder Rate of Application In bitumen-rubber work the residual binder is regarded as the bitumen + rubber. Cutter is not part of the residual binder. Rates of o application are expressed as rates of cold (15 C) nett binder (Le. bitumen + rubber) per square metre of surface. Percentage of Theoretical Voids to be Filled by Binder for Stress Absorbing Membrane (SAM) Actual Traffic 50 50-150 150-250 250-500 500-1000 1000-2000 2000 + Voids % 90 85 80 75 70 65 60 For Stress Absorbing Membrane Inter1ayers (SAMr) the percentage voids to be filled shall be 95 per cent regardless of traffic volume. Binder Allowance for Existing Surface Condition To be Added to the Calculated Rate of Hot Binder If no remedial work is done prior to reseal.
- 27 - Fatty - excess binder over aggregate Smooth - no absorption Lean - open textured slight absorption Hungry - very open textured high absorption Extremely Hungry - extremely open texture extremely absorptive Litres/Square Metre -0,10 to -0,30 Nil +0,10 +0,25 +0,40-0,50 Adjusted Rate of Application The calculated rate of binder application shall be increased by multiplying the basic rate by a Rubber-Bitumen Factor. For Stress Absorbing Membranes (SAM) the Rubber-Bitumen Factor is the average of two figures - the Rubber Content Factor and The Traffic Factor. Rubber-Content Factor This takes account of the expanded condition of the bitumen-nlbber blend when sprayed. % Rubber in Blend 20 25 Rubber Content Factor 2,0 2,4 Traffic Factor This takes into account the fact that the aggregate does not necessarily lie on the flattest side in bitumen-rubber seals. The degree to which it does so depends on the traffic intensity during the early life of the seal particularly the first few hours. Once the bitumen-rubber binder cools, re-orientation does not occur as readily as it does in normal seals. The Traffic Factor decreases with increased traffic volume as shown.
r - 28 - Traffic Volume (Corrected) (For Divided Carriageways consid~r one direction only) Traffic Factor 1 000 and below 5 000 10 000 15 000 20 000 and over 2,1 1,9 1,7 1,6 1,5 For Stress Absorbing Membrane lnterlayers (SAMl) the Rubber-Bitumen Factor is the Rubber Content Factor only and no allowance is made for traffic. All designs for bitumen-rubber seals and reseals must be approved by the Rehabilitation Laboratory before any work is put in hand.