Department For International D F ID ~~~~Development TITLE: by: The Performance of Asphaltic Pre-mix Surfacings in Indonesia H Dardak, T Toole, S Mahmud and A T DachlIan Transport Research Laboratory Crowthorne Berkshire RG45 6AU United Kingdom PAl 280192 DARDAK, H, T TOOLE, S MAHMUD, and A TATANG DACHLAN, 1992. The Performance of Asphaltic Pre-mix Surfacings in Indonesia. Proceedings of the Seventh REAAA Conference, Singapore, 22-2 6 June 1992. THE PERFORMANCE OF ASPHALTIC PRE-MIX SURFACINGS IN INDONESIA Dr. Hermanto Dardaik Head of Road Engineering Division, Institute of Road Engineering Indonesia T. Toole Princ ipal Scientific Officer, Transport and Road Research Laboratory United Kingdom S. Mahmud Researcher, Institute of Road -Engineering Indonesia A. Tatang Dachlan Researcher, Institute of Road Engineering Indonesia ABSTRACT Studies have been undertaken over the past three years of the performance of asphaltic overlays in Indonesia. The overlay materials were designed to accomodate a higher than normal bitumen content to improve mix durability and flexibility. The mixes are similar to the UK gap-graded Rolled Asphalt rather than continuously graded Asphaltic Concrete. The studies have involved an evaluation of performance, for which detailed condition surveys were carried out on more than 300 km of roadway. Subsequently, twenty five sections were selected for long-term study. The sections were chosen to cover the range of standard design parameters, including design traffic, overlay thickness and structural strength, road geometry and mix type and properties. The paper describes the evaluation methodology used and the preliminary results of the performance studies, including the relation between mix type and mix properties and behaviour. Recommuendations are given to improve the current Indonesian specifications for asphaltic surfacings. 1. INTRODUCTION -A comprehensive* research programme to evaluate the performance of road strengthening overlays used in Indonesia has been underway for over three years. The objectives of the studies are to enable improvements in mix design and specifications, construction practice and structural design and to produce road deteriorat ion relationships relevant to the Indonesian environment. Early results from the studies and details of their .design and i mplemfentation were reported (TOOLE et al, 1990 and 199 1). The types of overlay being evaluated include mixes produced to the 1986 "~Specification for. High Duirability Asphalt' (DGH 1986), known as Hot Rolled Sheet (HRS), and Asphaltic Concrete (AC) mixes produced 495 according to Marshall- design principles (ASPHALT INSTITUTE, 1983). The basis for the adoption of HRS in Indonesia was first described by Corme (1983). He described an alternative composite overlay system comprising an asphalt treated base (ATB) and HRS wearing course. These materials are designed to accommodate a larger amount of bitumen than AC-type mixes, the aim being to increase flexibility and durability. The HRS is similar to the gap-graded rolled asphalt (RA) used in the UK (BSI, 1985) and Southern Africa (NITRR, 1978). Two classes of IIRS are specified, Class A is used for light and medium traffic and Class B is used for heavy traffic and steep grades. Since the mid-nineteen eighties many thousands of kilometres of road have been overlaid but no quantitative comparison between HRS and AC overlays has been made. Such a comparison is of paramount importance to the Directorate General of Highways (DGH) so that rational decisions can be taken on future road betterment and maintenance schemes. A cooperative research programme between the Indonesian Institute of Road Engineering (IRE) and the Transport and Road Research Laboratory (TRRL) of the United Kingdom has been undertaken to achieve this and is described in this paper. 2. EVALUATION METHODOLOGY AND PERFORMANCE ASSESSMENT 2.1 Evaluation methodology The scale of the road betterment programme meant that a wide variety of roads had been overlaid under a wide range of conditions. It was anticipated that a study of road betterment projects completed by 1988 would enable the range of the key variables required by the study to be covered. This. necessitated .window-monitoring', whereby observations are taken over a limited period of time on sections of varying age. Any factors which could not be examined in this way could be investigated in studies of new overlays or specially constructed trial sections. The evaluation has so far involved 7 main stages:- (i) A desk study of design and construction p rocedures. (ii) Condition surveys of a sample of road-links to identify general performance trends and mechanisms of failure. (iii) Selection of a sample of short stretches of road (or monitoring sections) for detailed investigation. (iv) Determination of the effects of traffic, mix properties and road geometry on mix performance. (v) Evaluation of problems from the design stage through to construction. (vi) Identification of inadequate data sets and selection of relevant additional test sections on other completed or new projects. (vii) Development of improved specifications and design procedures. 2.2 Performance assessment Pavement condition surveys have been carried out at the following levels of detail (Toole et al, 1991I): 496 LEVEL 1 (PCS 1). Visual assessment and measurements within a lO0m length of road and for each lane. LEVEL 2 (PCS2). Detailed visual assessment and measurements within l1in sub-sections. LEVEL 3 (PCS3). Measurements taken at l1in intervals at specific chainage and wheelpath positions. The results given in this paper relate to the PCS2 and 3 surveys in which the structural and compositional properties of ten metre lengths of roadway have been related to performance. The principal performance measurements were rut depths measured at the test points and wheelpath cracking in each ten metre sub-section. The criteria used are given below: Rutting Sound < 10mm Critical 10-20mm Failed > 20mm Cracking Sound No cracks Initiated < 10% of length affected Critical 10-30% of length affected Failed > 30% of length affected 3. GENERAL DESCRIPTION OF THE STUDY SAMPLE The sample of roads included in the research comprises a total length of 360 km of road-link located in West Java from which twenty five representative sections, most of which were SO0m in length, were selected for detailed study. The road-links have been surveyed at intervals of two to three years whilst the monitoring sections were assessed at six monthly intervals. The types of overlay investigated include 22 sections of nomninal HRS and 3 sections of AC laid on asphalt base layers constructed to a common specification. Subsequent analysis of mnix properties has shown that 35% of the surfacings are of an AC-type, 50 % are of an intermediate gradation and 15% 9are of an RA-type. The thicknesses of the composite overlay ranged from 50mm to more than 200mm. The original surface prior to overlay was Penetration Macadam on 21 sections and natural granular materials for the remaining sections. Subgrade CBR's were between 1 and 6 per cent. 4. PROPERTIES OF THE ASPHALTIC SURFACINGS The properties of the asphaltic surfacings taken from each homogeneous road section, divided by mix specification, are given in Table 1. It is clear that mixes produced to the same target specification cover a wide spectrum and possess grading characteristics ranging from continuously- graded to gap-graded mixes. In order to distinguish between mixes a method of classifying them into gap-graded or continuously- graded types has been devised and involves determining the amount of material retained between the 2.36mmn and the 0.6mmu sieves. The method is based on the BS 594 Specification for RA (BSI 1985) in which an upper limit is placed on this amount. The limrit is set according to coarse aggregate content (CA) and ranges from 9 to 14 per 497 cent for CA of 55 and 30 per cent respectively. The following three classes of muix were identified in relation to the BS 594 upper limit: Gap-graded where the amount retained is less than the upper limit. Intermediate where the amount retained is up to 5% above the upper limit. Continuous where the amount retained is more than 5 % above the upper limit. Nearly 70 % of all mixes laid on medium and lightly trafficked roads are of the continuous type, the remainder are mainly intermediate. On heavily-trafficked roads, gap-graded, intermediate and continuos type mixes account for 30, 50 and 20 per cent respectively. Table I Properties of the asphaltic surfacings by target specification ontract TB B A F 1Vm (t) /b-- pakag (% (% (% (% (%) (%) urn) Ratio HOT-ROLLED SHEET CLASS A Specification > 7.3 > 6.8 20-40 47-67 5-9 3-6 >8 > 0.73, CP 39 8.1 7 53 30 8.4 2.4 7 1.2 (1.00) (1.00) (3.9) (3.2) (23) (1.4) (2) C- CP 40 7.0 5.8 54 30 8.9 6.1 7 1.5 (0.48) ,0.54) 27 214 (21) 2.5 (1 0.4) HOT-ROLLED SHEET CLASS B___ ___ Specification > 6.7 > 6.2 30-50 39-59 4.5-7.5 3-6 >8 > 0.73 CP 11 7.4 6.2 44 40 8.0 6.4 6 1.3 (0.37) (0.47) (5.5 (24 (3.2) (1 (0.45) CP 12 7.6 6.3 54 30 7.7 20 8 1.2 (0.40) (0.40) (2) ( ~ 19 L (0.30) CP 21 6.7 5.6 55 30 7.8 1.7 7 1.4 (056 (06)(1)(.) (19 J (0.25) CP 38 7.2 6.1 56 29 7.9 5.4 6 1.3 (0.52) (0.67) (4.7) (4.) (1.7) J32.5) (0.4) CP 39 7.2 5.8 52 30 10.2 5.2 5.4 1.9 ____________ (0.58) (.2 (3.5) 6~ (3.6) (.) .A!. .. i.. 'CP 40 7.4 5.8 52 33 7.8 5.7 7 1.3 ____________ (0.42) (03) (8.2) £(7.1 (21) (2.9) (1) j(0.3) ASP HALTIC CONCRETE C21 51.9) 48 518 28 (8.1) 3.5 6 1.7 0.39 0.50 (28) 27) (0.5) (2.3) 1) 0.19) Notes: 1. Mean and standard deviation (in brackets). 2 CA -Coarse aggregate fraction FA -Fine aggregate fraction FF -Filler fraction VIM -Voids in the mix(air voids) TBC -Total bitumen content EBC -Effective bitumen content Ft -Bitumen film thickness (effective) f/b -Filler- binder ratio (effective) 498 5. RELATIONSHI1P BETWEEN MIX PROPERTIES AND P.ERFORMANCE 5-1 General The data for locations where performance has been affected by factors such as poor road widening practice or inadequate pre-treatment of an existing damaged surface have been excluded from the analysis. The analysis has included the following: (i) Relationship between mix type and performace. (ii) The ability of two alternative specifications, namely 3S 594 (1985) and TRH 8 (1978) to identify satisfactory m.ixes. (iii) Influence of specified mix properties and composition (DGH 1986) on performance. Analysis (i) and (ii) above included data from all monitoring sections. Analysis (iii) has so far only included data from the most heavily-trafficked roads. These have carried 3,000 to 4,0,00 ESA per day for a period of five years, equivalent to 5.5-7.5 million ESA since overlay. The results should be considered as preliminary and their statistical base will be improved following the completion of further sampling and testing which will increase the sample five-fold. 5.2 Effect of mix type Time-series data for each sampled location has been examined and the percentage of locations giving satisfactory or unsatisfactory performance for each mix type and traffic level determined. The sample comprises 18 gap-graded mixes, 56 intermediate mixes and 45 continuously-graded mixes. The ages of the roads and the cumulative traffic loading correspond with between 35 and 50 per cent of the expected design life. Details are given in Table 2. Table 2 Performance of the overlays 499 Traf fic -All levels Heavy only mix Satisfactory Unsatisfactory Satisfactory Unsatisfactory (per cent) (per cent) ,.(per cent) (per cent) Gap graded 94 6 94 6 Intermediate 69 31 60 40 Continuously graded 71 29 so so The satisfactory category includes locations in a sound condition or where cracking has begun but extends to less than 10 per cent of the sampled location. In such cases normal routine or surfacing maintenance should be sufficient to arrest or retard deterioration. The incidence of severe cracking by mix type Is 6, 16 and -25 per cent for gap, intermediate and continuously-graded mixes resetively. The incidence of severe rutting is 0, 15 and 4 per cent respectively. The results confirm that gap-graded mixes generally give satisfactory performance. The performance of the other mix types is more variable. 5.3 Alternative specifications The performance of mixes which conform to two alternative specifications, BS 594 (BSI, 1985) and the South African (Guidelines for gap and semi gap-graded bituminous wearing courses (TRH 8) (NITRR, 1978), has been examined. Both specifications contain a combined grading envelope and recommend that a Marshall-design approach is followed to obtain the optimum binder content. BS 594 allows the use of a recipe approach where experience indicates that the mix will be satisfactory. Fifteen per cent of all mixes fulfill the BS 594 criteria for combined grading and the overall percentage success rate is the same as that for gap-graded mixes (see Table 2 above). Only in one case has failure occurred. When compared with the requirements of TRH 8, 39 per cent of all mixes fulfill the combined grading criteria. These comprise all gap-graded mixes and approximately 40 per cent of the intermediate mixes, again these mixes had high ratings for performance. 5.4 Effect of mix properties and composition The HRS specification requires the mix to conform to specified values for eight mix properties. It also states that the governing criteria should be to ensure that the air voids content lies near the centre o~f the specified range while the bitumen film thickness should be as high as possible (for maximum mix durability). The Marshall stability and Quotient criteria should also be met (for adequate mix strength). For practical purposes a minimum effective bitumen content and a range of mix design fractions are specified. Trial mixes are produced and adjusted until the criteria are met. It has not been possible to examine the original Marshall test criteria since records are not available but the effect of the following individual properties on performance has been examined In-place Air Voids Content (ie. After Trafficking) Bitumen Film Thickness Effective Bitumen Content Total Bitumen Content Coarse Aggregate Fraction (>2.36mmn) Fine Aggregate Fraction (75 um -2.36mm) Filler Fraction (<75 um) Effective Filler : Bitumen Ratio Per cent retained between 2.36mmn and 0.6mam sieves. 500 Table 3: Relationship between mix properties and performance for heavy traffic: Percentage of samples in each property range Effective bitumen content 4.7-5.19 5.2-5.69 5.7-6.19 6.2-6.69j 6.7-7.19 7.2-7.69 > 7.7 (%by weight of mix)I Sound 83 50 94 79 j 80 100 33 Cracks 17 17 6 0 20 0 0 Ruts 0 33 0 21 0 0 67 Total bitumen content 6.2-6.69 6.7-7.19 7.2-7.69 7.7-8.19 8.2-8.69 > 8.7 (%by. weight of mix) Sound 60 82 75 92 1 00 0 Cracks 20 9 5 0 0 0 Ruts 20 9 20 8 0 100 % retained 2.36inu-0.6mmn < 6 6-7 .99 8 -9 .9 9 10-11.99 12-13.99 Sound3 100 100 92 .75 63 Cracks 0 0 0 12 11 Ruts 0 0 8 13 26 501 The percentage of sound, cracked and rutted locations bave been determined for each property range and the results are shown in Table 3. Comments on these are summarised in Table 4. Table 4 Ef fect of mix properties on performance 6. CONCLUSIONS The main conclusions drawn from this re-search are as follows: 1) The performance of overlays laid to a number of target specification has been very'variable, with critical conditions being reached at the highest traffic levels within 35-50 per cent of the planned design lives. High variability in mix properties and non-mnix related factors have contributed to this result. 2) Relationships between the properties of the asphaltic surfacings and performance have shown that semui-gap to gap-graded mixes are superior to continuously graded asphaltic concretes. 502 MIX PROPERTY EFFECT COMMENTS SUGGESTED LIMITS Voids in Mix Rutting at low values. Use of 50 blow 2 -8 per ,Cracking at high values.-Marshall test cent probably underest ima ted traffic compaction. ____ ____ ___ ____ ___ Use 75 blow test _ _ _ _ _ _ _ Bitumen film Rutting at high values.- >6 microns thickness Cracking at low values. .Effective Rutting at high values. ~Rutting at low values >5.7 per .bitumen -Cracking at low values, coincides with continuously cent after content graded mixes. Cracking at allowing high values related to for plant ____________________high f/b ratio tolerances Total Bitumen- No clear >6.2 per content relationships. cent Coarse No clear trend - 45 -55 Aggregate _ _ _ _ _ _ _ _ _ _ _ _ Fine Low FA (<30 per cent - 30 -40 Aggregate equates with rutting) ______________ Filler More than 10 per cent - 4.5 -10 ________________equates with cracking Filler to More than 1.5 equates - 0.7 -1.3 binder ratio with cracking (0.7 -1.0 ________ _______ ____ __ _______ _______ _______p referred ) Material 2.36 Good performance at low - <13. to 0.6mm values. preferably Variable performance at <8. ___ ___ ___ ___ ___ high values _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ * for gap or semi-gap graded mixes 3) Two alternative specifications for gap and semi-gap graded mixes were assessed and found to relate closely to the properties of satisfactory mixes. 4) Further research is required to develop an improved HRS specification for use in the most heavily trafficked design situations. The results of this research will be used as the basis of a trial specification. 7. ACKNOWLEDGEMENTS The work described in this paper forms part of the collaborative research project being undertaken by the Indonesian Institute of Road Engineering and the Transport and Road Research Laboratory. The paper is published with the permission of the Chief Executive of TRRL- and the Director of IRE. Crown Copyright. The views expressed in this paper are not necessarily those of the Department of Transport,' United Kingdom. Extracts from the text may be reproduced except for commercial purposes, provided the source is acknowledged. 8. REFERENCES THE ASPHALT INSTITUTE (1983). Asphalt overlays for highway and street rehabilitation. Manual Series No. 17 (MS-17). The Asphalt Institute, College Park, Maryland, USA. BSI (1985). Hot rolled asphalt for roads and other paved areas. British Standard BS 594. London: British Standard Institution. CORNE, C.P. (1983). Optimising pavement overlay design for Indonesia. Proc. 4th Conif. of Road Eng. Assoc.' of Asia and Australia, Jakarta, Indonesia. DIRECTORATE GENERAL OF HIGHWAYS (1986). IBRD) Highway Betterment Programme: Specifications for High Durability Asphalts. Central Design Office, Jakarta, Indonesia. NITRR (1978). Selection and design of hot-mix asphalt surfacings for highways. National Institute for Transport and Road Research, Technical Recommendations for Highways, TRH 8. Pretoria, South Africa. TOOLE, T., S. MAHMUD and A. TATANG DACHLAN (1990). Research on Hot Rolled Sheets in Indonesia. In: 4th Conif. on Road Eng., Jakarta, 19-21 November, 1990. TOOLE, T., S. MAHMUD and A. TATANG DACHLAN (1991). Research on hot rolled sheets in Indonesia: Design and implementation of studies of existing overlays. IRE Research Report No. 11 .024.TJ.90. Bandung: Institute of Road Engineering. 503