Sutiace dre=ing in developing countries: research in Kenya by L. S. Hitch TRANSPORT and ROAD RESEARCH LABORATORY Department of the Environment Department of Transport TRRL LABORATORY REPORT 1019 SURFACE DRESSING IN DEVELOPING COUNTRIES: RESEARCH IN KENYA by L S Hitch The work described in this Report forms part of the programme carried out for the Overseas Development Administration, but any views expressed are not necessarily those of the Administration Overseas Unit Transport and Road Research Laboratory Crowthorne, Berkshire 1981 ISSN 0305–1293 CONTENTS Abstract 1. htroduction 2. Surface dressing design 3. me research progr~e 4. Materials used in the trials 4.1 Mppings 4.2 Binders 5. Equipment and workforce 5.1 Bitumen distributors 5.2 tipping spreaders (or gritters) 5.3 RoUers 5.4 Workforce 6. me trial sections 6.1 Selection of the sites 6.2 Design of experiments 6.2.1 Average least dimension and shape of chippings 6.2.2 Existing surface condition, traffic and ctimate 6.2.3 Design 6.3 Construction 6.3.1 Binder and road temperatures 6.3.2 Rate of spread measurements 6.3.3 Samphng and testing 7. Post construction measurements 8. Discussion and results 8.1 Condition of ctippings Page 1 1 2 3 3 3 6 7 7 7 8 8 8 8 8 8 10 10 10 10 10 11 11 11 11 8.2 Effect of road temperature on binder selection 8.3 Double surface dressing 8.4 Traffic 8.5 Surface texture measurements 8.6 Road hardness @robe depth test) 8.7 Subjective assessments 9. Conclusions 10. Acknowledgements 11. References 12. Appendix: Surface dressing design, Jackson method: a summary @CROWN COPYMG~ 1981 Page 12 13 14 17 19 20 20 22 22 36 Extracts from the text may be reproduced, except for commercial purposes, provided the source is acknowledged SURFACE DRESSING IN DEVELOPING COUNTRIES: RESEARCH IN KENYA h many developing countries surface dressing is used as a running surface on new low-cost roads as we~ as a maintenance treatment for existing roads. These countries areusuWy ckracterised bycombinations of road surface condition, traffic, cfimate, and materials, that are not found in Britain andthat arenotfavourable forgood surface dressing. Insuch cases the recommendations for road surface dressing given in Road Note 39 which is intended for use in Britain, are not apphcable. Jackson (1963) has proposed a method of surface dressing design based partly upon ideas put forward by Hanson (1935). This method was used to design a total length of 27’km of experimental sections of surface dressing in Kenya during 1975–77. This report describes the construction, monitoring and performance of the sections during the first 34 years trafficking. It is concluded that Jackson’s design method works satisfactorily in Kenya conditions over the range of chipping size used in these trials, namely 5 rnm to 13 mm average least dimension. 1. INTRODUCTION Surface dressing sealsthe surface of a road pavement with a fih of binder thus preventing the ingress of water and air. The binder fti is covered by a layer of stone chippings which provides a non-skid surface and replaces to some extent stone that may have been lost by attrition from, or embedment into, the existing surface. The surface dressing process is relatively simple and cheap in comparison with other resurfacing treatments, but nonetheless requires close attention to design and control. ~st surface dressing is an effective and appropriate surfacing for marry new paved roads in developing countries, its most important application is for the maintenance of existing bituminous surfaced roads. Road maintenance presents a growing problem in many developing countries notably because of the ever-increasing demands being made on the materkd and equipment resources avadable. Efficient use of the surface dressing process is therefore vital; the designs,materials and plant used must be such that dressings with optimum fife and performance are obtained. ~st many of the problems encountered with surface dressing in developing countries can be resolved by effective training, there are often problems that are due to the use of inappropriate designs or materials. The design process, particularly, needs to be apphed with special care, since there is often a greater possibfity of unfavorable combinations of surface conditions, traffic, and chmate than usutiy occur, for example, in Britain. To study the effect of these factors on the surface dressing.process, fufl-scde road trirds have been undertaken in Kenya under a cooperative research programme between the OverseasUnit of the Transport and Road Research bboratory and the Kenyan Ministry of Transport and Communications (MoTC). The prime objectives of this research programrne were: (i) to examine the vtidity of a selected surface dressing design method (ii) to examine the performance of binders of different viscosities (iii) to examine the hves of surface dressings constructed using chippings of different sizes. This report describes the construction of the tnd sections and givesthe results of their performance over 3 years of trafficking. 2. SURFACE DRESSING DESIGN In 1972 the Transport and Road Research hboratory issued Road Note 39 ‘Recommendations for road surface dressing’l. The content of this Note can be related directly to UK practice; the materials and conditions of develop~g countries and the tropics in particular are, however, not considered separately. Road surface temperatures in the tropics are generdy higher than in the UK and WWinfluence the embedment of cfippings. Wtison2 has reported temperature measurements made on 4 smti experimental pavements laid adjacent to the fuU-scde experiment at Mconbury by-pass; the area is chaotically typical of most of southern En@and. During a 12 month period June 1967 to May 1968, an ‘average’year, thermocouples at a depth of 20 mm recorded the foflowing for the 3 hottest months in 1967. Temperature range (°C) Duration of temperature range (hr) (average per day in brackets) June July August 25–30 109 (3.6) 126 (4.1) 99 (3.2) 40–45 18 (0.6) 59 (1.9) 8 (0.3) k contrast, road surface temperatures in Kenya recorded during 1976–77 were invariably higher than 30°C by 1000 hrs, and were usutiy at 40–45°C by 1200 hrs. A maximum of 52°C was recorded in the early afternoon at two sites wtich were approximately 500 km apart and which differed in altitude by some 1500m. It is thus clear that on temperature grounds done Road Note 39 is not directly apphcable to the tropical conditions found in many developing countries where, moreover, chippings are often of poor qutity and choice of binder is severely hfited. Severaltropical countries have developed national specifications for surface dressing that suit local conditions but there remains a need for a surface dressing design guide that is suitable for tropical developing countries in general. h 1935 F M Hanson3 propounded a general approach to the problem of the selection of optimum rates of spread for binder and chippings in surface dressing. This approach takes account of the volume of the voids that exist between the cMppings when they are spread, when they are ro~ed and when they are 2 trafficked, (typicafly 50,30 and 20 per cent respectively). Hanson considered the orientation of cfippings that occurs due to the effect of traffic and observed that chippings tend to settle in a position in which their least dimension is vertical, thus giving rise to the concept of the ‘averageleast dimension’ (Am) of the chippings. Hanson’smethod of design for surface dressing has been adapted by others working in this field4 and in 1963 Jacksons put forward a design method that incorporates the concept of average least dimension and which includes 4 factors representing: (i) the level of traffic, (ii) the state of the existing road surface, (iii) the chate, and (iv) the nature of the chippings. Jackson’s method of design is summarised in Appendix 1. The method is applicable to conditions that exist in both tropical and temperate cbates. In the cooperative research programme between the Werseas Unit, T~L and the Kenyan MoTCJackson’s method was apphed and evaluated in Kenyan conditions. 3. THE RESEARCH PROGRAMME Fufl-scde surface dressing trials were constructed at 14 sites in which the fo~owing factors were examined: (i) Binder viscosity: The effect of binders of different viscosities on the performance of surface dressings. (fi) Chipping size: The effect of different chipping sizes on the performance of surface dressings. (iii) Wuble surface dressings: The effect of different combinations of chipping size and the ‘spht’ application of binder. 4. MATERIALS USED IN THE TRIALS 4.1 Chippings In Kenya it was found that nominal %in (20 mm) tended to contain a substantial proportion of oversize material, wtist the nornind %in(14 mm) cbippings tended to be undersize, often severely so. ~lst the Kenyan MoTC specifications for chipping sizes are expressed in Metric units6, in practice during the trials quarries tended to supply nominal Imperird sizes7. Tables 1 and 2 give results of grading and other testsg performed on samples of the chippings used in the trirds. The average least dimensions of the chippings were obtained by the use of a nomograph methods. It wi~ be seen from Tables 1 and 2 that many of the samples failed the BS636 grading specification, the fdure rate being as foflows: Notid chipping size No. of samples that fded %in (20 mm) 5 (45%) %in (14 mm) 11 (92%) 3 * TABLE 1 %inch (20 mm) chippings suppfied for surface dressing, Kenya 1976/77 Province I Eastern ( Rift Vdey I Western I coast Expt/Project I KSD 1 14 I 15 10 1- 6.3 1- 5 1- Per cent nominal size Spec: 60 (tin)(l) I 29 Per cent oversize Spec: 15 (max)(l) I 56 Per cent undersize Spec: 7 (max)(l) nd Fltiess index Spec: 35 (rnax)(l) 20 Pass/Fd F Averageleast dimension (Am) (mm) I 13 Aggregate crushing value(2) 19 Pokhed stone value(2) – Notes: (1) BS63: (Part 2): 1971 (2) BS 812:1975 100 I 99 75 ] 94 14 I 24 * — 1 — — — — + 61 70 25 5 d I 1 18 I 30 11 I 10 –i– KSD KSD Re-seahng KSD KSD B.I. 7 9 (Kibwezi) 4/5 4/5 Ahero 100 I 100 I 100 I 100 I 100 I 99 [ I I I I 97 96 90 90 99 56 33 31 7 7 I 27 I 9 1 1 2 I 3 3 1 I 1 I I — l–l– l–l–l– — — I — I — — — I I I 64165183 183172]47 3 4 10 10 1 43 1 1 2 3 3 d 11 I 16 I 13 I 36 I 30 I 20 I I 1 I I P P P F P F 12 I 12 I 14 I 11 I 11 I 14 I 1 t 1 1 — — 19 I — — — -1-1-1-1-1- 100 I 100 213 18 I 14 I & 2 3 19 22 F P 12 I 13 1 — 26 I “, ,,,,, .,. , ““. .,.,, ,,, ,, ,., ,“. ,, . ,,, . ,,, .,,,,, ,,”. ., .,. TABLE 2 %inch (14 mm) ctippings supphed for surface dressing, Kenya 1976/77 Province I Expt/Project KSD 1 Sieve (mm) (% pass cum) 28 20 1- 14 I 100 10 54 6.3 14 5 1- Per cent nominal size Spec: 60 rnin(l) 46 Per cent oversize Spec: 15 (max)(l) Id Per cent undersize Spec: 7 (max)(l) 14 Flatiness index Spec: 35 (max)(l) 23 Pass/Fail IF Average least dimension (AD) (mm) I 6 Aggregate crushing value(2) I – Pohshed stone value(2) – Notes: (1) BS63: (Part 2): 1971 (2) BS812:1975 Eastern Central Rift Western Coast KSD KSD KSD KSD KSD Re-seting KSD KSD B.I. KSD KSD 2 3 3 7 9 (Kibwezi) 8 4/5 Ahero 12 14/15 — — — — — — — — — — — — — — — — 100 100 100 100 100 100 100 100 100 100 100 97 98 79 83 89 95 77 88 73 60 67 45 58 23 21 31 47 25 41 14 12 10 3 3 4 1 ‘– 20 8 20 4 10 2 1 2 2 – – – 23 12 27 40 33 55 40 56 62 68 48 d nil d nil d 3 2 21 17 11 5 25 41 14 12 10 3 3 4 1 nil 20 18 30 27 23 26 25 26 42 30 25 28 ‘F F F F F F F F F P F 6 5 6 6 6 8 7 7 8 8 7 — 21 21 20 – — — 19 – – – 52 – – 54 52 54 53 46 47 43 63 ,,,,,,,,,”,”“““,.,,,,,.,.,“““” 4.2 Binders Penetration grade bitumen, 80/100 pen, was normdy used for maintenance reserdingthroughout Kenya in 1975–77, the period during which trials were commenced. Since other grades were not obtainable this bitumen was used in aflthe trials, binders of lower viscosity being made from it by blending 80/100 pen bitumen with diesel fuel shortly before spraying (Nate 1) (kerosene was not avaflable for this purpose). The resultant binders were aflin the penetration grade range except for the trial at one site (KSD 8) where the binder viscosity was reduced to 225 sec STV at 40°C.(see Table 4). Figure 1 shows the viscosity/temperature relationships determined in the laboratory for 5 of the blends used. The viscosities were measured in centipoises (cp): however, within the relatively smd temperature range shown in Figure 1 these viscositiesmaybe equated without significant error to centistokes (dynatic viscosity (cp) = kinematic viscosity (CS)x density at temperature of measurement). It should be noted that although the blends used in the trials have been classified on the basis of ( penetration value they would not necessarily have the same curing properties in service as refinery-prepared bitumens of the same penetration which norma~y contain fluxing oils less volatile than diesel od. Figure 2 which includes the results of tests done on blends made in the laboratory and in the field during the trials shows the relationship between the penetration and the ddution with diesel oil of the 80/100 pen bitumen used in the trials. The scatter of the points can be attributed to: (i) inaccuracy of the contents gauges on the distributors (ii) the absence, in some cases, of accurate diesel dispensing equipment (iii) differences in the initial penetration value of the bitumen drawn from stocks at different depots. h order to investigate the relative rates of curing of the blended binders in Kenyan ctimatic conditions a series of curing trials was undertaken. Resources avtiable did not permit the removrd of specimens of surface dressingsfrom the roads at the required intervals, and thus a set of trays containing simulated surface dressings was exposed on a roof site in Nairobi over a period of 3 years. ~st these simulated surface dressingswould not reflect the effects of tyre rubber and other depositions on the road surface, neither would they be subject to the softening of binder caused by fuel and lubricating ofl spiflage on vehicles. Blends, approximately 3 fitres each were prepared in the laboratory as shown in Table 3, using either ~ diesel or kerosene (domestic ‘paraffin’). TABLE 3 Blends of 80/100 pen bitumen containing diesel or kerosene Per cent (by volume) Diesel Kerosene Pouring Temp. (°C) Penetration at 25°C 4 248 285 155–165 8 370 648 140-150 Viscosity (ST~ at 40°C 12 268 sec. 108 sec. 130–140 6 Galvanised steel trays, area 0.76m2, were used for the exposure tests. Binder was heated to pouring temperature and spread by scraper to an even fti; the weight of binder appfied in each case was such as to provide a coverage of approximately 1.6 kg/m2 (normal rese~ng range = 0.8 – 1.3 kg/m2). I The ttick fti apptied was necessary in order to provide adequate stied samples for the subsequent tests. Grtite chippings (1Omm to 14 mm) were spread on the hot binder to provide fu~ coverage and were rofled with a 2.5 kg steel cytider. A smafl hole was dfled at one corner of each tray above binder levelto dow rain water to escape. One-quarter of each tray’s contents was removed after 7, 14; 24 and 36 months and the soluble binder recovered using Method IP.1059 but with the omission of a fractionating column. (This item was not avtiable but it is considered untikely that significant amounts of cutter ofls would have been lost.) Penetration tests were made on the recovered binders; the results are shown in Figures 3 and 4. The faster rate of curing of the kerosene blends is wefi demonstrated by the results, wtich can be su~arised as foflows:– Kerosene Mesel , Penetration range of 3 blends after 6 months 60/90 80/140 Penetration range of 3 blends after 12 months 40/60 50/80 ‘ Penetration range of 3 blends after 24 months 20/30 20/40 : 5. EQUIPMENT AND WORKFORCE 5.1 Bitumen distributors The fo~owing types of machine were used in the trials: Type Manufacturer Country of manufacture Constant volume Etnyre USA Constant pressure Phoenix UK tinstant pressure Ashurst UK These machines au had a maximum capacity of approximately 5200titres(1100 Imp. gal:) and were, with one exception, unctibrated. The Etnyre machine is more complex than the constant pres~ure machines and as a result the inter-relationship of certain of the binder system controls was not fu~y understood by most operators. 5.2 Chipping spreaders (or gritters) I 1 Either Hornsey tailboard gritters or pushed metering spreaders were used for the trials, except for one site where a modern self-propefled chipping spreader was employed. The simple and basica~y effective ttiboard gritters had, in most cases, suffered distortion~of the steel floor below the rotary gate. The resulting irregular gap produced uneven distribution of chippings; in severe cases,streaks of bitumen remained unchipped. Much rake and brush work was required to redistribute stone. These devices,moreover, w~st being wefl suited to the standard 2.25m (7.5 ft) spraybar as normdy used in Great Britain, were not ideal for the 3.3m (1Oft) spraybar of the Etnyre machines, 7 and echelon-gritting was necessary. (To have reduced the spraybar width to match the Hornsey gritter would have produced unacceptable comphcations in calibration and subsequent routine use of the distributor.) The USA-made pushed gritters generdy worked wefi, although crews tended to operate them at too fast a speed and the significance of the metering system was not fufly understood. 5.3 Rollers Norma~y Hyster pneumatic tyred ro~ers were used. At most sites an 8–10 Mg steel-wheeled roHer was dso avaflable, but this was not used on the trial sections because it invariably cracked or crushed the cfippings. 5.4 Workforce A typical reseahg operation required some 25–30 men inclutig supetiory staff and drivers. Approtiately 10 general labourers were needed at each resealing site for general duties such as road sweeping which was undertaken by hand. 6. THE TRIAL SECTIONS 6.1 Sale@ion of the sites The trials were necessarily confined to sites and materials for which maintenance funds had aheady been Wocated. Since reseting units often commenced work at very short notice no detded preparatory work was possible at a given site, such as a detded texture survey. Nonetheless it was usuafly possible to select a length of road for the trial that was reasonably straight and without serious surface defect; gradients however were unavoidable. Table 4 gives details of the 14 trials. 6.2 Design of experiments 6.2.1 Average least dimension and shape of chippings. ALD was determined by a nomog~aph methods (see Figure 5). Occasionally when work commenced at short notice aggregate properties could not be determined in advance; in such casesthe design of the surface dressing was done on site, ALD being assessedvisu~y. Subsequent laboratory checks enabled the design to be corrected if necessary but no large discrepancies occurred. The visual classification of flakiness was more difficult. Haky aggregate is defined8 as having a least dimension wtich is lessthan 0.6 of its nominal size, this size being taken as the mean of the Mting sieve sizes determining the size fraction. The current British Standard specification for sin~e sized chippings6 Wows a maximum flakiness index of35; only 2 of the 23 samples examined fatied in this respect. However %in (20 mm) nominal sized ctippings tended to contain unacceptable quantities of oversized material, whikt 1Ain (14 mm) stone was usua~y undersized. Visual assessment was probably influenced to a large extent by the shape of the undersized material. Six out of 12 samples of% in (14 mm) nominal size stone contained unacceptable quantities of material passing% in (6 mm) which is inadmissible for flakiness index determination. 8 I TABLE 4 1 Surface dressing rese~ch, Kenya (1976–1977) 1 Summary of trial sections constructed to July 1977 Road Temp. Oc (20 mm depth) StOOepropefiies ~tes of spread (binder) ‘ Nominal Trial type (@/m2) Bindersused size Am Pen at 25°C mm In. nm. Dmii (original) Min Mu Chippingsize % 20 13 1.1 0.8 1.4 1 % 14 6 0.8 0.9 1.0 80/ 100 % 6 3 0.7 0.7 1.0 Binder % 20 11 1.0 0.9 1.1 200/300 tiscosity % 14 6 0.8 0.6 0.8 300/400 Double sed % 20 10 1.1 1st 0.7SA 1.3SA 180/200 % 14 5 0.8 1st 0.7SA 1.4SA 200/300 Expfientd numbers bcation A 104 Nmnga Road 11.4 km from the MombasaRoad KSD 1 28 T 24 50 43 T 43 KSD2 KSD3 A 104 NamangaRoad 24 km from the MombasaRoad A 104 Nmmga Road 27.2 km from the MOmb~aRoad KSD4 A 104 Makatano 18.2 km from MO1O Nofih csmp 31 KSD 5 A 104 Makatmo 19.4 km from MOIO North Csmp 28 — 45 + 10 5 0.8 0.7 0.8 ‘ Triple sed % 16 9 1.0 2nd 0.6S 0.9N 80/ 100 % 10 5 0.91 St. 0.9s 0.9N % 10 5 0.9 3rd 1.0s 1.ON KSD6 A 104 Kipkaren 11.8 km from Turbo Post Office Double seal % 20 12 1.3C 1.0 1.3 80/100 1.4UC 1.0 1.3 ; % 14 6 0.9C 0.7 1.1 1.Ouc 0.8 1.2 KSD 7 MachakOs-Wmun~ (at Makutano) 26 31 KSD S KSD9 KSD 10 D 407 Umuru–Nairobi (Tigoni) 3.3 km from Bata Co, Umum 25 A 109 (Atbi River) T Ntiobi-Mombaw Road 4.6 km after E 434 tum~ff A 109 (Atbi River) Ntirobi-Mombaaa Road Emulsion % 20 12 1.8 / 1.5 1.8 trial 1.3 I 1.3 1.3 : 29 46 14.4 km after E 434 tum~ff Ii%14 6 1.2 Ist ]1.1 I 1.3 I 60%emulsion 1.5 2nd 1.4 I 1.4 (CatiOnic) KSD 11 ABANDONED KSD 12 C 29 (banda–Siaya) 13.7 km from Double seal ‘A 20 12 1.5 0.7SA 1.8SA 200/300 28 50 B1 turn-off handa Vtige % 14 8 1.3 1st 1.0 1.2 1.2 2nd 1.2SA 1.8SA ‘ Ssnd – - 0.8 2nd 0.8 1.0 I hmwmge–Mugumo Double sed % 20 11 1.3C 1.0 1.5 , *ct. 1 beginsappmx 4 km Emt of Thunguri MOWdepot % 14 8 1.2 lstuc 1.1 1.4 200/300 26 50 1.2 2ndUC 0.8 1.3 , A 14 (TW1). Section 1begins Binder % 14 7 0.9 0.7 1.2 200/300 11.2 km from top of tikoni Ferry tiscosity Ramp (South) 300/400 30 52 400/500 A 109 Mm&hi. Section 1 begins Binder % 20 13 1.1 0.9 1.5 180/200 Bppmx5 km Mombasaside of tiscosity t Mariakani E 14 7 chippingsize 0.9 1st 0.7 1.2 200/300 30 45 1.0 2nd 0.7 1.0 doo/5oo tid – – 0.7 2nd 0.9 0.9 KSD 13 KSD 14 KSD15 Abbreviations: AD = Averageleast dimension UC = Uncoated cbippings SA = ‘SpfitApplication’;(rates apphed as a percentageof total STV = Standard tar viacometer binder both layers) c Pen. = Penetration of bitumen at 25°C = Gated cbippinga Notes S,N = South, North 1. Bindersused: Bindersother than 80/100 or emulsion we blends of 80/100 and diesel fuel. The penetration valuesobtained on these blends have been considered ss f~ng tithin the mngesshown, which are either arbitrary (eg 300/400) or represent a penetration range fikely to be found in bhder specificationsin Europe or USA. 1 2. Stone proxtiies: ~c metric sties quoted are the nearest equivalents; e~ct equivalentsare aa foUows:– %in – 19 mm, %in – 12.7 &, %in – 9:5 mm I 9 6.2.2 Existing surfaw rendition, traffic and climate. The condition of the existing surface and the traffic was assessedas required in the Jackson design method 5. The chmatic condition of most of the 14 experimental sites was assessed as semi arid (hot/dry) although two sites at approximately 2100m (7,000 ft) altitude were classified as ‘temperate’ and one other site (at sea level) as ‘tropical’ (hot/wet). 6.2.3 Design. The design rate of spread for the binder, D, was obtained from the design chart (Figure 6) using the overafl factor for the site (see Appendix) in conjunction with the ALD of the stone. It should be recorded however that Jackson’s recommendation that rate of spread of binder be reduced by 0.1 kg/m2 when using penetration grades was not implemented, because of the presence of volatde ofls in the blend. Measured rates of spread are reported as D t x per cent. (For double seals D represents the total design quantity for the 2 layers.) 6.3 Constru~ion h order to simphfy the monitoring of performance, a standard section length of 200m was adopted. Since in a typical experiment binder was appfied at the design rate D and D t 10 per cent a group of three such sections, ie 600 finear metres (average width 7m) could conveniently be constructed from one distributor-load of binder. In practice this represented one batch of blended binder. It was found to be most convenient to spray one half-width of road for three sections, then returning to the start point for the second half-width. It was normrd practice in Kenya for reseafing units to spray hdf the width of road for some distance, the completed work then being kept closed to traffic for several hours whilst rolling continued. It was considered however that experimental work should, as for example in Europe, be exposed to the early-fife bzards which affect routine reseating; thus when work moved on or ceased for the day, barriers were removed and W completed work was opened to traffic. At some sites vehicles entered the road from unsurfaced minor roads; the transported dust or mud adhering to the road could not be entirely removed by the manual means avadable and the uniformity of the finished work suffered as a consequence in these areas. Distributor drivers sometimes fded to overlap the first spray adequately; also, because of the gritting equipment available,it was not practicable to leave 150 mm of binder unchipped at the centre he during the first pass, to be overlapped on the second apphcation. These two factors produced unsprayed strips at the centre fine or double-sprayed and chipped centre tine joints which usutiy developed a ragged appearance under traffic. 6,3.1 Binder and road temperatures. The temperature of binder in distributors was checked regularly by Rotatherm thermometer at the loading hatch; appropriate spraying temperatures were selected for the nornind penetration ranges to which 80/100 pen had been blended. Air and road temperatures were recorded at hourly intervals as far as possible during experimental surface dressing work. Road temperatures were recorded using a smW mercury-in-glass thermometer (0–50°C) with its bulb immersed in an oil-fi~ed hole 15–20 mm deep (see Figure 7). 6.3.2 Rate of spread measurements. Lightmetal trays approximately 10 mm deep and 0.1m2 in area were used to check rates of spread. Three trays were placed for each 200m run of the distributor and the weight of binder deposited on each was recorded. The rate of spread, taken as the mean of three trays, assisted in the ctibration of the machine and verified the rate of binder actually sprayed. The 10 unsealed squares beneath the trays were repaired by hand in the earher trials but subsequently they were Mowed to remain thus providing a comparison between the original and the resealed surface. Unsealed squares were always repaired during work on new bases. The poor condition of some of the distributors sometimes prevented the required rates of spread from being obtained”consistently. 6.3.3 Sampling and testing. Samples of chippings taken during construction, usually from the gritting vehicles, were checked for ALD and other properties. A 5 fitre sample of binder was taken from the spray bar after spraying the first 200m pass and this was subsequently tested for penetration or viscosity. 7. POST CONSTRUCTION MEASUREMENTS The trials were Wowed to settle down under traffic before commencing post-construction measurements; the period varied considerably. Most sites were revisited within 2–3 months of construction, but two remote sites which carried very light traffic were not visited for 7 months. Two methods of measurement were used to assessthe performance of the surface dressings:– (i) (ii) The sand patch test. The test is described in TRRL Road Note No 27 10 and involves careful formation of a known volume of sand into a circular patch on the road surface (Plate 2). The diameter of the circle is recorded when the sand has been fufly acconunodated by the road texture, the depth of which is then calculated. Four tests were performed at each of 3 chainagesin the wheel tracks and 2 tests out of the wheel tracks at one chainage in each 200m long trial section. The texture depth of a surface dressing tends to decrease under the action of traffic, untfl an unacceptable level of smoothness is reached; it then becomes necessary to restore texture depth (and hence skid resistance) by resetig. The probe depth test. This test measures the hardness of the road surface so that chippings of 11 suitable stie can be selected for surface dressing. A Sod Assessment Cone Penetrometer , origina~y evolved for the rapid determination of soflstrength in situ, is fitted with a 4 m diameter hardened steel probe in place of the original cone. A pressure of 278 kg/cm2 is maintained for 10 seconds and the penetration of the tip into the road surface is recorded @late 3). The test is normafly performed before the surface dressing is designed but in this case the probe tests were done later, usua~y in the unsealed squares left by rate-of-spread trays. The test was not used at 3 sites where surface dressingswere appfied over new crushed-stone bases. h addition the surface appearance of the surface dressings has been photographed at yearly intervals; a tripod-mounted camera is focussed on a 1m2 frame alongside each sand patch test point. Contiguous pairs of photographs are taken so that each photo records one wheel track and part of the ofllane. Subjective assessments of each section have also been made annually and traffic counts taken. 8. DISCUSSION AND RESULTS 8.1 Condition of chippings ~ppings stock pfles often contained undesirable amounts of dust; at two sites near Nairobi, where rain fefl a few hours before the surface dressing was due to start, the dust was found to have formed a thin ‘slurry’ coating on the chippings. The condition was evidently recognised by local operators as a 11 potential source of faflure resulting from non-adhesion of the chippings. However whilst undesirable, this did not prevent the construction of a good surface dressing, provided that binder viscosity was appropriate to road temperature. ‘Slurrie& stone taken from the stock pile ody two hours after heavy rain had f~en dried out rapidly on the road and was wetted readdy by 200/300 and 300/400 pen blended binders. At a subsequent trial near Mombasa, very du$ty %in (14 mm) bestone ctippings were suppfied; the stock pfle was hosed down immediately before use and, again, good adhesion was obtained. 8.2 Effed of road tempertiure on binder seledion One of the primary objectives of the trials was to examine the performance of binders of different viscositiesin reserding. In effect this impties validation of the chart shown in Figure 8 which was prepared by SheUInternational Petroleum Company and which shows that, when surface dressing, the binder in use should have a viscosity at road temperature within the range 104 to 6 x 105 centistokes (for ‘heavy’ traffic conditions this range is somewhat sma~er). The significance of Figure 8 is that two conditions need to be satisfied simultaneously: (a) chippings must be wetted by the binder, (b) chippings once wetted must be retained. This chart is a guide to the use of hot petroleum bitumens ody: cord tars occur very infrequently in developing countries and are not therefore included. Bitumen emulsions on the other hand have a low viscosity; thus spraying and wetting occur readily. Upon ‘breaking’the emulsion yields a relatively high viscosity penetration grade bitumen wtich retains the chipping. The chart in Figure 8 therefore has fittle relevance to the use of emulsions. Ody four sections were constructed with an emulsion binder, which was a cationic (60 per cent) emulsion (site KSD 10). Their usefulness was largely lost because of faulty spraying equipment and no useful conclusions can therefore be drawn from these four sections. The maximum and minimum temperatures of roads on wtich different blends were sprayed in the Kenyan trials,have been plotted on Figure 8. Since the actual penetration values obtained on certain blends did not f~ within the close ranges maintained by commercial producers, the temperature data for these has been plotted on the nearest appropriate grade fine. Of the four points which appear below the 400/500 pen-he, two correspond to a 600 pen blend and two represent a blend having a viscosity of 225 sec STV at 40°C (equivalent to 9 x 104 centistokes at 40°C). Of the 46 points plotted, 20 (43 per cent) fd within the permissible range. The remainder would be classified as ‘poor wetting’ risks according to the chart. However these include binders used at 3 new construction sites and at 3 very tightly trafficked sites where prolonged rohg could proceed unhindered and traffic whip-off risk was virtudy non-existent. The softest binders worked excellently; a 600 pen blend held %in (20 mm) stone we~ under medium-heavy traffic on the Nairobi–Mombasa Road near Nairobi, w~t a 400/500 pen ,wasvery effective with a %in (20 mm) dusty stone on the same road near Mombasa. The 225 sec STV at 40°C blend was used successfu~y with %in (14 mm) chippings under medium/heavy traffic near Nairobi. 12 At the one main site where 80/100 pen binder was used surface temperatures were between 33–50°C, traffic wastight and chipping lorries were made to fo~ow close to the distributor. These conditions are much more favorable than those norma~y experienced in some developing countries and the resultant success using 80/ 100 pen at this site is not therefore remarkable. Figure 7 shows the relationships between air and road temperatures recorded at surface dressing sites during 1976/77. The scatter of points is due to various causes, eg heating/coohng cycles, changes of wind direction, etc. The chart should, however, assistin predicting road surface temperature from air temperature readings; prediction is Rely to be rehable to within plus or rrdnus 5°C of the true value in the road temperature range 25/45°C. This range represents virtuafly aflthe sites which were surface dressed in the research programrne and should therefore be applicable for routine reseahg work in Kenyan conditions. The significance of the predictable accuracy is that, if the binder to be used for reseting is selected such that the predicted road temperature (Figure 7) intersects the binder grade fine at 105 centistokes (Figure 8), then an error of* 5°C in prediction W be tolerated by the permissible range of working viscosities. Figure 7 dso shows the temperatures at which 4 selected binders have a viscosity of 105 centistokes. 8.3 Double surfam dressing Double surface dressings were constructed at four sites, the main features of which can be sutiarised as fouows: – Site No. Features KSD 3 tighttraffic; ‘sptit’ binder application, delayed and undelayed second seals; 2 stone size combinations. KSD 7 Very tight traffic; standard binder application; delayed and undelayed second seals; coated and uncoated chippings. KSD12 tight traffic; spfit and standard binder apphcation; undelayed second seals; chippings or sand for second seals. KSD 13 tight traffic, standard binder apphcation; undelayed second dressings; coated and uncoated chippings, 2 stone size combinations. The standard specification in Kenya for double surface dressing ca~s for %in (20 mm) chippings to be fo~owed by %in (14 m). Table 2 shows the extent to which nominal% in (14 mm) chippings supplied ftied to meet British Standard 636. It could be argued that some samples of the %in (14 mm) material more nearly represented + in (10 mm) chippings containing oversize material. Their performance when used as second sed cover aggregate however lends support to the proposition that the aggregate of the second sed should differ in size by at least 50 per cent from that of the first seal. It is suggested,therefore, that for double seals, %in (20 mm) chippings should be fo~owed by ~ in (1Ornrn) material. For example this would be suitable over a rich bituminous surfacing wfich would permit the larger chipping a degree of embedment. This is Wustrated by the experimental site KSD 15 where the old bituminous surfacing was extensively cracked; here the trial sections constructed showed that: (i) A sin~e serd(nominal %in (14 mm) chippings) was not sufficient to sed the badly cracked surface. A double sed %in (20 mm) plus a %in (14 mm) was in good condition after 2% years trafficking. 13 (ii) A double seal, %in (20 mm) chipping plus sand as second seal cover material, was also in good condition after 21Ayears and is potentia~y an effective treatment. This combination was also tried successfu~y at a new road site with a crushed rock base (KSD 12). Some engineers favour the use of double sealsfor all routine maintenance, irrespective of pavement condition or traffic. Assuming the existence of a sound pavement structure, the main parameter determining the fife of a surface dressing is likely to be rate of loss of surface texture, and it is questionable whether or not this is Wely to be enhanced by the use of a double rather than single seal. Observation of some completed double sealssuggeststhat, in the absence of good control, the chances of fatting-up are greater than with sin~e seals. Double seti are useful however for the maintenance of roads which are badly cracked or which have a rave~ed surface. In afl cases experience indicates that the performance of a double sed is improved by delaying the application of the second sed by at least 14 days, as opposed to applying mediately the first seal has been laid. The double surface dressing trials showed clearly the advantage of dewing the first sealto be trafficked before placing the second. The Jackson design method, applied individua~y to each layer appeared to produce a correct rate of spread of binder. (Note: the etisting surface condition for the second layer was rated as ‘verylean bituminous’, because of the asperity of the first layer.) In view of the success of this approach it is difficult to justify an arbitrary ‘spfit’of binder between the two layers. However; one authority12 suggeststhat the tot d application of residual binder be spfit: First layer 40 per cent Second layer 60 per cent This is likely to be safely achieved ody on new construction or on a very lightly trafficked road. In other situations the first dressing which is likely to be a % in (20 mm) material maybe whipped-off by fast vehicles. This occurred at experiment KSD 11 which was to have been a split-application trial on the Nairobi-Mombasa road (approximately 2500 vpd): sections chipped with a nominal %in (20 mm) material suffered severe whip-off and because of,an unavoidable delay the second seal could not be placed before much damage had occurred. Cost estimates for 6 reseafing projects in one province of Kenya show fiat 45 per cent of the total cost was attributable to the second sed; the potential saving of using single seals only is therefore significant. 8.4 Traffic For the purpose of designing the trial surface dressingstraffic levels were estimated. The results of traffic counts made subsequently are shown in Table 5. Table 6 shows the effect of differences between the estimated and actual traffic classifications on the rate of spread of binder. It WUbe seen that if traffic is wron~y classified by one traffic category then the design ROS (binder) wfll be affected approximately as fouows: Stone size (nominal) Effect of difference of one traffic category %in (20 mm) D *10 per cent lAin (14 mm) Ne~igible 14 TABLE 5 Experimental surface dressing, Kenya; traffic data (source: Kenya Ministry of Transport and Communications) Years Traffic class Survey site Total Mean totrd Actual I Asestimated for design Experiment Cars Mght goods Medium goods Heavy goods Buses KSD 1/3 KSD 1/3 187 189 a I 1976 and 7 140 139 10 8 484 473 fight I Very tight 1 b ! 1976 and 7 132 126 6 9 462 , KSD 4/5 C I 1976/77 219 126 97 32 35 Medium I Medium — 28 509 509 , d ] 1977/78 106 3 1082 I 1098 A Me&u Ught eavy KSD 8 KSD 8 579 366 e I 1977 —776 913 224 68 2 44 1114 z Me&u Heavy eavy Ught Very tight f I 1976/78 523 333 119 2338 I 2376 KSD 9/10 450 KSD 9/10 138 2413 KSD 12 g 1976 and 7 h 1976 —976 61 458 550 291 91 62 1 17 232 232 — No data: 100< Estimate <300 * KSD 13 1976 and 78 552 457 225 17 111 1362 1976 and 78 493 439 144 15 86 1177 t 1192 KSD 14 / I KSD 14 Medium KSD 14 k 1976 and 78 451 352 128 19 88 1038 1] 1976 and 78 421 250 180 252 138 1241 1 1369 1977 419 285 301 340 152 1497 1 /Heavy I KSD 15 m KSD 15 n ,,, ,,, .,., “., ,.. . . . . ,. ..,, ,.,,, .“”” TABLE 6 Effect of incorrect traffic classification on Design ROS (Binder) (Note: ‘%’and ‘W are nominal stone size,inches) Rate of spread (Binder) (D) (kg/m2) Original Revised %—1.1 1.0 %–0.8 0.8 i—l .0 1.0 lA–0.8 0.8 %—1.0 0.9 %–0.8 0.8 1A—1.0 0.9 Experiment Overti design factor Change in ROS (% of D) Original KSD 1 –1 Revised D–10 –3 — — — I KSD 2 I –1 –3 KSD 3 (bwer) –1 D–10 –3 — D–10 I t2 No change due to traffic classification (Revised during construction due to old surface) KSD 4/5 I t2 t2 D–n Dtl 1 Original road surface Dt17 different Sections 1-9 and Sections 10-19 D–7 D–8 i KSD 8 I –2 –4 lh—o.9 0.8 Sin@eseals KSD 9/10 only – %–5 %–9 –3 _7T 0.9 0.6 1.0 0.7 KSD 12 (hwer) %tlo t8 1.5 1.4 KSD 12 (Upper) ~t8 t6 1.2 1.1 KSD 13 I No survey KSD 14 %0 –1 0.9 I 0.9 1- ! KSD 15 No change required t Note: the Jackson design method5 does not actually include factors for these combinations of traffic/ surface/chmate/chipping. Normdy a quite rudimentary traffic count should be sufficient to avoid such errors in calculating ROS, which are less than those hkely to arise from such causes as bad distributor operation, poor mechanical condition, etc. 8.5 (i) (ii) (iii) (iv) Surfs@ texture measurements bss of texture can be caused by: embedment of chippings abrasion of chippings excess of binder deposition of foreign material. Embedment could be expected to occur on d but 3 sites, ie KSD 7/ 12/ 13, which had new crushed rock bases. The existing surfaces at the remainder of the sites were either premix or old surface dressings. Abrasion of chippings was most tikely to occur at only 2 sites, ie KSD 14 and 15, where soft local Emestone chippings were used. Deposition of foreign material can take a number of forms, the governing feature being (as far as it affects the sand patch test) whether the deposit consists of bound material (ie clayey or resinous materials) or sand and grit which have no permanent effect on surface texture udess they are retained by excess bitumen in the surfacing. bss of surface texture of the trhd sections has been calculated as foflows: (i) From tests at 3 chainages in each section the average texture depth for a given wheel track has been calculated for each time the measurements were made. (ii) The loss of texture depth for a given lane has been taken as the mean of the two wheel tracks in the lane. The percentage lossesin texture depth thus obtained have been plotted against age (months): Figure 9 shows a typical example. Table 7 shows the average texture depth recorded in a sin~e nearside wheel track for different dressing types in each experiment; this Table is presented in order to record relative magnitudes of texture depth for these dressing types. Table 8 shows the minimum/maximum percentages of texture depth lost under traffic; M experimental variations, ie stone size, binder ROS, wheel track and direction, are combined. In general terms, texture loss during the first 2 years appears to be: for tight traffic 5–55 per cent for medium/heavy traffic 35–85 per cent Table 8 shows clearly that texture loss is rapid during the initial 12–18 months whilst chippings become embedded and interlocked. Subsequent loss appears to be slow unless, asin KSD 15, an abradable stone has been used and/or the surface is relatively soft. 17 TABLE 7 Texture depths (mm). (Sin#e nearside wheel track ody) %U Period under traffic (months) Vehicle Traffic classification Dressing . per da~ type 2–44 54 — 3.66 — 1.99 473 — 0.71 — 3.82 473 — 1.82 — — — — 473 — — — — (in) (mm) Initial =A 20 5.17 1 % 14 2.83 % 6 0.95 3A 20 2 . 6.28 % 14 2.73 % 20 t% t 14 4.36 3A 20 3 t% + 14 4.11 % 14 t% t 14 2.73 tight tight Single Single Double 1 day delay Double 14 day delay Double 1 dav delav fight — I 1.54 1.34 % ,14 t% t 14 2.37 3A 20 5.68 5 % 14 3.50 3 10 1.86 8 % 14 3.27 =h 20 9 4.68 % 14 2.65 Double 14 day delay M — 1.25 1.02 1.01 2.28 – 1.78 1.10 – * — — — — 509 Medium Single ).77 ] 0.45 / – — — — — 1098 1.75 – 2376 1.25 – — — — — 232 — — — — — — est. a 100-30( — — 1.72 – 1192 — — — — 1369 — — — — Medium/heav Single 1 1 Medium/Heal Single [.35 1.11 I – H %t 20 2.66 sand % 20 12 +% t14 2.69 1.59 I 1.80 – !.04 1.80 – ..52 1.49 – !.04 1.98 – !.41 — — .67 – – L.65 1.65 – ).70 0.65 – tight Double Single 13 w hght Double * Medium/Heav Single I~;d120 1362 Double & % 20 15 +% t14 3-28 3A 20 4.48 1.08 0.94 – Medium/Hea~ Single I % ]14 ] 2.90 18 TABLE 8 Texture loss under traffic (to nearest 5 per cent of original texture depth) Experiment (KSD) = der traffic (months Veticles per day Traffic Dressing classification type 54 15–50 1 I 1 5–35 473 fight I Single 2 I I 5-45 20–60 I 25–60 473 Ught I Single , 31 115-55 + 473 Li&t I Double 5* I 25-75 I 50-85 509 Medium I Single 1089 Medium/Heavy I Single I + 2376 12~ I 0-40 I 5-50 232 estimated 100–300 13? I 0–50 I 14 ~ 35–65 I 35–65 1 1192 Sin~e + 15 Medium/Heavy 3 double I 1369 1 , * Athough traffic volumes were moderate, KSD 5 was subjected to some heavy Uganda-bound vehicles t KSD 12 ~d 13 were the o~y Sites at which surface dressings were appfied directly over a new crushed stone base 8.6 Road hardne% (probe depth teti) Figure 10 shows the probe penetration/road temperature relationships obtained; results for ordy one lane of the road are reported. Four characteristic curves are also shown in Figure 10; these are based on researchl 1 by the Materials Division, TRRL which has produced the foflowing penetration vahres correspondtig to the hardness categories given in Road Note 392: Category Penetration at 30°C (mm) Very hard o–2 Hard 2–5 Normal 5–8 soft 8–12 Very soft More than 12 (Testing at 20–35°C is recommended, foflowed by correction to 30°C using the curves shown.) ksuming that probe penetrations can be vtidy corrected from temperatures higher than 35°C the penetration values shown in Figure 10 indicate that ahnost W of the existing surfacings used for the trials would be classified as ‘normal’ or ‘hard’. 19 8.7 Subjetiive assessments The observations made during an inspection in 1979 are summarised for each experimental site in Table 9. h particular the apparent tolerance to under- or over-spray was noted. In assessingthe vfidity of the Jackson design method5 it is important to note than in the Kenyan trkds it was possible to use chippings only in the ALD range 5 to 13 mm. ~lst this might wefl represent the practical range of ALDs to be expected in most developing countries it should be noted that the design chart (Figure 6) was not tested for chippings with ALDs fting in the upper half of the chart. Table 9 shows that sections containing% in (14 mm) nominal size chppings appear to have an ROS (binder) tolerance of D t 10 per cent for most traffic situations. For cfimbing traffic a reduction of 10 per cent is beneficial whilst for fast or dow~ traffic an increase of 10 or even 20 per cent would seem desirable. The Jackson design method makes no provision for such adjustments. It is important that the needs of the whole pavement width be considered when, for example, assessing the performance of a given surface dressing; it would be misleading to relate the surface texture in wheel tracks only to rate of spread of binder. The ofl lane, centre tine and edges are elements of the surface dressing which are as much at risk as are the more frequently tracked bands of the dressing. It foflows that some enrichment, but not fatting up, may have to be accepted in the wheel tracks of channelised traffic if sufficient binder is to be provided for the dressing,asa whole. Blended binders wetted and retained chippings very effectively: possible exceptions were those used at KSD 4/5, which were the first blends attempted in the programme and the required viscosities were not attained. Differences in binder viscosity do not appear to have affected significantly the performances of comparable sections of surface dressing; however two blends which had viscosities at road temperature in the lower hdf of the permissible range (Figure 8) were very effective in wetting ~and retaining chippings. On no occasion were chippings lost due to low binder viscosity, although for safety reasons it was considered inadvisable to risk such a situation. 9. CONCLUSIONS The trirdsections of surface dressing constructed in Kenya in 1976–77 have now been trafficked for some 3 years. The fo~owing main conclusions can be drawn from their performance up to October 1979: 1. 2. 3. 20 The Jackson design procedure5 appears to provide a correct rate of spread of binder for Kenyan conditions for chippings which have an ALD in the range 5 to 13 mm. (This is the ALD range of the chippings used in the trials and it is adequate for most reseating operations in Kenya.) Athough not prescribed in the Jackson design method it is considered that, in general, the design ROS for binder should be increased by 10–15 per cent for fast or down~ traffic conditions and decreased by 10 per cent for slow or,up~ conditions. Men applying the Jackson design method it is imperative that au other conditions necessary for good surface dressing are also met, including the selection of correct binder viscosity, rapid and complete apphcation of chippings, good plant operation and job control. Without these provisions the use of any desi~ method is of little relevance. TABLE 9 Experimental surface dressings, Kenya; summarised observations October 1979 Site KSD 1 Trial type ~pping size Traffic catesory Hght Site characteristics bden lmries on Nairobi bound lane, partly uphill. Summarisedobaewations S0/100 pen, ody, used: for% in (20 m) chippins D to D+l 5 aweptable; D–10 is ctiticd downhill but acceptable up~. For %in (14 mm) cbippins, D to Dt10 aweptable: D+20 excessivebut tolerated. %in (6 mm) chipping, D+20 acwptable or sli*tly excessive. KSD 2 KSD 3 Bindertiscosity ~uble seal; ‘spht’binder apphcation tight As KSD 1 For Mh (20 mm) cfdppinss D–1Oprobably critiml, but rates up to Dtl 5 tolerated. For %in(14 mm) ctippinp D k correct or evensfightlyexcessive; D–20 not tolerated. BleOdedbindem.rm~e 3@40 nen. gavegood wetting and retentimr. Ught As KSD 1 Improvedappearance without 10SSof texture when 1stsed is trafficked for 14 days before applyins 2nd seal. NOadvantase apparent k dlspmpoflionate sptitting of total binder qumtity; finadtisable when mad already in use tis a tis new cOmtruction). DesignROSappears correct. KSD4/5 KSD 8 KSD9 KSD 10 Bindertiwosity md chipping ske BinderWscosity Bindertiscoaity Emulsion trial buble seal (new construction) Dnuble sed (new construction) Medium Vehiclemunt not high but site mrries very hea~ ades, includlng10r~/trafler units, en mute to Uganda. Pavementfdure developedin 16 months after hying tfids. Concluded that binder viscositycmdd be critial at this site (required ticosity not atttied). (SasentiaUy mdy in 1S0-200 pen rinse.) Cmrcludedthat: O) for Min (20 mm) cbippinss, ROS belnw ‘D wmdd not be tolerated, in etiting pOOradhesion conditions. Dt10 would bsve been benefickl but rates greater thm this wmdd lead to exceaaivetexture 10aa. ~i) D–20 tolerated for% in (14 m) chippinss and would hsve been correct for UPMU,S1OWlorry conditions. OverallD to D–10 acceptable. (i) 3 in (1Om) chippingaunsuitable for site cmrditionsbut D–20 would hve been adequate (Site abandoned 1978 due to extensive damase and reptim). Binders 178 and 317 pen, and 225 sec at 40°C (ST~; au wetted and held cbippkga well. D+20 appearsto be tolerated by dOwnhiUtraffic and D–20 by upbiUtraffic (only %in (14 m) ctippinga wed). Concluded tht D–1Owmdd be correct for upbiu and D for leveltraffic. Blendedbindersin ranse 200–400 pen, PIUSrme 600 pen. AUwetted md rettined cbippinss well. Concludedtht: for %in (20 m) cbippin~; D-20 probably tolerated in hea~ traffic environment, givenSood wetttis and traffic control. Good performanceWtb ‘D over 3 yeara. Dt10 beneficialfor dow~, m D–1Ofor UPMUtraffic. For %in (14 mm) chippkgs; over-spraysup to 30 per cent apparently tolerated. Cmrsideredto be due possibly to accommodation of understied material, in binder filmwith largerstmre ako held above it. NOfirm conclusions; problem of ‘drtinage’prior to bre& during constmction very apparent. Designrate, D, substantiated; 10 per cent reduction for clibins traffic appews correct. Enrichment in wheel tracka not taken to represent binder excessin rest of lane. The effect of constmction traffic may interfere tith interpretation of results; also designbased on unrepresentative samplesof cbippinss. Result of above is probable cause of apparent over.designof 10 per cent. Moreover,at one sectimr, wuffins wsa occurrhg in untmfficked portions on bend; illustrates need to consider whole width, not merely wheel tracks. In Senerd, experiment shows a‘ tolerance of D * 10 per cent. D i 10 ~r cent ap~am to be satiafacto~. PO*bkty of multiple layer of a~re~te formed due to large amount of understie material mrddust. Blendedbinders (21S, 373 and 506 pen) wetted a~regate effectively. (Note: agsregate hoseddown and appfiedstiUwet.) Medium/Hea~ Gradient overmost of site. Mtily caraand Ught/med goodsvehicles. Medium/Hea~ Undulatingmtin road (NairObi–MOmbma). Hea~ vebicle”(+trtier) combinatimrs. Short trial length (800m) ftirly level. Steep escqment for most of length. Medium/Hea~ KSD 12 tight KSD 13 HiUyregiOn; roadwith continuous chmges in vefiicd and horizontal geomet~. Ught (orvery fight) hvel, strtight road in cowtd belt. KSD 14 Bindertiwosity Medium/Hea~ KSD 15 ~pping size t double seal Medium/Hea~ Mtinly stmight main mad with wg curve. hublesed clearly necesw~to aedbadly crazedsurface. Calculated ROSfOr dOubleseals cOrrectifuphiU/dOwnhiUtraffic W0wedf0r. F0r%in(14m) cbippinga,atOlemnw 0fDf20(cf KSD14)appears t00prate. FOrMin(20m) chippings,tOleranceappearstObe D+lOtO D-20. (Note: cbippinssfrom ame source as for KSD 14.) 21 4. Binders with a viscosity at road temperature in the range 50,000–500,000 centistokes wetted chippings effectively; where excessively dusty cMpptigs occurred itproved to be beneficial to wet stock pfles thorougtiy with water, provided that good drying conditions prevded during the spraying and chipping operations. 5. The Jackson design method appears to apply equa~y weflto double seals as to single seals. In spite of poor qutity %in(14 mm) chippings, double seals employing 1Ain plus %in chippings are performing very we~: simdarly %in (20 mm) chippings foflowed by sand as second cover material have produced we~-grouted stable seals. Experience indicates that the performance of a double serdis improved by delaying the app~cation of the second seal by at least 14 days after the first seal, and Wowing traffic to use the first during this period. 6. It is recommended that each layer of a double seal is sprayed at its design ROS, rather than adopting an arbitrary ‘spfit’of the total binder required. 7. When any existing surface with a significant degree of surface texture is being resealed, use of the Jackson design method wfl require it to be rated as ‘very lean bituminous’. 10. ACKNOWLEDGEMENTS The work described in this report forms part of the research programme of the OverseasUnit (Unit Head: ~ J N Buhnan) of the Transport and Road Research Laboratory. The research was undertaken with the support and cooperation of the Ministry of Transport and Communications, Repubtic of Kenya. The cooperation and encouragement shown at dl levelsin the Roads Branch, is gratefu~y acknowledged and the assistance of the Maintenance Branch, the Materials Branch, and the Staff Training Department is much appreciated. 1. 2. 3. 4. 5. 11. REFERENCES TRANSPORT AND ROAD RESEARCH LABORATORY. Recommendations for road surface dressing. Department of the Environment, Road Note No. 39. London, 1972 (H M Stationery Office). WILSON, AH. The distribution of temperatures in experimental pavements at Nconbury by-pass. Department of the Environment, TRRL Report LR 719. Crowthorne, 1976 (Transport and Road Research Laboratory). HANSON,FM. Bituminous surface treatments of rural highways. Proceedings, New Zealand Society of CivilEngineers, Vol 21, 1934–35. THE ASPHALT INSTITUTE. Asphalt surface treatments and asphalt penetration macadam. Manual Series No 13 (MS–13) CoUege Park, Maryland, 1965 (The Asphalt Institute). JACKSON, G P. Surface dressing. bndon, 1963 (She~ International Petroleum Co Ltd.). 22 6. BMTISH STANDARDS INSTITUTION. Sin~e sized roadstone and chippings. British Standard 63. Part 2 (Metric Units). hndon, 1971 (British Standards kstitution). 7. BMTISH STANDARDS INSTITUTION. Sin#e sized roadstone and chippings. British Standard 63. Part 1 (Imperial Units). bndon, 1951 (British Standards Institution). 8. BWTISH STANDARDS INSTITUTION. Methods for samphg and testing of mineral a~regates, sands and fflers. British Standard 812: 1975. bndon, 1975 (British Standards Institution). 9. INSTITUTE OF PETROLEUM. I P Standards for petroleum and its products. Part I Methods for analysis and resring 39rh edition. hndon, 1980 (The Institute of Petroleum). 10. ROAD RESEARCH LABOWTORY. Instructions for using the portable stid resistance tester. Minisr~ of Transport, Road Nore 27. London, 1969 (H M Stationery Office). 11. WWGHT, N. Surface dressing; assessment of road surface hardness. Department of rhe Environment Department of Transport, TRRL Report SR 573. Crowthorne, 1981 (Transport and Road Research hboratory). 12. NATIONAL INSTITUTE FOR ROAD RESEARCH. Bituminous surface treatments for newly constructed rural roads. TRH 3.1971. fietoria (National hstitute for Road Research, Councfl for Scientific and kdustrial Research). 23 . 107 106 105 $ ~ 8 0 .9 > 104 /’” 103 102 Per cent diesel Pen at 25°C Nil 80 4 246 6 292 8 385 10 604 12 225s STV at 40°C \ b \ \ 30 35 40 50 60 70 80 90 100 Temperature (OC) Fig. 1 Blends of 80/100 pen+ diesel fuel UA for Sutia= dressing : viscosi~/temperature irelationships ---.... 0 00 0000 00 00: o 0 90( 80( 70C 60C 50( 40C 30C 20C smm wm c o.-+~ g 100 2 90 80 70 60 5a 40 30 2a 10 \ \ ,\ \ \\ \ \ ● ✼ ✼ ✼ ✼ ✼ ✼ Q ● 1 I I 1 I I I 1 0 4 8 12 16 20 24 28 32 36 Fig. 3 Rata of curing trials r~overed binder . . Exposure period (months) blends of 80/100 pen+ diesel fuel. Penetration tests on 900 800 700 600 500 400 300 200 100 90 80 70 60 50 40 30 20 10 . . \ \ ,\ \ \\.\ \ \\ \ .\ .\ \ \ \ ● ✼ 1 \ \\ \ \ \. \\ \ \ ● \ \\ \ .?, ● \\’* 8% o 4 8 12 1, ExposuI Fig. 4 Rate of curing trials : blends of [ recovered binder Kerosene % v/v *--* 12 W4 8 A. —d 4 ~o ! I 1 t 1 20 24 28 32 36 eriod (months) ‘1OO~n + kerosene. Penetration tests on A 6- 7 - 8 - 9 — 10 - z & ,“ - w .-N - c 12 - .-m U s ,3 _ 14 - 15 - 16 - 17 - 18 — B Method: Join A to C Read average least dimension on B c Fig. 5 Determination of average least dimension Chipping application rat 45 50 60 70 80 901 I I I I 1 I I 1 I I # 27 25 23 21 lg 17 15 1: I 1 I 1 I I I , I 1 I I I 1 r 20 19 18 17 16 15 14.13 12 11 10 0.6 0.7 0.80.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1 i I I 1 I 1 I I 9876 5 4 3.5 Binder application rate Fig. 6 Sutia~ dr 9 8 7 6 5 4 litre/m2~ / ( r / / . 19 18 / 17 / 16 15 14 13 12 11 10 \ 9 8 7 6 5 4 ‘ B3 1.7 1.8 1.9 2.0 2.1 2.2 kg/m2 I I 3 2.5yd2/gal ;sing design chati (in) 1 7/8 3/4 5f8 1f2 3/8 1f4 ● ✍✍✍ ● 00 ● ● ● ● 0 ● ● O ● ● 0 ● ● ● ● ● @ ● , ● ● ● 0 ● I ● Maximum and minimum road temperatures at which binders of the nominal grades shown were used I ❑ ✚✚✚✚✚✚✚✚✚✚✚✚ ❞✎✛ ❑ W Too stiff (poor wetting) ,,,,,,,,,,,,,, ✎✌.,, ::::::::::::::: ✎✌✎✌✎✌✎✌✎✌✎✌✎.,,,,,.,.,,, Too fluid (drainage and whip-off) .............. ., 2 105 8 6 4 2 2 103 8 6 4 2 4 --- Light traffic T----- 100 10 20 30 40 50 60 70 80 90 (“c) I 1 1 I I 1 1 50 77 100 122 140 160 180 (°F ) Surface temperature Fig. 8 Surfaw temperature/choice of binder for surface dressing omouo w * m 0 m w w m 0 (mu) uo!leJlauad aqo~d Plate 1 Preparing blended binder (80/1 OOpen and diesel fuel loaded simultaneously) Plate 2 Sand patch test Plate 3 TR R L Probe Test for assessing road hardness (Inset: probe tip) 12. APPENDIX SURFACE DRESSING DESIGN, JACKSON METHOD: A SUMMARY This Summary is based on the boo~ets ‘Surface Dressing’by G P Jackson (SheU International Petroleum Co Ltd). 1963. 1. Determination of average least tiension (ALD). Method A: Measure the least dimension of approximately 200 representative ctippings. Method B: i) Mot the sieve analysis for chipping sample on a grading chart. Read off theoretical sieve size through which 50 per cent WMpass; this is the median size. ii) Determine the Flakiness Index (BS 812: 19758). iii) Apply O) and (ii) to Figure 5. 2. Overd design factor to determine rate of spread for bitumen. i) Traffic Select from the fo~owing four tables a factor which, in each case characterises the site to be treated. (TRRL note: Until there is evidence of the effect of different tie loadings on surface dressings,the parameter %ehiclesper day’ should be interpreted literally without regard to commercial vehicles necessarily, since these wfl constitute a proportion of traffic on rdmost a~ roads. The presence of abnormal proportions of heavy vehicles, however, would clearly indicate a bias towards a heavier traffic category than that indicated by the gross vehicle count.) Very fight tight Medium Medium heavy Heavy Very heavy Veh/day o– 100 100– 500 500–1000 1000–3000 3000–6000 6000+ Existing surface Untreated/primed base 6 Very lean bituminous t4 Lean bituminous o Averagebituminous –1 Very rich bituminous –3 36 Constant t3 tl o –1 –3 –5 Type of chippings Round/dusty t2 tibical o flaky –2 Pre-coated –2 Ctimatic conditions North European (wet and cold) t2 Tropical (wet and hot) ts Temperate o Semi-arid (dry and hot) –1 Arid (very dry and very hot) –2 ii) Sum the four factors to obtain an overa~ factor. iii) Entering Figure 6 detefie the intercept of ALD and factor tine corresponding to (ii) above. A vertical dropped to the basefine ~ give the design rate of spread for binder. (TRRL note: Jackson recomends that, since the rates obtained from Figure 6 relate to the use of cut-back bitumens of the MC/RC800 or 3000 type or equivalent, when penetration grade bitumen is used rates should be reduced by approximately 0.1 kg/m2. It should be noted that this reduction was not made during the experimental work in Kenya.) iv) Determine the intercept of ALD and the he AB; a vertical tine drawn from this point to the 3 scales at the top of the chart wiflindicate the rate of apphcation for chippings, based on a loose density of 1.35. An a~owance of 10 per cent has been incorporated to a~ow for whipoff. (TRRL note: Since, with present day equipment the rate of spread for chippings cannot be pre-set quantitatively, these scalesshould be regarded as a guide when cdculattig quantities of stone required. The correct rate of spread wiflinvariably be a matter of adjustment on site. Equtiy, it is stressed that attempts to spread chippings at pre-cdculated excessive rates is pointless and constitutes a wasteful and dangerous practice.) 37 (1833) Dd8041301 1,400 10/81 HPLtd So’ton G1915 PRINTED IN ENGLAND ABSTRACT Surfam dressing in developing muntries: research in Kenya: L S HITCH: Department of the Environment Department of Transport, TRRL Laboratory Report 1019: Crowthorne, 1981 (Transport and Road Research Laboratory). In many developing countries surface dressing is used as a running surface on new low-cost roads as well as a maintenance treatment for existing roads. These countries are usually characterised by combinations of road surface condition, traffic, climate, and materials, that are not found in Britain and that are not favorable for good surface dressing. In such cases the recommendations for road surface dressing given in Road Note 39 which is intended for use in Britain, are not applicable. Jackson (1963) has proposed a method of surface dressing design based partly upon ideas put forward by Hanson (1935). This method was used to design a total length of 27 km of experimental sections of surface dressing in Kenya during 1975–77. This report describes the construction, monitoring and performance of the sections during the first 3–4 years trafficking. It is concluded that Jackson’s design method works satisfactorily in Kenya conditions over the range of chipping size used in these trials, namely 5 mm to 13 mm average least dimension. ISSN 0305–1293 ABSTRACT Surfam dressing in developing countries: research in Kenya: L S HITCH: Department of the Environment Department of Transport, TRRL Laboratory Report 1019: Crowthorne, 1981 (Transport and Road Research Laboratory). In many developing countries surface dressing is used as a running surface on new low-cost roads as well as a maintenance treatment for existing roads. These countries are usually characterised by combinations of road surface condition, traffic, climate, and materials, that are not found in Britain and that are not favorable for good surface dressing. In such cases the recommendations for road surface dressing given in Road Note 39 which is intended for use in Britain, are not applicable. Jackson (1963) has proposed a method of surface dressing design based partly upon ideas put forward by Hanson (1935). This method was used to design a total length of 27 km of experimental sections of surface dressing in Kenya during 1975–77. This report describes the construction, monitoring and performance of the sections during the first 3–4 years trafficking. It is concluded that Jackson’s design method works satisfactorily in Kenya conditions over the range of chipping size used in these trials, namely 5 mm to 13 mm average least dimension. ISSN 0305–1293