ABO~ ~L

The Transport Research Laboratory is an executive agency of the Department ofTransport. It provides technical help and advice based on researchto enable theGovernment to set standards for highway and vehicle design, to formulatepolicies on road safety, transport and the environment, and to encourage goodtraffic engineering practice.

TRLalso sells its services, acting as contractor, consultant or providing facilitiesandstaff ona fee-paying basis for customers in the private sector.

TRUs expertise ranges from the constructionof highways, bridges andtunnels, toall forms of road safety, traffic control and driver behaviour.

For instance, highways and structures research at TRL develops improvedmaterials and methods which enable earthworks, roads and bridges to bedesigned, built and maintained more cost-effectively. New ways of reinforcingearth can cut construction costs, while bridges can be modified to reducecorrosion. Road surfaces developed at TRL can reduce noise and cut wet-weatherroad spray from lorries by 90 per cent.

Safety research varies from monitoring the incidence of drinking and driving anddevising ways of reducing it, to improving junction designs and cooperating withEuropean partners on new standards for improved impact protection in vehicles.

Traffic research seeks to make the most of existing roads by, for instance,improving traffic signal coordination and devising systems which help driversavoid congestion. Other research looks at the effectiveness of parking controlsand improved crossings for pedestrians.

TRL research also informs Government transport policy bythe effects of bus deregulation and how land use interactstransport system.

studying, for example,with- the road and rail

TRL employs around 600 scientists, engineers and support staff. Its headquartersare at Crowthorne, Berkshire where its facilities include a 3.8km test track, astructures hall where bridge structures can be stressed to breaking point, a facilityfor carrying out accelerated tests on road structures, and advanced computersystems which are used to develop traffic control programs. TRL Scotland issituated in Livingston, near Edinburgh, where the staff are concerned mainlywith research and advisory work in the fields of ground engineering, bridges androad pavements. This unit has responsibility for all TRL work in Scotland.

A large proportion of the research is sub-contracted to industry, consultants anduniversities. The Laboratory also collaborates with local authorities and otherorganisations within Europe and elsewhere. In addition, TRL expertise isprovided to developing countries as part of Britain’s overseas aid programme.

For more information: TRL Public Relations, 0344770587

Old Wokingham RoadCrowthorneBerkshireRG11 6AU

OverseasDevelopment Administration94 Victoria Street

LondonSW1E 5JL

RESEARCH REPORT 381

CONCRETE PAVEMENT TRIALS IN ZIMBABWE

By J D Parry, N C Hewitt and T E Jones

Main subject area: Concrete Pavement Research

Theme: Urbanisation~ranspofi

Project title: PC Concrete Pavements

Project reference: ODN3.4

This document is an output from andeveloping countries.

ODA-funded research project, carried out for the benefit of

Crown copyright 1993. The views expressed in this repoti are not necessarily those of theOverseas Development Administration or the Department of Transport.

Overseas Centre, TRL, 1993ISSN 02665247

CONTENTS

Abstract

1.

2.

3.

4.

5.

6.

7.

8.

9.

Introduction

Pavement design

2.1 Joint design

2.2 Concrete mix design

Construction

3.1 Sub-base and subgrade matetials

3.2 Concrete road-base

3.3 Concrete characteristics

3.4 Concrete slab thickness

Post construction condition monitoring

4.1 Visual condition surveys

4.2 Joint condition

4.3 Road roughness

4.4 Skidding resistance and texture depth

4.5 Traffic loading and slab defects

4.6 Deflections

4.7 Warping

4.7.1 Slab temperatures

4.7.2 Warping measurements

4.7.3 Effects of warping

4.7.4 Slab length

4.8 Shoulders

Maintenance

Summary and Recommendations

Conclusions

Acknowledgements

References

Appendix A: Figures Al to A6Tables Al to A17

Appendix B: Cost analysis of concretepavements in Zimbabwe

Tables A18 and A19

Page

1

1

1

2

2

3

3

5

6

7

7

7

8

8

8

12

15

20

20

20

26

27

27

27

28

28

29

29

3036

45

46

CONCRETE

ABSTRACT

PAVEMENT TRIALS IN ZIMBABWE

In 1985 the Overseas Unit of the Transport ResearchLaboratory assisted the Ministy of Transpofl inZimbabwe to design and construct four trial sections ofconcrete road pavement with a total length of 1.42kilometres. The objective was to demonstrate the viabilityof concrete pavements in a region where they are notused and to generate enough data to prepare suitablespecifications for the conditions common in the drierregions of Africa. This report describes the designs,summarises the condition surveys that have takenplace since the road was opened to traffic and drawsinitial conclusions concerning design and materialspecifications.

1. INTRODUCTION

Most countties in Africa have very little experience of PCconcrete road pavements. By tradition, paved roads havebeen surfaced with asphaltic concrete or bituminoussutiace dressings. However, over the past decadeseveral of these countries have shown interest in the PCconcrete alternative. The reasons for this are given as:

a)

b)

c)

Few African countries have indigenous oil stocksand so pay for bitumen with scarce foreign ex-change which must be earned by expotiing primarycommodities. Many of them have limestone andcoal with which to make cement and most havesome cement making capability.

It is preferable, strategically, to produce roadbinders within a country rather than depend uponother sources outside the control of government.

Concrete pavements require less maintenance thanbituminous pavements. If true, this is an importantadvantage, especially where roads are failingprematurely for lack of routine and periodic mainte-nance (World Bank, 1988).

Because of this growing interest in concrete as a pavingmaterial, four trial sections were constructed on theGweru-Mvuma road in Zimbabwe in 1985 as a jointventure between the Zimbabwe Ministry of Transport(MOT) and the UK Transpod Research Laboratory (TRL).The objectives of the collaborative project were:

●

●

to demonstrate the viability of concrete pavementsas a low maintenance alternative and

to develop the most appropriate designs for theclimate and traffic in Zimbabwe, which is typical oflarge areas in Africa.

The average annual rainfall for the construction site is670mm whilst the traffic loading on the trial sections is inthe order of 50,000 equivalent standard axles (ESAS) peryear.

This report provides a description of the test sections andtheir performance over the first six years showingroughness and skidding resistance to be almost un-changed while deflections and joint condition havedeteriorated. The effect of slab warping is also found tobe significant under these climatic conditions.

Overseas Unit Working Paper 210 contains a detaileddescription of the construction of these pavements; OUWP 236, 261, 264 and PR/OSC/025/93 contain details ofsubsequent condition surveys, which are summarisedand discussed in this report.

2. PAVEMENT DESIGN

In order to predict the traffic load that the trial sectionswould sustain during a nominal 20 year life, a 24 hour, 7day axle load survey was carried out in February 1985.This gave a two-way total for the week of 570 ESAS,using the equation from Rolt (1981)

( )4.5

ESA (for each axle) =axle load in kg

8,160

Growth rate was difficult to predict because the site forthe trial sections formed part of a road which was likely togenerate traffic on completion in 1989 when the final paflwas due to be upgraded from a single lane to a two-lanecarriageway. No growth would result in approximately0.6 million ESAS over 20 years. It was estimated thatthe improvements would probably bring this total to 1million. The axle load spectrum, as measured in 1985and during subsequent suweys, is shown in the Appendixin Table Al with damaging factors calculated from theaxle loads. Table A2 contains traffic counts made by thelocal District Office.

For the relatively low design life of one million ESAS,three internationally used design methods for determiningslab thickness (CPCA, 1984; AASHTO, 1975; RRL,1970) gave the results shown in Table 1 for three differ-ent sub-base strengths, chosen as typical for the localmaterials.

For the Zimbabwe trials, four designs were preparedusing three different thicknesses of concrete slab, twosub-bases and different joint designs, with the purposesof determining the most economical slab/sub-basecombination, and the most suitable joint design for thematerials, climate and traffic. Figure 1 shows the vertical

1

TABLE 1

Slab design thickness

CBR of TRL Road AASHTO PortlandSub-base Note 29 Pt 2.5 Cement

(mm) (mm) Association (mm)

5 160 180 17526 150 170 15037 140 160 140

Pt Terminal Serviceability Index.

and horizontal alignment and Table 2 summarises thedesigns of the trial sections. Each section was designedto be 400m in length and 7m wide, with 1.5m surfacedressed shoulders. Later changes and difficulties withsub-base levels caused some variations from Table 2,which are described below.

2.1 JOINT DESIGN

Expansion joints were not required because:

a) none of the test sections was butting against a fixedstructure and

b) construction took place during a relatively hotseason.

Contraction joints were specified at either 5 metreintervals, or a mean of 5 metres ranging from 4.5 to 5.5,

in order to compare the drumming effect from vehicletyres passing over regularly spaced and random spacedjoints. For the same reason, some joints were cutperpendicular to the centre line of the road and someskewed by 100. Aggregate interlock across the joints wasconsidered to be sufficient to prevent stepping under theanticipated traffic loading but the stronger section, C3,was given dowelled joints in order to determine whetherthis more rigid joint prolongs the life of joint seals.

The specification called for the contraction joints to besawn as soon as the concrete was set sufficiently not totear, and to be sealed with an elastomeric joint sealant.Possible alternatives included a hot poured bitumen,bitumen and sand mixture and, if narrow joints could becut, no sealant at all. This was to establish if a cheaperspecification could be justified for relatively light traffic.

Construction joints tied in the conventional manner withdeformed steel tie bars 16mm in diameter at 600mmcentres were specified for the end of each day’s work orwhen a delay occurred. The same tie bars were used at1m intervals at the warping joint between the two car-riageway lanes. No groove or sealant was specified fortied joints. Contraction and warping joints are of the typesshown in Figure 2.

2.2 CONCRETE MIX DESIGN

Slab thicknesses in Table 1 were calculated using aconcrete characteristic flexural strength of 4MPa, orcorresponding minimum compressive strength of 30MPafor Road Note 29 (RRL, 1970) and average compressivestrength of 35MPa for the AASHTO design (AASHTO,1974). This strength corresponded with Class 30 con-

~ clII C2

EASTk J

I 320m I 320m

~

I

I I

&I 380m I

I I

& WEST

I 400m I

I 0ri9inal ground profile

YI

i 1’I I

Road pr;file

Fig. 1 Horizontal and vertical alignment of the trial sections

2

TABLE 2

Test section design

Section Slab Sub-Base Subgrade Remarks Joints

Thickness(mm)

cl 150 150mm Class 3,3 Imported On 800mm 10° SKEW

LS 9370 Tar Prime 150mm SG 9 Fill @ 5m spacing

C2 125 150mm Class 3,3 Imported On 800mm SQUARELS 9370 Tar Prime 150mm SG 9 Fill @ 5m spacing

C3 175 NIL Imported On Im Fill SQUARE150mm SG 9 Dowelled @

Tar Prime 5m spacing

C4 125 NIL Imported On 3m cut. SQUARE150mm SG 9 Levelled and variable

Tar Prime compacted. spacing4.5m-5.5m

Note: Class 3.3 - Natural gravel, Texas triaxial class 3.3 maxLS 937. - MOD AASHTOSG9 - Natural gravel of minimum CBR 970

crete in the Zimbabwe specifications. Trial concretemixes were prepared in the Zimbabwe MOT Test Labora-tories using the proposed materials from sources nearthe site. These samples were tested for compressive andflexural strength at both the Zimbabwe laboratories andat TRL. Details are shown in the Appendix, Tables A3and A4. The mix resulting from the trials is shown inTable 3 with the propoflions actually laid.

3. CONSTRUCTION

The subgrades and sub-bases were prepared by theMinisty of Transport and the concrete was laid by a localcontractor. There was no previous experience of con-structing concrete pavements in the country, apati from afew hard standings, but this contractor had used concretein building dams and reservoirs.

3.1 SUB-BASE AND SUBGRADEMATERIALS

Sections Cl, C2 and C3 were built on low embankmentsbut section C4 was in cut (Figure 1) and it was foundwhile cutting the profile of section C4 that the subgradelevel at one end of the section was on very weak clayover a length of 80m. This material was removed to adepth of 400mm and was replaced with gravel taken fromthe ditches at the other end of the same section. Thiswas compacted and covered with approximately 150mmof imported sub-base material.

The sub-base material was generally well graded andconformed to the standard specified in almost everyrespect, Table A5. However, the fraction passing the 75micron sieve was at the top of the permitted limit, or justabove it. This and the entry of water between the shoul-der and concrete slabs, or through damaged joint seals,has led to clay fractions being evacuated from beneaththe concrete by downward or rocking movements ofsome of the slabs. This process is termed ‘pumping’.

The subgrade material used under the sub-bases onsections Cl and C2 and directly under the concrete onsections C3 and C4 was also well graded, apart from theoversize material which was mostly rejected by hand. Butthis material also had a high fraction passing the 75micron sieve and has been eroded in a similar manner onsections C3 and C4.

Day-time rocking of many concrete slabs shows that,although both of these materials exhibit high strength interms of in-situ CBR (Table 4), they have proved to beinadequate for use as sub-bases under a jointed concretepavement where water is permitted to enter the pave-ment structure. A limited amount of mud staining hasbeen recorded during the visual surveys on the shouldersof all the sections (Figures Al to A4) but is most apparentat the end of section C4 (Figure A4) at the lower side of asuper-elevated section where rain water has drainedacross the road and penetrated the gap between pave-ment and shoulder as well as entering by damagedtransverse joint seals. Plate 1 shows the most pro-nounced example of mud staining on the shoulder due tofine material being pumped up through damaged joints

3

Contraction joints on Sections Cl, C2, & C4 without dowels

Bitumen de-bondng coat

/ I

1/01 ~\

/’ II \

(:Q

\

———— ——, .—— —— —— Q_)

Conkaction joints on Section C3 with dowels

Deformed tie bar

I

Warping joints on all sections

Fig. 2 Contraction joints and warping joints

TABLE 3

Concrete mix proportions

Aggregate CemenV Water/proportions (7.) aggregate cement

stone river pit (70) (70)sand sand

Recommended: 63 32 5 19 54

Laid: 63 32 5 22.4 47.564 29 7 22.7 45

4

and through the gap between the shoulder and theconcrete slabs. There is no apparent correlation betweenthe amount of pumping, according to stains on theshoulders, and the magnitude of the measured deflec-tions, but staining was more marked at the lower side ofall three super-elevated sections.

The loss of fine material from the sub-base results invoids under the concrete but is difficult to measurebecause of the effects of slab warping, which are dis-cussed below in Section 4.7. However, it is consideredthat this is one of the causes of the accelerated crackingof the slabs. Road Note 29 (RRL, 1970) states that nosub-base is required under concrete slabs for subgradesof CBR greater than 15 per cent. However, the current

TABLE 4

Sub-base and subgrade strength testing by DCP.

CBR “kSection Average Range

Cl Sub-base 52 33- 86

C2 Sub-base 72 36-125

C3 Subgrade 43 30- 70

C4 Subgrade 62 41-108

Departmental Standard, ‘Structural design of new roadpavements’ (DTp, 1987) specifies a cement-bound or‘lean concrete sub-base for all rigid pavements in order toprevent pumping from the sub-base and, in most cases,from the subgrade too.

3.2 CONCRETE ROAD-BASE

A water tower was set up on a site adjacent to the trialsections with stockpiles of the coarse and fine crushedstone, river sand and pit sand. These were fed through aweigh-hatcher to a half-cubic-metre mixer which dis-charged into a lorry-mounted skip. When five or six loadshad been discharged into the skip, it was driven to thelaying site and tipped between rolled steel channel formswhile a second skip was positioned to receive theconcrete from the mixer. Samples of the fresh concretewere taken three times a day for wet sieving analysis and

Plate 1 Pumping stains on the hard shoulder of test section C4

5

to fill cube and beam moulds. The rest was spreadmanually between the forms. Several vibrating pokersand two vibrating screeds were used to compact andlevel the concrete, which was then textured using apurpose-made broom and covered with wet hessian laidover wooden frames. After a few hours the frames wereremoved for use on fresh concrete and the hessian’waslaid directly on the surface. The following day the hessianwas replaced with sand, nominally 50mm deep, whichwas sprayed with water.

Most of the joints were cut with a diamond wheel within24 hours of laying. All were cut within 48 hours but someshrinkage cracks between the joints became apparentwithin days of laying; these are marked on the visualcondition repofls, Figures Al to A4.

When the contractor had left the site, the Ministry ofTransport constructed the shoulders with cement-stabilised natural gravel and sealed them with a bitumi-nous sutiace dressing.

3.3 CONCRETE CHARACTERISTICS

During the construction of the trial sections a record waskept of all mix proportions and the times between mixing,finishing and joint cutting. Notes were made of all inci-dents such as delays and breakdowns, and a record waskept of weather conditions, including air temperature andwind speed, as possible causes of variability in the qualityof the finished concrete.

The cube and beam specimens made at the laying siteswere tested at the Central Test Laborato~ in Harare. Theresults are listed in the Appendix in Tables A6 and A7.After the concrete construction was complete, cores werecut from the pavement for verification of concrete thick-ness, mixture quality and strength. These crushingstrengths are listed in Table A8. Wet sieving results areshown in Figure A5 with the recommended gradingenvelope for standard class 30 concrete in Zimbabwe.

The test results on cubes and beams made at the time ofconstruction show that in most cases the concrete meetsthe specification described in Section 2. The 90-dayequivalent cube strength of the cores cut from thepavement also indicate that the average strength was atleast as high as that specified, even taking into accountstrength gain from 28 to 90 days, which was in the orderof 1070.

The recommended cementiaggregate ratio resulting frommixture trials prior to construction was 1970 (Table 3).The contractor used a ratio varying from 22.470 to 22.70A.This resulted in higher strength but may also havecontributed to early shrinkage cracking by increasing theheat of hydration.

Plate 2 shows typical cores cut from section C3.Compaction is seen to be uniform and distribution ofaggregates is also good. The contractor achieved thesame concrete density in the pavement as in the samplesmade for laboratory testing.

Plate 2 Cores cut from test section C3

6

3.4 CONCRETE SLAB THICKNESS

The surface of the trial sections conforms closely to thedesign veflical alignment. However, the sub-base onsections Cl and C2, and the subgrade on sections C3and C4 were not finished as accurately. The thickness ofthe concrete slab is the difference between the finishedsurface and the sub-base level (or subgrade in the caseof C3 and C4) and varies according to the deviation ofthe actual sub-base level from the design profile. On allfour sections, levels were taken at 10 metre intervals atfive stations across the pavement, on the sub-base (orsubgrade) and on the finished concrete. These areshown in OU WP21 O. Table 5 summarises the readingsand shows that, instead of the three thicknesses of trialpavement as intended in the original design, there is avariation of slab thickness ranging from 80mm to 237mmover the total 1.42km. The variation in thickness ofindividual slabs is typically 15-20mm. This larger varia-tion, although not intended, will provide more scope toidentify the effects of thickness on the various modes ofdeterioration.

4. POST CONSTRUCTIONCONDITION MONITORING

4.1 VISUAL CONDITION SURVEYS

Each year since 1987 a visual condition survey has beenmade of the trial sections, in a similar manner to thatdescribed in the UK DTp Standard HD1 7/88 (1988). Onschematic maps of each section, defects such as crack-ing, joint damage and pumping stains on the shoulders,have been marked and later summarised in tables.Figures Al to A4 contain the condition assessmentsmade in 1992 and Tables A9, A1O and Al 1 show theprogression of deterioration from 1987 to 1992 of thejoints (see 4.2 below) and the slabs (see 4.5 below). Thecondition assessments in Tables A9, Al O and Al 1 aresubjective and consequently show some anomalies,particularly with joint sealant condition and transversecracking.

TABLE 5

Test sections - as built.

Section: c1 C2 C3 C4

Slab Thickness‘As built’ Average:

Range:

Design thickness:

Joint spacing:

Joint type:

Load transfer:

Sub-base:

Subgrade:

On Fill or Cut:

Chainage:

Length:

No. of slabs:

173 mm135-213 mm

150 mm

4.5 -5.5 m

10° Skew

150mmClass 3.3LS 937.

Imported150mm of

SG 9

On 800 mm Fill

17.480-17.800 km

320 m

128

124 mm80-189 mm

125 mm

5m

Square& 10° Skew

150mmClass 3.3LS 93%

Imported150mm of

SG 9

On 800 mm Fill

17.800-18.120 km

320 m

128

186 mm139-237 mm

175 mm

5m

Square

Dowels

Nil

Imported150mm of

SG 9

On 1 m Fill

18.580-18.960 km

380 m

152

151 mm96-233 mm

125 mm

4.5- 5.5 m

10° Skew

Nil

ImportedSG 9

+ Tar prime

On 3 m cut

19.360-19.760 km

400 m

163

Note: Class 3.3 - Natural gravel, Texas triaxial class 3.3 maxLS 93Y. - MOD AASHTOSG9 - Natural gravel of minimum CBR 970

4.2 JOINT CONDITION

The criteria used to judge sealant condition was:

GOOD - No visible damage

FAIR - Some loss of adhesion but stilleffective

BAD - The joint allows water to pass through tothe sub-base or subgrade.

Transverse cracking close to joints can be classified asjoint spalling when further deterioration takes place.

Table Al O shows most joints to be properly sealed in1988, the worst section, C4, having 270 of joints in badcondition and 24% showing signs of aging but not yetletting in water. However, 350 metres of joint sealant wasreplaced in 1987. Apparently the locally manufacturedmaterial was applied properly in some areas but not inothers. This figure includes 70 metres of bitumen/sandfiller, which was not successful and was replaced.

Joints in concrete pavements that are built on founda-tions with little susceptibility to water damage need not berepaired until they deteriorate beyond the ‘Yai~ state to“bad”. The sub-base material under sections Cl and C2and the subgrade under sections C3 and C4 have bothbeen eroded by water penetrating the joints so the badand fair joints should be renewed. It would also beeconomical to reseal the joints currently in good conditionat the same time, because their remaining life is verylikely to be shoti after seven years of service.

A design feature that increased the penetration of waterpast the damaged joint seals is the skew of the joints ontest sections Cl and C4. Plate 3 shows how waterrunning down the texture grooves is intercepted by theskewed joint and penetrates the joint far more easily thanit could if the joint were cut parallel to the texture grooves.Clearly, the more water that penetrates the joint, themore likely that damage will occur to the sub-base orsubgrade material.

It is arguable that very little water could penetrate poorlysealed joints that are installed on pavements parallel topronounced drainage grooves, as shown in the upperphotograph of Plate 3, irrespective of whether thegrooves and joints are set perpendicular or skew to thecentre line. On gradients, the grooves and joints would bebetter set skew in order to follow the line of greatestslope.

4.3 ROAD ROUGHNESS

Road roughness was measured once a year in eachwheelpath using the TRL Profile Beam (Abaynayaka,1984). The results, averaging near-side and offsidemeasurements, are given in Table A12 and illustrated inFigure 3.

Although road roughness, or riding quality, is only oneindicator of pavement condition, an increase of rough-

ness with time is an important indicator of pavementdeterioration (Smith et al 1980). On concrete pavementsthis increase is caused by uneven settlement of slabs orbroken parts of slabs, faulting at joints, or severe spallingat joints or cracks. The roughness measurements to-dateindicate no serious deterioration, although a few cases offaulting are recorded on trial section C4 (Figure A4).

Slab warping, which is discussed in more detail inSection 4.7, has a small effect on the longitudinal profile.However, a jointed pavement with slabs 5m long i.e. 200slabs per kilometre, each warping by approximately1.Omm during the da~ime hours of measurement, wouldregister a maximum roughness change of about 200mmper kilometre, which is little more than the variability ofmeasurement.

4.4 SKIDDING RESISTANCE ANDTEXTURE DEPTH

On surface dressed and asphaltic concrete surfaces,texture depth frequently affects the resistance to skid-ding. This is not the case on Portland cement concretesurfaces with texture of the type used on the trial sectionsand illustrated in several of the plates. Skidding resist-ance is primarily dependent on the micro-texture of thesurface (Franklin, 1988), which in this case is formed bythe sand particles in the concrete. The macro texture isdominated by the drainage grooves but on this type ofsurface, macro texture only affects skidding resistanceunder conditions of heavy rain.

Skidding resistance was measured using the TRLskidding resistance pendulum according to Road Note 27(Road Research Laboratoy, 1969). Measurements weremade in both wheelpaths at three fixed locations in eachdirection on each trial section. Table Al 3 summarises theresults and Figure 4 illustrates the measured variations.

The results suggest that there has been a decrease inthe skidding resistance value over the years 1990 to1992. However, there is a considerable fluctuation in thereadings recorded from year to year, primatily due tochanges in operator of the skidding resistance testerrather than real changes due to traffic and climaticinfluence on the road surface. The uncharacteristicallyhigh readings for the year 1990 appear to exaggerate thisdecrease. Road Note 27 (1969) recommends a minimumskidding resistance value of 45 for straight roads witheasy gradients and curves, and a value of at least 65 forgradients above 5%, which would include part of ttialsection C4. All the measurements recorded are in excessof these figures.

No reduction of texture depth, as measured by the sandpatch method, can be identified from Figure 5. This isbecause, although some of the micro texture may havebeen worn away, there has not been enough traffic onthe sections to expose the coarse aggregate or to weardown the material between the drainage grooves. Thesegrooves, shown in Plate 4, were set at an average pitchof 25mm and vary in depth from O to 7mm, averaging

8

Plate 3 Square and skew joints in relation to transversedrainage grooves

9

I ––—- Noflh side South side I

3000

I—-A

--- \\.- \

2500--- \

__ —- --- K c1—- \

2000

87 88 89 90 91 92

3000

I

-— --- ------------- - ----—--- -

2500 C2zz 2000g

avca5~ 3000E

2500 – _____ __ ------- --- C3

2000 –

I I I I I I87 88 89 90 91 92

3000

2500

20001

———— ——-— — ———— ———- - --------- C4

1 I I I I I I87 88 89 90 91 92

Year

Fig. 3 TRL profile beam roughness measurements

10

Texture depth (mm)o p

o m o b o

Shdding resistan= valueN * m w s

o 0 0 0 0 N Go 0 0

T

II

IIII

I

III

I

I I I

00

● 0

I

Plate 4 Sudace texture

about 3mm. Table Al 4 summarises the texture depthmeasurements made during each dry season.

Some concrete sutiaces have been criticised for tyrenoise, mostly on surfaces with pronounced texturegrooving, as on these test sections. No noise measure-ments have been made but subjective tests suggest thatthe tyre noise on the concrete text sections is less thanthat on the adjacent flexible test sections, which aresurface dressed with 19mm stone. It is thought thatrandomizing the pitch of the grooves, as shown inPlate 4, contributes to the reduction in noise levelcompared to that on pavements with regularly spacedgrooves. No change in tyre noise has been detectedsubjectively when driving from square to skew jointedsections or from slabs 5m long to slabs of randomlengths between 4.5 and 5.5m. This is probably becausethese differences are masked by the effects of the texturegrooves and because stepping is not prevalent.

4.5 TRAFFIC LOADING AND SLABDEFECTS

Traffic using the trial sections has increased since thecompletion of the upgraded road link in 1989. Thepresent trend with modest further increases would bringthe 20 year traffic loading close to 1 million ESAS, asestimated at the design stage.

The concrete was of good quality (section 3.3) anddefects were few. Of the 61 metres of transverse crack-

ing recorded in Table A9 during the first survey in 1987,most were due to shrinkage immediately after construc-tion, while the concrete was still weak. This level of defectis not unusual for 2.84 lane kilometres of concreteconstruction, considering the severe drop in ambienttemperature during the night at that season of the year.

Tables Al and A2 illustrate the spectrum of traffic usingthe test sections. The range of axle loads is typical for thecountry with a small but significant number exceeding 10tonnes (the legal axle load limit is 8.2 tonnes). Table Alincludes an estimated additional loading during 1987 and1988 due to vehicles carrying stone from a quarry to aroad construction site. These lorries were granted adispensation to exceed the legal limit and they wereknown to have axle loads in the region of 14 tonnes butwere not traveling during the period of any suweys. Theypassed over the north lanes of the test sections in a fullyloaded condition and are thought to have caused twotransverse cracks on section C2 and one on section C4.At these locations the slabs were 11Omm and 115mmthick (slabs C2 16 and 18 and slab C422 in Figures A2and A4).

The UK Department of Transport Advice Note HA35187(1987) states “The failure condition (of a jointed concretepavement) represents the end of the initial serviceablelife, and the point where the rate of cracking begins toincrease rapidly. At that stage the life of the pavementmight be extended by suitable strengthening measuresrather than excavation and complete reconstruction.”

12

The failure condition of individual slabs is described as acrack across the width or length at least 0.5mm wideaccompanied by a partial loss of interlock, or more than200mm of cracking at least 1.3mm wide. Up to 3070 ofslabs can be expected to have reached the failurecondition during the design life of the pavement.

The last column in Table A9 shows the number of “failedslabs according to the above definition. Most of thisdamage occurred before the road was opened to trafficand is in the form of transverse cracking. However, theTable also indicates that further damage has beenregistered in the 1992 survey and repairs are required toseveral slabs as detailed in Chapter 5 below.

It should be emphasised that the “failed slabs noted inTable A9 in no way affect the vehicles using this road.However, there are defects connected with the joints, i.e.

spalling and stepping, that also require attention,(Table Al 1). Plate 5 shows a joint with spalling problemsfrom the time of construction. The type of repair requiredis referred to in the DTp advice note HA35/87 (1987) as“arris or thin bonded repair.” Where transverse crackscompletely cross a slab close to the joint, full depthrepairs are appropriate. Where a narrow transverse crackcrosses in the central region of a slab without stepping,no repair is required.

Plate 6 illustrates shallow plastic cracking caused bydrying winds blowing over the wet concrete before thehessian was placed to cover it. This cracking is notstructural. Plate 7 shows the small corner cracking thatoccurred mainly on section C3 and Plate 8 shows atypical transverse crack; this one on section Cl wascaused by shrinkage before traffic loading stafled.

Plate 5 Spalled joint

13

Plate 6 Plastic cracking on test section C3

Plate 7 Corner cracking on test section C3

14

Plate 8 Transverse crack on test section Cl

4.6 DEFLECTIONS

Starting in 1988, measurements of deflection were madeusing Benkelman beams and a lorry with a rear axle loadof 6.36 tonnes. The aims were to measure interlockacross the joints, compare deflections from year to yearas a measure of pavement condition, and to detect slabrocking.

Figure 6 shows the initial position of the beams and lorryrear axle before it was driven forward. Records weremade of the initial gauge reading, the maximum reading,and the final reading after the lorry had passed at leasttwo slabs beyond the joint being examined.

The Benkelman beam readings are influenced by acomplex combination of concrete flexural strength, sub-base compressive strength, interlock across the joint and,in the case of the off-side wheel, the influence of thecentral warping joint. Furthermore, when used in thetraditional manner, the feet of the beams in the startingposition rest on the concrete slab well within the influenceof the lorry rear axle. This makes the calculation of actualdeflections quite difficult since corrections are necessary.

Benkelman beams are too long to be used convenientlywith their measuring probes resting on the corners of theslabs and with feet resting on the shoulder outside theinfluence of the lorry axle. In order to eliminate theinfluence of the lorry axle on the feet of the measuringinstrument, a deflection logger was developed and built in

the TRL workshops (Plate 9). This instrument uses ashotier arm and two linear transducers to measurevertical movement each side of a joint at the slab edge.Instead of the three readings recorded manually, (initial,maximum and final) the logger records fifty readings fromeach transducer over a set time. These may be printedlater andlor plot?ed to show curves of deflections on eachside of the joint under examination. Figure A6 is a typicalprint and plot of readings each side of a joint.

A summay of the vertical deflections measured by theBenkelman beams and the logger over successivesurveys is shown in Table A15 and plotted in Figure 7 forone section. Figure 8 shows a summary of all the loggerreadings taken in 1992 on trial section Cl south side. Theplot shows maximum minus final values of vetiicaldisplacement, V, on both sides of each joint. Where thepoints VI and V, are close e.g. at the right side of thegraph, interlock is still good. The loss of interlock at theother end may be due to a number of factors, includinglow temperature at the time of measurement and sub-base erosion. There is also a plot of VI - V, on eachlogger plot (Figure A6).

There appears to be a correlation in Figure 7 betweenmeasurements made by the verge side Benkelman beamand the logger but more data are required before this canbe defined with confidence, because the deflection is acombination of bending and rocking which affect the twoinstruments in different ways. Furthermore, there are bothsimilarities and disparities between repeated measure-

15

deflection logger

-——=

f—2

-–1I

I‘[I—— -

[

——— ———I

I III II v,

Benkelmanbeams

Fig. 6 Instrument positions for deflection measurements

ments made in 1992 on the same sections using theBenkelman beams (Figures 9 and 10). Examination ofFigures 8, 9 and 10 led to the conclusion that weatherconditions affect the measurements of deflections morethan had been anticipated. This is discussed below inSection 4.7.

The work to date has revealed no change of deflection orof aggregate interlock between the annual suweysbecause the effects of linear expansion and warping, asdiscussed below, have masked any changes that haveoccurred. Attempts to match conditions of measurementfrom year to year, by repeating the measurements during

16

the same month and hour of the day, have been frus-trated by the strong influences of cloud cover and wind.The repeat measurements made in June and July 1992shown in Table A15 and Figures 9 and 10 illustratea) that average deflection measurements can be accu-rately reproduced under similar conditions and b) thedisruptive effect on deflections of variations in radiationfrom the sun. Although it is expected that deflectionmeasurements will increase with time due to erosion ofthe material beneath the slab edges, the averages of allreadings taken each year show little evidence of this todate.

Plate 9a The TRL Deflection Logger

Plate 9b Close-up showing two vertical transducers and the gaugeto measure horizontal strain

17

— Benkbam ~N

Lqger UN

---- Benk bam ~S

—-— Lqger W

\

100

90

80

70

87 88 89 90 91 92 93

Year

Fig. 7 Average transient deflections - Section C4

Section Cl South July 92

— V1 Mm-Hnal --- V2 Ma-final — H M~-Min

b),\ \\

A AA“---zl ------A ‘4T,A-UA-42.A---:A:A AA A--- 2---- -+ .-A--- . AA A AA A A ‘--i--+ -44~li-X-f~i-4>--J-xdx4-~i-A!--A!-,

A I I IA

I 1 I A 1 I 1 I I

o 10 20 30 40

Fig. 8

50 60 70 80 90 100 110 120 130Slab number

Deflection logger readings

Section C2 North

.

1 I I I I I 1 1 I 1 I I

o 10 20 30 40 50 60 70 80 go 100 110 120 130Slab number

Fig. 9 Repeat beam deflection measurements. similar weather conditions

S=tion Cl South

I I I I I I I I I I I 1

0 10 20 30 40 50 60 70 80 go 100 110 120 130Slab number

Fig. 10 Repeat beam deflection measurements- different weather conditions

19

Recently a third channel has been added to the loggercircuits to record horizontal strain across a joint or crackin the pavement, measured by a modified Dmec gaugewith a linear transducer set horizontally. This is shownhand-held in Plate 9. No conclusions have been drawnyet from these readings although there may be somesignificance in the effect shown in Figure 8 where thehorizontal movement, H, is slightly larger on the left sideof the Figure where the interlock is poor i.e. dissimilarvertical deflections VI and V2 either side of the joints.

4.7 WARPING

4.7.1 Slab temperatures

The temperature at the top of a concrete road slab variesaccording to the weather and the time of day. The airtemperature and rain have some influence but the largesteffect is due to radiation, i.e. heat gain from the sun andheat loss to a clear sky at night. The temperatures atlower levels in the concrete vary less than at the top andfollow the trend set by the top surface. In order to meas-ure the effect of these temperature variations on theshape of the concrete slabs, and identify a connectionbetween warping and pavement deterioration, levelswere taken at nine points on the top surfaces of each offive slabs. The measurements were made on severaldays between sunrise and sunset with an optical leveland a rod set into dimples on discs glued to the slabs inthe positions shown in Figure 11. The height measure-ments and recording times for a typical slab are shown inTable Al 6. The temperatures at different depths in theconcrete were also recorded over several 24 hour

periods on two slabs of different thickness usingthermistors and an automatic recorder. Similar tempera-tures were recorded on both slabs. The results aresummarised in Figure 12 and Table A17. Detailed

[measurements of warping and temperature are listed inOverseas Unit Working Paper 289.

Figure 12 shows that over approximately eight hoursduring the day, the temperature is higher at the top of theslab than at the base; at night the reverse is true. Thisvariation in the temperature differential causes differentiallinear expansion between the top and the base surfaces.The effect of this is a tendency towards hogging in theday time and dishing at night, as illustrated in Figure 13.

Figure 12 also shows that the temperature at the topsurface is equal to that at the base at about 0900 and1730 hours on a typically clear day. Whatever the actualshape of the slab surface at these times, for the purposeof analysis the slab could be considered flat; the day-timehogging and night-time dishing would then be relative tothe actual shape at 0900 and 1730.

4.7.2 Warping measurements

Both transverse and longitudinal warping measurementscan be derived from analysis of the levels taken at thenine points on a slab. Figure 11 shows three transversepaths and three longitudinal paths across the surface of aslab. The height of each centre point relative to a straightline drawn between the two end points for each path areplotted for one slab, the transverse paths in Figure 14and the longitudinal paths in Figure 15, against time of

Fig. 11 Measuring points for slab warping

20

30

“— Top — — Md depth — Base

—

I I I I I I I I I 1 I

o 2 4 6 8 10 12 14 16 18 20 22 24

Time of day

Fig. 12 Clear sky temperatures of slab C4-163

Shoulder Shoulder

Night-time shape

Day-time shape

Fig. 13 Slab warping shapes

day. However, each concrete slab was cast with a slightlydifferent shape so, in order to compare the differentpaths, all the heights were measured with respect to thevalue at 0900, when the temperatures on the top andbottom surfaces were similar. Figure 14 shows that thethree transverse paths warp in a similar manner andFigure 15 shows for each of the three longitudinal paths alarger movement, but again all three paths are similar.

As the three transverse and the three longitudinal pathsare similar, each group was aggregated to compare thewarping on slabs of different length and different thick-ness. Figures 16 and 17 show very little difference in the

amount of transverse warp or longitudinal warp on a slab5m long and a slab 7.5m long. Figures 18 and” 19 showsimilar warping of two slabs, one approximately 100mmthick and the other approximately 200mm thick.

To this point the analysis has considered only day-timemovements. In order to calculate night-time effects, thetemperature difference between the top and base of theslabs was plotted against longitudinal warping on twoslabs 200mm thick, (Figure 20), and the resultant rela-tionship was used to predict longitudinal warping over a24 hour period. This predicted warping is illustrated inFigure 21 with measured results of day time warping of

21

1.5 I

1.0

0.5

0

0

1.5

1.0

0

-0.5

-1.0

-1.5

I — End — Centre -- End II 1

2 4 6 8 10 12 14 16 18 20 22 24

Tme of day

Fig. 14 Transverse warping of slab Cl-97

— Vergeside — Centre slab -- Centre roadI I

\

\/

o 2 4 6 8 10 12 14 16 18 20 22 24

Tme of day

Fig. 15 Longitudinal warping of slab Cl-97

22

1.5

1.0

-1.0

— C1-97(5m) -- C4-56 (7.5m) I

0 2 4 6 8 10 12 14 16 18 20 22 24

Tme of day

Fig. 16 Transverse warping of slabs of different lengths

— Cl-97 (5m) -- C4-56 (7.5m)

\\\\\\\\\\

-1.5

0 2 4 6 8 10 12 14 16 18 20 22 24

Tme of day

Fig. 17 Longitudinal warping of slabs of different lengths

23

— Cl-97 (200mm) --- C2-11 (100mm)

o 2 4

Fig. 18

6 8 10 14 16 18 20 22 24Tm~~f day

Transverse warping of slabs of different thickness

— Cl-97 (200mm) --- C2-11 (100mm)

\\\

\\\\\\\

\

\\

\

o 2 4 6 8 10 12 14 16 18 20 22 24

Tme of day

Fig. 19 Longitudinal warping of slabs of different thickness

24

200mm thick slsbs

y=o.119x +o.oo9x2–o.171(r2 correlation value= 0.93)

.7 I I I I I I I I 1 I:10 -8 -6 -4 -2 0 2 4 6 8 10

Temperature gradient, base-top (T)

Fig. 20 Longitudinal warping plotted againsttemperature gradient (dafiime)

The curve is the predictedwarping, derived from the w

temperature gradien~arpingrelationship, o

0 2 4 6 8 10 12 14 16 18 20 22 24

?me

Fig. 21 Longitudinal warping plottedagainst time (24 hours)

25

the two slabs, one dowelled at the ends, the otherundowelled. Calculations showed the relationship to besimilar for slabs of less thickness. Figure 22 shows thesame cume with measured results on slabs 100mm, and140mm thick. Clearly, variations in the weather will affectthe magnitude of the warping by causing the temperaturedifferential between the top and base of the concreteslabs to vary. This in turn also affects the deflectionmeasurements as the slabs rock under load.

4.7.3 Effects of warping

The general conclusion from this analysis is that concreteroad slabs deform over a 24 hour period, tending to leavea void beneath the centre in the day time and a voidunder the edges at night. Theoretical analysis of theeffects of temperature changes on concrete slabs(Croney, 1991; Road Research Laboratory, 1956)concludes that slab thickness, length, sub-base and otherfactors influence the movements and internal stresses.The measurements recorded here, taken under clear skyconditions in winter at a latitude of 20° south, show littledifference in the movement of slabs with thicknessesvarying from 10Omm to 200mm and lengths of 5m and7.5m. This is due largely to the effect of the concretemass, which acts against the warping tendency, increas-ing the internal stress.

Day-time lack of suppoti at the centre causes bendingstress in the slabs due to self weight, which is additionalto the bending stress due to applied loads, and leads to

2

1

0

-1

-2

transverse cracking followed by longitudinal cracking.

Night-time lack of support at the ends and sides of theslabs promotes rocking under load, joint wear, waterpenetration and hence pumping out of fines from the sub-base, which increases the rocking effect. Lack of edgesupport makes corner cracking more likely, as well astransverse cracking which starts at the outer edges.

Stresses caused in this manner cannot be deduced bysimple calculation because there is also, in practice, adeformation of the sub-base. Day-time hogging concen-trates all the weight of the slabs and applied load aroundthe edges, which tend to be wetter than the centre. Thiscauses some deformation. Night time dishing causesrocking and pumping, which also tend to erode the sub-base under the edges of the slabs. The effect of this canbe hogging of the sub-base under each slab and thisresults in a tendency towards reasonably uniform supportto the concrete slab during the day and more severe lackof support at the sides and ends at night. Under theseconditions shear loads at joints are increased further,accelerating wear and sealant failure, with subsequentfurther deterioration of sub-bases leading to slab rocking,even during the day when the slabs are hogged in shape.The degree of sub-base erosion is dependent on the typeof material, the age of construction and the amount ofwater that has been permitted to enter the sub-base.

Although they are not as extreme as under tropical deseflconditions, the diurnal temperature variations illustratedabove are more severe than those found in most coun-

O 100mm thick ● 140mm thick

The curve is the predicted- (

warping, derived from the ●temperature gradientiwarping ● -relationship.

0 2 4 6 8 10 12 14 16 18 20 22 24

Tme

Fig. 22 Longitudinal warping plotted

against time (24 hours)

26

—

tties, and indeed in Ambabwe during most of the year.However, a surface temperature vatiation in the order of17°C is shown to cause warping, i.e. a change betweenday and night shape in excess of 1mm, in road slabsmade with granite aggregate. Limited variations in slablength and thickness and the presence of dowels at thejoints had no measurable effect on this movement.

Warping is not the only cause of joint deterioration andsub-base erosion on concrete pavements, but therelationship between the diurnal temperature variation atthe slab surface and slab rocking makes jointed concretepavements less suitable in areas where these tempera-ture variations are high.

4.7.4 Slab length

Transverse cracking of jointed slabs occurs when thetotal bending stress due to applied loads, self weight andmanner of suppofi exceeds the flexural fatigue strengthof the concrete. Clearly the length of the slab has aneffect on the stress, both in the day-time hogging shapeand in the night-time dishing shape. The type of aggre-gate used in the concrete affects the slab design lengthbecause different aggregates affect the coefficient oflinear expansion. Most design manuals take this intoaccount (CPCA, 1984; AASHTO, 1974; RRL, 1970).The degree of warping too is dependent on slab length,but also on the diurnal temperature range within the slabsand therefore this should also be a factor to be consid-ered when specifying slab length.

4.8 SHOULDERS

During construction of the shoulders the natural gravelwas distributed on each side of the concrete pavement,graded and mixed by hand with 570 cement. It wascompacted using pedestrian operated vibrating rollersand was then sealed with a single bitumenised surfacedressing. The shoulder material soon shrank leaving agap of at least 3mm against the concrete pavement anda step of 10mm or more from the concrete down to theshoulder. A fillet of bitumenised material was applied butthe gap soon opened again. Unstabilised, well gradedgravel shoulder material would have been more success-ful because:

a)

b)

it would not have shrunk

it would have been plastic enough to maintaincontact against the concrete.

On concrete pavements, tied concrete shoulders aremore successful than shoulders built with unboundaggregates. More cement is required but the bitumenisedsurface dressing and the cement costs are similar.Furthermore, a reduction in concrete pavement thicknessranging from 10OAto 1570 is permitted when tied

shoulders at least one metre wide are specified (DTpStandard, 1987; CPCA, 1984).

5. MAINTENANCE

Over the first six and a half years of use, maintenance onthe test sections has been restricted to:

a) attempts to seal the gap between the concrete andthe stabilised shoulders by the application of abituminous fillet and

b) replacement of transverse joint seals that failed inearly life.

All the joint seals now require replacement. All theremaining sealant should be removed, the groovescleaned with an abrasive and resealed. There is a varietyof sealant material available but a hot applied sealant toBS 2499 (1973) is recommended as a first choice.

The gap between the concrete and the stabilised shoul-ders is more difficult to seal. However, the same sealantas used in the transverse joints would be more success-ful than the bitumen used to-date. Furthermore, thesealant would suffer less strain if a conventional groovewere cut in the shoulder material against the concrete toprovide a joint width in the order of 15mm to 20mm. Thiswould enable the surfaces to be cleaned before applyingthe sealant as well as reducing the strain in the materialduring temperature induced movements and slab rocking.However, since it is not known whether the sealant wouldadhere to the shoulder material, a trial length of about100 metres should be tried first.

The DTp Advice Note HA 35/87 (1987) states, “Duringthe life of the road, some replacement of the concreteslabs will be required in addition to the resealing of jointsand any arris and thin bonded repairs that may berequired.”

The following maintenance is recommended to take placein 1993, after 7 years and approximately 250,000 ESASof traffic:

i)

ii)

iii)

Replacement of seven entire slabs

Full depth repair or replacement of sixteen furtherslabs

Thin bonded arris or corner repairs at twenw fivelocations.

The above repairs should be carried out according to theDTp/C&CA publication by Mildenhall (1986). They arediscussed in detail in PRIOSC/025/93 and the locationsare marked on Figures Al to A4.

27

6. SUMMARY ANDRECOMMENDATIONS

1) The contractor proved that good quality concrete canbe produced using relatively small items of equipmentthat were either old or locally-made. The combination ofvibrating pokers and two light vibrating screeds was veryeffective in compacting a stiff concrete mix (generally15-30mm slump) throughout depths ranging from 80mmto 237mm.

2) A daily temperature variation at the top of theconcrete pavement of 17°C caused the slabs to warp bymore than 1mm (measured as a change in shape of theslab centre relative to the slab edges). This warpingmovement was not measurably affected by slab length(5m to 7.5m) or by slab thickness (100mm to 200mm) orthe use of dowels at transverse joints.

3) The weight of the concrete acting over the unsup-ported areas of a slab, which is warped to the extent thatit is not in full contact with the sub-base, acts against thewarping force and so stresses the material. This bendingstress is proportional to the length of unsuppofled slaband is additional to that resulting from traffic loads. Hencediurnal temperature variations should be consideredwhen designing slab lengths. Until more data becomeavailable from other sites, an interim recommendationrestricting unreinforced slab lengths to 4.5m (5m usinglimestone aggregate) would be advisable where topsurface temperatures vary by 15°C or more daily.

4) Stabilised gravel shoulders are not recommended foruse with concrete pavements because of the difficulty insealing the gap between the pavement and the shouldercaused by shrinkage of the stabilised material.

5) Deflections measured on jointed concrete pavementsare affected by the warping of the slabs, which in turn isinfluenced by weather conditions. Variations in weatherconditions have masked any trend in the deflectionmeasurements made over five years.

6) The electronic deflection logger that was developedas part of this programme has proved to be a useful andreliable tool for the measurement of vertical deflectionson slabs and horizontal strains at joints and cracks onthis and other concrete pavements.

7) A significant amount of fine material has been erodedfrom the subgrade directly below the slabs on sectionsC3 and C4, and to a lesser extent from the sub-basematerial on sections Cl and C2. It is concluded thatconcrete pavements require a non-erodible sub-base, ascurrently specified in the DTp Standard (1987) evenunder these conditions of relatively light traffic and mostlydry climate i.e. 50,000 ESAS and 670mm of rain per year.

to the standards listed in the DTp/C&CA publication byMildenhall (1986). Some slab replacement and repair isalso required.

9) The small amount of faulting that has occurred is notdetectable by a change of roughness as measured by theTRL Profile Beam.

10) Skidding resistance on all trial sections is well abovethe recommended minimum, although there is a smallreduction in the skidding resistance value measured inrecent years. There has been no change in the measure-ments of texture depth. The randomised spacing of thetexture grooves is believed to reduce tyre noise.

11) Transverse joints cut parallel to texture groovespermit vey little water to penetrate to the sub-base, evenwhen the seals are damaged. This advantage wouldclearly be reduced on longitudinal gradients, so bothjoints and texture grooves should be skewed to follow themaximum slope.

12) Local deviations in the finished levels of sub-base onSections Cl and C2 (and subgrade on Sections C3 andC4) and accurate finishing of the top sutiace of theconcrete slabs led to wide variations in the thickness ofthe concrete on all four trial sections. These variationstogether with slab and sub-base strength measurementswill be extremely useful, when enough deterioration hasoccurred, for conclusions to be drawn concerningthickness design of concrete pavements under similarconditions of traffic, climate and materials. To-date,cracking due to traffic is confined to two locations wherethe concrete is only 11Omm to 115mm thick, the damagemost likely being caused by the passage of specificvehicles with axles weighing about 14 tonnes in 1987 and1988.

7. CONCLUSIONS

The project has been managed from TRL and hasincluded annual site visits with excellent cooperation fromthe Research Branch and the Central Laboratoy of theZmbabwe Ministry of Transpoti. This collaborativemethod of working has proved to be effective andeconomical in the use of resources.

It has been demonstrated that concrete pavements,constructed using only basic machinery, are viable underthe conditions of traffic, climate and materials availabilityin Central Africa.

Much of the data and recommendations summarisedabove will be applicable to other regions also, and will beincorporated in a manual on the design and constructionof concrete roads in the Overseas Road Note series.

8) The transverse joints on all the trial sections shouldnow be resealed with an elastomeric material conforming

28

8. ACKNOWLEDGEMENTS

This work has been pati of a continuous programme ofcollaborative research undertaken by the ZimbabweMinistry of Transport and the Overseas Centre of the UKTranspofl Research Laboratory. Thanks are given toMr R Mitchell, Director of State Roads, and all theengineers and technicians in Zimbabwe who have madecontributions during the construction and subsequentcondition assessments.

9. REFERENCES

AASHTO, 1974. Interim guide for the design of pavementstructures. Washington: American Association of StateHighway and Transportation Officials.

ABAYNAYAKA, S W (1984). Calibrating and standardiz-ing road roughness measurements made with responsetype instruments. In: ENPC. International Conference onRoads and Development, Paris, May 1984, ppl 3-18.Presses de I’ecole nationale des ponts et chaussees,Paris.

BSI, 1973. Specification for hot applied joint sealants forconcrete pavements. BS2499: December 1973. London:British Standards Institution.

CRONEY, D and P CRONEY, 1991. The design andperformance of road pavements. London: McGraw-HillBook Company.

CPCA, 1984. Thickness design for concrete highway andstreet pavements. Ottawa: Canadian Portland CementAssociation.

DEPARTMENT OF TRANSPORT, 1987. Structuraldesign of new road pavements. Departmental Standard,HD 14/87. London: Department of Transport.

DEPARTMENT OF TRANSPORT, 1987. Structuraldesign of new road pavements. Advice Note HA 35/87.London: Department of Transport.

DEPARTMENT OF TRANSPORT, 1988. Visual conditionsuweys of concrete pavements. Depatimental StandardHD 17/88. London: Department of Transport.

FRANKLIN, R E, 1988. Skidding resistance of concrete -pedormance of limestone aggregate experiment after 10years. TRRL Research Repofl 144. Crowthorne: Trans-

port and Road Research Laboratory.

MILDENHALL, H S and G D S NORTHCOTE, 1986. Amanual for the maintenance and repair of concrete roads.London: H M Stationery Office.

ROAD RESEARCH LABORATORY, 1956. ConcreteRoads - design and construction. London: H M StationeryOffice.

ROAD RESEARCH LABORATORY, 1969. Instructionsfor using the portable skid resistance tester. Road Note27. London: H M Statione~ Office.

ROAD RESEARCH LABORATORY, 1970. A guide to thestructural design of pavements for new roads. Road Note29. London: H M Stationery Otice.

ROLT, J, 1981. Optimum axle loads of commercialvehicles in developing countries. TRRL LaboratoyRepoti 1002. Crowthorne, Transpoti and Road ResearchLaboratory.

SMITH, H Ret al, 1980. Performance of sections of theNairobi to Mombasa road in Kenya. TRRL Laborato~Repoti 886. Crowthorne, Transport and Road ResearchLaboratory.

WORLD BANK, 1988. Road deterioration in developingcountries. Washington: The International Bank forReconstruction and Development.

PARRY, J, 1985. Concrete roads in developing countries.The Journal of the Institution of Highways and Transpor-tation, (AugusVSeptember), 13-17.

29

FIGURE AlSECTION Cl

VISUAL CONDITION SURVEY DATE: 7-7-92\=cRAcK. \ =C0N51R”CT10N JOINT. JOINT CWO1TION G -GOOD. F - FAIR. 8 -8A0

RECOMMENDED REPAIRS SR - SLAB REPLACEMENT, FD – FULL DEPTH REPAIR, TB - THIN BONDED REPAlR

I 6 3 B 5 8 7j

F 9 B II B,3 T6g ,5 ~ ,7 ~ 76 ,9 ~

2 8 d g a!e

6 8 G 10 F 12 P 14 ~ 16 B 18 6

17+W

20 6

21B

23B

25 g 27 6 B 6 31 6 33 8 35 8 37 8 39 6

22 6 2A 3 26 B 28 B 30 6 32 B 34 B 36 B 38 8 40 B

41 6 43 6 456 47 3 4? $ 51 B 53 B 55 6 57 F 59 F

42 6 44 B 46 8 48 8 50 8 52 8 54 6 56 8 58 F a G

61 5 63 6 65 F 67 ~ 69 6 71 s 73 8 75 G 77 ~ 79 gn

62 6 64 B 66 F a~ ~e’ 7, ~

72 g 74 8 76 B 78 6 80 8

81’5 83 F 85 B 87 K 89 B 91 6 93 8 95 B 97 8 99 s

\82’8 84 6 86 8 m 8 90 e 92 3 94 B 96 ,6 98 6 100

~

10I 6 103 6 105 6 107 3 Iw B Ill $3 113 6 115 6 1!7 B 119 6—

102 6 104 ; 106 6 108 B 110 3 112 6 114 8 116 6 I 18 8 120 6

121 ‘ 8 123 8 125 F 127 F SOUTH S1OE NOTE s5uP&=vfl” rlL7N QNOU : 5arloM - VO@TH 5/DE

122 B 124 3 126 r 128 F NORTH SIDE M6HC4 &Y~SIVE <NOOLDEP Sk felNK4LE

17+Boo

SUGHT Hun <7AIUI N6 ON ~GT# SHOOLbE~<, UACKINL - T@dti<V~CE CQKK5 <e/ 1LED,—

u N5EfiLE D (SAP RETdEE N Sf100LDE~ & ~NCQ&7E .

>

FIGURE A2

SECTION C2VISUAL CONDITION SURVEY DATE: 7-7-92

\=cRAcK. \=coN5rRuflloNJOIN,. JO!N1 CWOITlON: G -GOOD, F - FAIR, O- ~AO

RECOMMENDED REPAIRS SR - SLAB REPLACEMENT, FD - FULL DEPTH REPAIR, TB - THIN BONDED REPAlR

1 ( 3 F 5 F 7 F 9 F IIG

13 6 15 F 17 F 19F

10

2 F 4 6 6 F 8 F 10 F FF ~>

12 Ff 7014 ‘k B

SR’8 f 20

17+m

21 F 23 6 25 F 27 F 29 F 31 F 33 B 35 F 37 ~ 39 $

es{22 8 24 t9 6 26 r 28 F

w ;b( ‘mfa F 36 F

/.538 r 40 e

FIGURE A3 VISUAL CONDITION SURVEYSECTION C3

DATE: 7-7-92\=cRAcK \ =CONSIRUC1lONJOINT. mtNT cwnrrION: G - cooo, F - fAIR. 8 -MO

RECOMMENDED REPAIRS SR – SLAB REPLACEMENT, FD – FULL OEPTH REPAIR, TB - THIN BONDED REPAIR

1 6 3 3 5 6 *U 7Q

B 9 6 II B 13 B Is 8 17 B 19 601 \ ,Q

2 B Brol~

4 6 801 JY

8 6 10 6 12 8 14 9 16 B 18 a 20 B

18+SW

21 9 23O*Q ~ 25 9 27 K 29 G 31 8 33 B 35 ~ 37 9 39 F

22GL ~

6 24 8 26 B 28 F w ~ ~%=F 34 0+ F<9{

36 & 38 8 40 F

61 9 63 & 65 B 67 g 69 B 71 E ?9 73 B 75 E 77 F 79 F

/

62 s 64 3 66 g <~ 6S F 70 6 72 B 74 E 76 F 78 F so F

F81 6 83 F 85 F 87 F 89 5 91 s 93 F 95 B 97 6 99 E

82 F 10 ctiL

“ G$86 & u F 90 F 92 F 94 F 96

698 F 100 F

101 6 103 5 10549 G

107 6 109 K*1

~ +9 Ill ~ 113 F 115 D 117B !19

/—

102 F 104 F 106 F 108 F I 10 3 112 B I 14 F 116 3 118 8 120 B

FIGURE A4 VISUAL CONDITION SURVEY DATE : 7-7-72SECTION C4 \=c@AcK. \.coNm8”a!oNJOINT. ,OIN1 CWOtllON: G ‘GOOD. F - FAIR. B -MD

RECOMMENDED REPAIRS SR - SLAB REPLF\CEMENT, FD - FULL DEPTH REPAIR. TB - THIN BONDED REPAIR

El B 83 F 85F

87 F 89 F 916

93F ~

I

F95 97 99

B

80 5 F F82

F f B ~-84 86 88 90 92 94

B%

5 986

Iol B 103 B 105 B 107 B 109 F Ill B 113 $ 115 6 117 5 119 B.

— ——

100 B 102 F 104 B u F B F B106 108

BI 10 112 I 14 116 118 B

100 I

II I

II

1

I I II 111 I

I

1

II I

Zimbabwe Ispecification - -

:

I /

I1 I I

I I Measured.result

I

II #I /

I II ,0 /4I

II 1

/’

//

4 I

I .4@I

##,f I,0 , -----

./ II /“ II /I

I0’1

/ (I / / /“ I

/0 0’I I~~’ I ,

b“- 1 1##@ / I#e I

/I

t II 1

75 150 300 600pm 1,18 2.36 5 10 26.5 37.5

British Standard sieve sizes

Fig. A5 Wet sieving results - concrete aggregates

34

Run No 36 Date 13/10/92 line 16:34:31 Sample Tiw 10 s Section 11 Chainage 111

VERTICAL 1 :0.00 0.00 0.010.02 0.03 0.040.15 0.12 0.07

-0.22 -0.24 -0.25-0.27 -0.27 -0.27

VERTICAL 2 :0.00 -0.00 -0.000.01 0.02 0.03O.m 0.04 -0.00

-0.28 -0.31 -0.30-0.32 -0,33 -0.32

VERTICAL 1 SWRY :INITIAL VALUE :WIW VALUE :HINIW VALUE :

FINAL VALUE :TWSIENT DEFLECTIW :

WRI ZDNTAL :0.00 0.00 0,00

-0.00 -0.00 -0.00-0.02 -0.01 -0.00

0.02 0.02 0.020.00 0.00 0.00

NDRIZWTAL SWRY :INITIAL VALUE :WIW VALUE :MINIW VALUE :

FIUL VALUE :TWSIENT DEFLECTION :

v] ● ***.* V2++++++Vertical 1 ais :

-1 0+-. ----... ----- .-.. ---. -.-+ ---------

Values in m

0.00 0.000.06 0.080.02 -0.02

-0.26 -0.27

D.01 0.01 0.01 0.010.10 0.13 0.17 0.18

-0.07 -0.11 -0.15 -0.18-0.27 -0.27 -0.27 -0.27

-0.000.04

-0.05-0.31-0.32

0.000.19

-0.28-0.260.32

O.ou-0.01

0.000.010.00

0.000.02

-0.030.000.02

H . .

-0.010.06

-0.10-0.32-0.31

-0.00 -0.00 0.00 0.000.08 0.10 0.12 0.12

-0.14 -0.18 -0.21 -0.24-0.33 -0.32 -0.32 -0.32-0.31 -0.32 -0.31 -0.31

VERTICAL 2 SWRY :lNITIAL VALUE : 0.00w IW VALUE : 0.13MINIW VALUE : -0.33

FINAL VALUE : -0.31TWSIENT DEFLECTIW : 0.29

0.00 -0.00 -0.00 -0.00 0.00-0.01 -0.02 -0.02 -0.03 -0.03

0.01 0.02 0.02 0.02 0.020.01 0.01 0.01 0.01 0.010.00 0.00 0.00 0.00 0.00

REMTIVE [V I-V2] SWRY :INITIAL VALUE : 0.00WIW iALUE : 0.07HININUN VALUE : 0.00

FINAL VALUE : 0.04

I 2.......................-. -.....-. -.-... ---+.---------- ..------ .----- +-------------------------i

Vertical 2 uis :-2 -1 0 1 2 3+-------------------------+-.---------.-------------+-------------------------+-------------------------+--------....-.--.--------+.

..

.

.

..

..

.

-0.27 -0.27 -0.27 -0.27 -0.26 -0.26

0.020.19

-0.20-0.28-0.26

0.010.13

-0.26-0.32-0.31

-0.00-0.030.020.010.00

Plot signification Factor = 4

3 4

.

.

.

....

..

..

..

,+++

+++,+++++

++

..

..

++

++

+. +

* +. +

. +. +

●☞

✎☞

✎☞

●☞

●☞

✎☞

✎☞

●☞

● ☞

●☞

✎ ☞

✎ ☞

✎ ☞

●☞

✎☞

4 ------------------------- + -------------- ----

0.6 0.4Horizontal uis :

.

,

-.

,

.---*-------

++

+

.----

+

+

..--.. --....-+--------I

-----------------+------------------.-.----4

+---.----- ..-------------- +---------------------0.2 0.0 -0.2 -0.4

.-. -,----------------------- .-+---------------- ..-. ---. -+-... ----- .-.. --. ---------+-4 -3 -z -1 0

Relative [V1-V2] axis :1

====== ====== ====== ====== ====== ====== ====== =.==== ====== ,= ========== ====== ====== ====== ====== ====== ====== ====== ====== ====== ==----------- .=. ----------

Fig. A6 Typical deflection logger print and plot

35

TABLE Al

Two-way axle load measurements

Axle Load (tonnes) 1985 1987 1988 1989 1990 1991 1992

C2

C3

C4

<5

<6

<7

C8

<9

<lo

<11

<12

<13

C14

14+

394669461421408207150

681530010

No. of axles from 7 day count

133 140 166 134

343 357 304 286

308 252 222 288

322 294 250 336238 217 355 457133 28 166 431

63 7 102 2860 21 63 195

0 1 22 807 0 7 90 0 1 30 0 0 00 0 0 00 0 0 0

400 240447 515393 513453 429482 695311 601252 341107 17733 81

6 152 30 20 05 0

Damaging factor 569 262 156 491 831 896 1070(ESAs/week)

Total Damaging factor 29,600 26,1OO* 20,600’ 25,500 43,200 46,600 55,600(ESAs/year)

Total ESAS 1987-92: 217,600

ESA = Equivalent Standard Axle.‘ 1987 and 1988 each include 12,500 ESAS as the estimated effect of aggregate lorriesJuly 1987 and 1988 were estimated from several 8 hour counts.

TABLE A2

Two way traffic counts (vehicles per day)

1986 1987 1988 1989 1990 1991 1992

All vehicles: 243 365 221 297 430 425 567

Buses and HGVS: 128 130 126 166(3:$.) (35%) (3;;.) (28/.) (30%) (3370) (2970)

36

TABLE A3

Results of beam test results at TRL - concrete mix trials

Beam No. 28 day Flexural Density Dynamic E Equivalent cubeStrength (MPa) (Kg/m3) (N/mm2 X103) strength (MPa) Notes

515253545659626465

4.534.143.913.624.043.953.763.923.87

240023752420241124062427241224192422

41.0

40.5

42.0

41.0

40.0

41.5

41.5

41.0

41.5

42.043.543.542.0 ● slow loading rate42.041.039.0 * large stone in38.0 fracture40.5

Average: 3.97 2410 41.1 41.3

TABLE A4

Results of beam tests in Zimbabwe - concrete mix trials.

Beam No. 28 day Flexural Equivalent cubeStrength (MPa) strength (MPa)

484950555758606163

3.544.044.054.164.054.313.97

Beam Damaged3.88

39.035.039.035.039.038.035.0

39.6

Average: 4.00 37.5

37

Sub-base Standard

Triaxial class 3.3

Y. passing

Sub-base Test Results

Section Chainage

cl 17.520 CSub-base 17.580 L

17.640 C17,700 R17.760 L

C2 17.840 C

Sub-base 17.900 L17.960 C18.020 R18.080 L

Subgrade Test Results

Section Chainage

cl 17.520 CSubgrade 17.580 L

17.640 C17.700 R17.760 L

C2 17.840 C

Subgrade 17.900 L17.960 C18.030 R18.080 L

C3 18.580 C

Subgrade 18.600 L18.680 R18.800 c18.880 L

C4 19.360 C

Subgrade 19.400 L19.440 R19,480 L19.592 R

TABLE A5

Sub-base and subgrade material propetiies

Y. Passing (sieve sizes inmm)

26.5 19 9.5 4.75 2.36 425p 75p

89-10075-10053-100 37-75 26-75 11-42 5-2

70 Passing (sieve sizes in mm)26.5 19 9.5 4.75 2.36 1.18 600L 300L 150K 75p

96 92 83 73 61 52 45 37 29 2597 95 88 80 63 48 41 35 30 2593 91 77 64 54 47 42 34 25 2095 92 80 66 55 47 41 34 25 2198 93 81 65 54 47 41 33 24 20

97 93 82 69 55 48 41 34 25 2299 92 83 71 62 55 44 32 28 2394 86 75 63 53 48 42 34 25 2098 96 90 82 71 57 49 42 34 3296 92 79 65 55 48 42 32 23 19

Y. Passing (sieve sizes in mm)26.5 19 9.5 4.75 2.36 1.18 600v 300u 150y 75p

9795969698

99nil999595

9696938792

8384838892

95 8793 8293 8493 8595 85

92 8698 9197 9192 8093 81

93 8694 8885 6779 6482 66

71 5171 5768 5179 6489 79

7771747476

72

83846867

79

81605455

4050425369

61 4958 5155 4256 4561 50

59 5174 6274 6159 5356 48

70 5871 5955 5049 46

50 46

35 3246 4239 3647 42

59 53

4144363944

4354534742

515246

4342

3039343949

3436313339

3446453833

4546434039

2836323644

2827262732

28

37352726

3536383534

2332283239

2222232328

2330292322

2928323029

2027232734—

Plasticity PlasticityIndex Product

nean =10 mean =150Norst =12 worst =200

lejection Liquid PlasticityIndex Limit Index

5 27 11

5 21 54 22 6

nil 23 11nil 23 9

2 23 81 23 13

nil 22 94 29 4

6 22 10

?ejection Liquid Plasticity

Index Limit Index

nil

4

2

3

8

192

nil

5

5

10

8313829

nil2935399

2529283533

29

30292322

29

28304542

4740424045

118

122016

13171598

1612122325

2814232123

Note: Allgradings have 100YOpassing 37.5mm sieveC = Centre, L = Lefi, R = Right

38

TABLE A6

Results of cube tests - construction

Section Density 28 day equivalent Cube(Kg/m3) Strength (MPa)

Average Range Average Range

Cl NorthCl South

24432446

2415-24592411-2475

44.041.8

32.1 -47.231.5 -46.4

C2 NorthC2 South

24452443

2426-24662407-2456

40.442.4

32.3 -45.132.5 -47.2

49.944.4

C3 NorthC3 South

24692458

2428-25072421-2493

39.8 -68.334.3 -50.4

C4 NotihC4 South

24602467

2404-24942431-2513

47.743.6

36.0 -58.828.9 -63.6

All Sections: 2454 2404-2513 44.2 28.9 -68.3

TABLE A7

Results of beam tests - construction

28 day equivalent 28 day equivalentSection Density Flexural Strength Cube Strength

(Kg/m3) (MPa) (M Pa)Average Range Average Range Average Range

Cl NorthCl South

2443 2420-24672460 2414-2518

4.82 4.17-5.454.46 3.36-5.07

48.7 27.8 -58.243.6 31.4 -51.9

C2 NorthC2 South

2442 2386-24922445 2418-2496

4.51 3.76-5.064.61 3.97-5.56

38.5 32.8 -43.947.4 32.1 -59.2

C3 NorthC3 South

2469 2346-25282472 2431-2516

4.79 4.17-5.744.64 3.95-5.48

52.8 35.5 -66.849.1 36.9 -82.4

C4 NorthC4 South

2483 2422-25262480 2443-2528

4.85 3.77-5.374.71 3.92-5.55

51.8 34.7 -62.849.8 28.4 -68.3

All Sections: 2462 2414-2528 4.67 3.36-5.74 47.7 27.8 -82.4

,

39

TABLE A8

Core test results

Section No. of Density 90 day equivalenttests (Kg/m3) Cube Strength (MPa)

Average Range Average Range

cl 3 2448 2416-2512 50.5 40.4 -59.1

C2 2* 2485 2482-2487 60.7 58.8 -62.5

C3 4 2427 2410-2451 57.8 52.8 -62.3

C4 3 2444 2423-2459 57.8 52.3 -60.9

All sections: 2451 2410-2512 56.7 40.4 -62.5

● These cores had a lower than recommended height to diameter ratio.

TABLE A9

Cracking summary

Failed

Section No. Average Longitudinal Transverse Corner slabs

slabs concrete 1992

thickness 87 88 89 90 91 92 87 88 89 90 91 92 87 88 89 90 91 92 No. y.

Cl North number 000000 111111 11112’2 0length (m): 000000 0.511111 --- ---

128 173 mmCl South

1number 000000 2222220111 11 1

length (m): 000000 7 77774 ---- --.

C2 North number 000001 4 866613 022225 6length (m): O 00001 9 151515.51530 - - - - - -

126 124 mm 7C2 South number 1 1 0000 111114 332334 3

length (m): 4 40000 3.5 3.5 3.5 3.5 3.5 11 - - - - - -

C3 North numbec OOlll 1 7 77aa5 a66797 1length (m): o 0 0.3 0.3 0.3 0.3 a a 910108 ---- --

152 laa mmC3 South

3number 000000 4 44445 555555 4

length (m): 000000 131314141313 --- ---

C4 North number: 001113 5 a77712 001133 7length (m): O 0 0.3 0.3 0.3 2.5 131517181836 --- ---

163 151 mmC4 South

5number 000112 2 22225 222334 1

length (m): o 0 0 0.3 0.3 1.5 7 777710 ---- --

40

TABLE A1O

Joint defect summa~ - sealant condition

Bad FairSection

Good88 89 90 91 92 88 89 90 91 92 88 89 90 91 92

Cl (n=64) No:70:

021 14 60

070 370 270 2270 9470

7 25 50 43 4ll% 39% 787. 6770 6~0

57371370897. 587. 2070 117. 07.

C2 (n=64) No:70:

1144352% 270 6Y. 6% 55V0

4 29 46 50 296% 4570 72% 78% 45%

59 34 14 10 092Y. 5370 22% 16% 0%

C3(n=75) No:

70:0 13 12 - 66

0% 17~0 16% - 88%18 43 61 - 9

247. 5770 8170 - 127057 19 2 - 0

767. 25% 3% - o%

C4(n=82) No:70:

225 13 63

2% 2% 6% 1670 7770

20 55 62 59 17

24% 677. 767. 72~o 2170

60 25 15 10 2

7370 31% 18% 1270 2%

TABLE Al 1

Jointdefectsumma~ -spelling &stepping

Spalling Spalling Stepping>5mmSection at joints at cracks at joints & cracks

(No.) (No.) (No.)1992 1992 1992

Cl NotihCl South

o1

00

00

C2 NorthC2 South

o1

42

00

C3 NorthC3 South

o2

00

40

C4 NorthC4 South

60

51

41

TABLE A12

Roughness measurements

Section Concrete Profile Beam Reference Roughness(mm/km)*Design 1987 1988 1989 1990 1991 1992

Cl NorthCl South

150mm on 2311sub-base 2322

23982597

24672432

27642486

30133152

23022408

C2 NotihC2 South

125mm on 2781sub-base 2534

28402568

29212538

30712874

2956 283725192686

C3 NorthC3 South

175mm 2342Dowelled 1997

2508 23672346

25042076

23131985

228420062242

C4 NotihC4 South

125mm 26372776

26502833

25762827

26372827

23942564

25592705

*UsingTRLeqn:RR= 2045XRMSD+ 68X(RMSD)2where RMSD= RootMean Square ofDeviationfor thoroughness measurements.

41

TABLE A13

Skidding resistance measurements

Section Skidding Resistance Value1987 1988 1989 1990 1991 1992

0990

8890

107104

120122

10689

7795

Cl NotihCl South

C2 NotihC2 South

C3 NorthC3 South

C4 NorthC4 South

1816

117116

10884

7597

9576

104106

0088

74106

9894

114116

9492

7178

9492

109107

113108

124118

9591

6886

Average: 102 93 104 118 95 81

TABLE A14

Sand patch/texture depth results

Section Average texture depth (mm)1987 1988 1989 1990 1991 1992

Cl North 1.0 1.1 0.8 0.9 0.9 1.0

Cl South 0.9 1.1 0.9 1.0 1.0 1.0

C2 NorthC2 South

C3 NorthC3 South

.0 1.0 0.7 1.0 1.1 1.0

.4 1.4 1.2 1.4 1.4 1.3

.2 1.7 1.1 1.4 1.4 1.3

.1 1.0 0.7 0.9 1.0 0.9

C4 North 1.5 1.6 1.3 1.4 1.6 1.7

C4 South 1.3 2.1 1.2 1.5 1.5 1.7

Average: 1.2 1.4 1.0 1.2 1.2 1.2

42

Section

Cl NorthCl South

C2 NorthC2 South

C3 NorthC3 South

C4 NorthC4 South

Average:

TABLE A15

Benkelman beam & Logger deflections

Average Transient Deflection (mm x 0.01)

Nearside wheelpath1988 1989 1990 1991 1992

BBBLBL BL18 14 19 (23) 20 (13) 29/22 (15)35 23 24 (18) 30 (24) 40/25 (34)

18 16 21 (20) 24 (15) 31/29 (17)23 20 23 (21) 43 (39) 26 (10)

13 14 14 (14) 20 (21) 17 (8)23 16 25 (30) 40 (37) 29 (19)

47 24 55 (92) 36 (30) 33 (19)

27 25 24 (19) 65 (69) 50 (42)

25 19 26 (30) 35 (31) 30 (20)

Offside wheelpath1989, 1990 1991 1992

B B B B16 20 24 31/2616 23 27 41123

20 21 29 36/3811 21 42 22

15 15 23 1912 15 28 23

12 40 35 3816 21 59 50

15 22 33 32

Note: fly indicates a repeated run.B = Benkelman Beam measurements.L = Deflection Logger measurements.

TABLE A16

Rod and level survey measurements for slab Cl-97

Position of level ~SHOULDER~measurements on slab:

::m::

July1992 Time Height (mm)Date: A B c D E F G H I

6 0615 131.0 163.5 116.2 59.0 88.3 106.0 161.5 209.2 192.66 0710 147.0 179.2 131.8 74.5 104.2 121.6 177.4 225.2 208.26 0800 127.3 160.0 112.8 55.3 84.7 102.4 158.0 206.0 189.07 0835 78.3 111.5 64.0 7.0 35.8 54.0 109.2 157.6 140.26 1055 196.5 231.0 183.2 126.5 155.0 174.0 229.0 277.5 259.27 1140 48.5 82.5 34.0 -21.5 6.0 25.0 80.0 129.0 110.56 1315 134.4 168.8 120.3 64.0 92.0 111.3 166.2 214.8 196.77 1320 103.2 138.0 89.4 33.2 61.2 80.2 135.8 183.8 166.04 1450 125.8 160.0 112.5 55.3 83.5 103.0 157.3 206.0 188.06 1525 196.3 230.2 182.2 126.0 154.5 173.5 229.0 277.0 259.04 1545 96.2 130.0 82.2 25.2 54.0 72.2 127.4 176.0 157.5

6 1630 185.0 219.0 171.0 114.3 142.8 161.5 216.8 265.0 247.36 1720 173.3 207.0 159.0 102.3 131.0 149.3 205.0 253.2 235.5

43

TABLE A17

Concrete temperature measurements

TIME Temperature of slab C4-163

on 6-7 July 1992

TOP MID DEPTH BASE(“c) (Oc) (Oc)

01:0002:0003:0004:0005:0006:0007:0008:0009:0010:0011:0012:0013:0014:0015:0016:0017:0018:0019:0020:0021:0022:0023:0024:00

10.49.89.38.88.37.87.69.1

11.214.417.520.022.123.324.123.621.919.117.115.614.313.212.211.3

14.313.713.112.612.111.711.211.211.613.114.916.818.720.221.522.122.121.320.219.018.017.016.115.3

15.915.314.714.313.813.413.012.712.512.713.514.515.817.018.219.119.719.919.619.118.517.917.216.6

MIN: 7.6 11.2 12.5MAX: 24.1 22.1 19.9

44

APPENDIX B: COST ANALYSIS OFCONCRETE PAVEMENTS IN

. condensed from a paper by W C Kuwaza,Research Branch, Ministry of Transport,Zimbabwe.

Construction costs

Five civil engineering contractors tendered for the projectin March 1955. The winning tender was pticed at Zim$263,134 but the final price proved to be Zim $230,076.

The construction of the concrete slabs costZim$18,000more than originally estimated. Of this additional costabout Zim $13,000 came from the 11YOincrease inconcrete volumes brought about by the problems withlevels. The rest of the increase was caused mainly byincreases in the price of cement and aggregate. The risein the cost of cement increased the cost of the concretepavement (including flexible shoulders) by aboutZim $6,000 per km. The corresponding increase forflexible pavements (270 cement in base 1) is about Zim

$1,000 per km. The net increase due to this factor is thusZim $5,000 per kilometre comparing concrete and flexiblepavements.

Comparison of construction costs - concrete vsflexible pavements

Table Al 8 gives a detailed cost breakdown of theconcrete pavement construction. Estimates for structures,earthworks, base 2 construction and shoulder construc-tion were obtained from the Road Design Branch of theMinistry for the 20-62 km section of the Mvuma - Gweruroad being funded by the African Development Bank(ADB). The costs were calculated by simple proportionfor the length involved in the concrete pavement project.The estimated cost of flexible pavement per kilometrewas taken from the aforementioned ADB funded section.

The following comparisons can be made:-

costisquare metre of concrete slab dm $23

costisquare metre of completed concrete pavement Zim $28

costisquare metre of flexible pavement Zim $23

The total cost per square metre of a concrete pavement

would be lower than the above figure if done on a large

scale as the cost of prelimina~ items would be lower perunit length of completed pavement. Using the abovefigures as a basis, concrete pavements cost 22% more atconstruction than flexible pavements. In the UnitedKingdom the construction costs depend on local condi-tions with concrete pavements being cheaper in certainareas and more expensive in others. In the Philippinesthe cost difference between the two methods of construc-tion is not significant. The same state of affairs was alsonoted in the Caribbean Island of Virgin Gorda (Parry,’1985).

Comparison of overall costs of concrete vsflexible pavements

In Western Europe and the USA, problems of ascertain-ing which pavement is the cheaper arise from thedifferent interest groups or lobbies. The concrete lobbyemphasizes the low periodic maintenance costs of thistype of pavement while playing down the usually highercost of construction. The flexible lobby on the other handhighlights the usually lower construction cost whileplaying down the usually more expensive maintenanceaspects. The two sides often use different methods ofeconomic analysis. The Portland Cement Association ofUSA have recommended the use of the “Cash flowmethod” which seems to favour concrete roads, in thatthe higher costs of construction are not discounted overthe life of the road. This coupled with the low mainte-nance costs makes concrete roads cheaper.

The flexible pavement lobby tend to favour the “presentworth” method by using an appropriate discount rate onthe cost of construction which often results in theirpavement being cheaper than concrete.

The Permanent International Association of RoadCongresses (PIARC) have shown in recent analyses bythe committee on concrete roads that no single count~has clearly shown the advantages in terms of costsbetween flexible and concrete pavements.

In Zimbabwe, Makoni (unpublished) used a present worthtype method, which showed that concrete roads weremarginally cheaper for the most heavily trafficked pave-ments but more expensive for the lower traffic categories.The Author has used the cash flow method on thepavement at Mvuma - Gweru and found concrete pave-ments to be substantially cheaper than flexible pave-ments (Table Al 9). These conflicting results give a vividillustration of the problems of economic comparison.

45

TABLE A18

Cost of rigid and flexible pavements at construction

A. Total cost for 1.42km concrete Davement Zimbabwe $

Prelimina~ Items

Structures (culverts)

Earthworks

Pavement Construction for 640m base 2 natural:

(a) Clear gravel pit

(b) Remove overburden

(c) Stockpile base 2

(d) Load, haul up to 1km, control compact

(e) Overhaul pavement material

(f) Water for controlled compaction

Prime (used on C4 only)

Construct shoulders, prime seal with bitumen

Construct asphaltic concrete at transition with flexible pavement (6m 3 total)

Construct concrete slab

Miscellaneous (Fencing, signs, markings)

39,650

40,910

76,850

240

260

2,400

3,460

3,520

890

1,340

23,000

2,000

190,430

12,340

TOTAL 397,290

TOTAL PER Km 280,000

B. Cost of flexible DavementEstimated cost per km 230,000

c. CornDare cost of constructing riaid and flexible pavementAdditional funds required for tigid pavement per km Zim $50,000

Rigid pavement costs 227. more than flexible pavement

46

TABLE A19

Cash flow method of payment for concrete and flexible pavements:Period of comparison 25 years at 10% annual increase in prices

A. Concrete Pavement per kilometre Zimbabwe $

Initial cost 280,000

Total routine maintenance cost: 106,560

At 10 years periodic maintenance: 18,160

At 20 years periodic maintenance: 47,090

TOTAL 451,810

B. Flexible Pavement per kilometre

Initial cost

Total routine maintenance cost:

At 10 years reseal:

Residual life of reseal at 20 years

Reconstruction at 20 years:

Salvage value of reconstruction at 25 years:

230,000

106,560

32,420

0

739,970

554,980

553,970

Therefore flexible pavement costs 23% more than concrete pavement.

NOTES ON ASSUMPTIONS MADE

1.

2.

3.

4.

5.

Concrete pavement has a life of 25 years; flexible 20 years.

Routine maintenance costs are the same for both types of pavement (Zim $985 p.s. based upon National TranspotiStudy 1985).

Periodic maintenance of concrete road cost Zim $7,000 p.s. at present prices, this being restricted to joint sealing.Seal every 10 years.

Reseal of flexible pavement costs Zim $12,500 at present prices and last 10 years.

Reconstruction of flexible pavement costsZim$110,000 at present prices. This is a mean value from asphalticconcrete, granular overlay and reconstruction costs as given in National Transport Study 1985.

Ptinled in the United Kingdom for HMSO

Dd K 63500 8/93 C5 Gw2 10170

47