NATIONAL UNIVERSITY OF SCIENCE AND TECHNOLOGY FACULTY OF INDUSTRIAL TECHNOLOGY DEPARTMENT OF CIVIL AND WATER ENGINEERING FINAL YEAR PROJECT Evaluation of Reclaimed Asphalt Pavement as Constituent Base Course Material

NATIONAL UNIVERSITY OF SCIENCE AND TECHNOLOGY
FACULTY OF INDUSTRIAL TECHNOLOGY
DEPARTMENT OF CIVIL AND WATER ENGINEERING
FINAL YEAR PROJECT
Evaluation of Reclaimed Asphalt Pavement as Constituent Base Course Material. A Case study of Chivi-Mandamabwe road.

(YEAR 2017-2018)
STUDENT NAMEMWOYOSVI DAVID KUSHINGA
STUDENT NUMBER N01310615Y
SUPERVISOR MR T.C. MDLONGWA
MAY 2018
NATIONAL UNIVERSITY OF SCIENCE AND TECHNOLOGY
Department of Civil and Water Engineering
Faculty of Industrial Technology
MAY 2017
Evaluation of Reclaimed Asphalt Pavement as Constituent Base Course Material. (A Case study of Chivi-Mandamabwe road)
“A project submitted to the Faculty of Industrial Technology,
National University of Science and Technology, in partial fulfillment of the requirements for the Degree of
Bachelor of Engineering Honors (B.Eng. Hons) in the Field of Civil and Water
Engineering”
MWOYOSVI DAVID KUSHINGA
(N01310615Y)

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

DECLARATIONI, Mwoyosvi David Kushinga declare that this project is my own work. It has never been submitted to any university or institution for any related degree.

Signed: ________________________ Date: ___________________

DEDICATIONThis work is dedicated to my family, friends and my colleagues.

ACKNOWLEDGEMENTSAll the credit goes to the staff at the Department of Roads, Masvingo and Road Lab, Bulawayo for their coordinated efforts in allowing me to use their facilities and assisting me technically. I would like to acknowledge with thanks, the time and effort given by my academic supervisor, Mr. T.C Mdlongwa in guiding me throughout the whole project from the inception stage to date. I would also want to express my sincere gratitude to my parents in assisting me financially and the Msimanga family as well. I would also like to thank the National University of Science and Technology especially the Civil and Water Department which equipped me with knowledge and essential skills in my area of study. Above all else, I give all the glory and honor to our God Almighty for guiding and protecting me this far.

ANNOTATIONS TOA h c “1” p
AADT: Average Annual Daily Traffic
CBR : California Bearing Ratio
CGF : Cummulative Growth Factor
D : Directional split
FDR : Full Depth Reclamation
GM : Grading Modulus
HCE : High Compactive Effort
HMA : Hot Mix Asphalt
ICE : Intermidiate Compactive Effort
IP : Plasticity Index
LCE : Low Compactive Effort
LDF : Lane Distribution Factor
M: Pavement standard in million standard axles
MDD :Maximum Dry Density
OMC: Optimum Moisture Content
RAP: Reclaimed Asphalt Pavement material
S: Swell of material
SG9: Material class of CBR>9
T5: Pavement treatment procedure
w: Moisture content
Yd: Dry density
YW: Wet Density

ABSTRACTMost roads in Zimbabwe are aged and the situation is now cost straining to the local authorities when they try to maintain them. Resurfacing and patching methods have been applied but pavement failures seem to perpetuate over especially in rain season. Rehabilitating an old pavement by means of pulverising an existing old pavement is referred to Full Depth Reclamation (FDR) and is often used in rehabilitating Low Volume Roads (LVR). This process is an economical rehabilitation technique that provides structural benefit, saves construction raw material and offers a quick return to service. The main objective of this project was to evaluate the use of RAP aggregates as base course material in low volume roads considering a case study of Chivi-Mandamabwe route.

Blend mixes of virgin aggregate and RAP were prepared and tested in accordance to Ministry of Transport, Part N and TMH1 of South-Africa. The tests done were sieve analysis, Atterberg limits test, Compaction test and CBR testing. These tests facilitated material characterisation and classification for design purposes. Blend mix ratios of virgin material: RAP tested were 100:0, 90:10, 70:30 and 50:50 and the 100:0 (unblended sample) was used as a control for the behavioural assessment of every blend mix as the amount of RAP increases. Results from the testing showed that the quality decreases as more RAP is introduced in a mix but however, all the blend mixes were structurally sound though the 50:50 was of low quality. The 50:50 was then chosen as the optimal blend mix and was used to design a Full Depth Reclamation pavement. The 100:0 was used to design a conventional rehabilitation pavement.

From the results of the two designs, the base layer thickness was used to carry out a cost analysis between the two pavements rehabilitation techniques. From the cost analysis, Full Depth Reclamation (FDR) was found economic because it has less haulage cost, consumes less time and is a cold in-place recycling technique whilst the conventional method was found costlier due to haulage costs and lengthy time to complete its works.

TABLE OF CONTENTS TOC o “1-1” h z “Heading 2,1,Heading 3,1,Heading 4,1,Heading 5,1,Balloon Text,1,Header,1,Footer,1,Default,1,Bibliography,1,Caption,1” DECLARATION PAGEREF _Toc512042867 h iDEDICATION PAGEREF _Toc512042868 h iiACKNOWLEDGEMENTS PAGEREF _Toc512042869 h iiiANNOTATIONS PAGEREF _Toc512042870 h ivABSTRACT PAGEREF _Toc512042871 h vTABLE OF CONTENTS PAGEREF _Toc512042872 h viiCHAPTER 1: INTRODUCTION PAGEREF _Toc512042873 h 11.1Background PAGEREF _Toc512042874 h 11.2Problem Statement PAGEREF _Toc512042875 h 21.3Rationale PAGEREF _Toc512042876 h 21.4Research questions PAGEREF _Toc512042877 h 21.5Objectives PAGEREF _Toc512042878 h 31.5.1Main objective PAGEREF _Toc512042879 h 31.5.2Specific objectives PAGEREF _Toc512042880 h 3CHAPTER 2: LITERATURE REVIEW PAGEREF _Toc512042881 h 42.1Introduction PAGEREF _Toc512042882 h 42.2RAP PAGEREF _Toc512042883 h 42.3Recycling of Asphalt pavements PAGEREF _Toc512042884 h 42.3.1Hot recycling of Asphalt pavements PAGEREF _Toc512042885 h 42.3.2Cold recycling of Asphalt pavements PAGEREF _Toc512042886 h 52.4Sources of RAP PAGEREF _Toc512042887 h 52.4.1Milling PAGEREF _Toc512042888 h 52.4.2Full depth reclamation (FDR) PAGEREF _Toc512042889 h 52.4.3RAP Variability PAGEREF _Toc512042890 h 62.5Stabilisation of RAP aggregates for road bases and road subbases PAGEREF _Toc512042891 h 72.5.1Mechanical Stabilisation PAGEREF _Toc512042892 h 72.5.2Bituminous Stabilisation PAGEREF _Toc512042893 h 82.5.3Chemical Stabilisation PAGEREF _Toc512042894 h 82.6Properties RAP PAGEREF _Toc512042895 h 92.6.1Gradation PAGEREF _Toc512042896 h 102.6.2Atterberg Limits PAGEREF _Toc512042897 h 112.6.3Optimum Moisture Content (OMC) and Maximum Dry Density (MDD) PAGEREF _Toc512042898 h 122.6.4Unconfined Compressive Strength Testing PAGEREF _Toc512042899 h 122.6.5Soaked CBR test PAGEREF _Toc512042900 h 132.7Pavements PAGEREF _Toc512042901 h 142.7.1Flexible pavements PAGEREF _Toc512042902 h 142.7.1.1Types of flexible pavements PAGEREF _Toc512042903 h 152.7.1.2Conventional flexible pavements PAGEREF _Toc512042904 h 152.7.1.3Full depth asphalt flexible pavement PAGEREF _Toc512042905 h 152.7.1.4Contained rock asphalt mats flexible pavements PAGEREF _Toc512042906 h 152.7.2Flexible Pavement Structural Layout PAGEREF _Toc512042907 h 152.7.2.1Sub-grade PAGEREF _Toc512042908 h 162.7.2.2Sub-base PAGEREF _Toc512042909 h 162.7.2.3Base PAGEREF _Toc512042910 h 162.7.2.4Binder course PAGEREF _Toc512042911 h 172.7.2.5Surface course PAGEREF _Toc512042912 h 172.7.2.6Prime coat PAGEREF _Toc512042913 h 172.7.2.7Tack coat PAGEREF _Toc512042914 h 172.7.2.8Seal coat PAGEREF _Toc512042915 h 172.8Pavement Design PAGEREF _Toc512042916 h 182.8.1Equivalent Standard axle load PAGEREF _Toc512042917 h 182.8.1.1Equivalency Factor PAGEREF _Toc512042918 h 182.8.1.2Design Traffic estimation PAGEREF _Toc512042919 h 182.8.1.3Design Traffic Class PAGEREF _Toc512042920 h 192.8.1.4Commercial Vehicle Distribution PAGEREF _Toc512042921 h 192.8.2In-situ Subgrade Properties PAGEREF _Toc512042922 h 202.8.3Drainage and Shoulders PAGEREF _Toc512042923 h 20CHAPTER 3: METHODOLOGY PAGEREF _Toc512042924 h 223.1Introduction PAGEREF _Toc512042925 h 223.2Materials PAGEREF _Toc512042926 h 223.3Sampling Plan PAGEREF _Toc512042927 h 223.3.1Sample Preparation for Testing PAGEREF _Toc512042928 h 233.3.1.1Methodology PAGEREF _Toc512042929 h 233.4Laboratory Tests PAGEREF _Toc512042930 h 233.4.1Sieve analysis PAGEREF _Toc512042931 h 233.4.2Atterberg Limits PAGEREF _Toc512042932 h 243.4.2.1Liquid Limit PAGEREF _Toc512042933 h 243.4.2.2Plasticity Index PAGEREF _Toc512042934 h 253.4.3Modified Proctor Tests PAGEREF _Toc512042935 h 253.5CBR tests and Optimal blend mix PAGEREF _Toc512042936 h 263.5.1CBR (California Bearing Ratio) Test PAGEREF _Toc512042937 h 263.5.2Optimal blend mix PAGEREF _Toc512042938 h 283.6Traffic PAGEREF _Toc512042939 h 293.7Cost Analysis PAGEREF _Toc512042940 h 29Methodology PAGEREF _Toc512042941 h 29CHAPTER 4: RESULTS ANALYSIS AND DISCUSSION PAGEREF _Toc512042942 h 314.1Chapter Overview PAGEREF _Toc512042943 h 314.2Test Results PAGEREF _Toc512042944 h 314.2.1Sieve analysis PAGEREF _Toc512042945 h 314.2.2Plasticity Index PAGEREF _Toc512042946 h 334.2.3Compaction PAGEREF _Toc512042947 h 344.2.4Soaked CBR Results PAGEREF _Toc512042948 h 364.3Material Classification for Pavement Design PAGEREF _Toc512042949 h 384.3.1Symbolisation of Layer Treatment PAGEREF _Toc512042950 h 39CHAPTER 5: PAVEMENT DESIGN PAGEREF _Toc512042951 h 405.1Pavement Standard PAGEREF _Toc512042952 h 405.2Pavement layers PAGEREF _Toc512042953 h 405.3: Design Calculation PAGEREF _Toc512042954 h 415.4: Bill of Quantities PAGEREF _Toc512042955 h 46CHAPTER 6: CONCLUSION AND RECOMENTATIONS PAGEREF _Toc512042956 h 47REFERENCES PAGEREF _Toc512042957 h 52

: INTRODUCTION Background
The environmental issues that rose in the last decades led designers to bring about ways that minimise impacts of road construction debris. The Federal Highway Administration’s (FHWA) recycled materials policy suggests that, same materials used to build the original highway system can be re-used to repair, reconstruct, and maintain them. Where appropriate, recycling of aggregates and other highway construction materials makes sound economic, environmental, and engineering sense.
According to (Missouri Asphalt Pavement Association(MAPA), 2007) the Federal Highway Administration(FHWA) estimates that about 100 million of HMA is milled every year and as a result of this vast amounts of reclaimed asphalt pavement(RAP) is produced. The Washington State Department of Transport(WSDOT) allows RAP to be recycled and they are carrying investigations together with other 12 state DOTs on the use of RAP for base course applicationCITATION Eri07 l 12297 (McGarrah, 2007). The two primary factors that influences the use of RAP in asphalt pavement reconstruction operations are economic savings and environmental benefits.
RAP is a useful alternative in road reconstruction compared to the use of virgin aggregate since it reduces the needy for virgin aggregates, which at times is scarce in some areasCITATION Aud11 l 12297 (Copeland, 2011). The use of RAP also conserves energy, lowers transportation costs required to obtain quality virgin aggregate, and preserves resources.
A lot of economic benefits are derived from RAP if used for base or sub-base applications and it is approximated that about 30% material cost savings could be realised if a 50:50 blend of RAP and virgin aggregate is usedCITATION Edw14 l 12297 (Hoppe, 2014). Additionally, using RAP decreases the amount of construction debris placed into landfills and does not deplete non-renewable natural resources such as virgin aggregate and asphalt binder. D’ Andrea et al (2001) suggest that the use of RAP for sub-base layer and subgrade is a useful solution in order to dispose of the large amount of waste produced during maintenance and rehabilitation activities.

In Zimbabwe, existing asphalt pavement materials are commonly removed during resurfacing, rehabilitation, or reconstruction operations but the construction debris is not recycled. The aim of this project is to analyse the use of reclaimed asphalt pavement (RAP) aggregate as base material in road rehabilitation of low volume traffic roads in Zimbabwe.

Problem StatementIn Zimbabwe, asphalt debris produced during road resurfacing, rehabilitation, or reconstruction operations of asphalt pavements is not recycled and this has caused depletion of non-renewable natural resources
RationaleThis study is relevant because asphalt waste material produced during reconstruction operations of asphalt roads is not being re-used and this is causing depletion of non-renewable virgin material like gravel yet recycling asphalt creates a cycle that economises the use of natural resources and sustains the asphalt pavement industry in Zimbabwe. Due to the limited land space in Zimbabwe, there exist a problem of finding new dump sites for the asphalt waste and other industrial waste products. The use of reclaimed asphalt pavement (RAP TA l “RAP: Reclaimed Asphalt Pavement material” s “RAP” c 1 ) aggregate as base material is economic and more environmentally friendly compared to the conventional materials used as base material.

Research questionsIs Reclaimed Asphalt Pavement (RAP) material a suitable virgin aggregate supplement?
What effect does RAP have on mechanical properties of a blended mixture?
Is all RAP homogeneous and how do you improve consistency in RAP?
What percentage of rap should be allowed in base course?
What method is going to be used to recycle the RAP?

Objectives Main objectiveTo evaluate the use of RAP aggregates as composite element in base course material in the construction of low volume traffic pavements
Specific objectivesTo determine a mix blend for application in base reconstruction.

To analyse strength and workability of each mix blend.

To design a Full Depth Reclamation(FDR TA l “FDR: Full Depth Reclamation” s “FDR” c 1 ) pavement basing on the blend found.

To draw a cost analysis between RAP and purely virgin base material as they are used in base reconstruction.

: LITERATURE REVIEW IntroductionThe main thrust of this chapter is to give an in-depth on the reclamation of asphalt pavements. It is also going to review studies that were carried in the past pertaining to reclamation of asphalt pavements. Furthermore, the chapter will explain different ways in which asphalt pavements are reclaimed putting much focus on Full Depth Reclamation(FDR) which is of great interest to the project since in is the most used recycling technique for low volume traffic roads.
RAPRAP is old asphalt pavement that is milled up or ripped off the roadwayCITATION REB01 l 12297 (McDaniel & Anderson, 2001). According to (Kearney, 1997) reclaimed asphalt pavement is the previous road pavement chunk material produced during reconstruction operations which have been damaged. RAP can be reused in new hot asphalt mixtures or cold asphalt mixtures since its components will still be valuable. Using RAP in new mixtures reduces quantities of imported new material to be added to it, reduces costs as well as depletion natural resources.

Recycling of Asphalt pavementsPavement rehabilitation and reconstruction generates large quantities of reclaimed asphalt pavement (RAP TA l “RAP:ASPHALT” s “RAP” c 1 ) aggregate. Recycling the generated aggregates into a new pavement creates a way that saves material, cost and time. In addition to this recycling process, disposal problems are reduced as well. A lot of benefits are derived from the recycling process. These include; reduced costs of construction, conservation of aggregate and binders, conservation of the existing pavement geometrics, preservation of the environment, conservation of energy and less user delay (Mallick, 1997). There are different ways in which asphalt pavements can be recycled.

Hot recycling of Asphalt pavementsHot recycling process involves combining RAP material with new material, sometimes to produce hot mix asphalt (HMA TA l “HMA: Hot Mix Asphalt” s “HMA” c 1 ) mixtures. The recycling can be done in place or offsite at a central plant (Mallick, 1997).

Cold recycling of Asphalt pavementsCold recycling is a process in which an existing old pavement material is reused without the application of heat. The process can be carried out in-place (cold in-place recycling) or offsite at a central plant (Mallick, 1997).

Sources of RAPRAP is obtained from different sources which include milling, full depth pavement reclamation and waste HMA materials generated at the plant. In RAP management, it is important to know when to keep RAP from a new source separate and when to combine RAP from different sourcesCITATION Aud11 l 12297 (Copeland, 2011).

MillingThis process involves removal of distressed upper layer(s) of an existing pavement to a given depth. The milling machines grind, pick up and load RAP into a truck for transportation. During operation, the milling speed should be controlled and kept uniform so as to enhance consistency in the resulting RAP. In some cases, it may be beneficial to mill the surface layers and intermediate layers separately from the asphalt base layer since the upper layers often contain aggregates with special characteristics such as polishing resistance and/or freeze/thaw durabilityCITATION Aud11 l 12297 (Copeland, 2011)
Figure STYLEREF 1 s 2. SEQ Figure * ARABIC s 1 1:Road Reclaimer performing FDR of an asphalt roadwayCITATION Gar17 l 12297 (D.Reeder, 2017) Full depth reclamation (FDR)Full depth reclamation (FDR) has been defined as a recycling method where all of the asphalt pavement section is pulverised and a predetermined amount of underlying subbase material are treated to produce a stabilized base courseCITATION Kan97 l 12297 (Kandhal & Mallick, 1997). The FDR technique is selected basing on several factors. These include the condition of existing pavement, availability of virgin aggregate, traffic loads and cost of reconstruction CITATION Pen12 l 12297 (Pennsylvania(DOT), 2012). According to CITATION Har11 l 12297 (Harrington & Adaska, 2011) FDR is most appropriate when pavement failure can no longer be rectified by simple resurfacing methods and also when the pavement distress indicates that the primary failure is within the base or subgrade. Further stabilisation is obtained through addition of aggregate and other stabilisers like Portland cement. The four main steps in this process are in-situ pulverization of existing pavement and underlying layers, uniform blending of pulverized material, grading, and compaction. If the in-place material is not sufficient to provide the desired depth of the treated base, new materials may be imported and included in the processing. New aggregates can also be added to the in-place material to obtain a particular gradation of material.
FDR rehabilitated roads are associated with quite a number of improvements compared to other rehablitation techniques. These include increased capacity (through road widening), increased structural strength (through proper stabilization depth and structural overlay), and improved road condition and service life CITATION Pen12 l 12297 (Pennsylvania(DOT), 2012).

Figure STYLEREF 1 s 2. SEQ Figure * ARABIC s 1 2: Heavy patching indicating end of service CITATION Wir10 l 12297 (WirtgenGmbH, 2010) RAP VariabilityRAP is highly variable due to different gradations, oil contents and milling processes. Blending virgin aggregate with RAP increases this variability causing a wide range of test results observed in the published literatureCITATION Eri07 l 12297 (McGarrah, 2007). According to CITATION Cal13 l 12297 (California(DOT), 2013) in-place material is subject to variations in their composition, uniformity and quality. RAP removed from an old pavement include original pavement materials, patches, chip seals and other maintenance treatments. It is recommended to stockpile RAP from different projects separately as means of minimising variations in consistency of materials. Good stockpile management practices should be followed to keep material variability in check. Research has shown that the variability of RAP can be controlled and may not be as great as expected.CITATION REB01 l 12297 (McDaniel & Anderson, 2001). Variability in bituminous layer thickness will result in the percentage of RAP changing and the best method to control these variations is best achieved through upfront testing and some design stages of FDR projects but however, even with careful planning variation in RAP will still be encounteredCITATION Tom12 l 12297 (Sebesta, 2012).

Stabilisation of RAP aggregates for road bases and road subbasesJust like virgin aggregates, RAP aggregates can function as structural components of a pavement if stabilised. This functioning is more pronounced when RAP-virgin aggregate mixture is further stabilised since the initial stabilisation (mechanical stabilisation) through the pulverisation process is not sufficient. According to CITATION Wir10 l 12297 (WirtgenGmbH, 2010) adding any stabilising agent binds material particles together in such a way that changes the behaviour of materials under loading thus the material will act as a slab with different stress patterns. The stabilisation is the same as that done in conventional base construction. There are several ways in which the stabilisation can be done. These include mechanical stabilisation (addition of aggregates), chemical stabilisation (addition of cement) and bituminous stabilisation (addition of asphalt binder).

Mechanical StabilisationMechanical stabilization is defined by the Asphalt Recycling and Reclaiming Association (ARRA, 2015) as pulverization, mixing, and densification of reclaimed materials with the addition of granular materials, if necessary, to produce the required degree of structural support. This type of stabilization relies most on particle interlocking between the pulverized mixture of the existing asphalt pavement and the subsurface layers. The pulverised mixture will be compacted to a specified density soon after the stabilisation processCITATION Gar17 l 12297 (D.Reeder, 2017).

Bituminous StabilisationBituminous stabilization is done by means of mixing the pulverized asphalt pavement and subsurface materials with an emulsified asphalt or foamed asphalt. Bituminous stabilization can be aided with other stabilizers such as Portland cement for it to meet the required optimal FDR performance. Materials stabilised with bitumen do not suffer from shrinkage cracking unlike cement stabilisers CITATION Wir10 l 12297 (WirtgenGmbH, 2010).

Chemical StabilisationChemical stabilization is done by means of mixing the pulverized asphalt pavement and subsurface materials with a chemical stabilizing material. The materials often used for chemical stabilization include: Portland cement, Lime, Class C or F fly ash, Cement kiln dust, Lime kiln dust, Calcium chloride and Magnesium chloride but of all, cement is widely used.
This is because cement-stabilized FDR mixtures have higher initial strength values and long-term strengths compared to other reclaimed mixtures stabilized with lime or bituminous agents. Cement stabilisation increases strength and stiffness but introduces shrinkage cracking CITATION Wir10 l 12297 (WirtgenGmbH, 2010). Unlike other stabilisations, cement stabilisation is versatile since it works on a wide range of existing materials and can also accommodate highly plastic soilsCITATION Gar17 l 12297 (D.Reeder, 2017). Other stabilisation methods are more limited in the types of existing materials they can effectively treat. Figure 6 clearly explains the versatility of cement stabilisation over others as far as Full Depth Reclamation (FDR) is concerned.

Figure STYLEREF 1 s 2. SEQ Figure * ARABIC s 1 3: Versatlity of Cement compared to other stabilising agents used with FDR, ARRA 2015.

Properties RAPJust like natural material, FDR mixtures has similar properties to those of virgin material that makes them suitable to be used as base material for low traffic volume roads. Various laboratory tests were conducted to determine the physical properties of RAP – virgin aggregates blends. These tests included gradation, Atterberg limits, moisture content, specific gravity, water absorption, sand equivalent and toughness (Taha et al.1999). The results of a laboratory study conducted at Sultan Qaboos University indicated that RAP aggregate could be expected to replace virgin aggregate in road subbases if RAP is mixed with other virgin aggregates. From the study (Taha et al.1999) recommends the use of RAP: virgin aggregate blends so as to attain proper bearing strength since the RAP: virgin aggregate blend bearing capacity is usually lower than that of conventional granular aggregates.

The State Organisation of Roads and Bridges CITATION SCR03 l 12297 (SCRB, 2003) in Iraq conducted a study on blending RAP with natural gravel. The material that was tested had a maximum dry density of 2235 kg/m3 at an optimum moisture content of 6.5% and when more RAP was added to the natural gravel the maximum dry density reduced to 2000 kg/m3 at an optimum content of 6.8%CITATION Sau13 l 12297 (Sultan, 2013).

Laboratory and field evaluations of the use of RAP in road base and subbase applications were conducted by Rutgers University (Maher and Popp 1997). Results of this study showed that RAP has a slightly higher resilient modulus and field elastic modulus than the dense-graded aggregate used in the State of New Jersey.

Table STYLEREF 1 s 2. SEQ Table * ARABIC s 1 1: Physical Properties of Reclaimed Asphalt Pavement and Virgin AggregatesCITATION Ram02 l 12297 (Taha, et al., 2002))
Property Reclaimed Asphalt Pavement Virgin Aggregate
Moisture content (%) 0.23 0.86
Specific Gravity (SSD) 2.12 ND
Water Absorption (%) 1.0 ND
Sand equivalence (%) 97 67
Los Angles Abrasion (%) 33.6 18.8
GradationSieve analyses were performed on RAP and virgin aggregates in accordance with AASHTO T27. RAP was generally classified as well-graded gravel (GW). The uniformity coefficient (Cu) was 6 and the coefficient of curvature (Cc) was 1.5. The virgin aggregate was a mixture of well-graded sands and gravelly sands with little or no fines (SW). The uniformity coefficient was 15, and the coefficient of curvature was 1.

Figure STYLEREF 1 s 2. SEQ Figure * ARABIC s 1 4: Particle size distribution for reclaimed asphalt pavement(RAP) and Virgin aggregates CITATION Ram02 l 12297 (Taha, et al., 2002).

Atterberg LimitsThe liquid limit test was performed in accordance with AASHTO T89 cup method. The liquid limit of RAP aggregate was 8, and it was essentially non-plastic. The Omani specifications for a base course are a maximum liquid limit of 25% and a maximum plasticity index of 6%. Similarly, the virgin aggregate was characterized as being non-plasticCITATION Ram02 l 12297 (Taha, et al., 2002).

Optimum Moisture Content (OMC) and Maximum Dry Density (MDD)
Figure STYLEREF 1 s 2. SEQ Figure * ARABIC s 1 5: Optimum Moisture Content and Maximum Dry Density Results for All Blends CITATION Ram02 l 12297 (Taha, et al., 2002)Figure 5 above shows a summary of the optimum moisture content and maximum dry density values presented in Table 2 extracted from the Journal of Material in Civil Engineering (2002). The data shows that as the percentage of cement increases for each mixture, the optimum moisture content and maximum dry density values will slightly increase. Similarly, maximum dry density will increase as more virgin aggregate is added to RAP due to the increase in the percentage of fines in each mixture.

Unconfined Compressive Strength TestingThe test was performed in accordance to ASTM D1633 using a 100 kN DARTEC machine with a loading rate of 0.5kN/s. For strength-based tests, similarsamples were prepared for each mixture. These samples were left for 3, 7, and 28 days in sealed plastic bags under room conditions. The data obtained showed that as virgin aggregate and cement contents in the blend increases, the strength value increases as well. This is clearly shown in figure 6.

Figure STYLEREF 1 s 2. SEQ Figure * ARABIC s 1 6: Unconfined compressive strength test results CITATION Ram02 l 12297 (Taha, et al., 2002)Soaked CBR testIncreasing the percentage of RAP in a mixture decreases the mechanical strength of the resulting mixture CITATION SCR03 l 12297 (SCRB, 2003). According to CITATION Sau13 l 12297 (Sultan, 2013) the minimum CBR TA l “CBR: California Bearing Ratio” s “CBR” c 1 for base or subbase should be more than 20% and not more than 40% of RAP is required in a mixture requiring a CBR>20% therefore, an optimal maximum percentage of 40 of RAP can be blended with local virgin material. The variation of CBR as the percentage of RAP increases is shown in the figure below.

Figure STYLEREF 1 s 2. SEQ Figure * ARABIC s 1 7: Relationship Between CBR and Percentage of Reclaimed Asphalt Pavement Materials (RAP) in the Blend of RAP and Natural Gravel Materials CITATION SCR03 l 12297 (SCRB, 2003) PavementsIn Zimbabwe, less attention is given to the maintenance of pavements and this has led to increased deterioration in most pavements. Failures in these pavements increases especially in rain seasons because of ease ingress of water into the pavement through cracks on roads or poorly finished patches. Pavements are designed in such a way that they withstand traffic loads from the vehicle tyres into the pavement layers without stressing the pavement.

A wide range of materials can be used for unbound bases. These include: crushed rock or stone, naturally occurring material as ‘dug’ gravels. These material should contain sufficient fines in order to produce a dense layer when compacted. The expected level of future maintenance and the total cost over the expected life of a pavement should be taken into account when designing it. The use of locally available materials is encouraged, particularly at low traffic volumes that is categories T1 and T2 CITATION SAT01 l 12297 (SATTC, 2001). There are two common types of pavements used in Zimbabwe namely rigid pavement and flexible pavement.
Flexible pavements are widely used in Zimbabwe compared to rigid pavements but however these pavements are more susceptible to failures compared to rigid pavements. There are situations where rigid pavements are preferable to flexible pavements. These include: areas prone to chemical attack, heavy breaking zones and highly trafficked areas.

Flexible pavementsThese are pavements which are capable of adjusting its position to the shape of the underlying layers without sustaining considerable damage (Tonias, 1999). Flexible pavements transmit wheel load stresses to the lower layers by grain-to-grain transfer through the points of contact in the granular structure. These wheel loads are distributed to a wider area of the, and the stress decreases with the depth. Due to the stress distribution characteristics, flexible pavements are normally layered. The design life of flexible pavements ranges from 15 to 20 years without major maintenance.
Types of flexible pavementsCITATION Tom06 l 12297 (Mathew, 2006) discussed that there are three main types of flexible pavements. These are convectional flexible pavement, full depth asphalt flexible pavement and contained rock asphalt mats.

Conventional flexible pavementsThese flexible pavements are characterized with high quality materials which are placed in the top layer where there is high stress intensity while low quality cheap materials are placed in lower layers as the stress intensity decrease with depth.

Full depth asphalt flexible pavementFull depth asphalt pavements are constructed in such way that the bituminous layers are placed directly on the soil subgrade. This type of pavement is usually used in areas where there are low traffic volumes and scarce of local materials for the subbases and bases in the area. This type pavement is conservative since the existing pavement is recycled and little quantities of natural material is used during construction of such a pavement. When RAP is used in FDR mixtures, there is need for taking it into account in the new designCITATION Gar17 l 12297 (D.Reeder, 2017).

Contained rock asphalt mats flexible pavementsThese pavements are constructed by means of placing dense graded material layers between two asphalt layers. As means of minimising vertical compressive strains generation from the top layers of the pavement, a modified dense graded asphalt concrete is added directly above the subgrade and this concrete layer will protect the subgrade from surface waterCITATION Tom06 l 12297 (Mathew, 2006).

Flexible Pavement Structural Layout Figure 2.8 illustrates all the components in a typical cross section of a flexible pavement.

Figure STYLEREF 1 s 2. SEQ Figure * ARABIC s 1 8: Typical cross section of a flexible pavementCITATION Tom06 l 12297 (Mathew, 2006)A typical pavement consists of the sub-grade, the sub-base course, base course, binder course, surface course, prime coat, tack coat and seal coat.

Sub-gradeThis is the underlying layer of soil in a pavement that is referred as to the foundation of a flexible pavement thus it withstands all the stresses coming from the top layers of the pavement to itself. The subgrade should be compacted to a desirable density, near the optimum moisture content
Sub-baseThis is a layer of material situated beneath the base course and its primary functions are to provide structural support, improve drainage and reduce the intrusion of fines from the sub-grade in the pavement structure. If the base course is open graded, then the sub-base course with more fines can serve as a filler between sub-grade and the base courseCITATION Tom06 l 12297 (Mathew, 2006). This layer maybe provided or not depending on the strength of the subgrade. If the subgrade is stiff to the desired level, this layer is not provided.

BaseThis is the layer of material that is directly beneath the binder course. It consists of aggregate materials such as gravel, crushed stone, sand or a combination of two more of these aggregates. These materials should be of high quality than those of the sub base since this layer is subjected to more stresses than the subbase and also contributes to the sub-surface drainage.

Binder courseThis layer provides the bulk of the asphalt concrete structure. The main purpose of this layer is to distribute load to the base course. The binder course generally consists of aggregates having less asphalt and doesn’t require quality material as high as of the surface course, so replacing a part of the surface course by the binder course results in more economical designCITATION Tom06 l 12297 (Mathew, 2006)Surface courseThis is the top layer material of a flexible pavement and where the traffic loads are transferred from the vehicle tyres to the pavement. The surface layer should be strong enough to sustain the pressure exerted by the vehicle tyres. The materials used to make up this course layer should be inert in all-weather conditions so as to prevent the layer from wearing under traffic loads. The layer should also be impermeable to prevent storm water from entering in to the flexible pavement. It must also provide smooth riding surface for the moving vehicles.

Prime coatPrime coat is a low viscous bitumen liquid which is spread on top of a mosaic finish of a clean base course. The prime will percolate and penetrate downward into the voids until it is totally absorbed. It will then form a water tight surface and also ease the binding of the base course and surface together. The prime will prevent seepage of water from sub-grade. The surface of the base where the bitumen prime is to be applied should be dry and clean.

Tack coatThe tack coat is used to provide ease bonding between two pavement surfaces usually a new wearing surface and an existing base or binder course surface. The existing surface where the coat is applied should be dry and clean. Chips used on this layer are of greater nominal size compared to those used for sealing.

Seal coatSeal coat consists of thick binder which is covered with fine aggregates applied on top of the surface course with aim of creating a water tight surface. This will prevent ingress of storm water into the pavement structure and it also provides skid resistance to vehicles.

Pavement Design Equivalent Standard axle loadThe pavement design process requires the estimation of the average daily number of Equivalent Standard Axles (ESAs) on one lane at the opening of the new road to traffic, which is then projected and cumulated over the design period to give the design traffic loading. Moving loads has damaging effect on road pavement and this can be compared to the damaging effect caused by a standard axle load. This load can be converted into equivalent number of standard axles loads. The extent of pavement damage is not proportional to the axles loads. A standard load of 80kN is usually used as the single axle load with a dual wheel arrangement. For vehicles using multi-axle configurations (such as tandems and tridems), some agencies introduce further factors to derive modified load equivalencies on the basis that these axle groupings may be less damaging than the sum of the individual axles CITATION SAT01 l 12297 (SATTC, 2001)Equivalency FactorThe power damage law equation is used to calculate the load equivalency factor (LEF) of any given axle load. The ESA or E80 is the unit for the equivalent standard axles. The load equivalency factor is given by the equation below.

LEF =P80n ………………………Equation STYLEREF 1 s 2. SEQ Equation * ARABIC s 1 1
Where LEF is the load equivalency factor P is any load for which the load equivalency is required n is damage exponent often used as 4 CITATION SAT01 l 12297 (SATTC, 2001).

Design Traffic estimationDesign traffic estimation considers heavy vehicles or commercial vehicles as the vehicles having a damaging effect on road pavement while light vehicles on the other hard are ignored as their structural damage effect is very minimal (Zofka 1990). A vehicle having a net mass exceeding 2300kg is considered to be a heavy vehicle. The annual traffic flow of these commercial vehicles is found by multiplying the daily traffic flow by 365days/year. The total number of vehicles that will use the road pavement during its entire design life can then be obtained by multiplying the annul traffic by the design life in years selected of the road pavement. The design life is the period during which the road is expected to carry traffic at a satisfactory level of service, without requiring major rehabilitation or repair work CITATION SAT01 l 12297 (SATTC, 2001). In order to come up with a design traffic flow, site surveys must be conducted and only commercial vehicles must be counted.

Design Traffic Class
Figure STYLEREF 1 s 2. SEQ Figure * ARABIC s 1 9: Design traffic classes CITATION SAT01 l 12297 (SATTC, 2001)The figure above shows the classification of design traffic and the category depends on cumulative ESAs projected. Category T1 shows design traffic less than 0.3 million ESAs is considered to be practical minimum since realistic layer thicknesses as well as materials specifications tend to preclude lighter structures for lesser traffic. This class caters for low volume traffic roads. Low volume roads include access roads in the residential areas are generally low volume roads with traffic volume less than or equal to 400 vehicles per day (Douglas, 2004).

Commercial Vehicle Distribution
Surveys carried out in the past indicates that most commercial vehicle drivers tend to use the outer slow lane in multi-lane carriage ways (Zofka, 1990). It is assumed that about 65% of the commercial vehicles use the outer slow lane although this percentage is not to exceed the traffic capacity of the lane and due to this, road shoulders should be designed in the same way as the actual lanes of the pavement as they continuously experience heavy traffic loads from commercial vehicles.

In-situ Subgrade PropertiesIn designing flexible and concrete pavements, the subgrade strength must be known. California Bearing Ratio (CBR) test is seldom used to determine the elastic modulus of subgrade for granular soils whereas for cohesive soils plasticity index is preferred. There are other un-destructive tests used for the assessment of the elastic modulus of the subgrade. These include Falling Weight Deflectometer (FWD) and Dynamic Cone Penetrometer (DCP). Any of these methods cannot be that accurate in evaluating CBR of the subgrades. Each method is best suited to different circumstances. The choice of a method to use is greatly influenced by type of material in addition to precision and cost of the materialCITATION Ken00 l 12297 (KentCountyCouncil, 2000).

CBR Tests (Granular Soils)The relationship between CBR values and the elastic modulus of the subgrade is given by the following relationship:
Es = 10 x CBR…………Equation STYLEREF 1 s 2. SEQ Equation * ARABIC s 1 2
where, Es = elastic modulus of the subgrade MPa, CBR = California Bearing Ratio %CITATION Div13 l 12297 (Division ,Research ; Development, 2013).

Plasticity Index (Cohesive Soils)For cohesive soils, the CBR test is not reliable. The following relationship is used to estimate the elastic modulus of the subgrade from the Plasticity Index:
Es = 70 – Ip ………………(Equation STYLEREF 1 s 2. SEQ Equation * ARABIC s 1 3
where, Es = elastic modulus of the subgrade MPa, Ip = Plasticity Index %CITATION Div13 l 12297 (Division ,Research ; Development, 2013).
Drainage and ShouldersDrainage within the pavement layers is a significant element of structural design since the strength of a subgrade used for design purposes is governed by moisture content and it is a critical issue to ensure that water quickly drain off from within the pavement. The performance of a road throughout its design life is greatly influenced by both its shoulders and drainage. Providing suitable drainage is needed since there is probability of pavement layers being affected by water and moisture ingress. Shoulders aids effective drainage of a pavement in conjunction with other road drainage features. In addition, shoulders provide lateral support to the pavement structure as means of preventing layer materials from deteriorating when trafficked CITATION SAT01 l 12297 (SATTC, 2001).

: METHODOLOGY IntroductionThis chapter discusses methods that were used to assess the soundness and strength of different blend mix ratios of RAP: natural aggregates and also methods to come up with pavement design parameters. These methods included experiments and procedures that were carried out under laboratory conditions in accordance the Standard Testing Procedures (STPs) of 1989 Ministry of Transport Material Testing Manual, Part N and TMH1 of South African Standard Testing method, 1986. The experiments done were Sieve Analysis, Atterberg limits tests, CBR test and Compaction Tests. All these tests were done on each of the following mix blends of virgin aggregates: RAP 100:0, 90:10, 70:30, 50:50. These blends were not stabilised. The chapter also discusses the cost analysis for the compared rehabilitation techniques thus conventional rehabilitation and Full Depth Reclamation (FDR).

MaterialsRAP and gravel samples used in the experiments were obtained along Chivi-Mandamabwe route which is the selected study area. The gravel pit used was found using a gravel pit plan for that route.

Sampling PlanDisturbed gravel was taken from one gravel stockpile of a pit situated along Chivi- Mandamabwe route. A gravel pit plan for that route was used to locate the pit. This gravel was blended with RAP with respect to the blend mix ratios mentioned above. The RAP used was obtained from rejected pothole material stockpiled on the sides of the road and also from patches under repair (gravel patching) and these were taken as trial pits. Samples of RAP were collected in diagonals as means of obtaining a uniform mixture which covers the whole area under analysis. The ratio indicated the percentage by mass with respect to a blend mix.
An unblended mix of purely virgin material (100:0) was used as control to assess the behaviour of parameters to be tested on how they vary as the percentage of RAP increases in all the prepared blend mixes. A batch of at least 55kg was prepared on each blend mix so as to accommodate experiments done on each blend mix. Four blend mixes analysed were 100:0, 90:10, 70:30 and 50:50.

Sample Preparation for TestingThe blend mix samples were treated as disturbed samples and were prepared according to STP1 of 1989 Part N Material testing manual of the Ministry of Transport.

MethodologyThe blending was done with respect to ratios stated earlier on in this chapter. Each blend mix sample prepared was left to dry in open air. The mass of each batch was at least 55kg. The whole sample was then back mixed to ensure thorough mixing of the sample. For Sieve analysis, a mass of 2000g was put aside for the determination of Atterberg limits and for compaction tests samples of 7000g were prepared after back mixing was done and a five-point testing method was used on each of the 7000g sample.

Laboratory TestsThe following tests were done on each blend mix sample prepared; Sieve Analysis, Atterberg Limits Tests, Compaction Tests and CBR test.

Sieve analysisThe test was carried out in accordance to method A.1(b) of TMH1 of South African Standard Testing method, 1986. This test was used to determine the particle size distribution and to classify the aggregate blend mix being analysed by means of determining the masses of successive amounts of material passing a series of standard B.S Sieves nested in descending order of their sizes and a collecting pan placed at be bottom of the nest. The percentage by mass of material retained on each sieve was recorded. For each blend mix the Grading Modulus (GM TA l “GM: Grading Modulus” s “GM” c 1 ) was calculated using equation 3.1:
GM=300-(P2mm+P0.425mm+P0.075)100…..… Equation STYLEREF 1 s 3. SEQ Equation * ARABIC s 1 1 CITATION TRH85 l 12297 (TRH04, 1996)Where, P is the percentage passing on the indicated sieve size and GM is the grading modulus.
Atterberg LimitsThe Atterberg limits will be determined according to the Material Testing Manual, Part N. These limits are plastic limit and liquid limit and they are to be determined on each of the ratios mentioned in earlier on in this chapter.

The values of the liquid and plastic limit are used to classify the fine grained soils and also to assess the behaviour of the aggregate blend under different moisture contents. These two values can will be used to calculate the plastic index, toughness index and these indexes will be used to predict the plasticity, cohesiveness, permeability, shear strength and compressibility of each aggregate blend.

Liquid LimitFor each sample, a blend mix sample that passed the 425µmBS sieve was thoroughly mixed with distilled water until a uniform paste was formed. A portion of the uniform paste was placed in a Casagrande’s cup and levelled parallel to the base to a depth of 10mm at the deepest point. The paste was then divided by drawing the grooving tool along the diameter through the centre of the hinge keeping the tool normal to the surface of the cup and a clear sharp groove will be formed. The crank of the device was turned at a rate of 2 revolutions per second and this made the cup to be lifted and dropped until the two parts of the paste sample came into are intimately close at the bottom of the groove. The procedure will be repeated at different moisture contents and all the values obtained will be recorded. These values recorded were corrected with corresponding factors of the blows and the corrected values were recorded for further calculations as shown in the table below.

NUMBER OF BLOWS FACTOR NUMBER OF BLOWS FACTOR NUMBER OF BLOWS FACTOR
15 0.95 21 0.98 27 1.01
16 0.96 22 0.99 28 1.01
17 0.96 23 0.99 29 1.01
18 0.97 24 0.99 30 1.02
19 0.97 25 1.00 31 1.02
20 0.98 26 1.00 32 1.02
Table STYLEREF 1 s 3. SEQ Table * ARABIC s 1 1: Factors for correction to 25 blows,CITATION Min89 l 12297 (Ministry of Transport: Part N, 1989) Plastic limit
A blend mix sample of the material passing the 425µm sieve will be mixed with water until it become plastic. The sample was divided into two portions, each approximately weighing 10g. The sample was then rolled between the glass plate and fingers until it becomes approximately 3mm in diameter. The samples were further broken into six pieces and re-rolled to threads approximately 3mm in diameter and also until the threads crumbled. The moisture content of the crumbed threads was determined. The procedure was repeated 2 times using fresh soil samples on each of the blend mixes.

Plasticity IndexFrom the recorded values of Liquid limit and Plastic limit results of the blend mix samples that were tested, the plasticity index was deduced and was reported as percentages. The plasticity index was calculated using the equation below:
Plasticity index(Ip TA l “IP: Plasticity Index” s “Ip” c 1 ) =Liquid Limit(LL) – Plastic Limit(PL)………… Equation STYLEREF 1 s 3. SEQ Equation * ARABIC s 1 2
Modified Proctor TestsThis test was done in accordance to STP 15 of Ministry of Transport of Zimbabwe 1989 Part N. The test was used to establish the relationship between Maximum Dry Density(MDD) and Optimum Moisture Content(OMC) in each blend mix ratio using High Compactive Effort (HCE) of approximately 2470kJ/m3 and is equivalent to AASHO method T180. Moulds were prepared using High Compactive Effort (HCE) varying moisture contents in each mould per sample. These moulds were weighed and the dry density of the compacted material was calculated using the equations belowCITATION Min89 l 12297 (Ministry of Transport: Part N, 1989).
Yw TA l “YW: Wet Density” s “Yw” c 1 =m2-m1vm…………………….Equation STYLEREF 1 s 3. SEQ Equation * ARABIC s 1 3
Where: Yw – is the wet density of the compacted material (kg/m3).

m1 – is the mass of the mould (kg).

m2 – is the mass of mould + material (kg).

vm – is the volume of mould in m3.

Yd TA l “Yd:Dry density” s “Yd” c 1 =Yw×100100+w …………………….Equation STYLEREF 1 s 3. SEQ Equation * ARABIC s 1 4
Where: Yw – is the wet density of the compacted material (kg/m3).

Yd – is the dry density of the compacted material (kg/m3).

w TA l “w: moisture content” s “w” c 1 – is the percentage moisture content in the material
The selection of a suitable moisture content was aided by plotting a rough dry density/moisture content curve.
The moisture content at the maximum of the parabolic curve was recorded and the density that corresponded to this moisture content was also recorded for each of the blend mix. These values were recorded as Maximum Dry Density(MDD TA l “MDD: Maximum Dry Density” s “MDD” c 1 ) and Optimum Moisture Content (OMC TA l “OMC: Optimum Moisture Content” s “OMC” c 1 ).
CBR tests and Optimal blend mixCBR (California Bearing Ratio) TestThe CBR test was done on each and every mix blend tested thus, 100:0, 90:10, 70:30 and 50:50. It was done in accordance to method A8 of TMH1 of South African Standard Testing method, 1986. The test was used to establish the relationship between CBR and Density as well as the percentage swell.

Approximately 25kg was used to conduct the tests and were thoroughly mixed by means of the quartering procedure. For the CBR, 21kg was used and two representative samples each weighing 500g were taken for the determination of the hygroscopic moisture content. Three moulds were prepared according to method A7 of TMH1 of South African Standard Testing method, 1986 and were compacted at different compactive efforts for five layers in each mould. The moulds represented MoD, NRB and Proctor density. The mould representing MoD was compacted using a 4,536 kg hammer at 55 blows per layer whereas the mould for NRB was compacted using the same compactive effort thus 4.536 hammer but at 25 blows per layer. The mould for Proctor was compacted using a 2.495kg hammer at 55 layers per layer. These moulds were soaked in water for 4days and penetration due to a 5.56kg surcharge weight in each mould was determined. The calculations were done using the following equations.
W=z(y-x)100+x …………………………Equation STYLEREF 1 s 3. SEQ Equation * ARABIC s 1 5
Where,
W = amount of water to be admixed
x = hygroscopic moisture content
y = required moisture content
z = mass of air dried sample
For the Swell percent the following equation was used:
s TA l “s: Swell of material” s “s” c 1 =(k-L)127×100……………………………..Equation STYLEREF 1 s 3. SEQ Equation * ARABIC s 1 6
Where,
S= swell expressed as a percentage of the height of the moulded material before soaking that is 127mm.

k= dial gauge reading after 4days of soaking
L= dial gauge before soaking
Using the zero, the load was read off at 2.54mm, 5.08mm and 7.62mm etc. The CBR found at these depths were expressed as percentages of the California Standard for that penetration shown in table 3.2.

Table STYLEREF 1 s 3. SEQ Table * ARABIC s 1 2: Surcharge Penetration depth(mm) with their respective California Standard (kN) CITATION TMH86 l 12297 (TMH1, 1986).

Penetration(mm) California Standard (kN)
2.54 13.344
5.08 20.016
7.62 25.354
The CBR at 2.54mm is generally used for assessing the quality of material CITATION TMH86 l 12297 (TMH1, 1986). The CBR values at this penetration depth was recorded for each mix blend and was plotted on a logarithmic scale for the three different compactive efforts used thus, High Compactive Effort (HCE TA l “HCE:High Compactive Effort” s “HCE” c 1 ), Intermediate Compactive Effort (ICE TA l “ICE: Intermidiate Compactive Effort” s “ICE” c 1 ) and Low Compactive Effort (LCE TA l “LCE: Low Compactive Effort” s “LCE” c 1 ).

Optimal blend mix
According to Texas Department of Transport (TxDOT) specifications, RAP can be limited to 50%. To prove this, the TxDOT team used FDR material at a constant design cement content but at varying percentages of RAP to find that percentage of RAP that fail to meet the minimum strength criterionCITATION Tom12 l 12297 (Sebesta, 2012). It is approximated that about 30% material cost savings could be realised if a 50:50 blend of RAP and virgin aggregate is used CITATION Edw14 l 12297 (Hoppe, 2014). If 60% of natural aggregate is used in a mixture of RAP: virgin aggregate in road reconstruction operations, a lot of economic and environmental benefits are derived (Abbas, 2013). In this project a mix blend ratio of 50:50 will be used for design purposes as well as costing of the FDR project. This ratio was chosen because it met the minimum CBR criterion required for selected layer construction thus base reconstruction. According to CITATION SAT01 l 12297 (SATTC, 2001), for selected layer construction, the minimum CBR required is 15% at 93% MOD AASHTO of the modified AASHTO MDD test method (T-180) at the highest anticipated moisture.

TrafficTraffic count data is required for the estimation of design traffic load of the pavement. The traffic load is usually predicted by carrying out a 24 hour, 7-day axle load survey (Parry, 1993) on a similar type of road to be designed. From this survey the Average Annual Daily Traffic (AADT) can now be computed and the total number of expected vehicles which are going to use the pavement throughout its design life can be computed by multiplying the AADT by 365/year and further multiplying by the number of years which is equal to the design life of the pavement. In this project, a secondary road(Chivi-Mandamabwe) that links a primary road(Masvingo-Mbabalala) and a regional road(Masvingo-Beitbridge) and is generally a low volume road. According to (Douglas, 2004) a low volume road has traffic volume less than or equal to 400 vehicles per day hence an AADT equal to 400 is going to be used for design of the pavements in this study and of this AADT the percentage of heavy vehicles was taken as 15%.

Cost AnalysisA cost analysis was done so as to come up with the costs of construction of the conventional methods used in flexible pavement rehabilitation and Full Depth Reclamation (FDR). The cost analysis compared one kilometre length road of width of 7m on both rehabilitation techniques.
Methodology
The work to be done per kilometre length will be quantified using the base layer thickness on design details for both types rehabilitation techniques. For each rehabilitation technique, an item rate built up will be prepared and several factors will be considered and these shall be specified in the item rate built up. As a summary of the item rate built up, a priced bill of quantities for both methods will. The quantified work and the rates will now be used to come up with a priced bill of quantities for both methods will be prepared. The duration for conventional rehabilitation on a 1km will be assumed to be 1month and normal working time of 9 hours per day will be used for costing labour.
For traffic signalling, the costing was based on new traffic signalling adopted in Zimbabwe thus SATTC codes. Codes for each sign were provided in the item rate build up for easy identification by the supplier. For costing the aggregates for surfacing the nominal rates given in table 4403/1 of CITATION SAT011 l 12297 (SATTC, 2001)shall be used for costing since they are used for tendering purposes.

Table STYLEREF 1 s 3. SEQ Table * ARABIC s 1 3: Nominal rates of application ,Table 4403/1 CITATION SAT011 l 12297 (SATTC, 2001)Nominal size of aggregates (mm) Nominal rates of application
Tack coat (litres of tar or netbitumen per m2) Aggregate(m3 per m2)
19 1.80 0.015
13.2 1.50 0.010
9.5 1.10 0.007
6.5 0.80 0.005
For the costing analysis of the FDR project, portable reclamation was considered since it is considered fast as the machinery used thus Asphalt Zipper is less costly compared to heavy reclamation machines used in conventional rehabilitation. The duration for the FDR was assumed at 2 weeks.
: RESULTS ANALYSIS AND DISCUSSIONChapter OverviewThis chapter will present laboratory testing and the results analysis of the collected data.

Test ResultsThe results of gradation, Atterberg limits testing, Density testing and CBR testing will be presented in the following sections of this chapter.

Sieve analysisA sieve analysis test was carried out on each mix blend with aim of finding the particle size distribution of each mix as well as the class of each mix basing on the Grading Modulus. The grading modulus indicates the amount of fines retained in the following sieves of sizes 2mm, 0.425mm and 0.075mm. Equation 3.1 was used to calculate the Grading Modulus.
Results in table 4.1 shows the Grading Modulus for all blend mix ratios. The GM values for all the mix blends lied within the acceptable standards required for base materials thus, 2.6 ? GM ? 1.2 CITATION COL09 l 12297 (COLTO, 2009). A lower value of GM=1.87 was noticed in the 50:50 mix blend but however, it lied in the given range as well as the grading envelope required for base material. Adding RAP to neat material is a form of mechanical stabilisation and it gives desired gradations. The trend observed was that, as the percentage of RAP in in a blend mix increases, the Grading Modulus(GM) decrease due to the decrease in fine content. A material with more fines gets easily lubricated if water is added to it in such a way that enhances formation of a densely packed material if compacted thus, a soil mass that is more workable and gives a better packing CITATION Aro04 l 12297 (Arora, 2004). The results in table 4.1 were plotted as shown in figure 4.1.
Table STYLEREF 1 s 4. SEQ Table * ARABIC s 1 1: Particle size distributions for different mix blends
Sieve Sizes
(mm) Percent Passing (%)
Mix Blends ( Natural aggregate: RAP)
100:0 90:10 70:30 50:50
53 – – – –
37.5 – – – –
26.5 100.0 100.0 100 100
19 92.0 94.0 92.0 93.7
13.2 77.0 82.0 84.0 81.1
4.75 64.0 67.0 69.0 68.1
2 28.9 30.0 40.2 48.0
SF:0.425 24.8 25.2 28.1 37
0.075 15.2 16.4 20.9 28
B:0.075 0.0 0.0 0.0 0.0
Grading Modulus 2.31 2.28 2.1 1.87

Figure STYLEREF 1 s 4. SEQ Figure * ARABIC s 1 1: Graph showing how Grading Modulus(GM) vary with the increase of RAP% in each blend mix.

Plasticity IndexTable 4.2 shows results that were obtained from the Atterberg limits tests done on each mix blend and Figure 4.2 is a graphical illustration of the results in table 4.2.

Table STYLEREF 1 s 4. SEQ Table * ARABIC s 1 2: Plasticity Index results for all the mix blends
Mix Blend
( Virgin Aggregate: RAP) Percentage of RAP Plasticity Index (PI)
100:0 0 10
90:10 10 9
70:30 30 7
50:50 50 5

Figure STYLEREF 1 s 4. SEQ Figure * ARABIC s 1 2: Plasticity index versus %RAP, 2018.

The results in table 4.2 show that as the percentage of RAP increases the resultant blend mix become less plastic since the introduction of RAP is mechanical stabilisation of the resultant mix. The strength of the mix blend decreases as the proportion of RAP increases in the blend mix but however, due to the mechanical stabilisation the Plasticity Index of the resultant mix decreased from neat material to the 50:50 mix blend.
According to Asphalt Recycling and Reclaiming Association (ARRA, 2015), mechanical stabilisation is defined as the pulverization, mixing, and densification of reclaimed materials with the addition of granular materials, if necessary, to produce the required degree of structural support. All the blend mixes, the Plasticity Index (Ip) decreased to values within the construction standards for bases CITATION COL09 l 12297 (COLTO, 2009) thus, Ip < 12 which makes all the blend mixes qualify for base material. According to CITATION Aro04 l 12297 (Arora, 2004) , base material should have low plasticity index to avoid consequences of excessive accumulation of water as well as loss of strength.
CompactionTable 4.3 shows Optimum Moisture Content (OMC) and Maximum Dry Density(MDD) values obtained for each mix blend using modified proctor energy and figure 4.3 and 4.4 are graphical representation of OMC and MDD versus RAP content respectively. The trend observed in tested specimens shows that an increment in RAP content in a mix blends follows a decrement in OMC and MDD values. This decrement is due to low fine content in blend mix specimens of high content of RAP. Addition of more RAP to neat material made the resultant mixes coarser which implied that, the grains adsorbed less water hence there was less lubrication of particles which facilitates better compaction. The fine content plays the most essential part in the densification of the resultant mix thus the higher the density the higher the bearing capacity of the material if used for base construction. RAP aggregates are encased in asphalt cement and the presence of Asphalt cement leads to the reduction of water needed to achieve the required MDD CITATION Coo05 l 12297 (Cooley, 2005). All the blend mixes proved to be workable as base material they can be compacted to desired strengths.Table STYLEREF 1 s 4. SEQ Table * ARABIC s 1 3: Compaction characteristics of all blend mixes, 2018.

Blend Mix Ratio Virgin
Material(%) RAP Content
(%) MDD
(Kg/m3) OMC
(%)
100:0 100 0 2175 10.8
90:10 90 10 2028 9.1
70:30 70 30 1960 8.8
50:50 50 50 1940 7.6

Figure STYLEREF 1 s 4. SEQ Figure * ARABIC s 1 3: Optimum Moisture Contents(OMC), 2018.

Figure STYLEREF 1 s 4. SEQ Figure * ARABIC s 1 4: Maximum Dry Densities, 2018
Soaked CBR ResultsTable 4.4 shows the values for the CBR obtained for each mix blend at different compactive efforts thus, 1,2,3 for High Compactive Effort (HCE), Intermediate Compactive Effort (ICE) and Low Compactive Effort (LCE) respectively. Figure 4.5 is a graphical representation of the worst case scenario for CBR thus the values obtained using relative Low Compactive Effort (LCE). The trend observed in the CBR values was similar to that observed in Maximum Dry Density values since the CBR test is controlled by MDD and OMC. At varied compactive efforts, HCE gave better results on all blend mixes thus, higher values of CBR because the resultant compacted material gave better structural support or bearing capacity at high compactive efforts in all blend mixes. The strength decreased as the compactive was reduced to ICE and LCE. The CBR values decreased as the RAP content increased thus the strength of each resultant mixture decreased. According to CITATION SCR03 l 12297 (SCRB, 2003), increasing RAP in a blend mix decreases the mechanical strength of the resultant mix blend. According to CITATION TMH86 l 12297 (TMH1, 1986), the CBR value corresponding to the penetration depth of 2.54 is generally used for assessing the quality of material. The load readings are expressed as a percentage of the California standard corresponding to the penetration. For a penetration of 2.54 the California standard is 13.344kN.
The following equation was used for calculating the CBR of all the mix blends for the penetration of 2.54 according to CITATION TMH86 l 12297 (TMH1, 1986).

CBR=Load reading for penetration of 2.54Carlifornia standard×100…………………….Equation STYLEREF 1 s 4. SEQ Equation * ARABIC s 1 1

Table STYLEREF 1 s 4. SEQ Table * ARABIC s 1 4: CBR test results, 2018.

Virgin Material(%) RAP
(%) Relative Compaction(%) Specimens MDD
(kg/m3) OMC
(%) CBR
(%)
100 0 100.3 1 2182 10.8 39
95.2 2 2071 10.8 34
90.2 3 1962 10.7 29
90 10 100.3 1 2030 9.1 34
95.2 2 1929 9.1 29
94.2 3 1827 9.0 24
70 30 99.9 1 1958 8.8 30
94.9 2 1860 8.8 25
90.1 3 1766 8.7 20
50 50 99.9 1 1938 7.6 26
94.9 2 1841 7.6 16
90.1 3 1748 7.5 11

Figure STYLEREF 1 s 4. SEQ Figure * ARABIC s 1 5: CBR plot for the worst case scenario, Low Compactive Effort(LCE) for all mix blends, 2018.

Material Classification for Pavement DesignAll the tests done facilitated the classification of all the resultant mix blends but however much focus will be put on the optimal blend mix of ratio 50:50 (virgin material: RAP). This blend was considered optimum because it met the minimum criterion required as per selected layer construction thus a minimum CBR of 15% at 93% MOD AASHTO CITATION SAT01 l 12297 (SATTC, 2001). At 94.9% relative compactive effort, the CBR for the 50:50 blend was 16% at MDD of 1841kg/m3 and OMC of 7.6%. this CBR value will be taken as CBRdesign. The dependants to RAP content for a 50:50 mix blend are shown in the table 4.5.

Table STYLEREF 1 s 4. SEQ Table * ARABIC s 1 5: Classification for design of Mix Blend(50:50) using (TRH, 1996) and (COLTO, 2009), 2018.

Material Property 50:50 Standard Specifications
TRH14, 1986 COLTO, 2009
Grading Modulus, GM 1.87 ?2.5 ?2.7
Plasticity Index , PI 5 <12 <10
Relative Compaction 94.9% 93% 93%
Strength (CBR) 16% ?15 ?15
Swell 0.16% ?1.5% ?1.5%
Both standard specifications classify the optimal mix blend of 50:50 as G6 which has the following material description according to CITATION rh l 12297 (TRH04, 1986) and CITATION COL09 l 12297 (COLTO, 2009). G6 material, is natural gravel or a mixture of natural gravel and crushed boulders. This describes the mix blend. Soaked CBR and Atterberg limits are mainly used to classify the material in addition to the above material properties tabulated in table 4.5.
According to Ministry of TransportCITATION Par98 l 12297 (Ministry of Transport :Part F, 1989) any soil with a CBRdesign of 9 or greater is considered as SG9 and on site the upper 150mm is usually compacted at 93% Mod AASHTO. The optimal mix blend was classified as SG9 since it has CBRdesign>9. Therefore, the pavement design will be based on SG9 material.

Symbolisation of Layer TreatmentEach material class has a corresponding treatment that makes it form a satisfactorily pavement foundation and the treatment is symbolised. The material in section 4.3 of this chapter also require a treatment and this treatment shall be specified on the pavement drawing for easy interpretation of the construction drawing. T5 is a treatment that is normally applied to SG5 material or betterCITATION Par98 l 12297 (Ministry of Transport :Part F, 1989). Therefore, the material in section 4.3 will be given a treatment symbolised as T5 since it is classified as SG9 according to Ministry of Transport CITATION Par98 l 12297 (Ministry of Transport :Part F, 1989) because the material has a CBR>9 thus 16>9.

: PAVEMENT DESIGNPavement StandardThe standard of a pavement is determined by the number of E80 axles that a pavement is expected to carry throughout its design lifeCITATION Par98 l 12297 (Ministry of Transport :Part F, 1989). Table 5.1 shows pavement standards used in Zimbabwe.

Table STYLEREF 1 s 5. SEQ Table * ARABIC s 1 1: Pavement Standards with their respective design traffic rangesCITATION Par98 l 12297 (Ministry of Transport :Part F, 1989)Pavement Standard Design (E80)
3M 1 to 3×106
1M 0.3 to 1×106
0.1M 0.05 to 0.1×106
0.05M Less 0.05×106
Pavement layersPavement layers are numbered as 1, 2, 3 etc. and will be referred to as bases in the pavement drawing provided in the appendix section. These layers were defined by their position in the pavement structure starting with base 1 as the uppermost layer followed by base 2. The minimum layer thickness corresponding to a pavement standard with suffix, M representing Million Standard Axles will be given in table 5.1.

Table STYLEREF 1 s 5. SEQ Table * ARABIC s 1 2: Minimum compacted layer thickness for pavement standardsCITATION Par98 l 12297 (Ministry of Transport :Part F, 1989)Pavement Standard Minimum base thickness (mm) Minimum Total
Pavement Thickness
Base 1 Base 2 3M 150 150 450
1M 150 120 270
0.3M 120 120 250
0.1M 120 120 150
0.05M 120 120 120
: Design CalculationBoth the Conventional rehabilitation pavement and Full Depth Reclamation Pavements are going to be designed using the Ministry of Transport pavement design charts (1989).
left1905NATIONAL UNIVERSITY OF SCIENCE AND TECHNOLOGY
CALCULATION SHEET
TITLE: Evaluation of reclaimed asphalt pavement as constituent base course material. A case study of Chivi-Mandamabwe road.

PROJECT SECTION: Flexible pavement design: Conventional Rehabilitation Sheet No. 1
Designed by: Mwoyosvi David. K
Checked by: Mr T.C. Mdlongwa
Date:
REF Designing a Flexible Pavement :Conventional Rehabilitation OUTPUT
Ministry
Of
Transport
Part F, Construction
Manual Design parameters for a flexible pavement
a) The CBRdesign=16 calculated from chapter 4 will be used. From the soil analysis done this design will be based on SG9 soils.
b) Calculating the design traffic load in ESA
Assumptions
i) From the methodology chapter, AADT =400
ii) Design period is 20 years
iii) Traffic growth rate is 4 %
iv) 15% of the vehicles in the traffic stream are heavy vehicles
v) The road is a two way with one lane in each direction i.e. DF is 0.5
vi) The lane distribution factor is 1
Average ESA for heavy vehicles is 1.5ESA
For a design life of 20 years the cumulative growth factor (CGF)
CFG=1+0.01RP-10.01RWhere P is the design life and R is the traffic growth rate
CGF =1+0.01×420-10.01×4 =29.8

Now calculating the design traffic load NDT
NDT =365xAADTxDFx%ageHVxLDFxCGFxNHVAG
Where
NDT- Design Traffic Load in ESA
AADT TA l “AADT: Average Annual Daily Traffic” s “AADT” c 1 – Average Annual Daily Traffic
D TA l “D: Directional split” s “D” c 1 – Directional split (2-way DF)
R- Traffic Growth Rate as a percentage
%age HV-Proportion of heavy Vehicles
NHVGA- Average ESA’s
P- Design life of the pavement
LDF TA l “LDF: Lane Distribution Factor” s “LDF” c 1 – Lane distribution factor
CGF TA l “CGF: Cummulative Growth Factor” s “CGF” c 1 – Cumulative Growth factor
NDT = 365x400x0.5×0.15x1x29.8×1.5
= 489 465 ESA’s
= 0.49MESA’s
= 0.49M
Structural Design
Since NDT lies in the range, 0.3M TA l “M: Pavement standard in million standard axles” s “M” c 1 ; NDT; 1M
Therefore, we assume a 1M pavement standard using the charts for 1M pavements we have:
Design summary
Subgrade Classification: SG9 TA l “SG9: Material class of CBR>9” s “SG9” c 1
Subgrade Treatment : T5 TA l “T5: Pavement treatment procedure” s “T5” c 1
Pavement Layer Base 1 Base 2
Thickness (mm) 150 150
Class Class 2.6 Class 3.3
Therefore, provide;
Wearing course 50mm
Base course (Base 1) 150mm
Subbase course (Base 2) 150mm
50mm hot mix asphalt layer can be reduced to 40mm where local experience shows this to be adequate CITATION SAT01 l 12297 (SATTC, 2001)OK.

left1905NATIONAL UNIVERSITY OF SCIENCE AND TECHNOLOGY
CALCULATION SHEET
TITLE: Evaluation of reclaimed asphalt pavement as constituent base course material. A case study of Chivi-Mandamabwe road.

PROJECT SECTION: Flexible pavement design: Full Depth Reclamation(FDR) Sheet No. 2
Designed by: Mwoyosvi David. K
Checked by: Mr T.C. Mdlongwa
Date:
REF Designing a Flexible Pavement : Full Depth Reclamation(FDR) OUTPUT
Ministry
Of
Transport
Part F, Construction
Manual a) The CBRdesign=16 calculated from chapter 4 will be used. From the soil analysis done this design will be based on SG9 soils.
b) Calculating the design traffic load in ESA
Assumptions
i) From the methodology chapter, AADT =400
ii) Design period is 10 years
iii) Traffic growth rate is 4 %
iv) 15% of the vehicles in the traffic stream are heavy vehicles
v) The road is a two way with one lane in each direction i.e. DF is 0.5
vi) The lane distribution factor is 1
Average ESA for heavy vehicles is 1.5ESA
For a design life of 20 years the cumulative growth factor (CGF)
CFG=1+0.01RP-10.01RWhere P is the design life and R is the traffic growth rate
CGF =1+0.01×410-10.01×4 =12

Now calculating the design traffic load NDT
NDT =365xAADTxDFx%ageHVxLDFxCGFxNHVAG
Where
NDT- Design Traffic Load in ESA
AADT- Average Annual Daily Traffic
D- Directional split (2-way DF)
R- Traffic Growth Rate as a percentage
%age HV-Proportion of heavy Vehicles
NHVGA- Average ESA’s
P TA l “P:Design Life of a pavement” s “P” c 1 – Design life of the pavement
LDF- Lane distribution factor
CGF- Cumulative Growth factor
NDT = 365x400x0.5×0.15x1x12x1.5
= 197 200 ESA’s
= 0.197MESA’s
= 0.20M
Structural Design
Since NDT lies in the range, 0.1M < NDT< 0.3M
Therefore, we assume a 0.3M pavement standard using the charts for 0.3M pavements we have:
Design summary
Subgrade Classification: SG9
Subgrade Treatment : T5
Pavement Layer Base 1 Base 2
Thickness (mm) 120 120
Class Class 2.8 Class 3.6
Therefore, provide;
Wearing course 50mm
Base course (Base 1) 120mm
Subbase course (Base 2) 120mm
50mm hot mix asphalt layer can be reduced to 40mm where local experience shows this to be adequate CITATION SAT01 l 12297 (SATTC, 2001)OK
: Bill of Quantities
: CONCLUSION AND RECOMENTATIONS
MATERIAL TESTING RESULTS

PAVEMENT DESIGN CHARTS

DRAWINGS

ITEM RATE BUILT UP

REFERENCES BIBLIOGRAPHY Arora, D. K., 2004. Soil Mechanics and Foundation Engineering, Delhi: Standard Publishers Distributors.

California(DOT), 2013. Division of Maintenance Pavement Program, s.l.: s.n.

COLTO, 2009. Standards Specifications for Road Construction Material. In: s.l.:MATRO LAB.

Cooley, D. A., 2005. “Effects of Reclaimed Asphalt Pavement on Mechanical Properties of Base Materials”, s.l.: s.n.

Copeland, A., 2011. Reclaimed Asphalt Pavement in Asphalt Mixtures: State of the Practice, s.l.: s.n.

D.Reeder, G., 2017. Guide to Full-Depth Reclamation (FDR) with cement, s.l.: s.n.

Division ,Research & Development, 2013. Pavement Design for Carriageway Construction. HONGKONG: s.n.

Harrington, D. & Adaska, W., 2011. Full-Depth Reclamation of Asphalt Pavements with Cement, s.l.: s.n.

Hoppe, E. J., 2014. Feasibility of Reclaimed Asphalt Pavement (RAP) Use As Road Base and Subbase Material, Virginia: s.n.

Kandhal & Mallick, 1997. s.l., s.n.

KentCountyCouncil, 2000. Road Pavement Design. s.l.:s.n.

Mathew, T. V., 2006. Transportation Engineering I. Mumbai: s.n.

McDaniel, R. & Anderson, R. M., 2001. NCHRP REPORT 452 :Recommended use of Reclaimed Asphalt Pavement in Superpave Mix Design Method: Technician’s Manual, Washington, D.C: National Academy Press.

McGarrah, 2007. Evaluation of Current Practices of Reclaimed Asphalt Pavement/Virgin Aggregate as Base Course Material, Washington D.C: s.n.

Ministry of Transport :Part F, 1989. Construction Manual, s.l.: s.n.

Ministry of Transport: Part N, 1989. Part N: Material Testing, s.l.: ARUP.

Pennsylvania(DOT), 2012. Developing Standards and Specifications for Full Depth Pavement Reclamation, s.l.: s.n.

SAPEM, (. A. P. E. M., 2014. Chapter 3, Materials Testing, s.l.: SOUTH AFRICAN NATIONAL ROADS AGENCY SOC LTD.

SATTC, 2001. Code of Practice for the Design of Road Pavements. s.l.:s.n.

SATTC, 2001. Standard Specifications for Roads and Bridges, s.l.: Division of Roads and Transport Technology, CSIR.

SCRB, 2003. State Corporation of Roads and Bridges, “Highway Design Manual”, Republic of Iraq.: s.n.

Sebesta, T. S. S., 2012. FULL-DEPTH RECLAMATION: NEW TEST PROCEDURES AND RECOMMENDED UPDATES TO SPECIFICATIONS, Texas: s.n.

Sultan, S. A., 2013. IMPROVEMENT OF THE MECHANICAL CHARACTERISTICS OF RECLAIMED ASPHALT PAVEMENT IN IRAQ. November .Volume Vol. 2, No. 4.

Taha, R. et al., 2002. Cement Stabilization of Reclaimed Asphalt Pavement Aggregate for Road Bases and Subbases. JOURNAL OF MATERIALS IN CIVIL ENGINEERING, pp. 239-245..

TMH1, 1986. Standard methods of testing road construction materials, s.l.: s.n.

TRH04, 1986. Guidlines for Road Construction Materials, s.l.: s.n.

TRH04, 1996. Structural design af flexible pavements, Pretoria: s.n.

WirtgenGmbH, 2010. Wirtgen Cold Recycling Technology. Windhagen: s.n.