Performance characteristics of pavements are obtained through laboratory-based materials testing in conjunction with field testing. The RLT test is used to evaluate pavement fatigue performance in the laboratory. Although the RLT test has been shown to be an effective tool used in predicting pavement performance, the methods of compaction used in laboratories globally to prepare specimens lack appropriate test standard repeatability and reliability and therefore correlation with field compaction. This can lead to serious consequences regarding specifications for field basecourse compaction and performance and can mislead engineers who rely principally on the results of laboratory tests for making decisions concerning the field compaction process.
Compaction increases the density of a material, increasing its load bearing capacity, reducing consolidation and providing stability. This chapter is aimed at comparing the effectiveness of kneading and vibratory compaction for RLT specimens and exploring variables in vibratory compaction that can be manipulated to increase the likelihood of specimen reproducibility. Empirical compacting procedures and equipment have been used in the past due to the simplicity of use and economic considerations. However, methods such as the Modified AASHTO compaction test procedure and NZS 4402 Test 4.1.3 do not accurately simulate field compaction (Chilukwa, 2013, Kelfkens, 2008).
A poorly compacted layer is characterised by Chilukwa (2013) as having a lower density, higher porosity and lower cyclic stiffness in comparison to a properly compacted layer. Poor compaction can result in rutting and eventually lead to water ingress in the pavement layer, which results in early failure of the pavement. The porosity of a layer is dependent on the degree of compaction, the size and shape of particles and dry density of aggregates which make up the layer. The PSD of the layer determines the distribution of sizes of aggregates within the layer. The key performance indicators for the compaction research will be the PSD and dry density of specimens.
Equation 4-1: Maximum Dry Density (MDD).
MDD (kg/m3) = Weight of dry compacted material (kg) / Volume of specimen (m3)
Particle Size Distribution
The stability of an in situ unbound granular layer is derived mainly from particle interlock and inter-particle friction, particularly for permeable basecourse layers, which have considerably lower fines than a typical basecourse layer. The PSD is, therefore, an important characteristic for strength and stiffness determination. The maximum particle size of an unbound layer is limited to the layer thickness due to roller compaction only being effective up to a depth of approximately 250 mm. Layer workability gets difficult if stones exceeding 30% of the layer thickness are present within the layer (Thom, 2008). Particle size and gradation are significant factors to be considered when constructing granular pavement layers. Salt (2011) recommends, the PSD of basecourse material tested in RLT tests and compacted in the field needs to be closely controlled for consistency and to avoid unpredictable performance, even though this is not common practice in New Zealand. A change in PSD has an influence in the degree of saturation and permeability of the basecourse, which has a significant influence on its performance (Salt, 2011). Salt (2011) goes on to say that contractors must conduct pre- and post-compaction PSD of basecourse layers in the field to minimise changes in PSD that can hinder pavement performance.
The density and hence stability of a pavement layer is influenced by the extent of fines within it. A greater fines content can lead to layer instability when traffic loading is applied. In the case of vibratory compaction, high energy compaction can lead to aggregate crushing and disintegration of larger aggregates, producing greater proportions of fines than expected. This, in turn, produces unpredictable results in laboratory-based tests and accelerated pavement deterioration in the field. Unrealistically high Maximum Dry Densities (MDDs) can be achieved in the laboratory due to the unrealistic lateral confinement conditions of the specimen mould.
In Figure 4-1‘case c’, a greater extent of fines can lead to layer instability when traffic loading is applied. Particle degradation in the laboratory can result from a combination of excessive force exerted to compact basecourse specimens, long compaction duration, and restricting aggregate re-orientation by using small diameter compaction moulds. Addition of fines also results in the plasticity of the material having a more profound effect on performance.
Chilukwa (2013) identified a 5% increase in fines (material passing 0.075mm sieve) when vibratory compaction was used to compact scalped (19mm down) G3 (high-quality basecourse in South Africa) material. However, only one test of pre and post-compaction PSD was undertaken. More replicate testing is required in order to provide a statistically robust conclusion. When the larger G7 (maximum particle size of 37.5mm) material was tested for PSD before and after compaction, there was a reduction of 18% of fines post vibratory compaction due to cementation of fines.
The PSD of material can be characterised by using Talbot’s n value. Talbot’s exponent is a measure of the degree to which the coarse particles float in a matrix of fines. Values less than 0.3 give unacceptable stability for unbound granular layers, whereas values greater than 0.4 are normally associated with good stability. Values higher than 0.6 become difficult to construct without causing segregation (Salt, 2011).
The PSD of the material used for basecourse construction in New Zealand should conform to the envelope limits defined in Tables 4-1 and 4-2 below.
Where an AP40 aggregate has a much higher than minimum crushing resistance, the standard grading (which includes provision for some gradation change during compaction), may accept materials that appear deficient in fines during construction. In such a case, it is appropriate to vary the specified grading by adding an amending clause to the job specification (TNZ(b), 2006). This is true for the permeable basecourse used in this research. The aggregates used have a very high crushing resistance as shown in Chapter 3: Section 2.1.1.
The PSD test was conducted to determine aggregate grading and compliance to M/4 envelopes defined by Transit NZ (TNZ, 2006).
Sieve analysis is used to determine the PSD of soils or granular materials. In the present study, sieve analysis was done following the procedures specified in NZS 4402: 1986 Test 6.1.1 and methods discussed by Head (1984). The methods of conducting sieve analysis are previously explained in Chapter 3: Section 3.0. Graphing the cumulative percentage of solids passing each sieve on a log graph enabled Talbot’s n value to be deduced.
Current M/4 specifications do not provide restrictions on the PSD of the basecourse once it has been compacted (Alabaster, 2012). Field compaction techniques implemented in New Zealand using high energy vibratory rollers are detrimental towards the aggregates, especially when target densities are difficult to satisfy. High energy compaction alters the PSD of the basecourse by crushing the larger aggregates, especially when crushing resistance is low. In the present study, degradation of aggregates was witnessed during laboratory compaction using high-pressure kneading, static, and vibratory compaction. This leads to unpredictable basecourse behaviour and facilitates premature basecourse failure.
Influence of Fines
Shackel (2006a) tested granular basecourse materials both as delivered and after removal of the material finer than either 0.600 mm or 1.18 mm. The gradations from which the 1.18 mm and 0.600 mm had been scalped were also tested after removal of particles larger than 13.2 mm. Scalping out the fines led to reductions in both the modified MDD and the corresponding OMC irrespective of the maximum particle size (Shackel, 2006a). The materials with fines smaller than 1.18mm removed, exhibited permeability almost 100 times greater than the unmodified material.
For the basecourse material used by Shackel (2001), fines were removed. It was evident that simple measures such as scalping out fines could greatly increase the permeability of basecourse materials. However, the influence of fines removal on the mechanical properties of the material needed further investigation. The resilient modulus, Mr, was selected by Shackel (2001) as the parameter that would best describe the mechanical properties of the materials tested.
The response of the materials to repeated triaxial loading depended on the degree of saturation during the test and the PSD. Irrespective of the repeated stress levels, the resilient modulus decreased with increase in the degree of saturation. Typically, an increase in saturation led to reductions in Mr between 40% and 70%, depending on the gradation and maximum particle size.
For the materials used in the study by Shackel (2001), Mr, increased with increase in Cu, the undrained strength of the material. The tests showed that the permeability of a typical crushed rock base material could be significantly increased by scalping out the finer fractions of the material. This was accompanied by a reduction in resilient modulus. Shackel (2001) suggested that removal of material smaller than 1.18mm reduced Mr by approximately 30% to 55%, whereas scalping material smaller than 0.600mm caused modulus reductions between 20% to 45%. The choice of unbound material for the permeable base is typically a compromise between permeability and modulus or structural capacity.
Methods of Laboratory Compaction
Compaction is a key process in the construction of pavement layers. It is significant in ensuring the structural integrity of the layer and has an influence on the engineering properties and performance of the pavement layer. It is important that field and laboratory compaction is undertaken correctly. For this reason, laboratory compaction methods have been developed to simulate the field compaction process in the laboratory.
There are five popular modes of compaction in the laboratory:
Gyratory Compaction, and
Laboratories that perform cyclic triaxial tests on unbound granular materials occasionally develop their own methods for specimen compaction and preparation. Hoff et al. (2004) evaluated the influence of gyratory compaction, modified Proctor impact hammer, vibratory hammer and vibratory table on the resilient modulus and permanent deformation of unbound granular material using cyclic loading. Triaxial cyclic loading was undertaken at a constant confining pressure and varying deviatoric stress per stage. Testing was conducted at a frequency of 10Hz. The results of the research showed that no systematic difference was found for the resilient modulus as a result of the different methods of compaction. However, significant differences in permanent deformation were found between specimens compacted to the same density, using different methods. Specimens compacted using a vibratory compaction method showed about 20 – 25 % higher resistance to incremental failure and 40 – 50 % higher limit values for purely elastic behaviour in comparison to specimens compacted using the modified Proctor impact hammer (Hoff, 2004).
For granular soils used in this research, impact compaction is inappropriate due to the cohesionless nature of the soil. During impact compaction, the particles have a tendency to dislodge following load drops, resulting in low dry densities (Shahin, 2011, Chilukwa, 2013, Kelfkens, 2008).
In New Zealand, when compacting granular cohesionless material, the modified AASHTO compaction method has been replaced by the vibratory hammer compaction method. The primary reason is that New Zealand has softer aggregates due to its “youthful” geological position in the world, and the heavy dynamic compaction of the modified AASHTO compaction method causes a change in the grading. This influences the MDD and OMC. The vibratory hammer compaction method was used to determine the moisture-density relationship of the material. Preliminary compaction of specimens used for determination of OMC in the present study indicated that vibratory compaction also causes changes in PSD of material, as shown by Figures 4-2 and 4-3. Some of the variations in PSD which is shown in Figures 4-2 and 4-3 can also be attributed to the compaction duration and confinement conditions of the specimens, which are further discussed in Sections 4.4 and 4.5 of this chapter respectively.
The degree of compaction achieved by the vibratory and modified AASHTO compaction methods was evaluated. The results, which are presented in Chapter 3: Section 5.1.1, showed that vibratory compaction could achieve a greater degree of compaction than the modified AASHTO or Proctor compaction method. Increased degree of compaction has the following effects on basecourse performance (van Nierkerk, 2002):
Influences cohesion values (especially at compaction levels greater than 100%), and
Has a positive influence on the magnitude of the resilient modulus.
Van Nierkerk (2002) goes on to say that gradation and composition of a granular base have an influence on pavement performance but are secondary in comparison to the influence of stress conditions, the degree of compaction and moisture content.
The influence of density, as described by the degree of compaction, has been regarded in previous studies as being significantly important in determining the long-term behaviour of granular materials (Barksdale, 1972, Thom, 1988c). Resistance to permanent deformation under repetitive loading is improved due to an increase in density. Barksdale (1972) studied the behaviour of several granular materials and observed an average of 185% more permanent axial strain when the material was compacted at 95% instead of 100% of maximum dry density.
The influence of grading was also studied by Thom and Brown (1988b) and was found to vary with the degree of compaction. When un-compacted, the specimens with uniform grading resulted in the least permanent strain. The resistance to plastic strain was similar for all gradings when the specimens were heavily compacted. The influence of the degree of compaction on strain is presented in Figure 4-4 below.
Conclusions from Thom and Brown (1988b) were disputed by Dawson (1996), who found the influence of grading on permanent deformation to be more significant than the degree of compaction. The differences in findings between Thom and Brown (1988b) and Dawson (1996) may be related to the wide range of densities and gradings adopted by Thom and Brown, which exceeded the range typically expected in any pavement. The influence of fines content was investigated by Barksdale (1972) and Thom and Brown (1988b), who concluded that permanent deformation resistance in granular materials is reduced as the fines content increases. However, opinion is divided as to whether the degree of compaction or grading has a greater influence on permanent deformation. Researchers agree that both factors significantly influence the permanent strain of basecourse aggregates.
Laboratory vibratory compaction if preferred mainly due to its ease of use and efficiency in effectively compacting specimens for a relatively low cost (Araya, 2010, Hoff, 2004, Chilukwa, 2013). Common methods of field compaction comprise of vibratory compaction, and researchers believe that vibratory compaction in the laboratory yields a better correlation between field and laboratory results (Patrick, 2010, Arnold, 2004, Chamblin, 1962). Field validation of laboratory compaction is achieved if the compaction tests in the laboratory were conducted under conditions representative of field compaction.
Laboratory vibratory compaction can be achieved by using a vibrating hammer or a vibrating table. Chilukwa (2013) and Chamblin (1962) identified that the vibratory hammer is capable of producing densities comparable to those produced from vibratory table compaction. For the purposes of this study, vibratory hammer compaction is considered due to the unavailability of a vibrating table compaction setup.
Repeatability and reproducibility are important variables when conducting RLT tests as they can greatly influence test results (Toan, 1975). Eliminating as much variability as possible is highly recommended when comparing results between specimens.
Repeatability tests conducted by Chilukwa (2013) showed that, the vibratory hammer compaction method was effective in compacting graded crushed stone material and that vibratory hammer compaction did not result in significant material disintegration. However, Shahin (2011) states that the New Zealand vibrating hammer compaction test procedure has been proven to provide inconsistent results. Shahin (2011) found up to 20% variation in vibratory compaction MDDs. The variations can be caused by operator error, natural properties, hammer energy, water content, mould size, oversized particles, aggregate degradation and aggregate segregation.
Table of Contents
Table of Contents
List of Figures
List of Tables
1.2 Problem Statement
1.3 Research Objectives and Scope.
1.4 Thesis Organisation
2.0 Literature Review
2.2 Typical Impermeable Pavement Structure
2.3 Permeable Pavement Structure
2.4 Repeated Load Triaxial
3.0 Experimental Design
3.2 Aggregate Selection and Acquisition
3.3 Sieve Analysis
3.5 Repeated Load Triaxial Tests
3.6 Permeability Tests
4.2 Methods of Laboratory Compaction
4.3 Field Compaction
4.4 Specimen Compaction
5.0 Deformation Characteristics of Basecourse Material
5.2 Pilot Small-Scale Repeated Load Triaxial Tests
5.3 Large-Scale Repeated Load Triaxial Tests
6.0 Permeability Characteristics of Basecourse Material
6.2 Flow Through Roading Aggregates
6.3 Theoretical Considerations of Permeability
6.4 Permeability of Basecourse
6.5 Relative Permeability Using Repeated Load Triaxial Equipment
7.0 Effect of Scale on Repeated Load Triaxial Tests
7.2 Influence of Specimen Size on Repeated Load Triaxial Test
7.3 Effects of Aggregate Scalping on Specimen Performance
7.4 Influence of Specimen Density on Repeated Load Triaxial Tests
8.0 Conclusions and Recommendations
8.4 Repeated Load Triaxial Tests
8.5 Future Testing
8.6 Recommendations to New Zealand Materials and Testing Specifications
GET THE COMPLETE PROJECT
Investigation of The Properties of Unbound Granular Basecourse Aggregate Using Laboratory Tests