Recycled glass is a mixture of different coloured glass particles with debris. It is the result of crushing the waste glass collected from residential and industrial areas. Construction and demolition (C&D) materials such as concrete, excavation stone (basalt) and brick make up a significant proportion of the waste materials present in landfills in Australia.
{{image2-a:r-w:450}}Substituting the quarry-produced crushed rock with recycled materials in road pavement sub-base applications would significantly reduce the demand for landfill sites and would potentially provide an opportunity to use recycled material as aggregates in parts of Victoria where aggregates sources are becoming scarce.
This paper discusses the suitability of using blends of recycled glass and crushed rock as road pavement sub-base materials with mixtures of 10 per cent to 50 per cent by mass of recycled glass. The experimental works undertaken in this study includes basic classification tests along with modified compaction, California bearing ratio and Los Angeles abrasion tests to assess the suitability of the blends.
The research indicates that initially up to 15 per cent ?recycled glass with the maximum particle size of 4.75mm? could be safely added to Class 3 crushed rock. The degree of breakdown occurring in the recycled glass blend is on the limit of what would be acceptable for this material.
Depending on the results of future field trials, it may be possible to increase the percentage of recycled glass.
Waste materials have been defined as any type of material by-product of human and industrial activity that has no lasting value (Tam & Tam, 2006). The escalating quantities and type of waste materials, shortage of landfill spaces and the likely shortfall of premium aggregate materials in the near future imposes pressure and urgency on finding innovative ways of recycling and reusing waste materials.
The recycling and subsequent reuse of waste materials will also reduce the demand for virgin natural resources, which consequently leads to less energy usage, lower greenhouse gas emissions and, ultimately, a more sustainable environment (Disfani et al, 2009a).
{{image3-a:l-w:300}}In Victoria, the total amount of recovered waste material is recorded as 6.36 million tonnes in the 2006-07 financial year, with about 62 per cent of the solid wastes recycled over that period (Sustainability Victoria, 2007). Eighty-two per cent of the material received for reprocessing during 2006?07 was sourced from commercial and industrial and C&D industries, with C&D material accounting for 49 per cent of all material (by weight) recovered.
Among C&D material, concrete was the major component, representing 52 per cent (by weight) of the total, followed by rock/excavation stone (16 per cent), brick/brick rubble (14 per cent), soil/sand (eight per cent), asphalt (six per cent), mixed C&D waste (three per cent) and plasterboard (one per cent).
Recycling of construction waste materials would clearly provide substantial benefits to the industry in terms of reduced material supply and waste disposal cost, increased sustainability and reduced environmental impact (Sivakumar et al, 2004).
Recycled glass is a mixture of different coloured glass particles and often comprises a wide range of debris (mainly paper, plastic, gravel, metals and food wastes). The presence of different coloured glass and diverse types of debris is the primary obstacle in reusing recycled glass in the bottle production industry.
Previous work reported by Disfani et al (2009b) indicated that a recycled glass source with maximum particle size of 4.75mm does not carry its original shape and in many geotechnical engineering aspects resembles natural and quarried aggregates. Recycled glass particles are generally angular-shaped and contain some flat and elongated particles. The degree of angularity and the quantity of flat and elongated particles mainly depend on the crushing process (FHWA, 1998).
The waste stream from which the glass bottles or glass particles have been produced is the main factor that controls the quality of the material, especially the amount of debris in the mixture (Landris, 2007). The gradation curve and amount of foreign material in the recycled glass is partly determined by the type of machinery and procedures used in crushing and sieving waste glass.
Therefore, the geotechnical engineering behaviour of recycled glass varies from one supplier to another (Landris, 2007). This has led to varying results for tests on the geotechnical characteristics of recycled glass studied around the world (Disfani et al, 2009b).
Current experimental works reported by Disfani et al (2009b) on crushed recycled glass sources passing the 4.75mm sieve show satisfactory geotechnical characteristics regarding usage in specific roadwork applications. The results confirmed the lack of cohesion among particles as a principle consideration in the shear strength of recycled glass.
This is likely due to smooth surfaces of crushed glass particles and lack of fine clay-size particles in the mixture. Currently, more than 250,000 tonnes of this source of recycled glass is stockpiled annually in Victoria.
{{image4-a:c-w:600}}To enhance the shear strength properties of recycled glass and to overcome its deficiencies when used by itself, different proportions of recycled glass and crushed rock were mixed and their geotechnical characteristics were examined in an extensive laboratory testing program.
This paper primarily focuses on the suitability of recycled glass-crushed rock blend as a pavement sub-base or light-duty base material. The engineering properties of recycled glass blended with crushed rock were investigated and compared with the local specifications for pavement sub-base applications.
The recycled glass (RG) and crushed rock (CR) mixtures were prepared in a range of proportions. Particle size distribution, specific gravity, water absorption, organic content, pH value, modified proctor compaction tests, particle size distribution after compaction tests, Los Angeles (LA) abrasion loss and California bearing ratio (CBR) tests were undertaken on all mixtures.
RECYCLED MATERIAL SOURCES
Samples of Class 3 crushed rock (20mm nominal size) produced from recycled excavation stone and recycled glass (4.75mm minus) were collected from the recycling site of Alex Fraser Group at Laverton, Victoria, located about 20km west of Melbourne. Sampling procedures in accordance with the relevant Australian Standards were adopted for collecting samples from the stockpiles.
All necessary precautions were taken to capture a sample containing representative particle sizes and representative amounts of all contaminants. Samples were placed in plastic bags and tightly sealed to retain the natural moisture content and then transported to a geomechanics laboratory at Swinburne University of Technology.
{{image5-a:c-w:600}}The crushed rock used in this study originated from newer basalt floaters or surface excavation rock, which commonly occurs near the surface to the west and north of Melbourne. The rock is often encountered in sub-divisional excavation for residential properties and in excavation works for drainage lines as well as other sub-surface infrastructure.
Traditionally this material would have been discarded as waste and often end up in landfill sites. However, because this rock is generally hard and durable, with the LA value typically less than 35, VicRoads has permitted its use for pavement sub-base (Class 3) and other uses under controlled conditions.
Recycled glass used in this research was derived from household waste collection as well as C&D activities. Recycled glass in storage may become segregated by size. Additionally, segregation of material from contaminants (such as closures, labels, or miscellaneous debris) may occur (Clean Washington Centre, 1996).
EXISTING SPECIFICATIONS
VicRoads is the statutory authority responsible for managing the road network, including roads and bridges in Victoria. VicRoads classifies crushed rock into four classes for use in pavements, namely Class 1, Class 2, Class 3 and Class 4. This classification is based on the physical and mechanical properties of crushed rock.
Class 1 and Class 2 crushed rock are usually specified for base-course applications, while Class 3 and Class 4 are restricted primarily to upper sub-base and lower sub-base applications respectively. Table 1 presents the physical properties of crushed rock normally specified by VicRoads.
Table 2 presents the particle size distribution requirements for the ?20mm Class 3 crushed rock before and after compaction.
GEOTECHNICAL CHARACTERISTICS OF PURE MATERIALS
Sieve analysis was conducted on representative specimens of as-received recycled glass and crushed rock samples to determine their gradation curve and soil classification according to the Australian Soil Classification System (ASCS). The essential difference between the Unified Soil Classification System (USCS) and the ASCS is the boundary between sand and gravel size particles.
While USCS assumes that 4.75mm (sieve No.4) is the border between sand and gravel, for ASCS 2.36mm has been selected as the border between sand and gravel particles. This means that particles larger than 2.36mm are considered gravel, according to ASCS. Table 3 presents the physical properties of as-received recycled glass and crushed rock samples obtained through a gradation curve analysis.
Recycled glass is classified as well-graded sand, according to the ASCS (Standards Australia, 1993). Crushed rock is classified as poorly graded gravel mixed with silt. This is because gravel particles are dominant in the coarse fraction of crushed rock and silt-size particles are also building up to 12 per cent of this recycled material.
After a single cycle of modified compaction, the compacted samples were extruded from the compaction mould and sieve analysis was performed on these post-compaction samples. Using gradation curve of post-compaction samples the relevant values are reported in Table 3.
The post-compaction sieve analysis results on recycled glass source showed negligible change in the gradation curve of the recycled glass before and after modified compaction, while for the crushed rock source, some changes in gradation curve before and after modified compaction was noticed.
The sand content of the as-received crushed rock sample increased from 28 per cent to 30.2 per cent after modified compaction effort. The gravel-size particles? percentage in the mixture decreased from 60 per cent for the as-received sample to 56.6 per cent after modified compaction test.
The change in the particle size distribution of crushed rock after the modified compaction test is due to the crushing of gravel-size particles under compaction energy and the subsequent decrease in their size to mainly sand-size and finer particles. The small changes in the particle size distribution of both the recycled glass and crushed rock sources after modified compaction indicate that recycled glass and crushed rock sources are relatively stable mixtures during the engineering operations, including handling, spreading and especially compaction.
LABORATORY TEST METHODS
The results of geotechnical engineering laboratory tests performed on the as-received recycled glass and crushed rock samples are presented in Table 4.
{{image6-a:r-w:450}}The results of particle density tests that correspond to the specific gravity value suggest that recycled glass possesses a specific gravity value about 15 per cent lower than the natural aggregate, while the specific gravity value of crushed rock is similar to those of natural aggregates.
The organic content test results proved that the organic content of both materials is below one per cent and is considered negligible. Both recycled glass and crushed rock sources contain negligible debris levels obtained in visual and weight methods.
It is believed that the primary reason recycled glass has a higher value of debris level obtained in the visual method, compared with crushed rock, is the presence of low density debris such as paper and plastic in the material.
{{image7-a:r-w:450}}A test method recommended for assessment of the durability and resistance of aggregate materials is the Los Angeles abrasion test (Wartman et al, 2004). LA abrasion test results suggest that recycled glass and crushed rock sources have similar LA abrasion values, although the LA abrasion value of recycled glass is slightly higher than crushed rock. This is believed to be due to the higher debris level of the recycled glass material and also the brittle nature of glass particles.
The brittle nature of the glass cullet particles will contribute to additional breakdown under attrition by the steel balls. Furthermore, the significant percentages of particles with a flat nature in the processed glass cullet will be a contributing factor.
Table 4 presents the results of modified proctor compaction tests and it is apparent that the maximum dry density value obtained for recycled glass source is about 10 per cent to 15 per cent lower than the values belonging to excavation stone crushed rock categorised in the same soil classification (Craig, 1992).
This is likely due to the lower specific gravity value of recycled glass source compared with natural aggregates within the crushed rock (Disfani et al, 2009b).
CBR tests were conducted on recycled glass and crushed rock test specimens compacted with the modified compaction effort inside the CBR moulds. The test specimens were soaked for four days inside the water under a surcharge of 4.5kg. Table 4 indicates that the CBR value of recycled glass source is lower than that of crushed rock source.
The lower CBR values for recycled glass are probably more related to grading, moisture sensitivity and the mobilisation of particles in a near-saturated condition. This trend seems to be related to higher values of maximum dry density obtained for the crushed rock source in the modified compaction test.
The higher maximum dry density for crushed rock is an indication of better compaction, which results in better particle contact and eventually better shear performance of the crushed rock source. The higher CBR value of the crushed rock source can be also attributed to its lower LA abrasion value.
Direct shear tests (DST) were conducted following the procedure recommended by British Standard (BS1377-7, 1990) on specimens compacted inside the shear box with optimum water content to achieve the maximum dry density obtained in modified compaction tests.
It is recommended that the size of the largest particle of the test specimen should not exceed one-tenth of the specimen height and, consequently, with the 100mm ? 100mm shear box having the effective depth of 40mm, only the recycled glass samples of particle size less than 4.75mm were tested. Five different normal stress levels were applied to the test samples.
The internal friction angle of recycled glass source declined from 55? to 46?, with normal stress increasing from 25 kilopascals (kPa) to 400 kPa. The internal friction angle of recycled glass is found to be similar to that of dense sand with angular grains (Das, 2008). This suggests that the recycled glass source exhibits the satisfactory friction characteristics for usage in some geotechnical engineering applications.
Laboratory test results shown in Table 4 suggest that recycled glass and crushed rock sources attain comparatively acceptable engineering characteristics to be used as alternatives for natural aggregate in appropriate geotechnical engineering applications. The DST results indicate that the recycled glass source possesses lack of cohesion resistance between particles. This is most probably the result of the smooth surface of the glass particles and the small number of fine particles in the mixture.
LABORATORY STUDY ON RECYCLED GLASS-CRUSHED ROCK BLENDS
The laboratory study on recycled glass-crushed rock blends were conducted on mixtures comprised of 10 per cent, 15 per cent, 20 per cent, 25 per cent, 30 per cent, 40 per cent and 50 per cent (by mass) RG mixed with Class 3 CR. The blends are respectively represented by RG10/CR90, RG15/CR85, RG20/CR80, RG30/CR70, RG40/CR60 and RG50/CR50. The gradation curves of recycled glass-crushed rock blends before compaction as compared with VicRoads requirements are shown in Figure 1. The gradation curves of blends after modified compaction in comparison with VicRoads requirements are shown in Figure 2.
A set of geotechnical engineering tests including Atterberg limit, particle density, water absorption, organic content, pH value, LA abrasion loss, modified compaction and CBR tests were conducted on recycled glass-crushed rock blends and the results are presented in Table 5.
{{image8-a:c-w:600}}The specific gravity of the blends is increasing with amplification in the crushed rock percentage of the mixtures, as presented in Table 5, except the RG15/CR85 blend, which might be due to the lower accuracy of the testing equipment, as the difference between values is small and possibly lower than the accuracy of the equipment.
This is the result of the higher specific gravity of the crushed rock as compared with the recycled glass. The change in water absorption values indicates that recycled glass has a lower water absorption value as compared with crushed rock, as expected. This is believed to be the result of the nature of the material and the impermeable smooth surfaces of recycled glass particles.
All the blends have a low percentage of fine particles, varying from eight per cent to 11 per cent. The majority of this fine content is made out of silt-size particles, as indicated by the very low PI values achieved in each blend. Atterberg limit tests conducted on all blends showed that the plastic limit and liquid limit cannot be obtained. The primary reason for this is that the Atterberg limit is directly related to clay mineralogy and, as such, very low fines content with silty materials results in immeasurable plasticity.
This aspect suggests that some difficulties may be experienced with the workability of the recycled glass blends, as cohesion of particles and a ?tight? prepared surface are usually sought after characteristics. A field trial of the recycled glass?crushed rock blends would best determine the level of difficulty that may ?be experienced. An addition of small quantities of clayey sand or plastic crushed fines may also be a good solution to this potential problem.
The organic content in the blends varies from 0.6 per cent to 0.8 per cent, indicating stable behaviour of the material. The pH values of all blends are much greater than seven, which shows the slightly alkaline nature of the blends.
Figure 3 shows the compaction curves for the recycled glass-crushed rock blends obtained from modified proctor compaction test results. From the compaction curves, it is evident that by increasing the recycled glass content in the mixtures, the maximum dry density (MDD) decreases. The compaction curves presented in Figure 3 are generally flatter than the compaction curve of excavation stone crushed rock.
{{image9-a:r-w:450}}This suggests that recycled glass-crushed rock blends can tolerate a greater number of variations in the moisture content without compromising much of the achieved density from compaction. Conversely, materials with sharp curves are very sensitive to the change in the moisture content and there is a need to ensure that the moisture content is close to the optimum value during compaction (Poon and Chan, 2006). The optimum moisture contents and their corresponding maximum dry densities are shown in Table 5.
Figure 4 presents the LA values and CBR values of the recycled glass-crushed rock blends. LA abrasion value is a useful indicator of the durability and hardness of aggregates during crushing and compaction. The LA abrasion values were found to be within the maximum value of 35 normally allowed by VicRoads for Class 3 sub-base materials as shown in Figure 4. As seen in Figure 4, all samples have CBR values higher than 80 per cent, which is the typical specification applied by VicRoads for Class 3 sub-base materials produced from recycled materials.
COMPARISONS WITH LOCAL SPECIFICATIONS
The grading limits of the recycled glass-crushed rock blends fall within the upper and lower boundaries of the grading envelopes required by VicRoads.
{{image10-a:r-w:450}}Some blends were noted to be just on the limits and do traverse from the fine grading limit towards the target mid-range; as the sieve sizes become smaller, this may translate to more difficulties being experienced in compaction of these materials.
The after-compaction gradation curves presented in Figure 2 show that some breakdown has occurred during compaction; however, compliance with normal after-compaction requirements is still achieved. In Figure 1 and Figure 2, gradation curves belonging to RG50/CR50 and RG40/CR60 blends are just at the boundary or outside the boundary normally specified by VicRoads.
This means some difficulties might be experienced with the workability of the recycled glass content greater than 30 per cent. A field trial of the recycled glass-crushed rock blends would best determine the degree of difficulties that may be experienced.
Figure 2 would suggest that a maximum of 30 per cent recycled glass should be added to a Class 3 sub-base to maintain an acceptable after-compaction gradation curve. The breakdown in grading is most probably happening in both the recycled glass and crushed rock components.
The materials appear to be remaining reasonably well-graded through the compaction process and this will generally aid compaction.
{{image11-a:r-w:450}}The results of geotechnical engineering tests were compared with existing Victorian specifications for pavement sub-base applications. The maximum LA abrasion value of 35 allowed by VicRoads for Class 3 sub-base pavement materials and the test results for all blends were found to satisfy the VicRoads requirements as shown in Figure 4.
All samples were also found to have CBR values higher than 80 per cent, which is the VicRoads requirement for Class 3 sub-base materials produced from recycled materials.
RECOMMENDATIONS
The laboratory studies carried out in this research have shown overall that the incorporation of up to 50 per cent recycled glass into crushed rock has low to minimal effect on the physical and mechanical properties of the original material but the particle size distribution curves of RG50/CR50 and RG40/CR60 blends are at the boundary or outside the boundary lines specified by VicRoads. As such, the recycled glass-crushed rock blends with the maximum percentage of 30 per cent of recycled glass were found to satisfactorily meet the current VicRoads requirements.
Findings of this research suggest that up to 15 per cent ?recycled glass with the maximum particle size of 4.75mm? could be safely added to Class 3 crushed rock. Test results show that the extent of breakdown occurring in the blends within excess of 15 per cent recycled glass is on the limit of what would be acceptable for this material. It might be possible to increase the percentages of recycled glass but this depends on the results of future field trials.
The grading limits of most of the recycled glass-crushed rock blends were well within the upper and lower bounds for crushed aggregates specified by VicRoads. The before and after compaction grading curves for RG50/CR50 was noted to be on the VicRoads upper limits.
CBR and LA abrasion values of all the mixtures were found to be satisfactory, according to VicRoads requirements for Class 3 sub-base material.
ACKNOWLEDGEMENTS
This article first appeared in Geomechanics 46 (1), March 2011, and is reprinted with kind permission. The authors thank Sustainability Victoria for funding this research project (Contract No. 5312) and the Alex Fraser Group for providing samples of recycled glass and crushed rock as well as its technical assistance on this project. The authors also thank VicRoads for permission to publish this paper and Alec Papanicolaou (geomechanics laboratory technician at Swinburne University of Technology) for his technical support during the experimental works.
CONTRIBUTORS
MM Younus Ali and MM Disfani, PhD candidates, Faculty of Engineering and Industrial Sciences, Swinburne University of Technology, Melbourne G Newman, VicRoads, Melbourne Associate Professor A Arulrajah, Faculty of Engineering and Industrial Sciences, Swinburne University of Technology, Melbourne
REFERENCES
British Standards. Methods of test for soils for civil engineering purposes – Part 7: Shear strength tests (total stress). (BS1377-1990 (7)). UK; 1990.
Clean Washington Center. Methods for sampling and testing recycled glass. Seattle: CWC; 1996.
Clean Washington Center. A toolkit for the use of post-consumer glass as a construction aggregate. Seattle: CWC; 1998.
Craig RF. Soil mechanics. London: Chapman and Hall; 1992.
Das BM. Advanced soil mechanics. 3rd ed. New York: Taylor and Francis; 2008.
Disfani MM, Arulrajah A, Suthagaran V, Bo MW. Geotechnical characteristics of recycled glass – Biosolids mixtures. Proceedings of the 17th International Conference on Soil Mechanics and Geotechnical Engineering. Alexandria, Egypt; 2009a: 201-204.
Disfani MM, Arulrajah A, Suthagaran V, Bo MW. Shear strength behaviour of recycled glass-biosolids mixtures. Proceedings of the 62nd Canadian Geotechnical Conference. Halifax, Canada; 2009b: ?812-819.
Federal Highway Administration of US Department of Transportation. User guidelines for waste and by-product materials in pavement construction. Washington DC; 1998.
Landris TL. Recycled glass and dredged materials. US Army Engineer Research and Development Center, Vicksburg, Mississippi; 2007.
Poon CS, Chan D. Feasible use of recycled concrete aggregates and crushed clay brick as unbound road sub-base. Construction and Building Materials 2006; Vol 20: 578-585.
Standards Australia. Geotechnical site investigation (AS1726-1993), Homebush, NSW; 1993.
Standards Australia. Method for sampling and testing aggregates (AS1141-1996). Homebush, NSW; 1996.
Standards Australia. Method of testing soils for engineering purposes (AS1289-1998). Standards Australia, Homebush, NSW; 1998.
Sivakumar V, McKinley JD, Ferguson D. Reuse of construction waste: Performance under repeated loading. ICE Journal of Geotechnical Engineering 2004; 157 (GE2): 91-96.
Sustainability Victoria. Annual Survey of Victorian Recycling Industries 2005-2006. Melbourne; 2007.
Tam VWY, Tam CM. A review on the viable technology for construction waste recycling. Journal of Resources, Conservation and Recycling 2006; 47 (3): 209-221.
VicRoads. Standard specifications for roadworks and bridge works (Section 304: Flexible pavement construction). Melbourne: February 2008.
VicRoads. Standard specifications for roadworks and bridge works (Section 820: Recycled crushed concrete for pavement sub-base and light duty base). Melbourne: January 2009.
VicRoads. Standard specifications for roadworks and bridge works (Section 812: Crushed rock for base and sub-base pavement). Melbourne: ?May 2010.
Wartman J, Grubb DG, Nasim ASM. Select engineering characteristics of crushed glass. Journal of Materials in Civil Engineering 2004; 16(6): 526-539.