Resources, Conservation and Recycling

35 (2002) 259–274 www.elsevier.com/locate/resconrec

Engineering properties of asphalt concrete

made with recycled glass
Nan Su *, J.S. Chen
Department of Construction Engineering, National Yunlin Uni6ersity of Science and Technology, 123,
Section 3, Uni6ersity Road, Touliu, Yunlin 640, Taiwan, ROC

Received 30 May 2001; accepted 23 January 2002

Abstract

Taiwan is an island nation with high population density: 23 million people living in a total
land area of 30 000 sq. km. The island has transformed from an agricultural society to an
industrial one in the past three decades. Coming with the transformation is a cumulated
waste problem. Taiwan produces nearly 5 million tons of waste each year, of which 10% is
of glass materials. Bureau of Highway Department funded a research program to investigate
ways of recycling the glass waste. This report presents information on the program and
laboratory/field test data. Materials used in the test program included 85/100 asphalt, Type
II modified asphalt, and treated glass waste. Four glass contents: 0, 5, 10, and 15%, in terms
of the total aggregate weight, were used in the mixture designs for casting series of 10 cm
diameter by 6.35 cm disk specimens. Tests including Marshall stability value, dry/wet
moisture damage, skid resistance, light reflection, water permeability, and compaction were
carried out in accordance with the ASTM and AASHTO procedures. The test results reveal
that glass waste is a viable material for asphalt concrete that has been widely used in
pavement that offers profound engineering and economic advantages. © 2002 Elsevier
Science B.V. All rights reserved.

Keywords: Asphalt concrete; Waste glass; Light reflection; Marshall test

* Corresponding author. Tel.: +886-5-534-4140; fax: +866-5-537-1703.

E-mail address: sun@pine.yuntech.edu.tw (N. Su).

0921-3449/02/$ - see front matter © 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 1 - 3 4 4 9 ( 0 2 ) 0 0 0 0 7 - 1
260 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274

1. Background

The remarkable technology advancement and economic growth in the last 30

years have ushered Taiwan to become a high consumption nation. An official
statistics released by the Environment Protection Agency (EPA) reveals that an
average daily waste per person has reached 1.14 kg close to that of the United
States if not surpassing. Based on the current population of 23 millions, the
national annual waste in Taiwan is 4 785 000 tons, of which approximately,
600 000 tons are of glass materials (Su and Chen, 2000). Being non-metallic and
inorganic, glass can be neither cremated nor decomposed by microbial. For this
reason, underground burial has been the only possible solution.
In 1989, the government assigned certain areas within major municipal and
county districts for burying. Today, more than two-third of these designated
areas have been used up and new burial grounds are hard to obtain largely
attributed to public oppositions and expensive land prices (Chang et al., 2000).
To help solve the waste problem, the related governmental agency has imple-
mented a recycling program similar to those carried out in Japan for years.
Waste recycling is a worldwide problem. The Luzon Institute of Waste Manage-
ment of Switzerland recommended the waste-recycling rate for measuring a na-
tion’s gross productivity.
In general, waste glass produced in Taiwan can be sorted out as:
“ building/automobile windows and doors;
“ glassware and bottles;
“ television tubes and light bulbs;
“ others such as mirror and clock covers.
Among them, the first two are the major sources. Several domestic glass
manufacturers have attempted to recycle the glass waste. EPA statistic data show
approximately 110 000 tons; about 20% of the annual waste-glass volume has
been recycled (Japan Asphalt Pavement Association, 1996). It is a poor standing
by the Luzon Institute recycling measurement.
In order to increase the recycling rate EPA established a task force in 1998 to
implement and enforce a cash-for-waste glass program; one NT dollar (equiva-
lent to 3 US cents) for 1 kg of waste glass. Unfortunately, the program did not
generate interest from glass manufacturers who experienced difficulties associated
with quality control (Su and Chen, 2000; Chang et al., 2000).
Japan, United States, and several European nations have utilized waste glass
to partially substitute fine aggregate of asphalt concrete (Japan Asphalt Pave-
ment Association, 1996; Su and Chen, 2000). The authors of this paper were
attracted by the concept and initiated a research program to investigate the
possibility of recycling the domestic waste glass in Taiwan. The fifth District of
the Engineering Administration of the Taiwan Highway Bureau granted a one-
year program in 2000 for the study. The obtained experimental data obtained
are presented in this paper.
 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274 261

2. Objectives

The primary objective of the test program is to establish a practical procedure of

recycling waste glass.

3. Scope of work

The major work of this program involved with laboratory and field tests.

3.1. Laboratory tests (Hseng and Su, 1997; Chen, 1998; Lin, 2000)

“ optimizing particle sizes of recycled glass, percentages of sand replacement, and

mix proportions through Marshall mix design and related materials property
tests;
“ determining the optimum asphalt content for 0, 5, 10 and 15% of glass
substitution cases; and
“ analyzing the measured test data.

3.2. Field tests (Faun, 1996; Ministry of Transportation, 1996; Taiwan Highway
Bureau, 1998; Chen and Chaou, 1999; Pang, 2000)

“ stability value;
“ roadway light reflection;
“ permeability;
“ skid resistance;
“ compaction;
“ performance evaluations.
All field tests were conducted on three test sites, one 140 m2 and two 510 m2, that
were constructed with glass asphalt concrete made in accordance with the mix
designs developed in the laboratory.

4. Basic information on asphalt, glass, and aggregate

Since asphalt concrete made with recycled glass has been relatively new to most
engineers, it is desirable to familiarize readers with some basic information on
bituminous materials, waste glass and aggregate at this point.

4.1. Bituminous materials

Bituminous material, the basic substance for asphalt, is a complex mixture of

hydrocarbons, residues of a petroleum production. It consists of asphaltenes,
262 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274

asphaltic resin, and oily constituents, solvable in a carbon disulfide (CS2) or

carbon tetrachloride (CCl4) solution. The colors of natural bituminous can be
brown, black, white, red, blue, or green depending on the addition of chromium
oxide (Cr2O3), ferric oxides (Fe2O3, FeO), or titanium oxide (TiO2) (Tsai, 1995).
Heated asphalt has a strong plasticity property while the cooled asphalt has
an oxidation property under a long weather exposure and load services. Asphalt
of low viscosity has a better workability than that of the high viscosity in terms
of wetting and coating the aggregate surfaces during the mixing process.
Nonetheless, asphalt with high viscosity provides strong binding forces among
aggregates and excellent weather resistance (Bureau of National Highway Engi-
neering, 1997).
Other properties such as specific gravity, water absorption, penetration ratio,
softening point, ductility, viscosity, temperature sensibility, and stiffness are also
important to asphalt applications and are discussed in the subsequent sections.

4.2. Glass

Glass is a non-metallic inorganic material made by sintering selected raw

materials comprising silicate and other minor oxides as shown in Table 1. By
cooling the molten glass on mold beds sheet glass of desirable sizes is formed.
A conventional glass is rather brittle, easily broken by a small impact. This
physical property has been used to crush the waste glass to form desirable
particles for mixing purpose. In general, broken glass has a specific gravity of
2.5, a bulk unit weight of 1.3– 1.4 tons per cubic meter, and a water absorption
rate of 0.3– 0.4 wt%. It has high volumetric stability under a temperature up to
700 °C. Its thermal expansion coefficient and softening point are 8.8– 9.2× 10 − 6
cm/cm/°C and 718 – 738 °C, respectively. Suitable particle sizes for asphalt con-
crete application as aggregate are 4.75 mm or smaller.

Table 1
Chemical composition of glass

Oxide Weight (%)

SiO2 70.87–72.83
Na2O 12.40–13.67
CaO 8.84–10.47
Al2O3 1.47–2.43
K2O 0.79–1.17
SO2 0.20–0.26
MgO 0.11–3.91
Fe2O3 0.03–0.37
TiO2 0.01–0.04
 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274 263

4.3. Aggregate

Both physical and chemical properties of coarse aggregates contribute significant

effects on the asphalt concrete. The surface characteristics of aggregate provides the
binding mechanism; greater the roughness, stronger the resistance to separation.
The chemical property has a profound effect on the endurance of asphalt
concrete. In general, aggregates can be classified as hydrophobic or hydrophilic of
which the former is acid attributed to the silicate (SiO2) constituent, while the later
is alkali due to the dominant calcium carbonate (CaCO3).

5. Test programs

Major tests carried out in this program followed the ASTM or AASHTO
standard procedures. Fig. 1 gives the general flow chart of the test programs; three
specimens were tested in each test. Following sections describe tests conducted in
the laboratory and fields.

5.1. Laboratory tests

The basic material properties of materials are prerequisite to the Marshall mix
and stability value tests.

5.1.1. Specific gra6ity and moisture

Materials used in the test program include 85/100-grade asphalt, modified
asphalt, and graded crushed glass, sand, and coarse aggregate. The specific gravity
of asphalt (85/100) and modified asphalt were determined following the ASTM
C-127 and 128 standard procedure. The ASTM C-566 was applied to measure the
moisture contents of aggregate. Table 2 lists the measured data.

5.1.2. Gradations
Particle-size distributions of combined coarse and fine aggregate were determined
following the ASTM C-136, standard sieve analysis method. Fig. 2 shows the
measured data of which the maximum coarse aggregate was 20 mm (3/4 in.).
The recycled glass used in the test program was first crushed and then pulverized
to the maximum size of 4.75 mm (c4 sieve size). Again ASTM C-136 was applied
to determine the particle size distribution and the results are shown in Fig. 3.

5.1.3. Combination of sie6ed aggregate and glass

In each test series a group of specimens were made without recycled glass (0%)
for the purpose of reference to specimens made with 5, 10, and 15% glass
substitutions. The substitution of glass was made on each sieve size from c 4 down
to c 200 in the designated substitution percentage. A 5% glass substitution, for
example, a 5% of c4 aggregate was removed and replaced by an equal amount of
c 4 glass particles and likewise for other sized aggregates.
264 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274

Fig. 1. Test flow chart.

5.1.4. Optimizing glass contents

Prior to the Marshall mixture tests the aggregate was dried in oven regulated at
a temperature of 170 °C for a minimum of 6 h until they reached an absolute dry
condition. The ASTM C-29 and AASHTO T19-80 were used to measure the unit
weights, void ratios, and void of mineral aggregate of fine and coarse aggregates,
while the ASTM D-1559 procedure using Marshall apparatus was applied to
measure the flow resistance of bituminous mixtures. The mixing temperature for
85/100 asphalt was controlled at 150 °C at which an asphalt viscosity is 0.17 Pa s
 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274 265

Table 2
Measured specific gravity and moisture contents

Material Asphalt (85/100) Modified asphalt Sand Coarse aggregate 20, 10, Glass
properties 5 mm

Specific gravity 1.025 1.028 2.66 2.63, 2.62, 2.64 2.54

Moisture content – – 1.64 0.92, 1.04, 1.40 0.34
(%)

as required by Section 4.3.1 of ASTM D1559. Similarly, the mixing temperature for
the modified asphalt was controlled at 170 °C.
Through series of tests, the optimum asphalt contents for 0, 5, 10, and 15% of
glass substitutions cases were assessed and the results are presented in Figs. 4 and
5 for 85/100 asphalt and modified asphalt, respectively.

5.1.5. Lime admixture

Asphalt concrete made with partial glass substitution tends to develop durability
problems in connection with permeability. To minimize the problem, proper
amounts of lime were added in mix designs to enhance the durability (Lin and Yu,
1997).

5.1.6. Stability tests

5.1.6.1. Wetted and dried specimens. A series of specimens, made in accordance with
the AASHTO T245 and AIMS-2 standard methods, were subjected to water
submerge for 4 days and subsequently removed from water to dry air for 24 h. The
dried specimens were then subjected to the stability value tests to determine the
maximum strength of each test specimen (10.4 cm diameter by 6.35 cm high). Fig.

Fig. 2. Aggregate sieve analysis data.

266 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274

Fig. 3. Sieve analysis of recycled pulverized glass.

6 gives the measured stability values of glass asphalt concrete made with added
limes varying from 0 to 3%.

5.1.6.2. O6en dried specimens. Two groups of specimens made in accordance with
the AASHTO T245 standard method were placed in an oven regulated at tempera-
ture of 609 1 °C for 10 and 20 consecutive days for the first and second group,
respectively. The 60 °C was adapted to simulate the roadway surface temperature
in hot summer days in Taiwan. The measured stability values of these two groups
together with the reference specimens without heating are shown in Fig. 7.

5.1.6.3. Specimens subjected to dry– wet cycling. Again, specimens made in accor-
dance with the AASHTO T245 standard method were placed in oven controlled at
60 °C for 12 h and then removed for water submerge for 12 h. The dry–wet cycle
was designed to simulate the typical tropical weather (Summer shower) in Taiwan.
Likewise, two groups of specimens were subjected to 24-h dry– wet cycles; for 10

Fig. 4. Optimum asphalt content of glass asphalt concrete – 85/100 asphalt.

 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274 267

Fig. 5. Optimum asphalt content of glass asphalt concrete.

and 20 days, respectively. The data measured including the reference specimens are
shown in Fig. 8.

5.1.7. Permeability tests

Permeability tests were conducted on cores drilled from pavement beds at two
different field test sites; Highways 156 and 158. Each test site consisted of various
asphalt concrete made with 0, 5, 10, and 15% glass contents. These tests were
carried out following the Japanese standard JIS A1218 for testing asphalt concrete.
The test was designed to measure the amount of water passing through the
cross-section of core specimen in a certain time period. Table 3 lists the calculated
permeability coefficients based on the measured data.
The data obtained indicate that the permeability increases with respect to the
amount of recycled glass used and are less than 1.0× 10 − 2 cm/s recommended by
the JIS A1218 standard. Because of the hydrophobic property of recycled glass, the
asphalt concrete made with glass appears to have a low permeability.
Likewise, the stability value and compatibility tests (AASHTO T230-68) were
conducted in laboratory on 10-cm diameter by 6.35-cm thick cores drilled from field

Fig. 6. Stability values of glass asphalt concrete with lime after 4 days of submerge.
268 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274

Fig. 7. Stability value of glass asphalt concrete with lime subjected to 60 °C heating.

test sections. Again, the standard Marshall procedure was used and the measured
stability values for specimens made with 85/100 and modified II asphalt and
variable glass contents are shown in Fig. 9. The compatibility was determined by
dividing the specific weight of tested specimen by that of a standard specimen are
also listed in Table 4.

5.2. Field tests

Field tests are an important part of this study program and were carried out
following the laboratory tests. Three test sites; one with an area of 140 m2 and two
with 510 m2 provided by the Highway Engineering Department, were used for the
tests.
Asphalt concrete used in the first test site consisted of 10% glass content while the
two others were constructed with asphalt concrete of glass contents varying from 0,
5, 10, and 15% and asphalt contents of 5.32, 5.20, 5.05, and 5.00%, respectively.
The two large test sites were designated for the stability value, light reflection, skid
resistance, and compact tests performed in accordance with the standard Marshall
test (ASTM D1559, AASHTO T245-97), Chinese National Standard (CNS 4346),

Fig. 8. Stability values of glass asphalt concrete made with lime subjected to dry – wet cycles.
 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274 269

Table 3
Permeability values of asphalt concrete at field test sections

Recycled glass content (%) Measured average permeability (cm/s)

Highway 156 Highway 158

0 1.66×10−3 1.06×10−3
5 3.34×10−3 2.24×10−3
10 4.77×10−3 3.91×10−3
15 6.27×10−3 4.84×10−3

and the ASTM E303 test procedures, respectively. The test sections made of glass
asphalt concrete reflect light in the night (Fig. 10). Table 5 summarizes the
measured light reflection values, while Table 6 lists measured skid data for cores
taken from Highways 156 (85/100 asphalt) and 158 (Modified asphalt).

5.3. Ser6iceability e6aluations

In order to understand the information of asphalt concrete pavement with

recycled glass, this study tested the light-reflection, permeability (JIS A1218), skid
resistance (ASTM E303), degree of compaction (AASHTO T230-68), and pavement
deflection (Benkelman beam, AASHTO T256-77) on test sections with 10% recycled
glass after one-year service-life.

6. Discussion

The results of the present study can be categorized discussed in accord with the
following aspects.

Fig. 9. Stability values of 85/100 and modified asphalt concrete specimen drilled from field test sites.
270 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274

Table 4
Measured average degree of compaction for glass asphalt concrete

Recycled glass content (%) Average compaction degree (%)

85/100 Asphalt highway 156 Modified asphalt highway 158

0 97.68 98.36
5 98.48 97.92
10 98.17 97.86
15 97.95 98.48

6.1. Laboratory tests

1. Test data reveal that the optimized asphalt content for a Marshall mixture
design varies with respect to the amount of recycled glass used. Mixtures that
composed of more glass require less asphalt binder. Figs. 4 and 5 confirm that
observation can be explained by the low water/asphalt absorption rate of glass
attributed to its surface characteristics.
2. Because of the low asphalt absorption rate the glass granule conceivably retain
thin asphalt coating on the surfaces and result to weak binding strength along
the interfaces between glass particles and asphalt. As the rule of thumb, a glass

Fig. 10. The glass asphalt concrete shows light reflection from test section in the night.
 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274 271

Table 5
Measured average light-reflection data

Recycled glass content (%) Light-reflection (mcd/lx/m2)

Highway 156 Highway 158

0 0.00 0.00
5 14.63 14.09
10 16.23 14.90
15 17.84 16.67

asphalt concrete has a weaker debonding resistance than asphalt concrete made
without glass substitution.
3. The shortcoming of glass asphalt concrete can be corrected and improved by
adding lime to the mixture. A series of tests on specimens made with 1– 2% by
weight of lime were performed in this program. The test results show that the
greater amount of lime used produces a stronger deformation resistance. In view
of recycling waste glass, we recommend 2% lime addition that offers both
engineering and economic advantages.
4. Data obtained from heating and dry/wet cycling tests are shown in Figs. 6 and
7. The degree of deformation or aging process increases with respect to the
increase of glass content. In the asphalt concrete practice, a term, ‘stability
value’ has been widely used to describe the concrete’s ability to withstand the
deformation. It is interesting to note that the stability value of glass asphalt
concrete gradually increases up to 10 days of heating and decrease slowly
thereafter as shown in Fig. 7.

Table 6
Calculated skid resistance coefficients based on measured data, R20

Highway 158

Recycled glass content (%) 85/100 asphalt Modified asphalt

Dry Wet Dry Wet

L T L T L T L T

0 84.4 87.4 42.6 44.6 85.4 88.8 45.4 47.2

5 85.6 88.2 45.2 47.0 88.2 93.8 43.6 48.2
10 896 92.0 48.0 51.4 98.8 101.2 50.0 53.2
15 95.2 98.0 53.8 57.2 99.6 103.0 53.8 57.2

Note: L and T stand for longitudinal and transverse directions.

272 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274

Table 7
Information on asphalt concrete test sections with 10% recycled glass after one-year use

Test Item Highway 156 Highway 158

New After 1 year New After 1 year

Light-reflection (mcd/lx/m2) 16.23 18.77 14.90 18.74

Skid resistance, R20 48.0 51.6 50.0 51.2
Degree of compaction (%) 98.4 98.8 97.9 98.4
Permeability (×10−3 cm/s) 4.77 2.08 3.91 1.02
Deflection (1/1000 cm) 41 45 41 68

6.2. Field tests

1. The stability values of tested glass asphalt concrete decrease with respect to
increasing volume of glass content. By comparing the 85/100 and modified
asphalt, it was found that the stability value for modified one is approximately
33% stronger than the regular asphalt. Nonetheless, the stability values for
either case are profoundly greater than the minimum value of 681 kg as required
by the standard Marshall test procedure.
2. Light reflection intensities as obtained from testing roadway in the night,
increase with respect to the amount of glass used. For the purpose of reference,
the reflection intensity of white roadway divide lines is of 200 mcd/lx/m2, which
is obviously much stronger than the measured intensities of 14.09–17.84 mcd/lx/
m2 of the tested glass asphalt concrete. It is a well-known fact that the reflection
intensity increases as the roadway age prolongs attributed to frictional wearing
of the roadway surface. Good roadway reflection can enhance drivers’ visibility
and make driving condition safer.
3. The skid resistance characteristic of asphalt concrete depends greatly upon the
materials used, construction quality, roadway age, dry/wet condition, surface
roughness, tire traction, and others. The tester used in this project was the
British portable skid-measuring unit. The test method follows the ASTM E303.
The investigation concludes that the skid resistance in both longitudinal and
transverse directions increases as the glass content in the asphalt concrete
increases.
4. The advantages of using glass in asphalt concrete can physically increase the
skid resistance and make driving safer in rainy days.

6.3. Ser6iceability e6aluations

1. Table 7 shows that the light-reflection and skid resistance of asphalt concrete
pavements on highways with recycled glass after one-year use are higher than
those of new pavements. This is because the glass granule has been progressively
 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274 273

exposed to the pavement surface due to the surface abrasion caused by the
vehicle traffic. As the recycled glass pavement shows enhanced light reflection
at night and skid resistance in rainy days, it improves the safety of vehicle
operations.
2. After one-year use, the degree of compaction of recycled asphalt concrete
pavements is higher and the permeability is lower than those of new pave-
ments. The pavement deflection (Benkelman beam) resulting from the applica-
tion of a known transient loading of pavement after one-year use is larger
than that of the new pavement.
3. The test section with 10% recycled glass after one-year use show well perfor-
mance (no spalling of glass granule, no rutting except small cracks on pave-
ment surface) as that without recycled glass.

7. Conclusions

The investigation concludes that adequately pulverized glass can be used to

substitute find aggregate for asphalt concrete mixtures.
The stability values of asphalt concrete made with 5–15% are lower than that
of asphalt concrete made without glass addition, but are well within the limit
specified in the related code requirement. An addition of suitable amount of
lime, approximately 2 wt%, to the mixture conceivably increases the stability
values of the asphalt concrete thus made. Asphalt concrete made with recycled
glass produces adequate light reflection enhancing night visibility.
Data obtained from testing field cores show low permeability and higher fric-
tional coefficients that are beneficial to the roadway safety. All field test sites
perform exceedingly well since a year ago. We were convinced that asphalt
concrete made with recycled glass offer a very practical solution to the glass
waste management, which will be of interest to industrial nations around the
world.

References

Bureau of National Highway Engineering. Specification of highway construction. Taiwan: Ministry of

Transportation; 1997.
Chang TJ, Chen JS, Pan W. Technologies for waste glass-recycling programs. Taiwan: National Taipei
Technology University; 2000. p. 1 – 49.
Chen FS. Pavement designs –construction, and quality control for asphalt concrete. Chung-li, Taiwan:
National Central University; 1998.
Chen SI, Chaou TD. Advanced highway construction materials. Taiwan: Taiwan Construction Research
Institute; 1999. p. 163 –89.
Faun CS. Skid tests and design criteria. J Mod Construct 1996;202:11 – 20.
Hseng DS, Su N. Engineering and economic evaluation of asphalt concrete made with sea sand. Taiwan:
Bureau of National Highway Engineering, Ministry of Transportation; 1997.
Japan Asphalt Pavement Association. Waste Glass Bottles for Pavements. Tokyo: Clean Japan Center;
1996. p. D32 – 38.
274 N. Su, J.S. Chen / Resources, Conser6ation and Recycling 35 (2002) 259–274

Lin ZL. Asphalt concrete made with waste concrete. Thesis. Taiwan: National Taiwan University of
Science and Technology; 2002.
Lin JD, Yu CZ. Durability of asphalt concrete made with lime admixture. Taiwan: National Central
University; 1997. p. 4 –137.
Ministry of Transportation. Construction specifications for highways. Taiwan: Lion Publication Com-
pany; 1996.
Pang CL. Designs and construction of asphalt concrete. Taiwan: Taiwan Construction Research
Institute; 2000.
Su N, Chen JS. Physical properties of glass asphalt concrete and its applications. The First Conference
on Roadway Pavement, Taiwan; 2000. p. 12 – 26.
Taiwan Highway Bureau. Handbook of highway construction materials. Taiwan Highway Materials
Laboratory; 1998. p. 637 –97.
Tsai PA. Asphalt concrete. Taiwan: San Ming Publication Company; 1995. p. 1 – 110.