Professional Documents
Culture Documents
a r t i c l e i n f o a b s t r a c t
Article history: After several years in service, porous asphalt looses permeability primarily due to clogging. This paper
Received 10 November 2010 presents a phenomenon, postulated as binder creep, as another source of permeability loss in porous
Received in revised form 24 October 2011 asphalt. To ascertain the binder creep phenomenon, samples were separately conditioned at 15 °C,
Accepted 26 November 2011
20 °C, 30 °C and 35 °C. Permeability measurements were carried out at regular intervals up to 60 days
Available online 29 December 2011
using a falling head water permeameter. Permeability reduction was quantified in terms of increased
time of flow and reduced coefficient of permeability over time. The results showed that, permeability loss
Keywords:
was more significant on samples conditioned at high temperatures. Statistical analyses showed that tem-
Porous asphalt
Binder creep
perature and time have significant effect on permeability reduction of the porous mixes tested. Over an
Conditioning temperature extended period of time and due to gravitational forces, it was postulated that binder creep in the mix
Time of flow had taken place which steadily disrupted air voids continuity and subsequently caused the mix perme-
Permeability ability to reduce. Permeability reduction took place more rapidly during the first week but later tend to
asymptote. This finding implies the need to specify a time frame within which porous asphalt permeabil-
ity measurements must be made. A recommended practice would be to conduct the permeability test not
exceeding 1 day after the porous asphalt samples were compacted to eliminate inaccuracies in perme-
ability measurements due to interference by binder creep.
Ó 2011 Elsevier Ltd. All rights reserved.
0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2011.11.038
M.O. Hamzah et al. / Construction and Building Materials 30 (2012) 10–15 11
% Passing
60
2. Materials and methods
40
2.1. Materials and mix design
20
The aggregate material used was granite, obtained from a local asphalt mixing
plant. A conventional bitumen 60/70 penetration grade supplied by Shell Ltd. was 0
used to prepare the porous asphalt specimen. The basic properties of the 60/70 bitu- 0.01 0.1 1 10 100
men used are summarised in Table 1. The Malaysian porous asphalt gradation [9] was Sieve Size (mm)
used in specimen preparation as shown in Fig. 1. The cylindrical specimens, prepared
at 4.7% binder contents, were compacted via impact mode at 50 blows per face. All Fig. 1. Aggregate grading used in this study [9].
compacted specimens were of similar heights. The samples were then left to cool at
ambient temperature for 4 h before immediately tested for permeability. After the
initial permeability was measured, the specimens were conditioned at the designated
conditioning temperature.
28-Aug
12-Jun
40
17-Jul
3-Jun
4-Sep
18-Aug
2-Jun
3-Jul
10-Jun
24-Jul
25-Aug
25 tion course (OGFC) layers with air void contents approaching 20%
7-Aug
1-Jun
16-Jun
1-Jul
16-Jul
31-Jul
15-Aug
30-Aug
14-Sep
Table 4
Percentage changes in F and k.
explanation for the permeability loss is binder creep. Creep is a Fig. 6 presents the individual plot of the relationship between k
time-dependent phenomenon. In this regards, creep can also take and time on all specimens tested. A statistical analysis is made on
place in an asphalt binder. The temperature which hastens creep the k values and the result is shown in Table 6. Temperature, time
to occur differs according to binder grades. Binders can also creep and interaction between them have a significant effect on the
or flow when acted upon by long loading time, which in this case is change in k of the mixes tested. In addition, the model fixed the
the force of gravity (self-weight of the binder). According to Evans data very well with a high coefficient of determination (R2),
and Wilshire [16], any material will creep upon approaching its 94.45%. Over time and at elevated temperatures, k decreases due
melting point. to the disruption in air voids continuity.
The term permeability generally refers to the ease in which a
permeant, water in this case; can permeate through a porous med- 3.4. Flow time ratio
ia. In porous asphalt, permeability depends on the amount of inter-
connected voids through which water can flow. The time of flow time ratio (f) is another parameter that can be
Bitumen is a visco-elastic material and exhibits viscous behav- used to evaluate the extent of voids continuity disruption due to
iour (hence flow) at elevated temperatures and long time of loading. binder creep. The ratio is expressed by:
In this investigation, some specimens were subjected to elevated
temperatures causing the bitumen to become less viscous and flow. f ¼ F n =F 0 ð2Þ
Additionally, the permeability was monitored over an extended
time up to 60 days. Being a rheological material, gravitational forces where f is the time of flow ratio, Fn is time of flow at day n(s), F0 is
act on the binder coating the aggregate causing it to flow or creep the initial time of flow just after compaction (s).
over such long time of loading. Though in reality, it was the asphalt The time of flow ratio result is plotted in Fig. 7. An ideal mix
mastic as a whole that has creep, the terminology binder creep is would have a time of flow ratio equal to 1.0 and this implicates
used throughout this paper. When binder creep takes place, it occu- that the air voids remain continuous without any interference.
pies the air voids within the porous mixes and subsequently disrupts All mixes, except those conditioned at 15 °C, exhibit a time of flow
voids continuity, thus causing the permeability to gradually reduce. ratio significantly greater than 1.0. The higher flow time ratio of
Ferguson [17] pointed out at a phenomenon similar to binder creep specimens tested at an elevated temperature can be related to
that took place after the porous asphalt was laid down. Accordingly, the extent of binder creep.
the asphalt binder migrated downward through the pores under the
influence of heat and gravity. During the hot summer days; the sur- 3.5. Laboratory versus field permeability
face of an asphalt pavement is notably hot, causing the heat-soften
binder to flow down from the surface until it meets the cooler level A schematic diagram of the laboratory experimental setup and
where it ceased to flow and stop filling up the pores. The laboratory direction of flow in this study is shown in Fig. 8. In the laboratory,
investigation results reported in this paper confirms the field obser- the specimen was confined inside the cylindrical mould while
vation made by Ferguson [17]. Binder creep has the potential to dis- direction of flow through the specimen is unidirectional (vertical).
rupt the hydraulic conductivity of the porous mix. In the field, the porous asphalt layer overlies an impervious base.
In this investigation, the force of gravity (time of loading) is con- As schematically shown in Fig. 9, initially flow takes place verti-
stant on all specimens. Only the conditioning temperatures differ. cally and then laterally through the porous asphalt wearing course
The higher permeability loss registered by specimens subjected to into side drainage system. It is highly likely that the vertical and
higher conditioning temperature is consistent with the fundamen- horizontal coefficient of permeabilities for the same mix will be
tal principle that the bitumen becomes less viscous at elevated different. During specimen compaction in the laboratory and con-
temperature, causing higher flow, more air voids continuity dis- struction in the field, the binder migrates to the lower layers
ruption, hence more permeability loss. At low temperature, the resulting in a bottom layer richer in binder compared to the leaner
bitumen is stiffer, binder creep is literally prevented and the mix upper layer. The binder film coating the aggregate at the upper
permeability largely remains. layer will be deficient in binder, hence thinner bitumen film coat-
ing and this layer will be more prone to ageing. This layer will be
3.3. Statistical analysis most susceptible to ravelling. However, the bottom layer which
is richer in binder will be more susceptible to binder creep and this
The effects of time and conditioned temperature on binder
creep of porous asphalt are further analysed using statistical anal-
ysis. The Two-Way Analysis of Variance (ANOVA) was used with a Table 5
Two-way ANOVA on time of flow.
confidence interval of 95% (a = 0.05) on the samples evaluated
within 60 days at the four test temperatures. The result of Two- Source DF SS MS F p-Value
Way ANOVA is tabulated in Table 5, which indicates that time Conditioned temperature 3 4310.4 1436.81 209.81 <0.01
and temperatures have a significant effect on the binder creep Time 12 6408.1 534.01 77.98 <0.01
properties of porous asphalt when the p-value is less than 0.01. Interaction 36 2715.6 75.43 11.02 <0.01
Error 52 356.1 6.85
The interaction effect between conditioned temperature and time
Total 103 6.85
factors also has a significant effect on the increment of time of
flow. R2 = 97.42%, R2(adj) = 94.89%.
14 M.O. Hamzah et al. / Construction and Building Materials 30 (2012) 10–15
0.15
0.14
0.12
0.11
0.10
0.09
0.08
0.07
0.06
Day 0 3 10 30 60 0 3 10 30 60 0 3 10 30 60 0 3 10 30 60
Temperature 15 20 30 35
Fig. 6. Individual plot of coefficient of permeability at different conditioning temperatures over the test duration.
Table 6
Two-way ANOVA on coefficient of permeability.
Source DF SS MS F p-Value
Conditioned temperature 3 0.0122 0.00405 90.83 <0.01
Time 12 0.0216 0.00180 40.28 <0.01
Interaction 36 0.0058 0.00016 3.60 <0.01
Error 52 0.0023 0.00004
Impervious Layer
Total 103 0.0418
R2 = 94.45%, R2(adj) = 89.01%. Fig. 9. Actual flow in field porous asphalt pavement.
(i) The results shown in Fig. 5 shows that for specimens condi-
tioned at 30 °C and 35 °C, permeability loss is expected to
take place beyond 60 days. To ascertain this, permeability
loss on samples is monitored up to 120 days.
(ii) Binders become stiffer due to ageing. Modified binders are
also known to be stiffer and aged lesser compared to con-
ventional binders. Therefore, similar studies are being car-
ried out but on specimens that have been subjected to
short term ageing and specimens incorporating PG76 modi-
fied binders. The extent of permeability loss should be lesser
Fig. 8. Vertical flow in laboratory porous asphalt sample. in those aged and modified mixes.
M.O. Hamzah et al. / Construction and Building Materials 30 (2012) 10–15 15