Artigo Isap Final
Artigo Isap Final
ABSTRACT
This laboratory study investigates the effect of aggregate-binder adhesion on the moisture
damage and fatigue life of asphalt mixtures. Three mixtures with the same aggregate gradation
and approximate volumetrics were investigated: one with pure binder, another with the same
binder and a commercial antistripping agent, and a third one with the same binder and grounded
plastic bags as an additive to improve aggregate-binder interface properties. Adhesion was
analyzed following the Brazilian standard NBR 12583/17 and subsequent digital image
processing (DIP). Moisture damage was also investigated considering NBR 15617/15, and
fatigue life of the mixes was determined by the diametral compression test under controlled
stress. The aggregate area covered by the binder, as determined by DIP a correlation with
moisture damage results. The results showed that both the commercial antistripping agent and
the plastic bags produce an increase in the resistance of asphalt mixtures against moisture
damage and also an increase in fatigue life when compared to the mixture with pure binder.
Keywords: Adhesion, stripping, cracking fatigue, moisture damage.
1. INTRODUCTION
2
2. MATERIALS AND METHODS
The aggregate used is phonolitic and the asphalt binder is processed by Petrobras/Lubnor
in the state of Ceará. The characterization of these materials is described in Table 1.
TABLE 1 Characterization of Materials
Characterization of asphaltic binder Characterization of aggregates
Penetration, 25ºC [dmm] 50 Test Coarse Fine
Softening Point [ºC] 49 Los Angeles abrasion [%] 20.26 -
Temperature Susceptible -1.4 Absorption [%] 0.874 1.218
Rotational Viscosity, 135ºC [Cp] 395 Adhesion Unsatisfactory
Mixing Temperature [ºC] 160 – 166 Actual Density - 2.547
Compaction Temperature [ºC] 148 – 152 Apparent density 2.394 -
Three asphalt mixtures were investigated using the Superpave Methodology. Mixture
M1-P (P from Pure binder) was produced with pure binder, whereas mixture M1-M (M from
Modifier) with a 0.2% commercial adhesion antistripping agent, and mixture M1-R (R from
Recycled) with grounded plastic bags (PB) from waste. Fine PB material reached 1 cm² ± 0.5
cm², and it was placed upon the aggregates heated at 160ºC. These residues were added at a
content of 1.0% of the total mass of the mixture. This technique enabled PB to melt over the
aggregates prior the subsequent addition of the asphalt binder. The mixing temperature was
160°C and the compaction temperature was 150°C. Figure 1 presents the sole aggregate
gradation curve used and Table 2 presents the volumetric parameters of the investigated
mixtures.
100
Passing percentage [%]
Zone C-DNIT
TABLE 2 Volumetric Parameters of the Specimens
80 Project Curve Mixtures
Control Points M1-P M1-R M1-M
60 Optimum binder
5.0 5.0 4.5
40 content CA [%]
Air voids [%] 3.73 4.14 3.80
20 Bulk specific gravity
2.253 2.186 2.214
0 [g/cm3]
0.01 1 100
Particle diameter [mm]
FIGURE 1 Aggregate Gradation Curve
The adhesion of the mixtures was measured following NBR 12583/17, and analysis of the
bitumen-coated aggregate area was also determined by digital image processing (DIP). Moisture
damage was analyzed following NBR 15617/15, and laboratory fatigue life was measured by the
diametral compression test under controlled stress. Testing procedures are briefly described
ahead.
NBR 12583/17 has traditionally been used in Brazil access the aggregate-binder
adhesion. It uses a sample of heated aggregate (12.5/19.0 mm) weighing 500 ± 1g and 17.5 ±
0.5g of asphalt cement at a temperature of 100°C for aggregates and 120ºC for the AC. The
mixture is homogenized until the bituminous film completely coats the aggregates, and it is then
3
cooled maintaining the aggregates separated one from another. Upon cooling, the mixture is
placed in a beaker completely submerged in distilled water and then heated in the oven at 40ºC
for 72 hours. After this period, the mix is removed from the beaker and a visual inspection is
performed. If the film continues to completely coat the aggregates, the result is considered
satisfactory. If one notes displacement of the film, the result is considered unsatisfactory.
Therefore, results are reported on a qualitative and subjective basis.
In order to also have a quantitative and objective analysis, a pixel counting methodology
was developed in this research to determine the coated area of bituminous film after the adhesion
test. After the test, the loose mix is placed on a white paper sheet and photographed using a 12
megapixel camera (resolution of 1280 x 720). The aggregates of the mix are then rotated, and
three images are taken, since the test has a three-dimensional result and an image is two-
dimensional. Pictures are then taken to Matlab R2017 and histogram evaluation techniques are
used to find thresholds, image segmentation, recognition, classification, and image analysis.
DIP allows counting pixels within the aggregates and the asphalt binder. With Equation
(1), it is possible to find the bitumen film coating area (𝐴𝐷𝐼𝑃 ), where 𝑃𝐵 denotes the number of
pixels of the binder and 𝑃𝐴 the number of pixels of the aggregates.
𝑃𝐵
𝐴𝐷𝐼𝑃 = ∙ 100 (1)
𝑃𝐴
Mixture specimens are compacted to 7±1% air voids and then divided into 2 groups. In
the first group, three specimens are submitted to evaluate the indirect tensile strength (𝐼𝑇𝑆) by
diametral compression following NBR 15087/12. In the second group, three more specimens are
placed under high humidity severity conditions. Specimens are submerged in water and saturated
under vacuum until they reach a saturation level between 55-80%. They are then kept in the
freezer at -18°C for 16h, taken to a hot water bath at 60°C for 16h, and finally for temperature
stabilization at 25ºC for 2h. After this procedure, specimens are subjected to the 𝐼𝑇𝑆 test. The
indirect tensile strength ratio (𝑇𝑆𝑅), expressed in percentage, is determined by Equation (2),
where 𝐼𝑇𝑆𝐶 (MPa) is the indirect tensile strength of the conditioned specimens and 𝐼𝑇𝑆(MPa) is
the indirect tensile strength of the non-conditioned specimens.
𝐼𝑇𝑆𝐶
𝑇𝑆𝑅 = ∙ 100 (2)
𝐼𝑇𝑆
2.3 Fatigue
The laboratorial test to evaluate the fatigue life most commonly performed in Brazil is the
diametral compression under controlled stress (CS), in which the load is applied by the
pneumatic equipment at a frequency of 1 Hz (0.1s load and 0.9s rest period).
4
The fatigue life is expressed by Equation (3), where N is the number of loads to specimen
failure, the difference between compression and tension stress in the middle of the sample,
and k and n constants determined from the experimental results.
1 𝑛
𝑁 = 𝑘∙( ) (3)
∆𝜎
In this study, it is tested the levels of stress of 20, 30 and 40% of the 𝐼𝑇𝑆 at 25ºC. For
each stress level, three specimens of each mixture were tested and then a fatigue curve was
plotted.
Figures 2 (a), (b) and (c) show images of the tested mixes after the adhesion test. The
results obtained after DIP are shown in Figures 2 (d), (e) and (f), respectively.
With DIP analysis it was determined that M1-P mixture had the worst aggregate-binder
adhesion, as expected since no additive was used. Mixture M1-M presented the best result, while
M1-R presented an intermediate result, as shown in Table 3, which indicates the potential of
using a waste such as grounded plastic bag as an antistripping agent.
The results following NBR 15617/15 demonstrate that the antistripping agent increased
considerably the compatibility of the materials, reducing the moisture damage.
Considering only ITS of the first three non-moisture-conditioned mixtures, the addition of
the antistripping agent by itself did not produce a significant increase on its ITS. From M1-R
results, one note PB produces a considerable increase of 𝐼𝑇𝑆, when compared to M1-P and M1-
M, as shown in Figure 3. However, M1-M is much more efficient in combating moisture
damage, when compared to other mixtures. This is because positive charges of antistripping
agent increase the binding strength at the aggregate-binder interface, making moisture infiltration
difficult.
1.6
ITS, ITSc [MPa]
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
M1-P M1-R M1-M
ITSc [MPa] 0.33 0.69 0.94
ITS [MPa] 1.10 1.44 1.18
TSR [%] 30 48 79
Figure 4 shows the relation of the quantitative results of the adhesion determined by DIP
as presented in Table 3 with the results of the moisture damage as indicated in Figure 3. Despite
the reduce number of mixtures thus far in this research, one note a stronger correlation between
the areas determined by DIP with the 𝐼𝑇𝑆𝑐 of the conditioned specimens, which is not observed
for the non-conditioned specimens.
Considering the three analyzed mixtures and a minimum 𝑇𝑆𝑅 of 70% [2], a criterion of
𝐴𝐷𝐼𝑃 ≥86% could be established to classify mixtures with minimally desirable adhesion, giving
adequate moisture damage resistance.
6
1.6 100
y = 0.1291x + 1.1464 90 y = 121.36x - 34.543
1.4 R² = 0.02 R² = 0.99
80
ITSc. ITS [MPa]
1.2
70
TSR [%]
1.0 60 UNSATISFACTORY
0.8 50 ADHESION
y = 1.4024x - 0.3377 40
0.6 R² = 0.97
30
0.4
20
0.2 10
0.0 0
0.0
0.0 20
0.2 40
0.4 60
0.6 80
0.8 100
1.0 0.0
0.0 20
0.2 40
0.4 60
0.6 80
0.8 100
1.0
APDI [%] APDI [%]
Conditioned specimens Non-conditioned specimens ITSc/ITS
3.4 Fatigue
Figure 5 and the experimental fatigue coefficients in Table 4 show that M1-M and M1-P
present similar trends. Because M1-R has a higher ITS (1.44MPa) than the other two mixtures,
the applied target load was higher and consequently the difference of stresses obtained was
higher. This is a bias of the diametral compression test under controlled stress [3].
The number of loads to failure for M1-R was much higher when compared to the other
two mixtures. It is observed that the antistripping agent increased the fatigue life when compared
to the mix with pure binder.
100000 M1-P TABLE 4 Fatigue Test Data
M1-R
M1-M M1-P M1-R M1-M
N (Cycles)
100
0.2 2.0
∆σ (MPa)
FIGURE 5 Fatigue Life Curves
3. CONCLUSIONS
Both the grounded PB and the investigated antistripping agent decreased the
displacement of the bituminous film, the latter with greater effectiveness. The digital image
processing makes it possible to objectively determine the bituminous film coated area after the
adhesion test and this measurement apparently may be used to access the results of moisture
damage. Further investigation is necessary with a larger number of mixtures.
7
It is noticed that the commercial antistripping agent strongly contributes to prevent
moisture damage. The mixture containing this additive had a satisfactory coated area, produced
the best response with respect to moisture-conditioned, and also presented a potential to increase
fatigue life.
Considering the PBs, it was noticed a trend of considerable improvement related to an
increase in fatigue life. This study identified the potential of the investigated waste both to
prevent moisture related damage and to increase fatigue life.
4. REFERENCES