Professional Documents
Culture Documents
CTRG2015 Submission 462
CTRG2015 Submission 462
4 Research Scholar, Department of Civil Engineering, Indian Institute of Technology, Roorkee, India, niks.iitkgp88@gmail.com.
5 Praveen Kumar, Department of Civil Engineering, Indian Institute of Technology, Roorkee, India, pkaerfce@iitr.ac.in
6 Abstract: Study of creep and recovery for bitumen is gaining a lot of importance in the current time to evaluate the
7 rutting susceptibility of asphalt mix. Multiple Stress Creep and Recovery (MSCR) test using Dynamic Shear
8 Rheometer (DSR) is presently used to compare and study the behavior of different binders with respect to rutting.
9 This paper focuses on the study of creep and recovery response of different binders by varying different test
10 parameters, judging the applicability of the Boltzmann superposition principle, modelling the creep and recovery
11 response using simple mathematical models, analyzing the model parameters and evaluating its possible correlation
12 with the relative conventional binder properties.
13 It was found that the Boltzmann superposition principle is applicable to creep but not to the recovery response. A
14 simple power law model was used to fit the creep response and was modified to incorporate the delayed elastic
15 response of binder. The ratio of the model parameters for different bitumen was found to be related to the intrinsic
16 binder properties like softening point and viscosity. This relation was transformed to simple correlation equation for
17 modelling the creep and recovery response of binders.
19 1. Introduction
20 Increase in loading, high temperature and introduction of new axle configuration, demands
21 effective strengthening of pavements. Use of modified binders are among the most common
22 techniques, employed to improve the strength characteristics of bituminous mixes(Becker and
23 Méndez 2001; Santagata et al. 2013; Cardone et al. 2014). Amid various forms of modification,
24 polymers tends to increase the viscoelastic response of binder and reduces its temperature
25 susceptibility(Airey 1997; Isacsson and Lu 1999; Lu and Ekblad 1999; Airey 2002; Sengoz and
26 Isikyakar 2008). It increases the stiffness of the binder at higher temperature, while upholding the
27 flexibility at lower temperatures.
36 Modelling the response of any material to the imposed stress/strain has always been considered as
37 an important tool to study the material properties at a more discrete level. Many researchers have
38 tried to model the creep and recovery behavior of bitumen using different mathematical and
3rd Conference of Transportation Research Group of India (3rd CTRG)
39 rheological models (Mui 2008; Delgadillo and Bahia 2010; Ye et al. 2010; Celauro et al. 2012;
40 Delgadillo et al. 2012; Merusi 2012; Bai et al. 2014). However, no models have correlated the
41 model parameters with the conventional binder properties. This study attempts to model the creep
42 and recovery response of both unmodified and modified bitumen using a simple mathematical
43 equation and establish correlations of the model parameters of different unmodified and modified
44 bitumen with their predetermined flow properties.
45 2. Materials
46 Four binders were used in the study. VG10 and VG 30 were the viscosity graded binder. VG 10
47 was being modified with Styrene Butadiene Styrene (SBS) and Ethylene Vinyl Acetate (EVA) at
48 different percentages. Previous studies done by the authors (Saboo and Kumar 2014) have shown
49 that the interlocked phase of polymer with the base binder can be obtained using 3% SBS and 5%
50 EVA. Higher percentages yielded binders which were susceptible to phase separation. Also, using
51 lower percentages did not fully optimize the properties of the base binder, which resulted in
52 uneconomical blend. So for comparison only 3% SBS and 5% EVA was considered in the study.
53 The conventional properties along with the high temperature grade PG specification and true
54 intermediate grade temperature for these binders can be seen in Table 1. In this paper the polymer
55 modified bitumen will be represented as PMB (E) and PMB (S) indicating modification with EVA
56 and SBS. The authors also studied the optimum blending requirements for both the polymers.
57 Following the study, PMB (S) was modified at a temperature of 180 °C using a high shear mixture
58 operated at 1500 rpm for 60 minutes. PMB (E) on the other hand was modified at 190 °C, at a
59 shear rate of 600 rpm for 30 minutes.
62 3. Experimental
63 Dynamic shear rheometer (DSR) was employed for conducting multiple stress creep and recovery
64 (MSCR) test at three different temperatures (40, 50 and 60 °C). These temperatures ranges from
65 average to maximum for Indian highways. Four different stress levels were chosen, viz. 100, 3200,
66 5000 and 10,000 Pa. Test was carried out for all the binders mentioned in section 2.0. These are
67 the most common virgin and polymer modified binders used in the country.
68 Loading of 1s and unloading for 9s was given to simulate creep and recovery for 10 cycles at each
69 stress level. The amplitude of stress was increased at the end of 10th cycle of each stress level, till
70 10,000 Pa. First cycle of each stress, at a particular temperature, was chosen to model creep and
71 recovery response separately. It was hence necessary to bring the data on a single scale of time.
72 The creep and recovery of the first cycle, at stress level of σi+n was shifted to σi by subtracting each
73 value of strain for creep and recovery of σi+n by the strain value of σi+n-1. Figure 1 shows the
74 procedure of shifting the axis. Applicability of Boltzmann superposition principle was checked for
75 the final shifted data. If Boltzmann superposition is assumed to be applicable, then
76 i k i n / i . i (1)
77 Where, ‘i’ is the known data and ‘n’ is the point for which the applicability has to be checked.
78 Checking this applicability using such technique would automatically enclose the effect of cycle
79 on creep and recovery, as the data for σi+k are shifted from higher cycle numbers.
80
93 For modelling creep, a Power Law Model was used of the form
94 (t ) K A.t B (2)
95 K is the assumed elastic response and the second term is the time dependent viscoelastic response.
96 A and B are the model shape parameters. It was found that this model gave excellent fit for all the
97 binders at all the temperatures and stress levels. Figure 3 shows the fit for all the binders at 60 °C.
98 Plot at other temperatures are not shown due for brevity. The value of parameter K was found to
99 be zero for all the considered cases. This implies that there is no dominance of the elastic response
100 of the binder at temperatures greater than 40 °C. The next task was to explicitly analyze the model
101 parameters A and B.
VG 10, 40 °C VG 10, 50 °C
1 kPa 10 1 kPa
10 3.2 kPa
3.2 kPa
5 kPa 5 kPa
10 kPa 10 kPa
1 kPa Sup 1 kPa Sup
Strain
3.2 kPa Sup
Strain
0.001 0.001
0 0.5 1 1.5 0 0.5 1 1.5
Time, Sec Time, Sec
102
VG 10, 60 °C VG 30, 40 °C
100 1 kPa 1 1 kPa
3.2 kPa 3.2 kPa
5 kPa 5 kPa
10 kPa 10 kPa
1 kPa Sup 0.1 1 kPa Sup
3.2 kPa Sup 3.2 kPa Sup
Strain
Strain
0.01 0.001
0 0.5 1 1.5 0 0.5 1 1.5
Time, Sec Time, Sec
103
VG 30, 50 °C VG 30, 60 °C
10 1 kPa 100 1 kPa
3.2 kPa 3.2 kPa
5 kPa 5 kPa
10 kPa 10 kPa
1 kPa Sup 1 kPa Sup
Strain
Strain
0.1 5 kPa Sup 1 5 kPa Sup
10 kPa Sup
0.001
0 0.5 1 1.5 0.01
0 0.2 0.4 0.6 0.8 1 1.2
Time, Sec Time, Sec
104
0.0001 0.001
0 0.5 1 1.5 0 0.2 0.4 0.6 0.8 1 1.2
Time, Sec Time, Sec
105
5 kPa
Strain
0.01
0.001 0.0001
0 0.5 1 1.5 0 0.5 1 1.5
Time, Sec Time, Sec
106
Strain
5 kPa Sup 0.1 5 kPa Sup
0.01
0.001 0.001
0 0.5 1 1.5 0 0.5 1 1.5
Time, Sec Time, Sec
107
108 Figure 2 Validating Boltzmann superposition principle for different bitumen. (*Sup-
109 Superimposed)
110
VG 10, 60 °C VG 30, 60 °C
100 1 kPa 100
3.2 kPa 1 kPa
5 kPa 3.2 kPa
10 kPa 10 5 kPa
10 10 kPa
1 kPa Fit
1 kPa Fit
Strain
0.1 0.1
0.01 0.01
0 0.5 1 1.5 0 0.5 1 1.5
Time, Sec Time, Sec
111
0.01 0.01
0.001 0.001
0 0.5 1 1.5 0 0.5 1 1.5
112 Time, Sec Time, Sec
115 The parameter B was found to be independent of the stress level at each temperature. However, it
116 showed a linear relationship with temperature. Table 2 presents the average values obtained for all
117 the type of binders at different test conditions. An interesting fact to be noted was that, the ratio of
118 their averages, was almost constant irrespective of any temperature, as can be seen in Table 3. The
119 ratio was taken corresponding to VG 10, just for referencing. This ratio was found to be same as
120 the ratio of the corresponding softening point for the binders. Figure 4 shows the correlation.
0.9
PMB (S)
0.7 PMB (E)
0.7 0.9 1.1
Ratio of Softening Point
125
127 Appreciating the above mentioned fact, the result can be now mathematically written as
SPk
128 B ,T Bk ,T (3)
SP
129 Where, is any binder for which the value is to be determined, provided value of B is known for
130 binder k at temperature T.
132 The parameter A was found to be function of both stress level and the temperature. It was found to
133 obey a power law behavior with respect to the stress level as shown in figure 5. Mathematically,
134 A M i N (4)
135 N represents the dependence of strain on the stress level. For linear behavior it remains close to 1.
136 It was found close to 1 except for stress levels higher than 5000 Pa.
Modelling A at 40 °C VG 10
Modelling A at 50 °C VG 10
VG 30 8 VG 30
1.4
SBS SBS
7
1.2 R² = 0.9999
EVA R² = 0.9999 EVA
Value of A
Power (VG 5 Power (VG
Value of A
30) 30)
0.8 Power R² = 0.9999 Power
(SBS) 4 (SBS)
Power Power
0.6 (EVA)
3 (EVA)
0.4 R² = 1 2
R² = 1
0.2 1
R² = 1 R² = 0.9997
0 0
0 5000 10000 15000 0 5000 10000 15000
Stress Level, Pa Stress Level, Pa
137
Modelling A at 60 °C VG 10
35 VG 30
R² = 0.9997 SBS
30
EVA
25
Value of A
Power
(VG 10)
Power
20 (VG 30)
R² = 0.9992 Power
(SBS)
15 Power
(EVA)
10
R² = 0.9998
5
R² = 0.9946
0
0 5000 10000 15000
Stress Level, Pa
138
140 Instead of going into the details of too many values obtained for different binders at various stress
141 levels and temperatures, ratio of A of VG 30, PMB (S) and PMB (E) were taken with respect to
142 VG 10. For a particular temperature this ratio was found to be invariant with stress level. Table 4
143 presents the average value obtained for all the binders.
146 The ratio, at a particular temperature showed excellent correlation with the inverse of the ratio of
147 Zero Shear Viscosity (ZSV) of the binders. This portrays the temperature sensitivity of PMBs in
148 comparison to the virgin binders. It can be seen that the increase in the ratio is higher for PMB (E),
149 especially at higher temperature. At 40 °C the value of PMB (S) and PMB (E) is almost close to
150 each other, but as the temperature increases, the corresponding values differ to a very high degree.
151 This shows the shift of behavior of other binders towards more viscous regime, while EVA
152 maintains a higher degree of stiffness with increase in temperature. Figure 6 shows the plot of both
153 the ratios at the three temperatures studied.
Ratio of A , Model
0.6 0.6
0.4 0.4
PMB (E) PMB (E)
0.2 0.2
PMB (S)
PMB (S)
0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
1/Ratio of ZSV 1/Ratio of ZSV
154
0.6
0.4
0 PMB (S)
0 0.2 0.4 0.6 0.8
1/Ratio of ZSV
155
ZSV ,T
158 A ,T Ak ,T (5)
i ZSVk ,T
159 So, if the value of A and B for any binder ‘k’ is known, the corresponding values for any other
160 binder ‘ ’ can be calculated using the above relationship.
ZSV ,T (
SPk
) Bk
162 (t )T , , Ak ,T t
SP
(6)
i
ZSV k ,T
163 4.2. Modelling Recovery
164 If Boltzmann superposition is to be valid for both creep and recovery phenomena, then the recovery
165 could be written as
167 Where, the suffix ‘r’ stands for the recovery region and t̕ is the time at which the load is removed.
168 Similar to creep the recovery phase for all the binders were found to deviate from the Boltzmann
169 superposition principle. These deviations were higher than those found in creep interval. This is
170 attributed to the delayed elastic response of the binders present in the recovery phase, which are
171 highly susceptible to change in temperature and stress levels.
172 To remove these small deviations between the modelled and experimental data a factor of ‘n’ was
173 introduced as a multiple of B, for modelling recovery of all the binders. The expression now can
174 be written as
176 This factor, n, accounts for the nonlinearity of all the binders. For linear materials the factor will
177 be equal to 1.
178 It was found that the above model gave excellent fit for all the binders, at all temperatures and
179 stress levels. As an example, the fit for PMB (S) and VG 30 is shown at 400C and 600C at two
180 similar stress levels in Figure 7. All the curves are not displayed due to space restrictions. But it is
181 worth mentioning that all data comply with the stated results. The parameter B.n (considered as α)
182 was found to have similar correlation as the factor B. so, the factor α could be written as
SPk
183 ,T k ,T (9)
SP
184 The model of recovery for any binder ψ, provided the model parameters of binder k is known, can
185 be written as
ZSV ,T (
SPk
) Bk ZSV ,T (
SPk
) k
186 (t )T , , Ak ,T t
SP
Ak ,T (t t )
' SP
(10)
i
ZSV k ,T ZSV k ,T
Strain
0.006 0.3
0.004 0.2
0.002 0.1
0 0
0 5 10 15 0 5 10 15
Strain
0.1 6 Experimental
Experimental
0.08
4
0.06
0.04 2
0.02
0
0
0 5 10 15
0 5 10 15
Time, Sec Time, Sec
188
100 Pa, 40 °C, PMB (S) 3200 Pa, 40 °C, PMB (S)
0.003 0.14
0.0025 Fit 0.12 Fit
0.1
0.002
Experimental Experimental
0.08
Strain
Strain
0.0015
0.06
0.001
0.04
0.0005 0.02
0 0
0 5 10 15 0 5 10 15
Time, Sec Time, Sec
189
5000 Pa, 60 °C, PMB (S) 10000 Pa, 60 °C, PMB (S)
2.5
0.04
Fit Fit
0.035 2
0.03
Experimental 1.5 Experimental
0.025
Strain
Strain
0.02
1
0.015
0.01 0.5
0.005
0 0
0 5 10 15 0 5 10 15
Time, Sec Time, Sec
190
191 Figure 7 Modified power model fit for VG 30 and SBS at 40 and 600C.
192 5. Conclusions
193 The creep and recovery test was conducted on four different bitumen’s. It was found that
194 Boltzmann superposition principle was applicable in the creep domain but deviated when applied
195 to recovery. A simple power model showed to give excellent fit to the measured creep response at
196 all the temperatures studied. The power law was modified to incorporate the delayed elastic
197 response for the recovery period.
198 The parameter B showed a linear relationship with temperature. The ratio of the averages, was
199 almost constant irrespective of any temperature. The ratio was found to be same as the ratio of the
200 corresponding softening point for the binders. The parameter A was found to be function of both
201 stress level and the temperature. It displayed a power law behavior with respect to the stress level.
202 For a particular temperature the ratio of A was found to be invariant with stress level. This ratio
203 was at a particular temperature showed excellent correlation with the inverse of the ratio of Zero
204 Shear Viscosity (ZSV) of the binders. The parameter B.n (considered as α) in the recovery model
205 was found to have similar correlation as the factor B. The equations relating the model parameters
206 with the intrinsic binder properties is presented and could be very useful to establish the creep and
207 recovery behavior of any binder through simple measurements of conventional properties.
208 Recommendation
209 The above mentioned technique is empirical and may change depending on the type of binder.
210 Validation is required to gain more confidence on such results. The change in strain with change
211 in temperature for a single binder was not correlated with any of the conventional test parameters,
212 as these parameters are normally calculated at a single set temperature. New empirical tests like
213 penetration at multiple temperatures, or measuring the time the ball will take to fall, at different
214 temperatures in a softening point apparatus can be useful to investigate and relate the complex
215 rheological properties to simple conventional properties.
216
217
218 References
219 Airey GD (2002) Rheological evaluation of ethylene vinyl acetate polymer modified bitumens.
220 16:473–487.
221 Airey GD (1997) Rheological Characteristics of Polymer Modified and Aged Bitumens.
222 University of Nottingham
223 Bahia HU, Zhai H, Zeng M, et al (2001) Development of binder specification parameters based on
224 characterization of damage behavior (with discussion). Asphalt Paving Technology 2001.
225 Bai F, Yang X, Zeng G (2014) Creep and recovery behavior characterization of asphalt mixture in
226 compression. Constr Build Mater 54:504–511. doi: 10.1016/j.conbuildmat.2013.12.088
228 Biro S, Gandhi T, Amirkhanian S (2009) Determination of zero shear viscosity of warm asphalt
229 binders. Constr Build Mater 23:2080–2086. doi: 10.1016/j.conbuildmat.2008.08.015
230 Cardone F, Ferrotti G, Frigio F, Canestrari F (2014) Influence of polymer modification on asphalt
231 binder dynamic and steady flow viscosities. Constr Build Mater 71:435–443. doi:
232 10.1016/j.conbuildmat.2014.08.043
233 Celauro C, Fecarotti C, Pirrotta a., Collop a. C (2012) Experimental validation of a fractional
234 model for creep/recovery testing of asphalt mixtures. Constr Build Mater 36:458–466. doi:
235 10.1016/j.conbuildmat.2012.04.028
236 D’Angelo J (2009a) Current Status of Superpave Binder Specification. Road Mater Pavement Des
237 10:13–24. doi: 10.1080/14680629.2009.9690233
238 D’Angelo J a. (2009b) The Relationship of the MSCR Test to Rutting. Road Mater Pavement Des
239 10:61–80. doi: 10.1080/14680629.2009.9690236
240 D’Angelo J, Reinke G, Bahia H, et al (2010) Development in Asphalt Binder Specifications (E-
241 C147).
242 Delgadillo R, Bahia HU (2010) The Relationship between Nonlinearity of Asphalt Binders and
243 Asphalt Mixture Permanent Deformation. Road Mater Pavement Des 11:653–680. doi:
244 10.1080/14680629.2010.9690298
245 Delgadillo R, Bahia HU, Lakes R (2012) A nonlinear constitutive relationship for asphalt binders.
246 Mater Struct 45:457–473. doi: 10.1617/s11527-011-9777-y
247 Isacsson ULF, Lu X (1999) Characterization of bitumens modified with SEBS , EVA and EBA
248 polymers. 4:3737–3745.
249 Lu X, Ekblad UI (1999) Rheological properties of SEBS , EVA and EBA polymer modified
250 bitumens. 32:131–139.
251 Merusi F (2012) Delayed mechanical response in modified asphalt binders. Characteristics,
252 modeling and engineering implications. Road Mater Pavement Des 13:321–345. doi:
253 10.1080/14680629.2012.657096
254 Mui J (2008) Viscoelastic-viscoplastic model to predict creep in a random chopped mat
255 thermoplastic composite.
256 Saboo N, Kumar P (2014) Optimum Blending Requirements for EVA Modified Binder.
257 Proceedings of 11th Transportation Planning and Implementation Methodologies for
258 Developing Countries. IIT Bombay, India,
259 Santagata E, Baglieri O, Dalmazzo D, Tsantilis L (2013) Evaluation of the anti-rutting potential
260 of polymer-modified binders by means of creep-recovery shear tests. Mater Struct 46:1673–
261 1682. doi: 10.1617/s11527-012-0006-0
262 Sengoz B, Isikyakar G (2008) Evaluation of the properties and microstructure of SBS and EVA
263 polymer modified bitumen. Constr Build Mater 22:1897–1905. doi:
264 10.1016/j.conbuildmat.2007.07.013
265 Ye Y, Yang X, Chen C (2010) Modified Schapery’s Model for Asphalt Sand. J Eng Mech
266 136:448–454. doi: 10.1061/(ASCE)EM.1943-7889.0000092
267
268
269