Estimating the Thermal Conductivity of Asphalt Binders

Abstract

Nowadays, there is a common worldwide interest in environmental issues and pavements. How to save energy and mitigate the urban heat island (UHI) effect are topics that are drawing the attention of different researches and industrial organizations. In road infrastructure, one of the important properties addressing environmental and UHI aspects of pavements is the determination of the thermal conductivity. Asphalt concrete represents the third most widely used resource in the world, with asphalt-paved roads being its principal usage. One of the most important components of asphalt concrete is bitumen. Bitumen is a viscoelastic material susceptible to temperature changes. The determination of the bitumen´s thermal conductivity becomes very important in understanding and improving its thermal performance. There are very few test methods and equipment to measure thermal conductivity of bitumen (asphalt binders). Some are expensive and require special equipment and instrumentation. This study developed and validated a simplified testing technique to measure thermal conductivity of asphalt binders. This test is a steady state-based method to estimate the thermal conductivity of asphalt binders using cylindrical samples poured into a silicon mold. The method was validated using material of known thermal conductivity. Eighteen samples of different binder grades were tested, and the test results were repeatable and within known thermal conductivity values. Sensitivity analysis and accuracy of the proposed method were validated modifying the asphalt binder with a material with a very low thermal conductivity. This method to estimate thermal conductivity of bitumen samples was found to provide an affordable alternative test procedure with good accuracy and precision.

Keywords: Thermal conductivity, bitumen, steady-state method, heat transfer rate

Full text

Estimating the Thermal Conductivity of Asphalt Binders
1
Carlos J. Obando, Ph.D. (c).1 and Kamil E. Kaloush, Ph.D., P.E.2
2
1School of Sustainable Engineering and the Built Environment, Arizona State University, P.O.
3
Box 873005, Tempe, Arizona 85287-3005; e-mail: cobandog@asu.edu
4
2School of Sustainable Engineering and the Built Environment, Arizona State University, P.O.
5
Box 873005, Tempe, Arizona 85287-3005; e-mail: kamil.kaloush@asu.edu
6
7
8
Published in: Journal of Testing and Evaluation (ASTM)
9
DOI: 10.1520/JTE20210208
10
11
12
ABSTRACT
13
Nowadays, there is a common worldwide interest in environmental issues and pavements.
14
How to save energy and mitigate the urban heat island (UHI) effect are topics that are drawing the
15
attention of different researches and industrial organizations. In road infrastructure, one of the
16
important properties addressing environmental and UHI aspects of pavements is the determination
17
of the thermal conductivity. Asphalt concrete represents the third most widely used resource in the
18
world, with asphalt-paved roads being its principal usage. One of the most important components
19
of asphalt concrete is bitumen. Bitumen is a viscoelastic material susceptible to temperature
20
changes. The determination of the bitumen´s thermal conductivity becomes very important in
21
understanding and improving its thermal performance. There are very few test methods and
22
equipment to measure thermal conductivity of bitumen (asphalt binders). Some are expensive and
23
require special equipment and instrumentation. This study developed and validated a simplified
24
testing technique to measure thermal conductivity of asphalt binders. This test is a steady state-
25
based method to estimate the thermal conductivity of asphalt binders using cylindrical samples
26
poured into a silicon mold. The method was validated using material of known thermal
27
conductivity. Eighteen samples of different binder grades were tested, and the test results were
28

2
repeatable and within known thermal conductivity values. Sensitivity analysis and accuracy of the
29
proposed method were validated modifying the asphalt binder with a material with a very low
30
thermal conductivity. This method to estimate thermal conductivity of bitumen samples was found
31
to provide an affordable alternative test procedure with good accuracy and precision.
32
33
Keywords: Thermal conductivity, bitumen, steady-state method, heat transfer rate
34
35
INTRODUCTION
36
There is a common worldwide interest in environmental issues and pavements. One aspect
37
is the mitigation of the urban heat island (UHI) effect. In road infrastructure, one of the important
38
materials properties in addressing the UHI of pavements is the determination of the thermal
39
conductivity. The thermal conductivity is a physical property that is also related to the performance
40
of the materials, which implies the energy transfer rate or heat transfer rate (Q) that occur when
41
bodies in contact have different temperatures1.
42
43
Heat can be transferred from one point to another by three different processes: conduction,
44
convection and radiation2. Conduction can occur in solids, and in liquids when there is no
45
macroscopic movement. Convection occurs when liquids are in movement, and radiation occurs
46
in the vacuum or air. These modes of heat transfer are governed by different laws, Fourier, Newton,
47
and Stefan-Boltzmann, respectively
48
49
This document addresses the conduction phenomenon of bitumen. In conduction, heat is
50
transmitted through a material medium and there is no transport of matter. The rate at which heat
51
is transferred through the material (dQ/dt) is represented by the letter Q and is called the heat flow
52

3
rate. Empirically, the heat flow rate is proportional to the cross-sectional area (A) to the direction
53
of the flow, to the temperature difference on both sides of the material (∆T), and inversely
54
proportional to the distance traveled from the place at the highest temperature (∆x) [3]. That is:
55
56


 
 (1)
57
58
To achieve the equality of the previous expression, a constant k is added, which is the
59
thermal conductivity, the intrinsic ability of a material to transfer or conduct heat3 4.
60
61
  
   
 (2)
62
63
The conduction into a cylindrical geometry, introduces the Equation 35 6.
64
65
   


(3)
66
67
Where:
68
69
Q: heat flow rate (W=joule/s)
70
A: the cross-sectional area (m2)
71
ΔT: the temperature gradient (oC)
72
Δx: the thickness (m)
73
k: the thermal conductivity (W/moK)
74
t: time (s)
75
h: length/height of the sample (m)
76

4
r1: inner radius (m)
77
r2: outer radius (m)
78
79
Thermal conductivity is inherent to each material and expresses the ability of a given
80
material to conduct heat3. Thermal conductivity can be affected by moisture, ambient temperature,
81
and the density of the material. If moisture, temperature, and density are increased, the thermal
82
conductivity rises too, so thermal conductivity is not constant1.
83
84
The exactitude of different methods for calculating thermal conductivity is extensively
85
debated in several fields. In addition, the wide range of thermal characteristics of different
86
materials generated several methods for the estimation of thermal conductivity7.
87
88
Along the last two decades, the accuracy and the understanding of the principles of heat
89
transfer have been improved for several materials. These techniques present different ranges of
90
thermal conductivity even for the same material, different accuracy, temperature ranges, and
91
specimen type8.
92
93
Although there are many methods to estimate thermal conductivity, there are few for
94
specific materials like bitumen, or asphalt binder. There are two basic methods. The first one is a
95
group of steady‐state methods, and the second one is called the transient or a group of non‐steady‐
96
state methods7 9. The implementation of each method depends on the characteristic of the materials.
97
All methods are based on electrical analogy and on the essential laws of heat conduction. Steady‐
98
state methods are mathematically simpler8, while transient heat transfer methods are efficient to
99

5
determine thermal diffusivity. However, steady‐state methods are known as the most accurate for
100
testing dry materials10.
101
102
The steady‐state technique is related to an equilibrium state, then, these methods consider
103
the data to do the calculations when a material reaches a constant temperature. As a disadvantage,
104
to reach a steady temperature takes a long time1. In addition, these methods involve expensive
105
equipment and difficult experimental set-up installation. Nonetheless, steady-state methods are the
106
most accurate and the main measurement methods. The non‐steady‐state or transient methods take
107
measurements during the heating progression. These techniques estimate thermal conductivity
108
using transient sensors. The time needed in these methods is relatively quick, which is the most
109
important advantage over the steady‐state systems11. Table 1 shows a summary of the principal
110
characteristics of the various methods.
111
112
In the Civil Engineering field, asphalt concrete represents the third most widely used
113
material in the world, with asphalt-paved roads being its principal usage. One of the most important
114
components of asphalt concrete is bitumen, a residue of oil distillation processes. Bitumen is a
115
highly susceptible viscoelastic material to temperature changes. This can be brittle as glass at low
116
temperature and flow like oil at high temperatures12. From this conception, the determination of
117
thermal conductivity of the bitumen becomes very important to understand and improve its thermal
118
performance.
119
120
Analytical models in different studies have been used to calculate the thermal conductivity.
121
However, the accuracy of each model and technique is constricted by the physical properties and
122

6
other factors of each material to test. Therefore, quantity and modeling of thermal conductivity are
123
complex and need high precision. The approaches and the models used to study thermal behavior
124
of materials must be clearly defined8.
125
126
Table 1. Summary of methods used for the determination of the thermal conductivity8
127
Method
Usage
Uncertainly
Estimation
Range of
Temperature
Advantages
Disadvantages
Steady‐ state methods
Guarded hot
plate
Solids, insulator
materials
2% –5%
−93°C – 127°C
High level of accuracy
Long measurement time,
low conductivity
materials, large
specimen size
Heat‐flow meter
Rocks, polymers,
insulations, plastics,
glasses, ceramics, some
metals
3% –10%
(normal), 0.5% –
2% (axial) and
3% –15% (radial)
−100°C–200°C
(normal), -183°C–
126°C (axial heat
flow), and 25°C –
2326°C (radial heat
flow)
Easy operation and
construction
Relative measurement,
uncertainly
Cylinder
Metals
2%
-269°C – 727°C
Simultaneous
estimation of electrical
conductivity, and
temperature range
Long measurement time
Pipe method
Calcium, silicates, solids,
refractory fiber blankets
and minerals
3% –20%
20°C – 2500°C
Good temperature range
Long measurement time,
specimen set up
Comparative
Plastics, metals, ceramics
10% –20%
20°C – 1300°C
Simple construction and
operation
Relative measurement,
uncertainly
Direct heating
Tubes of electrical
conductors, metals,
wires, rods
2% –10%
127°C – 2727°C
Easy and fast
measurements,
simultaneous estimation
of electrical
conductivity
Limited to electrically
conducting materials
Transient Methods
Hot disk (TPS
technique)
Solids, powders, liquids,
pastes
--
247°C – 927°C
Diverse thermal
properties
simultaneously, and
accuracy
Conducting or insulating
material
Hot wire
Hot strip
Hot wire: Solids, liquids,
glasses, plastics,
granules, powders
Hot strip: Ceramics,
glasses, foods
1% –10 % hot
wire
5% –15% hot
strip
20°C –2 000°C, −40–
1600°C for hot wire
and −50°C to 500°C
for hot strip, 25°C –
1527°C for hot wire
Fast, accuracy, and,
good temperature range
Only for low
conductivity materials
Photothermal
(PT)
Photoacoustic
Thin films, solids,
liquids, gases
1%–10 % for PT
−50°C – 1500°C, and
-73°C – 527°C for PT
Operational for liquids,
gases, and thin films
Unknown accuracy,
Nonstandard
Laser flash
Polymer, ceramics,
solids, liquids, powders,
metals
1.5% –5 %
−373°C – 3027°C
Good temperature
range, accuracy at high
temperature, small
specimens, fast
Expensive, not for
insulation materials
128
129
Based on the unique characteristics of bituminous materials, and the need to know their
130
thermal properties for better understanding the potential improvement when using various
131
modification techniques, this document presents an alternative method for determining the thermal
132
conductivity of bitumen, while addressing issues like cost and accuracy.
133

7
EXPERIMENTAL APPROACH
134
Calibration
135
The determination of the thermal conductivity of bitumen samples using the method
136
described in this document was first used on material of known characteristics and thermal
137
conductivity. The calibration sample used was acrylic glass (Plexiglas V045i), which has a known
138
thermal conductivity range between 0.17W/moK and 0.20W/moK13 14; in confirmation, and
139
following the method developed at The National Center of Excellence for SMART Innovations at
140
ASU15, the thermal conductivity of this material was estimated as 0.1852W/moK.
141
142
As it was explained above, thermal conductivity is related to the heat transfer rate, which
143
is central in the estimation of thermal conductivity in this method. Due to the unique characteristics
144
of the bitumen/asphaltic binder, the medium to transfer the heat was chosen as distilled water in
145
no macroscopic movement. Then, the temperature transfer from the outside to the sample is
146
realized using non-turbulent, distilled water. At the liquid-solid interface, the main mechanisms
147
contributing to heat transfer are convection and conduction. However, the present work restricts
148
the domain study to the sole solid sample. This assumption is sustained by the fact that the bitumen
149
is considered as a solid. Therefore, it is possible to restrict the heat transfer rate (Q) calculation to
150
a conduction-driven mechanism only, using Equation (3).
151
152
The heat flow rate is independent of radial location but varies depending on the temperature
153
of the water; therefore, it was necessary to calibrate the model measuring the heat flow rate at
154
several temperature points. This calibration method compares different water temperatures and the
155

8
resulting heat flow rate, knowing the thermal conductivity, and the acrylic-sample’s geometrical
156
features.
157
158
To determine the heat flow rate, we need to measure the two final steady temperatures in
159
the system. In this approach, the outer temperature is the water temperature being controlled by
160
the water bath, and the inner temperature is the one in the center of the acrylic sample. Note that
161
the “system” includes all instrumentation features like water bath, thermocouple types and
162
accuracy, and thermometers, which are described next.
163
164
Instrumentation
165
To avoid the interference of air currents that could alter the temperature readings and make
166
it more difficult reaching the steady state temperatures, the experimental setup was employed
167
inside a chamber conditioned at 25oC. To control the water temperature, a water bath (Thermo
168
Scientific, 180 Series, Model: Precision) was used. For temperature measurements, J type thermal
169
couples (-40 to 510 oC) were used, and a software LabVIEW 8.6 with a DAQ system were used to
170
record the temperature changes along with time. To check the accuracy of the temperature
171
readings, a high precision thermometer (Precision RTD Handheld Data Logger Thermometer) was
172
used.
173
174
The acrylic samples used to calibrate the model were cylindrical shaped. The samples are
175
40mm in diameter (r2, the outer radius is then 19mm), and 25mm in height (h), with a hole of 2mm
176
diameter (r1=1mm, which is the inner radius) in the center of the top circular face, extending to the
177
middle of the sample. Figure 1 shows the cylinder’s geometry.
178

9
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
Figure 1. Acrylic sample characteristics
194
195
196
To measure the thermal conductivity in steady state using a conduction method, it is
197
necessary to ensure that the heat flow goes only in one direction. A balsa wooden platform was
198
used to place the samples inside the water bath. This setup is needed to avoid the water outer
199
temperature affecting the inner temperature in the center of the acrylic cylinder. An isolator foam
200
was used on the top of the acrylic sample, and a high vacuum grease silicone on the bottom. This
201
particular grease has sealing ability and at the same time excellent resistance to water.
202
Additionally, because the relative high specific heat capacity, 2900 J/kgK16, a very low thermal
203
conductivity, 0.045W/moK17, of the balsa wood, the very low power in the system (e.g. 0.09W at
204
47oC), and the short time of the test (2 hours), it is considered that no significant heat enters from
205
the bottom of the sample. The samples were submerged into the water bath taking care that the
206
level of water goes near the edge of the top circular face. For temperatures above 50oC, it is
207
recommended to cover partially the water bath to avoid water evaporation. Figure 2 shows the
208
complete setup.
209
210

10
211
212
213
214
215
216
217
218
219
220
221
222
223
224
Figure 2. Complete set-up of the calibration test
225
226
227
The final data needed to calculate Q are the steady-state temperatures. Figure 3 shows
228
examples of the temperature change recorded for various samples versus time. The steady-state
229
temperatures are those when the inner (center of the sample) and outer (water) temperatures reach
230
a unchanging condition.
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
Figure 3. Temperature vs. Time
247
248
249
After certain time period, the temperatures get to the steady state. Note that the time needed
250
to get to the steady state may vary; however, for this setup the usual time was 1.5 hours. Once the
251

11
steady-state temperatures are reached, it is recommended to continue recording readings for at
252
least 30 minutes and calculate the average value.
253
254
From Equation 3, knowing the geometrical characteristics of the specimen (refer Figure 1),
255
the thermal conductivity constant of the acrylic material (k), and the difference between inner and
256
outer temperatures (ΔT), it is possible to calculate the heat flow rate for each temperature. Figure
257
4 shows Q for different water temperatures ranging between 31oC and 82oC.
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
Figure 4. Heat flow rate (Q) as a function of the temperature (oC)
275
276
277
Note that the acrylic material was used to estimate the heat flow rate of the system, which
278
has a thermal conductivity of 0.185W/moK. Equation 4 represents the results of the calibration
279
process to find Q as a function of water temperature; it can be used subsequently to calculate the
280
thermal conductivity (k) .
281
282
󰇛󰇜  (4)
283
284

12
Where:
285
Q: heat flow rate (W=joule/s)
286
T1: the outer temperature (water temperature) (oC)
287
288
The estimation of ¨Q¨ is an important step to determine the thermal conductivity of any test
289
samples of interest using Equation 3. It depends on the thermal conductivity of the acrylic
290
calibration sample being used. Therefore, the constant number 0.0153 in the exponential Equation
291
4 may change. In addition, the exponent part constant 0.0412 in the equation would remain the
292
same if the system components being used are kept unchanged. This is because the exponent
293
constant in the equation is dependent on the system configuration (e.g. water bath characteristics).
294
295
Thermal conductivity of bitumen/asphalt binder
296
To employ the above test procedure, it is needed to produce asphalt-binder samples with
297
similar dimensions to the acrylic cylinders. Therefore, special molds are needed to be made and
298
used to pour in them the hot asphalt binder. The material used to create the mold was a commercial
299
product that consists of two liquid substances. These substances need to be mixed in a specific
300
proportion to get the raw silicone material. This silicone material can support temperatures above
301
300oC. Figure 5 shows the silicone container / mold used to produce the asphalt binder samples
302
for testing.
303
304
305
306
307
308

13
309
310
311
312
313
314
315
316
317
318
319
320
321
Figure 5. Silicon mold used for asphalt binder samples production and testing (“h” corresponds to the
322
inner depth of the mold).
323
324
325
The use of silicone molds is very convenient due to their flexibility. Once the hot binder is
326
poured in the mold and cooled down, a 2mm diameter hole is drilled in the center from the top to
327
the middle of the cylinder height, similar to the acrylic cylinder test procedure described earlier.
328
The hole in the center is made using a heated metallic rod or a screwdriver, both with appropriate
329
diameters. As the air inside the samples can affect thermal conductivity, it is important to pour the
330
material in the mold as hot as possible and leave it to cool down slowly undisturbed at room
331
temperature. Before drilling or removing the samples from the mold, it is recommended to place
332
the asphalt binder samples inside a freezer for 20 minutes at -10 oC. Figure 6 shows how the binder
333
samples look like inside the silicone mold, and Figure 7 shows how those binder samples look like
334
when removed from the containers.
335
336
337
338
339
340

14
341
342
343
344
345
346
347
348
349
Figure 6. Asphalt binder sample inside the silicone mold
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
Figure 7. Asphalt binder samples ready to test
367
368
369
Thermocouples are placed in the center hole, and the samples are placed on wooden
370
platform. The isolator foam is placed on top, and high vacuum grease silicone on the bottom of
371
the circular face of each sample. The grease helps the samples get locked on the wooden platform,
372
isolating water at the bottom, and avoiding samples getting stuck. Figure 8 shows the final setup
373
of the test before adding the foam on top. Note that the level of the water is just at the edge of the
374
samples.
375
376
377
378

15
379
380
381
382
383
384
385
386
387
388
Figure 8. Setup of the thermal conductivity test on binders
389
390
391
Based on known thermal susceptibility of the asphalt binders, it is recommended to perform
392
the test between 28oC and 40oC to avoid the softening of the samples. The temperature of the water
393
would vary depending on the type of binder being evaluated. For softer binders such us PG58-22,
394
and PG64-16, the recommended maximum test temperature is 28oC, which is about 15oC below
395
their softening point measured with the ring and ball method (ASTM E28 – 67). For stiffer binders
396
such as PG76-22, the recommended maximum test temperature is 33oC. Binders modified with
397
polymers can be tested up to 40oC.
398
399
This method could be implemented using any type of water-bath following the calibration
400
step described earlier in this document. In earlier experiments, the authors had also good success
401
in using type K thermocouples with an automatic USB output thermometer, and/or manually
402
registering temperatures with time. Care in selecting, manipulating and calibrating the
403
thermocouples and water-bath will provide repeatable and accurate results.
404
405

16
This method can be used to calculate the thermal conductivity of any material with
406
impermeable properties, so the water in the system, which controls the outer temperature, cannot
407
get to the center of the sample where the inner temperature is taken.
408
409
RESULTS OF THE IMPLEMENTATION
410
Eighteen samples of different virgin binders (PG58-22, PG64-16 and PG76-22) provided
411
by HollyFrontier in Arizona, were tested using this developed method. Binders PG58-22 and
412
PG64-16 are unmodified bitumen used for hot mix asphalt, emulsion production or further
413
modification for higher temperature paving grades; whereas binder PG76-22 is a modified asphalt
414
cement used for hot mix asphalt. The softest of these asphalt binders is PG58-22 and the stiffest is
415
PG76-2218. Soft binders are more susceptible to temperatures changes and flow more at high
416
temperature than stiff binders. It is also noted that lower ability of the binders to conduct heat
417
(lower k) means better thermal resistance.
418
419
To get the heat flow rate (Q) for the three asphalt binders, Equation 4 of the base calibration
420
model were used. Once Q is found, thermal conductivity is calculated based on the Equation 3
421
solving for Thermal Conductivity (k). As it was mentioned before, the whole system is employed
422
inside a chamber setup at 25oC, and the resulting thermal conductivity is estimated under this
423
condition. Table 2 and Figure 9 show all the test results presenting the coefficient of variance
424
(COV) and the standard error respectively. The average test results for each binder grade produced
425
repeatable outcomes that are similar to known thermal conductivity values as shown below; the
426
coefficient of variation was also under 10% for each binder. While the average thermal
427
conductivity between the binder grades are statistically the same, there seem to be a trend of having
428

17
slightly lower thermal conductivity for stiffer binders. This result is rational as one would expect
429
a PG76-22 binder with a polymer modification should have lower thermal conductivity compared
430
to a conventional softer binder. Figure 9 present the results showing the standard error.
431
432
Table 2. Thermal conductivity of different binders
433
Binder Type
Sample No.
Sample's
Height
h (m)
Sample's
radius
r2 (m)
Sample's
radius
r1 (m)
Outer Temp.
(water) T1 (C)
Flow Rate Q
(W)
From Eq 4.
Sample's
Inner Temp.
T2 (C)
k (W/moK)
From Eq 3.
Av. k
(W/moK)
COV
Binder PG58-22
1
0.0250
0.019
0.010
31.37
0.05572
30.23
0.200
0.209
0.08
2
0.0250
0.019
0.010
31.15
0.05521
30.16
0.227
3
0.0250
0.019
0.010
31.15
0.05521
29.89
0.179
4
0.0250
0.019
0.010
31.20
0.05533
30.18
0.222
5
0.0250
0.019
0.010
31.00
0.05487
29.95
0.214
6
0.0250
0.019
0.010
31.40
0.05578
30.32
0.211
Binder PG64-16
7
0.0250
0.019
0.010
31.37
0.05572
30.16
0.188
0.204
0.07
8
0.0250
0.019
0.010
31.15
0.05521
30.1
0.210
9
0.0250
0.019
0.010
31.15
0.05521
29.9
0.183
10
0.0250
0.019
0.010
31.20
0.05533
30.15
0.215
11
0.0250
0.019
0.010
31.00
0.05487
29.98
0.220
12
0.0250
0.019
0.010
31.40
0.05578
30.31
0.209
Binder PG76-22
13
0.0250
0.019
0.010
31.37
0.05572
30.25
0.203
0.198
0.08
14
0.0250
0.019
0.010
31.15
0.05521
29.9
0.177
15
0.0250
0.019
0.010
31.15
0.05521
29.9
0.177
16
0.0250
0.019
0.010
31.20
0.05533
30.10
0.206
17
0.0250
0.019
0.010
31.00
0.05487
29.95
0.214
18
0.0250
0.019
0.010
31.40
0.05578
30.31
0.209
434
435
436
437
438
439
440
441
442
443
444
Figure 9. Average thermal conductivity of binder with the standard error
445
Salt remains

18
Thermal conductivity of asphalt binders can range between 0.17W/moK and 0.28W/moK
446
19 20. From Table 2, it is possible to see that k varies between 0.23W/moK and 0.18W/moK (note
447
that in the calculation, Celsius degrees are transformed to Kelvin). Thermal conductivity results of
448
the different binders could confirm the better thermal resistance of the binder PG76-22, one of the
449
reasons to choose hard binders for a better asphalt pavement performance in hot climates.
450
451
To better demonstrate the capability of the developed test method in capturing a different
452
thermal conductivity value for modified binders, the PG76-22 binder was modified with 5% Enova
453
Aerogel. Aerogel is a material with extremely low thermal conductivity of about 0.012 (W/ moK)21
454
22. The effect of the Aerogel’s low thermal conductivity when added to binder was studied using
455
the proposed method. For sample preparation, once the PG76-22 binder reached 165oC in the oven,
456
5% of Enova Aerogel by weight of binder was added and blended manually using a wooden stick
457
for about 1 minute. The Aerogel modified binder along with a control were tested using the
458
proposed method and the thermal conductivity results obtained are shown in Table 3.
459
460
461
Table 3. Thermal conductivity of unaged binder PG76-22 with different methods
462
Binder Type
Sample No.
Sample's
Height
h (m)
Sample's
radius
r2 (m)
Sample's
radius
r1 (m)
Outer
Temp.
(water) T1
(C)
Flow
Rate Q
(W)
From
Eq 4.
Sample's
Inner
Temp.
T2 (C)
k (W/moK)
From Eq 3.
Average k
(W/moK)
COV
Control
1
0.0250
0.019
0.010
33.91
0.06185
32.71
0.211
0.199
0.06
2
0.0250
0.019
0.010
33.91
0.06185
32.63
0.198
3
0.0250
0.019
0.010
33.91
0.06185
32.56
0.188
5% Aerogel
1
0.0250
0.019
0.010
33.91
0.06185
32.25
0.153
0.166
0.08
2
0.0250
0.019
0.010
33.91
0.06185
32.49
0.179
3
0.0250
0.019
0.010
33.91
0.06185
32.38
0.166
463

19
The results supported the capability of the proposed test method in capturing lower thermal
464
conductivity values, as expected, for the Aerogel modified binder. The precision for the modified
465
samples was slightly lower, most likely due to the difficulty in uniformly distributing the Aerogel
466
particles in the binder samples.
467
468
469
CONCLUDING REMARKS
470
The determination of thermal conductivity of the asphalt binders is very important in the
471
understanding and improvement of its thermal performance. There are very few test methods and
472
equipment to measure thermal conductivity of asphalt binders. Some of those are expensive and
473
require special equipment and instrumentation. This study developed and validated a simplified
474
alternative testing technique to measure thermal conductivity of asphalt binders. The determination
475
of the thermal conductivity of bitumen samples using the method described was validated on
476
material of known thermal conductivity. In addition, eighteen samples of different binder grades
477
were tested using the developed method. The average test results were repeatable and within
478
known thermal conductivity values reported in the literature; the coefficient of variation between
479
the various samples were in the 7 to 8% range. Additionally, the sensitivity and capability of the
480
proposed method to capture lower thermal conductivity values were proven by using an Aerogel
481
modified binder. This method to estimate thermal conductivity of bitumen samples was found to
482
provide an affordable alternative test procedure with good accuracy and precision.
483
484
DATA AVAILABILITY STATEMENT
485
The data that support the findings of this study are available from the corresponding author,
486
Carlos Obando, upon request.
487

20
ACKNOWLEDGMENTS
488
The authors would like to thank the Global Kaiteki Center at Arizona State University for
489
the funding support. Additional support was provided by The National Center of Excellence for
490
SMART Innovations and the Advanced Pavement Laboratory at ASU. Based on the Program
491
Colombia Cientifica focuses/challenges related to Sustainable Energy, this work serves as a tool
492
for Sustainable Construction and a Cleaner Transportation development. The authors would like
493
to acknowledge the invaluable support provided by the Colombian Program Colombia Cientifica
494
and the Scholarship Fulbright - Pasaporte a la Ciencia.
495
496
497
AUTHOR CONTRIBUTIONS
498
In this article, Carlos Obando performed all laboratory experiments, designed the
499
experimental plan, analysis of the results, and participated in the development of the research topic.
500
Kamil E. Kaloush provided an overall guidance for the research conduct, interpretation of the test
501
results, and editing the manuscript.
502
503
504
505
506
507
508
509
510

21
REFERENCES
511
1 A. Vimmrová and J. Výborný, Building Materials 10: Materials and Testing Methods,
512
Vydavatelství ČVUT, 2002, p. 98.
513
2 J. H. Lienhard IV and J. H. Lienhard V., A Heat Transfer Texbook, Cambridge,
514
Massachusetts, U.S.A.: Phlogiston Press, 2003.
515
3 J. R. Welty, C. E. Wicks, R. E. Wilson and G. L. Rorrer, Fundamentals of Momentum, Heat,
516
and Mass Transfer. 5th Edition, John Wiley & Sons, Inc., 2007.
517
4 J. G. Speight, Rules of Thumb for Petroleum Engineers, John Wiley & Sons, 2017.
518
5 A. Mills, Solution Manual for Heat Transfer, Second Edition, New Jersey: Prentice-Hall,
519
1998.
520
6 J. Holman, Heat Transfer, Tenth Edition ed., New York: McGraw-Hill, 2010.
521
7 N. Yüksel, "The investigation of structure and operating parameters effect on the heat transfer
522
coefficient in porous structures," Uludag University, Bursa, 2010.
523
8 N. Yüksel, "The Review of Some Commonly Used Methods and Techniques to Measure the
524
Thermal Conductivity of Insulation Materials," in Insulation Materials in Context of
525
Sustainability, IntechOpen, 2016. https://doi.org/10.5772/64157
526
9 N. Yüksel, A. Avcı and M. Kılıç, "The effective thermal conductivity of insulation materials
527
reinforced with aluminium foil at low temperatures," Heat and Mass Transfer, vol. 48, pp.
528
1569-1574, 2012. https://doi.org/10.1007/s00231-012-1001-2
529
10 N. Mohesnin, Ed., Thermal Properties of Food and Agricultural Materials, 1 ed., New York:
530
Gordon and Breach Science Publishers, 1980, p. 408.
531
11 H. Czichos , T. Saito and L. Smith, Eds., Springer Handbook of Materials Measurement
532
Methods, Springer Science & Business Media, 2006, p. 1208.
533
12 C. DeDene, J. Gorman, M. Marasteanu and E. Sparrow, "Thermal conductivity of reclaimed
534
asphalt pavement (RAP) and its constituents," International Journal of Pavement Engineering,
535
vol. 17, no. 5, pp. 435-439, 2016. http://dx.doi.org/10.1080/10298436.2014.993201
536
13 Goodfellow, ""Polymethylmethacrylate - online catalogue source - supplier of research
537
materials in small quantities," 2019. [Online]. Available:
538
http://web.archive.org/web/20210330050225/https://www.google-analytics.com/ga.js.
539
[Accessed 20 November 2019].
540
14 Engineering ToolBox, "Plastics - Thermal Conductivity Coefficients," 2011. [Online].
541
Available:
542
http://web.archive.org/web/20210330050557/https://www.engineeringtoolbox.com/thermal-
543
conductivity-plastics-d_1786.html. [Accessed 20 November 2019].
544
15 J. D. Carlson, R. Bhardwaj, P. E. Phelan, K. E. Kaloush and J. S. Golden, "Determining
545
Thermal Conductivity of Paving Materials Using Cylindrical Sample Geometry," ASCE
546
Journal of Materials in Civil Engineering, vol. 22, no. 2, pp. 186-195, 2010.
547
https://doi.org/10.1061/(ASCE)0899-1561(2010)22:2(186)
548
16 Engineering ToolBox, "Specific Heat of Solids," 2003. [Online]. Available:
549
http://web.archive.org/web/20210330044802/https://www.engineeringtoolbox.com/specific-
550
heat-solids-d_154.html. [Accessed 20 February 2021].
551
17 N. J. Kotlarewski, B. Ozarska and B. K. Gusamo, "Thermal Conductivity of Papua New
552
Guinea Balsa Wood Measured using the Needle Probe Procedure," BioResources, vol. 9, no.
553
4, pp. 5784-5793, 2014. https://doi.org/10.15376/biores.9.4.5784-5793
554

22
18 Marathon Petroleum, "Asphalt, Emultions and Roofing," 2018. [Online]. Available:
555
http://web.archive.org/web/20210612023356/https://www.mpcasphalt.com/Products/.
556
[Accessed 10 June 2021].
557
19 Engineering ToolBox, "Thermal Conductivity of selected Materials and Gases," 2003.
558
[Online]. Available:
559
http://web.archive.org/web/20210330045112/https://www.engineeringtoolbox.com/thermal-
560
conductivity-d_429.html. [Accessed 7 December 2019].
561
20 J. Co
te, V. Grosjean and J.-M. Konrad, "Thermal conductivity of bitumen concrete," Canadian
562
Journal of Civil Engineering, vol. 40, pp. 172-180, 2013. https://doi.org/10.1139/cjce-2012-
563
0159
564
21 R. Pekala, "Organic aerogels from the polycondensation of resorcinol with formaldehyde,"
565
Journal of Materials Science, vol. 24, p. 3221–3227, September 1989.
566
https://doi.org/10.1007/BF01139044
567
22 Cabot Corporation, "Enova Aerogel, Brochure," Billerica, USA, 2013.
568
569