The Performance and Distribution of Polyurethane-Modified Asphalt That Exhibits Different Molecular Weights

Abstract

To analyze the effect of polyol on polyurethane (PU)-modified asphalt, three different soft segments of polyurethane were synthesized, and we utilized the reaction of MDI (diphenylmethane diisocyanate) with PU650, PU1000, and PU1400. With respect to molecular weight, the effect of polyol on the performance of modified asphalt was analyzed, and the asphalt was modified by using three different polyurethanes. To analyze the PU samples, the Fourier transform infrared spectroscopy (FTIR) tests and gel permeation chromatography (GPC) tests were selected; by contrast, to analyze the rheological properties and modification mechanism of asphalt, the dynamic rheology test (DSR), low-temperature bending creep test (BBR), multi-stress repetitive creep test (MSCR), FTIR, and differential scanning calorimetry (DSC) were selected. The results indicate that the molecular weight of polyol affects the molecular structure of polyurethane, the distribution of soft and hard segments, the content of soft segments, and the distribution of asphaltene in asphalt; thus, the asphalt modification effect occurs differently. The storage stability and high-temperature stability of the polyurethane-modified asphalts that were synthesized using three different polyols (i.e., polyols that exhibit different molecular weights) did not differ considerably, and the PU1400-modified asphalt exhibited the best low-temperature performance.

1. Introduction

Asphalt pavements exhibit effective levelness, are comfortable driving surfaces, and generate low noise levels; thus, they have become prominent in China. However, due to overload and environmental factors (e.g., temperature, sunlight, water, and air), which affect aging, the performance of asphalt undergoes significant changes [1]. Currently, an increasing awareness of road safety issues and high pavement performance requirements (e.g., asphalt performance, material composition, grade composition, pavement structure design requirements, self-healing asphalt [2,3]) have crucially impacted the utilization of pavement performance specifications, and asphalt performance is directly related to the utilization of asphalt pavement performance specifications. Therefore, to enhance pavement performance, the asphalt performance should be improved [4]. Polymers can effectively enhance the properties of asphalt [4,5,6]; polyurethane-modified asphalt is an effective contemporary modification that can enhance the overall performance of asphalt [7,8], and its modification mechanism differs from that of rubber-modified asphalt.

Polyurethane (PU) resin exhibits excellent properties such as toughness, wear resistance, oil resistance, and aging resistance. It is widely utilized in binders, construction, packaging materials, and the automotive industry [4,8,9,10]. Polyurethane is mainly composed of urethane, which is produced by the reaction between polyol and isocyanate at high temperatures, and this reaction mainly entails chemical cross-linking. Urethane exhibits large cohesion energy and large polarity and rigidity. In addition, it easily forms hydrogen bonds and it exhibits aggregation (i.e., forms a hard segment micro-phase area); by contrast, soft chain segments (e.g., polyethers and polyesters) are less polar and they aggregate to form soft segment micro-phase regions. Elasticity and low-temperature properties are influenced by the soft segment micro-phase regions, whereas hardness and tensile strength are influenced by the hard segment micro-phase regions [4,11,12,13]. The existing literature indicates that PU-modified asphalt exhibits an effective performance and that the structure of the PU modifier considerably impacts the performance of the asphalt [9,14,15,16]; therefore, to meet the performance needs of different pavement structures, different structures of PU-modified asphalt can be prepared. Bazmara et al. [17] observed that during the modification of polyurethane, new chemical bonds were formed, chemical reactions occurred, and the modified asphalt exhibited a significantly enhanced high-temperature stability. Zhang et al. [14] demonstrated that the -NCO group that constitutes the polyurethane prepolymer reacted with the -OH group that constitutes the asphalt; thus, the high- and low-temperature properties and the mechanical properties of the modified asphalt were significantly enhanced. Xia et al. [18] observed that the utilization of a polyurethane prepolymer that exhibits a -NCO content of 3% to 6% as an asphalt modifier can effectively enhance the high- and low-temperature performance of asphalt. Zhang et al. [16] analyzed polyurethane-modified asphalt that is utilized in bridge deck pavements, and they observed that the modified asphalt exhibited satisfactory strength, flexibility, and storage stability; in addition, the low-temperature crack resistance of the mixture was significantly better than that of the epoxy asphalt mixture. Therefore, polyurethane-modified asphalt can effectively enhance certain properties of asphalt; however, differences in the raw materials can affect the performance of modified asphalt. Herein, we used isocyanate and polyol that exhibit different molecular weights, independently prepared a PU prepolymer, and analyzed the effect of molecular weight on polyurethane. In addition, we prepared PU-modified asphalts that exhibit different molecular weights, and to investigate the effect of different soft and hard chain segment structures and ratios on PU-modified asphalt, we measured its performance.

With respect to studies on polyurethane-modified asphalt, many researchers [14,16,17,19,20,21] focus on the preparation of polyurethane-modified asphalt that utilizes pre-polymerization or compounding. However, few researchers utilize different molecular weights of polyol and isocyanate to prepare modified asphalt and to analyze the effect that the molecular weight of polyol exerts on the performance of modified asphalt. For example, Zhang et al. [14] observed that the performance of polyurethane-modified asphalt that was prepared by the pre-polymerization system was significantly enhanced and that under optimal dosing conditions, the low-temperature performance and water stability of the modified asphalt mixture were better than those of the SBS-modified asphalt mixture. Zhang et al. [16] noted the following: modified asphalt that is prepared by compound modification (i.e., using polyurethane and epoxy asphalt) can effectively solve the cracking problem that affects flexible bridge decks under low-temperature conditions. Bazmara et al. [17] found that the high-temperature performance, penetration, and softening point of modified asphalt were increased when the PU content exceeds 3%, but the low-temperature performance had little effect. Martin et al. [20] found that modified asphalt with MDI-PPG prepolymer could change the structure of asphalt, and the research showed that chemical bonds can be formed between the modifier and polar substances in the asphalt, which can eventually form a three-dimensional network structure. Sun et al. [19] observed that MDI-PPG prepolymer could be used to modify asphalt. The results showed that PU-modified asphalt had better water resistance and deformation resistance compared with SBS-modified asphalt and virgin asphalt, it was a cost-effective material because its cost was similar to that of SBS modified asphalt. The epoxy resin and polyurethane composite-modified asphalt prepared by Sun et al. [21] could effectively enhance the high- and low-temperature properties of asphalt binders, and the modified asphalt exhibited certain economic and environmental advantages.

There are many types of polyurethane-modified asphalt, and the performance of the different types of polyurethane-modified asphalts varies greatly. Therefore, to comprehensively and systematically analyze the performance of different types of polyurethane-modified asphalt, we selected different molecular weights of polyol and isocyanate as raw materials, prepared different types of polyurethane-modified asphalt, and analyzed the effect of molecular weight on the performance of the modified asphalt.

Herein, we investigated polyurethanes that were synthesized using different polyols (i.e., polyols that exhibit different molecular weights) and their modification effect on asphalt. First, we explored the polyurethane synthesis process by using Fourier transform infrared spectroscopy (FTIR) and gel permeation chromatography (GPC) tests. Subsequently, penetration, softening point, ductility, and the rheological properties of prepared polyurethane-modified asphalt were evaluated. Furthermore, using differential scanning calorimetry (DSC), the properties of the polyurethane-modified asphalt were analyzed.

2. Raw Materials and Test Methods

2.1. Raw Materials

2.1.1. Polyol

Three different types of polyether polyols (molecular weights: 650, 1000, and 1400) were selected (BASF polyether polyol, Jining Huakai Resin Company, Jining, China). The polyether polyol was polytetrahydrofuran (PTHF), also known as polytetramethylene glycol (PTMG);

Table 1. Indices of polyether polyols that exhibit different molecular weights.

Projects	PTMG650	PTMG1000	PTMG1400
Appearance (23 °C)	White liquid	White liquid	White liquid
Viscosity (40 °C cp)	205	270	660
Hydroxyl value (mgKOH/g)	166.2–175.5	106.9–118.1	80.2–84.3
Average molar mass (g/mol)	650 ± 25	1000 ± 50	1500 ± 50
Melting point (Tm, °C)	19	24	27

illustrates the chemical indices.

The molecular structure formula of polyether polyol is illustrated in Figure 1.

2.1.2. Isocyanate

Next, a diphenylmethane diisocyanate mixture that exhibits 2,4′-MDI (liquid state; room temperature) was added. The mixture, which is a common isocyanate that is utilized in industrial production [22], is known as liquefied MDI-50 (C-MDI); specifically, we utilized the Wanhua brand isocyanate (liquefied MDI-50) that is produced by Yantai Wanhua Company, Yantai, China and the technical specifications are depicted in

Table 2. MDI-50 technical specifications.

Projects	Indicators	Test Results
Color and status	Colorless or yellowish transparent liquid	Slightly yellow, transparent liquid
Purity	≥99.6%	——
2,4′-MDI isomer content	50~54%	52%
-NCO mass fraction	30~35%	33.5%
Relative density (50 °C/4 °C)	1.19	——
Functionality	2	2
Hydrolysis chlorine mass fraction	≤0.005%	——

.

2.1.3. Asphalt

The Kyobo 70# virgin asphalt was selected, and the basic technical specifications are depicted in

Table 3. Technical specifications of the virgin asphalt.

Projects	Test Results	Technical Requirement	Test Method
Penetration (25 °C; 100 g; 5 s) (0.1 mm)	68.2	60~80	ASTM D5
Softening point (universal method) (°C)	46.9	44~54	ASTM D36
Density (g/cm3)	1.023	-	ASTM D70
Solubility (trichloroethylene) (%)	99.7	≥99.5	ASTM D2042
Quality loss/m%	0.052	−0.8~0.8	ASTM D6
Penetration ratio/%	63.7	≥58	ASTM D5
10 °C ductility/cm	25	≥20	ASTM D113

.

2.2. Preparation

2.2.1. Polyurethane Prepolymer

Urethane is the main structure that constitutes polyurethane, and it is mainly generated by the reaction between isocyanate and hydroxyl compounds, which is also the main reaction for the preparation of polyurethane materials; the chemical equation is depicted in Figure 2. Herein, the polyurethane prepolymer was obtained by homemade means. Polyurethane prepolymer that is prepared using a homemade method enables researchers to control the variables in the reaction process; in addition, it ensures the uniformity of the raw material batch that is utilized, which improves the accuracy and reliability of the test. To synthesize PU prepolymers that exhibit different molecular weights, three polyether polyols that exhibited different molecular weights (PTMG650, −1000, and −1400) and a polyisocyanate (MDI-50) were reacted with R = 2.

2.2.2. Polyurethane-Modified Asphalt

The virgin asphalt was heated to 135 °C in an oven, and the shear machine was used for shearing 20 min at a constant temperature. Subsequently, the appropriate amount of compatibilizer was added, and to achieve the grafting reaction, which occurs between the compatibilizer and the asphalt, and to change the material’s interfacial tension, the resultant mixture was sheared for 50 min. In addition, to fully dissolve the asphalt, a chain expansion crosslinker was added. Furthermore, the PU prepolymers (5% prepolymers of the asphalt mass) with different molecular weights were added and sheared at 135 °C for 10 min; thus, they were fully dispersed in the polymer system. Finally, to ensure that the asphalt modifier was fully modified and that the curing segregation phenomenon did not appear, it was placed in the oven for 2.5 h for low-temperature maintenance. The preparation process is shown in Figure 3.

2.3. Polyurethane Testing

2.3.1. FTIR Testing

The chemical structure and physical properties of polyurethane were analyzed using FTIR (Verter 70 type, Bruker, Karlsruhe, Germany) testing (resolution, 4 cm⁻¹; scanning range, 4000 cm⁻¹~400 cm⁻¹).

2.3.2. GPC Test

GPC (Waters 2695 type, Waters, MA, USA) is a method of separating polymer samples that entails passing polymeric materials through a porous gel column. The method is based on the volumetric difference that characterizes molecular hydrodynamics, and it involves varying the elution time.

2.4. Polymer-Modified Asphalt Performance Testing

2.4.1. Physical Characteristics Testing

The penetration at 25 °C and the softening point and ductility at 5 °C according to ASTM D5, ASTM D36, and ASTM D113 were tested, respectively; furthermore, using a viscosity test, we evaluated the physical properties of the modified asphalt.

2.4.2. Storage Stability

To observe the micro-phase structure of the modified asphalt, an LW300LET fluorescence microscope (Shanghai Jovian Photoelectric Technology Co., Ltd., Shanghai, China) with 1000 times magnification was selected. To analyze the storage stability of the modified asphalt, the prepared PU-modified asphalt was poured into an aluminum sample tube weighing approximately 50 g, placed vertically at 163 °C for 48 h, and stored in a freezer for more than 4 h. Subsequently, the aluminum tube was cut into three sections of equal length under the condition of soft asphalt, and based on the standard test, the softening point difference was tested; the difference should be less than 2.5 °C [4,14,24]. To ensure satisfactory storage stability, the test process ensured that the asphalt that was contained in the aluminum tube remained in an undisturbed state.

2.4.3. Rheological Performance Test

(1)

DSR test

With respect to high-temperature performance, the rutting resistance factor G*/sinδ and phase angle δ were utilized to evaluate the performance of the asphalt by the DSR (SmartPave102, Anton Paar, Graz, Austria) test. The test temperature was set to 58–82 °C, the heating rate was set to 2 °C/min, the strain value of the specimen was set to 1%, and the load action frequency was set to 10 rad/s.

(2)

BBR test

To evaluate the low-temperature performance of the asphalt in the linear viscoelasticity range, namely at −6 °C, −12 °C, and −18 °C, the BBR (TE-BBR type, Cannon, NS, USA) test utilized the flexural creep stiffness modulus (S) and creep rate (m).

(3)

Multi-stress repeat test (MSCR)

Using the multi-stress repetitive creep recovery (MSCR) test, the creep recovery of the asphalt under repeated loading and unloading was evaluated; thus, we evaluated the high-temperature performance of the asphalt. The MSCR (SmartPave102 type, Anton Paar, Graz, Austria) test was conducted on a dynamic shear rheometer (DSR) at an experimental temperature of 64 °C. Short-term aged asphalt specimens were tested at two stress levels, namely 0.1 kPa and 3.2 kPa, with 1 s of loading and 9 s of unloading at each stress level, which was repeated for 10 cycles.

The evaluation indices of the MSCR asphalt test were the deformation recovery rate (R), the irrecoverable creep flexibility (J_nr) at two stress levels (i.e., 0.1 kPa and 3.2 kPa), and the stress sensitivity parameters (i.e., J_nr-diff and R_diff) for the aforementioned indices, which were calculated as follows.

Non-recoverable creep flexibility (J_nr):

$${\mathrm{J}}_{\mathrm{nr}}\left(\mathsf{\sigma },\mathrm{N}\right)=\frac{{\mathsf{\epsilon }}_{\mathsf{\gamma }}-{\mathsf{\epsilon }}_0}{\mathsf{\sigma }}$$

(1)

Deformation recovery rate (R):

$$\mathrm{R}\left(\mathsf{\sigma },\mathrm{N}\right)=\frac{{\mathsf{\epsilon }}_{\mathrm{c}}-{\mathsf{\epsilon }}_{\mathsf{\gamma }}}{{\mathsf{\epsilon }}_{\mathrm{c}}-{\mathsf{\epsilon }}_0}\times 100\%$$

(2)

where $\mathsf{\sigma }$ denotes the applied stress level in kPa, representing either the 0.1 kPa or 3.2 kPa level; N denotes a certain loading cycle; R denotes the recovery rate; and ${\mathrm{J}}_{\mathrm{nr}}$ denotes the amount of unrecovered creep flexibility in kPa⁻¹.

Equations (1) and (2) were combined to calculate the average recovery rate (i.e., R_0.1 and R_3.2) and the irrecoverable flexibility (i.e., J_nr0.1 and J_nr3.2) for 10 load cycles under two stress levels; this represents the evaluation index of the high-temperature performance of the asphalt. Subsequently, the stress sensitivity index (i.e., R_diff and J_nr-diff) of the asphalt was calculated. In the actual test process, the creep recovery test at 0.1 kPa exhibited 20 repetitive load cycles. However, the first 10 load cycles of asphalt strain response had not yet stabilized; thus, the creep recovery test at 0.1 kPa could not accurately reflect the asphalt creep characteristics. To calculate the evaluation indices, the measured data pertaining to the last 10 cycles could be utilized. The calculation formula is as follows:

Average recovery rates R_0.1 and R_3.2

$${\mathrm{R}}_{0.1}={{\displaystyle \sum }}_{\mathrm{N}=11}^{20}\mathrm{R}\left(0.1,\mathrm{N}\right)÷10$$

(3)

$${\mathrm{R}}_{3.2}={{\displaystyle \sum }}_{\mathrm{N}=1}^{20}\mathrm{R}\left(3.2,\mathrm{N}\right)÷10$$

(4)

Unrecovered creep flexibility J_nr0.1 and J_nr3.2.

$${\mathrm{J}}_{\mathrm{nr}0.1}={{\displaystyle \sum }}_{\mathrm{N}=11}^{20}{\mathrm{J}}_{\mathrm{nr}}\left(0.1,\mathrm{N}\right)÷10$$

(5)

$${\mathrm{J}}_{\mathrm{nr}3.2}={{\displaystyle \sum }}_{\mathrm{N}=1}^{10}{\mathrm{J}}_{\mathrm{nr}}\left(3.2,\mathrm{N}\right)÷10$$

(6)

The stress sensitivity indices (i.e., R_diff and J_nr-diff) for asphalt specimens at two stress levels were as follows:

$${\mathrm{R}}_{\mathrm{diff}}=\frac{{\mathrm{R}}_{0.1}-{\mathrm{R}}_{3.2}}{{\mathrm{R}}_{0.1}}\times 100\%$$

(7)

$${\mathrm{J}}_{\mathrm{nr}\text{-}\mathrm{diff}}=\frac{{\mathrm{J}}_{\mathrm{nr}3.2}-{\mathrm{J}}_{\mathrm{nr}0.1}}{{\mathrm{J}}_{\mathrm{nr}0.1}}\times 100\%$$

(8)

2.4.4. Microscopic Performance Testing

To analyze the glass transition temperature of the polymer-modified asphalts, the DSC (200 F3 type, Netzsch, Bavaria, Germany) test was utilized; thus, the low-temperature properties of the polymer-modified asphalts could be characterized. Here, the heating rate was set to 10 k/min, there was nitrogen protection, the flow rate was set to 20 mL/min, and the test temperature range was set at −40~140 °C.

3. Results and Analysis

3.1. Polyurethane Test Results

3.1.1. FTIR Test Results

To analyze the changes that affect the position and intensity of the absorption peaks that characterize the functional groups of the materials, to qualitatively and quantitatively analyze the materials, and to study the modification mechanism of polyurethane and modified asphalt, we utilized the FTIR test. The FTIR curves that represent MDI50, PTMG1000, and MDI-PTMG prepolymer (PUP) and the FTIR spectral curves of prepolymers of three different molecular weights are illustrated in Figure 4.

From the comparison of the FTIR curves pertaining to the three materials (Figure 4), with respect to the RCOOH group that is unique to the polyol, the wave peaks of the prepolymer are mostly superimposed on the wave peaks of the isocyanate and polyol, namely at 2853 cm⁻¹, and similar wave peaks characterize the prepolymer. However, some emergent or peak size changes remain, which indicates that the functional group corresponding to this wave position changed and that a chemical reaction occurred. A change in the absorption peak was observed at 2272 cm⁻¹, where the characteristic absorption peak of -NCO that is unique to isocyanate was observed, and the peak of the prepolymer decreased significantly after the polymerization reaction, which indicates that the reaction consumed a large amount of -NCO groups. The C=O (i.e., stretching vibration) peak of urethane at 1730 cm⁻¹, which is present in only the prepolymer, indicates that the PU prepolymer that exhibits an -NCO capped end was experimentally synthesized.

In Figure 5, with respect to carbamate and ureidoformate, the stretching vibration peak of -NH appears at 3300 cm⁻¹. After zooming in, we observed that the peaks of the three prepolymers of different molecular weights exhibited a regular distribution, with the highest content of -NH groups in PUP650, with the smallest molecular weight, and when the molecular weight was increased, the content of -NH groups decreased. Methyl and methylene stretching vibrational peaks were observed at 2858 cm⁻¹, and with respect to the prepolymer, their content patterns contrast with the -NH group content.

3.1.2. GPC Test Results

The GPC distribution curves of the three prepolymers of different molecular weights are depicted in Figure 6, and the number average molecular weight (Mn), heavy average molecular weight (Mw), and z-average molecular weight (Mz) data are illustrated in

Table 4. Molecular weight statistics of PU prepolymers.

Molecule Type	Mn (Number Average Molecular Weight)	Mw (Heavy Average Molecular Weight)	MP (Peak Molecular Weight)	Mz (Z-Average Molecular Weight)	Dispersibility (d)
PUP650	5436	6749	6335	8115	1.2415
PUP1000	7147	8545	6710	9947	1.1956
PUP1400	7317	8265	6768	9310	1.1302

.

As illustrated in Figure 6, the three prepolymer isocyanates were composed of the same raw material, and when the molecular weight of the polyol increased, the molecular weight of the prepolymer increased. This indicates that during the synthesis of a polyurethane prepolymer, PTMG1400 can yield polyurethane that exhibits a high molecular weight; the PTMG1400 molecule contains more -OH, which enhances the -NCO reaction in MDI.

Table 4 indicates that the number average molecular weight was the highest in PUP1400, followed by PUP1000 and PUP650 in order of magnitude, which indicated the following: when the molecular weight was increased, the molecular weight of the synthesized polyurethane increased under the same preparation conditions.

3.2. Analysis of Polyurethane-Modified Asphalt Test Results

3.2.1. Physical Properties Test Results

The three main indicators of virgin asphalt and polyurethane-modified asphalt were tested, and the results are depicted in Figure 7.

Figure 7 indicates that compared with virgin asphalt, the three different types of polyurethane-modified asphalt exhibited superior performance. Figure 7a indicates that the penetration exhibited by the three polyurethane-modified asphalts did not differ considerably, which indicates that the effect of the three polyurethane-modified asphalts on the high-temperature performance was nearly constant. Figure 7b,c indicate that the softening point and ductility of the three polyurethane-modified asphalts were significantly different and that as the molecular weight increased, the softening point became larger and the ductility of the modified asphalt increased. When the molecular weight is increased, the -OH group that constitutes the polyol consumes the free -NCO group; thus, the number of -NCO groups decreases. In addition, polyether polyol is a soft satin; when the molecular weight increases, the soft segment becomes larger, and the low-temperature performance is enhanced.

The viscosity of the asphalt bond was evaluated [25], and the test results are depicted in Figure 7d. When the temperature was increased, the viscosity of the three polyurethane-modified asphalts decreased, and compared with the virgin asphalt, the viscosity of the modified asphalts improved to different degrees. This is mainly due to the formation of a three-dimensional network in the modified asphalts, which increased the resistance to molecular movement [4]. In addition, when the molecular weight increased, the viscosity gradually increased, which is mainly related to the degree of cross-linking; with the molecular weight increasing, the degree of cross-linking increases, the modified asphalt exhibits increased resistance to movement, and its viscosity increases [26].

3.2.2. Storage Stability Test Results

Due to the difference in solubility and density between the polymer and the asphalt, polymer-modified asphalt is prone to phase separation, which occurs during hot storage [27,28]. Therefore, to analyze the distribution of the modifier in the asphalt and to verify that the modified asphalt met the storage stability requirements, fluorescence microscopy and segregation tests were used.

Figure 8 indicates the fluorescence microscopy test results of the three PU-modified asphalt samples with different molecular weights. In the three samples, the modifiers, which were in the form of fine particles, were uniformly dispersed in the asphalt, and when the molecular weight was small, the dispersion became more uniform.

The storage stability test results of the polyurethane-modified asphalt samples with different molecular weights are illustrated in Figure 9. Usually, the limit of the softening point difference is 2.5 °C, and below this value, modified asphalt exhibits satisfactory storage stability [29,30]. Figure 8 indicates that when the molecular weight was increased, the softening point difference gradually became larger and that all the differences were below 2.5 °C. This finding coheres with the fluorescence microscopy results. Therefore, all three samples of polyurethane-modified asphalt exhibited satisfactory storage stability, despite polyurethane and asphalt being two different substances that exhibit a large difference in polarity, which results in different storage stability. The reaction between the compatibility agent and some components of the polyurethane enhanced the polarity of the asphalt and polyurethane, which improved the storage stability [31].

3.2.3. Rheological Performance Test Results

(1)

DSR test results

The DSR test, which utilizes the phase angle (δ) and rutting resistance factor (G*/sinδ), can be utilized to evaluate the viscoelasticity and high-temperature rutting resistance of polyurethane-modified asphalts of different molecular weights [32]. The test temperature ranged from 58 to 82 °C, and the results are as follows.

Figure 10a indicates that compared with the virgin asphalt, the rutting resistance factor (G*/sinδ) of the three polyurethane-modified asphalt samples was significantly enhanced, and the difference in the rutting resistance factor was not significant. Thus, with respect to enhancing the rutting resistance factor of asphalt, the difference in the effect of these three polyurethane samples was not significant.

Figure 10b indicates that with respect to the virgin asphalt, the phase angle of all three polyurethane-modified asphalt samples was significantly lower, which indicates that all three polyurethane-modified asphalt samples exhibited enhanced elasticity; due to the addition of polyurethane to the asphalt, a different structure was formed within the asphalt [8,31]. PUP650 exhibited the smallest phase angle, which indicates that polyurethane with a small molecular weight is more likely to improve the elasticity of asphalt at high temperatures.

(2)

Multi-stress repetitive creep test results

We observed that the high-temperature performance of polyurethane-modified asphalt can be effectively evaluated using an MSCR test [33,34,35], and to evaluate the elastic recovery performance of the three polyurethane-modified asphalt samples under high-temperature conditions, the modern multi-stress repetitive creep test was utilized. The test results of the modified asphalts at two stress levels (i.e., 0.1 kPa and 3.2 kPa) are illustrated below (Figure 11 and Figure 12).

The MSCR load cycle curves demonstrate that the irrecoverable creep flexibility of the three polyurethane-modified asphalt samples was smaller than that of the virgin asphalt, which indicates that polyurethane can ameliorate the elastic recovery of asphalt. With respect to the 0.1 kPa stress level, the three polyurethane-modified asphalt samples exhibited a significant recovery trend compared with the virgin asphalt, which indicated that the unloading phase curve of the modified asphalts was non-horizontal and that the curve pertaining to the virgin asphalt was nearly horizontal. With respect to the 3.2 kPa stress level, when the stress was continually increased, both the virgin asphalt and the modified asphalts exhibited poor elastic recovery. This finding is consistent with Figure 13a.

Figure 13b indicates the irrecoverable creep flexibility values (J_nr) of the asphalts at two different stress levels, namely 0.1 kPa and 0.32 kPa. The irrecoverable creep flexibility values (J_nr) can enable researchers to evaluate the cumulative strain under repeated loading [36,37], and they also effectively correlate with the rutting resistance of asphalt [38,39,40]. When the value of non-recoverable creep flexibility (J_nr) is small, the elastic recovery of the asphalt becomes enhanced. The J_nr values of the polyurethane-modified asphalt samples with different molecular weights were smaller than that of the virgin asphalt, which indicates that the addition of polyurethane improved the rutting resistance of the asphalt. The scale of the rutting resistance followed the order: PU1400 > PU650 > PU1000, which indicates that molecular weight affects the rutting resistance. Due to the difference in stress magnitude and loading between the two tests, this conclusion differs from the DSR test results; furthermore, when the asphalt is subjected to lower stress levels, DSR affects the performance of the modifier [4,34].

Table 5. Asphalt MSCR test evaluation data.

Projects	Base Asphalt (JZ)	PU650-Modified Asphalt	PU1000-Modified Asphalt	PU1400-Modified Asphalt
Rdiff (%)	76.65	88.34	85.81	86.31
Jnr-diff (%)	6.03	54.86	52.90	65.83

indicates that as the stress increased from 0.1 kPa to 0.32 kPa, the R_diff of the three samples of polyurethane-modified asphalt decreased by 88.34%, 85.81%, and 86.31%, respectively. By contrast, the J_nr-diff of the three samples of polyurethane-modified asphalt increased by 54.86%, 52.9%, and 65.83%, which indicates that the stress sensitivity of the three samples followed the order: PU1400 > PU650 > PU1000. Thus, the molecular weight has an impact on the stress sensitivity.

(3)

BBR test results

With respect to the BBR test, the bending creep stiffness modulus (S) and creep rate (m) are two crucial indicators that can enable researchers to evaluate the low-temperature performance of asphalt based on the online viscoelastic range, and Figure 13 depicts the results of this evaluation. Generally, when the value of S is large, the tensile stress becomes greater, and when the value of m is small, the stress relaxation becomes small [41]; therefore, for asphalt to exhibit a satisfactory low-temperature performance, the value of S should be small and the value of m should be large.

Figure 14 indicates that under different temperature conditions, the three samples of polyurethane-modified asphalt exhibit a markedly lower S-value, but an almost constant m-value compared with the virgin asphalt, which was higher, which indicates that the three polyurethanes could enhance the low-temperature performance of asphalt. With respect to the comparison of the three different polyurethanes, we observed that as the molecular weight increased, the S-value gradually decreased and the m-value gradually increased, which also indicates that the reaction between the -OH and -NCO groups, which constitute polyol, can enhance the low-temperature performance of asphalt. This conclusion coheres with the results of the asphalt ductility test.

(4)

DSC test results

The chemical composition and molecular weight of asphalts vary [42,43,44,45]; thus, with respect to performance, asphalts exhibit significant differences. By analyzing the glass transition temperature of asphalt, the DSC test evaluates the low-temperature cracking resistance exhibited by asphalt; when the glass transition temperature is low, the resistance to low-temperature cracking increases [4,46]. The DSC curves of the three different polyurethane-modified asphalt samples and the DSC curve of the virgin asphalt are depicted in Figure 15, where Tg denotes the stile midpoint position, which indicates the first step of the DSC curve.

The glass transition temperature (Tg) denotes the temperature that divides the highly elastic state and the glassy state of asphalt, and the two states correspond to the different mechanical properties of the material. When the temperature is higher than Tg, the asphalt transitions into the highly elastic state, and when the temperature is lower than Tg, the asphalt transitions into the glassy state [47,48].

Figure 15 indicates that the glass transition temperature of the three different polyurethane-modified asphalts was lower than that of the virgin asphalt; thus, all three polyurethanes can reduce the glass transition temperature of asphalt and enhance its low-temperature performance.

In addition, Figure 15 indicates that when the molecular weight increased, the glass transition temperature gradually decreased and the low-temperature performance of the asphalt gradually increased. The analysis indicates that when the molecular weight of the polyol increases, the flexible chain segment that characterizes the polyurethane increases, and the rigid chain segment decreases; furthermore, the flexible chain segment enhances the elasticity and low-temperature performance of polyurethane.

4. Discussion

Polyols have a small molecular weight and good high-temperature elasticity but poor low-temperature performance. Therefore, in order to ensure a normal service life and good road condition level, it is necessary to comprehensively consider the selection of polyol-modified asphalt with an appropriate molecular weight.

All three polyurethane-modified bitumen showed positivity storage stability, however, the microscopic morphology of the three modified asphalt storage stability tests needs to be further explored.

Here, only commonly utilized polyol materials were selected, and there are many other materials that require further analysis.

Our future recommendation is to further verify the road performance of materials through mixture testing, which can then be used to evaluate the appropriate molecular weight range of polyols.

5. Conclusions

Herein, we analyzed the effect of polyols with different molecular weights on the performance of polyurethane-modified asphalt, determined the synthesis process for polyurethane and asphalt modified with it, and analyzed the effect of different polyols on the performance of modified asphalt. Based on the results, the following conclusions can be drawn:

(1)

Polyols of different molecular weights (PTMG650, PTMG1000, and PTMG1400) exert similar modification effects on the penetration, softening point, and ductility of asphalt, and the effect on ductility is the most apparent.

(2)

The polyurethane modification that affects asphalt is not a simple physical blend; polyurethane and asphalt can generate a complex network structure that increases the viscosity of the modified asphalt, and when the molecular weight of the polyol is increased, the viscosity of the modified asphalt increases.

(3)

Polyurethane can enhance the deformation resistance of asphalt, and the difference in the rutting resistance of the three polyurethane-modified asphalt samples, which exhibited different molecular weights, was small.

(4)

During the reaction with MDI, when the molecular weight of the polyol was high, the number of soft segments that were generated increased, and the low-temperature performance of the modified asphalt was enhanced, which coheres with the results of the BBR test and DSC test.

(5)

The polyols selected in this paper are more expensive and have certain difficulties in actual engineering promotion, and in future research, the compound method can be used to prepare low-cost polyurethane for asphalt modification.