A Methodological Approach to the Study of Retroreflective Pavements

Abstract

Climate change, principally driven by human activities, has led to an increase in global temperature, which is predicted to continue rising in the coming years. This temperature increase is even more pronounced in urban areas due to the heat island effect. This phenomenon is highly influenced by the presence of paved streets made with bituminous mixtures, which are characterised by their high solar radiation absorption capacity. Bituminous mixtures retain and re-emit a large amount of heat that intensifies the urban heat island effect. The novelty of this work is to measure retroreflective properties of bituminous mixtures that present a highly textured surface. In this context, the aim of this study is to evaluate the retroreflectance of different bituminous mixtures for use as pavement surfaces, focusing on the influence of colour and different types of aggregates. For this, total and directional reflectance measurements were conducted to determine the retroreflectance of these mixtures, with the purpose of mitigating the heat island effect in urban environments without affecting users through reflected solar radiation. The results show the retroreflective capacity of the designed mixtures within the visible spectrum, especially those manufactured with light-coloured aggregates and synthetic binders pigmented with titanium dioxide. Thus, the retroreflectance of the lighter mixtures range from 37.9% at a 0° entrance angle to 68.9% at 60°, while the black mixtures exhibit values between 5.1% and 8.4%.

1. Introduction

One of the main challenges facing the current society is climate change [1]. This phenomenon is demonstrated by the increase in global temperature, which has a significant impact on the environment and society in general [2,3]. According to the Intergovernmental Panel on Climate Change (IPCC) [4], the Earth’s average surface temperature has increased by approximately 1.2 °C since the late 19th century, primarily driven by human activities such as the burning of fossil fuels and deforestation. Moreover, the IPCC predicts that the global temperature rise is likely to reach or exceed 1.5 °C between 2021 and 2040, with a 50% probability.

The temperature increase is even more pronounced in urban areas due to the phenomenon known as the urban heat island effect [5,6], with temperature differences between urban and rural areas observed to range from 1 to 14 °C [7,8,9]. This effect is intensified by several factors. High heat emissions result from concentrated human activity. Additionally, lower levels of vegetation and a high concentration of materials such as asphalt mixtures, concrete, and bricks contribute to this effect. These materials absorb solar radiation, causing their temperatures to rise [10,11]. They are characterised by a high thermal storage capacity and low solar reflectance. As a consequence, their surface temperature increases during the day and releases the absorbed heat at night, keeping temperatures high. This phenomenon not only increases energy demand for cooling but also raises mortality due to heat stress, exacerbated during heat waves, which will be more intense and frequent as a consequence of climate change [12,13,14]. Tackling the urban heat island effect problem is essential for improving the habitability and sustainability of cities in the context of global climate change.

The urban heat island effect is strongly influenced by streets paved with asphalt mixtures, which comprise approximately 29% to 39% of the urban surface area [15]. These dark pavements have a high capacity for solar radiation absorption due to their low albedo, which means they reflect little sunlight while retaining and re-emitting substantial heat [16]. According to [17], pavements absorb between 80% and 95% of incident solar energy. Thus, in the middle of summer, the surface temperature of bituminous pavements can exceed 70 °C, heating the local air [18,19]. Mitigating the impact of the heat island effect requires the implementation of new technological solutions that maintain lower surface temperatures on streets and pavements, such as cool pavements [20,21].

Within the different cool pavement technologies developed to date, reflective pavements can significantly contribute to reducing the urban heat island effect. These are characterised by high solar reflectance, defined as the fraction of incident solar radiation that is reflected. This property means reflective pavements store less heat, thereby maintaining a lower surface temperature compared to conventional pavements. Moreover, they can improve night-time visibility, reducing the need for artificial lighting.

The literature contains various studies on the development of reflective pavements. Hen et al. [22] conclude that increasing the albedo of the pavement by 0.1 reduces its average temperature by 2.1 °C. Research on reflective pavements has mainly focused on new chip seals and coatings. Shamsaei et al. [23] used the construction and demolition waste of different colours as aggregates to manufacture different chip seals with the aim of reducing the heat island effect. In their study, a spectrometer with sensitivity in ultra-violet, visible, and near-infrared spectral regions was used to determine the reflectance of the chip seals, finding that the red brick chip seal reflected the most solar radiation and released the least heat at night. Hu et al. [24] modified an asphalt binder with TiO₂ quantum dots, increasing the solar reflectance by up to 12%. Like the previous studies, the total reflectance was obtained from a UV-VIS-NIR spectrophotometer. Xie et al. [25] identified reflectance in the visible and near-infrared spectra as dominant factors affecting the thermal behaviour of coatings. These authors studied different coatings to improve the pavement reflectance. Other authors have also proposed solutions through the use of coatings [26,27]. However, these types of solutions can decrease the skid resistance of the surface course and present durability concerns. By these methods, the reflectance could be reduced by up to 50% within a single year [28].

Liu and Wang [29] filled the voids in a porous mixture with magnesium phosphate cement to obtain a reflective grouting compound pavement. They observed a surface temperature reduction of 4.5 °C compared to the reference mixture after 180 min of exposure. The reflectance in the near-infrared spectrum increased by up to 34.19% related to that of the asphalt. In the study of Dai et al. [30], the maximum surface temperature reduction was 20.2 °C. In this case, the authors used, in addition to magnesium phosphate cement, titanium dioxide as an additive to improve the reflectance. No recent studies have been found on the design of bituminous mixtures manufactured with synthetic binders and additives that change the colour of the mixtures to lighter shades, thus improving their reflectance.

The greatest drawback of reflective pavements is the glare caused by reflected radiation, which can be harmful to skin, cause visual effects on the cornea of the human eye, and affect drivers’ perception [31]. This is because reflective pavements reflect solar radiation in all directions. In this way, if the radiation is not reflected to the atmosphere, it can negatively affect the thermal comfort of citizens and affect vehicles and urban buildings. For this reason, reflective pavements will only effectively reduce the heat island effect if they reflect incident radiation to the atmosphere. This highlights the need to modify existing reflective pavement technology towards those capable of reflecting the incident solar radiation back in the same direction, that is, retroreflective pavements [32]. Thus, retroreflection refers to the ability of certain surfaces to reflect incoming light back towards the light source, regardless of the original angle of incidence [33]. In this manner, whereas solar reflectance reflects light in multiple directions and focuses on the amount of light reflected, retroreflectance is characterised by redirecting light directly back to its source, regardless of the angle of incidence.

Different authors have studied the retroreflective capacity of materials for their use in buildings, façades, and pavements, aiming to address the limitations of reflective materials [34,35]. Some findings indicate that for small angles of incidence (10°), the retroreflective materials predominantly reflect the light towards its source [36]. However, there are few studies in the literature focusing on retroreflective materials specifically for urban pavements. Rossi et al. [35] analysed the retroreflective capacity of an urban pavement within a canyon using a spectrophotometer with an integrating sphere, as well as an ad hoc pyranometer that enabled measurement of both the incident light and that reflected in the same direction. The results demonstrated an increase in the albedo of the retroreflective surface by 4.6% compared to white and beige diffusive materials.

Traditionally, a pyranometer or a spectrophotometer with an integrating sphere has been used to determine the optical properties of reflective materials [37]. The former is generally employed in field measurements, while the latter is used in laboratories to conduct measurements on small samples. However, it is not possible to measure the reflective properties directly in specific directions using this equipment to determine a material’s retroreflective capacity. Other authors, instead of determining the most energetic direction of the reflected light, estimated the retroreflectance of the studied materials through thermal distribution [38,39]. However, there are other measurement methods that allow for the precise distribution of the reflected light to be determined, particularly in the incident direction. Various measurement systems have been proposed in the literature, notable among which are the fibre optic system for emission and reception [40], the method for measuring the angular distribution of the reflected light [35], and the method for measuring bidirectional reflectivity using a goniometer [41]. Despite the availability of these methods for determining retroreflection, this remains a field for exploration, especially concerning retroreflective bituminous mixtures for urban pavements.

Considering the above, the aim of this research is to evaluate the retroreflectance of different bituminous mixtures that could be used in pavement surfaces, as well as the influence of colour and different types of aggregate on retroreflective properties. To achieve this, total reflectance measurements will be conducted in multiple directions to determine the feasibility of estimating the retroreflectance of these bituminous mixtures.

2. Materials and Methods

2.1. Materials

2.1.1. Aggregates

Three types of aggregates with different colours were used in this study, namely a dark-grey porphyry aggregate, a light-grey granite aggregate, and a white limestone aggregate (Figure 1).

The standardised emittances for each type of aggregate, measured in the near-infrared (NIR) spectral range according to ASTM E408-13 (2019), are shown in Figure 2 and

Table 1. Average value of the normalised emittance in the infrared range.

Aggregate	Normal Emittance
Limestone	0.64
Granitic	0.37
Porphyry	0.26

. These measurements were conducted using the MPA (Multi-Purpose Analyzer) FT-NIR Bruker brand apparatus (Bruker, Bremen, Germany, E.U.).

2.1.2. Bitumen

The bitumen used for the bituminous mixtures manufacturing was a conventional 50/70 and a pigmentable synthetic bitumen of similar penetration. The synthetic binder was a commercial binder made from a blend of resins, oils, and polymers. It was a colourless thin film and could be coloured with the use of pigments.

Table 2. Properties of the binders: 50/70 and pigmentable synthetic binders provided by the supplier.

Property	Standard	50/70	Synthetic
Penetration at 25 °C (0.1 mm)	EN 1426 (2015)	61	20–50
Softening Point (°C)	EN 1427 (2015)	50.9	≥85
Fraass Breaking Point (°C)	EN 12593 (2015)	−14	≤−20
Flash Point (°C)	EN ISO 2592 (2018)	280	>270
Residue after Ageing
Mass Variation (%)	EN 12607-1 (2015)	0.1	<1.5
Retained Penetration (% o.p.)	EN 1426 (2015)	66	>80
Softening Point Increase (°C)	EN 1427 (2015)	7.6	<10

shows the properties of the binders according to the supplier.

2.1.3. Pigments

Various types of pigments were used to give the desired colour to the binders. No pigment was used with the 50/70 conventional binder to produce a black surface, which served as a reference, while the following pigments (see Figure 3) were used with the pigmentable synthetic bitumen:

• Titanium dioxide (TiO₂) to give a white shade.
• Iron composites of different colours: ochre, brown, and red.
• Combination of silica, sulphur, aluminium, and sodium composites to a give a blue shade.

A qualitative elemental analysis was conducted using wavelength-dispersive X-ray fluorescence methodology (WDXRF). The analysis was conducted with a rapid scan (EZ analysis) across all elemental lines, and SQX calculations were employed to ascertain the elemental composition of each sample.

Table 3. Main components of the pigments obtained from the XRF analysis.

Element (% Mass) (*)	Pigment
	Ochre	Blue	White	Brown	Red
Fe	95–99	-	-	42–47	95–99
Ti	-	-	95–99	-	-
Ca	-	-	-	48–53	-
K	-	residual	-	-	-
S	residual	20–25	-	residual	-
Si	-	31–36	residual	residual	residual
Al	-	19–24	residual	residual	-
Mg	-	-	-	residual	-
Na	-	16–21	-	-	-

(*) It should be noted that the mass balance presented in this table does not account for the light elements that are not detectable by the XRF instrument (Rigaku Corporation, Tokyo, Japan). These include hydrogen (H), lithium (Li), beryllium (Be), boron (B), carbon (C), nitrogen (N), oxygen (O), and fluorine (F).

shows the main components of each pigment.

2.1.4. Asphalt Mixtures

Three types of bituminous mixtures were considered in this study, namely asphalt ultra-thin layers (AUTLs), open-graded (BBTM), and asphalt concrete (AC). For the first part of the research, only AUTL mixtures were used. They were applied in surface courses with thicknesses from 10 to 20 cm and were characterised by their discontinuous grading and open texture [42]. Specifically, AUTL 8 mixtures were manufactured with the considered aggregates, the two types of bitumen, and the selected pigments, with the aim of obtaining a large variety of shades. The different combinations of bituminous mixtures evaluated are shown in Figure 4.

In the last stage of the study, BBTM 8A and AC 16S mixtures were also considered to evaluate the effect of the surface texture on the reflectance. Both types of mixtures were manufactured with aggregates and binders to achieve the extreme colours among those studied, specifically white and black. Figure 5 shows a diagram of the BBTM 8A and AC 16S mixtures analysed.

The grading of the different types of bituminous mixtures used in this study is shown in Figure 6. The series of AUTL specimens were compacted following the guidelines outlined in article 545 of the Spanish Technical Specifications Document [43], with 25 blows per side to obtain Marshall-type cylindrical specimens of approximately 37.5 mm in height. The BBTM specimens were also compacted by impact, with 50 blows per side, while the AC specimens were compacted by applying 75 blows per side to obtain cylindrical samples of approximately 65 mm in height in both cases.

Table 4. Binder content of the mixture used for the manufacturing of different bituminous mixtures considered.

Bitumen	AUTL 8	BBTM 8A	AC 16S
Conventional, 50/70	5.5%	5.5%	4.8%
Synthetic	5.6%	5.6%	4.9%

shows the binder content, as a proportion of the total mixture, used for the manufacturing of the different bituminous mixtures considered in the study. For bituminous mixtures manufactured with synthetic bitumen, the pigment content was 2% of the aggregate. This quantity of pigment was subtracted from the filler content.

2.2. Methods

This study aims to analyse whether it is possible to evaluate the retroreflectance of the considered mixtures. Specifically, it focuses on the reflectance occurring in the same direction as the incident light on the surface. Additionally, the research seeks to determine the possible effect of the components of the mixture.

Two main types of measurements for evaluating a surface reflectance can be distinguished as follows:

• Integrated total reflectance measurement in all directions for a single incident light direction.
• Reflectance measurement in a specific direction for a single incident light direction. Both lighting and measurement directions can vary.

The equipment for measuring the total reflection has the advantage of minimising the effect of the irregularities of surfaces that are not completely smooth. Thereby, this equipment provides a total value of reflectance. In contrast, equipment that measures the reflection in a single direction is more sensitive to the surface texture. Its main advantage lies in the ability to vary both the direction of the incident light and the direction of the reflectance measurement. Thus, it allows the reflectance of a sample to be evaluated according to the direction.

2.2.1. Spectrophotometer with Integrating Sphere to Evaluate Total Reflectance

To evaluate the total reflectance, an AvaSpec-3648 spectrophotometer (Avantes, Oberursel, Germany, E.U.) and the AvaSphere-30 (Avantes, Germany, E.U.) integrating sphere were used, the spherical diameter of which was 30 mm and the sample port diameter was 6 mm (Figure 7). The measurements were carried out using an Avantes AvaLight-DH-S (Avantes, Germany, E.U.) light source that consisted of a deuterium light source (190–400 nm) and a halogen light source (360–2500 nm).

To ensure accurate measurements, the procedure was conducted in a dark room with only minimal light from the computer screen and any light that managed to penetrate through the window covers. Before starting the measurement, the light source needed to be turned on and allowed to warm up for approximately 20 min.

After the initialization of the equipment, the first step involved taking a blank measurement. This measurement utilised a reference white material, BaSO4, to obtain a low baseline that allows for the comprehensive evaluation of a solid diffusion spectrum. A small amount of BaSO4 was placed under a glass microscope slide to prevent any powder contamination in the integrating sphere. The blank reference measurement was then taken. Subsequently, a dark reference measurement was taken without removing the BaSO₄; from under the sphere. This was achieved by utilising the TTL shutter (Avantes, Germany, E.U.), which closes the beam from the source to the sphere. The spectrophotomer then took a measurement with any residual light present in the sphere, which would have entered through the sphere’s port.

Once the reference measurements were completed, the bituminous samples were analysed. For each sample, up to five measurements were taken from different positions on its surface, and the wavelength range of the measurements was from 190 to 1100 nm. Due to the non-homogeneous nature of the sample’s surface, a geometric pattern resembling a cross was followed for the measurements. The four points at the end vertices of the cross and the central point of the cross were measured. To maintain a stable horizontal position above the sample, a glass microscope slide was placed between the sample and the integrating sphere (see Figure 8).

The different reflectance spectra obtained from these measurements were calculated following the E903-20 (2020) ASTM Standard [44].

$$\rho (\lambda )={\displaystyle \frac{(S_{\lambda }-Z_{\lambda })}{({100}_{\lambda }-Z_{\lambda })}}$$

(1)

where S_λ is the recorded specimen reading, Z_λ is the zero line reading (dark reference), and 100_λ is the 100% correction obtained with the specimen port replaced by a sample having a coating and a curvature identical to the sphere wall (BaSO4), all at wavelength λ. Based on this spectrum, the average for the five measured points was calculated.

2.2.2. Spectroradiometer to Evaluate the Reflectance in a Particular Direction

A PR-715 SpectraScan spectroradiometer (Photo Research^®, New York, NY, USA) was used to measure the reflectance of the sample surface in a specific direction (Figure 9). In this study, the measurement range of this equipment was limited to the visible spectrum, between 400 and 700 nm.

The reflectance measurements were conducted in a VeriVide CAC 120 H4 colour assessment cabinet (VeriVide^® Ltd., Leicester, UK) with grey interior diffuse walls to minimise unwanted specular reflections. This observation booth was equipped with a D65 daylight simulator, provided by an F40/T12 fluorescent lamp for illumination. The lamp spectral power distribution was measured, determining its correlated colour temperature to be 6438 K (10° observer). To eliminate other potential sources of light influencing the results, the room was darkened during the reflectance measurements, ensuring that the booth lamp was the sole source of illumination. For the measurements, a calibrated PTFE white standard (Photo Research Reflectance Standard model SRS-3) was used as the reference. As in the previous case (Section 2.2.1), several measurements were taken across the surface of the specimens, following the diagram of positions 1 to 5 shown in Figure 8. The sample was moved and rotated so that the area to be measured remained in the same relative position with respect to the equipment.

2.2.3. Work Stages

In the first stage of the work, reflectance was evaluated using two procedures. A spectrophotometer with an integrating sphere was employed for the evaluation of total reflectance. Additionally, a field spectroradiometer was used to assess directional reflectance (see Figure 10). Although the spectrophotometer with an integrating sphere measures reflectance across the entire spectrum (ultraviolet, visible, and infrared), the spectroradiometer works only within the visible spectrum. Therefore, the comparison of results was made exclusively in the visible range (wavelengths between 400 and 700 nm).

In the second stage, the aim was to approximate the retroreflectance measurement. For this, the reflectance was evaluated using the spectroradiometer when both the emitter and receiver were approximately in the same position. This is when the observation angle, θ, was very small. The observation angle typically used between the direction of the entrance light and the measurement direction of the reflectance is 45°, with measurements taken perpendicularly to the sample surface. However, for the measurement of the retroreflectance, an observation angle close to 0° was considered while maintaining the same entrance angle of incident light on the surface (Figure 11). Nevertheless, all reflectance measurements on the bituminous mixtures could be affected by the texture of the sample itself.

Therefore, a third and final stage of the work was conducted. The purpose was to evaluate the effect of the entrance angle of light on the retroreflectance. In this stage, the emitter and the receiver were placed very close, so that the observation angle, θ, between the direction of the incident light and that of the reflectance measurement was nearly zero. The tilt of the sample was varied relative to the illumination axis, such that three different entrance angles, β, of 60°, 45°, and 0° were considered (see Figure 12). These measurements were carried out on the AUTL-PB and AUTL-LW mixtures corresponding to the extreme colours (black and white). To further highlight the effect of the texture of the mixtures, two additional types of mixtures were included, a BBTM type and an AC type (the latter being more compact) with the extreme colours (black and white). These mixtures are BBTM-PB, BBTM-LW, AC-PB, and AC-LW.

3. Results and Discussions

3.1. Stage 1: Total and Directional Reflectance

Figure 13 shows the average reflectance in the visible range of the different bituminous mixtures considered in the first stage of the work (see Figure 4). The reflectance was determined using a spectrophotometer with an integrating sphere to evaluate the integrated total reflectance in all directions and a spectroradiometer to assess the reflectance in a specific direction (Figure 10 configuration). The reflectance values in this range are presented in

Table 5. Reflectance values in the visible range obtained in the first work stage.

Bituminous Mixture	Global Reflectance (%)		Reflectance at 45° (%)		Reflectance Ratio 45°/Global
	Mean	Std. Dev.	Mean	Std. Dev.
AUTL-LW	79.7	4.6	67.8	2.5	0.9
AUTL-GW	63.2	6.6	58.8	2.4	0.9
AUTL-PW	65.4	14.3	40.3	8.3	0.6
AUTL-LO	38.7	1.7	23.6	1.1	0.6
AUTL-GO	31.9	3.8	19.5	1.4	0.6
AUTL-PO	35.0	5.2	19.0	1.0	0.5
AUTL-LBR	32.8	7.2	19.2	1.6	0.6
AUTL-GBR	26.0	3.9	15.3	1.8	0.6
AUTL-PBR	26.4	3.3	14.5	1.1	0.6
AUTL-LR	32.1	2.4	15.2	1.1	0.5
AUTL-GR	26.1	3.3	13.3	1.0	0.5
AUTL-PR	22.5	2.3	12.0	1.0	0.5
AUTL-LBL	23.2	1.9	10.6	0.6	0.5
AUTL-GBL	25.5	3.8	10.3	1.6	0.4
AUTL-PBL	23.3	1.8	9.7	2.5	0.4
AUTL-LB	20.0	1.0	5.1	0.8	0.3
AUTL-GB	17.9	1.0	5.1	0.6	0.3
AUTL-PB	20.9	3.8	4.9	1.1	0.2

.

When comparing the reflectance results obtained by the two pieces of equipment, it is observed that the trend throughout the evaluated wavelength range is the same for the two procedures. However, the values obtained with the integrating sphere are consistently higher than those from the spectroradiometer in all the cases considered. For example, for the mixture manufactured with synthetic bitumen and titanium dioxide with the limestone aggregate, the reflectance in the visual spectrum range is 79.7% using the integrating sphere and 67.8% using the spectroradiometer. For the granite aggregate, these values are 63.2% and 58.8%, respectively, while with the porphyry aggregate, the reflectance values are 65.4% and 40.3%, respectively. This difference is expected, as the integrating sphere measures the total reflectance in all directions, whereas the spectroradiometer does so in one direction only.

Furthermore, when comparing the differences between the two pieces of equipment based on the ratio between the reflectance measured by each, it is observed that the darker mixture yields lower ratios. Consequently, the differences between the measurements are greater. These ratios are relatively constant for each series of mixtures with the different aggregates, except the lighter mixtures, where the effect of the type of aggregate is more pronounced.

Another point to highlight when observing the variation in reflectance in the studied range (Figure 13) is how for any of the aggregates used, the reflectance increases with the wavelength as the infrared range is approached when the shade is white or warm. In contrast, for the blue shade (cold), the reflectance is higher at shorter wavelengths, closer to the ultraviolet range. For black, the reflectance remains practically constant throughout the visible range. This behaviour is due to the composition and light absorption and reflection characteristics of white or warm colours, which reflect more light in the spectrum near the infrared and are more efficient at reflecting light of a longer wavelength. Conversely, the blue light has a shorter wavelength, causing blue materials to reflect more light in the spectrum close to ultraviolet, as they absorb less light in the short wavelength range and reflect it more effectively than at longer wavelengths. Black materials, on the other hand, tend to absorb nearly all the visible light and reflect very little. The surfaces of this colour convert most of the light they receive into heat, explaining their low reflectance.

Finally, the effect of the type of aggregate can be seen. With the limestone aggregate (lighter), the spectral reflectance values are higher than those of the other two aggregates, with more pronounced differences for the lighter shades of the mixture (white, ochre, or brown). For darker shades (red, blue, or black), the differences between mixtures with different aggregates are minimal. For the granite and porphyry aggregates, the total reflectance values are very similar, whatever the shade of the mixture. For example, when measurements are carried out with the integrating sphere, the reflectance of the limestone aggregate mixture with a lighter shade (AUTL-LW) is 79.7%, while for the darker-shaded mixture (AUTL-LB), it is 20%. When using the spectroradiometer, the reflectance values were 67.8% and 5.1%, respectively. For the granite aggregate, the reflectance of the lighter-shaded mixture (AUTL-GW) measured with the integrating sphere was 63.2%, while that of the darker mixture (AUTL-GB) was 17.9%. The same trend is observed with the spectroradiometer, where the reflectance values for the lighter and darker-shaded mixtures are 58% and 5.1%, respectively. Similar differences are also observed for the mixture manufactured with porphyry aggregate, where the reflectance measured with the integrating sphere for extreme shade mixtures is 65.4% (AUTL-PW) and 20.9% (AUTL-PB), and with the spectroradiometer, it is 40.3% and 4.9%, respectively.

3.2. Stage 2: Approximation to the Retroreflectance Measurement

Figure 14 shows the variation in reflectance when varying the entrance angle between the direction of the incident light source and the measurement direction of the reflectance. These angles are 45° (typically used) and 0° (as shown in Figure 11). The reflectance values are given in

Table 6. Reflectance values in the visible spectrum for different measurement angles.

Bituminous Mixture		Reflectance (%)
	45°		0°
	Mean	Std. Dev.	Mean	Std. Dev.
AUTL-LW	52.4	2.3	78.0	9.7
AUTL-GW	49.1	2.6	82.7	11.6
AUTL-PW	59.5	10.2	78.2	10.9
AUTL-LO	20.5	1.5	32.4	5.8
AUTL-GO	16.0	2.5	26.4	4.3
AUTL-PO	23.3	2.4	28.7	9.1
AUTL-LBR	25.2	2.2	35.9	2.1
AUTL-GBR	21.9	2.3	36.1	1.6
AUTL-PBR	27.9	2.0	36.5	4.2
AUTL-LR	12.4	1.6	19.1	1.9
AUTL-GR	12.6	2.0	19.4	2.4
AUTL-PR	15.5	2.3	21.9	3.1
AUTL-LBL	23.4	1.6	38.4	2.1
AUTL-GBL	27.0	2.4	38.9	2.1
AUTL-PBL	26.4	1.3	36.9	8.5
AUTL-LB	8.9	0.6	14.3	2.7
AUTL-GB	9.5	1.4	11.0	3.1
AUTL-PB	7.6	2.7	10.4	2.2

. Firstly, it is observed that for both types of aggregate, reducing the entrance angle leads to an increase in the retroreflectance. This increase is nearly constant throughout the visible range for both white and black mixtures. However, for the mixtures with warm colours, the increase is more pronounced at wavelengths near the infrared range, while for the colder-shaded mixtures, it is more pronounced at wavelengths nearer the ultraviolet range, regardless of the type of aggregate used to manufacture the mixtures. This demonstrates the retroreflective capacity of the designed bituminous mixtures, suggesting that the adverse effects on the health of users could be reduced if the radiation reflected towards them is minimised.

The samples used in this stage, where the entrance angle is 45°, differ from those used for the same angle in the previous stage. In this case, the samples have been exposed to solar radiation for a long period of time. This has caused the lighter samples to be slightly darker than the original ones, while the darker ones appear slightly lighter. Therefore, the reflectance values are only directly comparable within each stage of the study. Nevertheless, as seen in the previous stage, the aggregate type influences only the lightest mixtures (white and ochre), among which the limestone aggregate exhibited slightly higher values in general.

It must be remembered that all the measurements with the spectroradiometer were conducted with the same light entrance angle with the surface (45°). Therefore, it is proposed to analyse the effect that the variation in this angle would have on the retroreflectance values. A priori, it could be thought that the texture of the mixture will also influence the measurements obtained; therefore, in addition to the AUTL-type mixtures, two further types of mixtures have been considered, a BBTM 8A and an AC 16S, both with different textures.

3.3. Stage 3: Effect of the Light Entrance Angle on the Retroreflectance

Figure 15 illustrates the variation in retroreflectance for the light source entrance angles of 60°, 45°, and 0° (as shown in Figure 12). The measurements were conducted on the AUTL mixture samples with the extreme colours, white and black, which were manufactured with limestone and porphyry aggregate, respectively. The retroreflectance values for the considered entrance angles are shown in

Table 7. Reflectance values in the visible spectrum for different measurement angles.

Bituminous Mixture	Retroreflectance (%)
	60°		45°		0°
	Mean	Std. Dev.	Mean	Std. Dev.	Mean	Std. Dev.
AUTL-LW	68.9	4.5	64.7	6.8	48.5	6.3
AUTL-PB	8.4	1.6	7.7	1.8	6.7	0.8
BBTM-LW	56.0	2.5	54.1	3.6	45.9	5.7
BBTM-PB	6.8	2.4	6.5	1.3	5.2	2.1
AC-LW	55.8	1.9	52.2	3.6	37.9	2.7
AC-PB	6.2	2.4	5.8	1.5	5.1	1.2

.

The results indicate that the retroreflectance values increase in all cases as the entrance angle is raised. This may be attributed to the hollows present on the sample surface, as none of the mixtures considered have a smooth surface. When the incident light is zenithal, the hollows appear darker and do not reflect the light. As the entrance angle increases, along with the observation angle, the darker hollows are perceived as being smaller, thereby reducing the extent of the darkest surface area. This effect could disappear when the light is incident at a very small grazing angle, resulting in a larger reflection area, as shown in Figure 16 and Figure 17.

The shade of the mixture has a significant influence. Thus, the retroreflectance of the white AUTL mixture varies from 48.5% at an entrance angle of 0° to 68.9% at 60°. In contrast, the retroreflectance for the black mixture ranges from 6.7% to 8.4%, with minimal differences due to the low values.

Finally, to analyse the effect of texture on these measurements, two additional mixture types were considered, a BBTM (thin mixture) and an AC16S (bituminous concrete-type mixture). Both types were made using the extreme colours achieved with the porphyry aggregates and 50/70 bitumen (black), as well as the limestone aggregate and synthetic bitumen with titanium oxide (white) (see Figure 5).

Normally, the void content of these mixtures is around 16% for the AUTL, 8% for the BBTM, and 4% for the AC mixtures. However, despite these mixtures usually having quite different textures, it is noteworthy that the white mixtures made with limestone aggregate exhibited a lower-than-usual void content. This is probably due to the low hardness of the limestone, which resulted in higher densification during the compaction process. Consequently, the textures of the different white mixtures are similar (see Figure 18). Nevertheless, the effect of the texture could only be appreciated in the white mixtures, since the reflectance values for the black mixtures are very similar for all the mixture types.

In any case, the AUTL mixtures have similar retroreflectance values (slightly higher) to those of the BBTM mixtures. Both types demonstrate values greater than those of the AC mixtures for all the considered entrance angles. For the highest grazing light incident angle (60°), the retroreflectance values obtained were 68.9%, 56.0%, and 55.8% for the AUTL, BBTM, and AC mixtures, respectively.

4. Conclusions

In this study, the retroreflectance of different bituminous mixtures for use in the top layer of pavements has been examined, considering the influence of colour, aggregate type, and surface texture. The investigation was carried out by measuring total reflection and directional reflectance. The main conclusions are detailed below as follows:

• When determining total reflectance using a spectrophotometer with an integrating sphere and directional reflectance at 45° with a spectroradiometer, similar trends were observed. The reflectance values measured by the spectroradiometer were lower than those obtained through the integrated reflectance across all directions. The ratio between these values for white mixtures is approximately 90%, while for black mixtures, it is around 30%. Nonetheless, both methods proved effective for analysing the reflectance.
• When the shade of the mixtures is white or warm, the reflectance increases with wavelength as it approaches the infrared range. However, for mixtures with a cooler shade (blue), the reflectance is higher at shorter wavelengths near the ultraviolet range.
• The effect of the aggregate depends on the shade of the mixture. For dark mixtures, the differences between aggregates are minimal (around 3% when measuring global reflectance and 0.2% for directional reflectance). For light-coloured mixtures, the reflectance values are higher when using a limestone aggregate (white). For the same shade, no differences in the reflectance were observed between granite and porphyry aggregates.
• By reducing the entrance angle between the incident light direction and the measurement direction, the spectroradiometer provides an approximation of the retroreflectance, which yield higher values than those obtained with the usual 45° angle.
• When measuring the retroreflectance and increasing the entrance angle of the light on the surface to a more grazing angle, the values increase, especially for the lighter mixtures (for AUTL mixtures, values increase by approximately 42%).
• The effect of the mixture type has been demonstrated, with higher retroreflectance values obtained for the AUTL mixtures manufactured with limestone aggregate and synthetic binders pigmented with titanium oxide (white). However, for light-shaded mixtures, the texture differences between the different mixtures were very small.

Consequently, light-coloured mixtures with high texture levels may be suitable for the construction and maintenance of urban roadways, particularly in areas susceptible to the urban heat island effect.