Asphalt

The data in this study were collected by staff working for one LEA, and the procedures have been detailed in the literature [23 (link),27 (link)]. While tests of other physical capacities (e.g., flexibility, linear speed, strength, etc.) would have been beneficial to include, this was not possible given the confines of time, equipment, and logistical restrictions provided by the LEA. Nonetheless, the fitness capacities that were assessed within this study are typical of law enforcement recruits in the literature [6 (link),20 ,23 (link),26 (link),27 (link),28 (link)]. The staff (~20 per testing session) were all trained by a certified TSAC-F who verified the proficiency of the staff members before each session, and all staff followed strict instructions (which will be detailed) to conduct each test. Each recruit’s age, height, and body mass were recorded at the start of academy training. Height was measured barefoot using a portable stadiometer (seca, Hamburg, Germany), while body mass was recorded by electronic digital scales (Health o Meter, Neosho, Missouri). As detailed by Lockie et al. [27 (link)], all tests were conducted outdoors on concrete or asphalt surfaces at the LEA’s training facility on a day scheduled by the staff for the LEA. Testing typically occurred between the hours of 09:00–14:00 depending on recruit availability, and recruits generally did not eat in the 2–3 h prior to their testing session as they were completing non-strenuous activity and employee-specific documentation for the LEA. The weather conditions for testing were typical of the climate of southern California during a calendar year. Although conducting testing outdoors is not ideal, there was no indoor testing facility available for this LEA and these procedures were typical of staff from the LEA [6 (link),27 (link)]. Recruits rotated through the assessments in small groups of 3–4 and were permitted to consume water as required during the testing session.

Basalt fibers with different lengths of 3, 6 and 9 mm were used in this study. According to the previous literature [9 (link),10 (link),15 (link),25 (link)], the percentage of basalt fiber to asphalt binder should not exceed 5%. Therefore, in order to investigate the influence of basalt fiber at the level of asphalt binder, their proportions added into SBS-modified asphalt were 0%, 1%, 2%, 2.5%, 3%, 3.5% and 4% by mass of SBS-modified asphalt, respectively. Based on previous research [10 (link),14 ,26 ], the detailed preparation procedures of SBS-modified asphalt binder reinforced with basalt fiber are as follows: (i) basalt fiber with different lengths and SBS-modified asphalt were heated in an oven at 170 °C until the constant weight and stable state; (ii) three kinds of basalt fiber with different lengths were added into SBS-modified asphalt at seven different proportions, and twenty-one experimental groups could be obtained; (iii) in order to ensure that basalt fibers can be distributed uniformly in asphalt binder, the mixture of basalt fiber and asphalt was placed in a shear homogenizer (KRH-I, Shanghai Konmix Mechanical & Electrical Equipment Technology Co. Ltd., Shanghai, China) after a preliminary manual blending. Then, shearing temperature and speed were set as 170 °C and 6000 rev/min, respectively. After mixing of asphalt with basalt fibers for one hour, SBS-modified asphalt containing basalt fiber was prepared.
Asphalt mixture specimens were produced to investigate the design optimization of SBS-modified asphalt mixture reinforced with basalt fiber. Figure 2 illustrates the gradation of SMA with a nominal maximum size of 13.2 mm. According to the Chinese specification JTG E20-2011 [27 ], the Marshall specimens of asphalt mixture with height of 63.5 mm and diameter of 101 mm were made by Marshall procedures, which were used for laboratory tests and optimization analysis based on RSM. The detailed preparation procedures of SBS-modified asphalt mixture reinforced with basalt fiber are as follows: (i) the aggregates and fillers were weighted and placed in an oven at 180 °C for two hours and SBS-modified asphalt was heated to 170 °C; (ii) the weighted aggregates and basalt fiber were blended together in a mixing pot and then asphalt was poured and mixed at 165 °C until the aggregates were coated; (iii) the weighted fillers were added and mixed well at 165 °C; (iv) asphalt mixtures were compacted with 50 blows of Marshall hammer per side for the target of 4% air void content.

Study sample. The initial sample consisted of 2,067 participants, 36–82 years of age, who were evaluated at baseline (2003–2006) within a population-based cohort of the REGICOR study (Grau et al. 2007 (link)), and who had answered a questionnaire on nighttime noise exposure at the bedroom at follow-up (2009–2011). Briefly, the baseline sample was a random selection of noninstitutionalized inhabitants of Girona who were called in a randomized order for the follow-up visit. Because the noise questionnaire referred to the residence at follow-up, we selected nonmovers from baseline to follow-up (93.3% of the follow-up sample) to ensure that responses referred to the same baseline residences.
The study was approved by Parc de Salut Mar ethics committee, and participants signed written informed consent.
Outcomes and health assessment. Participants were examined from 0800 to 1100 hours at the primary care center and after fasting for 10 hr but being allowed regular medication. Trained nurses measured BP and heart rate following the Joint National Committee VII recommendations (Chobanian et al. 2003 (link)), in sitting position, and with a calibrated automatic device (OMRON 711; Omron Healthcare, Lake Forest, IL, USA). Two measurements were done after at least 10 and 3 min of rest, respectively. If measurements differed by ≥ 5 mmHg, a third one was taken. To minimize the “white coat” effect, we used the last measurement. The nurses also measured weight and height and drew blood. The samples were coded, shipped to a central laboratory, and frozen at –80°C until the assay. Serum glucose, total cholesterol, and triglycerides were determined by enzymatic methods (Roche Diagnostics, Basel, Switzerland) in a Cobas Mira Plus autoanalyzer (Roche Diagnostics). Whenever triglycerides were < 300 mg/dL, LDL (low-density lipoprotein) cholesterol was calculated by the Friedewald equation. Quality control was performed with the External Quality Assessment–WHO Lipid Program [World Health Organization (WHO), Prague, Czech Republic] and Monitrol–Quality Control Program (Baxter Diagnostics, Dudingen, Switzerland).
We defined hypertension as having systolic (SBP) or diastolic (DBP) BP levels ≥ 140/90 mmHg, respectively (Chobanian et al. 2003 (link)), or reporting antihypertensive treatment with a positive response to the question “Do you take or have you taken any doctor prescribed medication to reduce blood pressure in the last two weeks?” For BP analyses, we defined a variable accounting for any “BP-lowering medication,” which included the self-reported antihypertensive treatment defined above or the use of “antihypertensives” or “beta-blockers” as coded by a physician from the medication list provided by participants, namely diuretics, ACE (angiotensin-converting enzyme) inhibitors, alpha or beta-blockers, angiotensin receptor II blockers, and calcium channel blockers. This variable was coded by a physician from the medication list provided by participants.
Exposure assessment. We derived individual long-term average levels of nighttime traffic noise (L_night, 2300 to 0700 hours) expressed in A-weighted decibels [dB(A)] at the geocoded residential addresses (hereafter called outdoor traffic L_night). Geocodes were separated 2 m from the postal address’s façade and located at the floor’s height of each dwelling. We derived the estimates with a detailed and validated city-specific traffic noise model (year 2005), described elsewhere (Foraster et al. 2011 (link)). This model complies with the European Noise Directive 2002/49/EC (END) (European Parliament and Council of the European Union 2002 ) and uses the interim European method NMPB routes-96 [CERTU (Centre d’Études sur les Réseaux, les Transports, l’Urbanisme et les Constructions Publiques) et al. 1997 ]. Estimates were computed at each receptor point by numerical calculations using CadnaA software (DataKustik, Greifenberg, Germany). The main input variables were speed limit, street slopes, type of asphalt, urban topography, and traffic density, also for small streets based on the Good Practice Guidelines for noise mapping (European Commission Working Group Assessment of Exposure to Noise 2003 ). Because railway noise may also be associated with BP (Dratva et al. 2012 (link)), and a single railway crosses dense traffic areas from North to South, we also derived individual residential railway noise estimates (L_night) from an END-based model according to the International Organization for Standardization (ISO; Geneva, Switzerland) standard 9613. The propagation model was built on source identification of railway noise with daytime and nighttime measurements of the noise frequencies (1/3-octave bands) and equivalent levels [in dB(A)] of freight and normal trains (a total of 72 measurements). Measurements were taken with an SC-30 sound level meter and a CB-5 calibrator (CESVA, Barcelona, Spain). Our study sample was not exposed to aircraft noise.
In a face-to-face interview we collected information on noise sensitivity (Weinstein 1978 (link))—a 10-item score based on a nonverbal 6-point scale—and traffic noise annoyance (Fields et al. 2001 (link))—nonverbal 11-point scales—in the bedroom during sleeping hours, as previously done (Babisch et al. 2012 (link)). We also evaluated a) type of glazing and type of window (single, double, laminated, or triple glazing; or double window), b) bedroom orientation (facing the postal address street/side street/backyard), and c) frequency of closing windows during sleeping hours (always/often/seldom/never). Availability of shutters and use of ear plugs was rarely reported and not used in this study.
We combined outdoor traffic L_night with the questionnaire data to calculate two estimates of “personal” noise exposure:
Outdoor traffic L_night at bedroom façade (step a). On the basis of refined modeling techniques for shielded areas (Salomons et al. 2009 (link)), we subtracted 20 dB(A) from the outdoor noise estimates at the postal address to obtain noise levels at the bedroom façade where participants slept. We left outdoor estimates unchanged for bedrooms facing the postal address street or a side street. Noise levels at the side street façade were difficult to quantify, and we assumed they were similar to those at the postal address street.
Indoor traffic L_night at the bedroom (step b). We corrected the outdoor traffic L_night levels at the bedroom façade (step a, above) by subtracting an insulation factor that we calculated according to the reported window types and the frequency of keeping windows closed at night. This is described in the Good Practice Guide on Noise Exposure and Potential Health Effects (European Environment Agency 2010 ). Levels of window insulation are commonly derived from laboratory acoustical measurements, and standard values are described in the Spanish Building Code and complementary technical information (Spanish Government 2010 ; Tremco Ltd. 2004 ). The insulation factors when “Always closing windows” (100% time) were –30 dB(A) for single and double glazing and –40 dB(A) for sound-proofed windows (triple or laminated glazing or double windows). If windows were “often” (75% of the time), “seldom” (25%), and “never” closed, the resulting insulation factors were –21 dB(A), –16 dB(A), and –15 dB(A), respectively, with no further contribution of the specific insulation of each window type.
We followed step b to obtain indoor railway L_night from outdoor estimates.
We also derived individual outdoor levels of annual average nitrogen dioxide (NO₂) concentrations (micrograms per cubic meter) at each geocoded address with a land use regression model (LUR) derived in 2010 for Girona, as described elsewhere (Rivera et al. 2013 (link)). Briefly, the LUR was based on a dense network of residential outdoor NO₂ measurements (years 2007–2009). The main predictor variables were the height above street and traffic-related variables within different buffers (from 25- to 1,000-m radii) around the sampling locations. The coefficient of determination (R²) of the model was 0.63.
Other data collection. Based on questionnaires we also assessed smoking (smoker/ex-smoker of > 1 year/never smoker), weekly leisure time physical activity (in metabolic equivalents) with Minnesota’s questionnaire (Elosua et al. 2000 (link)), daily alcohol intake (grams per day), adherence score to Mediterranean diet (lowest to highest, from 10 to 30) (Schröder et al. 2004 (link)), family history of cardiovascular disease (yes/no), living alone (yes/no), and hearing loss (no/mild/severe). We assessed socioeconomic status at the individual level with educational level (university/secondary/primary/illiterate) and occupation (employed/homemaker-inactive/retired/unemployed), and at the census tract of residences with the deprivation index (Domínguez-Berjón and Borrell 2005 (link)). We defined diabetes as fasting blood glucose levels ≥ 126 mg/dL or reported treatment with antidiabetic drugs; body mass index (BMI) as weight/height squared (kilograms per meter squared); intake of anxiolytics as having ever taken tranquilizers, sedatives, anxiety pills, sleeping pills, or muscle relaxants in the last two weeks (yes/no); and CVD as having ever had a cardiovascular event (myocardial infarction or stroke) or cardiovascular-related surgery intervention (yes/no).
We derived daily means of NO₂ (micrograms per cubic meter) and temperature (degrees Celsius) 0–3 days before the day of examination (lags 0–3) at an urban background station from the regional air quality and meteorology monitoring networks to control for the short-term effects of temperature and air pollution on BP (Servei de Vigilància i Control de l’aire 2008 ; Servei Meteorològic de Catalunya 2011 ). Season was categorized as winter (January–March), spring (April–June), summer (July–September), and autumn (October–December).
Statistical analysis. We performed descriptive analyses of all variables, assessed their linearity against the outcomes with generalized additive models, and transformed them accordingly. We excluded missing observations on the outcomes, exposure, and covariates of the main models (n = 141, 6.8%), resulting in 1,926 cases with characteristics similar to those of the original sample. The inclusion of confounders in the multivariate logistic regression (for hypertension) and linear regression models (for BP) was based on the hypothesized causal pathway of traffic noise and air pollution on hypertension (Fuks et al. 2011 (link)) and previous literature. All single and multi-exposure models were controlled for age, age squared, sex, educational level, physical activity, diet, alcohol consumption, smoking, diabetes, BMI, deprivation, railway noise, and short-term effects of daily temperature (lag 0) on measured BP. Occupational status, living alone, temperature at lags 1–3, instead of lag 0, and daily NO₂ (lags 0–3) did not contribute further to models (i.e., effect estimates changed < 10%). We additionally adjusted for BP-lowering treatment in models for BP and checked regression diagnostics. Effect estimates changed < 10% by further inclusion of potential intermediates (traffic noise annoyance, family history of cardiovascular death, heart rate, and CVD), so these were not considered (data not shown).
We also assessed linear threshold models assuming noise effects to start at 30 dB(A) indoors, the recommended indoor noise levels at night (WHO 2009 ). For this, we created a new variable by subtracting 30 dB(A) to the noise levels and giving the value zero to the resulting negative values. This new variable was then used as the exposure variable in the models.
We tested population characteristics that could modify the association between traffic noise (indoors) and hypertension by including an interaction term (i.e., evaluated categorical or continuous variable × indoor traffic noise) in multivariate models and checking its statistical significance (i.e., p-value of interaction term) as well as the stratum-specific effect estimate of the studied association. The evaluated ordinal variables were coded with consecutive numbers, multiplied by indoor traffic noise, and the resulting continuous variable was used in the models to test for trends. We evaluated age, sex, educational level, BMI, diabetes, traffic annoyance, noise sensitivity with a cut-off at the median, hearing loss, and intake of anxiolytic medication. Anxiolytics have been linked to transportation noise exposure (Floud et al. 2011 (link)), and their mechanism of action may directly affect the suggested stress pathway by which noise affects CVD.
Because of the rather high correlation between outdoor traffic noise and NO₂, we evaluated collinearity in two-exposure models with the variance inflation factor (VIF). A simulation study to assess the effects of collinearity on effect estimates was implemented by repeatedly (10,000 times) simulating data sets and fitting our final model. All final model predictors were simulated from a multivariate normal distribution with mean and covariance matrices as observed in the original data set; SBP was simulated using the regression equation obtained in our study plus normally distributed random error with mean zero and variance equal to the estimated residual variance in the original data set. The correlation between estimated coefficients for outdoor (or indoor) traffic L_night and NO₂ were calculated. We carried out the same procedure with indoor traffic L_night.
We reported estimated changes in the outcomes per 5 dB(A) for all noise indicators and per 10 μg/m³ for NO₂, unless otherwise specified. We defined statistical significance at an alpha level of 0.05.
Analyses were performed with Stata 12.0 (StataCorp, College Station, TX, USA) and R version 2.12 (http://www.r-project.org/).

In this study, six types of asphalt binder, as shown in Table 1, were selected to produce FAM mixtures. Binder A was #70 virgin asphalt; binders B to E were all virgin asphalt binder modified with rock asphalt (QC, UM, and Buton); the concentration of rock asphalt was selected based on the previous studies [2 (link)]. Binder F was a kind of high-modulus natural binder. The technical indices of the above asphalt binder tested according to JTG E20-2011 are listed in Table 2 [27 ].

In this study, two types of asphalt binders were used. The first is a conventional asphalt binder of type CA-24, which was used in the HMA asphalt mixtures. The second is a polymer-modified asphalt binder of type CA-60/80, which was used in the SMA asphalt mixtures. Both types of asphalt binders were classified according to the Chilean Standard [47 ]. The properties of these asphalt binders are shown in Table 4 and Table 5.

The asphalt binder utilized in this investigation was a specialized formulation incorporating LDPE as a modifying agent. LDPE, derived from recycled post-consumer plastic waste [23 (link)], was selected for its unique properties that have shown promise in enhancing the performance of asphalt binders. The LDPE modification process involved carefully blending the recycled polymer with the base asphalt binder (provided by Korean Asphalt company, Seoul, Republic of Korea) under controlled conditions. The selection of LDPE was aimed at capitalizing on its molecular structure, which imparts desirable attributes such as improved elasticity, flexibility, and resistance to temperature-induced aging. Figure 1 provides an overview of the research.
The addition of LDPE to the asphalt binder was expected to improve rheological properties, positively impacting the overall durability and fatigue resistance of the asphalt mixture. Furthermore, LDPE’s role in potentially mitigating the impact of wetting–drying cycles on asphalt performance was a key aspect of this study. The modified binder was subjected to thorough laboratory testing, including rheological analyses and microstructural evaluations, to assess the efficacy of LDPE in enhancing the asphalt binder’s properties and its subsequent impact on the overall performance of the asphalt mixture. The choice of LDPE-modified asphalt aligns with sustainable practices by repurposing plastic waste and has the potential to offer a resilient and eco-friendly solution for asphalt pavement applications.
In the utilization of recycled LDPE in the asphalt mixtures, the focus was on ensuring material homogeneity. The recycled LDPE underwent thorough pre-processing to minimize impurities and attain a more uniform composition. It is acknowledged that the recycled LDPE stream might contain traces of impurities inherent to the recycling process. Nevertheless, careful monitoring and control were exercised to maintain the desired quality standards for asphalt mixture production. The incorporation of recycled LDPE aligns with the commitment to sustainable practices and the reuse of materials in road pavement applications. The general properties of LDPE are shown in Table 1.