The data analysis conducted for each treatment response shows if the SR increased significantly after the application of the treatment and at what frequency, or if the SR maintained its value during the two weeks. Before studying the individual and general response of the treatments applied on the pavement surface, the data acquired from the Microgriptester for every week required filtering to find outliers. Results from the filtering process concluded that the data taken from every section under study required a 5% removal of outliers. After arranging the data set in increasing order, the analysis conducted did not include 5% of the lower and top parts considered as outliers.
4.1. Treatments General Response
The data collected every 5 cm with the Microgriptester showed discrepancies such as increment of SR without the application of a treatment or for the case of the control sections, SR increased and decreased from week to week. To reduce the risk of having errors in the response of the treated and untreated sections, a reduction of 5% of the outliers from the bottom and the top of the data set made possible a better and more accurate view of the results.
Table 4 shows the data considered in the study, including the quantity of data analyzed and the mean SR for each. Only the two untreated (control) sections considered the measured SR during the first week and the last week of the analysis and for the rest of the sections the included mean SR depended on the frequency of application.
The calculated mean SR for each section with and without outliers did not change more than 1% compared to the original mean SR, which allows us to have certainty that the outliers removed did not represent a major change in the average response of the treatment applied. Although the mean SR did not change significantly after the outlier removal,
Figure 2a,b show a different perspective of the impact of removing 5% of those outliers from each tested section.
The box plot and whiskers are shown in
Figure 2a, which includes the entire SR data set. The first two sets show the SR measured on the untreated sections during the first day (SR
0) and at the end of the study during the eighth week (SR
8). The following sets present the SR measured at treated sections right before the applied treatment (B
t) and after the applied treatment (A
t). Thus, for the section with water pressure applied weekly, A
t represents SR measured immediately after the treatment and B
t represents the SR measured a week later, just before the next treatment application. Although the whiskers plotted from the data acquired, from each section, show high variability, the plot shows the decreasing trend for each treatment and frequency. Both plots have the same scale on the y axle, which helps to visualize how the whiskers reduced drastically and are closer to the mean value after removing 5% of the outliers (
Figure 2b).
Figure 2b shows the unacceptable average SR from the control sections after two months, which proves the need for treatment. At the end of the second month, 63.8% and 68.2% of the SR measured at each untreated section were below 60.
Sections with water pressure applied weekly showed a good response in which 100% of the SR measurements maintained the SR above 60. Approximately 2.3% and 6.1% of SR measurements fall below 60 for the biweekly and monthly frequency, respectively.
SR measurements taken from sweeping weekly and biweekly were mostly above 60, and only 8.1% and 16%, respectively, were below. Although sweeping monthly achieved a higher mean SR compared to the other frequencies and treatment, 46.9% of the SR measurements were below 60 a month after the treatment.
Both treatments allow maintaining the average SR above 60, proving effectiveness at every frequency applied. For a cost-effective solution, the monthly application of water pressure maintains the average SR above the minimum limit. However, for the sweeping of the pavement surface, a monthly frequency will not prevent the SR from being below the required level for a short period. Sweeping biweekly will bring the needed performance to maintain the SR above the required level.
4.3. Water Pressure
Table 5 presents the results obtained from the one-tailed
t-test analysis conducted for the before and after treatment SR. Measurements were taken every week for each section; however, the analysis only considered the measure taken immediately before and after the treatment. For example, for a monthly frequency, the analysis included only the measurements taken at the end of each month. After the application of hydroblasting, the mean SR obtained from each section was considered as baseline (Mean
BL).
Results show that the null hypothesis is rejected since the t-statistic > t
0.05. There is statistically significant evidence at α = 0.05 that the mean SR measured before and after the treatment is greater than 60. Since the Mean
BL for each section is the same, a comparison can be made between frequencies. The frequency of treatment application shows a decreasing trend, in which the higher the frequency, the higher the mean SR.
Table 5 also shows that even though the section was not treated for an entire month, it still maintains the SR above 60.
4.4. Sweeping
Following the same procedure stated for water pressure,
Table 6 presents the results obtained from the one-tailed
t-test. The null hypothesis is rejected, therefore there is statistically significant evidence at α = 0.05 that the mean SR measured before and after the treatment is greater than 60. As a general observation, the mean SR for each section shows that for lower frequencies, the SR increases. Considering analyzing each section, unlike water pressure, the sections used for sweeping have different Mean
BL. This means that comparisons are possible between weekly and biweekly because the section with a monthly frequency started with a Mean
BL = 76.
The decreasing trend from the before treatment measurements proves that for lower frequencies, the SR reduces drastically, making it closer to 60. The after-treatment results suggest that sweeping the pavement surface frequently (weekly) does not maintain the SR higher than monthly. However, by comparing the MeanBL of each section with the calculated mean, a monthly frequency maintained the SR 4 friction numbers higher, while weekly and biweekly maintained the SR approximately 10 friction numbers higher.
Both treatments increased the average SR above 72 for every frequency. However, it does not prove that water pressure applied weekly is the feasible solution to maintain SR above 60 after the application of hydroblasting. When it comes to minimizing the impact to the traffic and finding an economic treatment application, water pressure applied monthly proves to be effective at maintaining SR above 60.
Every treated section maintained the SR value above the minimum required but sweeping applied monthly presented a better performance. A possible explanation could be due to the difference in MeanBL obtained after the hydroblasting application. The section with a monthly frequency started with an SR = 76, which was 10 friction numbers higher than the weekly and biweekly frequencies. The section performance was the best by increasing the SR after treatment, but when it comes to maintaining the SR after a month, 50% were below 60.