Evaluating the Effects of Transverse Bar on Vehicle Speed at an Arterial Road in Kuala Lumpur

Transcription

1 Asian Transport Studies, Volume 4, Issue 4 (2017), ATS All rights reserved Evaluating the Effects of Transverse Bar on Vehicle Speed at an Arterial Road in Kuala Lumpur Nur Shuhadah MOHD a, Abdul Azeez KADAR HAMSA b a Department of Urban and Regional Planning, International Islamic University Malaysia, Kuala Lumpur 50728, Malaysia; nurshuhadah.mohd@gmail.com b Department of Urban and Regional Planning, International Islamic University Malaysia, Kuala Lumpur 50728, Malaysia; azeez@iium.edu.my Abstract: Road safety is a major concern for road users. Vehicles traveling at speeds higher than the speed limit are one of the main reasons for the increased number of accidents along arterial roads in Malaysia. Transverse bars are considered as an effective measure in addressing the increased speed of traffic along arterial roads. This paper investigates the effects of transverse bars on the speed of the vehicles at a road segment along an arterial road in Kuala Lumpur. Two sets of transverse bars at an arterial road in Kuala Lumpur were selected. The design profiles and spot speed of the vehicles at the two selected transverse bars were measured. The speed of the vehicles before, on and after the transverse bars was analysed. The findings show despite the speed of the vehicles having decreased when approaching towards transverse bars, the speed still remains higher than the permissible speed limit. Keywords: Transverse Bars, Rumble Strip, Road Safety, Vehicle Speed, Arterial Road, Kuala Lumpur 1. INTRODUCTION Arterials and other major road networks in Malaysia have been experiencing an increase in the number of road accidents, resulting in high fatality rates. It was reported that an average of 18 fatal accidents were occurred everyday on major roads in Malaysia (Benjamin, 2015). Malaysia is also experiencing an average increase in traffic accidents of 9.7% per annum over the last three decades (Mustafa, 2005). The rapid growth in population, economic development, industrialization and motorization are some of the likely reasons for this trend. The Ministry of Transport (2012) and Royal Malaysian Police (2012) have indicated that the number of road accidents in Selangor and Kuala Lumpur had increased continuously from 2003 to 2012, especially along arterial roads. It clearly demonstrates how serious this problem is in developing cities. Vehicles travelling at a speed more than the stipulated speed limit is one of the main reasons for the increase in the number of accidents along these road networks. Mustafa (2005) noted that the number of vehicles exceeding the posted speed limit along major roads in Malaysia has increased over the years. According to the World Health Organisation (WHO) (as cited by NSW Centre for Road Safety, 2011), speeding "is Corresponding author. 723

2 unquestionably recognised as a major contributory factor towards both the number and severity of traffic crashes". Increase in vehicle speed has simultaneously led to the increase in the driver's reaction time (and thus stopping the vehicle less instantaneously to avoid collision), braking distance and the greater amount of kinetic energy to be absorbed by the impact of the crash. It is understandable that the amplification in the number of vehicles on the road would have negative impacts on the behaviour of the drivers. As indicated by the Ministry of Transport in Malaysia, the high number of motorcars on the road in Kuala Lumpur (more than one million motorcars) had caused discomfort among the drivers and eventually altered their driving behaviour. This has created opportunities for the drivers to speed whenever the road experiences less traffic volume, without considering the safety of other road users (Norsyahizan, 2007). The significance of this study is to ascertain the effectiveness of transverse bars on the change in vehicle speed as vehicles approach a change in road alignment along an arterial road. The extent to which vehicles reduce speed when approaching transverse bars is important to ascertain in determining speed-related accidents. The reduction in the number of accidents as a result of a decrease in the speed of vehicles would help to improve the road safety level of road users. It is considered as another significance of this study. The application of transverse bars at strategic locations along an arterial road is considered as one of the speed reduction measures to reduce the number of accidents especially fatal accidents (Carlson and Miles, 2003; Department for Transport, 2007; Minnesota Department of Transportation, 2009). Gorrill (2007) also agreed on it, however he indicated that, "...transverse bars are generally perceived to be effective in reducing intersection crashes when used appropriately, but there is no consensus on their effectiveness" (pp.3). Thus, it generally illustrates that the effectiveness of the transverse bars is subject to the appropriateness of its planning and design profiles. The purpose of this paper is to evaluate the effects of transverse bars on the speed of vehicles at certain stretches of an arterial road having certain roadway characteristics. This paper evaluates the suitability of transverse bars in terms of their planning and design characteristics on the reduction in the speed of the vehicles. Additionally, the speed trends of the vehicles near the transverse bars is also determined. The objectives of this study are: 1) to analyse the design characteristics of transverse bars and the geometrical design of the selected arterial road stretch, 2) to determine the speed characteristics of vehicles resulting from the effects of transverse bars at the selected arterial road stretch and 3) to evaluate the effects of design profiles of the transverse bars on the speed characteristics of the vehicles. The major contribution of this study is to devise an appropriate design and planning standard for the application of transverse bars along major roads in Malaysia to reduce the speed of the vehicles when approaching hazardous areas along the road. As a result, it would help to improve the road safety level for the road users in reducing the number of road accidents. In this paper, section 2 narrates the effects of transverse bars on vehicle speed from literatures, section 3 on transverse bars at the selected road stretch, section 4 on research approach used in this study, section 5 on analysis and findings and finally section 6 on conclusions. 2. TRANSVERSE BARS AND ITS EFFECTS ON VEHICLE SPEED O'Flaherty (1997a) described transverse bars as "attention-getting raised areas that are extended across the carriageway". It provides "... audible warning and physical vibration to alert the drivers..." (Gorill, 2007). There are several types of transverse bars that were 724

3 commonly installed around the world. The main function of transverse bars is to alert the drivers whenever the drivers encounter a cautionary situation on a road. The United States Federal Highway Administration has highlighted three types of rumble strips in general; road shoulder rumble strips, centreline rumble strips and transverse rumble strips (Robert Peccia and Associates, 2007). Transverse rumble strips are most commonly installed at important locations along major roads in Malaysia. Transverse rumble strips, also known as transverse bars in Malaysia, which include colour effects, are intended to alert the drivers with vehicle vibrations near hazardous points along an arterial road. They are generally installed well ahead of the hazardous point in order to give ample warning to the drivers to slow down their vehicle speed (Highway Planning Unit, 2002). O'Flaherty (1997b) stated that transverse bars are very effective in reducing the speed of the vehicles, especially with through traffic. Though transverse bars help to decrease the speed of the vehicles, on the negative side they tend to increase the noise level caused by the vibration of vehicles especially when vehicles cross over the transverse bars at high speeds. The planning and applications of any traffic calming measures, including transverse bars, are normally applied at a site according to its suitability as well as its surrounding environment. Nur Shazwani and Kadar Hamsa (2012) stated," the traffic calming plans need to consider the objectives of the neighbourhood, accessibility needs, safety and environmental standards". For a road with through traffic, transverse bars are generally considered as one of the most suitable traffic calming measures due to its design characteristics that do not give massive influence to vehicle speeds. It means that the suitability of a measure for implementation is highly important in ensuring its effectiveness to achieve the desired objectives. In Malaysia, the two types of transverse bars that are commonly used are typical transverse bars and the alert bars. Highway Planning Unit (2002) indicated that the typical transverse bars and the alert bars have similar design characteristics and specifications. The only difference between them is the pattern of spacing in between bars. The typical transverse bars have an equal spacing between each bar, while alert bars have unequal spacing (Ibid). The effectiveness of transverse bars in reducing speed of the vehicles varies according to its surrounding development. However, the impact of transverse bars is more or less the same, despite having differences in their design characteristics. A speed reduction ranging from 5 km/h to 7 km/h was found to be the immediate result after the installation of transverse bars (Adnan et al., 2002). Gorrill (2007) indicated that "...transverse bars are generally perceived to be effective in reducing intersection crashes when used appropriately, but there is no consensus on their effectiveness". Other than that, transverse bars also provide other positive impacts, especially to the road users: 1) improving the drivers alertness and thus effectively reducing accidents (Corkle et al., 2001; Carlson and Miles, 2003; Department for Transport, 2007), 2) economically feasible due to the positive cost-benefits of transverse bar construction or reconstruction project (Corkle et al., 2001), 3) provision of more time to respond towards any unexpected occurrences, thus improving the safety of drivers (Minnesota Department of Transportation, 2009), 4) noise and vibrations created through friction between car tyres and transverse bars surface could able to attract drivers attention towards any unusual vehicular movement, changes in road alignment and conditions that required stoppage of vehicles (Gorill, 2007), and 5) evaluation on the effectiveness of inducing compliance with traffic control devices (Carlson and Miles, 2003). 725

4 3. TRANSVERSE BARS AT WANGSA MELAWATI SEGMENT ALONG MRR2 The transverse bars on a road segment along an arterial Middle Ring Road 2 (MRR2) near Wangsa Melawati in Kuala Lumpur were selected for this study. This road stretch is located between south bound Klang Gate Interchange Exit 2813 and north bound Wangsa Melawati Interchange Exit This road stretch is located exactly at the border between Kuala Lumpur City Area and Ampang Jaya Area. Thus, it serves as a physical boundary between Kuala Lumpur City Hall and Ampang Jaya Municipal Council areas. Figure 1. Location of transverse bars on road segment along MRR2 near Wangsa Melawati The selected arterial road connects Gombak area in a northbound direction and Kuala Lumpur City Centre in a southbound direction. The MRR2 is also technically known as Federal Route 28 and has a total length of 65km, starting at Sungai Besi in the south of Kuala Lumpur and ending at Kepong in the north. The selected road stretch of MRR2 is a three-lane dual carriageway. The right-of-way of this road is 35m and the design speed is 80 km/h. The average traffic volume at the road segment along this arterial road was 7800 vehicles per hour in 2003 (Kadar Hamsa et.al., 2006). Transverse bars are the only vertical deflection provided on road stretches before hazardous points along MRR2. It is one of the important roads in Kuala Lumpur that enables traffic from outer areas of Kuala Lumpur to avoid entering into Kuala Lumpur city centre. Two sets of transverse bars were selected, one on each direction of the arterial road. Transverse Bar A is located before horizontal curve of the road with traffic heading in a northbound direction, while Transverse Bar B is located before the vertical curve with traffic heading in a southbound direction. Figure 1 shows the location of transverse bars A and B on the road segment along MRR2 near Wangsa Melawati. 4. RESEARCH APPROACH The following are the methods applied to collect data for this study. 726

5 4.1 Road Inventory Survey A road inventory survey was administered to collect data on the road geometrical details and design characteristics of the selected transverse bars. The data on road geometrical details include road design speed, road cross-sectional elements, and road alignment design. On the other hand, the data on the design characteristics of the transverse bars include height, width and length of the transverse bars and spacing between transverse bars. 4.2 Spot Speed Survey A spot speed survey was conducted to measure the speed of the vehicles near and on the selected transverse bars. The spot speed survey was administered on two different days, both on Friday, during off peak hours in order to get the actual speed of the vehicles. Direct-timing procedures by using a stopwatch were applied to record the total time taken by each vehicle when passing between two pre-determined fixed points. To compare the effects of transverse bars on the vehicle speed, the speeds of the vehicles were measured at three different points at each of the selected transverse bars. The first point was located at a distance of 50 meters before the transverse bars, the second on the transverse bars and the third 50 meters after the transverse bars. A pre-determined distance of 80 meters speed trap length (Currin, 2013), was set for recording the total time taken by each vehicle traversing over the specified length of the roadway. The selection of points is illustrated in Figure 2. Figure 2. Illustration of spot speed data collection using stopwatch method A total number of 200 motorcars was selected for the measurement of the spot speed. A systematic sampling method was used to select 200 samples over the given time period. The surveyors were positioned at a pedestrian over-bridge near the measurement points to ensure a clear view on the speed trap area, to ease communication among surveyors and to minimise the visibility of surveyors to the drivers of the vehicles in order to measure the actual speed of the vehicles. To ensure accuracy of the collected data, two separate sets of speed readings were collected for each sample at each of the three selected points. 5. ANALYSIS AND FINDINGS 5.1 Design Characteristics of Transverse Bar 727

6 The design characteristics of transverse bars A and B were found to be identical. However, the dimensions of each of the selected transverse bars were slightly different from the dimension provided by Highway Planning Unit (2002). Transverse bars A and B was slightly wider, and spaced closer between them when compared with specifications by the Highway Planning Unit. Table 1 shows the design characteristics of the selected transverse bars. There was also no signage ahead of the transverse bars to warn the road users about the presence of transverse bars. Table 1. Comparison between observed dimensions of transverse bars A and B and HPU design specifications. Dimension of Highway Planning Design Elements transverse bars Unit (HPU) A B Specification Number of bar Height of bar 3 mm 3 mm 3-5 mm Width of bar 600 mm 600 mm 300 mm Spacing between bars 2000 mm 2000 mm 2600 mm The wider dimension of bars has allowed vehicle movements to be smoother when passing over the transverse bars. Even though, the observed spacing of the transverse bars varied slightly with that of HPU specifications, it still yielded some impact on the vehicles. The "closer spacing allows vehicles to 'float' over the strips" (Department of Transport, 2007) and thus creates an insufficient level of vibration to the vehicles that passes by. Additionally, the 3 mm thickness of the bars is considered to be very marginal when compared with the vertical dimension of the transverse bars used in other countries, which some are up to 20 mm (Department of Transport and Main Road, 2002). As a result, the transverse bars have generated an insignificant level of vibrations and audible warning, which eventually allows only a slight reduction in the speed of the vehicles. 5.2 Spot Speed Analysis The analysis of spot speed at transverse bars A and B showed a different pattern when compared with each of the three selected points. Mostly, the speed of the vehicles at Point 1 of transverse bars A was higher than Point 2 and Point 3. Figure 3 illustrates the speed pattern at the selected points at transverse bar A. The highest and the lowest speeds at Point 1 of transverse bars A were 107 km/h and 62 km/h respectively. The highest and lowest speed of the vehicles at Point 2 were 83 km/h and 47 km/h respectively, and at Point 3 90km/h and 49km/h respectively. It shows that there is a reduction in vehicle speed, especially when vehicles travelled from Point 1 to Point 2. Thus, it indicates a change in the speed of the vehicles when approaching the transverse bars. 728

7 Figure 3. Spot speed at transverse bars A (before horizontal curve) Whereas at transverse bars B, the vehicles had increased speed when approaching the transverse bars. The speeds of the vehicles at Point 2 were higher than the speed at Point 1, in which the highest and lowest speeds at Point 2 are 104km/h and 53km/h and at Point 1 81km/h and 48km/h respectively. The speed trend at various points at transverse bar B is shown in Figure 4. The increase in speed from Point 1 (before the transverse bars) to Point 2 (on the transverse bars) is due to the location of the transverse bars, which are located before a vertical curve. The location of these transverse bars before a vertical curve on the road segment has caused the vehicle speeds to increase in order to manoeuvre the vertical curve. The increase in vehicle speed from point 1 to point 2 of these transverse bars was also noticed after collecting new data on the speed at the selected points on two different days of a week. The findings on spot speed on these two different days are given in Figure A1 and A2 in the appendix. On the other hand, surprisingly, the speed of the vehicles decreases when vehicles travelled from Point 2 to Point 3 (after the transverse bars). It is due to the downward slope of the road segment that follows immediately after the vertical curve. Figure 4. Spot speed at transverse bars B (before vertical curve) 729

8 When comparing vehicle speeds at transverse bar A and B, it was found that almost all vehicles at Point 2 (on the transverse bars) both at transverse bar A or B were travelling at speeds well above the ideal permissible speed limit of 55km/h as suggested by Highway Planning Unit (2002). Although the speed pattern has a downward trend as the vehicles move from one point to another point (especially at transverse bar A), the vehicles were, however, still travelling at speeds higher than the permissible ideal speed limit on transverse bars. 5.3 Spot Speed Characteristics Table 2 shows the spot speed characteristics of the vehicles at both transverse bars A and B. It can be seen from Table 2 that the mean speed of the vehicles from Point 1 to Point 2 at transverse bars A has decreased by 15 km/h, whereas it was almost the same from Point 2 to Point 3. On the other hand, at transverse bars B, the mean speed of the vehicles has increased by 18 km/h from Point 1 to Point 2, but decreased by 4 km/h from point 2 to point 3. Table 2. Spot speed characteristics of the vehicles at transverse bars A and B. Speed Characteristics Transverse bars A (km/h/) Transverse bars B (km/h/) P1 P2 P3 P1 P2 P3 Mean Mode Median th percentile spot speed Standard deviation Except at point 1 of transverse bar A and point 2 of transverse bar B, the median speed was less than the permissible speed limit (80km/h). On the other hand, the median speed of the vehicles at Point 2 was higher than the ideal permissible speed limit of 55 km/h when passing over transverse bars A and B. The 85 th percentile speed at Point 1 of transverse bars A was well above the acceptable speed limit of 80 km/h. The speed of the vehicles, however, was gradually reduced to below 80km/h at Point 2 and 3. On the contrary, the 85 th percentile speed at point 2 and 3 of transverse bar B was above 80km/h, but at point 1, it was below 80 km/h. These findings clearly demonstrate that, despite the speed of the vehicles decreasing when approaching transverse bars A, the vehicles were travelling well above the permissible speed limit of 55km/h. The speed of the vehicles increased when approaching transverse bars B. The 85th percentile speed of the vehicles on the transverse bar B and immediately after passing over this transverse bars was well above the permissible speed limit, but at point 1, the vehicles were travelling at a speed less than the permissible speed limit (80 km/h). It shows that the presence of a vertical curve near the road segment has made vehicles slow down before approaching the transverse bars. However, the speed of the vehicles increased as the vehicles approached the transverse bars due to vehicles accelerating to negotiate the vertical curve. The vehicle speed immediately after passing transverse bars B was above the speed limit of 80 km/h due to the presence of a downward slope. These findings have clearly demonstrated that the presence of the existing transverse bars is not really effective in reducing the speed of the vehicles to the desired acceptable permissible speed limit. 730

9 Referring to Figure 5, 98 percent of the samples were moving below the speed limit after passing transverse bars A (Point 3), when only 40 percent of samples were initially travelling below 80km/h at Point 1. It is due to the straight stretch of road with a length of approximately 200 meters prior to Point 1, which has encouraged drivers to accelerate before approaching transverse bars A. It is important to note that only 1.5 percent of the total samples (3 vehicles) were travelling below the ideal permissible speed limit (55 km/h) on the transverse bars as suggested by Highway Planning Unit (2002). It clearly indicates that the design profiles of the transverse bars are unable to reduce the speed of the vehicles to within the ideal permissible speed limit when travelling over the transverse bars. At transverse bars B, 198 samples (99%) were travelling below the speed limit at point 1 as shown in figure 6. It is due to the location of transverse bars as they are located before a vertical curve and the presence of another set of transverse bars approximately 150 meters before Point 1. Thus, the 85th percentile speed at Point 1 of transverse bars B was significantly lower than Point 1 of transverse bars A. Figure 5. Cumulative frequency curve for spot speed at transverse bars A Figure 6. Cumulative frequency curve for spot speed at transverse bars B The increase in vehicle speed, especially between Point 1 and Point 2 at transverse bar B is due to the location of this transverse bar as it is located before a vertical curve along a gradeseparated intersection. The presence of this vertical curve, somehow has encouraged drivers to accelerate to negotiate the ascending slope of the road segment. This was mentioned by O'Flaherty (1997b) as he explained that "...drivers usually accelerate upon entering an uphill section and the extra momentum can overcome the effect of gradient..." (pp.333). In addition, unlike other traffic calming measures, transverse bars provide a visual and audible warning to 731

10 the drivers for immediate and significant speed reduction. In the presence of road curvature, road marking warnings such as transverse bars have a less immediate result on drivers' speeds when compared with road sign warnings as transverse bars are less visible to drivers at a greater distance (Charlton, 2007). Thus, it indicates that the results of speed reduction from the installation of transverse bars can only be seen after the drivers pass through it. With reference to the findings on vehicle speed at transverse bar B, it can be seen that vehicles have started to decelerate only at Point 3, which is immediately after passing the transverse bars. Thus, this finding explains on how drivers behave in terms of vehicle speed at transverse bars when it is located near a vertical road curve. 5.4 Effects of Transverse Bars on Vehicle Speed At transverse bars A, the vehicles had significantly reduced speed when moving from Point 1 to Point 2. A maximum reduction in speed of 35 km/h was achieved when vehicles travelled from Point 1 to Point 2, as compared with the maximum reduction of 17km/h from Point 2 to Point 3. Figure 7 illustrates the reduction in speed of all the sampled vehicles from Point 1 and Point 2 and from Point 2 and Point 3 at transverse bars A. Whereas at transverse bars B, this trend was reversed as vehicles had increased speed in the range between 3 and 39 km/h when travelling from Point 1 to Point 2. The reason is due to the presence of a vertical curve located at a distance of less than 100 m from transverse bars B, which makes the vehicles accelerate when moving towards transverse bars B. Figure 8 illustrates the changes in speed of all the sampled vehicles from Point 1 and Point 2 and from Point 2 and Point 3 at transverse bars B. However, the vehicles had reduced speed by up to 30 km/h when travelling from Point 2 to Point 3 at transverse bars B. Figure A3 and A4 shows the changes in the vehicle speed at transverse bars B based on the new data collected on two different days of a week. Again, it shows that speed increases as vehicles moves from point 1 to point 2 of transverse bar B. Figure 7. Changes in spot speed of all the sample vehicles at transverse bars A. 732

11 Figure 8. Changes in spot speed of all the sample vehicles at transverse bars B. It should be noted that 53 percent of the samples at transverse bars A recorded speed reductions between 10 and 19 km/h (27% of vehicles experienced a reduction between 10 and 14 km/h, and 26% between 15 and 19 km/h) when travelling from Point 1 to Point 2. On the other hand, about 80% of the vehicles had accelerated when travelling towards transverse bars B from Point 1. The acceleration in speed at transverse bars B was between 10 and 24 km/h (24% between 10 and 14 km/h, 29% between 15 and 19 km/h and 27% between 20 and 24 km/h). It clearly shows that the reduction in vehicle speed towards transverse bars B is insignificant due to the presence of a vertical curve located after the transverse bars. Figures 9 and 10 illustrate the frequency of vehicle reduction at transverse bars A and B. Figure 9. Frequency of speed reduction at transverse bars A. 733

12 Figure 10. Frequency of speed reduction at transverse bars B. Based on these findings, it is clearly illustrated that effectiveness of transverse bars depends on its planning and design profiles. Additionally, the location of transverse bars and roadway characteristics is also the main cause for the difference in speed changes at transverse bar A and B. Charlton (2007) in his experiment on the impact of horizontal curves with road marking treatments on vehicles speed also found out positive speed reduction from the auditory feedback from the pavement marking of transverse bars. The positive speed reduction especially at transverse bar A from this study is similar with that of findings from Charlton (2007). Previous studies (Department for Transport, 2007; Minnesota Department of Transportation, 2009; Highway Planning Unit, 2002) also revealed that transverse bars had effectively reduced the speed of vehicles on arterial roads with an average reduction of 3 to 10 km/h. However, the impact of transverse bars on vehicle speed may differ occasionally depending on the steepness of gradients and sharpness of road curves (Räsänen, 2005; Charlton, 2007). Martens, Compte & Kaptein (1997) have stated that road gradient can affect driving speed similarly to that of road curvature due to the gravity effect. Additionally, Charlton also explained that the effect of transverse bars on speed is the greatest in the later portion of the hazardous area of the road, whether it is horizontal or vertical curve (Charlton, 2007). This finding is similar to that of the finding especially at transverse bar B in this paper as the vehicles tend to accelerate at the beginning and start to decelerate at the end of the transverse bars. 5.5 Testing the Differences in Spot Speed before and on Transverse Bars It is generally noticeable that the speed of the vehicles decreases when approaching the transverse bars. The difference in the speed of vehicles before and on the transverse bars was tested to determine whether there is a significant statistical change in the vehicle speeds. A t- test was applied to test the differences in speed at both transverse bars A and B. Table 3 shows the results of the t-test at transverse bars A, and Table 4 at transverse bars B. 734

13 Table 3. Paired samples t-test for vehicles speed at transverse bars A Paired Differences Mean Std. Dev. Std. Error Mean 95% Confidence interval of the difference Lower Upper t df Sig. (2- tailed) Pair Before - On Table 4. Paired samples t-test for vehicles speed at transverse bars B Paired Differences 95% Confidence interval of the t df Mean Std. Dev. Std. Error Mean difference Lower Upper Sig. (2- tailed) Pair Before - On The results show that there is a statistically significant difference in speed when the vehicles approached both transverse bars A and B. The t-value is with a significance value of (p<0.05) at transverse bars A, and and (p<0.05) at transverse bars B. The decrease in speed as the vehicles approached transverse bars A was ± 7.51 km/h, but the speed increased by18.30 ± 6.13 km/h when vehicles approached transverse bars B. The increase in speed at transverse bars B is due to the presence of a vertical curve at 100 m after the transverse bars. 6. CONCLUSIONS Arterial roads are frequently subjected to a higher number of accidents than other types of roads. One of the reasons for this trend is vehicles travelling at speeds higher than the permissible speed limit. The application of transverse bars near the changes in road curvature is one of the measures to reduce the speed of the vehicles, and the number of accidents. As a result, transverse bars can help increase the road safety of road users. The effectiveness of transverse bars, however, depends on the design profiles of the transverse bars and their location. Carlson and Miles (2003) have stated that the installation of transverse bars should consider several aspects, such as number of hazard points, installation distance, and types of transverse bar selected, to assure its positive impact on vehicle speed. One of the major findings from this paper shows that the speed of the vehicles decreased when vehicles approached transverse bar A, but not in the case of transverse bars B. The design profiles of transverse bars at A is similar to that of B. However, the location of transverse bar B is very close to a vertical curve and thus it requires vehicles to accelerate to overcome the vertical curve and at the same time to maintain the minimum accelerated speed to negotiate the vertical curve. It should also be noted that even though the speed of the vehicles decreased when approaching transverse bar A, the speed of vehicles on the 735

14 transverse bars was still above the ideal permissible speed limit speed (55 km/h) as suggested by Highway Planning Unit (2002). One of the main reasons for this trend is due to the height of the transverse bars, which was found to be lower than design specifications adopted in other countries. As a result, many vehicles were travelling at a speed higher than the ideal permissible speed limit speed on the transverse bars when approaching transverse bar A. Based on these findings, it is clearly illustrated that the effectiveness of transverse bars depends on their planning and design profiles. Additionally, the location of transverse bars and roadway characteristics is also the main cause for the difference in speed changes at transverse bar A and B. Charlton (2007) in his experiment on the impact of horizontal curves with road marking treatments on vehicle speeds also found a positive speed reduction from the auditory feedback from the pavement marking of transverse bars. The positive speed reduction especially at transverse bar A from this study is similar with that of findings from Charlton (2007). Previous studies (Department for Transport, 2007; Minnesota Department of Transportation, 2009; Highway Planning Unit, 2002) also revealed that transverse bars had effectively reduced the speed of vehicles on arterial roads with an average reduction of 3 to 10 km/h. However, the impact of transverse bars on vehicle speed may differ occasionally depending on the steepness of gradients and sharpness of road curves (Räsänen, 2005; Charlton, 2007). Martens, Compte & Kaptein (1997) have stated that road gradient can affect driving speed similarly to that of road curvature due to the gravity effect. Additionally, Charlton also explained that the effect of transverse bars on speed is the greatest in the later portion of the hazardous area of the road, whether it is a horizontal or vertical curve (Charlton, 2007). This finding is similar to that of the finding especially at transverse bar B in this paper as the vehicles tend to accelerate at the beginning and start to decelerate at the end of the transverse bars. As the effectiveness of transverse bars depends on their design and planning profiles, it is important to ensure that transverse bars are being installed appropriately based on the roadway characteristics. Othman et al. (2015) indicated that transverse bar guidelines practiced in Malaysia require further review especially on the planning and design profiles of the transverse bars to make them effective in dealing with the increasing speed of the vehicles along arterial roads. Thus, to increase the effectiveness of transverse bars, the height of the transverse bars should be increased to meet the design specifications as practiced in other countries. Additionally, the location of transverse bars should be avoided near any vertical curves of a road as it proves to be ineffective in reducing the speed of the vehicles. Thus, the location of transverse bars is crucial in order to increase their effectiveness in reducing the speed of the vehicles. It is clear that transverse bars can help reduce the speed of the vehicles, but their actual effectiveness in reducing the speed of vehicles still depends on the implementation of the design specifications and guidelines of transverse bars. The application of transverse bars with desirable design specifications, coupled with strict enforcement, will help reduce the speed of the vehicles and thus enhances the safety of the road users. 736

15 APPENDICES Figure A1. Spot speed at transverse bars B (before vertical curve) for the new data on day 1 Source: Primary source, 2017 Figure A2. Spot speed at transverse bars B (before vertical curve) for the new data on day 2 Source: Primary source,

16 Figure A3. Changes in spot speed of all the sample vehicles at transverse bars B for the new data collected on day 1 Source: Primary source, 2017 Figure A4: Changes in spot speed of all the sample vehicles at transverse bars B for the new data collected on day 2 Source: Primary source,

17 REFERENCES Adnan, M.A., Ahmad, J., Sulaiman, N. (2002). Evaluation of thermoplastic transverse bar in reducing traffic speed. Proceedings of Malaysian Universities Transport Research Forum, Kuala Lumpur, Malaysia, Benjamin, N. (2015). JPJ: People die on Malaysian roads every day. The Star. 27April Carlson, P.J., Miles, J.D. (2003). Effectiveness of rumble strips on Texas highways: First year report. College Station, Tex.: Texas Transportation Institute, Texas A&M University System. Charlton, S.G. (2007). The role of attention in horizontal curves: A comparison of advance warning, delineation, and road marking treatments. Accident Analysis & Prevention, 39(5), Corkle, J., Marti, M., Montebello, D. (2001). Synthesis on the effectiveness of rumble strips. Minnesota Local Road Research Board, Minnesota Department of Transportation. Currin, T.R. (2013). Introduction to Traffic Engineering: A Manual for Data Collection and Analysis. Second Edition. Cengage Learning, Stamford. Department of Transport (2007). Traffic Calming. Local Transport Note 1/07, Department for Transport, The Government of United Kingdom. Department of Transport and Main Road (2002). TRUM Technical Note.3.2 Rumble Strips. Department of transport and Main Road, Queensland Government. Gorrill, D. (2007). Transverse Rumble Strips. Transportation Research Synthesis, Minnesota Local Road Research Board, Minnesota Department of Transportation. Highway Planning Unit (2002). Traffic Calming Guidelines. Ministry of Works, Kuala Lumpur, Malaysia. Kadar Hamsa, A.A., Miura, M., Inokuma, S., Nishimura, Y. (2006). Evaluating the living environment in residential areas at Taman Melati, Kuala Lumpur. Journal of Asian Architecture and Building Engineering, 5(2), Martens, M.H., Compte, S., Kaptein, N.A. (1997). The Effects of Road Design on Speed Behaviour: A Literature Review. Transport RTD Programme of the 4th Framework Programme. European Commission. Ministry of Transport (2012). Transport Statistics Malaysia Ministry of Transport, The Government of Malaysia. Minnesota Department of Transportation (2009). The Effect of Rumble Strips on drivers Approaching Rural, Stop-controlled Intersections. Minnesota Local Road Research Board, Minnesota Department of Transportation. Mustafa, M.N. (2005). Overview of Current Road Safety Situation in Malaysia. Highway planning Unit, Road Safety Section, Ministry of Works, 5 9. NSW Centre for Road Safety. (2011). How does speeding increase the chances and severity of a crash? Speeding Fact Sheet 4 of 6. Transport Roads and Traffic Authority, New South Wales Government ( Norsyahizan, M.N. (2007). A Study on the Effects of Road Hump in Reducing Speed along Major Road of International Islamic University Malaysia, Gombak. Bachelor Degree Project Paper Report, International Islamic University Malaysia, Kuala Lumpur. Nur Shazwani, R., Kadar Hamsa, A.A. (2012). A Theoretical Review in Evaluating the Impact of Traffic-Calming Measures on the Residential Living Environment. Presented at International Conference on Green in the Built Environment, May 2012, International Islamic University Malaysia. 739

18 O'Flaherty, C A. (1997a). Physical Method of Traffic Control. In O'Flaherty, C A (ed.), Transport Planning and Traffic Engineering, Oxford, UK: Arnold. O'Flaherty, C A. (1997b). Geometric Design of Streets and Highways. In O'Flaherty, C A (ed.), Transport Planning and Traffic Engineering, Oxford, UK: Arnold. Othmana, M.H., Harona, Z., Haininb, M.R., Yahyaa, K., Yaacobb, H., Sanikc, M.E. (2015). Malaysian Transverse Rumble Strips: A Review and Recommendations for Practice. Jurnal Teknologi, 73(4), Räsänen, M. (2005). Effects of a rumble strip barrier line on lane keeping in a curve. Accident Analysis & Prevention, 37(3), Robert Peccia and Associates. (2007). Chapter 8: Traffic Calming. Greater Bozeman Area Transportation Plan. Bozeman City Government ( olicies/transportation.planfull.pdf). Royal Malaysian Police (2012). Malaysia Statistical Report Road Accident Kuala Lumpur, Malaysia. 740