Statistical Analysis of Truck Loading on Swedish Highways

Transcription

1 Statistical Analysis of Truck Loading on Swedish Highways Master s Degree Project Entesar Abdullah Ali Division of Highway and Railway Engineering Department of Civil and Architectural Engineering Royal Institute of Technology SE-1 Stockholm TRITA-VBT 11:1 ISSN X ISRN KTH/VBT-11/1-SE Stockholm 11

2

3 Statistical Analysis of Truck Loading on Swedish Highways Entesar Abdullah Ali Graduate Student Infrastructure Engineering Division of Highway and Railway Engineering School of Architecture and the Built Environment Royal Institute of Technology (KTH) SE- 1 Stockholm kth.se Abstract: Vehicle over loading, or single axle over loading, is one of the major causes of pavement deterioration. Trafik Verket (TV), the Swedish Transport Administration, recognized that the current process for estimating traffic volume should be reevaluated, and if possible improved. This degree project uses data from the Bridge Weigh in Motion (BWIM) system to study the actual loads applied to Swedish highways. The axle load spectrum is plotted with the conventional frequency distribution plots, and with a new cumulative distribution approach. The paper introduces the maximum allowable potential vehicle weight MAPVW concept, and uses this visual technique to identify overloads for different vehicle geometries. The paper concludes that for 5 and 6 axle trucks the triple axle is frequently overloaded, while for longer trucks one of the dual axles is often over loaded. The highest over loads tend to be on the driving axle, suggesting incorrect loading procedures. KEY WORDS: BWIM data; Axle load spectra; Outer axle distance; Inter axle distance; Gross vehicle weight (GVW); Maximum allowable potential vehicle weight (MAPVW); Steering Axle; Driving axle; Non driving axle; Single axle; Double Axle; Triple axle. i

4 ii

5 Acknowledgement It is a pleasure to thank the many people who made this thesis possible. After thanking God for giving me the opportunity to complete my master studies at Royal Institute of Technology (KTH), I would like to express my sincere gratitude and appreciation to my supervisors, Prof. Björn Birgisson and Dr. Michael Behn for giving me the chance to work with them, sharing their expertise, guidance, knowledge, beside their kindness and endless patience. Also I would like to thank Thomas Winnerholt for supplying us with data needed in this study. I wish to thank my husband and my family in Jordan especially my mother for their continuous support and encouragement. Finally I would like to thank everyone contributed to the completion of this research. iii

6 iv

7 List of Abbreviations AASHTO AGW ALS ATA BWIM Cost ESALs FAD FHWA GVW IAD ME MPAL NGWBS NOR OAD RMA TCF TV WAVE - American Association of State Highway and Transportation Officials - Actual Gross Weight - Axle Load Spectra - American Trucking Association - Bridge Weigh in Motion - Co-Operation in the Field of Scientific and Technical research - Equivalent Single Axle Load - Free of Axles Detectors - Federal Highway Administration - Gross Vehicle Weight - Inter Axle Distance - Mechanistic Empirical - Maximum Potential Axle Load - New Generation of Wide Base Single - Nothing on the Road - Outer Axle Distance - Rubber Manufacturers Association - Tire Configuration Factor - Trafik Verket - Weighing of Axles and Vehicles for Europe v

8 vi

9 Table of Contents Abstract i Acknowledgement ii List of Abbreviations v Table of Contents vi 1. Introduction Literature Review History of Super Single Tires Damage Caused by Super Single Tires BWIM Data Development of the BWIM Measurement System Allowable Gross Vehicle Weights in Europe Allowable Axle Loads in Sweden BWIM Data from Sweden Methodology Axle Load Spectrum (ALS) and Cumulative Frequency Distribution Graphs Maximum Allowable Potential Vehicle Weight (MAPVW) Graphs Comparison Between and Results ALS and Cumulative Frequency Distribution Analysis MAPVW and GVW Analysis Analysis of Triple Axle Characterization of the Magnitude of the Over Load Conclusion & Recommendation References Appendix A Appendix B Appendix C vii

10 viii

11 1. Introduction Vehicle over loading, or single axle over loading, is one of the major causes of pavement deterioration. Trafik Verket (TV), the Swedish Transport Administration, recognized that the current process for estimating traffic volume should be reevaluated, and if possible improved. This master s degree project uses data from the Bridge Weigh In Motion (BWIM) system to study the actual loads applied to Swedish highways. This study s main objectives are to analyze the data collected by the BWIM system, to develop a better understanding of the actual network loading, and to help to predict actual critical loading conditions for Swedish highways. The data was analyzed by separating out the different components of the axle load spectra, for the steering, driving, double, and triple axle loads. New approach was used to analyze the data by comparing the vehicle gross weight with the allowable load for different truck axle groups for each axle category. Chapter discusses the history of super single tires. Chapter 3 documents the BWIM system, and presents the allowable highway loads for Sweden and Europe. Chapter presents the methodology used to study vehicle and axle over loads. Chapter 5 discusses the analysis results. Chapter 6 presents the conclusions and recommendations. Appendices A, B and C present the complete analysis of each set of data analyzed. 1

12

13 . Literature Review.1. History of Super single tires During the last two decades there were many discussions concerning the advantages and the disadvantages of the use of super single tires, also known as wide base tires. Before going through those discussions let us first define what is meant by Super single tire, and since when has they been first used? Super single tires can be defined as a tire which is not intended to be mounted in dual assembly, but with a load bearing capacity of the same order of magnitude as a dual tire assembly. Initially Super single tires needed changes in the axle configuration for installation, but a new generation of Super single tires allow for a replacement of the dual tires without major changes. The new 5/5R.5 tire replaces 75/8R.5 duals, and the 55/55R.5 tire replaces either 11R.5 or 75/8R.5 tires (American Trucking Association, 7). 5/5R.5 stands for the tire width in mm/aspect ratio, the construction method, and the rim diameter in inches. The aspect ratio is the ratio of the height of the tire from the bead to the top of the tread to the tire width. Construction method R designates that the tire was made using radial construction. According to American Trucking Association (ATA), members and published sources, both Michelin and Bridgestone produced 5/5R.5 size, while only Michelin offered the 55/55R.5 size of these new tires (ATA, 7). A study conducted by AL-Qadi (Federal Highway Administration, 7), states that Super single tires have been used in Europe since the early 198 s. As of 1997, 65% of trailers in Germany used Super single tires. According to Modern Tire Dealer magazine (1993) based on the Rubber Manufacturers Association (RMA), the market penetration of super single tires for both original replacement tires in the US market was less than % through 199 as shown in Table 1. Thus the starting point of the super single tires is in the 198 s in Europe, and the 199 s in the US. According to a workshop organized by the Federal Highway Administration Super single tires are not widely used in South Africa (1-15%), because load limits laws restrict their use (FHA, 7). An article published in states that the Super single tire market is still limited because the drivers find it difficult to replace these tires on the road, and they are afraid to be immobilized if a failure occurs (Environmental Protection Agency, ). Truck operators found potential benefits from Super single tires, such as reduced weight and fuel economy. Tire manufacturers found Super single tire have lower costs, and increases stability on road. On the other hand pavement engineers and road owners found that this Super single tire contribute to increasing pavement distresses. 3

14 Year Original Equipment Replacement Table 1 Percent of Super Singles (Modern Tire Dealer magazine, 1993)... Damage Caused by Super single tires An investigation was carried out by the Ontario Ministry of Transportation to assess the potential impact of using Super single tires (Ontario Ministry of Transportation, 3). The study concluded that at an axle load of 73 kg, the impact on pavement caused by Super single tire and dual tire are the same. But if the axle load is less than 5 kg the damage due to wide base tire is % less than the dual tire; since the contact area at this loading (5 kg) is greater % for wide base tire than dual tire. In the other hand the impact on pavement caused by wide base tire at an axle load of 8 kg is the same impact caused by dual tire at 1 kg axle load. If the axle load on wide base tire is increased up to 9 kg the potential pavement damage could be up to 1 times more than the damage caused by conventional dual tire under the same loading. Another study about the impact of Super single tire on pavement was done by Al-Qadi (Federal Highway Administration, 7), testing at the Virginia smart road test site between years and. The study concluded that the fatigue failure is slightly greater for Super single tire configuration. The subgrade rutting is approximately equal for the wide base and dual tire. Finally, surface cracking is less for super singles than for dual tires. The European Co-Operation in the field of Scientific and Technical research (COST, 1) studied the effect of tires on pavement wear. The group carried out field tests and numerical studies to investigate effects of single tires, wide base tires, and dual tires on the wear of pavement structure (stresses and strains generated in the pavement by the passage of different wheel load configurations). The data was used to develop a set of equations for the so called Tire Configuration Factor (TCF), which expresses the amount of pavement wear relative to an arbitrarily chosen reference tire. The higher the TCF value, the higher the pavement wear caused by that tire.

15 The TCF number depends on tire configuration, size parameters, loading, and inflation pressure. It can be used to give an indication of the stresses caused by different kinds of tires, taking into consideration pavement thicknesses and distresses modes. Since the calculated TCF depends on the tire type, values were analyzed to find the distresses caused by current and future tires. This tool might guide the manufacturing of new tires based on criteria depending on the TCF number in order to limit future pavement wear. It is necessary to mention that the group stated that a large difference in relative pavement wear exists among dual tire assemblies, and among wide base single tires. Therefore, a single factor for the difference between wide base single and dual tires is not applicable. Comparisons between pavement wear effects can only be made if the detailed characteristics of tire configuration are taken into consideration. The study found that for rutting in thick and medium pavements this kind of tire does not differ significantly from dual tires, when the axle load, tread pattern width, contact area, tire diameter, and pressure ratio are equal. Table shows the results found by COST comparing between dual tires and wide base tires for different axles as shown. Axle Type Steering Axle: Reduction of contribution to Pavement Wear TCF Values for Primary Roads 36% TCF Values for Secondary Roads 5% Driving Axles: Dual Tires Wide Base Tires Increase of contribution to Pavement Wear 17% 6% Towed Axles: Dual Tires Wide Base Tires Increase of contribution to Pavement Wear 17% 97% Table Tire Configuration Factors for Super Single and Standard Tires. 5

16 The damage caused by different tire types was studied at KTH in 9, using the Florida Cracking Model. It was found that the damage increases from 15 to 3% when switching from dual to super single tires. Further results can be seen in Table 3. (Khavassefat, 9) Damage Ratio DCSE (kj/m 3 ) * 7-Axle Dual Tyre 7-Axle Super Single 8-Axle Dual Tyre 8-Axle Super Single 7-Axle Super Single +% Extra Load 7-Axle Dual Tyre,1 1 7-Axle Super Single,7 1,9 1 8-Axle Dual Tyre.3 1,1, Axle Super Single.7 1,9 1,1 1, Axle Super Single+% Extra Load.37 1,79 1,39 1,59 1,39 1 * DCSE is listed for the speed of 8 km/h Table 3 Damage Evaluation for Different Truck Type Calculated by FCM Method (Khavassefat, 9). 6

17 3. BWIM DATA 3.1. Development of the BWIM Measurement System The actual traffic loading is one of the most important input parameters for the analysis and design of the highways. If the loads are not well known, the pavements may be over designed, or may fail early. Both scenarios represent an uneconomical design. In the first one too much money was spend, in the second case, too little. Only by knowing the exact traffic load, can an optimum design be achieved. Furthermore, weight restrictions, or other legal limits on highway loading, are not always enforced, or enforceable. Two approaches can be used in pavement design, the ESAL approach and the Axle Load Spectra (ALS) approach. In the ESAL approach, all axle loads are converted into Equivalent number of Single Axle Loads. ALS approach does not convert loads into ESALs, it maintains data by axle configuration and weight. ALS indicates traffic loading which is used with the environmental conditions, to calculate the pavement response in the mechanistic empirical design. One way to measure the actual load spectrum on roads is the Bridge Weigh In Motion (BWIM) system, this system was first developed by Moses (1979), it has been used in Japan (Matsui and EL-HKIM, 1989), and was introduced to Sweden in. Since BWIM system was first developed in late 197 s, several research projects have been done to develop the system. One of these studies was done in late 199 s as a part of Weighing of Axles and Vehicles for Europe (WAVE) project. The WAVE project introduced the Nothing on the Road (NOR) measurement system. By mounting all instruments on the bottom of the bridge elements. The system indicates data by strain recordings that depends on the bridge structure as can be seen in Figure 1. All data in Figure 1 represent the response of structure during the passage of () Tones 5 axles tractor semi trailer with two single axles and one triple axle. BWIM system not only measures vehicles axles and gross weight, but also the vehicle speed. This can be done by placing strain transducers called axle detectors, on the top, or more typically, on the bottom of the bridge deck, along the entire width of the structure. That would be called (NOR), or Free of Axle Detectors (FAD). Axle detectors on the top of the pavement are difficult to install and maintain, and deteriorated over time and cause traffic delays during installation. It is important to note that different bridge types require individual analysis. Figure 1 (A) shows a long and thin bridge with high dynamic response which produces additional peaks into strain signal. Such peaks should be ignored and not treated as axles. Figure 1 (B) shows shorter bridge with thick slab deck, it has no dynamic response, but the peaks from group axles are smeared and difficult to find. Figure (C) shows the ideal bridge for NOR installation, which is short bridges or longer ones with secondary elements that divide the main span into shorter (sub-spans). Such cross beams or cross stiffeners, since in long bridges 7

18 Figure 1 (A) Bridge with High Dynamics, (B) Bridge with Thick Slab.8m (C) Orthotropic Deck Bridge, (D) Beam- Deck Bridge Instrumented on the Slab Between Beams (Kalin, 1). Figure Measured and Processed signal For 5 Axle Vehicle. 8

19 the measured strain signals represent joint contribution of several axles that are on the bridge at the time. Figure 1 (D) confirmed that NOR can be used on longer bridges but the strain transducers for detecting axles are attached to the slab between two beams close to the cross stiffener at the mid span. For bridges that have clear peaks, like the bridge in Figure 1(C), speed can be calculated from time difference of peaks from strain signals. For bridges with too many peaks or smeared peaks, like bridges in Figure 1(A) and 1(B), some Post processing of the measured data is needed. To determine the axle for these types of bridges, the signal should be filtered and processed. An example of measured and processed signal for 5 axle truck is shown in Figure. 3.. Allowable Gross Vehicle Weights in Europe Maximum allowable vehicle weights in Europe are determined for each country. Table shows the maximum weights for different axle types. It is noticeable that Sweden has the highest allowable weights for the trucks with axles, 5 axles and more than 5 axles. The Netherlands, Norway and Luxembourg also have higher than average allowable gross vehicle weights (GVW). Sweden limits the GVW for the road trucks with axles and greater by 6 tons which is the highest limit of all European countries. It is suggested that the reason behind this is the timber trade in Sweden. While Netherlands limited the load by 5 Tons, Norway and Luxembourg defined it by tons, which is still higher than the rest of the countries which limited the load by tons only Allowable Axle Loads in Sweden Table 5 shows loading limits specified for each kind of axle. To analyze BWIM data correctly we need to split the data set into equivalent categories for analysis. What are the basics that BWIM system uses for classifying these axle types, whether single, double, or triple? But before going through that, it may be helpful to explain some terms related to truck mechanism. Steering axle: The axle on which wheels turn left and right as controlled by the rotation of steering wheel in the driver s compartment. For this study, the steering axle is the first axle group in the front of the vehicle. Driving axle: The axle which is connected to the power train of the vehicle s motor and transmits tractive power to the wheels. The driving axle considered as the second axle group whether it is single, double, or triple axle. Single axle: One axle or two consecutive axles having an axle spread of less than 1. m. Outer Axle Distance: The distance between the centers of the outer wheels of a or 3 axle set. Double Axle: Two axles have their centers spaced no less than 1. m and no more than 1.7 m. A spacing of 1. m represents a practical lower bound for an axle spacing. Axles with a spacing greater than 1.7 m considered as two single axles. 9

20 Country Weight per Bearing Axle Weight per Drive Axle Lorry Axles Lorry 3 Axles Road Train Axles Road Train 5 Axles and + Articulated Vehicle 5 Axles and + Latvia Liechtenstein Lithuania Luxembourg Malta Moldova Netherlands Norway Poland Portugal Romania Russia Serbia Slovakia Slovenia Spain Sweden Switzerland Turkey Ukraine United Kingdom / Table Maximum Allowable Gross Vehicle Weights in Europe (Tons) (The Transport Manager s and Operator s Hand Book, 8). Triple Axle: Three axles with an outer axle distance ranging from 1.85 m but not exceeding 3.9 m. The outer axle distance is the distance between the centers of the first and third axles. Another condition was found while analyzing the spacing for triple axles, the spacing distance between the first and second or the second and third axles in the triple set must not exceeds 1.73 m. Otherwise the axle set is counted as a single and a double axles. Table 7 lists the geometric criteria. The axle group indicates the axle configuration. For example, an axle group classified as 13 consists of one steering axle, one double axle, and one triple axle. The second axle set is counted as the driving axle. 1

21 Axle configuration BK1(kN) BK(kN) BK3(kN) axle load non driven axle driving axle Tandem Inter axle distance < 1. m Axle load 1.<=inter axle distance<1.3 m <=inter axle distance<1.8 m <=inter axle distance<1.8 m and the driven axle has dual tire fitment and air suspension, or if the driven axle has dual tire and the load in each axle does not surpass 93. kn inter axle distance => 1.8m Tri axle The distance between outer axle<.6 m load The distance between outer axle=>.6 m kn =.1 Ton Table5 Maximum Allowable loads In Sweden for Different Axles (VV, 8) Number of axles Axle group Outer Axle Distance (OAD) distance range between 1st & nd Axles distance range between nd & 3rd Axles axles 13.6 m<=distance <=.86 m axles m<= Distance <= 3.9 m axles 7 axles 8 axles m<= Distance <= 3. m m<= Distance <= 3.1 m m<= Distance <= 3.6 m Table 6 Distances Criteria Used In BWIM System to Define Triple Axle. 11

22 3.. BWIM Data from Sweden Three sets of BWIM data were used in this study. The BWIM system recorded vehicles speeds, gross vehicles weight, number of axles, load of each axle, spacing between axle, and impact factor. Table 7 shows a summary of the data set analyzed in this study. The first data set is from E6 road northbound Löddekopinge station,. The total number of vehicles for one lane was 938. The data set included number of nonevents number which might be passenger cars. Only the 938 truck data was used in the analysis. The second data set was recorded for 9 days, from RV 5 at Gärdshyttan station, in 7. It included 671 vehicles on two lane road. The third data set was from E6 Löddekopinge station, in 9. The total number of vehicles was 159 for one lane. To analyze the data recorded by the BWIM system, the allowable loads on Swedish highways were identified. (VV,8) Table 5 shows a breakdown of allowable axle loads for different axle types, and spacing, for different road classifications. BK stands for the Bearing Class. All roads in this study were of the type BK1, the highest category for Sweden. Road and Location Date Number of Trucks E6 Northbound, Löddekopinge 938 RV 5, Gärdshyttan E6 Northbound, Löddekopinge Table 7 Location of Collected Data Sets. 1

23 . Methodology The complete series of tables and graphs from the three data sets are presented in the Appendices A, B, and C for the years, 7 and 9 respectively. Only selected figures will be discussed and repeated in this chapter. The first step used in this study was to break down the raw data according to the number of axles for each truck (two axles, three axles, etc.). Figures A1, B1, and C1 in the appendices show a pie chart of truck distribution by number of axles for each data set recorded in, 7, and 9 respectively. Figures A1 and C1 are for the same road way, and show an almost identical distribution of vehicles. Figure B1 shows great differences to figures A1 and C1. Data set B is from a location interior to Sweden, while data sets A and C are along the south eastern coast. The traffic patterns and road usage for these locations are clearly different..1. Axle load Spectrum (ALS) and Cumulative Frequency Distribution Graphs ALS data are often presented in an axle weight versus frequency plot, as shown in Figure 3 for the E6 Löddekopinge data. Figure shows the same data as a cumulative frequency plot. The first presentation is useful for showing the normal distribution, or in case of dual or triple axles, any potential bi-nodal distributions. The second presentation allows for an easier estimation of how many percent of the data exceed a certain axle weight. A line labeled BK1 is shown on both Figures 3 and. This represents the maximum allowable weight for this type of axle and road construction in Sweden. Any axles plotting to the right of this line exceed the allowable loads... Maximum Allowable Potential Vehicle Weight (MAPVW) Graphs The maximum Allowable Potential Vehicle Weight (MAPVW) graphs were developed to visualize the relationship between the measured Actual Gross Weight (AGW), and the maximum allowable potential vehicle weight. The MAPVW is defined as the sum of the maximum load allowed on each axle, as determined in Table 6. For example Figure 5 shows a 5 axle truck with an axle configuration of 1-1-3, would have a maximum allowable load of 6.7 kn. This is calculated by adding the allowable loads for each axle type: 98.1 kn for the first axle; 11.8 kn for the second driving axle; and 35.5 kn for the last triple axle with an outer axle distance greater or equal to.6 m. Figure 5 shows the MAPVW versus the GVW for the 5 axle truck from E6 Löddekopinge data set. The figure shows a diagonal line of equality. All data plotting on or below this line present trucks not exceeding the maximum allowable potential vehicle weight. All data plotting above present trucks exceeding the maximum allowable potential vehicle weight. The MAPVW graphs provide two visual checks for truck overloads, one based on the MAPVW concept, the other based on an overall weight limit. 13

24 Frequency % E6Löddekopinge, 938SteeringAxles BK1Limit.%Exceed thebk1limit Steering Axle Weight (kn) Figure 3 Axle Load Spectra for Steering Axle Cumulative Frequency % E6Löddekopinge, 938SteeringAxles BK1Limit.%Exceed thebk1limit Steering Axle Weight (kn) Figure Cumulative Distribution for Steering Axle 1

25 For example, for the truck with the configuration 9 out of 8 trucks exceeded the maximum allowable potential vehicle weight. Thus one or more axles must have exceeded the Swedish regulation. Since not all axles within the truck are likely to be loaded to the maximum allowable weight, the magnitude of the over load is such as to average out this effect. Figure 5 also shows a 6 ton limit line, which corresponds to the allowable load for the indicated number of axles as given in Table. Figure 6 shows the maximum allowable potential vehicle weight versus the actual gross weight for the 7 axle truck from E6 Löddekopinge data set. This figure also shows a 6 Ton limit line. This is the limit in the actual gross weight as shown in Table. All 7 axle truck types in this figure have some vehicles that exceeded the 6 ton limit..3. Comparisons between the and 9 Data For the E6 Löddekopinge station, two data sets were analyzed by the BWIM system. The first one was in, and the second one was in 9. This allowed for a comparison of axle loads over a 5 year period. Figure 7 shows a comparison of the axle load spectra for the steering axle for years and 9. Figures 7 and Figure 8 shows the axle load distributions for the steering axles are very similar for and 9. There is a slight rise in the top of the curve for the year 9, but it stills under the BK1 limit. Figure 7 shows that the ALS are very stable for time frame studied. Figure 8 shows the cumulative curves for the same location in both and 9 years. If we apply the same comparison for the rest of axle load spectra figures, we find a similar result as steering axle. Figures 9 and Figure 1 show there are however some minor differences. There is a slight rise in lightly loaded trucks, and a small increase in trucks having an outer axle distance less than.6 meter exceeding the BK1 limit. Overall the differences were considered minor and all data sets were used independently in the analysis. 15

26 8 7 E6Löddekopinge, AxlesTrucks AGW (kn) TonLimit MAPVW (kn) 113TruckOAD<.6;of 37FailByMAPVW 113TruckOAD=>.6;9 of8failbymapvw 111Truck;1of67Fails ByMAPVW 111Truck;1of185Fails ByMAPVW Figure 5 Maximum Allowable Potential Vehicle Weight for 5 Axles. AGW (kn) Truck;61of9FailbyGVW 113TruckOAD=>.6;8of31FailBy GVW 111TruckIAD<1.3;11of1Fail ByGVW 111TruckIAD=>1.3;1of69Fails ByMPAVWAnd7Of69FailByGVW 13111Truck;5Of7FailByGVW 6TonLimit E6Löddekopinge, 1177 AxlesTrucks MAPVW (kn) Figure 6 Maximum Allowable Potential Vehicle Weight for 7 Axles. 16

27 Frequency % E6Löddekopinge 9,159Axles,938Axles BK1Limit Steering Axle Weight (kn) Figure 7 Axle Load Spectra for Steering Axle in and 9 Cumulative Frequency % E6Löddekopinge 9,159Axles,938Axles BK1Limit Steering Axle Weight (kn) Figure 8 Cumulative Curves for Steering Axle in and 9 17

28 5 E6Löddekopinge 9,11Axles,3513Axles BK1Limit 5 Frequency % Triple Axle Weight (Outer Axle Distance =>.6) (kn) Figure 9 Axle Load Spectra for Triple Axle (OAD=>.6) in and 9. 5 E6Löddekopinge 9,3331Axles,59Axles BK1Limit 5 Frequency % Triple Axle Weight (Outer Axle Distance <.6) (kn) Figure 1 Axle Load Spectra For Triple Axle (OAD <.6) in 9 and. 18

29 5. Results The main objective of this study is to get a better understanding of truck loading on Swedish roads. This may lead us to know how does truck loading contributes to pavement deterioration. Four types of analysis were done to the three data sets used in this study. The first one based on the ALS and cumulative distribution figures. The second one analyze the MAPVW figures and the third one looks more specific to the triple axle sets, how does it affect the total truck loading compared with the dual axle sets. Finally the study attempts to characterize the type and magnitude of the over load. 5.1 ALS and Cumulative Frequency Distribution Analysis The ALS and Frequency Distribution figures presented in the appendices show the percentages of axles exceeding the BK1 limit for each axle type steering axle, driving axle, non driving single axle, non driving double axle and non driving triple axle. These values are shown in Table 8. By visual inspection it becomes apparent that many of the triple axles are over loaded. Only one of six data sets showed less than 8% over loads, and only one of the remaining 15 data sets had more than 8% over loads. To get more knowledge about the truck geometry and it s relation of these percent exceeding, the truck data were divided into separated groups based on the number of axles ( axles, 3 axles, 5 axle etc). Table 9 shows the percentage of trucks exceeding the BK1 limit for different truck types and axle configurations for the three data sets. The table also shows the value that can be calculated for the specific truck type and axle configuration. By applying the 8% criteria on the values in Table 9 it becomes clear that the most of the exceeding values are for the triple axle, especially when the outer axle distance is less than.6 m. Most of the exceeding values for triple axle were found in 5, 6, 7 and 8 axle trucks, but were very dominant in the 6 axle truck for the three data sets, 7 and 9. For the 6 and 7 axle data, two of the percent exceeding values are based on too small number of data. To be significant (6 axles, none driving triple axle, outer axle distance <.6m had 16 data points, and for the 7 axle, non driving triple axle, outer axle distance,.6m had data points).these values were put inside parenthesis to indicate a small data set. Other triple axles, in trucks other than 6 axles, also had exceeding values even though they are less than 8%. Thus the triple axle over loads seems to be a special issue for 6 axle trucks. The non driving double axle also has values exceeding the 8% value. They are found in all of the 7 axle group. There are some high percents in the 8 axle group as well, although only about 7% of the trucks exceed the allowable values. For 7 and 8 axle trucks the double axle over loads was a common problem. Since the triple axle had the highest percent exceeding, they were studied in greater detail. Table 1 shows the percent exceeding and the number of data for different truck types, and axle configuration for three data sets. Several of the 19

30 Axle type Axle spacing Year 7 9 Steering Axle.%.8% 1.% Driving Axle 5.8%.%.5% NDSA (a) 7.% 7.6% 8.% NDDA (b) 1.<=IAD< % 7.% 6.% NDDA (b) 1.3<=IAD< % 7.% 5.6% NDTA (c) OAD<.6 19.% 8.%.% NDTA (c) OAD=>.6 8.% 9.% 3.% (a) Non Driving Single Axle; (b) Non Driving Double Axle; (c) Non Driving Triple Axle. Table 8 % of Axles Exceeding the BK1 Limit for Each Axle Type. percentages are based on small data sets, and were not included in the conclusions. The triple axle with an outer axle distance <.6 m showed consistently high percent exceeding for 5, 6, 7, and 8 axle trucks. This axle configuration is a particular problem because there is so much over load in the influence zones beneath tire. The percent exceeding for the triple axles with an outer axle distance =>.6 m was somewhat lower. This difference seems artificial, since a small change in the spacing raises the allowable load by approximately 3 kn. For example, if we study the first tow data sets in the table by raising the allowable triple axle load for the outer axle distance <.6 m to 35.5 kn, the percent exceeding will drop from 1.% to 5%, conversely, if we lower the allowable triple axle load for the outer axle distance =>.6 to 6. kn the percent exceeding will raise from 9.1% to 3.6%. This shows that for the triple axle, the difference between outer axle distance =>.6 m and <.6 m may be arbitrary for this study.

31 Truck Type Axle Type % Percent Exceeding BK1 Allowable Triple Axle 7 9 Load (kn) for BK1 axle Steering Axle Driving Axle axle Steering Axle Driving Axle NDSA (a) axle Steering Axle Driving Axle NDDA (b) 1. <= IAD (c) < <= IAD (c) < axle Steering Axle Driving Axle NDSA (a) NDDA (b) 1. <= IAD (c) < <= IAD (c) < NDTA (d) OAD (e) < OAD (e) => (a) Non Driving Single Axle; (b) Non Driving Double Axle; (c) Inter Axle Distance; (d) Non Driving Triple Axle; (e) Outer Axle Distance; (f) Based on 16 Samples; (g) Based on Samples. Table 9 Percent of Axles Exceeding Loading Limits. 1

32 Table 9 (continued) 6 axle Steering Axle Driving Axle NDSA (a) NDDA (b) 1. <= IAD (c) < <= IAD (c) < NDTA (d) OAD (e) <.6. (1.5) (f) OAD (e) => axle Steering Axle Driving Axle NDSA (a) NDDA (b) 1. <= IAD (c) < <= IAD (c) < NDTA (d) OAD (e) <.6. (5.) (g) OAD (e) => axle Steering Axle Driving Axle NDSA (a) NDDA (b) 1. <= IAD (c) < <= IAD (c) < NDTA (d) OAD (e) < OAD (e) => (a) Non Driving Single Axle; (b) Non Driving Double Axle; (c) Inter Axle Distance; (d) Non Driving Triple Axle; (e) Outer Axle Distance; (f) Based on 16 Samples; (g) Based on Samples. Table 9 Percent of Axles Exceeding Loading Limits.

33 Truck Type Triple Axle Spacing Axle Group Year % Percent Exceeding # of Data Allowable Triple Axle Load (kn) 5 Axle OAD< % 11 of Axle OAD=> % 8 of Axle OAD< % 39 of Axle OAD<.6 13.% 1 of Axle OAD=> % 63 of Axle OAD=> % of Axle OAD< % of Axle OAD=> % 33 of Axle OAD< % 5 of Axle OAD=> % of Axle OAD=> % 7 of Axle OAD=> % 1 of Axle OAD=> % 1 of Axle OAD< % 1 of 6. 8 Axle OAD< % 1 of Axle OAD< % 1 of Axle OAD< % of 1 6. Table 1 % of Axles Exceeding MAPVW for Triple Axles. 3

34 5. MAPVW and GVW Analysis A second series of analysis were performed based on the gross vehicle weight (GVW), and the maximum allowable potential vehicle weight (MAPVW). The MAPVW was defined as the sum of the maximum load allowed for each axle or axle group, based on the axle geometry as listed in the Swedish regulations, as shown in Table 6. The MAPVW and GVW figures presented in the appendices show the percentage of trucks exceeding the limits on the Swedish highways. These values were tabulated and shown in Table 11 for, Table 1 for 7 and Table 13 for 9. Analysis of the data based on the GVW criteria showed that on the order of 19% of the 7 axle or longer trucks are over loaded. For trucks with 6 axle or fewer axles, the GVW is not a major concern. Only 3.% exceed the allowable limit. For the MAPVW limit only one value exceeded the 8% limit. For this axles data set the total number of samples was only 19, and it was not considered as representative. This can be explained by declaring that the MAPVW limit assuming the full loading of each axle of the axle group, where this rarely occurs. So the under loaded axles compensate the over loaded axles in final adding.

35 Year MAPVW GVW Allow. Allow. Truck Type AxleGroup Axle Spacing Limit (kn) % Exce. #ofdata Limit (kn) % Exce. #ofdata Axles of of16 3Axles of of of of665 Axles 11 IAD< of IAD=> of of of Axles 113 OAD< of OAD=> of IAD=> of of Axles 13 OAD=> of IAD< of of of of119 7Axles 1 IAD< of9 113 OAD=> of IAD< of1 111 IAD=> of of of7 IAD<1.3, 8Axles 113 OAD< of 13 IAD< of 1111 IAD< of5 9Axles 133 OAD< of1 IAD=>1.3, 133 OAD< of Table 11 % of Axles Exceeding MAPVW & GVW, Löddekopinge 5

36 Year 7 MAPVW GVW Allow. Allow. Truck Type AxleGroup Axle Spacing Limit (kn) % Exce. #ofdata Limit (kn) % Exce. #ofdata Axles of875 3Axles of of of3 Axles 11 IAD=> of of Axles 113 OAD=> of Axles 1111 IAD< Axles 111 IAD< of6 1 IAD=> of311 71of 111 IAD=> Axles 13 IAD=> of69 OAD=>.6 11 IAD< of 1311 IAD=> of5 9Axles 133 IAD< OAD<.6 Table 1 % of Axles Exceeding MAPVW & GVW, Gärdshyttan 7 6

37 Year 9 MAPVW GVW Truck Type AxleGroup Axle Spacing Allow. Limit (kn) % Exce. #ofdata Axle of 3Axle of Axle 11 IAD< of6 11 IAD=> of of393 5Axle 113 OAD< of OAD=> of3189 6Axle 13 OAD< of Axle 113 OAD< IAD< IAD< Axle 13 IAD< OAD< OAD< IAD< Axle 1311 IAD< IAD=> OAD< IAD< IAD=> Axle Axle 16 IAD=> Axle 17 IAD=> Axle 1311 IAD=> Axle 3115 IAD=> OAD<.6 Allow. Limit (kn) Table 13 % of Axles Exceeding MAPVW & GVW, Löddekopinge 9 % Exce #of Data 69of 15of 18 of of of36 69of of of of19 5of of of of of of of

38 5.3 Analysis of Triple Axle Based on the values from Table 11, 1 and 13, few values were picked to study the sensitivity of the axle spacing for the triple axle versus a combination of a single and double axle. Table 1 shows the effect of changing from a triple axle to a more widely spaced combination of dual axle and a single axle. The analysis shows that most of the exceeding values were eliminated by changing the spacing, thus the analysis is very sensitive to axle spacing. year Group Arrangement # of Exceeding Data for # of Exceeding Data for # of Exceeding Data for <=IAD< <IAD<= (a) of 37 of 37 of (a) 39 of 75 of 75 of (b) 9 of 8 9 of 8 of (b) 19 of of of (b) 9 of 3189 of of 3189 (a) Outer Axle Distance<.6; (b) Outer Axle distance =>.6. Table 1 Comparison between & Group Arrangement 8

39 5. Characterization of the Magnitude of the Over Loads The methodology used to describe the magnitude of the over loads is to show the relation between the number of the exceeding vehicles on the y-axis versus the exceedance weight above the allowable limit on the x-axis. This is done for the three data sets, 7 and 9 in Figures A.6, B.5 and C.7. Each figure represents the over load magnitude for steering axle, driving axle, non driving single axle, non driving double axle and non driving triple axle. Figure 11 shows the magnitude of the over loading for the year. By observing figures for the three data sets, we can notice that the most of the overloads are in the range of 5 to kn. The driving axle in and 9 has the highest number of data count followed by the non driving single axle. But in 7 the reverse condition appears, and the non driving single axle has the highest data count while the driving axle follows. 3 5 E6Löddekopinge 938Trucks SteeringAxle(6of938) DrivingAxle(819of161) DataCount 15 1 NonDrivingSingleAxle(3 of3537) NonDrivingDoubleAxle 1.<IAD<=1.3(57of66) NonDrivingDoubleAxle 1.3<IAD<=1.8(91of3) NonDrivingTripleAxle OAD=>.6(9of3513) NonDrivingTripleAxle OAD<.6(118of59) ExceedanceWeight(kN) Figure 11 Magnitude of the Over Loading, Löddekopinge. 9

40 The high over load in the driving axles observed in the and 9 in the E6 Löddekopinge road location, data may show that the vehicles are loaded incorrectly. Truckers prefer having high loads over the driving axle for better traction, and may thus over load this axle set. This may be especially true for truckers from Denmark, and Norway who are used to driving with lower allowable gross vehicle weight. 3

41 6. Conclusions and Recommendations Super single tire are used in Europe since the early 198 s, but not widely used in USA and South Africa. When the axle load for the super single tire is greater than 73 kg the impact on pavement is larger compared to the corresponding related to the dual tires. The Tire Configuration Factor (TCF) number might be a good tool to the manufacturing of new tires to reduce future pavement wear. BWIM technology is a useful tool to estimate the real traffic volume on roads. This system is used in this study in three stations, E6, Löddekopinge in, Rv 5, Gärdyshattan in 7 and E6, Löddekopinge in 9. These data were analyzed for over loads based on two tables of regulations for loading limits are used in this study. Sweden has the highest allowable vehicles load limit among European countries. Two approaches are used to visualize the data recorded by BWIM system, the Axle Load Spectrum (ALS) and the Cumulative Frequency Distribution method. The second approach allows for a better quantification of the overloads. The maximum allowable potential vehicle weight (MAPVW) is introduced as a visual tool, to characterize the overloads for different vehicle configurations. The method allows for comparing the GVW with the MAPVW as well as the allowable vehicle weight for different truck configurations. The analysis shows that the triple axle is frequently overloaded in 5 and 6 axles trucks, especially when the outer axle distance is less than.6 m. For longer 7 and 8 axle trucks the double axles are frequently over loaded. Analysis of the type and quantity of over loads shows that the typical overloads are between and kn. The highest overloads tend to be on the driving axle, which may be attributed to incorrect loading procedures. The use of BWIM data has been very helpful in identifying certain problem areas in truck loading on Swedish highways. These observations now need to be linked to a damage model to predict actual deterioration and cost to Swedish society. 31

42 3

43 References American Trucking Association (ATA), "New Generation Wide Base Single Tire", Revision 9-December 1, 7. PP 3. Available: /.../ Wide-Base Tires/Wide_Base_Summary-v9-ATA-whitepaper.pdf, Accessed: 5 June, 9. Anonymous, "Super Singles Hit the Road-hard, Modern Tire Dealer magazine AUG-1993.PP. Available: http: // coms/ summary _86-176_ITM; Accessed: June, 9. European Co-operation in the Field of Scientific and Technical Research COST 33, "Effects of Wide Single Tires and Dual Tires, Task group 3, Final Report, Version 9, November 1. PP 17. Federal Highway Administration, Washington, D.C., "International Workshop on the Use of Wide-Base Tires" October 5-6, 7. Available : www. arc.unr.edu/workshops/wide - Base Tires/International_ Workshop _ on _ the_use_of_widebase_tires-minutes-final.pdf; Accessed: 5 June, 9. Kalin, Jan, "Practical Implementation of Nothing on The Road Bridge Weigh In Motion System", Slovenian National Building and Civil Engineering Institute. Available :htt :// Conference Proceedings / ISHVWD_9_6/docs/pdfs/session%7/s7-%115.pdf; Accessed: 1 April, 1. Khavassefat, Parisa, Evaluation of Damage Induced by Vehicles on Pavements, Masters Degree Project, PP 3, KTH-Royal Institute of Technology, Stockholm, Sweden, 9. Lowe, David, "The Transport Manager s and Operator s Hand Book, 8", Kogan Page Limited, United Kingdom, PP 761, Chapter 1, 38 th edition. Available: British Library. Available: http :// ebooklink. Net /g/ download / 7957X/The%Transport%Manager%7s%&%Operator%7s %Handbook%8/; Accessed: 7 June, 9. Ontario Ministry of Transportation, "Use Of New Technology Single Wide-Base Tires: Impact on Pavements", Engineering Standard Branch, Executive Office November 1, 3.PP. Available: www. comt.ca /english/ programs/ trucking/on%wide%tire%study.pdf; Accessed: 5 June, 9. 33

44 Västmanlands läns författningssamling, " Traficförordningen (1998:176) ", 19FS 8:7. Available: DFC-AEA1-3E9FD968A//19FS87.doc; Accessed: 3 June, 9. Yamaguchi, E., Matsuki, Y., Naito, Y., "Traffic Flow of Heavy Trucks in Northern Kyushu, Japan". Kyushu Institute of Technology, Kitakyushushi, Fukuoka Ken, Japan. Available : lwk/ UploadFiles _8/88/ pdf. Accessed: 5 August, 9. 3

45 Appendix A The data in this appendix were collected using BWIM system starting from the 7 th of May to nd of June in, on road E6 Löddekopinge. The data set consists of 585 measurements collected over a week period, were nonevent numbers, presumably passenger cars, and 938 valid truck data are analyzed in this study. Figure A.1 shows a pie chart of the different truck types. FigureA. shows a plot of the total number of vehicles passing in each day of the week. Figure A.3 shows the number of passing vehicles in one day (Wednesday) for each hour during the day. Figures A., 6, 8, 1, 1, 1, 16 show the Axle Load Spectra for different axles. Figures A.5, 7, 9, 11, 13, 15, 17 show the Cumulative distribution for BK1 limits for Steering, Driving, and non driving axles whether single, double or triple. Figures A18, 19,, 1,, 3,, 5 represent the Maximum Allowable Potential Vehicle Weight (MAPVW) and Gross Vehicle Weight limits. Figure A6 shows the magnitude of exceedance for different axle categories. 7Axle 13% 8Axle % 9Axle % Axle 16% 6Axle 11% 3Axle 9% Axle 8% 5Axle 1% Figure A.1 Distribution By Number of Axles, E6 Löddekopinge,, 938 Trucks. 35

46 Truck Count Figure A. Total Number of Vehicles in Each Day. 5 Truck Count : 1: : 3: : 5: 6: 7: 8: 9: 1: 11: 1: 13: 1: 15: 16: 17: 18: 19: : 1: : 3: Figure A.3 Hourly Truck Count on Wednesday, /6/. 36

47 Frequency % E6Löddekopinge, 938SteeringAxles BK1Limit.%Exceed thebk1limit Steering Axle Weight (kn) Figure A. Axle Load Spectra for Steering Axle Weight (kn) Cumulative Frequency % E6Löddekopinge, 938SteeringAxles BK1Limit.%Exceed thebk1limit Steering Axle Weight (kn) Figure A.5 Cumulative Distribution for Steering Axle Weight (kn) 37

48 Frequency % BK1Limit Driving Axle Weight (kn) E6Löddekopinge, 161DrivingAxles 5.8%Exceed thebk1limit Figure A.6 Axle Load Spectra for Driving Axle Weight (kn) Cumulative Frequency % BK1Limit 5.8%Exceed thebk1limit E6Löddekopinge, 161DrivingAxles Driving Axle Weight (kn) Figure A.7 Cumulative Distribution for Driving Axle Weight (kn) 38

49 Frequency % BK1Limit E6Löddekopinge, 3537NonDrivingSingleAxles 7. %Exceed thebk1limit Non Driving Single Axle Weight (kn) Figure A.8 Axle Load Spectra for Non Driving Single Axle Weight (kn) Cumulative Frequency % BK1Limit 7.%Exceed thebk1limit E6 Löddekopinge, 3537NonDrivingSingleAxles Non Driving Single Axle Weight (kn) Figure A.9 Cumulative Distribution for Non Driving Single Axle Weight (kn) 39