Damage Factor Assessment through Measuring Direct Response of Asphalt Pavement due to Heavy Loading

Damage Factor Assessment through Measuring Direct Response of Asphalt Pavement due to Heavy Loading A arif Hamad, MSc. Candidate University of Calgary Department of Civil Engineering-Transportation Specialization 2500 University Drive N.W. Calgary, Alberta, Canada T2N 1N4 Phone: (403) 220-5970 Fax: (403) 220-7026 Email: ahamad@ucalgary.ca Lynne Cowe Falls, Ph.D. Assistant Professor University of Calgary Department of Civil Engineering 2500 University Drive N.W. Calgary, Alberta, Canada T2N 1N4 Phone: (403) 220-5505 Fax: (403) 220-7026 Email: lcowefal@ucalgary.ca Word count = 4745+10*200 = 6,745

Hamad, A. and Cowe Falls, L. ABSTRACT Road authorities in resource based economies are frequently challenged by the demands of heavy equipment operators requiring access to remote sites during sensitive spring thaw conditions. Access to these sites is rarely along high volume, structurally engineered pavements and many secondary pavements suffer premature deterioration as a result. To overcome this, agencies impose restrictions to overloads that can result in substantial costs to operators as equipment is broken down and re-assembled onto multiple flatbeds for transportation over provincial roads. This paper presents the results of a field study of the loads imposed by heavy oilfield cranes (with hydraulic suspensions and super single tires) on thin membrane asphalt pavements in Alberta. Three 50m test road sections (thin asphalt wearing course, bituminous surface treatment and granular surface) were built and instrumented for strain at the bottom of the asphalt layer, surface deflection, and subgrade pressures. Temperature and moisture profiles are also measured. Field testing involved controlled speed experiments of standard axle configurations and heavy (12,000kg) axle vehicles with and without hydraulic suspensions. The results will be used to evaluate the impact of overloading by these vehicles and will become part of a long term study to evaluate pavement performance and develop load equivalency factors.

Hamad, A. and Cowe Falls, L. 1 INTRODUCTION Road authorities in resource based economies are frequently challenged by the demands of heavy equipment operators requiring access to remote sites during sensitive spring thaw conditions. Access to these sites is frequently along low volume, structurally under-designed pavements and consequently many of these pavements suffer premature deterioration as a result of vehicle overloads. To overcome this, agencies impose restrictions that are based upon Load Equivalency Factors (LEF) which was initially developed at the AASHTO test road and has become the basis not only for overload permitting, but also for pavement design. Using LEFs, all vehicles using the road during the design period can be equated to a number of standard axles and the pavement structural thickness determined. LEFs reflect the expected damage imposed on the road by the vehicle, relative to a standard 80kN (18,000 lb) single axle (referred to as the Equivalent Single Axle Load (ESAL)). Traditionally, pavement engineers have designed the road for standard and/or common vehicle configurations such as those listed in Table 1. Recently, vehicle manufacturers have begun to design the vehicle to minimize pavement design through the use of larger tires, lower tire pressures, central tire inflation, hydraulic suspension systems and new axle configurations. These changes have altered the vehicle dynamics and a lack of understanding of the impact of these new mechanical systems (referred to as super singles) on pavement structural behavior means that existing Load Equivalency Factors may not be accurate for these vehicles. Extrapolation of the current LEFs for overload permitting has resulted in increased costs to the operators as agencies require owners of these new vehicles to break-down the vehicle onto multiple flatbeds and/or dollies to meet legal load limits. An indication of the costs to the operators of this break-down and reassembly was illustrated recently when a Canadian provincial highway agency allowed a heavy crane (in excess of 48,000kg on four axles) to travel between two sites during the winter months without multiple additional dollies. The operator had to move the crane approximately every two weeks for a period of four months and the cost savings were in excess of $200,000 CDN. Incidentally, many of these super single axle configurations are built and used extensively in Europe with little reported impact on overall pavement performance. The argument against super single usage in Canada has centered around the fact that many Canadian roads are thin membrane structures that cannot support heavy loadings. Figure 1 shows sample of these cranes. On the basis of these cost savings an initiative was begun in Alberta to investigate the impact of the new super single vehicles on typical pavements found throughout the province. This paper presents the results of a field study of the loads imposed by heavy cranes (with hydraulic suspensions and super single tires) on thin membrane asphalt pavements in Alberta. Three 50m test road sections (thin asphalt wearing course, bituminous surface treatment and granular surface) were built and instrumented for strain at the bottom of the asphalt layer, surface deflection, and subgrade pressures. Temperature and moisture profiles are also included in the test sections to capture seasonal differences. The long-term plan for the test road is to return to the site quarterly over the next five years to record and assess pavement deterioration and to develop load equivalency factors for the heavy cranes. BACKGROUND The use of instrumented test roads is not a new technique in the evaluation of the pavement response and in fact, similar studies to the one underway have been done in Alberta in the past. Christison at al. 1978 studied flexible pavement response under various axle loads and loading configurations (single axle dual tires loads ranging from 56 to 117 kn, single axle single tire load ranging from 9 to 53 kn, tandem axle dual tires load ranging form 95 to 334 kn, and single axle single tire load ranging from 71 to 295 kn on self-propelled earthmovers) on two full-depth asphalt pavements structures. The sites were instrumented with strain gauges to measure strains at the bottom of the asphalt concrete layer, pressure cells to measure stress at the top of the subgrade, and deflectometers to measure total deflection at the pavement surface. The test pavement structures were full-depth asphalt pavements that were standard in Alberta at the time of the test with 200 mm (7.7 in.) and 280 mm (11 in.) full depth asphalt concrete, respectively, constructed on a 1.2 to 1.5 m common borrow subgrade made of highly plastic soil. The study concluded that for the standard 80 kn axle load, the interfacial strains (at the bottom of the asphalt concrete layer) and surface deflections increase with the increase of the pavement age and pavement temperature and decrease with the increase of the vehicle velocity. Other findings were the agreement of the results for the LEFs values from this test with performance based factors from the AASHTO test road and theoretical values derived by other researchers. The study also indicated that the damage from one application of a 133.4 kn axle load is equivalent to the damage of one application of the standard load based on the predicted LEFs. Similarly, for the heavy wheel loads of selfpropelled earth movers used in the study, LEFs developed from deflection readings were higher in magnitude than

Hamad, A. and Cowe Falls, L. 2 those found by using the flexural distress concept. A comparison between similar axle loads on the single axle load of the self propelled earthmover and on a tandem axle showed that the LEFs on the tandem axle is less damaging by factors ranging from six times to 20-30 times depending on the distress criteria (six times for the interfacial strain and 20-30 times for the surface deflection). Christison 1990 investigated pavement response under heavy multi-axle crane units, similar to those used in the current study, to predict load equivalency factors. The test was conducted on two pavements: a 135 mm asphalt concrete pavement (ACP) on a 170 mm cement treated base and a 135 mm ACP on a 250 mm granular base layer, respectively. Subgrade conditions for both sites were similar with moisture contents ranging from 20 25% on medium to high plasticity glacial till. Both test sections were instrumented with deflectometers for surface deflection and strain gauges at the bottom of the ACP layer for measuring tensile strain. Results from this study indicated that LEFs for these cranes varied from 5.2 to 26.8 depending on test section structure, crane type, traveling speed, and basis for calculation (strain based or deflection based). Sebaaly et al. 2003 conducted a study to evaluate the effect of agricultural equipment on the response of low-volume roads in South Dakota. Four different agricultural vehicles were selected for this study and the tests were done on two different types of roads: a gravel pavement with a 100mm crushed aggregate base (CAB) and a blotter pavement with a 212mm CAB. Both pavements are on a silty subgrade. The roads were instrumented with pressure cells in the subgrade and the base layers and with deflection gauges for surface deflection measurements. The test were conducted quarterly (corresponding to winter, spring, summer, and fall) in order to evaluate the effect of environmental conditions on the response of the pavements. The research showed that most of the vehicles tested imposed more damage to the roads than the standard 80 kn axle load especially when they are fully loaded or exceeding their legal limit. The LEFs values found from this study ranged from 0.4 to 25.1 depending on the test vehicle type, pavement type, amount of loading (whether it is empty, fully loaded or exceeding the legal limit), and testing season. It was found that LEFs decrease as the testing season changed in the following order: summer (highest), fall, spring, and winter (lowest). Also found in this study was that the LEFs decreased with the increase of the base course layer thickness. OBJECTIVE OF THE STUDY The objective of this study is to study the impact of heavy axle loads on thin membrane flexible pavement structures. Specifically, the study has several long-term objectives including: 1. Update historical data Christison data on mag tape Load Equivalency Factors Dynamic loading Seasonal factors 2. Develop new performance models 3. Evaluate new technologies and equipment Super singles CTI 4. Update/validate existing road user regulations The study and assessment of the effect of these loads are achieved through the construction and instrumentation of a test road and a finite element analysis. The test road is funded by a consortium referred to as the Alberta Road Research Initiative, comprised of industry, academia and government. The test road was built in the County of Leduc - Nisku Industrial Business Park near to Edmonton International Airport. The Nisku Industrial Business Park is the home of many oilfield servicing companies, serving the Northern Alberta oil fields and consequently, is resident to many of the large vehicles of interest. Site selection criteria for the test road included geometric features that enable the testing to be both safe and practical due to the vehicle sizes and speeds involved in the program and to minimize disruption to other businesses in the area.

Hamad, A. and Cowe Falls, L. 3 DESIGN OF THE TEST ROAD The test road was designed to capture the Alberta provincial highway and county road standards with construction of three 150 meter sections with a road surface width of 9.0 meters. Three pavement structures: Hot Mix Asphalt Concrete (HMA), Cold Mix Asphalt Concrete (CMA) and Granular Base Course (GBC) were constructed on a tangent section of 8th St. The subgrade soil is a heavy plastic clay and the road prism has been constructed using a silty clay borrow material. All of the pavements are constructed on a standard 150 mm prepared subgrade and the pavement structure for each section is summarized in Table 2 and shown graphically in Figure 2. INSTRUMENTATION DESIGN The test road instrumentation was designed to capture a variety of axle widths (from 1.2 m to 2.8 m) and tire sizes (up to 390 mm wide). Instrumentation was also designed to capture a large amount of data at high speeds (up to 60 km/hr) and for redundancy / survivability over the duration of the test. Each array is repeated in the inner and outer wheel path and is also duplicated at two of the sites (HMA and CMA). The instrumentation layout in plan view and cross section are shown in Figure 2. The typical array in each of the HMA and CMA sections consists of two pressure cells, six strain transducers, and temperature and moisture profiles. The GBC section does not have strain transducers, however all other elements are measured. The instrumentation is described in the following sections. Pressure Cells Pressure cells measuring vertical pressure on the subgrade were installed in the inner and outer wheel paths, 300 mm below the top of the subgrade (which is 150mm below the prepared subgrade layer). The HMA and CMA sections have four pressure cells each (two in each wheel path), while the GBC section has three pressure cells (two placed in the inner and outer wheel path and one in the centre lane). The centre lane GBC pressure cell was placed to measure the combined effect of loading in the two wheel paths and to capture the intersection of the pressure cone beneath each wheel. This arrangement was used only in the GBC section as there is no bound wearing course and it is possible that it could be the worst case for the vertical pressure development. Strain Gauges Strain gauges at the interface between the asphalt concrete layer and the granular base course layer were placed for the HMA and CMA sections. The distribution for the strain gauges for the HMA and CMA sections was designed to suit the characteristics of a range of vehicle dimensions the tested vehicles. For redundancy, should any of the gauges deteriorate over the test duration, the arrays were duplicated at each site. Each array consists of six strain gauges, as seen in Figure 2, with the gauges distributed over three horizontal lines that have different transversal spacing forming a trapezoidal shape. The different widths allow for a variety of vehicle axle sizes and can measure tracking of multiple axles and/or dollies as they traverse the site. In each section two orientation were used for the strain gauges (along and perpendicular to the travel direction) in order to capture both the longitudinal and transversal strain at the asphalt interface. Linear Strain Transducers (LSCTs) Measuring vertical strain or deflection is problematic for pavement researchers as most installations require placement of a deflectometer in the pavement after construction. Placement of the deflectometers usually requires drilling of the pavement which was felt to be inappropriate for this test road as the introduction of a corehole on a thin membrane pavement introduces a discontinuity and a site for crack propagation. This was overcome by the construction of a surface deflectometers beam, similar in principle to the deflectometer beam attached to a Falling Weight Deflectometer. Four Linear Strain Transducers (LSCTs) were attached to a 3 m steel beam that was placed on the centrelane of the pavement and was used to capture the surface deflection due to loading as the vehicles crept over the site. Vertical clearance over the beam was very tight for the large crane in the study and it was decided to not use the deflectometers beam during the higher speed tests. Data Acquisition System (DAS) The Data Acquisition System (DAS) was comprised of a 32 channel data acquisition unit and dedicated computer capable of capturing and recording the dynamic readings with 500 Hz frequency. This enabled differentiation of each axle of the vehicles even at 60 km/hr. The system was duplicated and synchronized at the HMA and CMA sections; however, one of the sites had to run on AC power which produced some noise in the signal.

Hamad, A. and Cowe Falls, L. 4 Environmental Conditions Measurements Devices Temperature and moisture profiles were installed at all sections to measure the surrounding environmental conditions during the tests. Thermocouples wires were fixed on stakes at six levels (100, 200, 300, 400, 600, and 1000mm) and placed strategically on each site to capture the temperature profile throughout the pavement layers. Also, moisture profiles are measured at the same levels for the temperature around the strain gauge arrays for all sections to measure in-situ moisture in the pavement layers. The moisture is measured by insertion of a probe that has multiple sensors into an installed sleeve at the time of measurement with a stand-alone portable meter. Temperature and moisture readings were not collected dynamically during the test, as it was felt they would not change in the relatively short time that the test is completed per cycle. CONSTRUCTION OF THE TEST ROAD The construction of the test road was completed by the end of May 2005. The instrumentation work was implemented in between and after the construction stages. The placement of the pressure cells and the temperature thermocouples was done prior to placement of the granular base course and the strain gauges and thermocouples were placed after. The final step of the construction was done by placing the asphalt concrete layers. The remaining items for the instrumentation were the installation of the moisture sleeves and the preliminary test of all the devices to make sure that they were working properly prior to the vehicle testing. Extreme care was taken by the University of Calgary team to ensure proper and accurate installation of devices and a fourth order survey was completed for location referencing of the devices. The work of Timm D.H et al. 2004 served as a general guide for the instrumentation installation procedures. Under the supervision of the University of Calgary lead investigator, the same care was taken by the construction crew to ensure that no direct loading was done by the construction equipment on the instrumentation array in order not to minimize possible damage. Checks for the devices were done continuously during and after all stages that to check for survivability. TESTING PROGRAM The test plan that is decided for this test road will help in studying the effect of the heavy loading compared to the standard axle loading in conjunction with the following variables: Changes in the seasonal conditions. Different pavement structures. Different traveling speeds. The field testing plan includes five series of tests that will be performed around the year. These five series represent late spring, summer, fall, winter, and early spring seasons. The late spring series was performed in June 2005and the rest will be followed for the rest of 2005 and 2006. Each series of tests will include testing heavy cranes with different axle configurations and different speeds. A standard 80 kn truck will also be tested in each series for a comparison purposes. The testing plan for each series of tests is presented in Table 3. In the spring 2005 test, an 8 wheel tandem axle configuration was included in the test as this is most common, legal axle configuration in the province. Subsequent tests may not include this vehicle and other configurations, (in particular, heavier cranes and/or more axle cranes) may be substituted. Also, the number of passes may not be included in subsequent tests as the initial analysis may indicate only a few speeds may be necessary to characterize pavement response. As can be seen in Table 3, the equipment passed over the test site at selected varying speeds from creep to 60 km/hr at intervals of 5 kilometers per hour, initially and then 10 km/hr. Each speed was repeated five times for statistical significance and the vehicles traversed the section in alternating directions (south bound, north bound, south bound, north bound, south bound) to reduce any impact of the pavement transitions on the results. Vehicle dynamics as a result of discontinuities at the edge of the pavement and at the HMA/CMA transition may skew the results and traveling in two directions will capture this and allow for statistical averaging. Before commencement of the test run, the standard vehicle passed over the site five times to calibrate the sensors and establish a baseline. In between each speed interval, the standard vehicle again passed over the site again to reestablish the datum and to track any deterioration that was occurring as a result of the test. At the end of the tests, the standard vehicle repeated the initial 60 km/hr pass as a further calibration mechanism. Four vehicles were used in the initial test: a standard 80 kn single axle picker truck, an 8-wheel tandem 17,000kg axle, a 48,000 kg Speirings Crane operating on 8 wheels/ four axles and the same vehicle operating on 14 wheels/ seven axles. Each vehicle was weighed on site by Alberta Infrastructure and Transportation Inspection Services using a portable weigh scale system and the standard 80 kn and 8-wheel tandem vehicles were checked again at a nearby permanent weigh scale. Table 4 summarizes the axle weights for each vehicle.

Hamad, A. and Cowe Falls, L. 5 PRELIMINARY FIELD TEST RESULTS This section presents preliminary results of some of the data that are collected from the field test that is performed on June 26, 2005. There was some noise associated with some of the data collected and these data were excluded and will be presented in the future. As mentioned in the testing program, four types of vehicles were tested for different speeds ranging from very low speed (creep speed) until a 60 km/hr. the results presented below are related to three types of vehicles and these are, Standard 18 kip truck, heavy crane with four axle configuration, and the same heavy crane with seven axle configuration, referred to as ST18, CR01, and CR02 respectively, and for a speed of 60 km/hr. Strains at the Interface between the Asphalt Layer and the Granular Base Course Longitudinal Strains Table 5 presents a summary of the mean, plus one standard deviation, longitudinal strains under both wheel paths for the HMA and CMA sections. As expected, all vehicles generated a tensile strain in the longitudinal direction (i.e., in the direction of travel). The strains recorded for the CMA section were found to be considerably higher than for the HMA section due to the difference in the pavement strength (material and thickness). For example, for both types of cranes the strains were around 65% higher on the CMA section compared to the HMA section (the strains recorded for the ST18 truck however showed very high magnitudes for the CMA section compared to the HMA section.) The CMA results however, could be highly affected by the noise on the signal and therefore need further review. The maximum amount of total strain was recorded under the wheel paths as expected, and then the strain amount decreases gradually away from the wheel paths as recorded by the internal lines of strain gauges. For both types of cranes CR01 and CR02, the percentage of decrease in the HMA section was higher than the CMA section. In general a decrease of around 85% was noticed for the HMA section compared to around 65% for the CMA section. Figure 3 shows a typical strain profile under the Speiring crane (four axle configuration) and for three strain gauges, the highest peaks are for the strain gauge that is located under the wheel path. The comparison of strains between the cranes on different axles configuration however, showed slight difference (strains for the CR01 were around 10% higher than for the CR02 crane). These strains and after several tests will be used to asses the damage that these cranes cause to the pavement compared to the standard axle load through predicting a series of Load Equivalency Factors (LEFs) for each type of loading and according to the vehicle s other characteristics. Transverse strains The transverse strains that are generated on the direction perpendicular to the direction of travel indicate that the pavement is in both tension and compression depending on location in the lane. The data showed that for all the vehicles and for all the sections, tensile strains were generated under the wheel paths however, the area between the two axles was frequently in compression for both types of the cranes. The pattern was not clear however for the standard 18 kip truck. Table 6 shows this distribution for the transverse strain for the HMA section. Because of the noise in the signal for the CMA section no analysis is available at this time. It was found also that the magnitude for the transverse tensile strain is less than the longitudinal tensile strain at a percent that depends mainly on the vehicle type and to a lesser degree to the axle sequence. In general the CR01 crane generated transverse strains that were approx. 75-80 % of the value of the longitudinal strains, while for CR02 the difference is in the 55-75% range. This is probably due to the distribution of the vehicle weight onto more axles in the CR02 crane. This effect for the increased number of axles was not clear in the case for the longitudinal strain (differences were approx. 10 %) which may be related to the effect of the direction of travel on the amount of strains when the pavement is subject under both the axle weight and the dynamic action of the wheel movement. It was also noted that the amount of the tensile transverse strain under the outer wheel path is higher than that under the inner wheel path. Again here the vehicle type is a controlling factor. For CR01 strains under the outer wheel path were found to be 35-50% higher than under the inner wheel path, while for CR02 crane the increase was less at 15-30 %. The range in both cases is related to the axle sequence. This increase under the outer wheel path could be highly related to the edge effect of the pavement where less support is available under the outer wheel compared to the inner wheel. This may suggest in the future providing extra width from the outer wheel path to the edge or stronger shoulder structure in case a minimizing to these strains is required. The compressive strains that were generated between the outer and inner wheel paths however were more uniform in their distribution. Two things were consistent; first, compressive strain increases with the distance from the wheel path and second, compressive strains that are generated by CR01 crane are higher than those by CR02 and both are higher than ST18 vehicle.

Hamad, A. and Cowe Falls, L. 6 Vertical pressure in the subgrade layer Results for the pressure cells readings in the subgrade layer for the HMA section for three types of vehicles for the 60 km/hr speed are presented in Table 5. It is clear that the vertical stresses are higher for both cranes than for the standard truck. Vertical stresses measured under CR01 crane were 90-155% higher than for the ST18 truck, while for the CR02 crane, vertical stresses were 35-110% higher than for the ST18 depending on the axle number. Between the two cranes, vertical stresses measured under CR01 crane were 40-55% higher than for CR02 for the corresponded axle. This could be attributed to the load distribution change when adding three axles to carry the same load. Further analysis will be done for the other speeds and other pavement section to show the complete effect of the heavy cranes on the vertical stresses measured in the subgrade layer. Figure 4 shows typical vertical stress profile under the axle loading of Speiring crane (with the four axles configuration). CONCLUSIONS AND SUGGESTED FUTURE WORK The results that are presented in this paper represent only the first run of a long series; in addition to that the data that are collected are only three weeks old. Based on that what we presented here is a very preliminary result and therefore no clear conclusions could be drawn since no detailed analysis was done and therefore all the conclusions will be made when the complete data will be collected and analyzed. As for the future works, this test road could be adopted to perform more studies making use of the instrumentation design that made. Below are some ideas new studies. Studying the edge effect on the pavement response, making use of the strain gauges array design. Including the tire characteristics to the variables in the testing program such as tire size, tire type, tire pressure etc. Investigating the effect of any future work to the test road by comparing current results with the new ones.

Hamad, A. and Cowe Falls, L. 7 REFERENCES 1. Christison, J.T., Anderson, K.O., and Shields, B.P. In Situ Measurements of Strains and Deflections in a Full- Depth Asphaltic Concrete Pavement. Proceeding-Association of Asphalt Pavement Technologies, Vol.47, Minneapolis, Minn., 1978, pp.398-433. 2. Christison, J.T. Predicted Load Equivalency Factors for Two Multi-Axle Cranes. A report for Motor Transport Services, Alberta Transportation, 1990. 3. Sebaaly, P.E., Siddharthan, R., and Huft, D. Impact of Heavy Vehicles on Low-Volume Roads. In Transportation Research Records: Journal of the Transportation Research Board, No.1819, TRB, National Research Council, Washington, D.C., 2003, pp228-235. 4. Timm, D.H., Priest, A.L., and McEwan, T.V. Design and Instrumentation of the Structural pavement Experiment at the NCAT Test Track. National Center for Asphalt Technology NCAT, Auburn University, NCAT Report 04-01, 2004. 5. Transportation Association of Canada (TAC). Pavement Design and Management Guide. Ottawa, 1997.

Hamad, A. and Cowe Falls, L. 8 LIST OF TABLES AND FIGURES TABLE 1 Common Axle Configurations Found on Canadian Roads (TAC 1997) TABLE 2 Pavement Structure for the Different Sections in the Test Road TABLE 3 Testing Sequence TABLE 4 Test Vehicle Weights and Specifications TABLE 5 Longitudinal Interfacial Tensile Strains 1 and Vertical Stress 2 on Top of the Subgrade Layer Recorded Directly Under the Wheels Path for the Tested Vehicles (Speed = 60 km/hr) TABLE 6 Distribution of the Transversal Strains Within the Array for the HMA Section/All Vehicles FIGURE 1 Sample crane. FIGURE 2 The test road layout. FIGURE 3 Typical longitudinal strain profile under Speiring crane, HMA section, speed 60km/hr. FIGURE 4 Typical vertical stress profile under Speiring crane, HMA section, speed 60km/hr.

Hamad, A. and Cowe Falls, L. 9 TABLE 1 Common Axle Configurations Found on Canadian Roads (TAC 1997) Axle Tonnes Kips Steering 5.50 12.1 Single 9.10 20.1 Tandem 17.00 37.5 Tridem 1 21.00 46.3 23.00 50.7 24.00 52.9 Combinations 5 Axle Tractor Semi-Trailer 39.5 87.1 6 Axle Tractor Semi-Trailer 46.5 102.5 A Train Double 53.5 117.9 B Train Double 62.5 137.8 C Train Double 53.5 117.9 1 The three weights relate to different axle spacings.

Hamad, A. and Cowe Falls, L. 11 TABLE 2 Pavement Structure for the Different Sections in the Test Road Pavement Layers Description Test Section Asphalt Concrete Hot Asphalt Concrete Cold Mix Mix Granular Base Course Prepared Subgrade Borrow Material Hot mix asphalt concrete. ¾ in. Maximum crushed Prepared Silty Clay. Total thickness is 110 compacted granular base Thickness is 150 mm. scarified and compacted. mm. constructed in two course. Thickness is 50 lifts. First lift is 50 mm mm. primed with MC30. HMA and second lift (surface) is 60 mm. Maximum aggregate size is ½ in. Asphalt binder is 200-300A. Tack coat on the first lift is SS1. N/A CMA N/A GBC N/A N/A Cold mix asphalt concrete. Total thickness is 50 mm. Maximum aggregate size is ½ in. Asphalt binder is 200-300A. Prime coat on the compacted granular base is MC30. ¾ in. Maximum crushed compacted granular base course. Thickness is 50 mm. primed with MC30 ¾ in. Maximum crushed compacted granular base course. Thickness is 50 mm. primed with MC30. High Plasticity Clay with > 70% passing sieve #200 material. Liquid limit is 66%, Plastic Limit is 21%, and Plasticity Index (P.I.) is 45%. Maximum dry unit weight is 14.84 kn/m 3 and optimum Moisture Content is 21.46%.

Hamad, A. and Cowe Falls, L 12. TABLE 3 Testing Sequence Load Repetition No. 1 2 3 4 5 Standard 18kip @ 60 km/hr Standard 8 wheel Tandem (17,000kg) @ 60 km/hr Creep Standard 18 kip Creep Crane Standard 18kip 5km/hr Crane Standard 18kip 10 km/hr Crane Standard 18kip 15 km/hr Crane Standard 18kip 20 km/hr Crane Standard 18kip 30 km/hr Crane Standard 18kip 40 km/hr Crane Standard 18kip 50 km/hr Crane Standard 18kip 60 km/hr Crane 4 axle configuration Standard 18kip 60 km/hr Crane 7 axle configuration Standard 18kip

Hamad, A. and Cowe Falls, L 13. TABLE 4 Test Vehicle Weights and Specifications Vehicle Standard 80 kn 8 Wheel Tandem Speiring Crane Tire Specifications Michelin XZE M/275/80 R22.5 Michelin XDE M/275/80 R22.5 Goodyear Omnitrack MSS 445/75 R22.5 Inflation pressure 90 psi Axle 1 Axle 2 Axle 3 Axle 4 Total 4,980 7,800 - - 13,140 5,160 12290 17,040-34,490 11,700 12,400 10,650 9,200 43,950

Hamad, A. and Cowe Falls, L. 14 TABLE 5 Longitudinal Interfacial Tensile Strains 1 and Vertical Stress 2 on Top of the Subgrade Layer Recorded Directly Under the Wheels Path for the Tested Vehicles (Speed = 60 km/hr) Tested Vehicle Test Road Section 3 Standard 18 kip Truck (ST18) Axle # for the tested vehicle 1 2 3 4 5 6 7 Longitudinal interfacial strains Speiring Crane -Four Axle Configuration (CR01) HMA CMA 334 (379) 1057 (1704) 1053 (1362) 4 1051 (1345) 1703 (2238) 1806 (2380) 1031 (1361) 1677 (2110) 1049 (1392) 1792 (2309) - - - - - - Speiring Crane -Seven Axle Configuration (CR02) HMA CMA 334 (379) 1057 (1704) 923 (1164) 1732 (2261) 938 (1170) 1832 (2273) 932 (1177) 1844 (2343) 914 (1157) 1951 (2347) 865 (1123) 1783 (2278) 844 (1112) 1731 (2122) 819 (1054) 1766 (2226) Vertical stress on top of the subgrade layer Speiring Crane -Four Axle Configuration (CR01) 9.5 18.0 23.3 24.3 23.1 - - - HMA Speiring Crane -Seven Axle Configuration (CR02) 9.5 12.8 14.9 16.5 15.3 15.3 18.8 19.8 1 Average+ 1S.D. (mm/mm x 10-6 ) 2 Average+ 1 S.D. (kpa) 3 HMA=Hot Mix Asphalt, CMA= Cold Mix Asphalt 4 Values between ( ) are for total strains

Hamad, A. and Cowe Falls, L. 15 TABLE 6 Distribution of the Transversal Strains Within the Array for the HMA Section/All Vehicles O1 Strain Gauges Array I1 O2 Travel Direction I2 O3 I3 Vehicle Type Axle # O1 * O2 O3 I3 I2 I1 ST18 2 C C/T T C C/T T Speiring Crane -Four Axle Configuration (CR01) Speiring Crane -Seven Axle Configuration (CR02) 1 T T C C C T 2 T T C C C T 3 T T C C C T 4 T T C C C T 1 T T C C C T 2 T T C C C T 3 T T C C C T 4 T T C C C T 5 T T C C C T 6 T T C C C T 7 T T C C C T * O1, O2, and O3 refer to the strain gauges on the side of the outer wheel path, while the I1, I2, and I3 refer to the side of the inner wheel path

Hamad, A. and Cowe Falls, L. 16 FIGURE 1 Sample crane.

Hamad, A. and Cowe Falls, L. 17 PLAN VIEW Centre line Centre lane 390mm 390mm 1000mm 1000mm 1000mm 1000mm D.A.S Strain Transducer Temperature Profile Moisture Gauge Pressure Gauge CROSS-SECTION VIEW N.T.S. Hot Mix Asphalt 110 mm Cold Mix Asphalt 50 mm Granular Base Course 300 mm GBC 50 mm GBC 50 mm Subgrade Prep 150 mm 150 mm FIGURE 2 The test road layout.

Hamad, A. and Cowe Falls, L. 18 Longitudinal Interfacial Strain (H-CR01-1-60-P05) 0.0012 Axle1 Axle2 Axle3 0.001 Strain (mm/mm x E-06) 0.0008 0.0006 0.0004 0.0002 ACP 1SLO1 018 ACP 1SLO2 030 ACP 1SLO3 011 0-0.0002 0 0.2 0.4 0.6 0.8-0.0004 Time (Seconds) FIGURE 3 Typical longitudinal strain profile under Speiring crane, HMA section, speed 60km/hr.

Hamad, A. and Cowe Falls, L. 19 Vertical Stress in the Subgrade (H-CR01-1-60-P05) ACP 2PO 310 30 Axle1 Axle2 Axle3 25 20 Vertical Stress (kpa) 15 10 5 0 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Time (Seconds) FIGURE 4 Typical vertical stress profile under Speiring crane, HMA section, speed 60km/hr.