Pavement Trials at Ringour Quarry: Evaluating rutting in forest roads in Southern Scotland

Transcription

1 Pavement Trials at Ringour Quarry: Evaluating rutting in forest roads in Southern Scotland L. A. T. Brito & A. R. Dawson Nottingham Transportation Engineering Centre, University of Nottingham, England, UK R. W. W. Tyrrell Forestry Civil Engineering, Castle Douglas, Scotland, UK ABSTRACT: Among efforts being made by the forest authorities to minimize the impact of timber transportation on Scotland s rural and public roads, the Roads Under Timber Transport Project (R.U.T.T) has carried out full scale trials at the Ringour Quarry Facility, South-West Scotland. These trials are part of a two years project which aims to assess the development of permanent deformation of forest roads in Southern Scotland as a function of traffic, weather and season. The main goals of these pavement trials were to evaluate the performance of aggregate under saturated conditions and relative pavement deformation caused by different timber haulage vehicles. Ten different trials were carried out combining three different aggregate materials and five types of vehicles. Tyre fitment and vehicle configuration were also assessed and demonstrated to play an important role on the study of rutting development. Amongst the preliminary conclusion drawn from the results, it is suggested that management of tyre s inflation pressure could be one of the most economic means of managing pavement deterioration in the forest road network. 1 INTRODUCTION Permanent deformation is unquestionably one of the main distress mechanisms in low volume roads. Innumerous are the road segments that collapse after a small number of load cycles due to excessive rutting. This is especially true on those roads on which budget is limited, and therefore, construction standards are far from desirable, material quality is low, vehicle loads are constantly high and weather is not favorable. Forest roads in Scotland are certainly a paradigm case. In order to improve the forest roads condition and to help in finding an optimum of financial resource investment with cost/effective timber haulage, the Forestry Commission (FC) through the Forestry Civil Engineering and the University of Nottingham through the Nottingham Transportation Engineering Centre (NTEC), sponsored by the Strategic Timber Transport Scheme, have developed the R.U.T.T Roads Under Timber Transport research project. The R.U.T.T project, conceived as a two-year long study, aims to evaluate the damage in forest roads located in southern Scotland, by monitoring 21 forest road entrances and also in performing specific trafficking trials on purpose-built road construction Ringour Quarry Facility. The main objective of these trials were to evaluate the performance of aggregate under saturated conditions and the relative pavement deformation caused by different timber haulage vehicles. In particular different traffic types (axle arrangements, tyre arrangements, tyre inflation pressure) were investigated. The aim of the programme is to gather sufficient data to determine the principal controls affecting rutting and the sensitivity to changes in value of the road material type and condition. 2 TRIALS DESCRIPTION Ringour, located off the A762 road, four miles from New Galloway, was originally a quarry for aggregate used in forest roads construction. It was later transformed in a purpose built test facility by the Forestry Commission for full scale pavement tests. The site was used to construct specific road sections of approximately 30m in a more controlled fashion. A pair of concrete walls was provided to act as a confinement of the pavement edges and to provide a datum from which to monitor rutting under controlled trafficking (Fig. 1). A watering system was assembled to promote the soaking of the road surface, in order to accelerate the rutting process (Fig. 2). The trial section layout consisted of an unsurfaced pavement with two layers, a base course of 300mm, on which three different materials were tested, and a 300mm layer of sub-base.

2 Figure 1. Ringour Testing Facility before construction of the test section with major dimensions indicated and sections chainage Figure 2. Intermittent watering system during trial tyres controlled from the driver s cab. The TPCS system is of growing use in some countries in the forest business. The variable pressure in the tyres has a direct effect on the area in contact with the pavement, and, therefore, the contact pressure tyrepavement. At lower pressures the same vehicle may cross a softer pavement without causing rutting or wheel spin. The technique also seems to result in much less tyre wear and, perhaps, fuel use. The drawback is that the vehicle cannot travel at speed on conventional pavements without safety concerns and extra fuel use, so the pressure must then be increased to conventional levels. This system is credited in reducing the rutting damage in unsealed roads and also in improving ride quality for drivers. Figure 3. Tyre and axle arrangements for the vehicles used in the trial 2.1 Vehicles A variety of conventional and unconventional vehicle types were used to traffic the trial sections. The most common vehicles used were an articulated truck and trailer, the trailer being fitted with supersingle tyres, and the other a FC rigid body vehicle with twin tyres. A third employed vehicle was a rigid body lorry with a trailer, Lorry & Drag, with twin tyres fitment. Use has also been made of a bespoke vehicle currently used in Scotland, known as a Low Ground Pressure (LGP) vehicle, as a tool to reduced damage to forest roads by means of an evenly spaced configuration of tyres. Although the LGP vehicle has the same number of wheels as a twin-tyred lorry, the effect of load spreading caused by two separate sets of suspensions per axle and evenly spaced tyres across the chassis, allow a more evenly distribution load on the pavement, even if the road is cambered. In addition, it was used an articulated truck and trailer with a Tyre Pressure Control System (TPCS) fitted. This system is an alternative means of applying low ground pressure via the use of low pressure (a) (b) Figure 4. (a) FC Multi-lift truck trafficking a Ringour trial (b) Low ground pressure vehicle trailer with tyres equally spaced across pavement on stub-axles

3 Figure 3 illustrates a summary of the vehicles employed during the trials. Note, especially, the use of vehicles with super-single tyres (single tyres nominally 385mm wide as opposed to a pair of tyres each nominally 295mm wide). Figure 4a pictures a typical vehicle (FC Multilift) running on a trial; Figure 4b, a rear view of the LGP vehicle. 2.2 Materials The granular materials used in this study were: a. Type 1 standard material for base and sub-base in road works in the United Kingdom (Highways Agency 2007) from Morrington Quarry, b. a material from FC s Risk Quarry a lower quality material of proven lower performance in regard to rutting observed in previous trials (Tyrrell 2004), c. a Type 1 material from the FC s Craignell Quarry, d. a standard Type 3 (Highways Agency 2007) material is also used. The first three of these were used in several of the trials as the surfacing aggregate whereas the last one was only used beneath the other aggregate layers to provide a cushion and mitigate the effects of the hard quarry floor on the surfacing. More details of the material can be found elsewhere (Brito & Dawson 2007). 2.3 Trial procedure A standard compaction protocol was followed (Highways Agency 2005) in order to reduce variability in compactions levels among the various trials. Between each trial, the test road surfaced was prepared by loosing, re-grading and compacting the upper layer. The trials run basically consisted in 100 passes of one vehicle at a time in the test section, following the same wheel track. Measurements were taken of the changing surface profile at various number of passes using a laser device supported on a reference beam, with measurements recorded at 10mm spacing transversally in two different cross sections (T2 and T4 See Figure 1). In some trials the vehicle re-trafficked the same wheel track, either at different moisture content, first dry and then in wet condition (Trials 1 & 2, 9 &10), or in different tyre pressure, lower pressures first (Trial 7). Occasionally, two vehicles run simultaneously in different wheel tracks during the same trial (Trials 4, 6, 9 &10). Throughout the test, the watering system was turned on to soak the surface for a define period of time at each 10 vehicle passes. As weather varied amongst trials, and therefore evaporation and precipitation levels, moisture contents were determined in the aggregate material at 150mm of depth at the end of each trial. The purpose of the water is to weaken the material allowing higher levels of permanent deformation within 100 vehicle passes. The combinations of materials, material condition and trafficking arrangements is shown in Table 1. Vehicle s tyre pressure and dimensions were registered. A Dynamic Cone Penetrometer (DCP) and a Mini FWD device were used in some of the trials for the assessment of compaction level and materials strength. The results didn t yield reliable results, and are currently under further analysis. On the basis of the axle loadings of each vehicle, the equivalent damaging potential of each vehicle was estimated according to the familiar 4th Power Law and the number of ESALs (Equivalent Standard Axle Loads) per vehicle pass was computed (Table 2). Vehicle axle loads were obtained from weighbridges or by manufacturer s specifications. The computation also relies on the assumption that damage increases with the axle loading raised to the 4th power. This law is known not to apply to the rutting of pavements with no bound course, but the alternative methods are no more reliable, so the 4th power computation is used on the basis that it is, at least, familiar. One deficiency of the technique is immediately apparent. The TPCS (low tyre pressure) vehicle is computed to have the same damaging capacity as the same vehicle at higher tyre pressures because damage is assumed to be based solely on axle loading. Clearly this is erroneous as lowering tyre pressures unquestionably reduces pavement damage. However the same set of data also shows a relevant point the so-called Low ground pressure vehicle causes more damage per pass than conventional vehicles. This apparent anomaly is not because the special trailer doesn t spread the load better than conventional vehicles (it does) but because the truck pulling the trailer imposes significant damage which is greater than the damage saved by using the novel trailer. Table 1. Loading and moisture matrix of Ringour Trials VEHICLE TYPE MATERIAL Morrington Risk Quarry Craignell FC Type 1 Material Type 1 Dry Wet Dry Wet Dry Wet FC Multi-Lift with twin tyres #1 #2 #8* #6 #9 #10 Artic DAF with super singles #3 #6 #9 #10 Lorry & drag with super singles #4 LGP Vehicle #5 TPCS Vehicle #7 # represents the chronological order of the trial * Trial with tyre pressures of 70psi and after 110psi

4 Table 2. Vehicles ESAL (Equivalent Standard Axle Load) calculation according to 4 th Power Law Vehicle Trial 3 PERMANENT DEFORMATION MONITORING The permanent deformation monitoring was performed with a laser based device equipped with built-in Bluetooth transmitter, allowing connection with a handheld computer for data storage. Two cross sections monitoring stations were set 6m apart, centered in the trial lane, for reference level. On each station a portable 5.5m long aluminum beam allowed readings to be taken at 0.1m spacing across the road. In each cross section two rows of readings, 0.3m apart, were taken for consistency. All four measured cross section profiles were plotted and the permanent deformation recorded for each wheel path at 0, 10, 20, 40, 70 and 100 vehicles passes. Figure 5 show a typical result of the permanent deformation monitoring. Two methods were used to calculate the permanent deformation developed in the trials: rut depth and vertical surface deformation (VSD). The permanent deformation considers the upward shoving at the edges of the wheel path, and is determined considering a straight edge across the surface. The VSD is the difference between the current cross section level and the start reference level of the pavement. Although arguably the VSD may be considered best for describing surface deformation (Arnold 2004), the rut depth calculation yield better results for the analysis of the Ringour trials. This could be attributed to the high level of near surface deformation considerable amount of heaving in the wheel paths edges. This is likely to be due to the high levels of moisture in the surface caused by watering mechanism used. Furthermore, as the surfacing materials used in the trials were somewhat dense graded, the water is expected to have permeated at low speed. 4 RESULTS GTW (kn) Front Axles load (kn) Rear ESAL FC Multi-Lift # 1,2,4,6, FC Multi-Lift # Artiulated DAF # 3,6,9, Lorry & Drag # LGP # TPCS # The observed rutting in the all trials is summarized in Figure 6. Those trials on which the vehicle run at different configurations in the same wheel track were plotted forward the number of preceding vehicle passes. In those on which two vehicles run parallel to each other, the results are plotted from the origin. The results of the analysis of the permanent deformation as a function of number of ESALs are presented in Table 3. In this table the rutting rates are calculated considering only the last 60 vehicle passes in each trial after initial settlement has occurred and rutting rates are more stable. The table also summarizes the various variables monitored in the trials and presents a ranking according to the rutting rates which is illustrated in Figure 7. Figure 5. Example of a cross section profile measurement Pavement Elevation (m) Section T2B - Trial N= N=10 N=20 N=40 N=70 N= Transverse Distance (m) Figure 6. Rutting observed in trials at Ringour Rut (m) @758 6ML 9 Artic 7@689 7@517 4ML 8@483 6 Artic L&D 9ML 7@ ML 10 Artic Number of Vehicle Passes Trial 1 Trial 2 Trial 3 Trial 4 L&D Trial 4 FCE ML Trial 5 Trial 6 Artic Trial 6 FCE ML Trial 7 517kPa Trial 7 689kPa Trial 7 793kPa Trial 8 483kPa Trial 8 758kPa Trial 9 Dry - Artic Trial 9 Dry - ML Trial 10 Wet - Artic Trial 10 Wet - ML Figure 7. Rutting rates observed according to number of ESALs Rutting Rate (mm/esal) Trial 1 Trial 2 Trial 3 Trial 4A Trial 4B Trial 5 Trial 6A Trial 6B Trial 7A Trial 7B Trial 7C Trial 8A Trial 8B Trial 9A Trial 9B Trial 10A Trial 10B

5 Table 3. Ringour Trials summary chart Vehicle Mean Tyre Vehicle ESAL Tyre Fitment Pressure (kpa) Material tested Moisture Condition Moisture Content (%) Rutting (mm) N of ESALs Rutting Rate (mm/esal) Trial 1 Multi-Lift 3.63 Twin Tyres Morring. T1 Dry Trial 2 Multi-Lift 3.63 Twin Tyres Morring. T1 Wet Trial 3 Artic 4.94 Sup. Singles 786 Morring. T1 Wet Trial 4A Lorry & Drag 5.33 Twin Tyres 834 Morring. T1 Wet Trial 4B Multi-Lift 3.63 Twin Tyres Morring. T1 Wet Trial 5 LGP Twin Tyres 510 Morring. T1 Wet Trial 6A Artic 4.94 Sup. Singles 724 Risk NS Wet Trial 6B Multi-Lift 3.63 Twin Tyres 779 Risk NS Wet Trial 7A TPCS 4.94 Sup. Singles 517 Risk NS Wet Trial 7B TPCS 4.94 Sup. Singles 689 Risk NS Wet Trial 7C TPCS 4.94 Sup. Singles 793 Risk NS Wet Trial 8A Multi-Lift 3.63 Twin Tyres 483 Morring. T1 Wet Trial 8B Multi-Lift 3.63 Twin Tyres 758 Morring. T1 Wet Trial 9A Artic 4.94 Sup. Singles 689 Craignell FC T1 Dry Trial 9B Multi-Lift 5.15 Twin Tyres 689 Craignell FC T1 Dry Trial 10A Artic 4.94 Sup. Singles 689 Craignell FC T1 Wet Trial 10B Multi-Lift 5.15 Twin Tyres 689 Craignell FC T1 Wet Ranking* 5 DISCUSSION For the pavements with low total rutting there is normally an initial high rate of rutting followed by a stabilizing response (e.g. Trials 1, 9 & 10). This type of response is very common (Dawson, 2008), and is the response desired of the best pavements where traffic levels are to be highest and long-term performance must be assured. Other pavements show an on-going development of rutting. In some cases this development is very large and rapid (e.g. Trial 3) but, for the most part, it is moderately fast. Such a response is appropriate for a less frequently trafficked pavement or one that only needs to provide service for a limited time. Regarding the vehicle used, the much greater rutting that occurs under the FC s multi-lift vehicle is apparent when compared with that occurring due to the articulated vehicle used on the same pavements (Trials 6, 9 & 10). This is despite the fact that the articulated lorry has super-single tyres which, it is known from other studies, are more damaging than the twin tyres which fitted to the multi-lift vehicle. Why is there this disparity in the present study? For Trial 6 this might be explained due to tyre inflation problems. Only by the time of Trial 8, it was found that the FC Multi-Lift vehicle had run from Trials 1 to 6 with the inner tires of the rear most axle with very low pressure (69 & 276kPa). This was due to the difficult in reaching those tyres for pressure checking. Hence, as the rear axle didn t have tyres at the same pressures, this almost certainly resulted in local overloading and accelerated damage. For the other trials the reasoning is not so clear. It may relate to the more limited opportunity for load distribution along the length of a rigid-bodied lorry than along an articulated one. Table 2 does not indicate that the individual axles are very differently loaded, so a rogue axle loading does not seem to be the explanation in this case. The same explanation is almost certainly valid for the comparison of Trials 4A and 4B. Trial 4B caused more caused approximately twice as much rutting as 4A, and a three times higher rutting rate, when related to number of ESALs. Nevertheless, a close study shows that the VSD levels were very similar for both vehicles; the heaving effect on the right hand side wheel path trafficked by the FC Multi- Lift was responsible for increasing the rutting measurement, suggesting higher shear stresses near the surface. This may be explained by the later finding that the rear-most axle of the multi-lift was running virtually on two wheels a factor that is likely to be the responsible for a very high stress level between tyre and pavement, provoking higher shear stresses near the surface. With the so-called low ground pressure vehicle, the novelty of spreading the load across the pavement through multiple wheels is negated by the heavily loaded axles of the rear of the tractor unit.

6 Visibly, these appear to have been the chief cause of damage and the results in Figure 6 appear to confirm this although the lack of a direct comparison to the trafficking of the same pavement by another vehicle hinders interpretation. Table 3 seeks to interpret the data in a somewhat different manner by looking only at the rate of rutting during the last 60 passes of each trial. The aim is to be able to separate-out any bedding-down effects. The results (last column) yield a similar story regarding the multi-lift vehicle. Then the rate of rutting is either similar or greater for the multi lift (twin-tyred) than for the articulated vehicle (with super-single tyres on its trailer) Trials 6, 9 & 10. Regarding the low-ground pressure vehicle, however, the story is somewhat different. Despite applying a very high number of equivalent standard axles compared to other vehicles, the rate of rutting is very low. This suggests that the vehicle doesn t prevent rut initiation but limits ongoing damage due to an inherent kneading action occasioned by the multiple wheels over the road width. Nevertheless this does lead to some looser aggregate on the pavement surface which can collect between the wheel tracks allowing higher than expected rutting, even when the VSD reading is low. The story told by Table 3 concerning the effect of moisture and materials is interesting. There is a clear ranking of rutting from Risk (greatest) to Craignell (least). The relative moisture content has a lesser effect on rutting than the material chosen though some effect can be seen (e.g. compare Trials 9 & 10 where other parameters are constant). Trial 2 (ranked 13th in Table 3) behaved somewhat better than Trial 1 (ranked 11th), had the same lorry and the same construction, but was wet instead of dry. One possible reason for that is that, as the lorry trafficked the same wheel path during both trials, the initial movement that occurred in Trial 1 compacted the aggregate, limiting the opportunity for further deformation in Trial 2. Furthermore, following the assessment of Figure 6, it is possible to notice that the majority of the rutting from Trial 1 came from the early passes, resulting in a rather stable behaviour after 20 passes of the vehicle. Trial 2, however, presents an increasing rate of rutting throughout the test, resembling incremental collapse behaviour of the material. The relatively small permanent deformation registered is a valid response to the very low moisture content for a wet trial (3.7% - lowest of all trials). Considered is this way, the wetter condition is, after all, associated with poorer performance. Because the behaviour in the first trial probably affected the performance in the second, more disturbance and reconstruction was included between subsequent trials. 6 CONCLUSIONS Some conclusions from the results and analysis presented in this study case can be drawn: i. The means of loading and the type of aggregate used have a bigger influence on rutting performance than the relative moisture level of the aggregate (over the range of moisture conditions considered). ii. Reducing overloading of trucks and controlling tyre pressures may be a more economic means of managing pavement deterioration than by using higher quality aggregates (the other meaningful option). This suggests that some kind of policing or QA system for truck operators might be worth considering. iii. As far as rutting is concerned the type of vehicle appears to make some difference, but it has not been clearly established which factors of the different types of vehicles are most important in controlling propensity to rut. This aspect could usefully be investigated further, especially given the variety of options available and the need to have vehicles that can easily manoeuvre in the forest environment. ACKNOWLEDGEMENTS The work presented in this paper is part of the RUTT Project, funded by the Strategic Timber Transport Forum / Scotland. The authors wish to express their sincere thanks for all collaborants involved. REFERENCES Arnold G Rutting of Granular Pavements. PhD Thesis, University of Nottingham, UK. Brito, L.A.T & Dawson, A.R Roads Under Timber Traffic. Interim Research Project Report. NTEC Report No University of Nottingham. December Dawson, A.R. (2008 in press). Rut Accumulation in Low- Volume Pavements due to Mixed Traffic, paper accepted by Transportation Research Record. Highways Agency Manual Of Contract Documents For Highway Works - Volume 1 Specification For Highway Works. Series 800 Road Pavements - (11/04) Unbound, Cement And Other Hydraulically Bound Mixtures. Amendment May Highways Agency Manual of Contract Documents For Highway Works - Volume 2. Notes for Guidance On The Specification For Highway Works. Series Ng 800 Road Pavements - (11/04) Unbound, Cement And Other Hydraulically Bound Mixtures. Amendment November Tyrrell, W Pavement trials at Risk Quarry, Kirroughtree. 6 th International Symposium on Unbound Materials Unbar 6. Proceedings Intern. Symp., Nottingham, 6-8 July England: Balkema.