It is recommended that Culway calibrations are carried out at least once every two years as well as after each major repair or changes to the pavement and changes/repairs in the vicinity of the sensors. Replacement of any strain gauges also requires a re-calibration.
The calibration process requires the collection of controlled vehicle axle mass data, analysing and reporting the data to meet the following requirements:
Calibration data collection
This covers the provision of 'reference vehicles' used in the comparison of dynamic weights to that of the corresponding static weights.
The dynamic weights of the reference vehicle were measured by the Culway system and the static weights were measured by a static weighbridge. If a weighbridge is not easily accessible then portable scales may be used as long as they have been recently certified as being accurate and are used on a firm level surface (ideally a concrete slab).
The procedure used involves driving known statically weighed vehicles across the Culway site at normal highway speeds.
A number of Culway logged passes are made at varying static weights ranging from empty to fully loaded. On average 45 passes are made at each of the instrumented lanes at a site. Ideally, the more passes made at the various weights from all the reference vehicles the better one can predict the repeatability performance of each site.
The prescribed calibration process must be carried out for each of the lanes being calibrated, as no site/lane is the same.
A specially equipped semi-articulated truck is hired for the duration of the fieldwork. This vehicle travels between the sites to perform the required calibration runs.
The vehicle comes equipped with 16 one tonne calibrated steel blocks and a forklift, which is used to load and unload the weights.
This approach greatly improves the efficiency and accuracy of loading the vehicle to a specified weight range. The most significant advantage is the element of not having to arrange for suitable loading material and means of loading and unloading at each of the sites.
The rigid R11 truck requirement is met by hiring a suitable vehicle (at least 3.8 m wheel base). The steel block weights are used for loading the reference vehicles to the required weight.
The single drive rigid R11 is loaded with four different weight variations ranging from empty to fully loaded.
The articulated semi A123 is loaded with three different weight variations; empty, half loaded and fully loaded.
Figure: Calibration vehicles
Source: Pentagon Systems 2002
Processing and analysis of calibration data
This includes the downloading, validation, analysis and processing of the manually and electronically collected calibration data.
Derivation of calibration factors
These factors are derived from the calibration programs (e.g. WimCal):
- Sensitivity factor: this parameter relates tonnes per raw strain count for the culvert. This value would typically range from 0.005 for a highly sensitive culvert to 0.1 for a low sensitivity culvert. This factor is multiplied by the averaged strain count recorded from the logger to produce an axle mass.
- Power factor: this is a non-linear response compensation factor. A figure of unity is used for a linear transfer response. Positive figures fractionally less than or greater than 1.0 compensate for non-linear transfer functions. This factor was introduced because some Culway sites exhibited a non-linear response when loaded. As axle loads were increased so that the response from the culvert could either increase or decrease more than expected by using simple proportions. To compensate for this non-linear response the same algorithms are used to calculate axle weight.
- Unit influence line (UIL): When a wheel approaches and passes over a culvert, the response from beginning to end is described as the influence line. This is the effective longitudinal culvert length. This figure is approximated to the actual culvert length plus twice the depth of fill, in metres. This unit of length represents a distance in metres which axle group weights are adjusted to compensate for strain overlap. Any axles within a group being weighed with spacing less than the UIL will be adjusted. Any axles with spacing greater than the UIL value will not require adjustment.
- Steer factor: Culway has an inherent problem with continually under-weighing steer axles. To overcome the tendency to under-weigh steer axles compared to other axle groups a new calibration factor has been added to the analysis software. The steer factor is a value that is used to multiply the original calculated steer weight. Typically the value lies between 1.0 (if steers produce similar strains to other axles) and 1.5 (if steers produce smaller strains than the equivalent weight on another axle). Values of less than 1.0 can occur if steers are overweighed. This factor can be used to convert Culway strain counts to steer axle weight for single steer vehicles.
A final report is produced to provide an overview (brief summary) of the quality of the data collected, and the reliability/consistency/performance of each lane.
Time-dependent Drift between Calibrations
It has been known for some time that the Culway data can be affected by the time-of-day and week-of-year. The variation or drift is dependent on site and environment conditions. The strength of the bituminous surface on top of a culvert box is affected by temperature, and how well air and water are drained through the surface.
Austroads 2004 reported the study of two Culway sites in Victoria, making use of the property that the steering axle mass (SAM) of an articulated six-axle is the least affected by temperature and drainage conditions. This is because there is no direct load on the steering axle. It is therefore possible to use this constancy of average SAM to correct for diurnal and seasonal drift in WIM sensitivity.
Figure: Three-axle masses of an articulated six-axle vehicle (Class 9)
The results are shown in the figures below. The first figure shows the diurnal effect of hour for the sites – Western Ring Road in Melbourne and Yarra Glen just outside Melbourne. The second figure shows the seasonal variation over six months (July to December). The variation can be substantial in the case of the Western Ring Road. The correction factor varied from 0.93 (winter) to 1.1 (summer). The Yarra Glen site showed significantly less variation, possibly due to the use of a more traditional seal whereas the more porous surface (asphalt seal) of the Western Ring Road exhibited more variation with temperature and drainage.
Figure: Adjustment factors for diurnal variation of Culway axle-mass data
Figure: Adjustment factors for seasonal variation of Culway axle-mass data
A Culway system has its own temperature-compensating mechanism in its electronic hardware and provides reasonable results with acceptable accuracy levels but does not consider temperature/climatic variations of the pavement on top of the culvert box. It is therefore good practice to pay attention to site conditions and how seasonal changes in climate could affect a WIM installation. Some road agencies already allow for seasonal and diurnal drifts of Culway.
Normalisation of Culway WIM Data
The effects of 'calibration drift' at Culway sites are predominantly caused by fluctuations in temperature of the road surface/pavement.
To reduce the effects of calibration drift, DTEI SA has developed a procedure that can be applied to the data collected from any Culway site.
The procedure is based on the theory that the steer axle weight of a Class 9 vehicle barely changes regardless of the load being carried. This has been proven during the weighing of these vehicle types during calibration exercises and data from weighbridge stations.Therefore, the average steer axle weights of Class 9 vehicles at a given site should remain constant throughout the year (between 5.7 and 5.95 tonnes depending on location) for all seasons. The average steer axle weights (Class 9) for each Culway site have been determined and re-verified every three or four years with data from weighbridge sites.
The average steer axle weights for Class 9 vehicles at each Culway site are then used as a benchmark weight to check against the average for each file of data downloaded from a site (generally for one week). The sensitivity factor (which is the primary calibration factor) is then adjusted for the data files where a difference is identified.
A spreadsheet with macros has been developed by the traffic information unit to optimise the process of calculating an adjusted sensitivity factor for each data file so that this can then be used on all the data for that period. By doing this the users not only adjust the Class 9 steer axle weights but also proportionally adjust all the other axle groups for all vehicles in the data file accordingly (so that users get a more accurate weight estimate for all axle groups)
Pentagon Systems 2002, ‘Culway calibration procedure v1.0’, Pentagon Systems, Qld
Caruana, C 2006, ‘Culway weigh-in-motion (WIM) compensating for ‘calibration drift’ preliminary report’, internal report, Transport Information Management Section, Department for Transport, Energy and Infrastructure, Adelaide, SA.
Pearson, E 2010, ‘SA Culway calibration drift weigh-in-motion (WIM) site comparison report’, January 2010, internal report, Spatial Intelligence & Road Assets section, Department for Transport, Energy and Infrastructure, Adelaide, SA
Peters, RJ 1998, ‘Low cost calibration management’, Second European conference on weigh-in-motion of road vehicles, Lisbon, Portugal, Publications office of the European Union, Luxemberg, pp.153-60.