Piezoelectric WIM is also widely used in Australia.
A piezoelectric WIM system generally consists of at least one piezoelectric sensor and at least one inductive loop, embedded in road cuts or it may be portable. (Kistler piezoelectric sensors are encapsulated in an epoxy-filled metal channel, usually made of aluminum.)
Figure: A piezoelectric WIM site
Source: Kistler 2004
All piezoelectric sensors are placed in the travel lane perpendicular to the direction of travel enabling the wheels of one axle to hit the sensor at the same time. In the case of quartz piezoelectric sensors, one sensor is used for each of the two wheel paths in a lane.
The inductive loops are placed upstream and downstream of the sensor. One inductive loop is placed upstream from the scale to detect an approaching vehicle and triggers a sequence of events: WIM sensor signal detection, amplification, and collection. The other loop is placed downstream to determine the vehicle speed and axle spacing based on the time it takes the vehicle to traverse the distance between the loops. The distance between the two loops cannot be less than the required minimum distance.
Axle spacing, number of axles, vehicle length and weight enable the system to classify vehicles.
The main piezoelectric cable sensors applied in the Australian market include:
- Roadtrax® BL traffic sensor (supplied by MSI Sensors)
- Lineas® Quartz sensor (supplied by Kistler).
The HSWIM vehicle classification systems; RAKTEL 8000 (Mikros), Hi-Trac 100 (TDC) and ViperWIM (Applied Traffic) all use the Roadtrax BL traffic sensors while ARRB Express-Weigh uses the Kistler Lineas sensors.
The figure below shows an example of a piezoelectric WIM system consisting of a total of eight Kistler Lineas sensors, two rows of sensors, per traffic lane.
Figure: Layout of a piezoelectric WIM station
Source: Kistler 2004
The advantages of this set-up are:
- It captures all information about the vehicle.
- With two rows of four measuring channels information about the speed can be acquired and it is possible to evaluate the wheel loads separately. This also allows for evaluation and data processing of the precise gross vehicle weight, allowing for the differences between each wheel load (left and right). This will open up many evaluation possibilities, including
- different data processing for single axle load and group of axle loads – the cruise test between every axle in one group of axle
- precise vehicle classification
- asymmetric vehicle loading check (left to right)
- individual calibration coefficient for every channel
- high accuracy of measured data.
Other possibilities dependent on software development and electronic equipment such as an automatic self calibration procedure can also be developed for specific products.
The figure below details a Kistler Lineas piezoelectric cable sensor.
Figure: Diagram of the Kistler Lineas piezoelectric cable sensor station
Source: Kistler 2004
A wheel rolling over a piezoelectric cable sensor applies vertical forces to the quartz crystals in the sensor with virtually no deformation. The piezoelectric quartz disks yield an electrical charge proportional to the applied forces. The piezoelectric sensitivity is practically independent of temperature, time and speed.
The electric charge signals are converted by a charge amplifier into exactly proportional voltages which can be further processed as required. The accuracy of the measured wheel load is not influenced by tyre type, tyre quantity or tyre pressure. In the case of dual tyres, the piezoelectric cable sensor measures one signal and expresses it as one wheel load, which is equal to the sum of both wheel loads. Truck or car tyres with normal tread patterns will not affect the accuracy of the measurement.
As a tyre passes over the Lineas sensor, it generates horizontal, vertical and lateral forces between tyre and sensor. Due to the special sensor design, only vertical forces are measured. Lateral and horizontal side forces, between road and sensor, are decoupled by a special elastic material around the sensor. There are no errors caused by volume effects.
Spacing between two Kistler Lineas piezoelectric cable sensor rows
The spacing between two Lineas sensor rows depends mainly on the speed of the vehicles. As the main body oscillation frequency is between 1.8 and 3.5 Hz, it is recommended, from experience, that the spacing should be within 3 to 5 m. Both theoretical calculations and practical field implementations need to take this into account for spacing determination.
The frequency spectrum strongly depends on the vehicle. Air suspension vehicles are nearer to 1 Hz than 2.5 Hz for body oscillations and their damping characteristics are different from steel suspension types. Full or partial loading of vehicles severely affects the oscillation spectra and there may be remarkable differences according to various vehicle populations per road and country. Further to body vibrations, the axle hop and tyre natural frequencies are higher and important too.
The overall accuracy is potentially influenced more by pavement unevenness than by a spacing of 5 m instead of 4 m. Even the slightest bumps many metres ahead of the sensors may give rise to oscillation, so the flatness between sensor and surrounding pavement is essential. An important practical aspect is that signal integration requires the individual velocities between the two sensor rows for each vehicle. Thus the longer the spacing, the higher the errors caused by acceleration or deceleration, namely at lower speeds.
Kistler's Planning Manual provides recommendations on distance between two rows of piezoelectric cable sensors.
Table: Spacing distance estimation for piezoelectric cables
|Expected average speed||Recommended distance|
|45 - 75 km/h (30 to 45 mph)||3.5 m|
|75 - 95 km/h (45 to 60 mph)||4.0 m|
|> 95 km/h (>60 mph)||4.5 m|
Source: Kistler 2004
Kistler 2004, Planning manual: planning of a WIM (weigh in motion) station type 9195E, 002-300e-07.04 (200-348e), Kistler Holding AG, Winterthur, Switzerland.