More about SiWIM system

More about Hardware

Based on simple strain gauges, precise amplifiers, fast signal converters and a reliable computer, SiWIM system is a set of basic components, which together form a high-tech, advanced, reliable and scalable system for use in a wide range of different situations, both in the areas of WIM as well as for the analysis of bridge structures.
 
To weigh vehicles in motion, SiWIM® Bridge weigh-in-motion system uses instrumented existing bridges or culverts in the road network. In general, strains are measured on the main longitudinal members of the bridge to provide records describing the behaviour of the structure under the moving vehicle load.
 
Each strain transducer is equipped with 4 strain gauges in a full Wheatstone configuration. They measure strains, i.e. elongations and compressions of the structure (ΔL) between the two anchors placed approximately 200 mm apart (L), under the load applied. Strain transducers are bolted into steel anchors fixed in holes in concrete or on steel mounting plates glued onto the surface of the bridge.
 

More about Software

SiWIM’s proprietary software, divided into areas of collecting, adjusting, processing, controlling, analyzing and presenting data, is designed with the user in focus. Sophisticated algorithms are used to significantly ease up all necessary processes, while retaining on demand access to all parameters with extreme expandability.

SiWIM® software comprises 5 main independent components, coded for and operating in the 32-bit or 64-bit Microsoft Windows environment.
 
SiWIM® Engine (SiWIM-E) is a standalone application that runs on the on-site computer and performs and evaluates the measurements, calculates influence lines and stores raw and summarized data on the vehicles. It also collects activity, warning and other messages in the log and transmits them to the Front-End.
 
SiWIM® Front End (SiWIM-F) adjusts weighing parameters, displays the results and serves for site calibration, on- and off-site control and data collection.

SiWIM® Data processing software (SiWIM-D) is used for post-processing and evaluation of measured data.

SiWIM® Supervision software (SiWIM-S) provides web-based comprehensive checking, control and off-site analysis of the systems present, together with alarms processed by the software. It also provides a backup facility for aggregated data from all measured sites.

SiWIM® Monitoring software (SiWIM-M) is another web-based application used for preselection. It feeds the live traffic data flow from a connected site to a web browser.

Its main tasks are to acquire data from strain transducers, to process the signals, to calculate the influence lines, axle loads, axle spacings, vehicle classes and categories, and to process and transmit data to SIWIM-F.

An enforcement officer sets the initial parameters and then the system selects overloaded trucks from the traffic flow.

Report types, chart designs and data ranges can be selected according to the needs and are configurable.

System setup

As entirely portable, the SiWIM® system can be installed and configured in just a few hours, without any interfering in or even closing the traffic. It also does not interfere with the sensitive wearing course of the pavement, and despite that it is capable of monitoring, collecting and transferring data according to the requirements on the day of installation.

Every WIM system has to be installed prior to operation. The pavement WIM systems are, regardless of the sensor technology applied installed into the pavement, which involves stopping the traffic for several hours in each lane and use of heavy machinery such as pavement cutters and milling machines, thus extending the installation process over several days or even weeks. Apart from that, installation process requires cutting and milling the pavement. It also depends on dry weather and temperatures well above zero degrees centigrade.
 
SiWIM® system`s main advantage is its portability and fast installation which is typically completed in less than 4 hours. The sensors are attached to the bridge soffit either by means of anchors in holes or, when drilling holes is not allowed, by attaching the strain transducers to steel mounting plates that are glued on the surface of the superstructure.

Bridge Selection

Vast majority of bridges are suitable for WIM measurements, but it is vital to specify and understand the user requirements before selecting a bridge. Even if characteristics for suitability are not fully fulfilled, proprietary software, supported by advanced algorithms, elaborated calibration and post-processing procedures, considerably extends the range of suitable structures.

Algorithm, hardware and software design and implementation allows the SiWIM® bridge WIM system to instrument different types of bridges, ranging from short slabs to very long span bridges. An important bridge selection factor is its ability to perform bridge WIM measurements without axle detectors (Free-of-Axle Detector or Nothing-on-the-Road).

Characteristics that define suitability for bridge WIM measurements include bridge structural material, bridge and/or span lengths, boundary conditions, thickness of the superstructure, type of structure, and susceptibility to temperature effects, road roughness, skew, and susceptibility to dynamic loading.

A general rule of thumb is that the optimal span length of bridges used for SiWIM® measurements is between 6 to 20 m. However, due to other factors that influence the measurements, longer spans can provide equivalent or even superior accuracy and reliability of results. For spans with secondary elements, spans over 30 meters can provide results of similar quality, especially where traffic density is reasonably low and where the primary focus is on the vehicle gross weight.

Structural type

From the structural point of view, most bridge types can be used for SiWIM® measurements, provided that certain limitations about the geometry, pavement conditions and user requirements are observed. The acceptable structural types of a bridge include:

— slab bridges,
— beam-deck bridges,
— box girder bridges,
— orthotropic deck bridges.

System installation

After measuring the bridge dimensions, selection of location and installation of strain transducers is the single most important procedure of the system installation. Their exact position depends on the type of the structure and configuration of the traffic lanes. Cabling, temperature sensor and cabinet position only round up the whole procedure.

Every bridge WIM measurements start with bridge instrumentation. The usual procedure is as follows:

— Measurement of the bridge dimensions
— Installation of ST-503 strain transducers
— Installation of temperature sensors
— Installation of system cabling
— Installation of the SiWIM® cabinet

Measurement of bridge dimensions
Before installing the strain transducers, axle detectors, electronics and other SiWIM components, it is necessary to measure all key dimensions of the bridge.

These include:

— span(s) length(s),
— depth/thickness of the superstructure (slab, beams with deck, etc.),
— width of the superstructure, from below to define sensor positions and from the top to estimate the vehicle trails which may influence the sensor locations,
— angle between the abutments and direction of driving.

To improve efficiency of work, it is recommended to supplement the use of a measuring tape with the use of a laser measuring device. Based on the measurements, a detailed drawing or sketch should be done.

Calibration

In order to avoid significant errors in determining the load undergone by a pavement and thus calculating the expected pavement life, the data collected at the SiWIM® sites need to be as accurate as possible. Trucks of known weight are used to calibrate the SiWIM® system under any of the COST323 specified test plans to achieve the highest possible confidence level and targeted accuracy.

SiWIM® system, like any other weigh-in-motion system, has to be calibrated. Trucks of known weights are used to calibrate the SiWIM® system according to the COST323 specified test plans, to achieve the highest possible confidence level and target accuracy. Calibration is used to ensure that the static weight estimates produced by the weigh-in-motion (WIM) system are as close as possible to the reference weights.

Calibration also mitigates the effects of site conditions, such as varying pavement temperature, its conditions and vehicle speed. These factors can affect considerably the calculated axle loads. System accuracy can be increased by defining calibration parameters for different groups of characteristic vehicles (i.e. rigid, articulated, buses etc.).

Like with all other WIM sensors, only calibration of the entire measuring site (bridge and SiWIM® setup) provides a clear indication of the accuracy of the weighing results. In the scope of the calibration procedure, the axle loads of reference statically weighed vehicles are compared to weighing results from multiple passes of the same vehicles over the SiWIM® system. Trucks of known weight are used for this purpose, according to one of the test plans set out in the European WIM specifi cation (COST323). 

The test procedure is selected based on the target accuracy and confidence in the results. The higher the demands, the more elaborate, timeconsuming and expensive test plan is needed. For simpler calibrations, recommended for periodical checks or short-term installations with accuracy class C(15) or higher, measurements in repeatability conditions (with 1 calibration vehicle only) are sufficient. If possible, number of calibration runs should be more than 10, as specified in the COST323 specifications, especially if influence of  vehicle speed on accuracy is observed.

Initial calibration of an installation, especially for target accuracies of class B(10) or better, requires calibration in limited reproducibility conditions (with 2 to 10 pre-weighed vehicles). Depending on the target confidence level, this results in 20 to 110 calibration runs.

Vehicles extracted from the running traffic by the police and weighed statically can also be used for SiWIM® calibration. If static weighing sessions are repeated, the static weighing results are used for permanent recalibration of the SiWIM system.

Accuracy of the measurement

The accuracy of the SiWIM® system results depends on the type of the structure, the quality of installation, the type of calibration and particularly on the evenness of the pavement on the approach to the bridge. Achievable accuracy classes range from the excellent class A(5) on ideal sites to the still acceptable class D(25) on less ideal bridges. Under average circumstances classes B(10) or C(15) are to be expected.

The weigh-in-motion system accuracy depends on two primary factors: variance inherent to the WIM system and site conditions. The inherent variance of the WIM system is a function of the specific technology employed within the WIM system. In the case of SiWIM®, site conditions mean bridge characteristics, including its structural type, size and deterioration, susceptibility of the bridge to temperature variations, pavement smoothness, etc. The latter excites bridge-vehicle interaction, through dynamic oscillations of the vehicles, the intensity of which also depends on the vehicles’ suspension type and condition, their weight, speed, etc.
 
The accuracy of WIM systems is defined by the closeness or degree of agreement of an estimated measured value to an accepted reference value, defined typically within a 95% confidence interval. In other words, to be within a certain accuracy class, 95% of measurements errors should be less than those specified for that accuracy class.

SiWIM® results have a potential to be more accurate than any other type of WIM system, due to a much longer weighing platform. Unevenness (bumps or settlements) in front of the bridge are the most common cause of reduced accuracy as they are the main source of velocity-dependent results of weighing: vehicles are pushed up depending on their suspension and speed which affects their actions on the bridge. SiWIM® software provides tools to attenuate the dependence of the results on the vehicle velocity.

To evaluate measurement accuracy, WIM results should be compared to the weighing results of the same vehicles obtained on a scale with an accuracy of at least 3 classes superior than that of the WIM system, which means that the only realistic option for the reference weighing of accurate WIM systems is to use static scales. Static weighing is therefore a general requirement for calibration and accuracy checks.
 
SiWIM® system offers built-in tools for calculation of accuracy classes according to the European WIM specifications - COST323. These specifications denote an accuracy class with a letter and a number in the parentheses. Class A(5) is the most accurate one and is followed by classes B+(7), B(10), C(15), D+(20), D(25) and E(30).
Different applications of WIM data require different minimum levels of accuracy of WIM results. For example, assessment of effects of traffic loading on pavements often involves mathematical formulas with fourth or even higher power of the axle loadings (the equivalent single axle loads- ESALs). This heavily amplifies the effects of poor calibration and leads to significant errors in determining the load effects, expected pavement life, etc. For example, with the most commonly used power of 4, a 5% error in the calibration factor results in a 20% error of estimated load effects, a 10% error produces a 45% error of estimated load effects, etc.