In the United Kingdom, Heavy Goods Vehicles, by law, have to have a gross weight limit which cannot be exceeded for safety reasons. In order to ensure that drivers stay within the law, their vehicles must be accurately weighed after they are loaded and before they are driven. There is also a need for providing automatic checks on the weight of loaded vehicles as they are being driven. Although weighbridges exist in most major towns and cities for this purpose, there are times when weighing must take place in remote areas where a weighbridge is not available.

To overcome this difficulty, portable weig ing apparatus has been developed, such apparatus comprising weighing pads, the arrangement being that to weigh a vehicle one of the said pads is located under each wheel. Each pad has a load cell transducer which reacts to the weight of the vehicle to give a reading of the load thereon, whereby the total weight of the vehicle (loaded or unloaded) can be indicated.

The load cells which are used in the known pads are based on existing technology and invariably comprise either of t o readily available transducer types, namely

1) Strain guage elements

2) Piezoresistive sensors both of which suffer from drawbacks which render them unsuitable for applications where lightweight portable

load cells are required.

Strain gauge elements must be precisely mounted on carefully designed spring like supports and both strain gauge elements and piezoresi stiv e sensors usually require complicated temperature compensation in order to operate over a reasonable range. In addition, neither transducer can measure the large loads involved without the use of a mechanical transformer (usually a lever system). These transformers reduce the direct load applied to the sensor but result in a bulky, heavy pad with a vertical dimension which is so large that ramp devices have to be used in order to mount the vehicle on the pads, and such mounting is a difficult operation. The mounting of the pads can cause destabili sation of the load being weighed, and result in costly delays in the weighing process. As well as the aforementioned shortcomings of the existing apparatus, the task of accurately aligning the pads with the vehicle wheels is difficult and time consuming.

An object of the present invention is to provide a weighing apparatus which, at least in its preferred form is free of the disadvantages of the kno apparatus.

Certain specific objectives of the preferred form of the invention are that the apparatus should

1) be capable of weighing loads in the range of 0 to 9000 kg, with an of 0.5% fsd (45 kg in 9000 kg)

2) have a height in the range of 30 mm or less

3) have associated hardware to be capable of processing data from up to six apparatus

H) be capable of operation in a temperature range up to military standards

/GB88/00750

5) be of a cost comparable with or less than existing apparatus

According to the present invention there is provided weighing apparatus comprising an optical transducer material which' is stressed depending upon the load being weighed so as to vary the output of a beam of light passing through the transducer material in accordance with the load applied, characterised in that the apparatus is for measuring vehicle weights and loads and comprises a flat plate structure for placement under a vehicle wheel, so that the vehicle load is applied transverse to the plane of the plate structure, and including means for projecting a beam of light through the optical transducer material in a direction in the plane of the -plate structure, and beam receiving means for receiving the beam of light as it leaves the optical transducer material for the analysing of same.

It is known to use optical transducer material for load or stress measuring and examples of known apparatus are disclosed in the following Patent Specifications; German Patent Specifications 3,129,847 and 2,633,178; European Patent Specifications Nos. 0119592 and 0120999; U.S. Patent Specification 3,971,934; British Patent Specification No. 2,111,227; and Japanese Patent Specification No. 61-41933, but in such prior art arrangements the utilisation of an optical transducer in a flat plate for vehicle weighing has not been indicated or suggested, and the inventors of these prior art arrangements have not conceived the attendant advantages which can be achieved by the utilisation of an optical transducer in a flat plate structure.

In a preferred feature of the invention, the said means for projecting comprises a light source providing a

beam of plane polarised light.

Preferably also, the said optical transducer material is a birefringence material which doubly refracts the polarised light causing the polarised light to travel at different velocities in two principal planes, so producing optical interference.

The photosensitive means comprises an output polarizer for determining the degree of polarisation of the light emerging from the material, and a photosensitive diode arrangement.

Preferably, the light source is arranged so that the plane of polarisation is set at an angle to the plane of- the plate structure, and*the output polarizer is set at 90° to the said plane of polarisation of the- light source to give zero light output for zero load on the material.

By the invention, a relatively thin plate structure can be used to form the basis of a weighing apparatus for vehicles, and initial tests suggest that such a plate structure will not be required to have mechanical transformers. It is therefore believed that the overall height of the plate structure need be no greater than 30 mm, and at such height it is not necessary for additional ramps to be used to enable a vehicle drive to mount the plate structure from a standing start.

A suitable optical transducer material which is befringent comprises the plastics material PERSPEX pexiglas GS, but it is believed that other plastic polymer sheets may be used. Experiment has shown however that polymers which have a tendency to creep under load are not suitable as they do not repeatedly, over a period of time, give sufficiently accurate

results .

The basic features and an embodiment of the present invention will now be described, by way of example, with the assistance of and with reference to, the accompanying drawings, wherein:-

Fig. 1 is a graph of a compression characteristic of a polymer material ;

Fig. 2 is a perspective view of a plate of polymer located betweeen two polarisors;

Fig. 3 is an end view of the arrangement of Fig. 2 looking in the direction of arrow A and illustrates the vector magnitudes of the two rays as they progress through the arrangement of Fig. 2;

Fig. 4 is a perspective diagram showing how the two waves of polarised light are out of phase;

Fig. 5 is a perspective view of a test apparatus set up to test the inventive concept;

Fig. 6 is a circuit diagram of the pin diode shown in Fig. 5;

Fig. 7 is a graph of load output voltage as obtained from the tests carried out on the apparatus of Fig. 5;

Fig. 8 is the graph of the sine squared waveforms of the graphs of Fig. 7 ;

Fig. 9 is a graph similar to Fig. 7 but shows the results obtained when using load cells of increased surface area ;

Fig. 10 is a graph similar to Fig. 7 but shows the results obtained when a quarter waveplate is used;

Fig. 11 is a diagrammatic perspective view of a weighing pad according to the invention;

Fig. 12 is a sectional elevation of part of the pad sho n in Fig. 11;

Fig. 13 is an enlarged perspective view of the optical transducer system shown in Fig. 12;

Fig. 14 is a side view showing how the pad of Fig. 11 is used in weighing a vehicle; and

Fig. 15 is a graph of a load 'cell c aracteristic showing how only part of the characteristic may be used.

A stressible translucent or transparent material is necessary for the present invention, and as glass is unsuitable, a plastic polymer has to be selected, and when the expression "optical transducer material" is used herein it is intended to mean only such material which possesses the necessary characteristics to enable the weighing apparatus to perform satisfactorily.

A fundamental' property of semi-crystalline polymers is that when they are subjected to pressure they compress and spread over a period of time as their molecules slip relative to one another. This spreading action, known as creep, means that a static load applied to a polymer causes the material to initially o ey Hookes Law, and then continue compressing for a period of time which can vary from fractions of seconds to months or years depending on the polymer type. To a first approximation, the compression characte istics of a

polymer follow the first order response of 'charge and decay' as shown in Figure 1. This info mation is included in order to indicate that, as will be clear from the following, materials which have excessive creep are unsuitable for the present invention.

The invention uses a light transducer for detecting loads for providing an indication of weight, and specifically if a polarised light is projected through a piece of birefringent material and the state of polarisation of the beam after it has left the beam is determined and analysed it can yield information on the amount of stress placed upon the material. This is a known phenomenon and in the following birefringence is explained as is the adaptability thereof to use in a weighing apparatus.

A birefringent material divities an entering beam of plane polarised light into t o beams which are polarised at right angles to each other, if the direction of entry of the light is not parallel to the optic axis of the material. One beam will obey Snell's Law and will propogate through the material at a constant velocity, independent of direction and external forces, and is termed the 0-ray (Ordinary Ray). The other beam violates this Law and propogates with a velocity which can be influenced by optical, electrical or mechanical means, and is termed the E-ray (Extraordinary Ray). If a sample of birefringent material is cut as a plane-parallel plate, and the beam- is made to pass through the plate in its plane the E and 0 rays will recombine upon emergence to form an elliptically polarised beam. By applying stress to the material, the state of polarisation can be modulated and it is this modulation which is used in the preferred embodiment of the invention to measure weight .

Figure 2 shows a narrow plate of birefringent material placed between two polarisers which are orientated with their planes of polarisation at 90° to each other and respectively at angles øand 90° -0 to the direction of applied stress.

As the polarised light enters the material it splits into two orthogonal waves which propogate through the material at different speeds due to birefringence. Figure 3 shows an end view on A which illutrates the vector magnitudes of the t o rays as they progress through the system.

Vector 0A is the incident light polarised at angle to applied stress. This has component parts OB and 0C given" by

OB ₌ OA.sin €^ 0C = OA.cos

After travelling through the material and reaching the second polariser, only those components of OB and 0C in its plane of polarisation pass through to result in two terminal vectors of magnitudes

0D = OB. cos OE = OC.sin ®

which in terms of the incident light 0A are

OD = OA.sinβ .cos0= (0A.sin2θ)/2 OE = OA.cos© .sin©= (0A.sin2θ)/2

Although the above terms give magnitude information on the t o light rays, phase information is alsc needed before the nature of the stress related interference

can be determined. The difference in speeds between the two waves as they pass through the material gives rise to a phase difference £ as shown in Figure 4. As OD and OE are components of the two waves shown, they also experience a relative phase shift which is dependant on the applied str.ess. If the light is of wavelength then the relative retardation R produces a phase shift of

£f ^' = 2.pi.R/7

and the two resultant vectors OD and OE become

OD = (0A.sin2S)sin(wt) 12

OE - (0A.sin2Q ϊsin(wt-2.pi.R/' )/%

The relative retardation R depends on the stress-optic properties of the material- and is given by

R = C.l.S

where

C = Stress optic coefficient of the material 1 = Optical path length in the material S = Applied stress

resulting in

OD = (0A.sin2Q ) .sin(wt) 12

OE = (0A.sin2Q ) . sinfwt - 2. pi.C. )/ _Q

By vector addition the light output from the second polariser will be OF = OD - OE

= (0A.sin2 β ) .sin(wt) 12

- (0A.sin2&) .sin(wt-2.pi.C.l.S -fi.)/

= 0A.sin2 &.sin(pi.C.l.S/* ) .cos( wt - pi.C.l. S /A )/

It can thus be seen that the light output is harmonic in nature with an amplitude that varies with applied stress. As light intensity is dependent on the amount of energy that the light posesses and the energy of a harmonic system is proportional to the maximum amplitude that the system can sustain, then the output intensity is given by

Int = 0A ² .sin ² (2θ).sin ² (pi.C.1.SΛ^ )/ι ⁽ 3)

A transducer based on the. principles described above was constructed and tested to determine its suitability for use in a thin plate structure as a weig ing apparatus (or load cell). In order to simulate the large weights needed, a hydraulic press with a range of 0 kg to 30,000 kg was used throughout the tests.

Equation 3, above, shows that the theoretical output from a load cell which utilises optical biref ringance as a measurement method is cyclic and dependent on the cell's dimensions, stress-optic characteristics, the optical arrangement of the light source and detector and the amount of stress or load applied to the cell.

As the transducer effectively measures stress and is not dependent on relatively large dimensional changes as in the other types of load cells such as capacitance and piezoelectric load cells, the material used in the cell should ideally have a large modulus of elasticity and exhibit a low creep factor. Tests were carried out on samples of transparent polymer to determine their mechanical suitability, after which further tests were performed to both verify the response of such a load cell and determine the stress-optic coefficients of the materials .

Samples of four different types of polymer were initially considered for use as the birefringent material in the transducer. Each sample was placed under an initial load of 5000 kg in a press to determine how much indication of apparent variation in applied force would occur over a given time period. This variation was used to give a measure of both the modulus of elasticity E and the creep factor of the materials. The c aracte istics of the materials differed widely, from 100% creep in 10 minutes for one material to less than 5% creep in 72 hours for another. Since the output from the load cell must be constant over a period of time (for a constant load), the latter of the above materials, a grade of PERSPEX was selected and subjected to tests to determine its optical characteristics. ■

The theoretical response of the material as described above assumes that the incident light is normal to the material edge face and that no losses occur due to reflection or refraction. In order to tr*y and satisfy this assumption, three test load Oells with the path lengths shown in table 1 were prepared by cutting, filing and smoothing the edge faces where light would enter and leave the material. Finally proprietry metal polish was used to remove the fine surface imperfections and produce a flat polished surface. Once the test cell preparation was complete, the experimental set up illustrated in Figure 5, and described hereinafter, was assembled.

Table 1

T _h e light source used was a 5 mW laser having a plane polarised output, and was mounted on an optical bench set at an angle of 42° to the plane of applied stress, w _h ilst the polariser used to determine the state of polarisation of the output beam was orientated at 90° to the laser and -48° to the plane of stress. This degre.e of orientation between the laser and the polariser was. chosen so that for zero stress the light output from the system would also be zero, ^' as explained in the theoretical. - discussion. The intensity of the output beam was determined by a PIN DIODE detector, which converts light intensity to voltage in an op-amp feedback arrangement, as shown in Figure 6.

With the PIN DIODE in reverse bias as shown in Fig. 6, the diode 'leakage' current is linearly dependant on light and the output voltage is also a linear function of the light intensity, and is given by

Vout = R x 0.6 A/W

which for a 15v output (supply voltage) from an input of 5 mW (laser output power) results in a resistor value R of 4.7 koh s.

Each of the prepared load cells was positioned in the press and subjected to loads ranging from 0 kg to 2000 kg in 100 kg steps. Five minutes were allowed between

each load increment to test the cell response as a function of time. The load versus output oltage characteristic of each cell is plotted in Figure 7 and shows several interesting points.

1) The transducer output is cyclic but has a fairly restricted range as only up to the first peak is useable in practice.

2) Zero load did not produce zero-output volts as predicted .

3) The maximum output from the system was different for each cell tested.

4) The output voltage did not vary as a function of time, only with load.

Figure 8 shows a graph of a sine squared waveform with a maximum amplitude of 1 and a range of 90°. If the first half cycle of each of the three test cell outputs are normalised and plotted on Figure 8, then i can be seen that the load cell response ties in very closely with the expected response. In fact the responses do not vary by more than 5% at any time and even this can be attributed to innacuracies in the press load gauge and parallex error when reading said meter. By plotting the load at which the maximum output occurs as a function of path length or cell width, it can be shown that the range of the cell is inversely proportional to path length as predicted by equation 3. This implies that one way to increase the range of the transducer is to shorten the path length but as has already been explained, this load cell measures stress and as stress has units of kg per ² , another ay of increasing the measurement range would be to increase the cell area. In the test equipment, the press had a surface area of 0.0016m ² , but this was increased to 0.0123m ² by inserting suitably shaped mild steel blocks between the test cells and the press faces. Before the tests were duplicated with this large surface area, the stress-

optic coefficient C of equation 3 was determined from the output data already obtained.

The only relevant parameter not known in equation 3 is the stress-optic coefficient C. Path length, applied load and light wavelength were all known when the tests were performed. As the sine squared term has a peak at 90° or pi/2 the stress-optic coefficient was determined by equating the respective path length and load range to pi/2 as follows.

C.l.S.pi/ _/ K= pi/2

at peak load. Thus

C =Α/2.1.S

The S.I. units for stress are kg/m ² , so the applied load was normalised by dividing the load (in kg), by the press surface area (in m ² ) . This gave a stress range of

S = 370kg/0.0016 = 231250 kg/m ²

and stress-optic coefficient

C = 9.12 x E~ ¹² m ² /kg

With the increased surface area the three load cells were again positioned in the press and subjected to loads, this time in the range 0 kg to 10,000 kg in 500 kg increments. The results from this series if tests are plotted in Figure 9 and show an increase in transducer range from 570 kg to 2900 kg for the 150 mm long cell and 1100 kg to 8600 kg for the 50 mm path length. The offset in the origin of the characteristics

noticed in the first series of tests was also present during this series and appears to be a phase shift at zero stress, while the difference in the value of maximum output for each cell indicates that the amount of light power lost in each cell is different. The latter effect does not appear to be a function of path length and is probably insertion loss when the laser light enters the transducer but the phase shift at zero stress does vary with path length and could be caused by natural birefringence not related to stress. In order to gauge the effect of such zero stress phase shift on the transducer characteristics, a quarter wave plate was inserted between the laser output and the tes.t load cells which were again loaded as in the previous tests.

The quarter wave plate alters plane polarised light to

•elliptically polarised light by introducing a 90° phase shift between the two orthogonal light vectors as described above. The effect of this on the theoretical response of equation 3 is to offset the characteristic so that zero stress produces an intensity of Imax/2 as shown below.

Int = 0.A ² .sin ©.sin ² (C.1.S.pi ^+ pi/4)

as the 90° phase shift effectively produces a 45° shift in the expression due to the factor of 0.5 introduced into the equation as described above. At zero stress the expression reduces to

Int = O.A ² _i sin θ .sin (pi/4) = Imax.sin ² (pi/4) = Imax/2.

With the quarter waveplate in position, each cell was again tested and the results are plotted in Figure 10.

A comparison between these results and of Figure 9 shows that the output follows the same sine squared function but there is an additional offset of 45° as expected.

Each of the three load cells was loaded and unloaded three times to deter ine if the output was both repeatable and free of hysterises. The results from these tests are presented in tabular form in tables 2, 3 and 4 as a graphical representation would be confusin-g and difficult to interprate. The analogue meter on the press probably introduces sufficient error in the load readings to account for the small discrepencies apparent in the results. This, coupled with the fact that zero stress appears to produce the same output suggests that the transducer response is both repeatable and does not suffer from hysterisis.

LOAD kg Vout 1 Vout 2 Vout 3

B B

Column A shows oscilloscope readings while load was being applied from 0 kg to 9500 kg.

Column B shows' oscilloscope readings while load was being released from 9500 kg to 0 kg.

LOAD kg Vout 1 Vout 2 Vout 3

A B A B A B

Column A shows oscilloscope readings while load was being applied from 0 kg to 9500 kg.

Column B shows oscilloscope readings while load was being released from 9500 kg to 0 kg.

LOAD kg Vout 1 Vout 2 Vout 3

A B A B A B

Column A shows oscilloscope readings while load was being applied from 0 kg to 9500 kg.

Column B shows oscilloscope readings while load was being released from 9500 kg to 0 kg.

It can be seen from the above therefore, that the present invention conceives the utilisation of an optical transducer in a plate structure for the weighing of vehicles, a preferred feature of the invention being that the plate structure will be as thin as possible to enable the vehicle from a standing start to mount the plate structure in a much simpler fashion than is possible with the present plate structure weighing devices.

The invention furthermore envisages that the variation in light intensity of a beam which is passed through the plate structure in the direction of its plane is utilised, preferably with polarisors as described and as further explained with reference to an embodiment, to provide electrical output signals indicative of the stress applied to the plate structure by virtue of the vehicle load. The processing of the electrical output signals ca be effected ϊ-n any suitable manner by an appropriate microprocessor which it is believed will be capable of design and construction by a person skilled in the art. The processor can provide any desired form of output, but typically may give a direct reading of vehicle weight on each wheel, and in this connection the complete assembly may require a plurality of the pads or weighing apparatus according to the invention, appropriately linked to the microprocessor. The microprocessor may furthermore integrate the readings from the respective pads to provide a total indication of weight load and weight distribution throughout the vehicle. Figure 14 shows diag amm atically how the vehicle stands on the pads when weighing is taking place.

In an alternative arrangement, the pads can in fact be embodied in a road surface so as to detect the weight or impact loading of vehicles passing over the pads,

and so that such vehicles can in fact be detected when travelling on the highway.

The actual pads which are constructed according to the invention can take any form withi the general principles of the invention, and for example a pad may be made up of a thin sheet of optical transducer material, but as an alternative construction as shown in Fig. 11, a pad 10 comprises a sandwich of upper and lower metallic sheets 12, 14 between which are four optical transducer devices 16, of identical construction, and connected in parallel so as to receive input light from an input fibre optical cable 18, and to provide output signals through output fibre optic cable _, means 20. Each transducer operates in identical fashion, and the input signals and output signals may be sequenced so that the transducers are scanned in turn. The remaining space between the plates 12 and 14 may be filled by any suitable material such as an epoxy resin or by a steel plate or the like.

The form of transducer arrangement is shown in Figs. 12 and 13, and will be seen to comprise a block 22 of the optical transducer material of the type for example as described herein, such block being provided with a raised stressing area 24 which is in contact with the plate 12, so as to receive the loading stress when the vehicle sits on the pa ^* d as shown in Fig. 14. At the ^' left or input end, the block receives a fibre optic cable 18 and a plane polarising device 26 is received in a slot 28 in the block 22 to the left hand side of the central portion of the block having a face 24. At the other side, the output side a recess 30 receives a quarter wave retarding plate 32, and polarising beam splitter 34 which separates the t o terminal vectors of plane polarised light as discussed hereinbefore so that they can respectively be outputted on two fibre optic

22

cables 36, 38 and the intensity of said vectors of plane polarised light are sensed by two pin diode detectors 40 and 42 which are outputted to amplifier 44, 46, and thence to an appropriate microprocessing apparatus for processing the signals as hereinbefore described.

Fig. 13 shows the arrangement in perspective elevation.

Operation of the transducer will be well understood from the foregoing, but briefly, the output of a light emitting source 48 is applied to a polymer fibre optic cable 18 the output of which is applied in turn to the respective transducers 16 in the plate structure 10 and, referring to the drawing of a single transducer as shown in Fig. 13, the light from a fibre optic cable 1 _. 8 is passed through the plane polariser 26 then through the operative portion of the block 22 of optical transducer material where it is subjected to a division as explained h'erein dependent upon the stress applied on face 24 due to the weight the vehicle, and then through a quarter wave retarding plate 32 and finally to a polarising beam splitter 34 to separate said terminal vectors of plane polarised light. The intensities of these two vectors are detected by the diodes 40 and 42 and appropriate electrical signals are generated via the amplifiers 44 and 46 and are processed to give an indication of the weight of the vehicle.

The work carried out on the optical transducer described herein indicates that the invention provides a new advanced form of weighing apparatus, for weighing heavy goods vehicles. The non-stress related offset of the transducers characteristic needs to be taken into account in designing a load cell employing an optical transducer in accordance with the invention to quantify

the transducers.

The sine squared output of the cells (Fig. 8) is probably too insensitive to load changes around the lower end of the characteristics to be usable without linearisation. By int oducing an offset in the ^' characteristics (by choosing a suitable path length of inserting a quarter wave plate between the first polariser and material as described herein), zero load can be made to produce an output at the mid-point of the cell characteristic. If the load cell range is larg.e enough the output could be made practically linear over the desired operating range of 0 kg to 9000 kg (see Figure 15).

As has been shown herein, the cell ra.nge ..can be extended by increasing the surface, area, reducing the path length and altering the wavelength of the light source. For a practical load cell using the material tested during this investigation, the surface area can be easily increased to extend the range to 25,000 kg, assuming the load can be evenly applied across the area needed. If successful, the voltage output from the system can be easily altered with operational amplifier arrangements to still produce zero volts at zero load and 2.56 volts at maximum load. In order to produce a compact, practical load cell, an alternative light source to the laser as described in relation to the embodiment given herein, would be used. Since the light beam must be fixed in relation to the cell and detector, the source, polarisers and quarter wave plate should be precisely mounted. One way of accomplishing this is to use fiber optic cable to couple the source and detector to the load cell. This arrangement would increase the cost of the cell, but it could be justified if it simplified the construction and setting up procedure of the finished product. As any laser

arrangement is expensive, an alternative light source could consist of a high power diode with separate polariser .

The work done has indicated that a grade of PMMA (which is more commonly known by its trade mark Perspex) known as pexiglas GS which offers excellent low creep characteristics is suitable for use in the optical load cell. The optical cell has the advantage that its output parameter is a function of bulk density changes rather than dimensional changes as required in for example a capacitance cell. The cell has been shown to be capable of measuring extreme static loads with little or not hysteresis and favourable temperature stability. Furthermore, this type of cell offers the advantage that operational errors due to ageing, temperature offsets etc may be nullified under microprocessor control since the working range of the cell is always constant.

The maximum cell dimension in the thickness direction may be as little as 6 mm, well within the preferred specific objectives of the invention, although the final manufactured design may prove to be approximately 20 mm. The polymer substrate used in this cell is extremely robust, lightweight, cheap and can be easily shaped. Furthermore, the cell offers a high resistance to mechanical shock and chemical attack by acids and many common solvents which makes it uniquely well suited for use in hostile environments.

A complete set of apparatus will comprise a plurality of load cells for location under vehicle wheels and a common processor, but producing a production version would ^' not appear to provide any special problems. It is recognised however that there will be load plates to each side of the polymer sheet or plate.

2

The weighing apparatus according to the invention can, without modification, because of its thin plate structure nature, be used for the weighing of static loads such as tanks and the like especially where such loads are located in areas having little extra head space, because the thinness of the plate structure enables it to be placed ^' under the load without substantial loss in head space.