WO2004065970A1

WO2004065970A1 - Method and device for calibrating the speedometer of a vehicle

Info

Publication number: WO2004065970A1
Application number: PCT/FR2003/003733
Authority: WO
Inventors: Marius Renoult
Original assignee: Marius Renoult
Priority date: 2002-12-18
Filing date: 2003-12-16
Publication date: 2004-08-05
Also published as: FR2849207A1; FR2849207B1

Abstract

The invention relates to a method and device for calibrating the speedometer of a vehicle. The inventive device involves the use of a simulator consisting of a roller test bench comprising at least one roller which is driven by a wheel (RR) of the vehicle, and a computer unit (PC) which receives information relating to the rotation of the roller and the rotation of the wheel (RR). The computer unit (PC) comprises means for determining a calibration curve using the aforementioned information and vehicle-related data, said curve being obtained from the sum of the on-road creep rate and the on-road tangential creep rate.

Description

Method and device for calibrating a speedometer fitted to a vehicle

The present invention relates to a method and a device for calibrating a speedometer, for example a tachograph, fitted to a vehicle, this calibration being able to be carried out, either during the installation of the tachograph on the vehicle , or during periodic inspections, it being understood that in both cases the inspection procedures are slightly

15 different from the fact that:

- during the installation of a tachograph in a vehicle, calibration requires knowledge of some of the characteristics of the vehicle as well as its equipment ^pneumatic, 20 - during a periodic inspection, the test results in of the installation or of a previous periodic control can be advantageously used.

The invention is not limited to a particular type of vehicle. It concerns both trucks equipped with tachographs as well as public transport vehicles (bus, coach), taxis (light vehicles) equipped with taximeters - or even private vehicles (insofar as they are equipped with devices. speed measurement and / or recording).

30 In general, we know that at present the speed indicators provided by equipment manufacturers are designed to provide indications of speed obtained by counting the number of revolutions per unit time of the vehicle wheels or of a rotating member of the transmission chain between the engine and the drive wheels of the vehicle; for example, the gearbox output shaft.

The calibration of these indicators requires the use of a simulator comprising at least one endless roller or belt hometrainer test bench of which at least one roller (or band) is designed to be rotated by a wheel of the vehicle, and a unit for calculating and managing the simulator which receives the information relating to the rotation of the roller or the strip and information relating to the rotation of the wheel from the tachograph. This calculation unit measures the difference between the speed of rotation of the roller and the speed of rotation of the wheel and acts on the settings of the speedometer so as to reduce this difference to zero.

It turns out that, physically, the method and the means used until now to carry out this calibration are suspect of representativeness, since a precision is to be respected in all the conditions of use. Indeed :

- the rolling circumference measured is that on the rollers (without correction),

- the vehicle is empty,

- the load on the roller is a half load, oblique and variable obliquity with the pneumatic equipment, - the mechanical contact takes place on a curved surface (stretched tread),

- the rollers turn "crazy" on their axis,

- there is no rolling drag and aerodynamic drag.

In fact, the reactive torsor in a test bench bearing is different from the reactive torsor in a road test bearing. In an attempt to correct this defect, it has been proposed to carry out a calibration by carrying out no-load and load-on-road tests, at various speeds. However, this process is too long and too complex to apply to all vehicles on the road (either millions or tens of millions of vehicles).

A more particular object of the invention is therefore a calibration process which can be carried out on conventional test benches with implementation times compatible with high control rates comparable to those practiced by the hour. current.

It starts from the observation that taking information from the rollover (the instrumental errors being neglected) all the anomalies or errors come from the conditions in which the rollover takes place, and in particular from the slippage (or micro-slippage of structure).

It considers three types of slippage giving rise to the following three “creep” rates:

. the rate of “creep” (free rolling) ζxR on the road which varies linearly with the load and is independent of the speed,. the rate of tangential (propulsive) creep ζxt on the road which varies non-linearly with the load and the speed,. the rate of total “creep” route ζxT which is equal to the sum of the rate of “creep” route and the rate of tangential “creep”.

It suggests using the total route creep rate as the error rate for each speed (in percentage).

To this end, the method according to the invention comprises the determination of a table and / or a calibration curve providing the value of an error rate ε% as a function of speed, this error rate consisting for each speed in the sum of the rate of “creep” (free rolling) on the road ζxR and the rate of tangential “creep” (propulsive) ζxt on the road.

Thus, in order to be able to take advantage of this mode of determination, the method according to the invention provides for the following successive phases:

- A preliminary phase during which, after having characterized the vehicle, the operator enters, on the computer system associated with the test bench, the characteristic data of the vehicle, in particular the characteristics of the tires (which are usually found on these tires) to be able to have parameters such as sidewall height, wheel radius, crushing, front / rear load indices, vertical stiffness Kp.

- A test phase during which the vehicle is placed in position on the test bench where it is first weighed at the front and rear axle (in the case where this information has not been entered during the preliminary phase from the information provided in the vehicle documents) this test phase comprising:

"a first calculation step allowing the vehicle to be personalized by determining in particular:

. the crushing of the front wheels, this parameter being obtained by making the ratio between the half weight of the front axle and the front vertical rigidity,

. the radius of the front wheels with crushing (rolling radius), ie the difference between the radius of the front wheels and the crushing previously calculated,. front running drag (product of the ratio between the weight of the front axle and the radius of the front wheels with flattening, by a bearing parameter entered by the operator and empirically defined by the manufacturers),. the vertical stiffness AR, Kp. the rear transverse stiffness, Kq. the crushing of the rear wheels (possibly twinned) (ratio between the weight of the rear axle over n times the vertical stiffness, n being the number of rear wheels),. the radius of the rear wheels with crushing (rolling radius), which is equal to the radius of the rear wheel minus the crushing,. the rear running drag (product of the ratio between the weight of the rear axle and the radius of the rear wheels with crushing, by the above parameter),

. the total rolling area which is equal to the sum of the front drag and the rear drag, knowing that there can be several rear trains, tractors or carriers with a carrier plus a tractor, it being understood that the rolling drag in practice does not vary with speed. On the other hand, it is necessary to add the aerodynamic drag which varies with the speed. This drag is normalized in a file for the chosen test speeds 20 km / h, 50 km / h, 80 km / h, 100 km / h with or without deflectors on the roof, thus reducing aerodynamic drag,

^• * • a second calculation step carried out when all the vehicle-specific preliminary calculations have been made, this second calculation step comprising the calculation of “creep” rates:

. "creep" rate (free rolling) ζxR on the road,. tangential creep rate (propulsive or traction) ζxt on the road, then the calculation of the error in%, that is ε% which, for each speed, consists in performing the sum ζxR + ζxt _.

ε% = ζxR + ζχt

This process enables calibration by simulation in which:

- a rolling distance is determined which constitutes a standard length,

- We then consider that this length runs under the wheels of the vehicle and we deduce by calculation the length which should be displayed on the tachograph as well as the error ε% for each speed and for each load in a road trip.

This process makes it possible in particular to demonstrate:

a) that the results obtained by simulation (from physical characteristics) are more precise than those obtained in an approved test center using rollers, b) that the results obtained on the road by the usual speed measuring instruments are erroneous, but that we can deduce the real value from the calculation.

Of course, the invention also relates to a device for implementing the method defined above, this device involving a simulator comprising a test bench with endless rollers (or belts) of which at least one roller (or a belt ) is designed to be driven in rotation by a wheel of the vehicle, and a calculator and management unit of the simulator which receives the information relating to the rotation of the roller or the strip and information relating to the rotation of the wheel coming from of the speedometer, this calculation unit comprising means making it possible to determine, from said information as well as information relating to the vehicle, a calibration table or curve providing the value of an error rate ε % as a function of speed, this error rate ε% being obtained for each speed by the sum of the “creep” rate (free rolling) on the road ζxR and the tangential “creep” rate (propulsive or traction) ζxt on the road.

Advantageously, this unit for calculating and managing the simulator may include means making it possible to enter data characteristic of the vehicle as supplied by the vehicle and tire manufacturer, this data notably concerning the radius of the wheel, the height of the sidewall, overwriting, front / rear load indices, as well as means for determining parameters comprising in particular:. crushing the front wheels,

. the radius of the wheels with crushing (rolling radius),. the front rolling drag,. rear vertical stiffness,. rear transverse stiffness,. crushing the rear wheels,. the radius of the rear wheels with crushing,. the rear rolling drag,. the total rolling distance which is the sum of the front rolling drag and the rear rolling drag and, optionally, the rolling drag of one or more trailers.

Likewise, the calculation unit may include means for determining a rolling distance on the rollers of the test bench and calculating the length which must be displayed on the speedometer, as well as the error ε for each speed and for each load.

An embodiment of the method according to the invention will be described below, by way of nonlimiting example, with reference to the appended drawings in which: Figure 1 is a schematic view of a test bench for the calibration of vehicles equipped with chronotochy graphs.

FIG. 2 shows in section a bearing scale bearing equipped with measuring instruments used in the test bench represented in FIG. 1.

Figure 3 is a schematic representation showing a wheel arranged on the rollers of a test bench.

FIG. 4 is a diagram representing the displacements at the level of the wheel / roller contact zone (if the contact was wireless).

Figure 5 is a schematic representation illustrating the Winkler model (the contact being non-radio).

FIG. 6 is an error diagram as a function of the speed, for a vehicle running empty, laden and laden with a trailer.

As previously mentioned, the method according to the invention applies to the calibration of speed measurement installations involving a tachograph connected to the output of the gearbox of a vehicle by means of adaptation means. adjustable designed to convert a magnitude W representative of the number of turns made by the output shaft

, of the gearbox, in a quantity K representative of a number of electrical pulses per kilometer according to the relationship:

at

W

where a is the adaptation coefficient. To measure the quantity W before adapting it to the tachograph, the vehicle is brought to a simulation bench such as that illustrated in FIG. 1. This test bench comprises drive wheel support modules in ME of the wheels RA, RR each comprising two pairs PA _l5 PA ₂ of support rollers Gi, G ₂ respectively assigned to the wheels RR, RE of the same train. The two rollers Gi, G ₂ of the same pair PAi, PA ₂ are oriented parallel to one another. These two rollers Gi, G ₂ are intended to self-center one or more wheels (here two RR twin wheels) of the vehicle as shown in Figure 3.

These rollers G _{] 5} G ₂ are mounted on bearings of which at least one is a weighing bearing PP (mass measurement) with bearing equipped with RI measurement means intended to allow the measurement of the speed and the distances traveled during the simulation.

Furthermore, each support module includes a CC sensor making it possible to count the revolutions made by the wheel RR during the simulation. All this information is transmitted to a computer system fitted to a PC control station.

In a process of this type all the disturbances originating from running affect the quantity w and then the quantity K in the same ratio and consequently on the information recorded by the tachograph.

The reactive torsor in driving on a surface with a curvature is different from that of road driving. Likewise, it is different depending on many other factors (type of tires, empty or laden vehicle, aerodynamic deflector, etc.)

The mechanical study of the contact surfaces between the wheels and the rolling surfaces shows that a rolling can be considered as a combination rolling, sliding and rotating, with the understanding that vehicle tires:

- can roll without sliding or rotation,

- can slide without rolling (locked wheels) - can roll and be rotated by turning.

After loading the vehicle creating defoπnations, it is necessary to take into account the tangential actions if there is a driving force transmitted between the two contact surfaces.

By definition, “bearings free” are bearings without tangential, driving or friction action.

The stresses in the tread consist of tensions when the wheel rolls on a curved surface and of compressions when it rolls on the road.

The particles of material of the elements in contact flow during a bearing with a speed V of rolling plus a speed δ V due to the deformations. The speed of the material elements of the deformed surface can be calculated using a Eulerian model.

The velocity field being stationary, the velocity at a point is independent of time, but is a function of the point f (M, t) which can be followed in the displacement.

- = - + U • grâd f relation 1 dt δt

Point function • different from that in Lagrangian description and which it would identify with partial derivatives. 004/065970

11

This equation makes it possible to determine the relations without dimensions.

relation 2

6 x ά y and represent the relative speed (non-dimensional) at a point

V V arbitrary of the contact surface taking into account the rate of "creep" and the deformations in Eulerian description of the movement according to the relation l.

The equality â 'x = ά' y - 0 corresponds in a region of adhesion to a rate of "creep" due to the microgliding of the elements in contact.

When two cylindrical bodies rotate on each other, their speed is obtained by the following relation:

V _R1 (l + ξ%) = V _Rl relation 3

V _R2 being the speed of the wheel fitted with a tire V _R] being the speed of the roller ξ% being the speed error expressed in%

In an elastic transmission, the speed of the driving body is greater than the speed of the driven body.

The reactive torsor in taxiing on a test bench is different from the reactive torsor in taxiing on a plane (for example a road). To express these results by a very simple example:

. in a simulation by running on a test bench which does not take into account tangential actions, the real result concerning the speed or the distance covered:

d_

d _m being the distance measured on the measuring roller.

It therefore appears that the problem which the invention aims to solve is solved insofar as it is possible to calculate the “creep” rates ζ.

For this purpose, taking into account the fact that the contact to be considered is not wireless, a Winkler model whose expression given below is used:

1 2 2

Uz, (x) = 0, UzJx) = δ ² = relation 4

^{1 2} 2R 2R

R: radius of the tire a: Vi length of the contact area

Because the materials in contact have very different elastic constants, there will be tangential displacements. The stress and displacement calculations will be carried out from the Winkler model according to the following expression:

Kn p (x) (a ² -x ² ) relation 5

2Rh p (x) is the load 004/065970

13

Kp: vertical stiffness h: sidewall height

Kp

P = - Ï (a ² -x ² ) dx

2rh

for a half contact width with a distribution (x) on the roller, the value of a is:

relation 6

with r = radius of the roller

The tangential surface displacements such as Ux, Uy are linked to the components q _x and q _y of tangential traction by the relation:

Uy relation 7

with Kq: transverse stiffness

By considering only a direction Ox in the equations of relation 2, as well as the condition that the traction is zero on the leading edge as used to find the distribution of the tangential traction along the contact surface, we obtain:

â x dUx

-o -ξ _x +

V dx

By substituting U _x , U _x = \ ξ _x dx 004/065970

14 we obtain:

f) < ^«« > relation 8

since the traction increases linearly from the leading edge to the driven edge, the expression Q becomes:

Kq 2Kq relation 9 h ~ x J-a h

However, in practice a slip occurs at the driven edge where the pressure drops to zero. For this reason, an adherent region of width 2c extends from the leading edge. The sliding then begins at point x = (2c-a).

With a parabolic load distribution on the contact:

p (x) - ^ (a ² -x ² ) hR

,, Ap, 4Kp q (x) = - -x 2c = - (a - c) c relation 10 ^y h hR

Kq_ R r y, = 2 Kp a a)

relationship 11

004/065970

The invention proposes more precisely 4iîf5e, nt to determine the error rate from the rate of "creep" (free rolling) ζxR on the road and the rate of tangential "creep" (propulsive) ζxt on the road so as to be able to correct the speed values recorded by the tachograph.

An example of calculation will be described below for an unladen vehicle whose parameters entered or calculated are as follows:

. Characteristics: wheels: 315/80 -22.5 inches - prescribed inflation pressure; 8 bars sidewall height: 315/80/100 = 252 mm wheel radius: (22.5 x 12.7 mm + 252 mm) l, 017 = 546.89 mm reference crush: 546.89 - 506 = 40.89

Load index AR 150 = 3,350 kg Load index AN 154 = 3,750 kg

37500N

Vertical stiffness AN: Kp -9l7,055N / mm

40.89mm

- AN transverse stiffness: Kq = 917,055 x - = 611,370 N / mm

- Train weight AN = 5,920 kg = 59,200 Ν

- Weight of rear axle = 4,460 kg = 44,600 Ν

- Crushing of wheels AN = 32.27 mm

- Radius of AN wheels with crushing (rolling radius) = 514.61 mm

- AN running drag with parameter 2,5 = 304.84 Ν

- AR vertical stiffness: Kp = 819.93 Ν / mm

- AR transverse rigidity: Kq = 546.15 Ν / mm

- Crushing of rear wheels (twin) = 13.61 mm

- Radius of rear wheels with flattening (rolling radius) = 533.28 mm

- AR bearing processing (parameter 2.65) = 221.62 Ν 04/065970

• - Rollover treatment = Trailed AN + AR = 526.46 Ν

Calculation of creep rates:

. The total drag brought back to a driving wheel and to the load on a wheel is equal to:

8Qx _ Rolling drag + aerodynamic drag I number of rear wheels μP μx rear weight I number of rear wheels

= λ - 6λ + 12λ (relation 11)

The relation 10 becomes

from which we deduce

_ζ ^ E.ël _λ ___ _{w λ}

Kq R R

a being the length of ^the contact on a plane with

a = jδc (D - δc)

D: diameter of the tire δc: crushing when empty = 13.61 mm

The tangential creep rate is then for 20 km / h (<r - _o ⁼ 0 _> 003078 ¹⁷

According to a similar calculation made for 50, 80 and 100 km / h we find

(ζ _xl ) ₅ y 0.005582, (ζ _a ) _κ ^≈ 0.010177, {ζ _xl ) _m = 0.01433

The calculation of the rate of “creep” route is carried out according to the relation:

^rf 3R 3x546,891

The total creep rate, obtained according to the relation relationxT = ζxR + ζxt allows to obtain the following errors ε%:

(ε%) ₂₀ = 1.1374% (ε%) ₅₀ = 1.3877%

(ε%) ₈₀ = 1.8473% (ε%) ₁₀₀ = 2.2626%

It then becomes possible to draw an empty calibration curve.

The correspondence between the road results of the no-load calibration and the results obtained by a test bench bearing can then be checked. It then suffices to use the relationship between the “creep” rates:

D _rnule = - ^ - [l ₊ (ç _{xR +} ς _x! )]

(l + ζxr)

D _mute ~ distance on road

^ _nm i _ea ⁼ simulated distance on rollers ς _xl = rate of "creep" on rollers 4/065970

18

relationship which becomes, because the creep rates are considered to be infinitely small:

D _rmae ≈ d _rouleoux [l + (ζxR + ζxt + ζcr)] relation 12

When this relationship is used, synergy is ensured between the reactive torsors. Thus, a calibration simulation is carried out since instead of moving the vehicle at several speeds on a fixed base, the vehicle remains fixed and it is the calibrated base (here a roller) which moves under the wheels. In this case, we go back by calculating the displacement of the roller when moving the road.

As regards the determination of the load curve, the parameters determined by the manufacturer (data sheets delivered or values recorded on the test bench) are entered in a manner similar to that described above, namely:. AN weighing: 6,920 kg

. Front wheel crushing: = 37.729 mm

2x 9l7.55N / mm. Wheel radius with crushing (rolling radius): 509.162 mm π- - - _Λ , -. _{Λ Γ} 69200 Nx 2.65 mm „ _^ _ - _r

. Front undercarriage: = 360,160 N

509.162 mm

. Bearing parameters entered by hand. Rear weighing: 12,460 kg

. Rear crushing: = 38.023 mm

4x 819.93. Kp obtained from the load index: 3,350 kg 33,500 N _Q , _QQ ~ _{T /}

. Kq = = 819.93N / mm

40.821 mm

. Wheel-crush radius (rolling radius): 546.891 mm - 38.023 mm 04/065970

= 508.868 mm

_" . -,, _T , 126 000Nx2.65 w _{r A aa}

. Rear bearing drag: '= 648.870 N

508,868mm

. Rolling drag: 360,160 + 648,870 Ν = 1,009.31 Ν

From these parameters, the total drag (rolling drag + aerodynamic drag) calculated for various speeds is calculated. This total drag is then related to a wheel and to the load on the wheel to obtain a new Qx (rolling drag + aerodynamic drag for each wheel) and the Qx / μP ratio for different speeds:

Qx _ Rolling drag + aerodynamic drag I number of rear wheels μP μ x rear weight I number of rear wheels

We thus obtain

& - at (20 km / h) = ((1009.031 + 97) / 4) / (0.7 x 124 600/4) = 0.012680935 μP

- at (50 km / h) = ((1009.031 + 610) / 4) / (0.7 x 124 600/4) = 0.018562614 μP

- at (80 km / h) = ((1009.031 + 1566) / 4) / (0.7 x 124 600/4) = 0.029523402 μP

- ^ to (100 km / h) = ((1009.031 + 2446) / 4) / (0.7 x 124 600/4) = 0.039612831 μP

We then determine successively:

the value a of the half contact, that is

a = ^ rear crush x (rear wheel diameter - rear crush) = 738.023 x ((546.891 x 2) -38.023) = 200.3582035 , _^ _. „_, _N 004/065970

20

. the value of ζxt (tangential):

ζxt = cxaxλ (at x km / h) / wheel radius,

with c = vertical stiffness x μ / transverse stiffness = 819,236x07 / 546,157 = 1.05

is :

20 km / h - »1.05x200.358 / 546.891x0.0086 50 km / h -M, 05x 200.358 / 546.891x0.01342 80 km / h -» 1.05x200.358 / 546.891x0.02229 100 km / h - »1.05 x 200.358 / 546.891 x 0.0304

. the value of ζxR (route)

ζxR = rear crushing / (3 x wheel radius) = 30.023 / (3x546.891) = 0.023175351

. the value of total Ksi ζxT ζxT = (ζxR + ζxt) xl00

is

20 km / h - ζxT = (0.02317 + 0.00323) x 100 - = 2.641432233 = (ε%) 20 50 kmh - ζxT = (0.02317 + 0.00472) x 100 = 2.790686508 = (ε%) 50 80 km / h - ζxT = (0.02317 + 0.00749) x 100 = 3.06726853 = (ε%) 80 100 km / h - ζxT = (0.02317 + 0.01002) x 100 = 3.320000641 = (ε%) 100 004/065970

In a similar way to the previous one, it becomes possible to draw a calibration curve under load.

Thanks to the arrangements described above, it appears that it is possible to carry out a calibration:

1. in static, by calculation, from the physical parameters of the wheel-plane contact,

2. by rolling on rollers ensuring synergy by calculation between reactive rollers. The “creep” rates ζx _r , ζx _R , ζχ _t having been memorized during the installation of the tachograph (or other speed or distance measuring instrument), the periodic control can be carried out in application of relation 12 on a fictitious road distance with the recording of a control disc at one or more speeds in order to verify that the measurement error remains within the limit of the maximum tolerated errors.

It should be noted that so far it has been assumed that the number of pulses per km (gross W), delivered by the sensor provided at the output of the gearbox, corresponded to that K of the measuring instrument. In fact, this is not necessarily true so it is necessary to make an adaptation.

Claims

Claims 22

1. Method for calibrating a speedometer fitted to a vehicle, as a function of speed and load, characterized in that it comprises determining a table and / or a calibration curve providing the value of an error rate ε% as a function of the speed, this error rate ε% being obtained for each speed by the sum of the “creep” rate (free rolling) on the road ξx _R and the tangential creep rate (propulsive or traction) ζ _Xt on the road.

2. Method according to claim 1, characterized in that it comprises a preliminary phase, a test phase on a test bench, and a calculation phase:

- the preliminary phase comprising the entry by the computer system of the test bench, of the characteristic data of the vehicle as provided by the vehicle and tire manufacturer, these data concerning in particular the radius of the wheel, the sidewall height, the '' crushing, front / rear load indices ...,

the test phase comprising the placing of the vehicle on the test bench and the weighing at the level of the front axle and the rear axle, this test phase comprising a determination step, in particular:

. crushing the front wheels,

. wheel radius with crushing (rolling radius)

. front rolling drag,. rear vertical stiffness,

. rear transverse stiffness,

. crushing the rear wheels,

. the radius of the rear wheels with crushing,

. rear rolling drag,. of the total running drag which is the sum of the front rolling drag and the rear rolling drag as well that possibly the rolling bearing of one or more trailers,

- the calculation phase comprising the determination of a series of successive speeds and, for each of these speeds, the calculation of the rate of “creep” (free rolling) on road ζxR, the rate of tangential “creep” (propulsive or traction) ) on the road ζxt and the calculation of the error ε in particular as a percentage.

3. Method according to one of claims 1 and 2, characterized in that it comprises determining a rolling distance on the rollers of the test bench and calculating the length which must be displayed on the indicator speed, as well as the error ε for each speed and for each load.

4. Method according to one of the preceding claims, characterized in that:

- for the purpose of calculating creep rates, the total drag which is equal to the sum of the rolling drag and the aerodynamic drag, calculated for various speeds, is related to one wheel and to the load on a drive wheel,

- the ratio is determined:

8Qx _ Rolling drag + aerodynamic drag I number of rear wheels μP μ x weight rear / number of rear wheels

= λ ³ - 6λ ² + 12λ

we deduce the tangential creep rate ζxt

5. Method according to one of the preceding claims, characterized in that the rate of “creep” route is determined by means of the relation 004/065970

24

_ δc _ crushing AR

C_XR - - -

3R 3 x wheel radius

6. Method according to one of the preceding claims, characterized in that it comprises a step of verifying the correspondence between the results on the road of the no-load calibration and a rolling test bench by means of the relationship :

^ _roUle ≈ d _rmkaux [l + (ζxR + ζxt + ζxr)] relation 12

_Road D being the distance on the road, d _rou ι _eawc the length of roller bearing.

7. Device for calibrating a speedometer fitted to a vehicle in accordance with the method according to one of the preceding claims, this device involving a simulator comprising a roller test bench (Gj, G ₂ ) (or endless belts of which at least one roller (or a band) is designed to be rotated by a wheel (RR) of the vehicle, and a calculation and management unit (PC) of the simulator which receives the information relating to the rotation of the roller (G_, G ₂ ) or of the strip and information relating to the rotation of the wheel from the speedometer, characterized in that in order to carry out said calibration the calculation unit ( PC) comprises means making it possible to determine, from said information as well as information relating to the vehicle, a calibration table or curve providing the value of an error rate ε% as a function of the speed, this rate error ε% being obtained p or each of the speeds by the sum of the “creep” rate (free rolling) on the road ζxR and the tangential “creep” rate (propulsive or traction) ζxt on the road. 004/065970

25

8. Device according to claim 7, characterized in that the calculating and management unit (PC) of the simulator comprises means making it possible to enter data characteristic of the vehicle as supplied by the vehicle and tire manufacturer, this data concerning in particular the radius of the wheel, the height of the sidewall, the crushing, the front / rear load indices, as well as means for determining parameters comprising in particular:

. crushing the front wheels,. the radius of the wheels with crushing (rolling radius)

. the front rolling drag,

. rear vertical stiffness,

. rear transverse stiffness,

. crushing the rear wheels,. the radius of the rear wheels with crushing,

. rear rolling drag,

. the total drag which is the sum of the front drag and the rear drag and, optionally, the drag of one or more trailers.

9. Device according to one of claims 7 and 8, characterized in that the calculation unit comprises means for determining a rolling distance on the rollers (G_, G ₂ ) of the test bench and the calculation of the length which must be displayed on the speedometer, as well as the error ε for each speed and for each load.