WO2011101306A1

WO2011101306A1 - Method for testing the life-time of gear drives and the like

Info

Publication number: WO2011101306A1
Application number: PCT/EP2011/052107
Authority: WO
Inventors: Giovanni Castellani
Original assignee: Dinamic Oil S.P.A.
Priority date: 2010-02-17
Filing date: 2011-02-14
Publication date: 2011-08-25
Also published as: ITBO20100084A1; IT1398482B1

Abstract

The method for testing the life-time of gear drives and the like provides for prearranging the calculated life-time for a gear drive (1) to be tested, prearranging a scale model of said drive (1) to be tested according to a determined scale ratio, predetermining a desired testing time shorter than said calculated life-time. The method further provides for calculating equivalent operating conditions to be applied to said scale model during said desired testing time, suitable to determine, upon calculation, the same risk of damage determined by the design operating conditions of the said drive (1) to be tested. Finally, the equivalent operating conditions are applied to the scale model, and the damage condition thereof is monitored.

Description

METHOD FOR TESTING THE LIFE-TIME OF GEAR DRIVES AND THE LIKE

Technical Field

[01 ] The present invention regards a method for testing the life-time of gear drives and the like.

Background Art

[02] Digital simulation methods suitable to calculate the behaviour of machines or parts thereof in response to the stresses expected according to the design specifications and based on the type of application have been known for a wh ile. For example, such simulation methods provide for building a virtual geometrical model of the mechanical component to be verified and for performing the virtual subdivision thereof in a multiplicity of finite elements. The proper mechanical properties of the material provided by the design for the component are attributed to the so-discretized virtual model, and finally the stresses to be applied in the simulation calculation are defined on the basis of a physical model of the specific application.

[03] Generally, simulation methods enable to calculate the tensions and deformations of the virtual model at discrete points, known as "nodes", of the aforesaid finite elements. In such a way it is possible to obtain a generally rather precise prediction of the behaviour of the mechanical component to be verified.

[04] Nevertheless such methods, even if able to build high precision virtual models, representative not only of the geometrical features, but also of the technologies of the mechanical components, have limits in the prediction of the behaviour of some particularly complex mechanical components. This is specifically the case of the gears, for their particular complexity both constructive and of application . In particular, for the applications which involve gears, such simulation methods are not able to calculate in a sufficiently reliable way the working life until the arising of a determined damage event or a surface failure due to fatigue, the so-called "pitting", or a failure at the tooth base, and then the breaking.

[05] Rapid prototyping methods are also known, which provide for realizing small-scale mod el s of the mechanical components to be verified. Nevertheless such scale prototypes are especially designed to the verification of the manufacturing and geometrical aspects of the design of mechan ical com ponents, more than of the dynam ic and structu ral behaviours. Therefore prototyping is generally not suitable for simulating the behaviour as to the stresses, in particular to the fatigue stresses to which the gear teeth are usually subjected, to predict their working life or life-time.

[06] Therefore in order to verify that the life-time required by the design for a determined gear drive is actually guaranteed and in a sufficiently secure way, tests are usually carried out on at least one specimen of the gear drive to be verified. This turns out to be particularly costly, especially in the case wherein the gear drive is complex and/or has high dimensions and the required life-time determines very long test times.

[07] As an example, EP 1 930 855 discloses a method for estimating the life of a gear box using a simulation model.

[08] Moreover, DE 197 13 583 discloses a general method for evaluating the remaining working life of a machine. [09] Furthermore, DE 10 2007 017614 discloses a method of estimating the life of a gear device by determining stresses in several sections of the gear device.

Disclosure

[10] The task of the present invention is that of solving the aforementioned problems, devising a method for testing the life-time of gear drives and the like, the method being able to perform in an efficient and reliable way the testing of the life-time of gears until the arising of a damage or failure.

[1 1 ] Within such task, it is a further scope of the present invention that of providing a testing method which allows to carry out the testing of the lifetime of gears in shorter times than the design life-time.

[12] Another object of the present invention is that of providing a use of a scale model of a gear drive for actuating the method for testing the life-time of gear drives according to the invention.

[13] The above mentioned scopes are attained, according to the present invention, by the method for testing the life-time of gear drives and the like according to claim 1 , as well as by the use of a scale model according to claim 9.

Description of Drawings

[14] Details of the invention shall be more apparent from the detailed description of a preferred embodiment of the method for testing the life-time of gear drives and the like, illustrated for indicative purposes in the attached drawings, wherein:

[15] figure 1 shows a front view of a gear drive suitable to be tested by means of the method according to the invention; [16] figure 2 shows a longitudinal section view of the gear drive in figure 1 ;

[17] figure 3 shows a diagram of the damage curve defined by the standard ISO 6336.

Best Mode

[18] With particular reference to such figures, a gear drive suitable to be tested by means of the testing method according to the invention is indicated in its entirety with 1 .

[19] The gear drive 1 illustrated in figures 1 and 2 is of the epicyclic type, but different types can be also provided . In particular, the gear drive 1 illustrated for indicative purpose comprises four planet wheels 2 carried rotating by a planetary member 3. The gear drive 1 further comprises a solar pinion 4, mounted centrally and geared with the planet wheels 2. The satellites 2, the planetary member 3 and the central pinion 4 are mounted inside an internal-toothing crown wheel or ring 5, which is in its turn geared with the planet wheels 2. The central pinion 4 is mounted at the end of a shaft 6, which according to the application can be driving or driven. The planetary member 3 is mounted freely rotating and coaxial to the shaft 6, such that the rotational motion can be transferred from the central pinion 4 to the planetary member 3 or to the ring 5 or vice versa as a function of the type of application and of the desired reduction ratio. More precisely, in the case wherein the ring 5 is mobile, driven or driving, the planetary member 3 is fixed, so the functioning does not result epicyclic.

[20] The testing method according to the invention provides for verifying the lifetime of the gears of the real gear drive 1 through the building of a scale model of the same gear drive 1 or of the pair of wheels which results most stressed from the calculation. To this aim, the scale model is suitable to be subjected to a direct test with operating cond itions equivalent to the conditions provided by the design for the real gear drive 1 , as better explained in the following. Such verification suitable to be performed not on the real gear drive specimen, but on the corresponding scale model, is therefore an indirect verification aimed at testing a pair of wheels consisting of the central pinion 4 and any of the geared planet wheels 2 for example.

[21 ] I n su bstance, the scale model is subjected to operating conditions equivalent to the conditions provided by the design, that is suitable to cause, on the basis of the theoretical calculation, the same level of damage. In case the entity of the load is only handled as operating condition, the method provides for prearranging at first for the model a first load value and for calculating the life-time thereof until a damage, then for identifying an incremental coefficient to be applied to such first load value to obtain an equivalent load value, any other operating condition remaining the same, for example the angular speed , such as to obtain through calculation a calculated life-time until the damage which is substantially equal to a desired testing time and shorter than the initially calculated lifetime for the real gear drive to be tested.

[22] The method preferably provides for calcu lating in an intermed iate computational step, by means of an intermediate incremental coefficient, an intermediate load value suitable to determine the same life-time calculated for the gear drive to be tested, until the arising of the damage event. In such a case, the method provides for successively identifying a corresponding incremental coefficient suitable to determine the cited shorter calculated life-time for the scale model, substantially equal to the desired testing time. By carrying out a direct test on the scale model at the so-calculated equivalent operating conditions, it is therefore possible to verify the expected calculated life-time for the real gear drive to be tested in considerably shorter times.

[23] The method also applies in the case wherein the design provides variable loads, in particular different load levels that can be represented according to a determined load histogram. As for a single load and according to the same modal ity, it is possible to calculate for the model corresponding equivalent load levels, for corresponding time intervals, in percentage with respect to the total functioning time of the required life-time, in such a way that the corresponding load histogram has the same trend and, for the cited intermediate calculation step, the total life-time calculated for the model is the same as the gear drive of design. Successively, equivalent load levels are calculated, incremented preferably in proportional way, such as to determine a calculated life-time which is reduced to a desired shorter time. The model is then subjected to the incremented load levels, for the reduced calculated life-time that represents the testing time for the scale model.

[24] In particular, the cited equivalence is based on a calculation proceeding which takes into account all the usual circumstances which condition the life-time of the gears, on the basis of the determinations for example of the standard ISO 6336/2006, which allows to identify life-time curves for the gears, and correction factors to take the particular surface conditions of the gears to be verified into account.

[25] With particular reference to the verification of the life-time to the pitting of the pair central pinion 4 and planet wheel 2, one has to keep in mind that such a type of teeth damage is a surface fatig ue phenomenon, the probability of which or risk of failure is represented in a very precise way by a curve in a logarithmic diagram having the application cycles N_L in abscissa and a load factor Q_w in ordinates, Q_w being representative of the same load in terms of force or torque (see figure 3). In particular it is to observe that the first tract of the curve, from the point indicated with the W letter until the so-called conventional fatigue limit indicated with the V letter, schematically represented by a straight line, is defined in an unequivocal manner by the cited standard ISO 6336. In fact such standard defines for such tract the cited surface factors that enable to take into account all the circumstances of the different applications which can affect the arising of the damage. Using the life-time curve usually said "life curve" it is thus possible to obtain, given the design load, a calculated life-time for the gear until the arising of the pitting that is the "surface durability" or "resistance to pitting". Analogous considerations can apply for the calculation of the resistance to breaking.

[26] The cited equivalence concept of the method according to the invention also refers to the concept of damage D_n for a determined load level n, the only one provided in the design or the nth of a series of multiple load levels, which can be expressed such as the ratio between the required life-time N_L, for example expressed in number of functioning hours, and the calculated life-time until the arising of the damaging failure N_Lf or failure number of cycles expressed in hours as well, according to the formula [27] The life-time to the arising of the damaging failure N_Lf for the calculation of the damage D_n can be calculated, known the unique provided load level or each load level of the multiple provided ones, from the life-time curve of the type illustrated in figure 3, for the WV tract relative to the only finite life-time (see figure 3). In the case wherein various load levels are provided, it is possible to calculate a cumulative damage, assuming that each load level causes a partial damage which can be calculated according to the previous formula, the cumulative damage being given by the sum of such partial damages, according to the known Miner hypothesis.

[28] Once the life-time calculation is performed for the real toothing, for the pair central pinion 4 and planet wheel 2, the result is a calculated life-time until the arising of the fatigue failure, for example the so-called "pitting" or failure at the tooth root, that is until breaking. Such a calculated life-time is longer than the minimal life-time required for the given real gear drive 1 , by a factor representing the security level of the design.

[29] Successively, first operating cond itions for the scale model wh ich is intended to be realized are applied to the same calculation method. Such conditions are defined by a single load level or by a series of multiple load levels, according to the case, and determine a corresponding calculated life-time. In an intermediate step of calculation it is possible to identify intermediate incremental coefficients of the cited first operating conditions suitable to determine for the scale model a calculated life-time substantially equal to the one calculated for the real gear drive 1 .

[30] The first determined operating conditions, or the intermediate ones if provided, a re then incremented, through successive iterations of the calculation, until obtaining equivalent operating conditions for a reduced calculated life-time, that is such as to lead to a calculated life-time equal to a desired testing time, for example 500 hours maximum. In other words, fixed the desired number of hours for the test, it is possible to obtain through successive iterations of calculation the entity of the equivalent operating conditions, for example in terms of load or of the multiple load levels, to be applied to the scale model so that the predetermined life-time is equivalent, as for the damage effects on the model, to the life-time of the real gear drive. In other words these incremented operating conditions, as an example in terms of incremented loads, or in alternative of incremented rotational speeds or of both conditions, allow to perform on the scale model an indirect but reliable verification of the life-time of the real gear drive 1 in testing times reduced, that is shorter, at will.

[31 ] According to known formulas, for example and preferably according to the formulas proposed by the standard ISO 6336, it is possible to characterize specific values or surface "corrective" factors also for the scale model, in particular for the roughness, the speed and lubrication respectively, so that the life-time curve exemplified in figure 3, is adapted to the specific case of application and, therefore, the functioning test on the scale model for the desired time interval and to the equ ivalent operating conditions, can actually be equivalent, in terms of probability of arising of the damage, to a direct functioning test carried out on the real gear drive 1 for the calculated life-time of the real gear drive 1 .

[32] More precisely, on the basis of the type of application, of the loads and of the geometrical and technological features, it is possible to set the value of the corrective factors which, according to known formulas, enable to calculate the involved Hertz pressures. Such corrective factors regard different circumstances that influence the pressures acting on each tooth, such as, for example, the rigidity.

[33] It is also possible and known to introduce a power distribution factor or "sharing factor" to take into account the greater incidence of the applied load or loads, as an example in the case wherein a same wheel is geared with more than one pinion or vice versa. Once the Hertz pressures are obtained, it is possible to deduce the conventional fatigue limit, through which it is possible to derivate a calculated life-time value of the pair wheel- pinion under examination and so, known the required life-time, a value of the damage D_n for the evaluation of the equivalence.

[34] It is to note that the cited formulas known for the calculation of the resistance of the gears to the pitting in particular, also provide a corrective factor Zx to take into account the effect of the dimensions of the pairs wheels-pinions. But, actually, it is known that such a factor is considered equal to unity, confirming the fact that the test carried out on the scale model can be considered reliable for the verification of the damaging effects of equivalent operating conditions on the real gear drive 1 .

[35] In particular it is known that the empiric basis for the calculation of the resistance to the pitting of gears is given by a series of experimental results for the so-called "surface factors", obtained by the FZG institute on test gears with a centre distance equal to 91 ,5 mm. It has been demonstrated that such experimental results can be applied on gears of any dimension without distinction.

[36] With regard to the manufacturing of the scale model, it is preferable that the manufacturing techniques of the gears of the scale model are the same as the corresponding wheels of the real gear drive 1 to be tested . To such regard, for example, it is admissible that the mechanical workings are carried out by different specimens of tool machines, but of the same type, for example by means of gear cutting machines or grinding machines.

[37] The geometrical features of the gears of the model to be realized for the test must be in a fixed scale ratio with respect to the features of the corresponding gears of the real gear drive 1 . That means, for example, that the number of teeth, the dimension of the centre distance of the pair of gears to be tested, as well as the face width and the corresponding tip diameters must be in the same scale ratio with respect to the design values of the real gear drive 1 .

[38] The construction material is the same for the scale model and for the real gear drive 1 .

[39] As for the thermal treatments, if provided for one or both wheels of the gear to be tested , they must be of the same type, to obtain the desired equivalence, namely the same probability of failure or damage level in the reduced scale model and in the real gear drive 1 .

[40] Nevertheless, there are features for which it is not possible to maintain the same scale ratio for the real gear drive 1 and for the model because they depend for example on the tolerances of the mechanical workings. As an example, this regards the surface roughness, but also, in the particular case of cemented gears, the so-called "efficient cementation thickness". In this latter case for example it is preferable to choose for the scale model an efficient thickness such as to obtain an equal ideal compression stress at the internal limit of the efficient thickness.

[41 ] On this matter it is noted that the method can be applied to any type of material and of surface treatment provided for the pair of wheels of the real gear drive 1 to be tested. The material of one or both wheels of the pair to be tested can in fact lack of any hardening treatment, as in the case wherein for example the pin ion of the pair results cemented, while the wheel suitable to be coupled thereto is made of simply drained steel.

[42] In the same way as the surface roughness and as the possible hardening thickness, also the rigidity of the carrying structures of the model is not reduced in scale. In fact, it is necessary to provide bearing structures having greater dimensions with respect to the dimensions that can be calculated on the basis of the scale ratio of the model, in order to resist to the high loads which generally result from the cited calculation of the equivalent operating conditions, for a reliable life-time test in relatively short time. In the same way the addendum corrections are not made in scale, but maintain the same character, preferably with a gradual variation of the profile, in order to avoid the abrupt decrease of the contact ratio.

[43] Finally, it is preferable to maintain the same accuracy grade for the gears.

[44] The functioning of the method for testing the life-time of gear drives and the like according to the invention turns out to be easy to understand from the preceding description.

[45] For example, it is to be tested, with regard to pitting, an epicyclic gear drive 1 designed with centre distance equal to 203 mm, net face width of 190 mm, modulus equal to 10, tooth number equal to 19, 20 respectively for the central pinion 4 and for a planet wheel 2 coupled thereto.

[46] From the application of the cited calculation methods, it is possible to calculate the pitting resistance of the different toothings in terms of life-time, namely the number of total hours of functioning until the arising of the damage. To accelerate the computational times it is preferable to apply the cited calculation methods with the aid of computational proceed ings assisted by electronic computer which is able to give an answer in terms of life-time, as the calculation proceedings known with the name of RHF. On the basis of the known design loads, for example variable and precisely differentiated on three increasing load levels each with a determined required functioning life-time , it is possible to calculate the Hertzian pressures and so to identify in the design gear the weakest wheel, having the shortest calculated life-time until the damage.

[47] As can be seen in the following tables 1 and 2, in the illustrated case, for the design torques a calculated total life-time to the pitting of the central pinion 4 equal to 6420 hours has been obtained.

Table 1 real gear

Total life until pitting load factor QH

[hours] pinion wheel pinion wheel

Level a 2,441 2,399

Level b 1 ,709 1 ,679 6420 14900

Level c 0,367 0,36

Table 2

Table 3

Table 4 [48] The calculation takes into account all the factors previously described and, in particular, the geometrical, load and surface factors. For a more accurate verification, the load sharing has also been taken into account, introducing the already cited corresponding corrective factor, representative of the influence, a s fo r t h e l ife-time, of the number of planet wheels 2 simultaneously geared with the central pinion 4. It is also suitable to take into account, in the calculation of the life-time to pitting, the directionality of the motion for the determined application that, for example in the case wherein a bidirectional motion is provided, has the effect of reducing the surface fatigue by sharing it between both flanks of every tooth.

[49] In the illustrated case it is intended to perform a direct test on a reduced scale model of the real gear drive 1 , having modulus equal to 4, functioning centre distance equal to 81 ,2 mm, face width equal to 76 mm . More precisely, the geometrical data of the pair constituting the model are determined according to a reduction scale equal to the ratio between the corresponding modulus of the model and of the designed real gear drive 1 , that is equal to 4/10.

[50] As previously described, the manufacturing technology of the gear and the accuracy grade must be, for the scale model, the same as for the real gear drive 1 . According to the same principles, it has to be determined, in the case a surface hardening treatment is provided for one or both wheels of the real gear drive 1 , as an example a case-hardening treatment, the corresponding effective case-hardened thickness of the scale model, which geometrically depends on the diameter of the wheels and on the face width. It is preferable to provide, for the calculation , that the effective case- hardened thickness depends also on the value of the load or on the applied load levels, confirming the fact that as a geometrical data to be considered for the scale model , the case-hardened thickness is not proportional, according to the adopted scale, to the case-hardened thickness of the real gear drive.

[51 ] Obviously it is possible to apply the method also for testing gears obtained through different surface hardening technologies, such as for example the induction surface tempering or the gas or plasma nitriding. But it is to be considered that gas or plasma nitriding in particular allows to obtain an effective thickness comparable to the one of the case-hardening only for small-dimension gears, for which the realization of reduced scale models does not presumably result advantageous.

[52] The corrective factors are successively calculated as a function of the dimensions of the scale model, preferably according to the standard ISO 6336, in order to characterize the slope of the life-time curve in the WV tract of the finite life resistance for the three first loads, at least one of which must result comprised between the points W and V of the same curve (fig. 3), in order to obtain, as previously indicated, the corresponding partial damage D_n.

[53] To deduce the percentage of application time of the different load levels, the desired life-times, for both the real gear drive and the model , are introduced in the calculation, as input data.

[54] In the illustrated case, once identified the load levels for the model, such as to determine the same total calculated life-time to pitting of 6420 hours (see table 3), by successive iterations the incremented load levels are calculated in order to obtain the desired total reduced, that is shorter, life-time.

[55] Practically, the desired testing life-time is fixed, for example equal to 500 hours, and the sa me l ife-time calculation process is applied, with incremental load coefficients greater at each calculation iterated step, until a test incremental coefficient, in the case equal to 1 ,597, which determines a calculated total life-time equal to the testing life-time desired, namely around 500 hours (see table 4 of figure 4).

[56] Then, the operative step of direct testing on the scale model is carried out, applying thereto the calculated incremented load levels, for the desired testing life-time.

[57] At the end of the test the surface condition of the scale model is observed, to evaluate the level of surface damage thereof. A positive result of the test, namely the absence of pitting or the presence of an initial pitting only, constitutes an indirect verification that the central pinion 4 of the real gear drive 1 will work without pitting for the design number of hours, which according to a determined resulting safety level will be sufficiently distant from the number of hours corresponding to the calculated life-time, in the illustrated case equal to 6420 hours.

[58] In alternative to the described actuation modality, it is possible to provide for shortening the testing times, calculating equivalent functioning speeds instead of equivalent loads. Again, it is possible to provide for varying both, with respect to the project data, so as to obtain damage effects of equal risk, and therefore equivalent.

[59] The method for testing the life-time of gear drives and the like according to the invention allows to perform in an efficient and reliable way the testing of the life-time of gear drives. Such a scope is attained mainly through the use of a scale model of the gear drive to be tested, to which are applied, accord ing to the method, equivalent functioning conditions for a predetermined testing time, preferably reduced at will with respect to the calculated life-time until the damage for the designed gear drive 1 .

[60] A characteristic of the invention is therefore the fact that it allows to perform an indirect test of the resistance and of the functioning of the real gear drive 1 , on the model of which constant or variable loads according to the case are applied, or rotation speeds, proportional or equivalent to the values of the corresponding operating conditions accord ing to the design. Such equivalence, as previously described, refers to the evaluation of the same damage for each load of the real design and for each load to be applied to the model.

[61 ] An important advantage of the method in hand is constituted by the fact that testing times can be reduced considerably and at will. Furthermore, in spite of the use of higher load and/or speed levels with respect to the real case, the testing method in hand enables to reduce the total energy waste. In the illustrated example, with the model in a scale reduced according to the ratio of 4/10, to reduce the life-time to 500 hours, the equivalent loads result incremented by 59,7% with respect to the ones wh ich wou ld g ive a calculated total life-time, for the pinion, equal to 6420 hours. In spite of this, the power on the scale model is reduced until about 0,10 with respect to the one that would be required for the direct testing of the real gear drive at the design real conditions.

[62] It is also important to observe that the use of a model in reduced scale with respect to the real gear drive is particularly advantageous for cumbersome realizations which would require very high costs for manufacturing as well as testing . The model, even if built with the described precautions to respect the proportionality conditions, undoubtedly results cheaper and easy to verify in any testing laboratory.

[63] In practice, the embodiment of the invention, the materials used, as well as the shape and dimensions, may vary depending on the requirements.

[64] Should the technical characteristics mentioned in each claim be followed by reference signs, such reference signs were included strictly with the aim of enhancing the understanding the claims and hence they shall not be deemed restrictive in any manner whatsoever on the scope of each element identified for exemplifying purposes by such reference signs.

Claims

1 . Method for testing the life-time of gear drives and the like, characterized in that it comprises the following operating steps:

a. prearranging a calculated life-time until a determined damage (D_n) for a gear drive (1 ) to be tested on the basis of load data and/or design operating conditions;

b. prearranging according to a determined scale ratio a scale model of said drive (1 ) to be tested;

c. predetermining a desired testing time shortened with respect to said calculated life-time;

d. calculating equivalent operating conditions to be applied to said scale model d u ri ng sa id desi red testing time, such equivalent operating conditions being identified by operating conditions suitable to determine, upon calculation, the said damage (D_n) calculated for said design operating conditions of the said drive (1 ) to be tested;

e. applying to said scale model said calculated equivalent operating conditions for said desired test time;

f. monitoring said scale model in order to evaluate the damage condition thereof, considering the outcomes of said test performed on said scale model as indirect test on said drive (1 ) to be tested.

2. Method accord ing to claim 1 , characterized in that said operating conditions are selected in a group constituted by external loads, by the rotation speed of the gears forming part of the drive (1 ) or by the combination of said loads and said speeds.

Method according to claim 2, characterized in that said operational step of d. calculating equivalent operating conditions comprises the further step of prearranging for the computation first operating conditions for said scale model su itable to obtain for said scale model a corresponding first calculated life-time until said damage (D_n), and of applying to said first operating conditions for said scale model incremental coefficients to obtain for said scale model a reduced life-time substantially equal to said desired testing time.

Method according to claim 3, characterized in that said operational step of applying incremental coefficients to said first operating conditions for the said scale model comprises the intermediate step of obtaining intermediate operating conditions calculated for the said scale model through intermed iate incremental coefficients, sa id intermed iate operating conditions being suitable to obtain for the said scale model a corresponding life-time until said damage (D_n) substantially equal to said calculated lifetime for said drive (1 ) to be tested.

Method according to claim 3, characterized in that said scale model is made through the same type of manufacturing and thermal treatments as said drive (1 ) to be tested.

Method according to claim 5, characterized in that the roughness, lubrication and surface hardness factors for calculating the equivalent operating conditions to be applied to the scale model for performing said test are the ones deducible from the scale model itself.

Method according to claim 6, characterized in that said scale model is made through the same surface hardening treatment as the drive (1 ) to be tested, the corresponding effective thickness being calculated for said model in relation to the diameters at least of the corresponding gears in said model.

8. Method according to claim 1 , characterized in that said damage (D_n) is the arising of the gear surface fatigue phenomenon called "pitting" in at least one gear of said drive (1 ) to be tested.

9. Method accord i ng to cl a im 1 , characterized in that said damage determines the breakage of the tooth of at least one gear of said drive (1 ) to be tested.

10. Use of a scale model of a gear drive for testing the life-time of said drive (1 ) according to the testing method according to one or more of the previous claims.