WET CLUTCH FOR MOTORIZED EQUIPMENT
The wet clutch design addresses torque oscillations in hybrid vehicles by using a high-stiffness actuating piston and low-stiffness support means to minimize noise and mechanical stress, enhancing comfort and efficiency.
Patent Information
- Authority / Receiving Office
- FR · FR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2022-11-25
- Publication Date
- 2026-03-20
AI Technical Summary
Hybrid vehicles experience significant torsional vibration and noise issues due to torque oscillations in conventional wet clutches, which compromise acoustic comfort and require a technical compromise in parameter settings.
A wet clutch design with a high-stiffness actuating piston and low-stiffness support means that dynamically adjust contact points with the multi-disc assembly to mitigate torque oscillations, featuring a simplified architecture.
The clutch effectively reduces torque oscillations, improving acoustic comfort and reducing mechanical stress, while maintaining efficient torque transmission.
Smart Images
Figure 00000015_0000 
Figure 00000015_0001 
Figure 00000015_0002
Abstract
Description
Title of the invention: Wet clutch for motorized equipment
[0001] The present invention relates to a wet clutch, particularly for motorized equipment, for example, a motor vehicle, a motorcycle, a droid, or any other vehicle comprising at least one rotating electrical machine. When the motorized equipment, in a particular example a motor vehicle, also comprises a heat engine coupled to the electric machine, the vehicle is said to be "hybrid" because the vehicle's propulsion can be achieved either purely electrically, purely thermally, or in a hybrid manner by using both types of energy simultaneously.
[0002] Compared to conventional motor vehicles, hybrid vehicles require an even lower level of vibration because the noise of the internal combustion engine no longer masks other noises when the vehicle is stationary. The issue of torsional vibration under low torque within the wet clutch is one of the most concerning.
[0003] When using a wet clutch in a transmission, it is necessary to reconcile a large number of parameters, such as the torque to be transmitted, the speed of clutch opening and closing, and the comfort of actuation under low torque. Generally, respecting these various parameters entails accepting a technical compromise, particularly regarding the axial stiffness of the components constituting the wet clutch, which can have detrimental consequences on acoustic comfort.
[0004] The invention aims to remedy these technical problems by proposing a wet clutch capable of mitigating these torque oscillation phenomena while presenting a simplified architecture compared to known wet clutches in the prior art.
[0005] The invention also aims to provide a simple, effective and economical solution to these problems.
[0006] To this end, the present invention proposes a wet clutch for a motorized vehicle, comprising:
[0007] - a torque input disc holder arranged to be rotationally linked to a shaft leading,
[0008] - a torque output disc carrier arranged to be rotationally linked to a driven shaft,
[0009] - a multi-disc assembly radially interposed between the input disk holder of torque and the torque output disc carrier comprising friction discs and plates alternately linked in rotation to the torque input disc holder and the torque output disc holder,
[0010] - an actuating piston that is axially movable along the axis of rotation X of the wet clutch which is controlled by movement to alternately clamp axially in an engaged position the multi-disc assembly against reaction means and release axially in a disengaged position the multi-disc assembly.
[0011] The wet clutch according to the invention is remarkable in that the actuating piston comprises a main body having a high axial stiffness Kl greater than 1000 N / mm and support means on the multi-disc assembly having a low axial stiffness K2 between 500 and 5000 N / mm, the support means being able to come into contact with the multi-disc assembly at the beginning of the movement of the actuating piston during the clutching phase and the main body being able to come into contact thereafter with the multi-disc assembly beyond a threshold value of actuating piston movement between 0.2 mm and 0.5 mm.
[0012] This wet clutch, according to the invention, has the advantage of being relatively insensitive, or even insensitive, to the torque oscillation phenomenon that generates noise within the transmission of the motorized vehicle. At the beginning of the engagement phase, the stiffness of the actuating piston, which bears against the multi-disc assembly, is very low. This allows for an artificial reduction in the stiffness of the multi-disc assembly and, consequently, a positive effect on the torque oscillation phenomenon, as shown in the study of the influencing parameters available later in the description. Beyond a certain displacement threshold value, the stiffness of the actuating piston becomes high to reduce the closing stroke of the wet clutch and thus reduce the time required for its closure.
[0013] Preferably, beyond the displacement threshold value, the support means and the main body can be simultaneously in contact with the multi-disc assembly, the equivalent stiffness of the actuating piston being equal to the axial stiffness Kl of the main body. In this way, the support means no longer undergo any deformation other than that resulting from the initial part of the actuating piston's movement. The mechanical stresses resulting from the deformation no longer increase. The fatigue resistance of the support means is thus improved.
[0014] Advantageously, below the displacement threshold value, only the support means are in contact with the multi-disc assembly, the equivalent stiffness of the actuating piston being equal to:
[0015] Kequi = Al+A'2
[0016] Preferably, the actuating piston can apply a load in Newtons on the multi-disc assembly which is constantly increasing during the phase clutch with a discontinuity of stiffness measurable at the level of the displacement threshold value.
[0017] According to a variant of the invention, the contact diameter of the support means on the multi-disc assembly may be greater than the contact diameter of the main body.
[0018] According to another variant, the contact diameter of the support means on the multi-disc assembly may be less than the contact diameter of the main body.
[0019] Preferably, the support means can be a series of elastic legs distributed angularly around the main body.
[0020] According to a variant of the invention, the support means can be made of the same material as the main body.
[0021] According to another variant, the support means can be attached to the main body, for example by fitting or crimping or riveting so as to form a sub-assembly.
[0022] According to another variant, the support means may be a secondary support ring coming into contact with the multi-disc assembly at the beginning of the movement of the actuating piston during the clutch phase.
[0023] Advantageously, the friction discs and the plates can have an angular reference frame and be oriented angularly with respect to each other.
[0024] Preferably, the actuating piston can be moved along the X axis in the direction of the multi-disc assembly within a control chamber made in a hub, the hub being rotationally fixed to the torque input disc carrier.
[0025] According to another variant, the wet clutch can be integrated into a dual clutch or a triple clutch.
[0026] The invention will be better understood upon reading the following description, given solely by way of example and made with reference to the accompanying drawings in which:
[0027] - Figure [Fig. 1] is an axial cross-sectional view of a wet clutch for a machine conventional motorized system including a multi-disc assembly;
[0028] - Figure [Fig.2] is an isometric view of one of the plates of the mul assembly tidisques of the [Fig.l];
[0029] - Figure [Fig.3] is another schematic view of the numerical model of the assembly multidiscs of the [Fig.l];
[0030] - Figure [Fig.4] is a detailed view of the digital model of the mul assembly tidiscs of the [Fig.3];
[0031] - Figure [Fig.5] is a view of the compression test of the multi-disk assembly of the [Fig.l];
[0032] - Figure [Fig.6] is a graph comparing the compression tests with the digital model of the multidisk set of [Fig.1];
[0033] - Figure [Fig.7] is a graph showing the parameters influencing the torque oscillation phenomenon to which the multi-disc assembly of [Fig.1] is subjected;
[0034] - Figure [Fig.8] is an axial cross-sectional view of a wet clutch according to a first method of implementing the invention;
[0035] - Figure [Fig.9] is an axial cross-sectional view of a wet clutch according to a second method of implementing the invention.
[0036] In the following description and claims, the terms "front" or "rear" will be used, without limitation and to facilitate understanding, according to the direction with respect to an axial orientation determined by the main X axis of rotation of the transmission of the motor vehicle and the terms "inside / internal" or "outside / external" with respect to the X axis and along a radial orientation, orthogonal to said axial orientation.
[0037] The present invention relates to fluid-cooled wet clutches, for example oil, which are subject to torque oscillations. In order to understand the invention, we will first identify the sources of these oscillations by identifying the main parameters that govern them, and then we will describe the different solutions for controlling the phenomenon.
[0038] The study of the torque oscillation phenomenon focuses first on a numerical model that incorporates a wet clutch comprising a realistic multi-disc assembly, for example: a layered assembly of friction discs 12 and steel plates 11. The geometries of the friction discs and plates are derived from measurements, and each of these discs is discretized using finite elements, making the entire system flexible. Finally, the management of nodal contacts allows for the precise consideration of the compression behavior of the multi-disc assembly. Thanks to computer simulation software, it is possible to fully parameterize the numerical model and obtain a rapid calculation, facilitating the identification of the parameters governing the torque oscillation phenomenon.
[0039] The study of the torque oscillation phenomenon then focuses on correlating the model with experiments on a complete wet clutch system. The good correlation, in terms of torque oscillation, indicates that the multi-disc assembly model captures the essence of the torque oscillation generation mechanism. The parameters governing the phenomenon are then linked to the multi-disc assembly itself (geometry, rigidity, etc.).
[0040] We will first describe the wet clutch. [Fig.1] illustrates the classic architecture of a wet clutch 1. This performs exactly the same function as a dry clutch, namely to transmit the torque from the drive motor to the transmission, for example a gearbox or a reducer, and to help to shift smoothly from one gear to another in the transmission.
[0041] The wet clutch 1 is kinematically linked in rotation along an axis X with a first driving shaft, for example a rotor of a rotating electrical machine and also kinematically linked in rotation with a second driving shaft of a transmission, for example a torque output shaft.
[0042] The wet clutch 1 includes in particular:
[0043] - a torque input disc holder 2 arranged to be rotationally linked to the first rotating electrical machine via a gear system,
[0044] - a torque output disc holder 3 rotationally linked to the torque output shaft, and
[0045] - a multi-disc assembly 10 radially interposed between the input disk holder of couple 2 and the output disc carrier of couple 3.
[0046] The wet clutch assembly 10 comprises plates 11 rotationally connected to the torque input disc carrier and friction discs 12 rotationally connected to the torque output disc carrier. The friction discs 12 are axially interposed between two successive plates 11. The friction discs 12 and the plates 11 are thus alternately rotationally connected to the torque input disc carrier and the torque output disc carrier.
[0047] To actuate the wet clutch 1, an actuating piston 5 is moved along the X-axis towards the multi-disc assembly within a control chamber 9 formed in a hub 8. The control chamber 9 is sealed by means of a sealing gasket 6 integral with the actuating piston 5. By injecting a pressurized fluid into the control chamber 9 via a supply channel 7, the actuating piston 5 is moved. In the absence of pressurized fluid, the actuating piston 5 is returned to a rest position, corresponding to the opening of the wet clutch under the effect of a helical return spring 4.
[0048] The multi-disc assembly is assumed to be the source of the torque oscillation phenomenon. Geometric defects in the steel plates and friction discs have been identified as sources of perturbations on the transmitted torque. These can cause "static" perturbations, i.e., perturbations in which there is a direct transfer function between the torque oscillation and the circumferential defects and where the amplitude of the torque oscillation is independent of speed, or "dynamic" perturbations, for example, self-excited vibrations where the amplitude of the torque oscillation depends on speed. Tribological aspects can also play a major role in the torque perturbation.Friction paper is an organic material whose operational sliding profile has been shown to depend on both load and operating conditions, resulting in dynamic instabilities. To narrow the scope of investigation, a systematic approach is applied to identify and prioritize the root causes of torque oscillations.
[0049] To highlight the torque oscillation phenomenon, a numerical model of the multi-disc assembly is created. The following sections briefly describe the main steps for developing a complete model, which are then discussed in more detail. First, the geometries of each component, including the actuating piston, are precisely measured.
[0050] Next, the geometries are discretized using a finite element formalism. Then, contact management of the friction disks and plates is implemented to manage the internal interactions of the multi-disk assembly under load. Finally, a load case is applied and the torque oscillation phenomenon is monitored during the simulation.
[0051] The main objective is to obtain a correlation between the numerical model and the experiments. This implies that the numerical model must be as close as possible to the test sample in terms of geometry. For this reason, both sides of each platform 11 are precisely measured in order to derive the thickness and variations of the median plane 13, as illustrated in [Fig. 2]. The measurement is performed by laser sensors located on the mean radius of the platform 11. The sensors are fixed relative to the ground S, and the platform 11 is rotated to acquire the altitude over a complete revolution. Note that a filtering process is necessary to eliminate noise from the signals and obtain usable information as input to the numerical model.
[0052] A finite element formalism is used to construct the numerical model of the multidisk assembly. Each friction disk 12 of the multidisk assembly is represented by its neutral fiber, which is a 3D curve whose shape is given by the median height extracted from the measurements. The discretization of the neutral fiber is performed using 3D Timoshenko beams sharing the same cross-section as the physical parts. Figure 3 shows the complete numerical model of the multidisk assembly.
[0053] The finite element model of the multi-disc assembly consists of a piston 5, steel plates 11 and friction discs 12.
[0054] At this stage, the layers of the model are completely independent of each other, i.e., the stiffness matrices of the layers are decoupled. Coupling is achieved using a node-to-node penalty contact N, as shown in the left-hand portion of [Fig. 4], ensuring the transmission of internal forces. The penalty contact method consists of activating a node-to-node spring R if a node-to-node penetration is detected, as shown in the right-hand portion of [Fig. 4].
[0055] In [Fig. 4], the contact is detected before nodal penetration. The 3D Timoshenko F-beams are mapped along the neutral fiber, but the detection of the The contact must take into account the thickness information mentioned previously.
[0056] When penetration is detected, a node-to-node spring R is activated. However, a spring R must have a stiffness characteristic, which must be provided by the user. This is a drawback of the penalty method, as the contact stiffness is generally difficult to identify. To overcome this drawback, the contact stiffness is identified from the compression behavior of the multi-disk assembly measured during experiments. The nodal contact stiffness (assumed to be homogeneous for all nodal pairs in the model) is then adjusted (for example, by an optimization procedure) so that the compression behavior of the numerical model corresponds to the force-displacement law of the real system, as shown in [Fig. 5].
[0057] The originality here lies in the fact that the behavior of the organic friction material is modeled by the contact itself. The organic paper of the friction discs 12 is not taken into account in the model, and only the properties of steel are applied to the beam elements. The idea is to take advantage of the contact penalty, which exhibits low penetration at equilibrium. This gives physical meaning to the penetration by considering it as representing a crushing of the paper in the thickness direction. In practice, a smooth bilinear law is used to represent the nodal contact force as a function of the contact penetration.
[0058] The left part of [Fig.5] shows a compression test of the multi-disc assembly, and the right part of [Fig.5] illustrates a comparison of the measurement from the compression test, corresponding to curve Cl, and the numerical model, corresponding to curve C2, with the calibrated contact stiffness law.
[0059] It is interesting to note that this guiding equation is nonlinear. This arises from the fact that Fcont is unknown since the contact conditions are unknown a priori. Furthermore, Fcont depends on U, which is also an unknown of the system. The first equation is then solved classically by a first-order Taylor series expansion in the neighborhood of an initial assumption of U. Consequently, the overall nonlinear system is solved as a succession of linear systems. In practice, the Newton-Raphson method is used.
[0060] Finally, the complete model is governed by the following first equation:
[0061] K,MU + Fml(U)=F„,
[0062] Where Kstruct is the global stiffness matrix of the model, U is the nodal displacement vector, Fcont is the internal contact force vector and Fext is the external force vector.
[0063] It is interesting to note that this guiding equation is non-linear. This arises from the fact that Fcont is unknown since the contact conditions are a priori unknown. Furthermore, Fcont depends on U, which is also an unknown in the system. The first equation is then solved classically by a first-order Taylor series expansion in the neighborhood of an initial assumption of U. Consequently, the overall nonlinear system is solved as a succession of linear systems. In practice, the Newton-Raphson method is used.
[0064] Considering a real wet clutch system, the load on the multi-disc assembly is supplied by fluid pressure on the side of the actuating piston 5, as illustrated in [Fig. 1]. In this study of the torque oscillation phenomenon, the fluid is assumed to be incompressible, and the actuating piston 5 and the steel plate 11 are assumed to be much more rigid than the friction discs 12. Based on these assumptions, for a given operating point and a constant average torque, the stack of the multi-disc assembly is assumed to be constrained between two rigid (but geometrically realistic) walls during the sliding phase. The effects of friction and oil are not taken into account in this numerical model.
[0065] The complete simulation consists of a series of nonlinear static calculations (see first equation) for each increment of rotation of the friction discs 12. The output of the procedure is the overall axial force of the multi-disc assembly as a function of the friction disc angle over one complete revolution. This allows the evolution of the transmitted torque to be constructed a posteriori using a standard Coulomb law whose coefficient is identified experimentally.
[0066] At the end of the procedure, the performance of the oscillating torque CO is evaluated by calculating the amplitude at each order (harmonic) of the rotational speed. In practice, the oscillating torque CO takes the form of a Fourier series such as the following: [°067] C(?(e) = COQ + £^6^080 + COsi sine
[0068] Where COo is the average torque, COci and COsi being the cosine and sine parts, respectively, of order i and theta being the angle of rotation of the friction disks.
[0069] Finally, COci is the amplitude of the i-th order oscillating torque given by: [0°7°] ca = + C(^
[0071] A comparison between the prediction of the numerical model and the experimental results is presented in [Fig. 6]. Then, some parameters governing the torque oscillations are highlighted with the numerical model.
[0072] Figure 6 shows the correlation between the numerical model and the measurements from the compression test. In particular, the left column presents the results of the tests over thirty repetitions, and the right column the results of the numerical model. Orders 1, 2, and 3 are found in rows 1, 2, and 3, respectively. It can be seen that The amplitude of the torque oscillation represents a very small fraction of the average transmitted torque, and the proposed model is capable of predicting oscillations in the first and second order ranges. The predicted third order is slightly higher than the maximum value observed experimentally, but the order of magnitude is correct. The model is then validated for "static" torque oscillations, i.e., those related to geometric defects.
[0073] The numerical torque oscillation model allows the identification of two main parameters, shown in [Fig. 7]. It is clear from the left-hand side of [Fig. 7] that the higher the axial stiffness of the multi-disc assembly, the higher the torque oscillation. This is due to the fact that the internal geometric deviations of the multi-disc assembly interact during the rotation of the friction discs 12 relative to the steel plates 11 and impact the internal axial forces through the stiffness of the clutch pack.This results in a more or less high oscillating torque, depending on the rigidity. The second important parameter is the orientation of the friction discs 12, as shown in [Fig. 7]. For this exercise, only the orientation of the friction discs 12 was changed, while the orientation of the steel plates 11 remained unchanged. All other things being equal, this parameter has a significant impact on the oscillating torque. It is worth noting that this parameter does not require any design updates and that the orientation of the multi-disc assembly can be managed on the production line during the assembly phase.
[0074] In conclusion, the phenomenon of torque oscillations occurring in a lubricated automotive clutch is studied, and some of the parameters governing it are identified. A numerical model using finite elements of the multi-disc assembly is developed and tested experimentally. The originality of the approach lies in the fact that the behavior of the friction paper is taken into account at the contact management level by using a representative contact law. This allows the dimensions of the numerical model to be restricted while ensuring representative compression behavior. Dynamic contributions to the torque oscillation are neglected, and the oscillating torque is calculated through a succession of nonlinear static load cases.
[0075] Correlation with measurements from the compression test shows good agreement in the amplitudes of the torque oscillation over the first three orders. This shows that the mechanism of the torque oscillation is mainly due to interactions between geometric defects that influence the internal forces of the multi-disc assembly and therefore the transmitted torque. The validated numerical model is able to identify the two main governing parameters. The first is the axial stiffness of the wet clutch multi-disc assembly, which must be reduced as much as possible to minimize the torque oscillation phenomenon. The second is the orientation of the Elements of the multi-disc assembly, particularly the orientation of the friction discs, are crucial. It has been demonstrated that proper orientation results in minimal torque oscillation. This parameter is perhaps the most important from a production standpoint, as it requires no design modifications and can be directly controlled during assembly.
[0076] The main object of the invention is therefore to provide an improved wet clutch taking into account the parameters influencing the appearance of the torque oscillation phenomenon so that it is insensitive to this phenomenon.
[0077] We will now describe in more detail the arrangement of a wet clutch 1, according to one of the preferred embodiments of the invention, with reference to [Fig.8].
[0078] The wet clutch 1 includes, in particular:
[0079] - a torque input disc holder 2 arranged to be rotationally linked to a shaft leading,
[0080] - a torque output disc holder 3 arranged to be rotationally linked to a shaft led,
[0081] - a multi-disk assembly 10 radially interposed between the input disk holder of torque 2 and the torque output disc carrier 3 comprising friction discs 12 and plates 11 alternately connected in rotation to the torque input disc carrier and the torque output disc carrier,
[0082] - an actuating piston 5 that is axially movable about the axis of rotation X of the wet clutch 1 which is controlled by movement to alternately clamp axially in an engaged position the multi-disc assembly 10 against reaction means 15 and release axially in a disengaged position the multi-disc assembly.
[0083] In this example, the reaction means 15 consist of an end plate 12a having a thickness greater than those present in the multidisc assembly and an open elastic ring 16 inserted into a groove of the torque input disc holder 2.
[0084] The actuating piston 5 moves along the X axis towards the multi-disc assembly 10 within a control chamber 9 made in a hub 8, the hub 8 being rotationally fixed with the torque input disc holder 2.
[0085] According to the invention, the actuating piston 5 comprises a main body 51 having a high axial stiffness K1 greater than 1000 N / mm and support means 52 on the multi-disc assembly 10 having a low axial stiffness K2 between 500 and 5000 N / mm. In this example, the support means 52 come into contact with the multi-disc assembly at the beginning of the actuating piston's movement during the clutch phase, and the main body 51 subsequently comes into contact with the multi-disc assembly beyond a threshold value of displacement V of the actuation piston between 0.2 mm and 0.5 mm.
[0086] The support means 52 are a series of elastic legs distributed angularly around the main body 51. The elastic legs are in particular made of the same material as the main body 51. For example, the actuating piston comprises 6 elastic legs distributed angularly at 60° around the X axis.
[0087] Figure 8 illustrates the wet clutch 1 at the beginning of the engagement phase; the actuating piston 5 is below the displacement threshold value V. At this stage of clutch closure, only the support means 52 are in contact with the multi-disc assembly 10; the equivalent stiffness Kequi of the actuating piston is then equal to:
[0088] Kequi = yy
[0089] During the engagement phase and beyond the displacement threshold value V, the support means 52 and the main body 51 are simultaneously in contact with the multi-disc assembly. The equivalent stiffness Kequi of the actuating piston 5 is then equal to the axial stiffness Kl of the main body. In this second part of the actuating piston 5's movement, the elastic tabs no longer undergo any deformation other than that resulting from the first part of the actuating piston's movement. The mechanical stresses resulting from the deformation no longer increase. The fatigue resistance of the elastic tabs is thus improved.
[0090] As illustrated, the contact diameter dm of the support means 52 on the multi-disc assembly is greater than the contact diameter of the main body 51.
[0091] Figure 5 shows the load curve C3 representing the effort applied by the actuating piston 5 on the multi-disc assembly 10. The load curve C3, expressed in Newtons, exhibits a discontinuity at the displacement threshold value V corresponding to the contact of the main body 51 with the multi-disc assembly and, consequently, to the change in equivalent stiffness Kequi of the actuating piston. The displacement threshold value V of the actuating piston is between 0.2 mm and 0.5 mm.
[0092] We will now describe, with reference to [Fig. 9], a wet clutch 1 according to a second embodiment of the invention substantially similar to the previous one. This second embodiment of the invention is distinguished by the fact that the support means 52 are attached to the main body.
[0093] In this example, the support means 52 are a series of elastic tabs attached to the inner bore of the main body 51. For example, the elastic tabs can be fitted or crimped or riveted so as to form a subassembly with the actuating piston.
[0094] In order to reduce the phenomenon of torque oscillation, the geometric defects of the steel plates 11 and the friction discs 12 are identified and marked. An angular reference mark, for example a punch, is applied to each of the plates and friction discs. Then, the plates 11 and friction discs 12 are oriented angularly relative to each other to compensate for the geometric defects.
[0095] The invention is not limited to the examples of implementation of the invention just described. For example, the support means may be a secondary support ring coming into contact with the multi-disc assembly at the beginning of the movement of the actuating piston during the clutch phase.
[0096] According to one embodiment of the invention, the contact diameter dm of the secondary support ring on the multi-disc assembly may be greater than the contact diameter of the main body. Preliminary support on the larger diameter dm prevents the friction discs from conizing.
[0097] The invention is not limited to the examples just described. The wet clutch 1 can be integrated into an electric vehicle transmission but can also be integrated into a so-called "hybrid" motorized vehicle transmission comprising a rotating electric machine that can be coupled to the power transmission from an internal combustion engine.
Claims
Demands
1. A wet clutch (1) for a motorized machine, comprising: - a torque input disc carrier (2) arranged to be rotationally linked to a driving shaft, - a torque output disc carrier (3) arranged to be rotationally linked to a driven shaft, - a multi-disc assembly (10) radially interposed between the torque input disc carrier (2) and the torque output disc carrier (3) comprising friction discs (12) and plates (11) alternately rotationally linked to the torque input disc carrier and the torque output disc carrier, - an actuating piston (5) axially movable about the axis of rotation (X) of the wet clutch which is actuated in displacement to alternately axially clamp the multi-disc assembly (10) against reaction means (15) in an engaged position and axially release the multi-disc assembly in a disengaged position,characterized in that the actuating piston (5) comprises a main body (51) having a high axial stiffness (K1) greater than 1000 N / mm and support means (52) on the multi-disc assembly having a low axial stiffness (K2) between 500 and 5000 N / mm, the support means (52) being able to come into contact with the multi-disc assembly at the beginning of the movement of the actuating piston during the clutch phase and the main body (51) being able to come into contact thereafter with the multi-disc assembly beyond a threshold value of displacement (V) of the actuating piston of between 0.2 mm and 0.5 mm.
2. Wet clutch (1) according to claim 1, characterized in that beyond the displacement threshold value (V), the support means (52) and the main body (51) are simultaneously in contact with the multi-disc assembly (10), the equivalent stiffness (Kequi) of the actuating piston (5) being equal to the axial stiffness (Kl) of the main body (51).
3. Wet clutch (1) according to claim 2 or 3, characterized in that below the displacement threshold value (V), only the support means (52) are in contact with the multi-disc assembly (10), the equivalent stiffness (Kequi) of the actuating piston (5) being equal to: Kequi = [7] / [7] Kl
4. Wet clutch (1) according to any one of the preceding claims, characterized in that the contact diameter (dm) of the support means (52) on the multi-disc assembly is greater than the contact diameter (de) of the main body (51).
5. Wet clutch (1) according to any one of claims 1 to 3, characterized in that the contact diameter (dm) of the support means (52) on the multi-disc assembly is less than the contact diameter (de) of the main body (51).
6. Wet clutch (1) according to any one of the preceding claims, characterized in that the support means (52) are a series of elastic tabs distributed angularly around the main body (51).
7. Wet clutch (1) according to claim 6, characterized in that the support means (52) are formed from material with the main body (51).
8. Wet clutch (1) according to claim 6, characterized in that the support means (52) are attached to the main body (51), for example by fitting or crimping or riveting so as to form a sub-assembly.
9. Wet clutch (1) according to any one of claims 3 to 8, characterized in that the actuating piston (5) applies a load in Newtons on the multi-disc assembly (10) which is constantly increasing during the clutch phase with a discontinuity of stiffness measurable at the level of the displacement threshold value (V).
10. Wet clutch (1) according to any one of the preceding claims, characterized in that the friction discs (12) and the plates (11) have an angular reference frame and are oriented angularly with respect to each other.