Вы находитесь на странице: 1из 6

Downloaded from SAE International by Vellore Inst of Technology, Thursday, January 09, 2014 10:28:21 PM

2013-01-1895 Published 05/13/2013 Copyright 2013 SAE International doi:10.4271/2013-01-1895 saepcmech.saejournals.org

Multi-Point Mesh Modeling and Nonlinear Multi-Body Dynamics of Hypoid Geared System
Yawen Wang, Teik C. Lim and Junyi Yang
University of Cincinnati
ABSTRACT A multi-point hypoid gear mesh model based on 3-dimensional loaded tooth contact analysis is incorporated into a coupled multi-body dynamic and vibration hypoid gear model to predict more detailed dynamic behavior of each tooth pair. To validate the accuracy of the proposed model, the time-averaged mesh parameters are applied to linear timeinvariant (LTI) analysis and the dynamic responses, such as dynamic mesh force, dynamic transmission error, are computed, which demonstrates good agreement with that predicted by single-point mesh model. Furthermore, a nonlinear time-varying (NLTV) dynamic analysis is performed considering the effect of backlash nonlinearity and time-varying mesh parameters, such as mesh stiffness, transmission error, mesh point and line-of-action. Simulation results show that the time history of the mesh parameters and dynamic mesh force for each pair of teeth within a full engagement cycle can be simulated. This capability enables the analysis of durability of the gear pair and more accurate prediction of the system response. CITATION: Wang, Y., Lim, T. and Yang, J., "Multi-Point Mesh Modeling and Nonlinear Multi-Body Dynamics of Hypoid Geared System," SAE Int. J. Passeng. Cars - Mech. Syst. 6(2):2013, doi:10.4271/2013-01-1895. ____________________________________

INTRODUCTION
Hypoid and spiral bevel gears are widely used in the rear axle power-train systems of vehicles. The dynamic behavior of hypoid and spiral bevel gear pair is important mainly in two reasons: (1) the vibration and noise problems are related to the dynamic mesh force, which is believed to be excited by gear transmission error (TE) and parametric variation of mesh characteristics, (2) the prediction of dynamic mesh force is critical to the study of gear fatigue and gear life. Moreover, the vibration of the meshing gear pair is transmitted through shaft-bearing assembly into gear housing, which affects the overall system dynamics of driveline system. An accurate mesh model is critical for the prediction of dynamic mesh force. The spring-damper model is usually used to represent the mesh characteristics of gear pairs. Over the past 20 years, there are a number of studies assuming a time and spatial-invariant spring damper model. With the development of tooth surface contact program, there are more studies recently based on the 3-dimensinal gear tooth geometry results [1, 2, 3, 4, 5]. In their models, the timevarying mesh characteristics, including mesh stiffness, transmission error, mesh point and line-of-action (LOA), are obtained from the contact results of quasi-static loaded tooth contact analysis (LTCA), and the contact results are synthesized to a single spring-damper coupling, which
1127

represents the gear mesh. Besides, Wang [3] also investigated the multiple spring-damper coupling and compared the mesh parameters and dynamic responses with those by assuming single spring-damper coupling. The basic frame work of hypoid and spiral bevel gear dynamics is established by Cheng and Lim [1], a nonlinear time-varying (NLTV) model is applied, both backlash nonlinearity and time-varying mesh parameters are considered. Later, Peng [4] derived the coupled multi-body dynamic and vibration model by including the large shaft rotation. Yang [6-7] further investigated the nonlinear dynamic response of hypoid gearbox with time-varying bearing stiffness and elastic housing. Previous hypoid gear mesh model based on single-point coupling may lose accuracy when gear teeth are large and contact ratio is high. The multi-point coupling mesh model is believed to be able to catch the true mesh characteristics with less assumption. In addition, the lack of the detailed analysis of the time history of dynamic mesh force on each engaging tooth pair, limited the further study on gear surface wear and fatigue model. In this paper, a multi-point hypoid gear mesh model based on 3-dimensional loaded tooth contact analysis is incorporated into a coupled multi-body dynamic and vibration hypoid gear model to predict more detailed dynamic behavior of each tooth pair. Linear dynamic responses are

Downloaded from SAE International by Vellore Inst of Technology, Thursday, January 09, 2014 10:28:21 PM 1128 Wang et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)

compared between single-point and multi-point mesh coupling to validate the model. The nonlinear dynamic analysis is also performed considering the effect of backlash nonlinearity and time-varying mesh parameters. More detailed dynamic responses on each engaging gear tooth pair are examined, thus it can be applied to investigate the interaction between gear dynamics and surface wear [8].

(1) The line-of-action vector L(nx, ny, nz) can be obtained from

MESH AND DYNAMIC MODEL Gear Mesh Model


The hypoid gear mesh model, including mesh point, lineof-action, mesh stiffness and kinematic TE can be derived from the load distribution results calculated by a 3dimensional quasi-static loaded tooth contact analysis program for hypoid and spiral bevel gears [9]. This program combines the semi-analytical theory with finite element method [10], which can solve the gear tooth contact problem very efficiently. A hypoid gear pair with multiple contact interfaces is shown in Figure 1(a), and Figure 1(b) shows contact cells on each tooth pair. The position vector of the contact cell in the mesh coordinate system is ri(rix, riy, riz), the contact force is fi, and the normal vector is ni(nix, niy, niz). (2) The total contact moment is given by

(3) The mesh point R(xm, ym, zm) can be obtained from

(4a)

(4b) (4c) The translational loaded and unloaded transmission errors eL and e0 are the projections of corresponding angular transmission error along the line of action. Finally, the mesh stiffness is defined by (5) Above formulation is applied to each contact interface, thus the multi-point mesh parameters in one mesh cycle can be obtain and applied to the dynamic model. The number of contact tooth pairs at any given time is a function of load, pinion roll angle and tooth geometry.

Dynamic Formulation
The coupled multi-body dynamic and vibration model, proposed by Peng [4], is employed in this paper. The pinion, gear, engine and load component are considered as rigid rotors. The engine and load component each has only one rotational coordinate. The shaft-bearing assembly stiffness is used to represent the flexibility at both pinion and gear lumped support. Multi-point mesh coupling is applied between the engaging gear teeth. The equations of motion in matrix form can be written as

Figure 1. Illustrations of: (a) multi-point coupling of hypoid gear pair, and (b) contact cells on engaging tooth surface. The total contact force is calculated by summing the contact forces on each contact cell (assuming there are N contact cells)

Downloaded from SAE International by Vellore Inst of Technology, Thursday, January 09, 2014 10:28:21 PM Wang et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013) 1129

(1) where

(2)

(3) The stiffness matrixis [K] the lumped shaft-bearing assembly support stiffness. The damping matrix [C] is assumed to be viscous type and derived from damping ratio. The gyroscopic matrix [G] and [Ga] are associated with the absolute rolling velocities and accelerations of pinion and gear respectively. The force vector can be written as Figure 2. 14-DOF hypoid geared rotor system model Finally, the dynamic transmission error d is given by

(5) The dynamic mesh force Fm transformation vectors hl are defined as and directional

(11) The mesh parameters km, cm and 0 are obtained from multi-point mesh model, thus the dynamic mesh force for each contact gear pair can be calculated. Numerical integration is applied to solve the nonlinear equations in the time domain.

SIMULATION RESULTS
(6)

Mesh Analysis
Mesh parameters, such as kinematic TE, mesh stiffness, mesh point and line-of-action vector, that vary with pinion roll angle are obtained from quasi-static tooth contact analysis. At any instant of time, there are variable number of mesh interfaces linking the pinion and gear. Each mesh interface is represented by its mesh point, LOA and stiffness. The time-varying mesh point and line of action for each tooth pair are synthesized from the load distribution results. Unlike most other gears, the normal force on each tooth pair acts in different directions for hypoid gears. As is shown in Figure 3, the mesh point and line of action in multi-point mesh coupling presents stronger time-variant characteristics than that in the single-point mesh model. However, the mean values of multi-point mesh points and line of action are approximately the same as those of single-point mesh model. Figure 4 shows the time-varying mesh stiffness for singlepoint and multi-point mesh model. It can be seen that the mesh stiffness of each tooth pair is changing greatly in one mesh cycle, which makes mesh force varies largely. The mesh stiffness of single-point has much smaller variation.

(7) Where bc is the gear backlash, 0 is unloaded transmission error and rotational radius are defined as (8)

(9) (10) where Ll and Rl is the line of action and mesh point in pinion coordinate system (l = p) or gear coordinate system (l = g), and Xl, Yl, Zl are the unit vectors in the pinion or gear coordinate system.

Downloaded from SAE International by Vellore Inst of Technology, Thursday, January 09, 2014 10:28:21 PM 1130 Wang et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)

that the dynamic responses by using multi-point mesh model are in reasonably good agreement with single-point mesh model results. The advantage of multi-point mesh model is that it can predict the mesh force for each tooth pair. Figure 7 shows the dynamic mesh force for each engaging tooth pair. The tooth pair which stays engaged in one mesh cycle contributes the most to the total dynamic mesh force; the other two tooth pairs have nearly same contribution. Then the time-varying mesh parameters and backlash are taken into account, to examine the time history of dynamic responses under nonlinear time-varying condition. Figure 8 compares the angular DTE for single-point and multi-point mesh model, by using torsional angular displacement. The translational DTE then can be calculated and projected to the LOA of each mesh interface, as is shown in Figure 9. There are some variations of projected DTE at some certain frequencies, but the basic trends are the same.

Figure 3. Time-varying mesh point coordinates (a), (b), (c) and LOA directional cosines (d), (e), (f) single-point, mesh point 1, mesh point 2, mesh point 3.

Figure 5. Dynamic transmission error singlepoint mesh model, multi-point mesh model

Figure 4. Time-varying mesh stiffness in one mesh cycle single-point, mesh point 1, mesh point 2, mesh point 3.

Figure 6. Dynamic mesh force single-point mesh model, multi-point mesh model

Dynamic Analysis
The time-averaged multi-point and single-point mesh parameters are applied to linear time-invariant (LTI) analysis. Figures 5 and 6 show the comparison of dynamic transmission error and net mesh force, respectively, from single-point and multi-point mesh model. The results indicate

Downloaded from SAE International by Vellore Inst of Technology, Thursday, January 09, 2014 10:28:21 PM Wang et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013) 1131

Figure 7. Dynamic mesh force of each tooth pair mesh point 1, mesh point 2, mesh point 3

are changing with time, the mean value of mesh force on each tooth pair varies greatly. Thus the net mesh force in the time domain is first obtained by summing the forcing vectors, and then the individual force is obtained by the projection of net mesh force in the frequency domain. It is noticed that singlepoint mesh model tends to overpredict the mesh force at some local ranges, compared to multi-point mesh model. In the overall frequency range, they are still in good agreement with each other. In Figure 12 and 13 the time history of net mesh force and individual mesh forces are shown at 100Hz and 1700Hz, respectively, to compare the time history of dynamic mesh force at a non-resonant frequency and a resonant frequency. It can be observed that (1) the individual mesh forces always have large variations, whereas the total mesh force only has large variation at frequencies that are near resonance; (2) the individual forces vary in the same manner within a full engagement cycle at a certain mesh frequency; (3) the variation ranges of individual forces are almost the same, but in different manners at different frequencies.

Figure 8. Angular transmission error singlepoint mesh model, multi-point mesh model

Figure 10. Dynamic mesh force single-point mesh model, multi-point mesh model

Figure 9. Dynamic transmission error projected to the LOA of each mesh interface single point, mesh point 1, mesh point 2, mesh point 3. Figure 10 and 11 show the net mesh force predicted by single-point and multi-point mesh model, and the individual force for each contact tooth pair. For single-point mesh model, we define the dynamic mesh force as the variation around the mean value in the time domain. However, for the multi-point mesh model, since the load sharing characteristics Figure 11. Dynamic mesh force mesh point 2, mesh point 1, mesh point 3.

Downloaded from SAE International by Vellore Inst of Technology, Thursday, January 09, 2014 10:28:21 PM 1132 Wang et al / SAE Int. J. Passeng. Cars - Mech. Syst. / Volume 6, Issue 2(July 2013)

REFERENCES
Cheng, Y., Lim, T.C., Vibration Analysis of Hypoid Transmissions Applying an Exact Geometry-based Gear Mesh Theory, Journal of Sound and Vibration 240, 519-543 (2001). 2. Cheng, Y., Dynamics of High-Speed Hypoid and Bevel Geared Rotor Systems, PhD Dissertation, Ohio State University, 2000. 3. Wang, H., Gear mesh characteristics and dynamics of hypoid gear rotor system, PhD Dissertation, University of Alabama, 2002. 4. Peng, T., Coupled multi-body dynamic and vibration analysis of hypoid and bevel geared rotor system,, PhD Dissertation University of Cincinnati, 2010. 5. Wang J., Nonlinear Time-varying Gear Mesh and Dynamic Analysis of Hypoid and Bevel Geared Rotor System, PhD Dissertation, University of Cincinnati, 2007 6. Yang, J. and Lim, T., Dynamics of Coupled Nonlinear Hypoid Gear Mesh and Time-varying Bearing Stiffness Systems, SAE Int. J. Passeng. Cars - Mech. Syst. 4(2):1039-1049, 2011, doi: 10.4271/2011-01-1548. 7. Yang, J. and Lim T. C. (2011). Nonlinear Dynamic Simulation of Hypoid Gearbox with Elastic Housing. ASME Conference Proceedings 2011(54853): 437-447. 8. Ding, H. and Kahraman A. (2007). Interactions between Nonlinear Spur Gear Dynamics and Surface Wear. Journal of Sound and Vibration 307(3-5): 662-679. 9. Vijayakar, S. (2003). Hypoid Face Milled User's Manual, (Advanced Numerical Solutions LLC.). 10. Vijayakar, S., A Combined surface integral and finite element solution for a three-dimensional contact problem, International Journal for Numerical Methods in Engineering, 31, 525-545, 1991 1.

Figure 12. Time history of dynamic mesh force at 100Hz net mesh force mesh point 1, mesh point 2 mesh point 3.

CONTACT INFORMATION
Name: Teik C. Lim, PhD, PE, Fellows (ASME, SAE) Position: Herman Schneider Professor of Mechanical Engineering, and Dean of College of Engineering and Applied Science Director, Vibro-Acoustics and Sound Quality Research Laboratory Director, Hypoid Gear Mesh and Dynamic Modeling Consortium Director, UC Simulation Center Address: College of Engineering and Applied Science, Univ. of Cincinnati 801 Engineering Research Center (ERC) P.O. Box 210018, Cincinnati, OH 45221, USA teik.lim@uc.edu

Figure 13. Time history of dynamic mesh force at 1700Hz net mesh force mesh point 1, mesh point 2 mesh point 3.

SUMMARY
The accuracy of the conventional single-point mesh model is challenged by heavy torque load and high contact ratio. To obtain the detailed dynamic response of each engaging tooth pair and to predict the system responses with less assumption, a multi-point mesh model is applied to the coupled multi-body dynamic and vibration formulation of gear rotor system. The mesh parameters and dynamic responses are compared for single-point and multi-point mesh model, and good agreement is obtained. Dynamic mesh force for each engaging tooth pair can be calculated by the projection of net mesh force. The time history of net mesh force and individual forces within a full engagement cycle can be simulated, and results show that multi-point mesh model is able to give more insights on the dynamic behavior of each pair of teeth. This result can be further applied to investigate the gear surface wear model. The effect of drive load and backlash nonlinearity on the individual dynamic mesh force can also be further studied.

ACKNOWLEDGMENTS
The author wishes to express grateful thanks to VibroAcoustics and Sound Quality Research Lab of the College of Engineering and Applied Science at University of Cincinnati for support of this effort.

Вам также может понравиться