Search Results

The engine start-up process introduces speed-dependent transient vibration problems in ground vehicle drivelines as the torsional system passes through the critical speeds during the acceleration process. Accordingly, a numerical study is proposed to gain more insights about this transient vibration issue, and the focus is on nonlinear analysis. First, a new nonlinear model of a multi-staged clutch damper is developed and validated by a transient experiment. Second, a simplified nonlinear torsional oscillator model with the multi-staged clutch damper, representing the low frequency dynamics of a typical vehicle driveline, is developed. The flywheel velocity measured during the typical engine start-up process is utilized as an excitation. The envelope function of the speed-dependent response amplification is estimated via the Hilbert transform technique. Finally, the envelope function is effectively utilized to examine the effect of multi-staged clutch damper properties.

The transient vibration phenomenon in a vehicle powertrain system during the start-up (or shut-down) process is studied with focus on the development and experimental validation of the nonlinear powertrain models. First, a new nonlinear four-degree-of-freedom torsional powertrain model for this transient event, under instantaneous flywheel motion input, is developed and then validated with a vehicle start-up experiment. Second, the interactions between the clutch damper and the transmission transients are established via transient metrics. Third, a single-degree-of-freedom nonlinear model, focusing on the multi-staged clutch damper, is developed and its utility is then verified.

Many powertrain structural sub-systems are often tested under steady state conditions on a dynamometer or in a full vehicle. This process (while necessary) is costly and time intensive, especially when evaluating the effect of component properties on transient phenomena, such as driveline clunk. This paper proposes a laboratory experiment that provides the following: 1) a bench experiment that demonstrates transient behavior of a non-linear clutch damper under non-rotating conditions, 2) a process to efficiently evaluate multiple non-linear clutch dampers, and 3) generates benchmark time domain data for validation of non-linear driveline simulation codes. The design of this experiment is based on a previous experimental work on clunk. A commercially available non-linear clutch damper is selected and the experiment is sized accordingly. The stiffness and hysteresis properties of the clutch damper are assumed from the measured quasi-static torque curve provided by the manufacturer.

Combined active and passive damping is a recent trend that can be an effective solution to challenging NVH problems, especially for lightweight vehicle components that demand advanced noise and vibration treatments. Compact patches are of particular interest due to their small size and cost, however, improved modeling techniques are needed at the design stage for such methods. This paper presents a refined modeling procedure for side-by-side active and passive damping patches applied to thin, plate-like, powertrain casing structures. As an example, a plate with fixed boundaries is modeled as this is representative of real-life transmission covers which often require damping treatments. The proposed model is then utilized to examine several cases of active and passive patch location, and vibration reduction is determined in terms of insertion loss for each case.

Elastomeric joints are utilized in many automotive applications, and exhibit frequency and excitation amplitude dependent properties. Current methods commonly identify only the cross-point joint property using displacement excitation at stepped single frequencies. This process is often time consuming and is limited to measuring a single dynamic stiffness term of the joint stiffness matrix. This study focuses on developing tractable laboratory inverse experiments to identify frequency dependent stiffness matrices up to 1000 Hz. Direct measurements are performed on a commercial elastomer test system and an inverse experiment consisting of an elastic beam (with a square cross section) attached to a cylindrical elastomeric joint. Sources of error in the inverse methodology are thoroughly examined and explained through simulation which include ill-conditioning of matrices and the sensitivity to modeling error.

Decoupler nonlinearities of the automotive hydraulic engine mount affect its isolation performance and the transmission of structure-borne noise. The kinematic gap nonlinearity of the decoupler is examined in considerable detail in the context of the quarter car model. It is shown that while modeling it with a “softened” nonlinear expression may only moderately affect predicted system behavior at the excitation frequency, it can significantly after it at higher harmonics, changing the predicted level of structure-borne noise transmission. Studies of multi-harmonic motion and vibratory power transmission under sinusoidal and composite excitation conditions confirm that, in fact, use of a decoupler with a “softened” nonlinearity improves performance.

Many automotive components and sub-systems require viscoelastic damping treatments to control noise and vibration characteristics. To aid the dynamic design process, new approaches are needed for modeling of partial damping treatments and characterization of the overall dynamic behavior. The analytical component of the design process is illustrated via the transmission casing cover, along with supporting experiments. First, the vibration response of production casing plates is examined, with and without the constrained layer treatment. A modified flat plate is employed along with a generic housing that provides the realistic boundary conditions for subsequent work. A simplified analytical damping model for constrained viscoelastic layer damping is suggested based on assumed modal functions. Using the analytical model, design guidelines in terms of optimal patch shapes and locations are suggested.

The knowledge of frequency-dependent dynamic stiffnesses of mounts, in axial and flexural motions, is needed to determine the behavior of many automotive sub-systems. Consequently, characterization and modeling of vibration isolators is increasingly becoming more important in mid and high frequency regimes where very few methods are known to exist. This paper critically examines some of the approximate identification methods that have been proposed in the literature. Then we present a new experimental identification method that yields frequency-dependent multi-dimensional dynamic stiffnesses of an isolator. The scope is however limited to a linear time-invariant system and our analysis is restricted to the frequency domain. The new characterization method uses two inertial elements at both ends of an isolator and free boundary conditions are maintained during testing.

In many manual transmissions, conditions for the onset of vibro-impacts from an unloaded gear pair are more likely than from an engaged set. Although some of the general characteristics of neutral gear rattle are known, no specific analytical models are available in the literature that can explain interactions between unloaded and loaded gear pairs in the drive rattle mode. For the sake of illustration, a particular problem for a light duty truck is studied in this paper and dynamic interactions are investigated. Some experimental measurements are first presented to define the unloaded gear rattle problem. Linear and non-linear mathematical models of the driveline are developed to understand, quantify and control the rattle problem. Trends predicted by simulations are compared with those observed in experiments. The effects of various gear run-ups and vibratory drag torques are investigated.

A vibration isolator or mount is often modeled by the Voigt model describing uni-axial (longitudinal) motion with frequency-invariant parameters. However, wave effects due to the mass distribution within the isolator are observed as the frequency is increased. Further, flexural stiffness components play an important role, leading to off-axis and coupling effects. Thus, the simplified mount models could lead to erroneous predictions of the dynamic behavior of an overall system such as automotive powertrain or chassis mounting systems. This article compares various approximate isolator models using a multi-dimensional mobility model that is based on the continuous system theory. Harmonic force and moment excitations are separately applied to a rigid body source to investigate the multi-dimensional vibratory behavior. Analysis is however limited to a linear time-invariant system and the mobility synthesis method is utilized to predict the frequency domain behavior.

Modal sound radiation of a brake rotor is expressed in terms of analytical solutions of a generic thick annular disk having similar geometric dimensions. Finite element method is used to determine structural modes and response. Vibro-acoustic responses such as surface velocities and radiated sound pressures due to a multi-modal excitation are calculated from synthesized structural modes and modal acoustic radiation of the rotor using the modal expansion technique. In addition, acoustic power and radiation efficiency spectra corresponding to a specific force excitation are obtained from the sound pressure data. Accuracy of the new semi-analytical method has been confirmed by purely numerical analyses based on finite element and boundary element models. Our method should lead to an improved understanding of the sound radiation from a brake rotor and strategies to minimize squeal noise radiation could be formulated.

Accurate quantification of welded or adhesive joints in automotive chassis structures is necessary before reliable models can be developed. Such joints undergo shear and rotational deformations, which must be characterized via diagonal stiffness elements and cross-stiffness terms in order to describe static and dynamic problems. In this paper, a frequency domain decomposition technique is employed to extract static stiffness and viscous damping matrices of dimension 2 via analytical, computational or experimental models. Methods are applied to a laboratory fixture and alternate joints are compared.

Performance characteristics of passive, active, and broadband adaptive engine mounts are compared over a wide frequency range up to 250 Hz in the context of a quarter-vehicle heave model. The optimal damping coefficient of a rubber-metal mount is determined using random vibration theory. The small-scale active mount employs proportional-plus-integral control based on linear optimal control theory. The new adaptive hydraulic mount system implements an on-off damping control mode by using engine intake-manifold vacuum and a microprocessor-based solenoid valve controller. Through analytical methods, it is observed that this adaptive mount provides most desirable dynamic performance with regard to the engine-bounce control, shock absorption and vibration isolation performance requirements. Although technical prospects of the proposed adaptive system appear promising, in-situ performance needs to be evaluated.

Many discrete tonal type noise and vibration problems in automotive systems and other physical structures require passive multi-layer visco-elastic damping treatments in mid to high frequency regimes. To address such issues, experimental modal analysis and dynamic finite element methods are suggested as suitable tools. Results are presented in terms of several test structures (four thin elastic beams, a thick elastic plate and an automotive brake pad) with free-free boundary conditions. Composite modal loss factors are measured and predicted for two different damping insulators consisting of adhesive, steel and coating combinations. Special attention is paid to the elastic deformation modes of test structures and spectral scaling of material properties for the finite element models.

The sources of gear rattle in automotive transmission and other components are clearance non-linearities, which include backlashes, multi-valued springs, hysteresis, etc. Periodic vibro-impacts are generated because of the single sided or double sided impacts. It is obviously desirable to develop an appropriate computer simulation model which can aid in the control of transient noise and vibration signatures. Such models can also be used to design experiments and to interpret measured data. Several approaches have been attempted in the past, but most common is the digital simulation of the governing differential equations describing the non-linear, torsional dynamics of the drivetrain. Some progress has been made in understanding the basic rattle phenomenon and in developing suitable mathematical models. This article intends to be a status report on the mathematical or computer models including pre- and post-processing considerations.

The vibration isolation effectiveness of powertrain mount configurations is examined for electric vehicle application by considering the effect introduced by internal mount resonances. Unlike internal combustion engines where mounts are typically designed only for static support and low frequency dynamics, electric motors have higher excitation frequencies in a range where mount resonances often occur. The problem is first analytically formulated by considering a simple 3-dimensional powertrain system, and the vibration isolation effectiveness significantly deteriorates at the mount resonance(s). It is shown that by modifying the mount shape, the mount resonance(s) can be shifted while maintaining the same static rate, tuning the frequency away from any engine excitation or natural frequencies. Further, internal mount resonances are utilized to improve vibration isolation over a narrow frequency range, using non-identical mounts to split mount resonance peaks.

The elastomeric joints (bushings or mounts) in vehicle structural frames are usually described as uncoupled springs (only with diagonal terms) in large scale system models. The off-diagonal terms of an elastomeric joint have been previously ignored as they are often unknown since their properties cannot be measured in a uniaxial elastomer test system. This paper overcomes this deficiency via a scientific study of a laboratory frame that is designed to maintain a high fidelity with real-world vehicle body subframes in terms of natural modes under free boundaries. The steel beam construction of the laboratory frame, with four elastomeric mounts at the corners, permits the development of a highly accurate, yet simple, beam finite element model. This allows for a correlation study between the experiment and model that helps shed light upon the underlying physical phenomenon.