Robert B Randall | The University of New South Wales (original) (raw)

Papers by Robert B Randall

Lecture Notes in Computer Science, 1995

This paper provides an inspection maintenance model for a production system which is subjected to... more This paper provides an inspection maintenance model for a production system which is subjected to deterioration and random failures. A Markov model for preventive maintenance of such a continuously operating device whose condition deteriorates with time in service has been proposed by Sim and Endrenyi (1993). This model incorporates deterioration and so-called Poisson failures, minimal repair, periodic minimal maintenances, and

In this paper, the nonlinear process of amplitude modulation is modelled as the multiplication of... more In this paper, the nonlinear process of amplitude modulation is modelled as the multiplication of a real positive enveloping function (the modulating function) with another function (the modulated function) which is assumed to be a linear combination of all the resonances excited by mechanical faults (if any) in a particular system under consideration. These resonances appear as multiple carriers in the modulation process. Rectifying the multiplied signal is shown to decompose it into two additive functions. The first of these is the scaled modulating function alone and the second is a complicated function involving the modulating function, sum and difference frequencies of the original multiple carriers (and additive frequency components such as background noise) and higher order terms of the sum and difference frequencies. Because the modulating function is scaled up and at low frequencies while the second, complicated function is at higher frequencies, the characteristic frequenc...

Journal of Sound and Vibration, 2002

... Received 30 May 2001; accepted 17 December 2001. Available online 28 October 2002. Abstract. ... more ... Received 30 May 2001; accepted 17 December 2001. Available online 28 October 2002. Abstract. This part addresses the feasibility of reconstructing the pressure trace in IC engine from non-intrusive structural vibration measurements by means of an optimal inverse filter. ...

Mechanical Systems and Signal Processing, 2001

Journal bearings used in IC engines are prone to premature failures and are likely to fail earlie... more Journal bearings used in IC engines are prone to premature failures and are likely to fail earlier than the rated life due to highly impulsive and unstable operating conditions and frequent starts/stops. Vibration signature extraction and wear debris analysis techniques are prevalent in industry for condition monitoring of rotary machinery. However, both techniques involve a great deal of technical expertise, time, and cost. Limited literature is available on the application of these techniques for fault detection in reciprocating machinery, due to the complex nature of impact forces that confounds the extraction of fault signals for vibration-based analysis and wear prediction. In present study, a simulation model was developed to investigate the bearing wear behaviour, resulting because of different operating conditions, to complement the vibration analysis. In current simulation, the dynamics of the engine was established first, based on which the hydrodynamic journal bearing for...

This paper presents the findings of an investigation into the structural properties of a double d... more This paper presents the findings of an investigation into the structural properties of a double deck passenger railcar. The objective of the investigation was to develop an in-service dynamic model about an operational point where the car was travelling at 80km/h over track used in everyday service. A detailed grid of response measurements, providing a spatial definition of the car, were recorded consisting of nearly three hundred accelerometer records in three axes distributed over the structure in more than forty measurement batches while the car was running down the track. Using these measurements, an in-service modal model was constructed by operational modal analysis using the enhanced frequency domain decomposition technique. Of particular interest to this investigation were the primary vertical and lateral bending modes which were found to be coincident, with the lateral mode dominating. The application of whole body vibration weightings however, produced an equivalence of th...

INVESTIGACION & DESARROLLO, 2004

The main objective of this paper is to perform a comparison of several curve-fitting methods for ... more The main objective of this paper is to perform a comparison of several curve-fitting methods for extraction of the modal parameters from response vibration measurements, and in particular the best damping estimates. Measurements were carried out on a steel beam to which a constrained layer had been added to make the damping more similar to that of vehicle structural components. Two shakers with different excitation signals, a periodic impulsive and a random signal, respectively, excited the structure, but after separation, only the random part was analysed for the results of this paper. This study compares a number of common curve fitting methods, viz: The Rational Fraction Polynomial Method, the Complex Exponential Method, the Complex Cepstrum Method, the Hilbert Envelope Method and the Ibrahim Time Domain method. The most accurate results for detection of the damping and natural frequencies were obtained by using the Ibrahim Time Domain Method, with the Rational Fraction Polynomial method very similar. The Hilbert Envelope method gave comparable damping estimates. The Cepstrum and Complex Exponential methods gave reasonable results for the frequencies, but not for the damping. RESUMEN El objetivo principal de este trabajo es realizar una comparación de varios métodos de ajuste de curvas para la extracción de los parámetros modales a partir de mediciones de vibración de respuesta y, en particular, definir las mejores estimaciones de amortiguación. Las mediciones se llevaron a cabo en una viga de acero y se adhirió una capa de material de amortiguación, esto para hacer que la amortiguación experimental sea similar a la de los componentes estructurales en los vehículos. Dos agitadores "skakers" con diferentes señales de excitación se aplicaron, una impulsiva periódica y una señal aleatoria, respectivamente. En este trabajo de investigación, después dela separación de las señales; sólo la señal aleatoria se utilizó para futuros análisis. Este estudio se realizó la comparación de métodos más comunes de ajuste de curvas: El método de la fracción racional de polinomio, el método de exponencial compleja, el Método Cepstrum Complejo, el método de Hilbert Envoluta y el método Ibrahim dominio de tiempo. Los resultados más precisos obtenidos para la detección de las frecuencias naturales y de amortiguación fue con el método de Ibrahim en el dominio de tiempo, y con resultados similares con el método de fracción racional de polinomio. El método de Hilbert Envoluta dio estimaciones de amortiguación comparables. El método de Cepstrum y el Método Exponencial Compleja dieron resultados comparativos para las frecuencias pero no para los valores de amortiguación.

Applied Mechanics - Progress and Applications - Proceedings of the Third Australasian Congress on Applied Mechanics, 2002

Proceedings of the Institution of Mechanical Engineers, Part F: Journal of Rail and Rapid Transit, 2008

ABSTRACT

Mechanical Systems and Signal Processing, 2010

This work aims at monitoring large diesel engines by analyzing the crankshaft angular speed varia... more This work aims at monitoring large diesel engines by analyzing the crankshaft angular speed variations. It focuses on a powerful 20-cylinder diesel engine with crankshaft natural frequencies within the operating speed range. First, the angular speed variations are modeled at the crankshaft free end. This includes modeling both the crankshaft dynamical behavior and the excitation torques. As the engine is

Mechanical Systems and Signal Processing, 2004

Vibration analysis of machines gains greatly in efficiency if periodic vibrations can be separate... more Vibration analysis of machines gains greatly in efficiency if periodic vibrations can be separated out from non-deterministic ones. Part II of this paper introduces a solution in the frequency domain, which is faster and simpler to use than adapative algorithms (part I). The performance of this new algorithm is thoroughly investigated and compared to those of the self-adaptive noise cancellation algorithm. Finally, convincing examples are given of the application to the separation of actual bearing signals from gear signals in gearboxes.

Australian Journal of Mechanical Engineering, 2011

Abstract Lumped parameter models (LPMs) are widely utilised to predict the dynamic behaviour of m... more Abstract Lumped parameter models (LPMs) are widely utilised to predict the dynamic behaviour of mechanical systems such as gearboxes. LPM gives reasonable representation of the dynamics of the system if masses can be lumped at certain locations, such as gears, shafts, bearings, etc. LPM have the advantage of simulating the structure using a limited number of degrees-of-freedom (DOF), which facilitates studying the behaviour of gears and bearings in the presence of nonlinearities and geometrical faults. However, it is difficult to account for the casing flexibility in the LPM models which is an important consideration in the lightweight structures such as in aircraft applications and this results in poor spectral matching over a wide frequency range. In the case of continuous systems, where masses are distributed equally over the structure (gearbox casing), other methods, such as finite element analysis, are often used to study the behaviour of the structure. The use of finite element model (FEM) results in a large number of DOF, which in turn complicates simulating the whole system’s response to the presence of nonlinearities and to gears and bearing faults. This in turn limits the validity of the simulated results and restricts their later usage in the diagnostics and prognostics of the gears and bearings. This paper describes the use of dynamic reduction technique to reduce the FEM of a gearbox casing into manageable and well representative DOF of the casing. The reduced model of the casing is embedded with the LPM of the internals (shafts, gears and bearings), which was previously developed with the aid of Simulink and has the capability of capturing stiffness nonlinearities arising from gears and bearings and has also the capability of simulating geometrical faults (spalls) for both gears and bearings. This paper mainly deals with the simulation of localised and extended inner race bearing faults. The results show the improvements obtained through combining the reduced model of the casing with the LPM model and a much better correspondence with measured signals.

International Journal of Condition Monitoring, 2013

ABSTRACT The dynamic behaviour of a gearbox system was previously analysed using a lumped paramet... more ABSTRACT The dynamic behaviour of a gearbox system was previously analysed using a lumped parameter model (LPM), a reduced nite element model (FEM) and a combined LPM and FEM model of the gearbox casing, and the results were presented at last year's conference. Further investigations into the simulation model indicated the presence of aliasing components in the response spectra. The simulation model has since been updated by using a revised FE model of the casing and addressing the errors detected in the previous model, including aliasing. The FE model reduction was based on the Craig-Bampton method of component mode synthesis (CMS), which has the ability to include the modal properties of the gearbox casing and internals up to a much higher frequency. The reduced mass and stiffness matrices extracted om the FE model were imported into the dynamic model of the gearbox developed in Matlab/Simulink®, which has the capability of simulating time-varying stiffness non-linearities and geometric faults in the gears and bearings. The results of the updated simulation models are compared with a new set of measurements in the presence of extended inner and outer race faults in the bearings. This paper highlights the application of fault simulation techniques in machine condition monitoring (MCM) and in the training of neural networks, by being able to generate a large amount of vibration data without the need to conduct a large number of costly and time-consuming experiments.