Andrew Guzzomi

Variable inertia effects of an engine including piston friction and a crank or gudgeon pin offset

Abstract In order to obtain greater accuracy in simulation, more sophisticated models are often required. When it comes to the torsional vibration of reciprocating mechanisms the effect of inertia variation is very important. It has been shown that the inclusion of this variation increases model accuracy for both single-cylinder and multi-cylinder engine torsional vibration predictions. Recent work by the present authors has revealed that piston-to-cylinder friction may modify an engines ‘apparent’ inertia function. Kinematic analysis also shows that the piston side force and the dynamic piston-to-cylinder friction are interdependent. This has implications for engine vibration modelling. Most modern engines employ a gudgeon pin offset, and there is a growing interest in pursuing large crank offsets; hence, the effect of these on inertia variation is also of interest. This paper presents the derivation of the inertia function for a single engine mechanism, including both piston-to-cylinder friction and crank or gudgeon pin offset, and investigates the effect of each through predictions. The effect of crank offset on the variable inertia function is also verified by experiment.

The effect of piston friction on engine block dynamics

Abstract When developing vibration models, so as to reduce model complexity, it is typically expected that good prediction accuracy can be achieved by ignoring the complication of friction. In this paper, the significance of friction between the piston and cylinder on engine block dynamics is shown through simulation in both the time and frequency domains. Simulations and experiments indicate that large differences exist between model predictions for the engine block moment if this friction is not accounted for. This is especially true at low crankshaft rotational speeds when dynamic inertia effects are small. Experiments on a motored single cylinder engine at different average rotational speeds confirm the theory and very good tie-up with predictions is obtained. It is expected that these findings will also have implications for the torsional vibration of the engine.

Some torsional stiffness and damping characteristics of a small pneumatic tyre and the implications for powertrain dynamics

Abstract Torsional vibration modelling of powertrains is hindered by the lack of understanding of tyre low-frequency torsional stiffness and damping characteristics. Torsional here means rotation about the wheel (spin) axis so that the rotation of the wheel hub is different from that of the periphery of the tyre. This paper explains a simple, practical test rig design that can be used to determine a tyres torsional characteristics via simple modelling. The model assumes a complex stiffness that allows the damping type to be determined as a function of the frequency. The measurement of some torsional stiffness and damping characteristics of a small pneumatic tyre while under load and rolling is presented. The tyre was excited torsionally and, from the results obtained, the tyres torsional stiffness and damping characteristics were calculated. The investigations were concerned with the low-frequency characteristics. It was found that there was a large torsional vibration associated with the tread pattern. Over the frequency range tested (10–50Hz), it was found that the damping is more accurately represented as hysteretic and not viscous when lightly loaded. The effects of rotational speed, preload, and inflation pressure on the stiffness and damping were also investigated. The damping type for the tyre (hysteretic or viscous) was found to vary depending on the loading and excitation. Tyre properties were then included in a lumped-inertia, frequency-domain powertrain model and the steady-state response predicted. The results indicated that the viscous or hysteretic damping assumption has significant effects on the predicted frequency response of the powertrain. The need for further studies investigating the damping characteristics of automotive tyres in torsion is therefore recommended.

Torsional Receptances and Variable Inertia of a Two-Inertia Model of a Universal Joint

This paper presents the derivation of the torsional receptances for a universal joint. The joint is modeled as two inertias connected by a massless cross-piece. The equations of motion and resulting receptances reveal that the universal joint may be represented as a variable inertia. The inertia is a function of misalignment angle and angular position. It is shown that the inertia variation may be approximated by a second order cosine for typical misalignment angles. Systems with variable inertia are known to exhibit non-linear vibration behaviour.

A Mobility Assignment with Industry Relevance

In an attempt to improve the way students prepare for industry a 3rd year mechanisms and multibody systems class was given an assignment with a non-traditional scope and marking guide. Although the team based assignment, like those of prior years, involved mobility analysis of real historical systems in the form of a formal report, it was proposed that the assignment be marked to industry expectations. This meant 100% if the conclusions were satisfactory and 0% if they were not. This experience produced many interesting outcomes and these are discussed. The paper describes the process that led to the implementation of this new assignment structure. The methodology used in its development and reflections of both the authors and the students are discussed.

Monitoring single-stage double-suction pump efficiency using vibration indicators

Efficiency tests were conducted on two single-stage double-suction split-case pumps in a clean water distribution facility. Efficiency was monitored via two techniques, one established and the other proposed in this study. Measurements of the former were taken via the standard technique using a Yatesmeter that required time to calibrate and that probes be inserted into the clean water flow; the latter method was non-intrusive, being based solely on accelerometer measurements at key locations on the volute/split casing. Both techniques required some post-processing of data. Through comparison of the techniques, it is shown that for the pumps that were analysed it is possible to extract indicators within the vibration signatures that permit good correlation with efficiency data. In both cases only a few recorded measurement points were possible and the vibration indicator-based curve fitting technique was able to locate the best efficiency point more accurately than using the curve fitted to the Yatesmeter measured data. This short communication reports on the initial tests which are showing promising results.