Mr M de Smidt
2009 "Internal vibration monitoring of a planetary gearbox
Vibration monitoring is widely used to determine the condition of various mechanical systems. With epicyclic gearboxes, traditional techniques are not advisable. This is in part due to the fact that the planet gears rotate internally on a planet carrier. Special techniques have been developed to extract a viable data signal from the measured vibration signal. This work investigates the possibility of mounting transducers internally on the rotating planet carrier. Mounting transducers at this location removes the relative motion that would be present if traditional measurement techniques were used. An epicyclic gearbox is modified to facilitate the internal mounting of the accelerometers. It is shown that vibration data can be reliably measured internally, by means of accelerometers mounted on the planet carrier. The internally measured data is compared to data obtained by other special epicyclic gearbox measurement techniques and shown to be equally adept in detecting deterioration of a planet gear tooth.
Supervisor: Prof PS Heyns
Co-Supervisor: Dr CJ Stander
Mr C Becker
2009 "Profiling of rough terrain
In the automotive industry one of the methods utilized in accelerating the design, testing and development of a system or a vehicle is the use of virtual vehicle simulations. The simulations cut costs in the form of fewer prototypes required for actual testing and accelerated fault finding in the design of a system. The simulation results are very dependent on the model used for the simulation and the inputs to the system. Feasible results can often be obtained with a simplified model if the correct input data is supplied to the simulation.
In South Africa, the commercial, military and off-road vehicle industries may use the test tracks at the Gerotek Test Facilities for ride comfort and durability tests over repeatable terrains. Terrain profiles of these tracks are not available and cannot be measured using commercially available inertial profilometers due to the severe roughness of the terrain.
This study concentrates on obtaining the input data in the form of the terrain profile used for vehicle simulations and field tests in which a vehicle is driving on rough terrains. The input data is referred to as the profile of the terrain and the profiled terrains are the actual terrains used for testing.
Three different methods are utilized in measuring the profile of the terrain namely a mechanical profilometer, photogrammetry and a 3-D scanner using a laser displacement sensor. These methods are evaluated by profiling the same section of the Belgian paving and calculating the Displacement Spectral Densities. The most efficient method is used to profile additional terrains.
The terrain profiles thus obtained is used as input to an existing off-road vehicle simulation model built in MSC Adams View. This model has previously been verified over discrete obstacles where excellent correlation with experimental results was obtained. Comparison between simulated and measured results over the terrains profiled in this study also gives good correlation, establishing further confidence in the measured terrain profiles.
Supervisor: Prof Schalk Els
Mr P Cronje
2009 "Improving off-road vehicle handling using an active anti-roll bar
This thesis investigates the use of an active anti-roll bar as a means of improving the handling of an off-road vehicle.
The active anti-roll bar consists of a stiff anti-roll bar and a hydraulic actuator at the one end between the anti-roll bar and the rear axle of the vehicle. The system is designed so that the anti-roll bar can be preloaded in both directions by the actuator. The displacement of the hydraulic actuator is close loop controlled to be a function of the lateral acceleration of the vehicle, which is measured by an accelerometer.
For this study, full vehicle simulations were done in ADAMS/View to predict the response of the proposed solutions. A Land Rover Defender 110 was used as the test vehicle to verify the results of the simulations.
Constant radius tests and the severe double-lane-change manoeuvre, which are standard handling tests, were used to determine the vehicle's handling performance. Handling performance was quantified by measuring the body roll angle during the manoeuvre and noting the maximum roll angle. The effect of the active anti-roll bar on ride comfort was measured by driving over Belgian paving at a constant speed.
The results show that the proposed system reduces the body roll angle to zero up to a lateral acceleration of 0.4 g during steady state handling and provided a 74% improvement in maximum body roll angle during a double-lane-change-manoeuvre at 70 km/h. The system has no detrimental effect on the ride comfort of the vehicle.
Supervisor: Prof Schalk Els
Mr T Wilcocks
2009 "Investigation into a system that can detect improper combustion in a diesel engine before significant damage can occur
An alarming number of compression ignition (CI) engines in the transport, mining and heavy engineering environments have been failing due to combustion irregularities within their combustion chambers. It has been found that diesel fuels containing contaminants or diesel fuels with poor lubricity characteristics lead to stickiness of diesel injector needles, which badly affects injector spray patterns resulting in the phenomenon of “cold combustion”. This study has been undertaken to develop a technique for detecting and preventing the damage resulting from this deviation in the combustion of a diesel engine. The technique has been formulated with a view to being as non intrusive as possible, so as not to require major modification of an existing test engine to accommodate the technique. The practice of monitoring individual cylinder exhaust gas temperatures (EGTs) proved to be an effective way of determining whether potentially destructive combustion abnormalities were taking place within the diesel engine. By recording these temperatures at certain stages during the engine’s operation, taking their average, and comparing each one to this average it is also possible to isolate the location of the combustion abnormality. This method proved to be most effective at full loads and maximum fuel delivery where combustion temperatures are highest and the effects of poor combustion are most noticeable and potentially damaging. The second goal was to develop a small, portable electronic device that makes use of the monitoring technique developed and provides a visual and audible alarm to notify a vehicle operator or technician of a combustion fault within a diesel engine. A Combustion Monitoring System (CMS) prototype was developed and tested on a small naturally aspirated engine at the University of Pretoria’s engine testing facilities. The prototype met its primary goal of detecting simulated combustion abnormalities under a variety of test conditions. It is envisaged that the monitoring techniques applied in developing the CMS unit may eventually be incorporated into the powerful processing abilities of the modern diesel Engine Control Unit (ECU). In its current form the CMS prototype is a useful tool in sensing combustion related malfunctions within a diesel engine and preventing damage from occurring.
Supervisors: Prof NDL Burger
and Mr AJ von Wielligh.
Mr Botha M
2009: "A Comparative Study of Reynolds-Average Navier-Strokes and Semi-Empirical Thermal Solutions of a Gas Turbine Nozzle guide Vane
In a typical modern gas turbine engine, the nozzle guide vanes (NGVs) endure the highest operating temperatures. There exists a great drive in the turbine industry to increase the turbine inlet temperatures leading to higher thermal efficiency. This has led to a drive to increase turbine vane- and blade-cooling. Numerical modelling has to a large degree replaced empirical codes and models as the main research tool regarding simulation of blade-cooling. Outdated empirical solvers have made way for commercial solvers such as FLUENT, a Reynolds-averaged Navier-Stokes solver. One such empirical solver, TACT1, has until recently still proved to yield acceptable results. A comparative study has been done using the T56 NGV blade to establish the differences, advantages and disadvantages of these 2 codes. The engine and subsequent NGV blade were analysed using NREC, STAN5, LOSS3D and TACT1. FLUENT simulations were found to be computationally expensive. TACT1 yielded acceptable results compared with computational cost. For modern-day designers, FLUENT would be the preferred tool.
Supervisors: Prof Josua Meyer
and Dr Danie de Kock
Mr Kat C 2009
: "Suspension forces on a tri-axle air suspended semi-trailer
The aim of this study is to investigate the use of multi-body vehicle simulation models to predict the suspension forces acting on the chassis of the vehicle, in order to perform durability analyses.
Traditionally, durability of vehicles is evaluated with proving ground tests. This implies that a physical prototype of the vehicle is required before its durability can be evaluated. If we were able to evaluate the durability of the vehicle without any physical part or a full prototype of the vehicle available, great cost and time savings may be gained. These possible gains have lead to the use of computer aided engineering (CAE) tools. These tools have supplemented the proving ground durability test by using historical measured data and/or predicted data from vehicle simulation models, as input to the durability analyses i.e. Finite Element Analyses (FEA). The usefulness of the historical test data is limited and many of the vehicle simulation models that are used to predict the input data, have not been validated.
In this study a validated mathematical model of a 40 ton flat bed tri-axle semi-trailer, able to predict the suspension forces, is created. The validation of the full vehicle model includes correlations for displacements, velocities, accelerations and forces of various vehicle parameters. A validated mathematical model of the air springs, that includes mass transfer and flow effects for use in full vehicle dynamic simulations, is also developed.
The results obtained indicate that the air spring model, integrated into the full vehicle model, is able to give relative accurate predictions of displacements, velocities, accelerations and forces of various vehicle parameters, over a discrete road event and over a rough road.
Supervisor: Prof Schalk Els