Daniel Nicolas Wilke 2010 "Approaches to accommodate remeshing in shape optimization"

This study proposes novel optimization methodologies for the optimization of problems that reveal non-physical step discontinuities. More specifically, it is proposed to use gradient-only techniques that do not use any zero-th order information at all for step discontinuous problems. A step discontinuous problem of note is the shape optimization problem in the presence of remeshing strategies, since changes in mesh topologies may, and normally do, introduce non-physical step discontinuities. These discontinuities may in turn manifest themselves as non-physical local minima in which optimization algorithms may become trapped.

Conventional optimization approaches for step discontinuous problems include evolutionary strategies, and design of experiment (DoE) techniques. These conventional approaches typically rely on the exclusive use of zero-th order information to overcome the discontinuities, but are characterized by two important shortcomings: Firstly, the computational demands of zero order methods may be very high, since many function values are in general required. Secondly, the use of zero order information only does not necessarily guarantee that the algorithms will not terminate in highly unfit local minima. In contrast, the methodologies proposed herein use only first order information, rather than only zero-th order information. The motivation for this approach is that associated gradient information in the presence of remeshing remains accurately and uniquely computable, notwithstanding the presence of discontinuities. From a computational effort point of view, a gradient-only approach is of course comparable to conventional gradient based techniques. In addition, the step discontinuities do not manifest themselves as local minima.
 

Dr A Oberholster 2010  "The application of Eulerian laser Doppler vibrometry to the on-line condition monitoring of axial-flow turbomachinery blades.".

The on-line condition monitoring of turbomachinery blades is of utmost importance to ensure the long term health and availability of such machines and as such has been an area of study since the late 1960's. As a result a number of on-line blade vibration measurement techniques are available, each with its own associated advantages and shortcomings. In general, on-blade sensor measurement techniques suffer from sensor lifespan, whereas non-contact techniques usually have measurement bandwidth limitations. One non-contact measurement technique that yields improvements in the area of measurement bandwidth is laser Doppler vibrometry.

This thesis presents results and findings from utilizing laser Doppler vibrometry in an Eulerian fashion (i.e. a fixed reference frame) to measure on-line blade vibrations in axial-flow turbomachinery. With this measurement approach, the laser beam is focussed at a fixed point in space and measurements are available for the periods during which each blade sweeps through the beam. The characteristics of the measurement technique are studied analytically with an Euler-Bernoulli cantilever beam and experimental verification is performed. An approach for the numerical simulation of the measurement technique is then presented.

Associated with the presented measurement technique are the short periods during which each blade is exposed to the laser beam. This characteristic yields traditional frequency domain signal processing techniques unsuitable for providing useful blade health indicators. To obtain frequency domain information from such short signals, it is necessary to employ non-standard signal processing techniques such as non-harmonic Fourier analysis.

Results from experimental testing on a single-blade test rotor at a single rotor speed are presented in the form of phase angle trends obtained with non-harmonic Fourier analysis. Considering the maximum of absolute unwrapped phase angle trends around various reference frequencies, good indicators of blade health deterioration were obtained. These indicators were verified numerically.

To extend the application of this condition monitoring approach, measurements were repeated on a five-blade test rotor at four different rotor speeds. Various damage cases were considered as well as different ELDV measurement positions. Using statistical parameters of the abovementioned indicators as well as time domain parameters, it is shown that with this condition monitoring approach, blade damage can successfully be identified and quantified with the aid of artificial neural networks.
Supervisors: Prof P S Heyns
 

Dr J A Olivier 2009  "Single-phase heat transfer and pressure drop inside horizontal circular smooth and enhanced tubes with different inlet configurations for transitional flow".

It is common practice to design water chiller units and heat exchangers in such a way that
they do not operate within the transition region. This is mainly due to the perceived chaotic
behaviour as well as the paucity of information in this region. Due to design constraints or
change of operating conditions, however, exchangers are often forced to operate in this region.

This is even worse for enhanced tubes as much less information within this region is available.
It is also well known that the entrance has an influence on where transition occurs, adding to
the woes of available information.

The purpose of this study is thus to obtain heat transfer and friction factor data in the
transition region of fully developed and developing flows inside smooth and enhanced tubes,
using water as the working fluid, and to develop correlations from these results. The use of
different inlets, tube diameters and enhanced tubes was also investigated with regards to the
commencement of transition.

Heat transfer and pressure drop data were obtained from six different types of tubes with
diameters of 15.88 mm (5/8!!) and 19.02 mm (3/4!!). Low fin enhanced tubes with a fin height
to diameter ratio of 0.4 and helix angles of 18" and 27" were investigated. Heat transfer was
obtained by means of an in-tube heat exchanger with the cooling of water being used as the
test fluid. Reynolds numbers ranged between 1 000 and 20 000 while Prandtl numbers were
in the order of 4 to 6. Uncertainties in heat transfer coefficient and friction factors were on
average below 2.5% and 10% respectively.

Adiabatic friction factor results showed that the use of different inlets influenced the commencement
of transition. The smoother the inlet profile the more transition was delayed,
confirming previous work done. The effect of increasing tube diameters had a slight delay in
transition. Enhanced tubes caused transition to occur at lower Reynolds numbers which was
accounted for by the fin height and not the helix angle. Heat transfer results showed that
transition occurred at approximately the same Reynolds number for all the different inlets
and enhanced tubes. This was attributed to the secondary flow forces influencing the growing
hydrodynamic boundary layer. These secondary flow forces also influenced the laminar heat
transfer and diabatic friction factors with both these parameters being higher. Turbulent enhanced
tube heat transfer results were higher than those of the smooth tube, with the tube
with the greatest helix angle showing the greatest increase. Correlations were developed for all
the tubes and their inlets and predicted all the data on average to within 3%.

Supervisors: Prof J P Meyer and Prof L Liebenberg
 

Dr  O S Motsamai  2009 "Optimisation Techniques For Combustor Design"

 Dr J  Wannenburg: 2007 "A study of fatigue loading on automotive and transport structures".

Defective structural designs are mostly caused by insufficient knowledge of input data, such as material properties or loading, rather than inadequate analysis or testing methods. In particular, loads associated with automotive and transport (trucks, trailers, containers, trains) structures are nontrivial to quantify. Such loads arise from stochastic and ill-defined processes such as driver/operator actions and structure-terrain interaction. The fundamental processes involved with the determination of input loading are measurements, surveys, simulation, estimation and calculation from field failures. These processes result in design criteria, code requirements and/or testing requirements. The present study deals with methods for the establishment of input loading for automotive and transport structures. It is attempted to generalize and unify new and existing techniques into a cohesive methodology. This is achieved by combining researched current theory and best practices, with lessons learnt during application on, as well as new techniques developed for, a number of complex case studies, involving road tanker vehicles, light commercial vehicles, industrial vehicles, as well as tank containers. 

Supervisor: Prof P S Heyns
Co-supervisor: Dr A D Raath

(Full text of thesis can be found here)

Dr M Thoresson: 2007 "Efficient Gradient-Based Optimisation of Suspension Characteristics for an Off-Road Vehicle"

The efficient optimisation of vehicle suspension systems is of increasing interest to vehicle manufacturers. The main aim of this thesis is to develop a methodology for efficiently optimising an off-road vehicle's suspension for both ride comfort and handling, using gradient-based optimisation. Good ride comfort of a vehicle traditionally requires a soft suspension setup, while good handling requires a hard suspension setup. The suspension system being optimised is a semi-active suspension system that has the ability to switch between a ride comfort and a handling setting. This optimisation is performed using the gradient-based optimisation algorithm Dynamic-Q. In order to perform the optimisation, the vehicle had to be accurately modelled in a multi-body dynamics package. This model, although very accurate, exhibited a high degree of non-linearity, resulting in a computationally expensive model that exhibited severe numerical noise. In order to perform handling optimisation, a novel closed loop driver model was developed that made use of the Magic Formula to describe the gain parameter for the single point driver model's steering gain. This non-linear gain allowed the successful implementation of a single point preview driver model for the closed loop double lane change manoeuvre, close to the vehicle's handling limit.

Due to the high levels of numerical noise present in the simulation model's objective and constraint functions, the use of central finite differencing for the determination of gradient information was investigated, and found to improve the optimisation convergence history. The full simulation model, however, had to be used for the determination of this gradient information, making the optimisation process prohibitively expensive, when many design variables are considered. The use of carefully chosen, simplified, two-dimensional, non-linear models were investigated for the determination of this gradient information. It was found that this substantially reduced the total number of expensive full simulation evaluations required, thereby speeding up the optimisation time.

It was, however, found that as more design variables were considered, some variables exhibited a lower level of sensitivity than the other design variables, resulting in the optimisation algorithm terminating at sub-optimal points in the design space. A novel automatic scaling procedure is proposed for scaling the design variables when Dynamic-Q is used. This scaling methodology attempts to make the n-dimensional design space more spherical in nature, ensuring the better performance of Dynamic-Q, which makes spherical approximations of the optimisation problem at each iteration step. The results of this study indicate that gradient-based mathematical optimisation methods may indeed be successfully integrated with a multibody dynamics analysis computer programme for the optimisation of a vehicle's suspension system. Methods for avoiding the negative effects of numerical noise in the optimisation process have been proposed and successfully implemented, resulting in an improved methodology for gradient-based optimisation of vehicle suspension systems.

Supervisor: Dr P E Uys
Co-Supervisor: Dr P S Els

(Full text of thesis can be found here)

Dr PS Els:  2006   "The Ride Comfort vs. Handling Compromise for Off-Road Vehicles."

Schalk Els examines the classic ride comfort vs. handling compromise when designing a vehicle suspension system using mathematical modelling and field tests.

The full vehicle, non-linear mathematical model, built in MSC ADAMS software, is verified against test data, modified to incorporate hydropneumatic springs and used to obtain optimised spring and damper characteristics for ride comfort and handling respectively. It is found that these optimised results are at opposite corners of the design space, i.e. ride comfort requires a soft suspension while handling requires a stiff suspension. The ride comfort vs. handling compromise can only be eliminated by having a controllable suspension system that can switch between a soft and a stiff spring, as well as low and high damping. This switching must occur rapidly and automatically without driver intervention.

A prototype 4 State Semi-active Suspension System (4S4) is designed, manufactured, tested and modelled mathematically. This system enables switching between low and high damping, as well as between soft and stiff springs in less than 100 milliseconds.

A control strategy to switch the suspension system between the “ride” mode and the “handling” mode is proposed, implemented on a test vehicle and evaluated during vehicle tests over various on- and off-road terrains and for various handling manoeuvres. The control strategy is found to be simple and cost -effective to implement and works extremely well. Improvements in the order of 50% can be achieved for both ride comfort and handling. 

Supervisor: Prof N J Theron

(Full text of thesis can be found here)

Dr NDL Burger:  2006   "Failure analysis of ultra high molecular weight polyethylene acetabular cups."