|
The list of problems will be updated as problems are submitted and further information about the present problems and additional problems becomes available. Please keep watching these pages. 网易体育 five problems will be selected depending on the expertise and interests of the participants.
- Problem: Minimising the effects of rockbursts?
Industry: Mining industry
Industry representative: Professor Richard Stacey
Moderator:
Student Moderator:
Description:
 Rockbursts in mines are very violent events. Violent ejection of rock from the surfaces of underground openings is usually observed, and the velocities of such ejection can be of the order of 1 to 10 m/sec. After a rockburst event in 2005, it was noticed that, in one localized area, concrete in the floor had been subjected to heave (as a result of some ejection), but that an immediately adjacent mud-filled area was unaffected. This is illustrated in the photograph on the right. The violent loading of the steel arch by the concrete can also be observed in the photograph, but the adjacent arches were clearly not similarly loaded.
In other tunnels similar observations were made ? where there had been water or mud on the floor, there was no observable rockburst damage. It is expected that this behaviour could be the effect of compressive wave reflection at the free surface (concrete, mud or water) and the mud or water has effectively served as a tensile wave trap. The behaviour raises the following questions:
- Could the effects of rockbursts be mitigated in some cases by means of a soft coating or ?gel? covering on the rock surface?
- If there is such a benefit, would it be wave frequency dependent?
- Would the thickness of the ?soft? layer have any effect on the beneficial behaviour?
- What about the effects of other waves, and interactions of waves?
- Problem: Wrinkling of paper labels on beer bottles
Industry: Beverage industry
Industry representative: Thabo Ntlamelle
Moderator: Astri Sjoberg
Student Moderator:
 Description: SAB Ltd produces beer packaged in one way and returnable bottles. All the beer bottles have paper labels applied to them. The current problem is that the labels tend to wrinkle after application to the container, and the wrinkling effect is inconsistent. There are times when there is absolutely no wrinkling on labelled bottles in the same environment as other bottles with serious wrinkling. The degree of wrinkling may also differ from brand to brand, brewery to brewery and from Pack to Pack.
SAB Ltd uses the same paper and adhesive specification across the country, however the combination of adhesive and label suppliers differs, since we have a number of sources of paper that are processed by various printers and in addition we have 3-4 adhesive suppliers. This results in a fairly random combination of label and adhesive.
In the past months, SAB Ltd has gone through problem solving processes to eliminate the wrinkling effect of paper labels, but the problem still persists, although to a less frequent extent. We have evaluated numerous probable causes (see diagram attached) through investigative and benchmarking processes and have come to the conclusion that it may be the interaction between the adhesive (quality and application processes) and the paper (condition and quality wrt finish characteristics, printing process and the paper weight (G/cm2)).
To date the frequency of wrinkling has reduced but not disappeared. As a market driven company, SAB Ltd will endeavour to resolve this problem in order to give the consumer a perfectly labelled beer bottle filled with the best beer in the industry to quench their thirst.
The outcome that would be sought includes:
- Understanding what factors contribute to label wrinkling (beyond or including what SAB Ltd has identified to date).
- Identifying the root causes of our paper label wrinkling effect.
- Recommendations to eliminate paper label wrinkling.
problemstatementwrinklinglabels.doc
- Problem: Modelling notch flow of viscous non-Newtonian fluids
Industry: Mining and Nuclear
Industry representative: Paul Slatter
Moderator: Tim Myers
Student Moderator:
Description: Notches are used routinely for measuring the flow rate of water with free surface flows.
As far as can be ascertained, no work has been done to investigate the performance of these flow rate measurement devices for non-Newtonian fluids. The Flow Process Research Centre at the Cape Peninsula University of Technology has investigated the flow rate of kaolin and bentonite suspensions from a tank for two notch shapes, rectangular and triangular, and for 3 different concentrations of each suspension (Haldenwang et. al., 2007). The notches were calibrated using water and then kaolin and bentonite suspensions were tested in them.
Classical modelling analysis was then applied to the results. Briefly, what this means is that a somewhat simplistic energy balance (Bernoulli) is applied assuming a linear velocity profile, and zero energy losses. Integration over the flow area is then performed to give the fundamental relationship between flow rate and sill depth. Non-idealities and energy losses associated with the flow are accommodated for by introducing a coefficient of discharge Cd, the value of which is obtained by calibration of a real notch. Values for Cd classically lie in the range 0.6<Cd<0.8, comfortably accounting for losses in energy associated with non-linearity of the velocity profile, and losses associated with a boundary layer, flow constriction, etc. Implicit in the conceptualisation of Cd is that it is a loss accounting device, and the physical reality that it can never exceed unity.
Analysis of the experimental results using the classical approach indicates a significant increase in discharge coefficient with decreasing flow rate, and increase in suspension concentration. A particularly acute problem is that the discharge coefficient can significantly exceed unity, apparently defying fundamental energy considerations.
Clearly, the classical modelling approach is unable to model the flow adequately, and an improved modelling approach is called for, which will accommodate the effect of, the increase in, and non-linearity of, viscosity.
Reference: HALDENWANG R, ERASMUS R, SLATTER PT and CHHABRA RP (2007) Flow rate measurement of non-Newtonian suspensions from a tank using notches. Hydrotransport 17. The 17th International Conference on the Hydraulic Transport of Solids, The Southern African Institute of Mining and Metallurgy and the BHR Group; Cape Town, 7-11 May, 2007. ISBN 978 1 85598 084 6, pp301-316.
- Problem: Pre-split blasting technique in surface mines
Industry: Mining industry
Industry representative: Alan Naismith
Moderator:
Student Moderator:
Description:
A method of protecting permanent faces from blasting damage in surface mines is to use the pre-split blasting technique. A series of closely spaced vertical and parallel holes is drilled in the rock and a fracture surface is created by a simultaneous and equal energy input into each hole. This technique appears to be successful in both brittle rock types such as sandstones and basalts and in softer materials such as kimberlite. However, there have been situations recently in which the pre-split technique is effective for about 80% of its height and then it appears to go away. Photograph 1 is attached which shows the general effect.
The problem is to explore analytically and numerically the development of a fracture surface in the pre-split blasting technique. The outcomes that would be sought include:
- Development of a relationship between energy input and fracture development for different rock brittleness.
- Establishing the maximum hole spacing at which fracture propagation will occur.
- Establishing a cause for the apparently anomalous fracture behaviour at the unconfined end of the hole.
- Investigating the method of fracture propagation within a hole, in particular, determining the conditions under which ?multi-nodal? fracture initiation may occur as shown in Photograph 2.
- Development of methods to control the direction and extent of fracture development, particularly in the upper part of the hole and where other fractures intersect the hole.
 
-
Problem: Compression of Hyperspectral Image Data
Industry: Mining Industry
Industry and Academic Representatives: Professor Steve Damelin (Georgia Southern University, US nd Wits), Dr Neil Pendock (Anglo American and Wits) and Professor Michael Sears (Anglo American and Wits).
Moderator:
Student Moderator:
Description:
Hyperspectral image data is obtained from airborne, spaceborne and terrestrial systems where the reflectance from each image point is sampled at a high number of spectral values (typically over a hundred) ranging across the visible and infrared regions of the electromagnetic spectrum. The information obtained allows the identification of, for example, vegetation types and mineral species, and the clustering of other image components.
The processing, interpretation, and storage of this data is complicated by the huge volumes of high dimensional data obtained from even relatively small images. For example, the high spatial and spectral resolution core imaging system (Hyperspectral Core Imager) owned by AngloGold Ashanti and developed by SpecTerra Systems in Australia , which obtains hyperspectral image data from drill core for mineral identification and spatial distribution, acquires a gigabyte of data from just three metres of scanned core!
Our problem concerns the compression of the massive amount of image data which are acquired by such systems. Can the data be compressed in such a way that essentially no information is lost, by exploiting the overlapping (multi-modular) spatial, and high redundancy of the spectral information?
In the presentation, we shall first explain the problem of compression of Hyperspectral Core Imager data, why this is important, how the data s acquired, and why the data is multi-modular. For example we will explain how one typically finds double redundancy in core data with respect to both spatial and spectral pixels. Indeed, overlapping arises both from physical sampling and from the detectors themselves. Thus there is both sampling and data redundancy to deal with in trying to understand the nature of both the sample taken and the performance of the imaging instrument.
Multi-modality may also arise because of different sources measuring different aspects of the data at the same time. For example, core data presents us with a combination of qualitative and categorical variants. In the case of double redundancy occurring naturally in overlapping data samples, say hyperspectral data from core images, we will suggest compressive sampling and wavelet techniques to remove the redundancy which is necessary for storage and transmission of images. Finally we will describe the challenge of developing methods to extract, cluster and predict patterns in an exploratory manner. In this case, we are expected to build graphs on the data by way of kernels which are adaptive, for example signature and prim-like kernels.
-
Problem: Time domain dynamic simulations of rivers and water networks
Industry:
Industry representative: Tony Lange, Optin
Moderator:
Student Moderator:
Description:
There are three models.
(i) River model
Time domain dynamic simulations of rivers and water networks need to be performed. We would like to model a river as being made up of river segments, where water flows from one segment to the next. The river segments have a number of attributes all functions of time and geometry as follows:
geometry ? length, cross-section area, surface area
volume flows ? inflow, outflow, evaporation, seepage into ground, runofffrom sides, abstraction or pumping out of water.
The river segment models need to be quite robust and should if possible avoid iterative solutions to converge on solutions. If possible, pollution or chemical contamination should be considered as well as possible hydraulic conveyance of silt.
A real time expert system is used as the main computing platform. The platform is object oriented and understands connectivity and forward and backward chaining. We need to determine the basic formulae required per water segment. The river network will then be taken care of. The whole system should be highly flexible and behave like an actual river system rather than as a huge set of simultaneous equations. The question of how complex or simple the river segments need to be is not a trivial problem which is the key objective.
This approach to modelling river systems is a major change from the traditional approach which is in terms of large matrices. The problem is to decide how one embeds the water dynamics in the interconnected river segments. What behaviour is the river segment given. How are the river segments linked together, what is fed from one segment to another, what internal equations are engineered into each segment. As simple equations as possible are needed to minimise computation.
(ii) Dam model
Would the dam model be a special case of the river segment.
(iii) Model of high resolution river flows from low resolution river flows.
We have a number of time domain sequences of river flows in month resolutions going back a large number of years. The object is to use this data to feed time domain dynamic simulation but we need to simulate at daily resolution versus monthly resolution. Corresponding daily river flows or even rainfalls for some small time spans for the same river are available. The problem is to determine how we can use these few daily river flows as examples to somehow deconvolve the monthly river flows into daily river flows.
Existing approaches to the deconvolvation problem results in a huge set of matrix equations that are solved iteratively. We are looking for, if possible, simpler sets of equations.
This model will be used to feed the river network model described in (i) above
|
