http://hdl.handle.net/10149/56733 Tue, 21 May 2013 02:34:59 GMT 2013-05-21T02:34:59Z http://hdl.handle.net/10149/194590 Title: Innovative methodology for location-based scheduling and visualisation of earthworks in road construction projects Authors: Shah, R. K. (Raj) Abstract: This thesis focuses on the development of an innovative location-based scheduling methodology and a computer-based model for improving earthwork operations in road construction projects. Analysis of existing planning and scheduling practices in road construction projects conducted in the course of this research concluded that planning, scheduling and resource allocation are largely dependent on subjective decisions. Also, shortcomings exist due to the distinct characteristics of earthworks, e.g. one-off projects with uncertain site conditions and soil characteristics, causing delays and cost overruns of projects. The literature review found that existing linear scheduling methods provide inaccurate location-based information about earthworks and fail to integrate different productivity rates. A survey was used to capture and analyse industrial practices and issues related to delays and cost overruns. This analysis revealed that the accurate location-based information is vital for efficient resource planning and progress monitoring. Following these findings, a theoretical framework and specification were developed to automate location-based scheduling and visualisation of information. A prototype model was developed by integrating road design data, sectional quantities, productivity rates, unit cost, site access points, and arithmetic algorithms. The algorithms underpinning the model enable the generation of time-location plans automatically as a key output of the model. Weekly progress profiles, space congestion plans, and cost S-curves are the other outputs. A cut-fill algorithm was developed to identify optimum quantities of earthwork and its associated costs. Experiments were conducted with design data provided by a road construction company to demonstrate the model‟s functionality. Sensitivity analysis was used to identify the critical factors relating to earthwork scheduling. It was found that the model is capable of generating time-location plans, considering the critical factors and location aspects. Finally, the model was evaluated using a case study and validated by road construction professionals using an indirect comparison method. It was concluded that the model is a valuable tool for producing location-based scheduling, optimising resource planning and assisting in the communication of scheduling information from the location viewpoints in the earthwork projects. Tue, 15 Nov 2011 00:00:00 GMT http://hdl.handle.net/10149/194590 2011-11-15T00:00:00Z http://hdl.handle.net/10149/117946 Title: Development of simulation-based genetic algorithms model for crew allocation in the precast industry Authors: Al-Bazi, A. F. J. (Ammar) Abstract: The focus of this thesis is on the precast concrete products manufacturing industry, which as one of the labour-intensive industries requires a substantial number of highly skilled operators in terms of crews to produce the final product. A crew is a group of multi-skilled chargehands and operators that have various skills and experience necessary to conduct an activity in a professional way. The high cost of skilled operators and the apparent inefficiencies of utilising such skilled operators in the industry are the major driving force. To achieve this, optimal crew allocation is required. Crew allocation is complex because of the multi-criteria nature of the problem and availability of thousands of possibilities and allocation alternatives. There is a gap in previous research efforts associated with crew allocation planning in the precast industry. Current practices suggest that the crew allocation process is carried out intuitively and the allocation of crews to production processes is subjective. This has led to high process-waiting times, improper allocation of skilled operators and ultimately higher production costs. In this context, the aim of this research is to propose an effective crew allocation methodology and a computer-based intelligent simulation model for its implementation. The objective of the approach is to guarantee a better workflow through minimising process-waiting time, optimising operator utilisation, and subsequently reducing the allocation cost. This research develops a holistic and integrated methodology for modelling crew allocation problems by reviewing state-of-art resource allocation techniques, structured interviews with production managers, site visits and a detailed case study. The methodology is developed using an IDEF0 process model and a generic process map for both the business and the production processes of the precast manufacturing system. A multi-layered genetic algorithm model is developed in conjunction with a process-simulation model to form a hybrid allocation system dubbed ‘SIM_Crew’. The model incorporates databases (Excel and MS Access), a simulation model (developed using Arena 12.0) and genetic algorithms (developed using Visual Basic for Applications) to facilitate the generation and evaluation of various “what-if” crew allocation scenarios. A number of performance criteria have been developed to evaluate the allocation plans. ‘SIM_Crew’ enables the investigation and analysis of allocating possible schedules and provides a facility to visualise the production processes. ‘SIM_Crew’ was validated using real life case study data and it was concluded that the allocation of crews to precast processes using genetic algorithm improves the throughput time and reduces the allocation cost as compared with real life production data. It is anticipated that future use of this research will solve the crew allocation problem in the precast industry. Fri, 11 Jun 2010 00:00:00 GMT http://hdl.handle.net/10149/117946 2010-06-11T00:00:00Z http://hdl.handle.net/10149/117905 Title: Novel applications of optical analytical techniques Authors: Seetohul, L. N. (Nitin) Abstract: Novel applications of optical analytical techniques have been demonstrated in three general areas, namely application of broadband cavity enhanced absorption spectroscopy (BBCEAS) to the detection of liquid phase analytes, the use of total luminescence spectroscopy to discriminate between different type of teas and the development of an optical sensor to detect ammonia gas, based on the fluorescence quenching of a dye immobilised in a sol gel matrix. A simple BBCEAS setup has been developed with a view to perform sensitive visible wavelength measurements on liquid phase solutions. In the present work a simple low-cost experimental setup has been demonstrated for the measurement of the visible spectra of representative liquid-phase analytes in a 2 mm quartz cuvette placed at normal incidence to the cavity mirrors. Measurements on Ho3+ and sudan black with a white LED and the R ≥ 0.99 mirrors covered a broad wavelength range (~250 nm) and represents the largest wavelength range covered to date in a single BBCEAS experiment. The sensitivity of the technique as determined by the best αmin value was 5.1 x 10-5 cm-1 and was obtained using the R ≥ 0.99 mirrors. The best limit of detection (LOD) for the strong absorber brilliant blue-R, was approximately 620 pM. The optical setup was then optimised for the application of BBCEAS detection to an HPLC system. A 1 cm pathlength HPLC cell with a nominal volume of 70 l was used in this study. The cavity was formed by two R ≥ 0.99 plano-concave mirrors with a bandwidth of ~ 420 – 670 nm. Two analytes rhodamine 6G and rhodamine B were chosen for separation by HPLC, as they were chemically similar species with distinctive visible spectra and would co-elute in an isocratic separation. The lowest value of min obtained was 1.9 x 10-5 cm-1. The most significant advantage of the HPLC-BBCEAS study over previous studies arose from the recording of the absorption spectrum over a range of wavelengths. It was demonstrated that the spectral data collected could be represented as a contour plot which was useful in visualising analytes which nearly co-eluted. The LOD values for the two analytes studied indicated that the developed HPLC-BBCEAS setup was between 54 and 77 times more sensitive than a commercial HPLC system. For improved sensitivity and lower detection limits the low cost BBCEAS setup was used with a significantly longer 20 cm pathlength cell where the mirrors were in direct contact with the liquid phase analyte. This also reduced interface losses. The experiments were carried out using both R  0.99 and R  0.999 mirrors. The lowest αmin value obtained in this study was 2.8 x 10-7 cm-1 which is the lowest reported value to date for a liquid phase measurement, making this study the most sensitive liquid phase absorption measurement reported. The lowest LOD recorded was 4.6 pM, and was obtained for methylene blue with the R  0.999 mirrors. A novel application of total luminescence spectroscopy to discriminate between different types of teas objectively was also investigated. A pattern recognition technique based on principal component analysis (PCA) was applied to the data collected and resulted in discrimination between both geographically similar and dissimilar teas. This work has shown the potential of fluorescence spectroscopy to distinguish between seven types of teas from Africa, India, Sri Lanka and Japan. Geographically similar black teas from 15 different plantation estates in Sri Lanka were also studied. The visualisation technique allowed the separation of all 11 types of teas when the first two principal components were utilised. The final part of the thesis describes the development of an optical sensor for the detection of ammonia gas. The operation of the sensor depended on the fluorescence quenching of the dye 9 amino acridine hydrochloride (9 AAH) immobilised in a sol gel matrix. It was also shown that the sensor response was not affected by the presence of acidic gases such as HCl and SO2. The final version of the sensor made use of dual channel monitoring to improve the sensitivity of the sensor. Measurements using diluted mixtures of ammonia gas in the range 5 -70 ppm showed that the response of the sensor was nonlinear, with the sensitivity increasing at lower concentrations. The measurement of the baseline noise allowed the LOD to be estimated at ~400 ppb. Fri, 12 Jun 2009 00:00:00 GMT http://hdl.handle.net/10149/117905 2009-06-12T00:00:00Z http://hdl.handle.net/10149/112685 Title: Development of water-based core technology for light alloys Authors: Ramegowda, M. (Manu) Abstract: This thesis describes, in a manufacturing context, the development of new waterbased core technology for light alloys. Cores used for steel casting are made from fused silica and are removed using hot sodium hydroxide under refluxing (pressurising hot acids). However, aluminium and other light alloys are attacked by sodium hydroxide. Currently there is no good core system for aluminium and other light alloys. It is therefore desirable to find an alternative material/leaching agent combination for casting aluminium and other light alloys. The recent research review has shown that ceramic cores are mostly made by fused silica with different additives. The previous research has suggested using fused silica (different mesh size) as filling material and using magnesium oxide to control the slurry working life of core mixes. Calcium silicate assists core leaching in dilute acid. The plaster (calcium sulphate) in the form of proprietary plasters (Crystcal R, Fine Casting Plaster) is used to create bond and gives strength to the core. Lithium carbonate acts has an accelerator, improving the strengthening effect of the plaster in the cores. The binder (Ludox® AM) and water act (as added materials) to bind the composition Core compositions were made with different core trials to produce a core, which records suitable strength and quick leaching properties for light alloys. Core trials were individually mixed and poured into a wooden core box. Cores were pre-dried for twenty-four hours. Cores were fired to different temperatures for two hours, followed by two hours cooling. The cores were subjected to computerised three-point bend test to record the Modulus of rupture (MOR). The plain strain fracture toughness and Weibull parameters were calculated. The Weibull parameter was plotted using Minitab analysis software. Using the cores, gravity die casting process were carried out. The subsequent castings was dipped in diluted nitric, citric and acetic acid to leach out the core. Using different core compositions, core trials were mixed, poured, dried, tested and leached. The high amount of plaster in core trials records high MOR when cores were fired between 200°- 400° C and the opposite result when fired between 600°- 800° C. The different grade of plasters(CRP,FCP) do not influence the strength. One percent of magnesium oxide gives a very short working life. High amount of binder(Ludox® AM) in core produce strong cores. Workable MOR results can be obtained depending on composition allowing manual handling or a waxing process. The fracture toughness is typical of a brittle material, with matching Weibull parameters. The casting process suggests that the new materials are sufficiently refractory. The cores are leached out using diluted nitric, acetic and citric acid at rates compatible with commercial manufacture. This methodology has successfully produced a core using fused silica with plaster and magnesium oxide for aluminium and possibly for light alloys. Different core trials can be used depending on the specific industrial application relating to strength and removal with acid attacking the metal. Further work is needed to fine tune optimum leaching conditions. Fri, 01 Aug 2008 00:00:00 GMT http://hdl.handle.net/10149/112685 2008-08-01T00:00:00Z