Showing papers in "International Journal of Experimental Design and Process Optimisation in 2019"

Engineering optimisation of process parameters for polymers: an overview

Abstract: Polymers are one of the most extensively used materials in the manufacturing industry. Modified to the requirement or specification for a particular application, a variety of methods may be used in processing these materials. To fulfil the requirement of the application and improve the performance of end product, an optimal combination of process parameters is required. This may be achieved through optimisation, a promising tool, which provides better performance at a reduced cost. By employing a suitable optimisation technique, the properties of polymers can be predicted without performing experiments, which would be very beneficial in terms of time and money saving by preserving materials normally consumed during the experimental optimisation phase. In this study, different optimisation techniques such as genetic algorithm (GA), artificial neural network (ANN), particle swarm optimisation (PSO), Taguchi method and others, which have been used to optimise the process parameters in polymers, are discussed in detail. In addition, the detailed algorithm and mathematical expressions used to apply these optimisation techniques have also been presented.

Quantifying the effect of construction site factors on concrete compressive strength using designed experiments

Abstract: Both structured and unstructured factors affect concrete compressive strength. Structured factors, e.g., raw materials, affect concrete production. Unstructured factors, e.g., local conditions, affect concrete quality during the construction phase. The effects of structured factors on concrete metrics are well understood, while there is limited understanding on the effects of unstructured factors. A full 25 factorial design was conducted to quantify the effect of five construction site factors including crew experience, compaction method, mixing time, curing humidity and curing temperature. A linear regression model was developed considering significant affecting factors. Model adequacy was evaluated through residual plots. The results indicate that compaction method, mixing time, curing humidity and curing temperature affect concrete compressive strength significantly, with curing temperature having the highest percent contribution (50.0%). The final regression model was used to create a decision-support tool that enables construction workers to find operating conditions and take corrective actions to preserve concrete quality.

Optimisation of fuel in fire tube saturated steam boiler

Abstract: Considerable improvements have been made to minimise the fuel consumption in industrial boilers since a sizable portion of the operational cost can be reduced with fuel optimisation. In this study a combination of air preheater and condensing economiser is proposed to optimise the fuel usage in a fire tube heat boiler. The process utilises 10% excess air that is preheated to 96°C in an air preheater using the heat of stack flue gases. This improves the boilers efficiency by 3%. Moreover, makeup water is also heated to 84°C in the condensing economiser using the heat of stack flue gases coming from the air preheater. This proposed assembly extracts most of the energy from stack flue gases before it start to condense. A simulation using ASPEN HYSYS® shows that using preheated make up water and preheated combustion air, fuel demand in the boiler is reduced by 10%, thus making the process more economical.

Optimum design for an incomplete second degree mixture model

Abstract: Mixture experiments form an important part of research in areas like pharmaceutical, biometrics, agriculture, etc. Optimum mixture designs have been derived by many authors for mixture models due to Scheffe (1958) and others. In most of these models, there is no discrimination among the components. However, in practice, the situation may be quite different. In this paper, we have considered a situation where some but not all of the interaction terms are present. Optimum designs have been determined for the estimation of parameters of a second degree model. Finally, optimum designs for prediction have also been considered.

Non-sequential augmentation strategies to address separation in logistic regression

Abstract: Previous research on small sample multi-factor D-optimal designs for the logistic regression model has demonstrated that these designs are prone to encountering separation, a phenomenon where the responses are completely or quasi-completely separable by a hyperplane in the design space. Separation causes the non-existence of maximum likelihood parameter estimates and represents a serious problem for model fitting purposes. In this paper, several non-sequential design augmentation strategies, where additional experimental trials are performed following an initial experiment that has encountered separation, are investigated. Small local and Bayesian D-optimal initial designs are generated for several representative logistic regression models, and a Monte Carlo simulation methodology is then used to investigate the effectiveness of each augmentation strategy in eliminating separation. Results of the simulation study show that augmenting design runs (trials) in regions of maximum prediction variance (MPV) is the most effective strategy for eliminating separation. However, MPV augmentation tends to produce designs with lower D-efficiencies. The paper illustrates that MPV augmentation reliably eliminates separation and can be used in practice to obtain usable parameter estimates for the logistic regression model.