The efficiency of a drinking fountain

In this experiment the efficiency of a water drinking fountain was studied. Efficiency was defined in terms of the percentage of water actually consumed in relation to the water that flows from the drinking fountain, the drinking time, the water consumed per user and the electricity required to cool the water. For this purpose four different spout diameters were used and the water pressure was regulated to give two different water trajectory heights for each spout diameter setting. This resulted in eight different flow rates. The results of the study indicated that the water fountain is a highly inefficient water transfer mechanism and that extreme flow rates worsened the efficiency of the system. The intermediate spout nozzle diameters proved to be superior with regard to all the efficiency statistics.     

Experimental Liquidus Surface Projection of the Ni-Ru-Zr System

SEM with EDX and XRD analyses were used to study 21 as-cast Ni-Ru-Zr alloys. Phases were distinguished by BSE image contrast, and the primary phases were determined by their dendritic morphology, with the solidification sequences derived from the microstructure. Identification of the phases was done by EDX analyses, and they were verified by XRD. From these analyses, the solidus and liquidus surface projections were drawn. Three ternary phases were found: τ₁ of composition ~Zr₂₄Ru₂₂Ni₄₄ (at.%), τ₂ at ~Zr₇₄Ru₄Ni₂₂ (at.%), and τ₃ at ~Zr₃₅Ru₃Ni₆₂ (at.%). Sixteen invariant reactions were identified, and the largest liquidus surface was for ~ZrRu.     

Parallel neural net training on the KSR1

In modern day pattern recognition, neural nets are used extensively. General use of a feedforward neural net consists of a training phase followed by a classification phase. Classification of an unknown test vector is very fast and only consists of the propagation of the test vector through the neural net. Training involves an optimization procedure and is very time-consuming since a feasible local minimum is sought in high-dimensional weight space. In this paper we present an analysis of a parallel implementation of the backpropagation training algorithm using conjugate gradient optimization for a three-layered, feedforward neural network, on the KSR1 parallel shared-memory machine. We implement two parallel neural net training versions on the KSR1, one using native code, the other using P4, a library of macros and functions. A speedup model is presented which we use to clarify our experimental results. We identify the general requirements which render the parallel implementation useful, compared to the sequential execution of the same neural net training procedure. We determine the usefulness of a library of functions (such as P4) developed to ease the task of the programmer. Using experimental results we further identify the limits in processor utilization for our parallel training algorithm.     

Biscuit-Making Quality Prediction Using Heritability Estimates and Correlations

The aim of this study was to predict biscuit-making quality by using correlations between and, heritabilities of, the most important soft wheat quality characteristics. A full diallele cross was made between six spring wheats. The parents and 30 F ₁ hybrids were used as experimental material. Percentage vitreous kernels, breakflour yield, farinograph absorption, farinograph development time and alkaline water retention capacity were highly heritable and were also significantly correlated with sugar-snap biscuit diameter. Selection for these characteristics should lead to improvement of biscuit-making quality. In general, softer kernels and a weak dough with a low water absorption were desirable for good biscuit-making quality.     

Generalised theory of a non-TEM transmission line taper without discontinuities

The optimum Chebyshev dispersive tapered transmission line requires impedance steps at the taper ends, which limits its application. A generalised theory for the synthesis of a transmission line taper without impedance steps and supporting non-TEM modes is presented. The impedance discontinuities are avoided by using a near-optimum frequency response.<>     

Extraction of carbon 14-labeled compounds from plant tissue during processing for electron microscopy

Loss of ¹⁴C-labeled compounds from bean leaf tissue was monitored during all the stages of routine specimen preparation. No significant differences in extraction were associated with the use of acetone, ethanol, or dioxane as dehydration fluids. Fixation at low temperature increased the loss of label. Prolonged fixation in glutaraldehyde increased the loss, but fixation in osmium solutions for periods as long as 4 hr had no influence on extraction. Buffer rinses and dehydration fluids caused appreciable amounts of label to be extracted. The use of propylene oxide as transition fluid resulted in low extraction. Some embedding media caused the loss of small amounts of labeled compounds, but one of the media tested (LR-white) extracted significant amounts of label.     

Development of a model to predict fission product behaviour in spherical fuel elements during water ingress events

At PBMR gaseous fission product releases from spherical fuel elements under normal conditions are calculated by the code NOBLEG. The ability of NOBLEG to calculate noble gas and halogen release under oxidizing conditions during water ingress was developed. Observations made during the water vapour injection tests performed during the irradiation experiment HFR-K6, were used to determine simple relations that can be used to predict gaseous fission product release from spherical fuel elements under oxidizing conditions caused by small water ingress events, for PBMR operational temperatures. A new model was proposed to explain peculiarities observed during the water injection tests.