Search

The sheet metal spinning is frequently employed in the sheet metal manufacturing industry due to its great capability to form a wide range of materials for a variety of industrial products. During the sheet metal spinning process, a flat workpiece is rotated at a high speed to enable forming of the final shape of a spun component. The workpiece is deformed by moving a forming tool gradually onto the mandrel surface using either single-pass or multiple-passes. Due to the nature of localised material deformation, the sheet metal spinning has inherent advantages, such as low production costs, high material utilisation, good dimensional accuracy and enhanced mechanical properties. It is suitable for manufacturing light-weight components with flexibility in many manufacturing industries such as for aerospace, automotive, medical equipment applications. Nevertheless, the spinning process design, i.e. toolpath design and selection of forming process parameters, still highly relies on empirical knowledge and trial-and-error practice to obtain correct process setting and hence, avoid the component defects, especially wrinkling failure. Without guidance for the selection of spinning parameters, the manufacturing efficiency and product quality are severely restricted. This study aims to develop an analytical wrinkling prediction model for the first-pass conventional spinning therefore to provide a guidance of parameter setting and to support the process design of spinning. In this study, the multiple-passes of the forming tool commonly used in the conventional spinning are parameterised by using the forming depth and the radius of workpiece after each pass. The cubic Bézier curve is employed to parameterise a specific toolpath profile including linear path, convex path and concave path. The methods of calculating the control points of the cubic Bézier curve for the final toolpath profile and the first toolpath profile are proposed, respectively. The capability and accuracy of the cubic Bézier curve are examined by comparing the geometrical shapes of different spun components. Furthermore, the contact path created by the tool movement on the workpiece surface is plotted to analyse the effect of the feed ratio on the contact between tool and workpiece. Wrinkling failure in the first-pass conventional spinning process can be influenced by both process parameters, such as feed ratio and forming depth, and material parameters, such as spin ratio and workpiece thickness. Finite Element (FE) models of the first-pass conventional spinning are developed to analyse the influences of these parameters on wrinkling failure using the convex toolpath profile for the material of mild steel (DC01). The significance and relative significance of these parameters on wrinkling initiation in spinning are investigated by developing FE models using the Taguchi method. The FE results show that the workpiece thickness is the most significant factor affecting wrinkling in the first-pass conventional spinning, followed by the forming depth, the spin ratio, and the feed ratio. A mathematical wrinkling wave function is established considering the features of wrinkling waveform. This is validated by comparing the geometrical features between the model produced by the wrinkling wave function and the wrinkled FE model. An analytical wrinkling prediction method based on the conservation of energy is established for the first-pass conventional spinning with the convex toolpath profile. The strain energy (internal energy) and work done by the forming tool (external energy) are calculated, respectively. The critical condition of wrinkling occurrence can be determined when the external energy is equal to the internal energy therefore to enable the critical forming depth to be calculated. The influences of various parameters, including spin ratio, feed ratio, and workpiece thickness, on the critical forming depth is systematically investigated by the wrinkling prediction method using the Box-Behnken Design of the analytical models. It is found that the critical forming depth is affected by the spin ratio significantly, followed by the workpiece thickness and the feed ratio. The experimental testing of the first-pass conventional spinning using the convex toolpath profile is conducted to validate the developed analytical wrinkling prediction model for the material of DC01. A time-domain measurement system using the vibration sensor is employed to record the initiation and development of wrinkling failure during spinning tests. The accuracy of the analytical wrinkling prediction model is validated based on a small range of spinning parameters firstly in the preliminary experiment. Combining the results from the FE models and the full experiment, the developed analytical wrinkling prediction model is further validated under various spinning conditions. The processing map to predict wrinkling failure is produced for the first-pass conventional spinning using the convex toolpath profile. The contribution to knowledge of this study is an analytical model for the prediction of wrinkles in the first-pass conventional spinning process using the convex toolpath. The efficiency and accuracy of this model have been validated by both experiment and FE modelling approaches. This model has potential to be used in evaluating the effect of the key spinning parameters on wrinkling occurrence thus to provide guidance for the selection of spinning parameters in the actual production of spun components.

Vapour compression refrigeration (VCR) system is one of the most commonly used refrigeration methods for domestic refrigeration, heat pump, and automobile air conditioning. The ever rapid rise in the number of VCR units causes a dramatic increase in energy demand and greenhouse emissions, which results in a significant contribution to the global warming. To mitigate the greenhouse gas emission and improve energy efficiency, high efficiency and low carbon oil-free refrigeration system with capacity modulation can be considered as a potential solution for VCR units. A novel oil-free refrigeration system (ORS) is introduced in this thesis. The performance of the ORS has been experimentally and numerically investigated. The ORS consisting of two balanced novel oil-free linear compressors, an off-the-shelf water-cooled coaxial condenser and an evaporator with an electric heater has been instrumented. A series of experiments have been carried out over a wide range of test conditions to investigate the performance of the linear compressor using R134a. At a compressor stroke of 11 mm and a pressure ratio of 2.0, the oil-free linear compressor can achieve a volumetric efficiency of 71%. The oil-free linear compressor is capable of capacity modulation in response to heat load. Part load efficiency is even higher. An analytical model of the linear compressor has been developed. The drop-in performance of R1234yf for R134a system also has been experimentally investigated using the ORS test rig. At a condenser temperature of 40 °C, the coefficient of performance (CoP) of R1234yf is 5-20% lower than R134a. The effect of the refrigerant charge and refrigerant distribution for the ORS using R134a has been experimentally evaluated. The experimental results show that an optimal refrigerant charge, which allows the system to achieve the highest efficiency, varies with operating conditions. A higher compressor stroke and a lower pressure ratio tend to have a higher optimal refrigerant charge. To provide the numerical ORS model with evaporation heat transfer and pressure drop of refrigerants, two existing correlations for heat transfer and pressure drop for R1234yf, R152a and R134a have been improved based on the experimental data. Overall, the modelling results agree well with the measurements. A comprehensive numerical system model for the ORS consisting of linear compressor model, a heat exchanger model, and a refrigerant distribution model has been proposed using MATLAB and Simulink. The model is able to predict mass flow rate, power input, pressure drop and heat transfer in heat exchangers, and CoP of the refrigeration system. Overall, the modelling results agree well with measurements. The numerical model can be used for future work on low charge ORS design using oil-free linear compressor and microchannel heat exchangers for various applications.