Implementation of Convolutional Neural Networks for Warp Detection in 3D Printed Components manufactured via Fused Filament Fabrication: A Bayesian-Based Automated Approach

Fused Filament Fabrication (FFF) is an additive manufacturing technique commonly used in industry to produce complicated structures sustainably. Although promising, the technology frequently suffers from defects, including warp deformation compromising the structural integrity of the component and, in extreme cases, the printer itself. To avoid the adverse effects of warp deformation, this thesis explores the implementation of deep neural networks to form a closed-loop in-process monitoring architecture using Convolutional Neural Networks (CNN) capable of pausing a printer once a warp is detected. Any neural network, including CNNs, depend on their hyperparameters. Hyperparameters can either be optimized using a manual or an automated approach. A manual approach, although easier to program, is often time-consuming, inaccurate and computationally inefficient, necessitating an automated approach. To evaluate this statement, classification models were optimized through both approaches and tested in a laboratory scaled manufacturing environment. The automated approach utilized a Bayesianbased optimizer yielding a mean accuracy of 100% significantly higher than 36% achieved by the other approach.

Implementation of Convolutional Neural Networks for Warp Detection in 3D Printed Components manufactured via Fused Filament Fabrication: A Bayesian-Based Automated Approach

Fused Filament Fabrication (FFF) is an additive manufacturing technique commonly used in industry to produce complicated structures sustainably. Although promising, the technology frequently suffers from defects, including warp deformation compromising the structural integrity of the component and, in extreme cases, the printer itself. To avoid the adverse effects of warp deformation, this thesis explores the implementation of deep neural networks to form a closed-loop in-process monitoring architecture using Convolutional Neural Networks (CNN) capable of pausing a printer once a warp is detected. Any neural network, including CNNs, depend on their hyperparameters. Hyperparameters can either be optimized using a manual or an automated approach. A manual approach, although easier to program, is often time-consuming, inaccurate and computationally inefficient, necessitating an automated approach. To evaluate this statement, classification models were optimized through both approaches and tested in a laboratory scaled manufacturing environment. The automated approach utilized a Bayesianbased optimizer yielding a mean accuracy of 100% significantly higher than 36% achieved by the other approach.

Investigations on the Dissimilar Metal Joints Capability between Aluminum Alloy and Steel under CMT Welding

Sustainable lightweight vehicle design has becoming a key trend in the future. A new way to join aluminum alloys and steel sheet. Cold metal transfer (CMT) welding of thin sheet metal products. Determination of equilibrium wire-feed-speeds welding-speeds and arc length. Seam formation and microstructure of CMT were researched. Not only that, but optimum welding parameters were analyzed. Study of welding joint microstructure of compound layer on welding joint. Comparing the steel side of welding seam. The aluminum alloy siding grain size was larger. After tensile test, joint was fractured on the aluminum alloy side. Warp deformation occurred on aluminum side. Displacement of warp deformation became larger along with the increase of welding current. The results reviewed in this article indicate that the aluminum alloy fractured is more preferable to steel. CMT welding has found applications in automobile industries, defence sectors and power plants as a method of additive manufacturing.

Optimization Analysis of the Injection Molding Process Based on Orthogonal Test Method and Uniform Design

A car battery bracket was taken as the research object and the warp deformation of plastic products was as the quality evaluation index, through the range analysis of orthogonal test method, the influence degree of the warp deformation from the melt temperature, mold temperature, injection time, packing pressure and dwell time was determined, the preliminary optimization of process parameters combination was obtained. Further uniform design was applied for local optimization by taking the preliminary optimized parameters as the basis, the optimal process parameters combination was obtained and further verified by simulation.

Analysis of Cylindroid Shell Subject to Internal Linearly-Increased Pressure

Close cylindroid shells are widely used in many industrial branches. Membrane theory of shells is used to take an analytical solution to investigate the internal force distributions and deformation laws of such shells. The result shows that, under the condition of two-point simple supports, among three force components(meridional forceT1, circumferential forceT2, and shear forceT12),T1is the dominant one, which is negative (compressional) in the vicinity of the neutral axis, and becomes positive (tensional) after being away from the neutral axis. The shear forceT12is rather like a sine curve, which changes its sign at the neutral axis. This type of shear force distribution leads to a warp deformation within the cylinder.T2is always the tensional force, and when comparing to the other two components, it is too small to be dominant in shell designing. Somewhat similar to the three force components, among the three deformation components, the normal displacementwis the extreme one, and also it varies acutely. The circumferential displacementvis much less thanw, which is compressive below the neutral axis, and becomes tensile above the neutral axis. In the nearby of neutral axis,vis nearly zero. Compared towandv, the meridional displacementuis always the minimal.