Abstract:Plasma transferred arc additive manufacturing (PTA-AM) is a high volume and high speed directed energy deposition (DED) technique that uses metal powder as a feedstock material to build up 3-dimensional (3D) components, layer by layer. In the PTA-AM process, the quality of the build is a function of several parameters, including but not limited to the powder feed rate, the torch travel speed, stand-off distance (SOD), and plasma arc current. Unlike the laser DED systems, the melt pool and laser visualization are not possible through inline imaging systems. These processes have difficulty sustaining uniform evenly distributed powder flow between the nozzle and the substrate. For the constant operating parameters, the process requires in-depth attention to inspect the quality and geometrical features of bead deposition, which may alter due to continuous mechanical connection with the deposition process. Surprisingly, no work is reported in plasma-based DED systems to determine the process parameter influence on build quality using image processing technique. Therefore, in this work, high-speed imaging is employed to analyze and quantify the process parameters of the PTA-AM and their effects on the bead quality and geometry. The experiments of single-track multi-layer bead deposition of Tungsten Carbide (WC)-reinforced NiCrBSi composites are carried out and recorded with a high-speed camera. Based on captured 2D image data, the image processing techniques are developed using Matlab and Image-Pro premier software. The proposed techniques optimize and validate the SOD to print the stable bead and maintain the higher powder catchment efficiency at the deposition zone. It is also used to assess the process stability (uniform powder flow) and the impact of powder flow variation on the bead height. The powder stream distribution, the statistical dispersion of powder particle count, and their effect on the build quality are described based on powder flow images, using an image processing approach. The significance of relationships between bead height variation and powder particle counts is analyzed using statistical methodologies such as analysis of variance (ANOVA) and two-tailed t-test. The effect of PTA-AM process parameters on bead geometry reveals that the plasma arc current and powder particle count significantly affect the bead geometry. This work validates the benefit of using 7-mm SOD for improving part quality. Using image analysis, the uniformity of powder flow is evaluated. In addition, the finding shows that the increment in plasma current by 40% increases the mean bead width by 85% and decreases its height by 24%. By determining the effect of processing parameters on the bead quality and geometry, this PTA-AM process can build near-net-shape parts and increase the reliability and reproducibility of the process. Thus, the proposed in situ image processing methods can be adopted in the PTA-AM process to analyze manufactured parts’ quality and characterize the process parameters responsible for the deviation in surface quality and dimensional geometry.

Surface roughness prediction in micro-plasma transferred arc metal additive manufacturing process using K-nearest neighbors algorithm

Prediction of Metal Additively Manufactured Surface Roughness Using Deep Neural Network

Implementing machine learning in robotic wire arc additive manufacturing for minimizing surface roughness

Thermal Field Prediction for Welding Paths in Multi-Layer Gas Metal Arc Welding-Based Additive Manufacturing: A Machine Learning Approach

Machine learning for multi-dimensional performance optimization and predictive modelling of nanopowder-mixed electric discharge machining (EDM)

Prediction of surface roughness and depth of cut in abrasive waterjet milling of alumina ceramic using Machine learning algorithms

Experimental investigation and development of a deep learning framework to predict process-induced surface roughness in additively manufactured aluminum alloys

Surface quality prediction by machine learning methods and process parameter optimization in ultra-precision machining of AISI D2 using CBN tool

Machine learning approach towards laser powder bed fusion manufactured AlSi10Mg thin tubes in laser shock peening

A new intelligent approach of surface roughness measurement in sustainable machining of AM-316L stainless steel with deep learning models

Machine learning approach for prediction analysis of aluminium alloy on the surface roughness using CO2 laser machining

Comprehensive stratified K-fold machine learning for surface roughness prediction with 3D printed electrodes: Analyzing dielectric variability and heat treatment effects

Image Data-Based Surface Texture Characterization and Prediction Using Machine Learning Approaches for Additive Manufacturing

The Temperature Field Prediction and Estimation of Ti-Al Alloy Twin-Wire Plasma Arc Additive Manufacturing Using a One-Dimensional Convolution Neural Network

Prediction of Surface Roughness Using Machine Learning Approach in MQL Turning of AISI 304 Steel by Varying Nanoparticle Size in the Cutting Fluid

Downskin Surface Roughness Prediction with Machine Learning for As-Built CM247LC Fabricated Via Powder Bed Fusion with a Laser Beam

A universal predictor-based machine learning model for optimal process maps in laser powder bed fusion process

Prediction of Metal Additively Manufactured Bead Geometry Using Deep Neural Network

Modelling and optimization of weld bead geometry in robotic gas metal arc-based additive manufacturing using machine learning, finite-element modelling and graph theory and matrix approach

The development of an ANN surface roughness prediction system of multiple materials in CNC turning

In situ image processing for process parameter-build quality dependency of plasma transferred arc additive manufacturing