Contribution to Computer Control and Optimization of Hydro-Abrasive Jet Machining Process

Publisher Website
Google Scholar
Abstract
Hydro-Abrasive Jet Machining (HAJM) has demonstrated its suitability for several applications in the machining of a wide spectrum of materials (metals, polymers, ceramics, fibre reinforced composites, etc.). The paper is a contribution to the computer control, integration and optimization of HAJM process in order to establish a hierarchical control architecture and a platform for the implementation of a real-time Adaptive Control Optimization (ACO) module. The paper presents the approach followed and the main results obtained during the development and implementation of a HAJM cell and its computerized controller. A critical analysis of the process variables available in the literature is presented, in order to identify the process variables and to define a process model suitable for HAJM real-time control and optimization. Besides for HAJM computer control, in order to correlate process variables and parameters with machining results, a process model and an optimization procedure are necessary in order to avoid expensive and time-consuming experiments for the determination of optimal machining conditions. The paper presents the configuration of the cell and the specific components adopted in order to make possible a fully computerized control of the process, and the architecture of the controller, capable to manage the several logical and analogical signals from the different modules of the cell, for multiprogramming, process monitoring, controlling, process parameters predetermination, process condition multiobjective optimization. A prediction and an optimization model is presented allowing the identification of optimal machining conditions using multiobjective programming. This model is based on the definition of an economy function and a productivity function, with suitable constraints relevant to the required machining quality, the required kerfing depth and the available resources. A test case based on experimental results is discussed in order to validate the model.

This publication has 0 references indexed in Scilit: