Energy efficiency of milling machining: Component modeling and online optimization of cutting parameters. (10th September 2017)

• Record Type: Journal Article
• Authors: Shin, Seung-Jun; Woo, Jungyub; Rachuri, Sudarsan
• Is Part Of: Journal of cleaner production. Volume 161(2017)
• Page Start: 12
• Page End: 29
• Publication Date: 2017-09-10
• Subjects: Metal cutting -- Predictive modeling -- Optimization -- Energy efficiency -- STEP-NC -- MTConnect; Factory and trade waste -- Management -- Periodicals; Manufactures -- Environmental aspects -- Periodicals; Déchets industriels -- Gestion -- Périodiques; Usines -- Aspect de l'environnement -- Périodiques; 628.5
• DOI: 10.1016/j.jclepro.2017.05.013
• Languages: English
• ISSNs: 0959-6526

Abstract

Abstract: Energy consumption is a major sustainability focus in the metal cutting industry. As a result, process planning is increasingly concerned with reducing energy consumption in machine tools. The relevant literature has been categorized into two research areas. The first includes energy prediction models, which characterize the relationships between cutting parameters – the main outputs of process planning - and energy consumption. The second involves energy-consumption optimization, which uses the prediction models to find the cutting parameters that minimize energy use. However, previous energy prediction models are limited to predict energy for tool paths coded in a Numerical Control (NC) program. Previous energy optimization methods typically do not use online optimization, which enables fast optimization decision-making for supporting on-demand process planning and real-time machine control. This paper presents a component-based energy-modeling methodology to implement the online optimization needed for real-time control. Models that can predict energy up to the tool path-level at specific machining configurations are called component-models in this paper. These component-models are created using historical data that includes process plans, NC programs, and machine-monitoring data. The online optimization is implemented using a dynamic composition of component-models together with a divide-and-conquer technique. The feasibility and effectiveness of our methodology has been demonstrated in a milling-machine example. Highlights: A component-based methodology is presented for energy prediction and optimization modeling. The methodology enables energy prediction for a numerical control program. The methodology implements online optimization for supporting real-time control. The methodology uses historical machining data for energy component models.