Feedrate Optimization for Computer Numerically Controlled Machine Tools Using Modeled and Measured Process Constraints

Feedrate optimization for computer numerically controlled (CNC) machine tools is a challenging task that is growing in importance as manufacturing industry demands faster machine tools. The majority of research in this area focusses on optimizing feedrate using modeled process constraints. Some researchers have suggested using measured process parameters instead. The former approach suffers from uncertainties in the modeled pro- cess data that is the starting point of the optimization. The latter approach has difficulties achieving high levels of optimality. This study proposes the combination of both modeled and measured process data. To this end, a control architecture is described that allows combining measured and modeled process constraints. Within this architecture, a new algorithm to determine time optimum feedrates using modeled velocity and acceleration constraints is proposed. The new control structure including the novel feedrate optimiza- tion algorithm is verified experimentally on a high speed biaxial table.

Keywords: feedrate optimization, CNC machine tools, modeled process constraints, measured process constraints

1 Introduction

In modern manufacturing systems, the process planning system provides G-code motion commands to CNC machine tools. These commands are used by interpolators to create reference trajecto- ries for lines, circles, and free form surfaces. The trajectories are comprised of a tool path and a feedrate that indicates the speed that the path is traversed at. To reduce acceleration levels at the start and end of the motion segments, acceleration/deceleration algorithms are commonly incorporated into the interpolators for lines and circles in commercial machine tools [1].

The machining literature shows two main approaches to further improve machining trajectories. The first approach uses model knowledge of the machine tool and the machining process in order to constrain velocities, accelerations, and jerks during a prepro- cessing operation [2–17]. The second approach uses measured process data in order to limit axis accelerations and velocities by adjusting the overall feedrate in real-time [18–24].

The purpose of this study is to establish a feedrate optimization algorithm that can integrate both modeled and measured process data. This approach provides highly optimal motion trajectories while allowing for significant model uncertainties.

The remainder of the paper is organized as follows: Sec. 2 pro- vides background and a literature review for this work. Section 3 describes the proposed architecture that allows combining mod- eled and measured process constraints for feedrate optimization. Section 4 proposes a new algorithm to determine time optimal feedrates for modeled velocity and acceleration constraints. Sections 5 and 6 provide experimental results and conclusions.

Manuscript received March 4, 2016; final manuscript received June 10, 2016; published online August 15, 2016. Assoc. Editor: Radu Pavel.

2 Background

Conventional CNC control architectures receive G-code motion commands from the process planning system. An interpolator transfers the motion commands into motion increments, DX and DY that are used as the reference positions of the axis position controllers. This control architecture can run into difficulties when the process planning system provides motion commands that are beyond the capabilities of the machine tool. Typically, both the feedrate and the acceleration of a machine tool are limited due to voltage and current constraints within the axis drives [25]. In addi- tion, the cutting process can further limit the maximum allowable feedrate due to constraints such as tool edge breakage, shank breakage, spindle torque, spindle power, tool deflection, surface finish, or chatter [26–33].