Improving Energy Efficiency in CNC Machining

Get this Project:

Fields with * are mandatory

INTRODUCTION

Although still electrically-powered and operating on the same basic principles, modern CNC milling machines are significantly different from older milling machines. [Kordonowy 2002] studied several generations of milling machines in order to compare their electrical performance. Their work confirms that not all CNC milling machines are equivalent when electricity consumption is considered. Figure 1.1 is plotted to compare three select results from Kordonowy’s thesis simultaneously.

It can be seen that the oldest Bridgeport TCS, a 3-axis machine, spends most of the consumed energy (65.8%) on material removal processes, while in the case of the newest Haas 5-axis machine, only 24.2% of the total drawn energy is spent on material removal. The Cincinnati machine falls in between the two, spending about a half of the drawn energy on material removal.

Figure 1.1: Energy consumption profiles of three different CNC machines

Figure 1.1: Energy consumption profiles of three different CNC machines

BACKGROUND AND RELATED WORK

Figure 2.1 : Power drawn by a CNC machine as a function of time (symbolic) (a) 3-axis machine, (b) 5-axis machine

Figure 2.1 : Power drawn by a CNC machine as a function of time (symbolic) (a) 3-axis machine, (b) 5-axis machine

Figure 2.1(a) and (b) show simplified graphs of power drawn by typical 3 and 5-axis CNC machines at various times in a typical machining operation. Once a machine is turned on, its lights, computer fans, and the controller begin consuming electricity at their rated power until the machine is switched off (idle power state). Additional energy is required to warm up the machine (jog) by running a predefined set of commands during initialization.

MODELING ENERGY CONSUMPTION IN 3-AXIS MILLING

Figure 3.1: Computation of t block

Figure 3.1: Computation of tblock

Most CNC machines used in the industry and academia have this limiting value for axis acceleration, both in our observation and as reported in the literature. Thus, as shown in Fig. 3.1, we consider a trapezoidal profile for feed rate, to make realistic time computation for each block, computing tblock as the addition of the time required for the axis to accelerate to the rated feed, move at the rated feed for the programmed distance, and decelerate (if needed) to the rated feed of the next block.

Figure 3.3 : Typical power against time curve for a machining operation with a single linear demonstrating both, spindle and axis motor spike

Figure 3.3 : Typical power against time curve for a machining operation with a single linear demonstrating both, spindle and axis motor spike

Figure 3.3, a separate smaller peak corresponding to starting of an axis motor is clearly seen for another machining operation involving a single linear cut. In this version of the energy analyzer, we ignore the spindle related peaks when computing energy consumption of an NC code because that event (spindle starting from rest) happens only a few times in an NC code.

ENERGY MODEL APPLICATIONS

Figure 4.2: Face-milling a rectangular face

Figure 4.2: Face-milling a rectangular face

Figure 4.3 : Generating and verifying toolpath for a facing operation in MasterCAM

Figure 4.3 : Generating and verifying toolpath for a facing operation in MasterCAM

The software internally uses Google Charts API [Google Inc. 2010-2014] to plot time and energy estimate graphs in a browser during output display. Following is a description of a typical run of a software to illustrate various steps in an analysis of an example NC code for milling a face of a rectangular block.

An analysis task begins with a user generated G/M code. For this example run, we analyze an NC code for a toolpath to mill a rectangular face by a single directional toolpath, parallel to X-axis direction (Fig. 4.2). We generated the toolpath in MasterCAM and verified its correctness, as seen in a screen-capture (Fig. 4.3). The complete NC code obtained from MasterCAM is attached as Appendix B.

ENERGY ANALYZER: EXAMPLES

Figure 5.1: Face-milling a square face with cutting primarily in (a) X-axis direction (b) Y-axis direction

Figure 5.1: Face-milling a square face with cutting primarily in (a) X-axis direction (b) Y-axis direction

We deliberately generate NC codes from two toolpaths such that other than the primary direction of cut, all other geometric and mechanical parameters of the two toolpaths being compared are identical. Toolpath A, seen in Fig. 5.1a, has the majority of cutting in X-axis parallel direction while toolpath B, seen in Fig. 5.1b, has the majority of cutting in Y-axis parallel direction. Since this case study is for a square block, both toolpath alternatives have nearly identical machining time for the same cutter diameter and overlap.

Figure 5.5 : (a) Contour-parallel (b) Spiral toolpath for a facing operation for a face with islands

Figure 5.5 : (a) Contour-parallel (b) Spiral toolpath for a facing operation for a face with islands

We again generate the required toolpaths in MasterCAM, one with a contour-parallel toolpath strategy and the other with a spiral (inwards) toolpath strategy with the same percentage overlap. The toolpath based on contour-parallel strategy has significantly more 90-degree turns or corners (Figure 5.5a), while the one based on the spiral strategy has virtually no sharp turns but has many lifts resulting from the islands. Figure 5.5b.

TOWARD ENERGY EFFICIENT TOOLPATHS

Figure 6.1 : Examples of micrographic text arrangements as reported

Figure 6.1 : Examples of micrographic text arrangements as reported

The digital micrography technique focuses on generating aesthetically pleasing text arrangements that convey an underlying 2D shape while still being readable. Micrographic text arrangements, as shown in Figure 6.1, are a special type of calligraphic arrangement of text. These are highly artistic arrangements of text that can take hours to create manually. Maharik et al. generate complex micrographic arrangements automatically, by developing an algorithm to automatically place the text.

Figure 6.4 : 6 alternative strategies from MasterCAM (A) Constant overlap spiral (B) One-way, X-parallel

Figure 6.4 : 6 alternative strategies from MasterCAM (A) Constant overlap spiral (B) One-way, X-parallel

MasterCAM provides several alternative pocketing strategies (shown in Figure 6.4). Although it is difficult to select an equivalent toolpath to compare against that obtained from digital micrography, we decided to fix the depth of the pocket and select a MasterCAM alternative toolpath that removed the planned material—the bear shape, 0.025inch deep—in the least amount of time. Thus, in essence, we compared the most time-efficient MasterCAM toolpath for the same amount of material removed (the most resultant MRR) against the toolpath based on digital micrography.

TOOLPATH PLANNING FOR EFFICIENCY IN 5-AXIS CNC MACHINING

Figure 7.1: Lower dimensional problem setup: (a) design curve and cutter; (b) range of possible cutter orientations

Figure 7.1: Lower dimensional problem setup: (a) design curve and cutter; (b) range of possible cutter orientations

A planar, sufficiently long, ball-nose (semi-circle nose) cutter of radius ρ, that can translate and rotate in the plane. The cutter’s vertical orientation is denoted by orientation θ = 0 and the cutter can be rotated to any angle between −90 to +90 as shown in Fig. 7.1(b).

Figure 7.4 : Results obtained after optimization (a) With equal weights to machining time and MRR (b) With less importance to MRR and more importance to machining time

Figure 7.4 : Results obtained after optimization (a) With equal weights to machining time and MRR

In the second example, Fig. 7.4(b), we set the weight for MRR lower than the weight for orientation change. As seen in the plotted toolpath, the obtained toolpath was significantly different, much “flatter” (indicating fewer orientation changes) and only somewhat influenced by the surface normal curve, but still remaining inside the shaded region.

CONCLUSIONS AND FUTURE WORK

In this dissertation, we address energy efficiency in manufacturing specific to multi-axis CNC milling machines. To improve energy efficiency, we focus on reducing the energy consumed in machining. We identify and illustrate the inherent differences in energy consumption behaviors of 3- and 5-axis CNC machines and thus treat them differently, to analyze, model, and suggest ways to improve energy consumption behavior. The knowledge, techniques, and software tools developed in this work can be readily used in manufacturing industry at various stages in the manufacturing cycle of a part, to save energy (electricity). The major contributions of this dissertations can be summarized as:

1. We demonstrate that energy consumption is an important, independent parameter to consider when planning CNC manufacturing processes—a view that goes beyond simply associating energy consumption to the cycle (operation) time or MRR alone.
2. For 3-axis CNC machines:
(a) We propose an analytical model for energy consumption by 3-axis CNC machines (Chapter 3) and provide evidence based on machining trials that validates the predictions made by the model. The novelty of our energy consumption model is that it predicts energy usage for an NC code without the need of any handbook data or machine specifications, purely based on a one-time customization trial that customizes the model to a new CNC machine.

(b) We also report the development of an energy analyzer software tool to compare electrical performance of NC codes (toolpaths) using the proposed energy model. The energy analyzer (described in Chapter 4) estimates machining time and electricity consumption for a provided NC code, when executed on the specified CNC machine. Using the energy estimates for various alternative toolpaths for a component, process planners can determine the most energy efficient toolpath for an operation for a specific CNC machine. To our knowledge, no existing tool can perform such a customized analysis of toolpaths for a specific CNC machine to predict energy consumption requirements.

Source: University of California
Author: Sushrut S. Pavanaskar

Download Project

Get this Project:

Fields with * are mandatory

Leave a Comment

Your email address will not be published. Required fields are marked *