毕业设计外文翻译原文

ADDRESSING PROCESS PLANNING AND VERIFICATION ISSUES

WITH MTCONNECT

Athulan Vijayaraghavan, Lucie Huet, and David Dornfeld

Department of Mechanical Engineering

University of California Berkeley, CA 94720-1740

William Sobel Artisanal Software Oakland, CA 94611

Bill Blomquist and Mark Conley Remmele Engineering Inc.

Big Lake, MN

KEYWORDS

Process planning verification, machine tool interoperability, MTConnect.

ABSTRACT

Robust interoperability methods are needed in manufacturing systems to implement computer-aided process planning algorithms and to verify their effectiveness. In this paper we discuss applying MTConnect, an open-source standard for data exchange in manufacturing systems, in addressing two specific issues in process planning and verification. We use data from an MTConnect-compliant machine tool to estimate the cycle time required for machining complex parts in that machine. MTConnect data is also used in verifying the conformance of toolpaths to the required part features by comparing the features created by the actual tool positions to the required part features using CAD tools. We demonstrate the capabilities of MTConnect in easily enabling process planning and verification in an industrial environment.

INTRODUCTION

Automated process planning methods are a critical component in the design and planning of

manufacturing processes for complex parts. This is especially the case with high speed machining, as the complex interactions between the tool and the workpiece necessitates careful selection of the process parameters and the toolpath design. However, to improve the effectiveness of these methods, they need to be integrated tightly with machines and systems in industrial environments. To enable this, we need robust interoperability standards for data exchange between the different entities in manufacturing systems.

In this paper, we discuss using MTConnect – an open source standard for data exchange in manufacturing systems – to address issues in process planning and verification in machining. We discuss two examples of using MTConnect for better process planning: in estimating the cycle time for high speed machining, and in verifying the effectiveness of toolpath planning for machining complex features. As MTConnect standardizes the exchange of manufacturing process data, process planning applications can be developed independent of the specific equipment used (Vijayaraghavan, 2008). This allowed us to develop the process planning applications and implement them in an industrial setting with minimal overhead. The experiments discussed in this paper were developed at UC

Berkeley and implemented at Remmele Engineering Inc.

The next section presents a brief introduction to MTConnect, highlighting its applicability in manufacturing process monitoring. We then discuss two applications of MTConnect – in computing cycle time estimates and in verifying toolpath planning effectiveness.

MTCONNECT

MTConnect is an open software standard for data exchange and communication between manufacturing equipment (MTConnect, 2008a). The MTConnect protocol defines a common language and structure for communication in manufacturing equipment, and enables interoperability by allowing access to manufacturing data using standardized interfaces. MTConnect does not define methods for data transmission or use, and is not intended to replace the functionality of existing products and/or data standards. It enhances the data acquisition capabiltiies of devices and applications, moving towards a plug-and-play environment that can reduce the cost of integration. MTConnect is built upon prevalent standards in the manufacturing and software industry, which maximizes the number of tools available for its implementation and provides a high level of interoperability with other standards and tools in these industries.

MTConnect is an XML-based standard and messages are encoded using XML (eXtensible Markup Language), which has been used extensively as a portable way of specifying data interchange formats (W3C, 2008). A machine-readable XML schema defines the format of MTConnect messages and how the data items within those messages are represented. At the time of publication, the latest version of the MTConnect standard defining the schema is 1.0 (MTConnect, 2008b).

The MTConnect protocol includes the following information about a device: • Identity of a device

• Identity of all the independent components of the device

• Design characteristics of the device

• Data occurring in real or near real-time by the device that can be utilized by other devices or applications. The types of data that can be addressed includes:

• Physical and actual device design data • Measurement or calibration data • Near-real time data from the device

Figure 1 shows an example of a data gathering setup using MTConnect. Data is gathered in near-time from a machine tool and from thermal sensors attached to it. The data stored by the MTConnect protocol for this setup is shown in Table 1. Specialized adaptors are used to parse the data from the machine tool and from the sensor devices into a format that can be understood by the MTConnect agent, which in turn organizes the data into the MTConnect XML schema. Software tools can be developed which operate on the XML data from the agent. Since the XML schema is standardized, the software tools can be blind to the specific configuration of the equipment from where the data is gathered.

FIGURE 1: MTCONNECT SETUP.

TABLE 1:MTCONNECT PROTOCOL INFORMATION FOR MACHINE TOOL IN FIGURE 1.

Device identity “3-Axis Milling Machine” Device

components

1 X Axis; 1 Y Axis; 1 Z Axis; 2 Thermal Sensors

Device design X Axis Travel: 6” characteristics Y Axis Travel: 6”

Z Axis Travel: 12”

Max Spindle RPM: 24000 Data occurring Tool position: (0,0,0); in device Spindle RPM: 1000

Alarm Status: OFF Temp Sensor 1: 90ºF Temp Sensor 2: 120ºF

An added benefit of XML is that it is a hierarchical representation, and this is exploited by designing the hierarchy of the MTConnect schema to resemble that of a conventional machine tool. The schema itself functions as a metaphor for the machine tool and makes the parsing and encoding of messages intuitive. Data items are grouped based on their logical organization, and not on their physical organization. For example, Figure 2 shows the XML schema associated with the setup shown in Figure 1. Although the temperature sensors operate independant of the machine tool (with its own adaptor), the data from the sensors are associated with specific components of the machine tool, and hence the temperature data is a member of the hierarchy of the machine tool. The next section discusses applying MTConnect in estimating cycle time in high-speed machining.

ACCURATE CYCLE TIME ESTIMATES

In high speed machining processes there can be discrepancies between the actual feedrates during cutting and the required (or commanded) feedrates. These discrepancies are dependent on the design of the controller used in the machine tool and the toolpath geometry. While there have been innovative controller designs that minimize the feedrate discrepancy (Sencer, 2008), most machine tools used in conventional industrial facilities have commercial off-the-shelf controllers that demonstrate some discrepancies in the feedrates, especially when machining complex geometries at high speeds. There is a need for simple tools to estimate the discrepancy in these machining conditions.

Apart from influencing the surface quality of the machined parts, feedrate variation can lead to inaccurate estimates of the cycle time during machining. Accurate estimates of the cycle time is a critical requirement in planning for complex machining operations in manufacturing facilities. The cycle time is needed for both scheduling the part in a job shop, as well as for costing the part. Inaccurate cycle time estimates (especially when the feed is overestimated) can lead to uncompetitive estimates for the cost of the part and unrealistic estimates for the cycle time.

Related Work

de Souza and Coelho (2007) presented a comprehensive set of experiments to demonstrate feedrate limitations during the machining of freeform surfaces. They identified the causes of feedrate variation as dynamic limitations of the machine, block processing time

FIGURE 2: MTCONNECT HIERARCHY.

for the CNC, and the feature size in the toolpaths. Significant discrepancies were observed between the actual and commanded feeds when machining with linear interpolation (G01). The authors used a custom monitoring and data logging system to capture the feedrate variation in the CNC controller during machining.

Sencer et al. (2008) presented feed scheduling algorithms to minimize the machining time for 5-axis contour machining of sculptured surfaces. The algorithm optimized the profile of the feedrate for minimum machining time, while observing constrains on the smoothness of the feedrate, acceleration and jerk of the machine tool drives. This follows earlier work in minimizing the machining time in 3-axis milling using similar feed scheduling techniques (Altintas, 2003). While these methods are very effective in improving the cycle time of complex machining operations, they can be difficult to apply in conventional factory environments as they require specialized control systems. The methods we discuss in this paper do not address the optimization of cycle time during machining. Instead, we provide simple tools to estimate the discrepancy in feedrates during machining and use this in estimating the cycle time for arbitrary parts.

Methodology

During G01 linear interpolation the chief determinant of the maximum feedrate achievable is the spacing between adjacent points (G01 step size). We focus on G01 interpolation as this is used extensively when machining simultaneously in 3 or more axes. The cycle time for this machine tool to machine an arbitrary part (using linear interpolation) is estimated based on the maximum feed achievable by the machine tool at a given path spacing. MTConnect is a key enabler in this process as it standardizes both data collection as well as the analysis.

The maximum feedrate achievable is estimated using a standardized test G-code program. This program consists of machining a simple shape with progressively varying G01 path spacings. The program is executed on an MTConnect-compliant machine tool, and the position and feed data from the machine tool is logged in near-real time. The feedrate during cutting at the different spacings is then analyzed, and a machine tool “calibration” curve is developed,

which identifies the maximum feedrate possible at a given path spacing.

FIGURE 3: METHODOLOGY FOR ESTIMATING

CYCLE TIME.

Conventionally, the cycle time for a given toolpath is estimated by summing the time taken for the machine tool to process each block of G-code, which is calculated as the distance travelled in that block divided by the feedrate of the block. For a given arbitrary part G-code to be executed on a machine tool, the cycle time is estimated using the calibration curve as follows. For each G01 block executed in the program, the size of the step is calculated (this is the distance between the points the machine tool is interpolating) and the maximum feedrate possible at this step size is looked up from the calibration curve. If the maximum feedrate is smaller than the commanded feedrate, this line of the G-code is modified to machine at the (lower) actual feedrate, if the maximum feedrate is greater, then the line is left unmodified. This is performed for all G01 lines in the program, and finally, the cycle time of the modified G-code program is estimated the conventional way. This methodology is shown in Figure 3. The next section discusses an example applying this methodology on a machine tool.

Results

We implemented the cycle time estimation method on a 3-axis machine tool with a conventional controller. The calibration curve of this machine tool was computed by machining a simple circular feature at the following linear

spacings: 0.0001”, 0.00025”, 0.0005”, 0.00075”, 0.001”, 0.0025”, 0.005”, 0.0075”, 0.01”. We confirmed that the radius of the circle (that is, the curvature in the toolpath) had no effect on the feedrate achieved by testing with circular features of radius 0.5”, 1.0”, and 2.0”, and observing the same maximum feedrate in all cases. Table 2 shows the maximum achievable feedrate at each path spacing when using a circle of radius 1”. We can see from the table that the maximum feedrate achievable is a linear function of the path spacing. Using a linear fit, the calibration curve for this machine tool can be estimated. Figure 4 plots the calibration curve for this machine tool. The relationship between the feedrate and the path spacing is linear as the block processing time of the machine tool controller is constant at all feedrates. The block processing time determines the maximum federate achievable for a given spacing as it is the time the machine tool takes to interpolate one block of G-code. As the path spacing (or interpolatory distance) linearly increases, the speed at which it can be interpolated also increases linearly. The relationship for the data in Figure 4 is:

MAX FEED (in/min) = 14847 * SPACING (in)

TABLE 2: MAXIMUM ACHIEVABLE FEEDRATE AT VARYING PATH SPACING

Spacing Maximum Feedrate 0.0001” 0.7 0.00025” 3.6 0.0005” 7.2 0.00075” 10.5 0.001” 14.6 0.0025” 35.6 0.005” 71.2 0.0075” 106.7 0.01” 147.68

We also noticed that the maximum feedrate for a given spacing was unaffected by the commanded feedrate, as long as it was lesser than the commanded feedrate. This means that it was adequate to compute the calibration curve by commanding the maximum possible feedrate in the machine tool.

FIGURE 4: CALIBRATION CURVE FOR MACHINE

TOOL.

REFERENCES

Altintas, Y., and Erkormaz, K., 2003, “Feedrate Optimization for Spline Interpolation In High Speed Machine Tools”, CIRP Annals –Manufacturing Technology, 52(1), pp. 297-302.de Souza, A. F., and Coelho, R. T., 2007, “Experimental Investigation Limitations on High Speed

of Feedrate Milling Aimed at Industrial Applications”, Int. J. of Afv. Manuf. Tech, 32(11), pp. 1104–1114.

Elber, G., 1995, “Freeform Surface Region Optimization for 3-Axis and 5-Axis Milling”, Computer-Aided Design, 27(6), pp. 465–470. MTConnectTM, 2008a, www.mtconnect.org

.

MTConnectTM, 2008b, MTConnectTM Standard, v1.0

Sencer, B., Altintas, Y., and Croft, E., 2008, “Feed Optimization for Five-axis CNC Machine Tools with Drive Constraints”, Int. J. of Mach. Tools and Manuf., 48(7), pp. 733–745.

Vijayaraghavan, A., Sobel, W., Fox, A., Warndorf, P., Dornfeld, D. A., 2008, “Improving Machine Tool Interoperability with Standardized Interface Protocols”, Proceedings of ISFA.

Vijayaraghavan, A., Hoover, A., Hartnett, J., and

Dornfeld, D. A., 2009, “Improving Endmilling Surface Finish by Workpiece Rotation and Adaptive Toolpath Spacing”, Int. J. of Mach. Tools and Manuf., 49(1), pp. –98.

World Wide Web Consortium (W3C), 2008, “Extensible Markup Language (XML),” http://www.w3.org/XML/.

Wright, P. K., Dornfeld, D. A., Sundararajan, V., and Misra, D., 2004, “Tool Path Generation for Finish Machining of Freeform Surfaces in the Cybercut Process Planning Pipeline”, Trans. of NAMRI/SME, 32, 159–166.