Printer Friendly

Tracking machine-tool error.

Tracking machine-tool error

As machining accuracy becomes increasingly important, checks of machine-tool accuracy need to be both more frequent and more thorough. It is no longer sufficient to rely solely on the machine's accuracy tests performed when it was delivered. Machining a test part has the drawbacks that errors are many and additive--the machine geometry, the control, and any environmental factors--and that different parts will yield different results, as will subsequent tests of the same type part. Machining a test piece is also complicated by the need to choose a truly representative part size and shape and measure it accurately.

Errors in the controls are clearly different from errors in machine geometry and need to be differentiated. A new circular test system can do this with a relatively simple test. The Cary Bidim K, developed by Meseltron SA, Switzerland, and offered here by Movomatic USA, Greenville, RI, can detect and evaluate the most important dynamic errors in NC machine tools.

The circular test can distinguish between geometry error and control error. In four hours, the 21 geometrical machine errors over three axes can be checked and the size of each error measured, including testing the NC, repeatability, and the effects of speed.

Every machine tool with a minimum of two CNC linear axes can be measured--i.e., any linear-plane movement with interpolation can be related to another axis. Besides the obvious milling machines and machining centers, this also includes grinders, where table movement is measured relative to grinding-wheel movement. The test permits you to measure the movement of the machine relative to the cutter path, taking into account circular interpolation, slackness in the drive, twists or warp in the table, as well as servo performance.

Running the test

The test system consists of a two-dimensional probe, an X-Y recorder, a set of accessories, and one or more certified ring-gage standards or "master pieces." The probe is mounted in the machine spindle, with spindle rotation blocked. The machine controls are programmed to "machine" a circle in the plane of the master piece, coincident with one of its radii. Deviations between the "machined" circle and the gage are enlarged 500 or 1000 times and plotted by the X-Y recorder. This exaggerated plot can then be interpreted to identify geometry or control errors associated with those two axes. The process is then repeated with the gage in other planes.

The geometrical errors this will reveal for each axis include: positioning, straightness, perpendicularity, pitch movement, yaw movement, and roll movement. Because the spindle is fixed, its own error does not affect the measuring process.

Functional checks can be carried out with dynamic and repetitive measurements. These yield: accuracy of NC interpolation at different contouring speeds, influence of backlash, differences between different NC control units used on the same machine, effect of machine slideway lubrication, and effect of load.

Using a small test circle maximizes the effect of NC-control error while minimizing the effects of machine geometry. Using a large test circle, ideally two-thirds of machine travel, maximizes geometry error and minimizes control error. Ring-gage outer OD is generally 7" and used for machine travels of 8" to 10". For larger travels, you can simply test the table in several areas down its length. Obviously, much larger ring gages would be very expensive to produce and certify, and gage accuracy would vary with size.

Nine different mounting positions for the gage are required to do a complete check of a three-axis machine. Three with the small circle for checking the NC controls (one for each principal plane), and six (two per plane) with the large circle to check machine geometry. If the small and large circle are combined in the same master piece, only six positionings are required. Several tests are made with the small circle at different speeds and directions.

No electrical connections are needed between the machine NC or scales and the test device. The ring gage can be mounted and centered in 15 min, and each test sequence is only as long as it takes the machine to trace a circle. No complicated calculations are required to evaluate results--multiplication and division on a pocket calculator are sufficient.

Spindle measurement

In addition to the above tests, measuring spindle runout can be done by simply rotating the spindle with the probe inside a known, pre-centered bore to plot deviation, side to side. To measure spindle-perpendicularity error, you can reorient the probe on its side and use it as a trammeling device to measure the squareness of the spindle centerline to the flat table. You use a radius bar in the spindle, parallel to the surface of the table, with the probe mounted on its tip. Turning the spindle slowly, you plot points around the table to get an evaluation of full 360-deg spindle perpendicularity to the flat table.

System accuracy

System accuracy for the Bidim K is 1 micron, although much smaller variations than that can be detected. This includes the accumulative error of the ring gage, the probe, and the recorder. Accuracy of the master-piece ring gage itself is 0.000 008" or 0.2 micron in roundness, and 0.000 004" or 0.1 micron in diameter. Of the two, roundness is more important in determining the machine's ability to do circular interpolation.

Machine repeatability

Repeatability checks are simply a matter of running the same test twice without changing anything and superimposing the two measurements. Says Cary's Wolfgang Knapp, "It is astonishing that most of the machines we have checked have shown repeatabilities within 0.001 mm (0.000 040"). Only a small X-Y carriage with defective roller bearings had a repeatability significantly worse.

"This excellent repeatability of machine movements shows that the errors of machine tools are above all systematic. In other words, machine accuracy could be significantly improved by software compensation."

Cost of the system --probe head, master piece, and recorder--is about $28,000.

For more information from Movomatic, circle 340.

PHOTO : The Cary Bidim K traces the ID of the smallest radius on a ring-gage master piece to reveal a [+ or -] 20-micron plot of an NC machine tool's circular-interpolation error.

PHOTO : Contouring speed comparison: comparison of the ability of two machine tools to contour a 40-mm dia at 500 mm/min. The left machine deviated from circular form by only 0.008 mm, while the right machine's deviation was 0.04 mm. (The left trace is actually two traces, demonstrating a repeatability of under 0.001 mm.)

PHOTO : Effect of interpolation accuracy: the thicker the line (radial peaks), the greater the deviation between a true circle and each straight line of interpolation by the machine's control. For the same machine tool and the same contouring speed of 60 mm/min, the left trace was with 0.001 mm interpolation, and the right trace was at 0.01 mm. (Magnification = 1000.)
COPYRIGHT 1992 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1992 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Sprow, Eugene
Publication:Tooling & Production
Date:Jan 1, 1992
Previous Article:Shear spinning for the pressroom.
Next Article:Solving solenoid burnout.

Related Articles Adds Major Site Enhancement With the Arrival of a ''New Community Forum''. Now Features a Comprehensive Onsite Auction Area. Announces Major New Online Offerings in Conjunction with IMTS.
People News. launches financing center.
The proliferation of the dot com.
Single control source cuts it for saw maker.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters