Producing high quality precision parts is highly dependent on the accuracy of machine tools. Even parts which are not themselves machined often depend on machining to produce accurate moulds and jigs. For parts produced by casting or additive manufacturing often require secondary milling, grinding and turning to obtain the required tolerances for interfacing surfaces. In my previous articles I’ve detailed the sources of errors in machine tools including kinematic errors that are built into the machine, thermo-mechanical errors that cause distortions with temperature changes, load and dynamic forces that cause further distortions, spindle errors, tool errors, motion control and software errors. I’ve also covered methods of measuring and correcting these errors, such as rapid verification, multi-axis calibration and simultaneous error parameter estimation.
Modern laser calibration of machine tools provides a relatively fast, accurate and convenient way of correcting for errors in alignment and positioning. Most errors are adjusted using software parameters within the machine tool controller. This means that the measurements are made once, and a single calibration file can then be uploaded to the controller to implement the corrections. This approach is ideal for achieving the highest level of accuracy and for regular calibrations of machine tools. However, it is still necessary to get the machine reasonably well aligned using physical adjustments, before software corrections can be effective. More traditional alignment processes still have a role to play at this stage. This article will cover the basic techniques of machine tool alignment and look at how they differ from the laser calibration approach.
Traditional machine tool alignment methods have their origin in the work of the first machine tool builders such as Henry Maudslay. These became standardized to the point where they were described in Georg Schlesinger’s Test book for machine tools, published in 1927. This in turn formed the basis for the ISO standard 230 which in its current edition also includes many more modern methods and remains in use to this day. Part 1 of the ISO standard describes the basic tests for machine setup and alignment. Methods are divided into preliminary operations (levelling and temperature regulation), machining tests (producing sample parts and measuring them), and geometric tests. This article is primarily interested in the geometric tests. Geometric tests consist of measuring straightness, flatness, parallelism, squareness and rotation.
These methods deal with kinematic errors. These are built into the geometry-defining components of the machine, as a result of manufacturing defects and clearances in linear slide ways and rotary bearings. Kinematic errors do not depend on external factors such as temperature and forces and can, therefore, be corrected for. They are also often located at a particular position on a surface or axis. However, certain kinematic effects such as pay in drives, slideways and rotary bearings can result in hysteresis effects, where the errors are dependent on the direction of travel, such as backlash and lateral play.
Measuring Straightness in Machine Tools
Straightness may be measured in a number of ways – straightness of a line on a planar surface, straightness of components or straightness of motion. The actual method used may involve measuring the distance between the object or motion being measured and a reference straight artifact, or in can involve angle measurement.
There are many ways of measuring straightness using the distance from a reference straightedge artefact. For example, by moving a dial indicator along a straightedge, observing a taut-wire through a microscope, observing a target through an alignment telescope, measuring the position of a laser using a detector array, using a Wollaston prism to measuring laser deflection. Straightness can also be measured in terms of angles using a precision level or autocollimator.
Moving a dial indicator along a straightedge is probably the most established and easily understood form of straightness measurement. Despite the apparent simplicity of this method, there are a number of subtle considerations. Firstly, the straightedge should be supported by two blocks to ensure that it is stable, but these should be located at the points on the straightedge which will minimize deflection due to gravity. These locations may be marked on a high accuracy straightedge. With the straightedge located as a fixed reference, the dial indicator is then moved along the edge being measured. If a linear slideway is being directly measured, then a support with three points of contact should be used. The axis of the dial indicator should be directly in line with one of these contact points and it is this point which is moved along the measurement line. A second straightedge may be required to ensure that the dial indicator is moved along a straight line. However, if it is the straight line motion of a machine tool which is being checked for straightness then the later considerations will not be relevant since the dial indicator will be attached to the machine’s spindle.
The accuracy of this method can be improved by correcting for bends in the straightedge. This is done by making a first set of measurements, then rotating the straightedge 180 degrees about its longitudinal axis and taking a second set of measurements. The difference between the two sets of measurements is the corrected measurement. This method can correct for bends in the straightedge but not for deflection due to gravity or surface defects.
A similar method that can achieve greater accuracy is to use a taut wire in place of the straightedge and a microscope in place of the dial indicator. Typically a steel wire with a diameter of approximately 0.1 mm is used and the microscope is equipped with an etched glass scale to enable the measurement of distance from the wire. It is usually best to use this method over distances where the sag in the wire is negligible as it can be very difficult to accurately predict and correct for the sag.
Measuring Flatness in Machine Tools
Flatness measurement is important in machine tools as the machine bed is an important reference surface, used as a datum for machine setup. Methods of measuring flatness are similar to those used for straightness. The fundamental difference is that while straightness only requires measurements to be made along a single line, flatness requires measurements in a grid pattern.
The simplest method of measuring flatness is to arrange three straight edges. Two straightedges are first positioned parallel to one another, along opposite edges of the flat surface, the third is then placed on top of the first two, perpendicular to them so that it spans across the surface being measured. The upper straightedge can then be moved along the length of the two parallel straightedges to create a series of parallel reference lines. The dial indicator is moved along the upper straightedge, in each of these reference locations, to create a grid of measurements.
What complicates this is the need to establish a consistent reference plane. Three of the corners of the surface being measured as selected as the datums which define the theoretical reference plane. These points may be designated a, b and c. Three gauge blocks of equal thickness are then placed at these points. A straightedge is then placed on the blocks at a and c so that it diagonally spans the centre of the surface, point e. A stack of gauge blocks is then created to fit under the midpoint of the straightedge so that the upper surfaces of the blocks located at a, b, c and e all lie in the same plane, parallel to the theoretical reference plane. The straightedge is then placed along the other diagonal, so that is sits on the blocks at a and e and is suspended over d. A stack of blocks is then made at the correct height so that it presents a surface in the same plane at point d. Having setup supports at each corner, the two parallel straightedges can be positioned and the procedure is then essentially the same as for a straightness measurement but measurements are taken along multiple lines forming a grid.
Measuring Machine Tool Parallelism
The measurement of parallelism between two nominally parallel surfaces or trajectories can be carried out relatively simply using a dial indicator. The dial indicator is attached to a moving part of the machine so that the stylus touches a point on a part of the machine which will move with the axis of interest. As the axis is actuated, the stylus will slide along the surface, measuring deviations due to the straightness of the path on the surface and also due to the parallelism of the surfaces. By best-fitting a line through the measured points the angular deviation can be separated from these results.
Measuring Machine Tool Squareness
In practice, squareness is not measured directly. Instead a reference is made perpendicular to one axis. The measurement then becomes the parallelism between the reference and the second axis.
To measure the squareness between two linear axes, a square is positioned along one axis so that it is nominally parallel to the second axis. The measurement then becomes a measurement of parallelism between the square and the second axis. A dial indicator is attached to one axis, so that it measures the distance perpendicular to the face of the square. The axis is then translated, so that the dial indicator moves along the face of the square. The angle of deviation from squareness can be calculated from the change in the distance, as measured by the dial indicator, and the distance which the axis was translated.
In order to remove the effects of straightness from the calculation of the angle. A line may be best-fit to a number of measured points along the axis. It should be noted that although this may remove the effect of straightness along the actuated axis, the square is mounted to the other axis with its arbitrary position on that surface also resulting in some alignment error due to straightness or flatness.
To measure the squareness of a rotary axis, an arm is offset from the axis and measurements are made at the end of the arm, in the direction parallel to the axis of rotation. As the axis is rotated, the point of measurement describes a circle in a plane perpendicular to the axis of rotation. The deviation from parallelism measured between this circular path and the reference plane, for example the machine bed, gives the squareness of the axis. The actual measurements may be made using a dial indicator.
There may be a tendency for an arm attached to a rotary axis to slip axially as measurements are being taken. The resulting errors can be eliminated by using an apparatus with two arms of equal length so that two measurements can be made simultaneously with an offset of 180 degrees.
While the methods described here can still be useful in the initial build and setup of machine tools. Normal calibration can be carried out much more efficiently using modern laser calibration or even a multiline system for simultaneous error parameter estimation.