Error Motion and Microinch Machining
The error motion of a spindle is measured by installing a master part and adjusting it for minimum indicator reading (runout). The runout observed will be the composite of the error motion of the spindle, the error of the master, and the residual eccentricity. A full description of spindle error motion requires independent evaluation of axial, radial, and tilt error.
Axial Error Motion is measured by placing the indicator on the axis of rotation of the spindle and measuring the runout at the top of the test ball. Typical BLOCK-HEAD® spindles will have less than .5 microinch axial error motion.
Radial Error Motion of BLOCK-HEAD® spindles is measured with a 2″ diameter test ball placed about 2” above the top of one of the thrust plates. This puts it 4” above the center line of the spindle body. At this position, the typical minimum total indicator reading is about .5 microinch. Occasionally, it is possible to observe less than .2 microinch total indicator reading. Independent measurements by other laboratories have also shown less than .2 microinch error motion in similar tests. At this level, it is very difficult to separate errors of the spindle from errors of the master ball. This measurement is called Radial Error Motion. However, since the indicator is 4″ above the center line of the spindle, the measured error is actually a composite of radial error motion plus whatever tilt error occurs in 4″.
Tilt Error Motion is less than .1 microradian. This equals one microinch for every 10″ from the spindle center. One way to distinguish radial from tilt error motion is to put the test ball on a long riser. For example, a 42″ riser has been used on a model 10B BLOCK-HEAD®. Since the total radial plus tilt error motion was less than four micro-inches, it follows that the tilt error motion is less than one microinch per 10″.
An alternative way to measure tilt error motion would be to measure the flatness of a circular element on the face of a large diameter flat. This would give a composite of axial plus tilt error motion. This is an important attribute of spindle behavior, but difficult to measure as the master ring or flat would have to be accurate to a microinch or so at a diameter of several feet. The largest diameter circular element we have measured is a 54″ diameter steel plate that was ground on a 10″ BLOCK-HEAD® spindle and found to have less than 10 microinches total indicator reading when measured 180° from the line of contact of the grinding wheel.
By measuring on the opposite side from the grinding position, the tilt error can be expected to show double on the indicator. In this case, a substantial portion of the total indicator reading was attributed to the surface finish rather than geometry. It seems reasonable to assume that the maximum tilt error motion of the spindle in this particular case was something less than 10 microinches in 108″.
This demonstration illustrates a spindle attribute which is more pertinent than error motion alone. For a spindle to be useful in machining, it must have adequate stiffness in addition to accuracy. If a part is to be machined to one microinch, the spindle must not only rotate accurate to one microinch, but must also be stiff enough so that when a one microinch high spot comes under the cutting tool, it will be cut off. Obviously, if the spindle is soft, the tool will simply push the work away rather than cutting off the high spot.
The cutting test, in other words, is a dynamic check of accuracy and stiffness of the complete structural loop of the machine as measured from the tip of the cutting edge to the workpiece surface by way of all the support members, the sliding ways, and the spindle, workpiece and toolholders.
Probably the most searching test that is normally made of a machining system is to use a single-point diamond tool to cut an interrupted surface. Machining with a single-point tool faithfully records all relative motions between the tool and the workpiece. For example, when the tool first contacts the work, it inevitably moves in response to the force which is applied. The result will be a turned-down edge.
The depth of the turn-down on the edge is a result of the stiffness of the system and the amount of cutting force applied. The width of the affected zone is a measure of the cutting speed and the response time of the machine.
If the natural frequency of the structural loop is high and the damping factor is strong, the tool will quickly settle down and cut a straight line. If the system is not well-damped, the workpiece will show chatter marks. The spacing of the chatter marks is a measure of the natural frequency. In a highly-damped systems such as found on certain machines employing BLOCK-HEAD® spindles and very compact structures, the rate of damping is so fast that the tool essentially makes one accommodating motion and then stays put throughout the length of the cut.
In summary, microinch machining requires not only microinch accuracy of the spindle, but also stiffness and damping; not just of the spindle itself, but of every member of the entire structural loop. It should be noted that grinding, being an averaging process, is more forgiving of certain kinds of structural motion. For instance, holes can be routinely ground round within four microinches (1/10 micron) on I.D. grinders in every ordinary condition, provided that the workpiece and the wheel are supported on good air bearings.