George Jaffe
Executive Vice President
Steinmeyer GmbH
www.steinmeyer.com

Use centers for all measurements. But before using them, verify that center holes are clean and undamaged. Just check a ground diameter that is close. Runout should be near zero.

Use centers for all measurements. But before using them, verify that center holes are clean and undamaged. Just check a ground diameter that is close. Runout should be near zero.


To check a shaft's straightness, hold the nut and rotate the shaft. Readings from the nut (ground) OD reveal how concentric the ball tracks are. The same method should be used for aligning a screw with a machine axis. Indicating off the nut rather than the OD of the screw eliminates concentricity errors the shaft OS might have.

To check a shaft's straightness, hold the nut and rotate the shaft. Readings from the nut (ground) OD reveal how concentric the ball tracks are. The same method should be used for aligning a screw with a machine axis. Indicating off the nut rather than the OD of the screw eliminates concentricity errors the shaft OS might have.


The bearing journal should be the reference for all measurements. It determines the axis the screw will rotate around. For better repeatability of measurements, use the axis defined by the center holes. But the difference should be minimal. If not, look for reasons.

The bearing journal should be the reference for all measurements. It determines the axis the screw will rotate around. For better repeatability of measurements, use the axis defined by the center holes. But the difference should be minimal. If not, look for reasons.


This measurement reads the shaft's ball-track eccentricity, plus (or minus) the eccentricity of the nut. Repeat it at least three times, each time turning the nut 120° relative to the shaft. The maximum reading is equal to shaft runout plus nut runout.

This measurement reads the shaft's ball-track eccentricity, plus (or minus) the eccentricity of the nut. Repeat it at least three times, each time turning the nut 120° relative to the shaft. The maximum reading is equal to shaft runout plus nut runout.


The drive and bearing journal should be concentric within a few microns or the drive journal may receive substantial side loads. Use two indicators, or mark the high point and note the readings, then compare them to other journal and determine the difference.

The drive and bearing journal should be concentric within a few microns or the drive journal may receive substantial side loads. Use two indicators, or mark the high point and note the readings, then compare them to other journal and determine the difference.


Nut squareness helps avoid cocking. When using V-blocks, a ball can be inserted into one center hole of the screw and the screw pushed against a vertical surface to keep it from sliding.

Nut squareness helps avoid cocking. When using V-blocks, a ball can be inserted into one center hole of the screw and the screw pushed against a vertical surface to keep it from sliding.


It's relatively simple to determine the lead accuracy of a ball screw before you buy it. Just ask the manufacturer and get a protocol. If you just want to know if a screw is straight and true so you don't have problems during installation, you can check it yourself. And torque measurements are always an easy way to make sure new screws meet specifications and to see if old screws need replacing.

But checking a few geometric tolerances on ball screws isn't quite as easy. Here's the problem: It is impossible to take any measurements using the ball track, the most relevant surface for determining ball-screw performance and accuracy, without using a sophisticated 3Dcoordinate-measuring machine. Fortunately, there are simple alternatives.

Most screws have a center hole that can be used as a datum for geometric measurements. However, you may not have a bench long enough to hold the screw between centers. And if you use V-blocks, you might use a surface (the screw's OD, for example) that has no function and may not be concentric with the ball track. Resulting measurements may yield completely irrelevant data.

So, what is a valid datum? Not knowing which surfaces the manufacturer uses as a datum during production leaves only one choice — you must use surfaces relevant to operating the screw. These are screw surfaces that interface with the environment, such as the bearing journals, the nut's pilot diameter (if any), and the flange or other connecting surfaces. So if there are bearing journals at both ends of the screw with the same diameter, you only need a pair of V-blocks and a sufficiently long, flat surface.

It gets a little more complicated when screws with only one journal, or journals with different diameters, prevent screws from laying parallel in V-blocks. In this case, there is no line contact and measurements will be inaccurate.

In that case, the only thing to do is use the OD and do the best you can to eliminate errors it may introduce.

A third option is to use the actual support bearings. Maybe they are already on the screw, complete with bearing blocks. However, be careful: In some designs, removing support bearings damages them, so you may have only one chance. If you do find errors, then you may not know which one to blame — the support bearing or the ball screw.

As a general rule, only components that move have an impact. So we are dealing with relative measurements. The only gauge needed is a dial indicator. Here are tips for measuring nut runout, screw straightness, drive-journal eccentricity, and surface squareness.

Nut runout (concentricity): The nut's pilot diameter (if any) determines the nut's location. If the slide has a bore for the nut body, however, the pilot determines nut location. And if the screw's support bearings are fixed in place, then the pilot also determines alignment and possible side loads. In most machinetool applications, this check is not necessary because the nut is allowed to "float" into its proper position. But this measurement is undoubtedly most important if the nut is the rotating member.

During this measurement, only the nut turns; the shaft is stationary. Thus it doesn't matter how the shaft is supported. It could even be clamped in a vice.

Screw straightness: This measurement requires rotating the shaft. Otherwise we would need a straight-edge reference.Rotating the shaft turns any straightness error into a wobble you can measure with the dial indicator. But make certain that the axis the shaft rotates around is the same as when the screw is installed. Any eccentricity in the surface-supporting screw will show up in the measurement as well.

Drive-journal eccentricity: The drive journal is typically close to the bearing journal, so any eccentricity generates significant forces that either overload the motor's shaft bearings or cause the journal or coupling to fail.

Squareness of bearing shoulder and nut flange: Bad bearing shoulders can actually bend the screw when the bearing's locknut is tightened. And a nut flange with unacceptable squareness error will introduce substantial angular forces that almost certainly reduce screw life.