Checking Roundness Between Centers

A centers-based method can be useful for checking for out-of-roundness without the need for a dedicated roundness machine.


Facebook Share Icon LinkedIn Share Icon Twitter Share Icon Share by EMail icon Print Icon

A dedicated roundness machine is the method of choice for achieving the most proper and complete measurement of roundness. However, for some layout work or when an indication of the roundness of a part is required, the bench centers radial method is a useful plan B. This measuring method supplies the most important component in the measurement of roundness: the variation of the radius length around the entire perimeter of the part.

The process for measuring roundness is supposed to be based on the axis of the part and by mounting it on the same centers used to manufacture the part. This is correct according to general interpretations of the concept known as roundness. But what happens in principle and in reality can be different. Several factors influence the reliability of the roundness measurement; the major one is that the true axis of the part is most likely the one that will be established by mounting the part in centers. Whether it be a single factor or a combination of them, the result is a measurement that generally will be larger than the true out-of-roundness. 

These discrepancies represent the major drawbacks for using this type of system. However, if you understand the elements you are dealing with, then you can set up a basis for part qualification when the tolerances are in your favor.

Roundness measurements between centers are typically carried out with a set of bench centers and some readout device. Most centers will have a fixed reference center and a retractable “live” center. The readout could be a mechanical test, dial indicator or electronic equivalent. After clamping the part between the centers, the probe is brought into contact with the surface of the diameter to be measured. Typically, the test indicator or amplifier is zeroed out at any point on the part, then the part is manually rotated 360 degrees and the dial is watched to measure the total indicator reading (TIR), the span between the maximum and minimum value observed, which is referred to as the radial out-of-roundness.

Some of the sources of errors in this method are pretty obvious, and some may be subtler. Here is what is important when trying to obtain the best results: the center holes on the part should be made to the specification for angle and depth. Otherwise, they are not apt to match up to the more precision angle, likely found in the centers. 

Speaking of the precision centers, they both should be precision tapers, and they must be at the same height and parallel to the surface. Since these components are setting up the part reference, the accuracy is critical to control both the flatness of the surfaces for the centers and the height of the centers. In most cases, these are held to tolerances less than 0.00025 inch. From these tolerances, you can start to see the basic errors that are built into the setup. If your part roundness tolerances are anywhere near the level of accuracy built into the fixturing, then a move to a roundness machine will be required. If your roundness tolerance exceeds the threshold, however, then you may be able to use this method for guidance.

Additional mechanical sources of errors in this setup may arise if 1) the force of the centers is not enough to truly hold the part between the centers or 2) this pressure is not enough to maintain the part in the aligned position.

For very small-diameter parts, this should not be so much that the part distorts. Besides potential height misalignment, one should also ensure the centers are aligned side to side. Often there will be precision slots in the table to help with this, but ensure both centers are alighted to the same side of the rail. Dirt, chips, burrs or rust on the part or centers will affect the results. Angles on the centers and different parts can skew measurements. If the angles of the center or on the part are not concentric or are themselves not round, then errors can occur.

All of these sources of errors will affect the results of this centers-based out-of-roundness checking method. Understanding these conditions and knowing when this is the appropriate method will allow you to make informed decisions about part qualification from the generated results.