The Average Key Length Standard

Posted on: December 4th, 2012 by admin No Comments

The Average Key Length Standard

Manufacturers, end-users (and even many motor shops) seem to be unfamiliar with the Average Key Length standard as it relates to imbalance. There are a number of factors that can minimize the impact of not understanding this principal, but everyone who is responsible for the extended operation of rotating equipment should be familiar with it. Let’s take a quick look at a real-world example.

A couple of years ago a customer called up to have us check a large motor on a belt-driven air compressor. When I arrived at the job-site I found two air compressors sitting alongside each other. Each had a 150HPLincolnmotor belted to the large shaft of a high speed compressor.

One of the motors had experienced a winding failure and needed to be pulled. As I was about to find out, it was only the symptom of the failure.

As we discussed what would be needed to remove this 1500 pound motor from its base five feet above the floor, the other compressor suddenly turned on. The whole building seemed to shake as the compressor came to pressure then, just as suddenly, shut off. I asked the customer how long the compressor had been shaking that violently.

“Ever since it was installed. This one’s just as bad” he said as he pointed toward the unit we had been discussing. “We have to come in every so often and re-weld the supports ‘cause it breaks ‘em” he added.

While in the shop we set the rotor up in the balance stand to check the balance condition. The problem was soon apparent.

The motor shaft was 3.375” in diameter. The keyway was 7/8” wide x 8.375” long. The pulley that had been mounted on the shaft by the compressor manufacturer had a bore that was approximately 2.5” long.

The problem related to the “Average Key Length Standard”. More on that below, but first –

Let’s take a look at the manufacturing process to understand where this situation takes root.

When an electric motor is produced the piece of shaft material starts out very nearly perfectly balanced. As the material is machined, the various steps and shoulders appear as the result of material removed from the shaft. As these changes are virtually concentric (nothing is perfect) very little change is made to the balance condition of the shaft. However, two of the final steps introduce large changes to the concentricity of the rotor assembly.

When the rotor is pressed onto the shaft it brings with it a greater degree of eccentricity. The cast aluminum rotors used today throughout the industry can never be absolutely concentric. Air pockets and non-concentric shaping cause imbalance. The addition of a large non-concentric mass at a greater diameter than the shaft has a large influence on the overall balance of the rotating mass.

The final change is the cutting of a keyway on one side of the shaft. Material is removed to produce a large trough to accommodate the shaft key. As the material is removed, weight is removed and imbalance is introduced to the assembly.

To meet industry standards the final assembly is then dynamically balanced. To do so the machinist mounts the assembly in a balance stand and spins it. But the problem they face is as follows:

The motor manufacturer does not know what will be mounted on the shaft when the motor is placed into service. From a balance point of view the material that was removed when the keyway was machined must be compensated for.

By convention, one half (approximately) of the height of the key is considered to be material in the shaft, the other half is considered to be part of the attachment (the coupling or pulley that mounts on the shaft). When the rotor is balanced a weight equal to roughly 50% of the weight of the key that will be shipped with the motor is placed in the keyway. The assembly is spun and any necessary final adjustment to the balance is made by adding or subtracting weight at the proper position.

Simple enough – right?

But what about the “attachment” manufacturer? When the keyway is cut in the bore of his assembly it must be compensated for in the balance procedure. But they do not know what shaft their piece will be mounted on. The pulley and coupling makers must also supply a key, half of which is considered the material of their assembly, the other half is the compensation for the keyway to which it mounts.

Do you see the problem yet?

In the case of the air compressor motor mentioned above the pulley manufacturer supplied a key that matched the length of the bore of the pulley (approximately 2.5”). Their key weighed approximately 250 grams, half of which was to keep the balance of the pulley within specifications.

But the poor motor manufacturer had balanced the rotor assuming that their key would be used (7/8”x7/8”x8.375”). Their key weighed a whopping 837.5 grams. So when the pulley was mounted on the shaft an imbalance of 293.75 grams was immediately introduced into the rotating system at a distance of approximately 1.68” from the center of the shaft. The influence on balance increases with both the weight of the key and the distance from the center of the shaft. The imbalance is magnified by speed.

So how do you work around this situation?

 The Average Key Length Standard


Measure the full length of both keyways, add the two dimensions together, and divide by two. In the discussion above we have 8.375” + 2.5”/2 = 5.43”

Cut a piece of key stock as close as possible to the calculated length and you will have corrected the majority of your imbalance in many cases. It is important to know that the motor manufacturer supplies a full length key with every motor. It is the customer’s responsibility to use the correct portion of that key, whether the customer is an O.E.M. or an end user.

In this specific case, the imbalance dropped form 6.0 mils to 1.0 mils displacement. Even if you have never had to balance a rotating piece of equipment, you can probably deduce that this is quite an improvement.


 After returning to the job site with the freshly rewound and rebuilt motor I examined the pulley/shaft and key size on the driven shaft. The same condition existed there with one additional factor thrown in. The pulley ratios were such that the compressor shaft was rotating even faster and the keyway in the shaft was even longer, magnifying the effect on the driven end.

Instead of the manufacturer’s representative having to re-balance the compressor and motor, a change of keys was all that was needed to stop breaking steel supports and their welds. The customer was happy and is now fully aware that when replacing a motor or the shaft attachment, care should be taken to cut the new key to the appropriate length, instead of simply using the full length of the supplied key.

Leave a Reply