The shaft deflection criteria is an outgrowth of experience before modern computational methods were applied. One major criteria is the locking device:

A) It must not allow walking on the shaft.

B) Its structural integrity must not be compromised.

C) It must not create excessive stress risers on the connecting interface with the shaft.

D) Its many varieties require differing deflection design criteria

E) Deflection criteria must include compatibility between shaft diameter, end disk stiffness, and locking device bending moment capacity

The type of locking device then becomes part of the calculus. Locking devices differ in their ability to carry the shaft rotation, under deflection. The tapered compression locking hub had variants that included limiting the deflection to 0.0023 (MPTA) and the 0.0015 in./in. or mm/mm for engineered class designs at shaft diameters > 200 mm. Then later Ringfeder type locking devices were applied and new criteria developed based, in large part, on failures of shaft or compression locking device. Again there are many types and criteria. The designer and manufacturer have been caught in a cycle of trial and error until modern analysis methods have given clear understanding.

In addition, the end disk stiffness must be considered. Excessively stiff end disk and flexible shaft lead to premature locking device or shaft failure from the stress concentration at the shoulder of the locking device at its initial contact with the shaft. Many early Ringfeder narrow devices led to such failures.

Today, with modern analytic tools, new criteria have developed that protect against these failures. More flexible end disk and stiffer shafts ( lower deflection slope) protect against the new generation of failures as belt tensile loads increase and pulleys are designed with wider and smaller diameters.

Still these are evolving criteria. Many such design criteria are applied with ignorance to guarantee against failure. This is the price a client pays in selecting firms that practices overdesign. ■

I will only supplement Larry's comments as I am in total agreement. Limiting shaft deflection (rotation) at the pulley hubs, which is calculated as free shaft deflection (that is not what actually happens), is an artificial way to limit the moment at the hub and end disc. I remember a good customer telling me that he was having end disc failures. I told him his shafts were too small. He replied "I haven't failed any shafts". Though we have always known this to be an indirect way to accomplish an end it was the ringfeder walking occurrence that brought the problem to the forefront. Following the rash of such occurrences guidelines were issued that essentially required the hub juncture to transmit the full moment through the hub into the end disc, treating the shaft as two cantilevered stub shafts. For a time many followed those guidelines. I never accepted them and immediately developed a moment distribution approach the calculated the actual moment distribution; 1.) through the hub to the end disc and to the pulley shell and 2.) to the interior of the shaft. With such an approach the free shaft deflection criteria can be discarded altogether.

Joe Dos Santos ■