Page 15: of Maritime Reporter Magazine (August 15, 1973)

Shafting Eccentricity— (Continued from page 14) rest is generally used on short shafts (shafts up to about 30 feet long), whereas two steady rests are required on longer shafts such as stern tube and propeller shafts.

The machining process is begun by locating the rough work in lathe cen- ters, then with the shaft allowed to sag freely, a cut is made at the steady rest positions desired. When a smooth surface is obtained, the steady rest is put in place and shifted to posi- tion the shaft in a straight line. Then the remaining length of the shaft can be cut without tool chatter.

After each cut, there is a possibility that the material removed from the shaft may have contained residual stresses such that their removal would have tended to cause the shaft to bow; consequently, the steady rests must be removed so that the steady rest positions can be checked and corrected as required. If these steps are not carefully followed, it is pos- sible to machine a shaft that is per- fectly true with a steady rest in place, but in fact contain a significant bow when the steady rest is removed. This consideration has been recognized by

NavSec personnel, and it is reported that the use of steady rests when taking eccentricity measurements will not be permitted by the next revision of the NavShips Standard.

It may be noted that tolerance 6 allows a shaft outside diameter ec- centricity of 0.0025 inches in any 48 inches with an eccentricity upper limit of 0.015 inches. This means that an 8-foot shaft would be permitted to have an eccentricity of 0.0025 inches at the shaft midpoint (i.e., a 0.0025- inch change in eccentricity going from one lathe center to the shaft midpoint which is counteracted by a 0.0025- inch change in eccentricity going to the other lathe center). Similarly, a 16-foot shaft could have an eccen- tricity of 0.005 inches, and the ec- centricity of a 24-foot shaft could be 0.0075 inches. But, irrespective of the shaft length, the 0.0075-inch eccentricity cannot be exceeded, and this is when the trouble begins.

Shaft length is an important con- sideration insofar as the measured eccentricity of shaft diameters is con- cerned. Figure 2 shows that the allowable eccentricity of an 8-foot shaft is 0.0025 inches; however, such a tolerance for an 8-foot shaft would be so loose as to be meaningless (un- less the shaft had some assistance in becoming eccentric). On the other hand, a 0.0075-inch permissible ec- centricity is reasonable for a 32-foot shaft, and it is conjectured that this is approximately the length upon which the tolerance was established.

However, by linearly reducing the allowable eccentricity for shorter lengths, the tolerance becomes exces- sively relaxed (to be consistent) ; and for longer shaft lengths, the tolerance becomes too severe by a considerable margin when it is held constant.

If, as has been done in the past, the position of bearings in the shaft- ing arrangement are to be ignored and the length of shafting sections is the only criterion considered in establishing the allowable eccentricity of the shaft outside diameter, then another approach may be considered, i.e., a shaft with a continuous bend.

It can be shown that for any radius of curvature that is established to be acceptable, the maximum shaft eccen- tricity allowable should vary with the shaft length squared. This gen- eral concept is easily visualized. Con- sider, for example, that a 48-foot shaft was planned but difficulty was expected in meeting the 0.0075-inch eccentricity requirement. The prob- lem could easily be avoided by simply making two shafts, each of a 24-foot length. The eccentricity permitted would still be 0.0075 inches for each of the two shorter shafts, but theory and experience indicate that the 0.0075 inches for each of the two shorter shafts will be four times eas- ier to meet with the shorter shafts than with shafts twice as long. The shorter shafts would be easier to manufacture, but shaft reliability would be degraded (due to the addi- tional flanged joint introduced); con- sequently, this alternative is undesired.

Several observations can be made relative to the eccentricity of shaft bodies (not bearing surfaces) ; but the one which is most relevant is that when a balance tolerance is specified and journal and flange eccentricities are specified, the shaft body maxi- mum eccentricity tolerance appears to serve no practical purpose. Shaft eccentricity, as such, does not denote unbalance. Therefore, aside from the possibility of the off-center mass of the shafting eccentricity (etc.), no inherently objectionable aspects of shafting eccentricity can be identi- fied. However, studies show that it is inconceivable that a shafting sec- tion would be eccentric to such a degree that the eccentric mass of the shaft would result in an additional deflection of a significant magnitude.

While there is no practical justifi- cation for specifying a maximum per- missible shafting runout (e.g., 0.015 inches), there is some degree of justi- fication for maintaining the speci- fied maximum permissible change in eccentricity of 0.0025 inches in any 48 inches of shaft length. This latter criterion would tend to identify ab- normal conditions, and consequently, may be of value.

Taper Eccentricity

In order to establish the shafting taper eccentricity requirements which should be specified, it is necessary to ascertain the degree of machining accuracy required to ensure satisfac- tory service performance with proper regard to cost considerations.

In the past, the practice has been to specify the eccentricity of the pro- peller shaft taper with respect to the position in which the shaft is sup- ported in the lathe. However, this eccentricity measurement has no prac- tical significance. On the other hand, the eccentricity of the propeller shaft taper centerline with respect to its running position when installed in the ship is an important dimension.

The shaft taper eccentricity toler- ance of 0.0025 inches, as specified in "Shafting, Propulsion, and Compo- nents," BuShips Standard Drawing 5000-S4300-F-1385661, March 1955, would not inordinately increase the 14 12 10 to 2

E o