Five seal variables, viz., flex section thickness, lip length, trim diameter, material, and lip force, and two shaft variables, shaft out-of-round (OOR) and number of lobes, were investigated by means of fractional-factorial experimentation to determine the effects of shaft OOR and number of lobes on seal performance and to determine the interactions between the five seal design factors and the two shaft variables. A 1/4 replicate, 27 fractional-factorial experiment was conducted with leakage as the dependent variable to determine these effects. For the levels of OOR, lobing and shaft speed chosen, the tests indicated that shaft OOR and number of lobes in themselves had no effect. However, they appeared in significant interactions with several seal design factors. Also, at higher speeds, both OOR and lobing are extremely important. Analysis of the interactions and further test work with lobed shafts indicate that due to the interactions between several seal design factors and shaft OOR, the setting of an acceptable shaft OOR tolerance is extremely difficult. However, a general recommendation of 200 microin. maximum shaft OOR, with a minimum number of lobes, is supported by the data and is still valid.

This content is only available via PDF.
You do not currently have access to this content.