Abstract: There are lots of incomplete data and zero failure data in the real engineering world. A theory of incomplete data rank distributions and zero failure data rank distributions is established. Thus the percentile ranks and mean ranks of the failure data and suspension data in them are given. On the basis of the theory, the incomplete data and zero failure data can satisfactorily be treated. The optimum design and analysis of engineering experiments may be provided in coordination with the constraints of the experimental time or cost. It is simple and convenient for application in engineering.

Key words: rank distribution, incomplete data, zero tailure data, percentile rank, mean rank