Fuels, Lubricants, Engines, and Experimental Design

Abstract

Well-designed experimental programs yield results with broader applicability and better precision at less cost than poorly designed ones. Although every experimental program is “designed” to some extent, knowledge of certain formalized techniques aid the experimenter in developing really good designs. Complete understanding of them is formidable and can be attained only by a specialist. However, the basic ideas are simple and can be grasped by any experimenter. This paper serves as an introduction to:

1. 1.
   Scope of experimentation.
2. 2.
   Factorial plans.
3. 3.
   Lattice plans
4. 4.
   Fractional replication.
5. 5.
   Error hierarchy.
6. 6.
   Analysis of variance.
7. 7.
   7. Precision of inferences.

With understanding of these basic ideas, augmented by some reading on details, any experimenter can design good programs of moderate scope. He also can recognize cases for which he should obtain help of a specialist, and his greater knowledge will enable him to use this help more advantageously.