The scientific method revolves around four key activities: observation,
hypothesis, testing (experiment), and theory. Experiments are never conducted in a random fashion. Experimentation is a very organized, systematic, and planned activity. Planning of experiments may involve following three steps: (a) Parameter design plan involves experiments objective and identification of process variables and parameters that are to be measured. (b) System and tolerance design plan deals with selection of a measurement technique, test equipment, and procedure based on some preconceived tolerance limits for error. (c) Data reduction design plan involves presentation, use, and analysis of data to answer the question (or to prove the hypothesis etc.). Lecture 3 Process Variables For parameter design plan, process variables can be of following types: (1) Independent variable: any variable that can be changed independently of other variables. (2) Dependent variable: a variable that is affected by changes in one or more variables is known as dependent variable. (3) Discrete variable: a variable that describes an entity or a specific item. (4) Continuous variable: a continuous variable identifies a continuous quantity. Engineering quantities such as temperature, pressure, and strain are continuous quantities and hence are represented by continuous variables. (5) Extraneous variable: those variables that are not or can not be controlled during measurements but affect the value of the measured variable are termed extraneous variables. Requirements of a Good Experiment Absence of systematic error Precision Range of validity Simplicity Calculation of uncertainty Concept of Repetition and Replication Repeated measurements made during single test run or on a single batch (of products) are called repetitions. The estimated value of a measured variable improves with the number of measurements. (For example, a bearing manufacturer would obtain a better estimate of the mean diameter and the variation in the diameter of a batch of bearings by measuring many bearings rather than just a few.) An independent duplication of a set of measurements using similar operating conditions is referred to as a replication. Replication allows for quantifying the variation in a measured variable as it occurs between different tests, each having the same nominal values of operating conditions. Concomitant Methods Concomitant methods are alternative methods of carrying out the same experiment. The objective is to obtain two or more estimates for the result, each based on a different method, which can be compared as a check for agreement. Random Test While an experiment is being conducted, the independent variables are varied to record their effects on the dependent (measured) variable. The independent variables are usually varied and applied to the test in a sequential manner. At the same time the test is affected by variations in extraneous variables (a possibility that can not be fully eliminated). A coupling of a relatively slow and uncontrolled variation in the extraneous variables with the sequential and controlled application of the values of independent variables, will impose a false and untypical trend on the value of the dependent (measured) variable. The possibility of this false trend can be minimized by using a random test strategy. A random test is characterized by a measurement matrix that sets a random order in the values of the independent variables applied. Reference/Further Reading R. S. Figliola, D. E. Beasley, Theory and Design for Mechanical Measurements, John Wiley, Singapore, 2004.