The Use of Experimental Design and Computerized Data Analysis in Elastomer Development Studies.

Abstract

The rapidly expanding field of computerized data reduction has produced many special purpose programs and routines designed to perform recurring calculations involving complex mathematical techniques such as variance and least squares regression analysis. These basic techniques, when combined with the theory of experimental design, can be employed by the rubber compounder and can result in the derivation of extensive meaningful data from a minimum amount of actual compounding and physical testing. The report describes a computer program designed specifically to: (a) select the proper experimental design for any matrix of up to five compounding variables, each at up to five levels, according to the principle of defining contrasts; (b) analyze physical test data according to the method of least squares regression; (c) determine the relative significance of compounding factors through use of the F Ratio Tests; (d) predict the properties of all untested compounds derived from the design matrix; and (e) scan within the limits of the compounding variables to predict the formulation and physical properties of compounds conforming to a given set of specification or performance requirements. (Author)

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1973
Accession Number
AD0759902

Entities

People

  • Paul E. Gatza
  • Robert C. Mcmillan

Tags

DTIC Thesaurus Topics

  • Computer Programs
  • Computers
  • Data Analysis
  • Data Reduction
  • Data Science
  • Experimental Design
  • Information Science
  • Physical Properties
  • Regression Analysis

Readers

  • Polymer Science and Engineering.
  • Regression Analysis.
  • Systems Analysis and Design