A statistical method for predicting insulation life

TO EVALUATE INSULATION SYSTEMS, periods of humidification and operation at elevated temperatures (i.e, temperatures in the range of 175 to 275 C) are necessary if experimental data on the life of the system are to be realized in a reasonable length of time. After insulation life data has been obtained under several different operating conditions, the problem then is to evaluate the system in terms of its suitability for commercial use. A statistical technique whereby this evaluation can be achieved is known as regression analysis, which is concerned with the study of the relationship between two or more variables. In applying this analysis, it may be necessary to edit and transform the data; but in essence, this analysis involves the determination and significance testing of the functional relationship and the determination of the confidence envelope for the estimated relationship.