PRECISION OF LABORATORY MEASUREMENTS OF BREAKING STRENGTH OF TEXTILES
Warp breaking strength was measured in five test strips from each of a series of duplicate pieces of cotton duck in rotproofing experiments. Breaks occurring at the jaws of the machine, which amounted to some 16% of the total, gave results about 3.5% lower and 20% more variable than non-jaw breaks. Discrepancies in non-jaw breaks of similarly treated fabric arose partly from variance in test strips from the same piece and partly from additional variance between duplicate pieces. Weathering tended to reduce intra-piece, and soil burial to increase both intra- and inter-piece variability. The tests as conducted were capable of discriminating differences of the order of 15 to 20%. Greater gains in precision would have resulted from increasing the number of replicate pieces than from increasing the number of strips tested per piece in the same ratio. However, determination of the most economical test procedure for specified precision in each type of material also required consideration of the cost factor, which was nearly three times as great per piece as per strip.