Reliability analysis of organic fibres using limited data

This paper shows the results of research work regarding analysis procedures for prototype test results, based on heterogeneous prototype characteristics and limited data within the early phase of product development. The research activities were applied within a case study “Organic fibre prototype analysis” from the chemical process industry. An example of an organic fibre is cotton. The research goal was the analysis of reliability parameters of heterogeneous fibre samples as a function of treatment. The treatment types were a) untreated, b) thermal treatment and c) chemical treatment. The reliability parameter used herein is the number of cycles to failure (separation) of the fibre element under load in a linearly reciprocating test facility. The facility parameters are load (grams), stroke (mm) and frequency (Hz). The procedure used was to a) determine the optimum facility settings, b) perform the test on a sample set to determine cycles to failure, and c) perform a statistical analysis of results. The main results are: 1. Facility adjustment concerning maximal differentiability is stroke 4 mm, frequency 1 Hz and weight 10 g. 2. Suggested facility adjustment concerning maximal differentiability is stroke 4 mm, frequency 2 Hz and weight 10g. 3. Thermal treatment has a significant decreasing influence on reliability characteristics of organic fibres. 4. Chemical treatment has a low decreasing influence on reliability characteristics of organic fibres. 5. Chemical treatment can be used for improving durability of organic fibres in contrary to thermal treatment. The used approach shows that the combined application of both nonparametric and parametric statistical methods is successful. The transfer of the approach to similar technical problems is feasible. The future scope of the research work will be the analysis of various treatments using the suggested facility adjustments and the appropriate determination of their influence on reliability characteristics.