The subject of the presented work was the analysis of the influence of the distance between the electrodes using in the coating process on the tribological properties of oxide coatings. Oxide coatings were prepared on EN AW-5251 aluminum alloy samples. The samples surfaces were subjected to hard anodizing process in a multicomponent electrolyte based on sulfuric acid with an addition of organic acids. Anodizing was carried out with a constant electric charge density of 180 A·min/dm2. The distances between the electrodes for subsequent samples increased every 0.125 m up to 1 m. The tribological partner in a sliding couple with oxide layers was pin of PEEK/BG. Tribological tests were conducted on a T-17 tester in reciprocating motion, in technically dry friction conditions. Before and after tribological test, examination of the geometrical structure of counter-specimens’ surface was carried out using the Form Talysurf contact profilographometer, via a 3D method. The most satisfactory tribological parameters were obtained for the PEEK/BG association with the coating produced at a distance between the electrodes equal to 0.25 m.
The design of experiment (DoE) is a methodology originated from early 1920s when Fisher’s papers created the analysis of variance and first known experimental designs: latin squares. It is focused on a construction of empirical models based on measurements obtained from specifically structured and driven experiments. Its development resulted in the constitution of four distinctive branches recognized by the industry: factorials (full or fractional), Taguchi’s robust design, Shainin’s Red-X®and a response surface methodology (RSM). On one hand, the well-known success stories of this methodology implementations promise great benefits, while on other hand, the mathematical complexity of mathematical and statistical assumptions very often lead to improper use and wrong inferences. The possible solution to avoid such mistakes is the expert system supporting the design of experiments and subsequently the analysis of obtained data. The authors propose the outline of such system and provides the general analysis of the ontology and related inference rules.
Jacek Pietraszek, Renata Dwornicka and Andrii Goroshko
Shainin's component search procedure uses variability source detection based on specific median test. This approach has only two triple subsets and the certainty of inference can be weak for this reason. This paper checks this approach by series of numerical simulations.
Renata Dwornicka, Andrii Goroshko and Jacek Pietraszek
The bootstrap method is a well-known method to gather a full probability distribution from the dataset of a small sample. The simple bootstrap i.e. resampling from the raw dataset often leads to a significant irregularities in a shape of resulting empirical distribution due to the discontinuity of a support. The remedy for these irregularities is the smoothed bootstrap: a small random shift of source points before each resampling. This shift is controlled by specifically selected distributions. The key issue is such parameter settings of these distributions to achieve the desired characteristics of the empirical distribution. This paper describes an example of this procedure.
Renata Dwornicka, Norbert Radek and Jacek Pietraszek
The paper considers the use of the bootstrap method to improve the determination of confidence intervals identified by the DOE (design of experiment) procedure. Two different approaches have been used: one that is appropriate for factorial designs and the other one relevant to the methodology of the response surface. Both approaches were tested on the real experiment datasets and compared with the results obtained from the classical statistical expressions based on well-known asymptotic formulas derived from the t distribution.