Wyniki wyszukiwania - BazTech

Ograniczanie wyników

2 Journal of Achievements in Materials and Manufacturing Engineering

2 2010

Powiadomienia systemowe

Sesja wygasła!
Sesja wygasła!
Sesja wygasła!
Sesja wygasła!

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Sortuj według:

Ogranicz wyniki do:

Application of multiple regression and neural networks to synthesize a model for peen forming process planning

Delijaicov S., Fleury A.T., Martins F. P. R.

Journal of Achievements in Materials and Manufacturing Engineering

2010

Vol. 43, nr 2

651--656

Purpose: this paper aims to present a simple method to synthesize an empirically-based model that permit to estimate the maximum displacement of a plate when a shotpeening process values are known. Design/methodology/approach: This approach regards the difficulty to develop a mathematical model to describe the relationship between the shot peening process variables (shot diameter, impact velocity, static preload and coverage) and the curvature of the piece. Such a model was generated through the application of statistical inference methods - multivariable regression and neural networks – to a set of experimental data concerning the application of peen forming processes to a group of 215 aluminium 7050 alloy rectangular plates. Findings: Although the estimated displacements from both models comply reasonably well with the experimental data, the obtained results exposed the superiority of the regressive model concerning accuracy. Research limitations/implications: Shot peen forming, a die less forming process, is one of the most successful methods to produce slight and smooth curvatures on large panels and plates. Through the application of a regulated blast of small round steel shot on the piece surface, a thin internal layer of residual compressive stress causes the elastic stretching of the shotted surface, giving rise to a permanent non-plastic deformation of the whole piece. Although this forming process has been used since the fifties, especially by the aerospatial industry, a scientific method for peen forming process planning has not been developed yet. Originality/value: The referred model can be used as an engineering tool to aid setting up a peen forming process in order to produce a desired curvature on a given plate.

Application of computer vision methods to estimate the coverage of peen formed plates

Vieira L. C., Almeida de R. H. Z., Martins F. P. R., Fleury A.T.

Journal of Achievements in Materials and Manufacturing Engineering

2010

Vol. 43, nr 2

743--749

Purpose: this paper aims to present a simple method that allows for a systematic estimation of coverage of peen aluminum workpiece submitted to a peen form process. Design/methodology/approach: This approach is based on the application of computer vision techniques for segmenting amplified images of the shot peening processed surface. The work has employed two combined methods of image segmentation – inductive algorithm generated rule segmentation and a multiagent segmentation system. Findings: The two combined methods of image segmentation has allowed for an estimation of low coverage plates as well as done by human expert. Furthermore a model of the spatial shot distribution was also achieved. Research limitations/implications: The surrogated method is suitable for plates with relative low coverages, circa 50 %. Originality/value: The model can be regarded as useful by acelerating the coverage evaluation in comparison with conventional industrial approach.