Nasza strona zapisuje niewielkie pliki tekstowe, nazywane ciasteczkami (ang. cookies) na Twoim urządzeniu w celu lepszego dostosowania treści oraz dla celów statystycznych. Możesz wyłączyć możliwość ich zapisu, zmieniając ustawienia Twojej przeglądarki. Korzystanie z naszej strony bez zmiany ustawień oznacza zgodę na przechowywanie cookies w Twoim urządzeniu.
Experimental determination of the properties of polymer composites can be a valuable source of
information about the behavior of a given material, and the use of modern information technologies,
such as artiﬁcial neural networks, gives the possibility of forecasting its properties while reducing the
number of observations carried, bringing both economic and environmental beneﬁts. The aim of the
publication is to consider the possibility of using artiﬁcial neural networks as a model describing the
impact of the percentage mass friction modiﬁer on selected mechanical properties and the formation
of abrasive wear of polymer composites used in aviation. Using of models of ANN will allow obtaining good generalizing (generalizing) properties resistant to the input of incorrect input data. The authors
chosed ANNs for accurate recognizing the relationship between any sets of inputs and outputs (from
tribological experiments) without formulating a physical model of a phenomenon under consideration.
The research was carried out for aviation polymer composite are investigated with the matrix of L285-
cured hardener H286 and six reinforcement layers of carbon fabric GG 280T and a physical modiﬁer of
friction in the form of alundum with changed mass share and grain size. The experiment data were
used to build artiﬁcial neural network (ANN). The diﬀerent various percentage mass shares parameters
and grain sizes of alundum were used as inputs and the weight loss between the cycles and selected
mechanical properties as output of the model. The predicted values of the ANN model were veriﬁed with
the actual values. The proposed solution is an attempt to describe a complex nonlinear system which is
tribological system. The authors see the potential of this method in planning research experiments. The
model can help to reduce the number of samples produced. This is a time, economic and environmental
advantage. However, it should be noted that the eﬀectiveness of the ANN model crucially depends on
the amount and generality of the training data.