Scholarly record

USE OF ARTIFICIAL NEURAL NETWORK IN PREDICTING THE WEAR RESISTANCE OF SOME IRON-BASED POWDER METALLURGY MATERIALS SUBJECTED TO A THERMOCHEMICAL TREATMENT

Mihaela MARIN

Abstract

In this paper, the artificial neural network (ANN) technique is used to predict the wear resistance e under dry sliding conditions of some iron-based powder metallurgy materials subjected to a thermochemical treatment. The analyzed specimens were obtained using the conventional powder metallurgy route: mixing the raw powders with 1% zinc stearate, cold compacting at a pressure of 400 MPa and sintering at a temperature of 1150°C for 60 minutes. Following the sintering step, the specimens were subjected to carburizing in fluidized bed in a laboratory furnace at 930°C for 30 and 60 minutes. The compacts disc dimensions obtained are O8x6 mm. Density, porosity, particle size and carburizing time were used as the inputs parameters. Wear rate and hardness were used as the outputs parameters.The wear tests results were compared, analysed and predicted by applying the ANN technique. It is observed that the ANN predicted values are in good agreement with the experimental values obtained. The results show that ANNs can be use as an alternative technique to evaluate the wear rate and harndess values of the materials and also can reduce the large experimental work. The applications of proposed model not only reduce the overall experimental test but also reduce the cost by optimizing the process parameters.

Publication Impact Profile

• Captures
• Mendeley - Readers: 3

PLUMX see details

DOI resolver: doi.org/10.5593/sgem2019/2.1/s07.090

Publication details

Back to publication list