Processing and Tribo-Performance Analysis of Polymer Composites Filled with Linz-Donawitz Sludge

Abstract

Integrated steel plants, in general, produce large amounts of solid wastes during the production of iron and steel. Linz–Donawitz (LD) sludge are the fine solid particles recovered after wet cleaning of the gas emerging from LD converters during steel making. In general 8 kg of LD sludge is generated per ton of crude steel production. This solid waste can have many valuable product-oriented applications if processed economically such as recovery of metal values, recycling in sinter plant etc. But its use as a particulate filler in polymer composites is not yet established. This work attempts to utilize this LD sludge as the filler material in polymers like epoxy and polypropylene for composite making. The sludge used in this work is collected from Rourkela Steel Plant located in the eastern part of India. This research reports the processing details and the wear characteristics of a series of epoxy-LD sludge and polypropylene-LD sludge composites. The composites are characterized in regard to their physical and mechanical properties. A wealth of property data has been generated by conducting different characterization tests under controlled laboratory conditions.
In the present work, theoretical models have also been developed and correlations have been proposed for estimating sliding and erosion wear rates of particulate filled polymer composites respectively. Tests have been carried out for validation of these theoretical models. Dry sliding and solid particle erosion wear characteristics of LD sludge filled polymer composites have been successfully analyzed using Taguchi experimental design and response surface methodology (RSM). The material variables and the experimental factors that are found to be influencing the respective wear rates are identified. It is found that the influence of LD sludge particles in reducing the wear rates is significant. The microstructural features of the worn surfaces of various particulate filled composite specimens are examined using scanning electron microscopy in order to ascertain the wear mechanisms.
Two predictive models; one based on Taguchi approach and the other on artificial neural networks approach are proposed in this work. It is demonstrated that these models well reflect the effects of various factors on the wear loss and their predictive results are in good agreement with the experimental observations. Neural computation is successfully applied in this investigation to predict and simulate the wear responses of these composites under various test conditions within and beyond the experimental domain. The predicted and the experimental values of sliding and erosion wear rates exhibit good agreement and validate the remarkable capability of a well-trained neural network for these kinds of tribological processes.