A NOVEL CLUSTERING ALGORITHM FOR FLAW IDENTIFICATION IN PLUG VALVES THROUGH NON -DESTRUCTIVE TESTING AND THERMAL IMAGING

  • Jacintha.V, Dr.S.Karthikeyan,
Keywords: Dye Penetrating Test (DPT), Passive Infrared Thermography (PIT), Plug valve Defects, surface roughness, Fuzzy Clustering

Abstract

The objective of this paper is to identify the defects in Plug valves. The thermal images of the Plug valve are capture utilizing a Thermal Fluke camera[TiS20]. Plug valve thermal images are utilize for detection of defects such as Crack, Porosity, Internal defects. The thermal images can solely identify surface flaws and unable to identify subsurface flaws in plug valves. The identification of subsurface defects in Plug valves seems to be a challenge. To overcome this drawback, we apply Dye Penetrating Test (DPT) prior to Thermal Imaging. In DPT, a thinned paint is apply on the surface of the metal that exposes porosity defects and fine cracks. The Plug valve thermal images that are fetch post DPT, enables identification of variations in heat flux and defects on the defective metal surface. After DPT, the Plug valve thermal images are processed using Fuzzy algorithm to identify Internal Plug valve defects. Hence, we are proposing a combination of Dye Penetrating Test – Infrared Thermography (DPT-IT) with fuzzy algorithm to detect plug valve defects more effectively.  The Plug valve defects quantify through Slicing, Precision, Recall, F-score, Jaccard similarity coefficient. The parameters show relative changes with respect to surface roughness and metal defects. The DPT-IT and Fuzzy algorithm detect metal flaws with 88.6% accuracy.

Published
2021-07-07
How to Cite
Dr.S.Karthikeyan, J. (2021). A NOVEL CLUSTERING ALGORITHM FOR FLAW IDENTIFICATION IN PLUG VALVES THROUGH NON -DESTRUCTIVE TESTING AND THERMAL IMAGING. Design Engineering, 2023- 2033. Retrieved from http://thedesignengineering.com/index.php/DE/article/view/2527
Section
Articles