Share Email Print
cover

Proceedings Paper

Visual analysis of ductility/brittleness of welding fracture points on charpy test specimens using graphical blocks on field programmable gate arrays
Author(s): Andrew J. Tickle; Anyela Camargo-Rodriguez; Jeremy S. Smith
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

The charpy impact is a technique used to evaluate the toughness of an engineering material that determines the amount of energy absorbed by it during fracture. Initially, measurements were estimated manually and later replaced by a PC version. This study reports the development of the Field Programmable Gate Array (FPGA) portable version. The FPGA based version allows easy analysis of samples without the need of sending them to a lab for analysis. The process, presented here, as the original, is based on measuring the percent of crystal in the test sample after impact, to determine if the material is ductile or brittle. The FPGA version, adapted under the MATLAB Simulink environment, shows a graphical block representation of the charpy impact PC version. An important asset of the FPGA version is its portability, it has to be easily modified and downloaded onto a device to estimate the percent of brittle fracture of the broken Charpy surface. The beauty of the DSP Builder programme is that it allows the model to be compiled to various types of optimised code for any Altera FPGA device. To provide a firm basis for scientific comparison to the new FPGA system, images already analysed via the PC based Java system were also used for testing and comparison purposes. The FPGA system converts the image into an 8 bit grayscale image and analyses it in a 5x5 sampling window. This produces texture features that can be used in a comparison system, similar to the Support Vector Machine (SVM) used in the original. The output is a signal that states the material being tested is brittle or not via an output of '1' for brittle and a '0' for ductile. A detailed pixel by pixel analysis of the various output images is then investigated to state the percentage difference between the PC and FPGA based systems.

Paper Details

Date Published: 27 September 2008
PDF: 11 pages
Proc. SPIE 7100, Optical Design and Engineering III, 71002M (27 September 2008); doi: 10.1117/12.797720
Show Author Affiliations
Andrew J. Tickle, The Univ. of Liverpool (United Kingdom)
Anyela Camargo-Rodriguez, Univ. of Ulster (United Kingdom)
Jeremy S. Smith, The Univ. of Liverpool (United Kingdom)


Published in SPIE Proceedings Vol. 7100:
Optical Design and Engineering III
Laurent Mazuray; Rolf Wartmann; Andrew Wood; Jean-Luc Tissot; Jeffrey M. Raynor, Editor(s)

© SPIE. Terms of Use
Back to Top