Development of image analysis techniques to assist evaluation of both air void structure and aggregate shape factors in concrete

Abstract

Significant reduction in both strength and durability of concrete is brought about by voids left within the concrete once it has hardened. These voids can arise from a number of sources both intentional, in the case of air-entrained voids introduced by admixtures batched to provide a structure that can withstand frost attack and unintentional in the case of entrapped voids that arise due to characteristic of the sand and aggregates, excess water added and a lack of consolidation whilst plastic. Aggregate form, shape and texture are known to influence the way particles pack together and therefore the amount of space left between and amongst the particles. The second phase of this study has used desktop flat-bed scanning to record aggregate profiles, both raw particles and aggregate shape profiles taken from the curved surface of core samples, to classify the shape and then provide a protocol for defining the shape. This study has shown the photographs provided originally in Concrete Society Technical Report 11 and now recently re-introduced into the UK Annex of BS EN 12504-1:2009 Testing concrete in structures – Core specimens – Taking, examining and testing in compression, provide a misrepresentation of the curved surface of the core. A curved surface cannot be recorded faithfully by a 2D camera image. An accurate representation of the curved surface of concrete core samples has been obtained by the use of conventional desktop scanners, albeit using relatively high image resolution. By a novel yet simple modification, concrete core samples have been mechanically rolled above a modified flatbed desktop scanner driven by the crosshead so as to align directly above the cold cathode fluorescent (CCF) light source and the scanning charged coupled cross-head image recording device. A method of assessing the amount of voidage found within the curved surface of concrete core samples has been developed. A freely distributed software programme was used to process all images to determine percentage voidage and voids size distribution among other attributes. A second freely available statistics software package has been used to analyse the results. iii The second phase of the study has used the same scanning technique to classify 2D aggregate profile as used for voidage shape recognition taken from the curved surface of core samples. Three simple shape factors have been used, one developed specifically for this study. By means of Riley Circularity, Percentage Concavity and Aspect Ratio core surface aggregate profiles, raw aggregate shapes and voidage found on core samples have been classified. The objective being to determine if the aggregate within a sampled concrete has changed from that intended, possibly due to crushing oversize material or just changes within the source that would provide a means of assessing any influence aggregate shape change may have had on entrapped voidage and the effect that has had on the strength of the concrete. Scanner performance and calibration has been checked using high resolution calibration sheets. The image resolution was found to be accurate to 2.5% at 0.5mm diameter when scanned at 1200 dpi. This enabling the threshold to be investigated between entrained and entrapped air voids so as to allow discrimination between any combinations of the two found on a core sample. The equipment, equipment modifications, procedures, test protocols as well as the imaging software and statistical analyses packages included in this study have been chosen so as to allow others to utilise the benefits such analyses offers. The recent acceptance and drive to recycle materials for use as construction aggregate can benefit from classification by these procedures that until now have not been included in any published Standards. The procedures developed during this study have been published in the Magazine of Concrete Research, Dec 2010 and presentations given by invitation to joint meetings of the Concrete Society/Institute of Concrete Technology at Loughborough and in London, 2011.