ANN modeling of pull-off adhesion of concrete layers. (November 2015)

• Record Type: Journal Article
• Authors: Sadowski, Łukasz; Hoła, Jerzy
• Is Part Of: Advances in engineering software. Volume 89(2015)
• Page Start: 17
• Page End: 27
• Publication Date: 2015-11
• Subjects: Floors -- Concrete layers -- Pull-off adhesion -- Non-destructive methods -- Artificial neural networks -- Interlayer bond; Computer-aided engineering -- Periodicals; Engineering -- Computer programs -- Periodicals; Engineering -- Software -- Periodicals; Periodicals; 620.0028553
• DOI: 10.1016/j.advengsoft.2015.06.013
• Languages: English
• ISSNs: 0965-9978

Abstract

Abstract: When making and repairing concrete floors it is vital to properly prepare the interlayer bonding surface. The measure of the bond is the value of pull-off adhesion f b experimentally determined in building practice by the semi-destructive (SDT) pull-off method. In this paper it is proposed to assess pull-off adhesion by jointly the optical laser triangulation method and the acoustic impulse response method, using artificial neural networks (ANN), on the basis of a few parameters (independent of top layer thickness) determined by these methods. The proposed non-destructive (NDT) pull-off adhesion assessment method is devoid of the drawbacks and inconveniences of the pull-off method and makes possible the reliable mapping of the adhesion level on the tested surface without local damage to the latter.