City Research Online

Abstract

In the developed world, acid rain attack, chloride contamination, and inadequate design of quality control during the construction stage, has contributed to the damage of reinforced concrete buildings and structures. The importance of non-destructive testing (NDT) of concrete is substantial, so that consistent and effective maintenance of the structures is implemented correctly. This requires assessment methods that can identify the initiation of defects so that appropriate actions may be taken to prevent large scale deterioration.

The impact echo test method is one of the most recent of the NDT used on concrete for the detection of damage, and has been applied in this research. However, the interpretation and analysis of the output requires subjective judgement. This research proposes to use an artificial neural network in impact echo data output analysis and interpretation. Neural networks comprise of numerical processing elements which are linked in a way that the network generally can learn by examples and store such experience for later use. They are trained using past data records from the output, so that an appropriate trained network is able to generalise when presented with inputs not appearing in the training data. The major feature of neural networks for this application is that it does not involve subjective judgement, and provides a fast and efficient method for analysis of large quantity of data.

In the research, the finite element method is used to build a simulation of the NDT method. A model of a wall with voids, for example, is built where a force or pressure is used to represent the impact, the response measured at a nearby point. As well as providing theoretical analysis, this numerical method allows creation of reliable data for use in neural networks training. However, due to the time limit, the author was not able to use finite element data in the neural network analysis.

Publication Type:	Thesis (Doctoral)
Subjects:	T Technology > TA Engineering (General). Civil engineering (General)
Departments:	School of Science & Technology > Engineering

Export

Downloads