[Home] Return to Home Page

"Extending the Life of Bridges, Civil & Building Structures"

Structural Faults & Repair,
London, July 1995

"Quantifying the losses in cover-meter accuracy due to congestion of reinforcement"

John C Alldred MA, MSc, Protovale (Oxford) Ltd

Abstract

All commercially-available cover-measuring instruments are calibrated to indicate cover to a single bar. When multiple bars are present within the detection range of the instrument, the increased signal received from the increased mass of steel invariably causes the instrument to indicate a cover which is shallower than the true cover. This fact is well known by experienced users, yet not adequately quantified in published instructional literature; even the relevant British Standard (BS1881:204:1988) merely alludes to possible errors of up to ±5mm or ±15%, but does not suggest any values for bar spacing at which these errors may be expected.

This paper presents the results of systematic measurements of bars at realistic values of cover and bar pitch, using a variety of different instruments. It was found that, although the previously-published value for the minimum pitch-to-depth ratio for reliable operation is indeed a useful "rule of thumb", there were nevertheless wide variations in error at that ratio between different instruments and detector-head technologies. The testing methods that were used may be applied by any Civil Engineer or Surveyor who wishes to assess the suitability of an instrument to fulfil a particular cover measurement requirement.

Introduction

All cover meters are electromagnetic in operation, in that electric currents in a coil winding in a search head generate a magnetic field which propagates down through the concrete, and can interact with any metal present, such as reinforcing steel. The interaction will be due to either or both of two physical properties of the steel: its magnetic permeability and its electrical conductivity. The interaction will cause a secondary magnetic field to propagate back to the head, which can in principle be detected by a second coil winding or possibly the same winding (or alternatively, the interaction can be thought of as modifying the primary field).

The signal received will increase with increasing bar size, and decrease with increasing bar distance (cover). By making certain assumptions about the bar - and specifically by assuming that only one bar is present within the spatial extent of the primary magnetic field - an instrument can be calibrated to convert signal strength to distance, and hence indicate the depth of cover of concrete over the bar.

If in fact there are two or more bars within detection range (or one bar and some scaffolding), the instrument will receive a greater signal, and inevitably indicate a cover which is shallower than the true cover. This effect is well-known to users, but the magnitude of the likely error is a subject which is rarely publicised by manufacturers, and the aim of this paper is to quantify this effect.

British standard BS1881:204:1988 "Recommendations on the Use of Electromagnetic Cover Measuring Devices" [Ref.3] requires that when measuring cover to a single bar under ideal laboratory conditions, the error in indicated cover should be no more than ±5% or ±2mm, whichever is the greater.
It goes on to say that under favourable site conditions, an accuracy "approaching that obtainable in the laboratory" should be achieved. No numerical value is given, but it is interesting that an earlier 1984 draft of the Standard suggested a figure of ±3mm (but no percentage alternative); this value can usefully be employed here as a "desirable" accuracy target at shallow covers, augmented by an alternative of ±8% (derived by interpolation) for deeper covers.
The Standard continues by suggesting that "the accuracy of measurement likely to be obtained on the average site is within [the greater of] ±15% or ±5mm".

These values can be used to classify the performance of an instrument according to the error observed and the "conditions" of measurement, and this is summarised in Table 1.

The Standard also lists a number of extraneous factors which are potential sources of error. Those concerned with magnetic effects from the aggregate or matrix of the concrete, and variations in the properties of the steel and cross-sectional shape of the bars, should have but little influence on a well-designed cover meter, and will not be discussed here. Those concerned with multiple bars, and the proximity of neighbouring steel such as metal window frames or scaffolding, will be discussed in detail.

In this context, it should be appreciated that in practice, errors in indicated cover which are comparable to the surface roughness of the concrete, or the projecting height of the ribs of high-tensile bars, should be treated as inconsequential.

The bars used in the tests were all ribbed high-tensile steel, and therefore representative of the types of bars to be expected in concrete structures. The "concrete cover" was represented by machined plastic blocks, which ensured that the actual cover was specified from the surface of the search head to the surface of the bar's rib-profile, equivalent to the concrete cover to the nearest part of the bar.

The instruments to be compared were restricted to those developed within the last ten years, comprising four models, three of which are available with two sizes of search head; and are listed below in alphabetical order of manufacturer:
James Instruments HR7500;
Kolectric Microcovermeter, with both Maxiprobe and Miniprobe;
Proceq Profometer 3, with both Depth and Spot probes; and
Protovale CM5 CoverMaster, with both Standard and Midget heads.
In the tables, these have been abbreviated to:
JHR; KMax, Kmin; P3Dep, P3spot; CM5Std and CM5sm, respectively.

[Home] Return to Home Page
Back to Technical Library: Papers
[Reply] Information request