Elsevier

Engineering Structures

Volume 225, 15 December 2020, 111309
Engineering Structures

Combined application of distributed fibre optical and digital image correlation measurements to structural concrete experiments

https://doi.org/10.1016/j.engstruct.2020.111309Get rights and content

Highlights

  • Combined use of distributed fibre optics and digital image correlation is explored.

  • Procedure to quantify DIC uncertainty in large scale structural tests is presented.

  • DIC uncertainty is highly dependent on the quality of the calibration.

  • Noise of distributed fibre optics is dependent on the absolute strain level.

  • Crack locations given by both measurements agree in series of concrete panel tests.

Abstract

The combined application of distributed fibre optical strain measurements on reinforcing bars and digital image correlation (DIC) measurements on the concrete surface has a great potential to increase knowledge in many fields of structural concrete. This paper explores the advantages of these measurement techniques for concrete tests and the key aspects to be considered in order to obtain reliable measurements suitable for quantitative analysis. The uncertainty of DIC analysis is highly dependent on the test conditions and user carefulness, and should be assessed for each test. A procedure to quantify the DIC uncertainty in large scale structural tests is presented, showing that it is highly dependent on the quality of the calibration. Comparative tests on distributed fibre optical strain measurements with different fibre coatings show that polyimide-coated fibres capture properly high strain gradients and, therefore, should be used when instrumenting reinforcing bars in RC specimens. Moreover, the measuring noise was found to be dependent on the absolute strain level. Combined plots of crack kinematics and reinforcement strains, stresses and forces are shown for the results of a series of two concrete panel tests subjected to diagonal tension. Crack locations predicted by both measurements match perfectly in these experiments.

Introduction

The knowledge in structural concrete is perfectly developed for the dimensioning of new structures using conventional reinforcement and construction methods. Such structures can be designed to provide sufficient deformation capacity (i.e. by ensuring that failure is governed by yielding of the reinforcement) in order to guarantee a safe application of lower bound limit analysis methods [1], [2], [3]. However, there is a rising need and interest in the assessment of existing structures as well as in the use of non-conventional reinforcement (e.g. textile and fibre reinforcement) and digital fabrication methods [4], [5], for which limit analysis often is not applicable. Therefore, the development of sound mechanical models for existing structures and non-conventional materials is key to avoid unnecessary retrofitting measures and to open the way for the design of leaner new structures. Many such mechanical models have already been proposed for conventional reinforcement [6], [7], fibre reinforcement [8], [9] and even textile reinforcement [10], but all have been derived based on experiments with very limited directly measured information (typically only loads and deflections, sometimes average strains). Due to this limitation, existing models are prone to containing inconsistencies and/or empirical factors only applicable for the particular case they were derived [11], what hinders their practical application. Novel instrumentation techniques, such as optical systems, allowing distributed measurements on the surface of, and inside the concrete will be essential for the validation and further development of these models. These novel technologies complement standard measurement technologies providing insight into the mechanical behaviour of structural concrete (in particular the kinematics) that hardly has been possible up to now.

Most structural concrete experiments conducted in the past were instrumented using standard measurement technologies. These standard measurements consisted of load-cells to measure applied loads or reactions, and linear displacement sensors (e.g. LVDTs, potentiometers and strain gauge based sensors) to measure selected deflections and average deformations between two points. Strain gauges were sometimes used as well to record strains on reinforcing bars and on the concrete surface, but only a limited number of points could be instrumented (not only because of cost reasons, but also because sensors affect bond behaviour and change the member’s characteristic behaviour when applied to reinforcing bars). Moreover, discrete strain measurements on reinforcing bars provide information of limited practical use unless the gauges are located exactly at a crack (which cannot be ensured when placing them except in very particular cases [12]). Using these standard measurement techniques it was possible to determine the load-deformation behaviour and to estimate the global failure mode of the tested members. While this information was sufficient for developing design approaches based on the lower bound limit analysis, knowing more detailed information is essential to fully understand the behaviour of elements with low deformation capacity (e.g. members with very low transversal reinforcement ratios, or with non-conventional reinforcement of limited ductility).

Driven by the impossibility to gather detailed information in real-life structural members, researchers developed simplified test setups to study isolated structural phenomena matching the limitations of existing instrumentation. For instance, aggregate interlock relationships were derived typically based on push-off tests [13], bond relationships from pull-out experiments [14] and tension-stiffening mostly from direct tension or pure bending tests [15]. While these derivations provide valuable insight into these phenomena, there remains an uncertainty about their validity when applied for conditions different than those for which they were derived. For example, aggregate interlock relationships derived in push-off experiments with preformed cracks and theoretical uniform crack openings [16], might not be representative for members containing real cracks with significant macro-roughness and non-uniform crack openings. Therefore, the implementation of empirical relationships derived from simple tests into mechanical models with a general scope is prone to potential inconsistencies [11], [17]. To avoid these inconsistencies, refined measurements of non-oversimplified tests representing real-life structures are essential.

Some early attempts in this direction include the use of photoelasticity to quantify the strain field and visualise the cracking pattern [18] or the visual inspection of cracks on the concrete surface. The latter is still the most popular refined measurement nowadays, in which crack patterns are marked by eye inspection and the width of selected cracks is estimated by visual comparison with printed line-widths. The tendency to measure the widest cracks might bias the statistical results obtained with this approach. A more precise measurement consists in tracking the variation of distance between metallic targets glued to the concrete surface by means of a demountable mechanical strain gauge (DEMEC) [19]. Besides crack opening, this procedure also measures compressive strains and crack sliding (provided a grid of triangular targets is tracked), enabling an advanced analysis based on the experimental results (e.g. estimation of the contribution of the different shear-transfer actions [20]). However, all these methods (i) are labour-intensive and error-prone (many manual measurements are required at each measuring step), (ii) only provide discrete information for the concrete surface at certain load steps, and (iii) often cannot be used close to failure due to safety reasons. As discussed in this paper, recent advances in optical measuring techniques overcome these limitations and are therefore replacing rapidly these classical measuring techniques (considered as state-of-the-art until very recently).

The present work addresses the potential of combining in structural concrete experiments two of these novel technologies, i.e. full field digital image correlation (DIC) and distributed fibre optical strain sensing, as schematically illustrated in Fig. 1. These instrumentations are able to monitor crack kinematics and concrete strains on the concrete surface and strains along reinforcing bars quasi-continuously (in time and space) without disturbing the structural behaviour of the specimens: DIC measurements are contactless and fibre optics uses tiny glass fibres as sensors with negligible impact on bond. The recorded results allow determining the internal flow of forces of the specimens, as shown in this paper for a series of panel tests. Moreover, the forces transferred across cracks can be obtained directly from equilibrium at the crack locations. An important challenge when using these instrumentation methods lies in the large amount of collected data. This can be overcome by developing tools to automatically transform the acquired data into useful structural information. It should be noted that the knowledge on the accuracy of these technologies and the dependency of the results on the application conditions is still very limited, despite of their potential and their increasing use in structural concrete experiments (particularly for the case of DIC). This uncertainty compromises the reliability of the results except for qualitative interpretations. To improve this situation, several tests to estimate the measurement uncertainty of DIC and fibre optics when applied to large scale structural concrete experiments are discussed in this paper.

Section snippets

Digital image correlation

Digital image correlation (DIC) is a novel but already well-established optical method that allows tracking the displacements of a surface based on pattern matching of digital images [21]. This paper focuses in stereoscopic applications of DIC (3D-DIC), which provide three-dimensional displacements as well as the shape of the surface, which allows avoiding errors due to out-of-plane displacements inherent to two-dimensional DIC (2D-DIC) applications with a single camera. Measuring with DIC

Measurement uncertainty

In order to collect and evaluate reliable data in structural experiments, it is of utmost importance to assess the uncertainty of the novel optical measurements presented in Section 2. Assessing the uncertainty of a measurement comprises the quantitative estimation of the accuracy, i.e. assessing the deviation of the measured value to the true value. Such deviation from the true value is caused by a combination of systematic errors (bias) and random errors (noise). As it is impossible to

Application to concrete panels

This section presents and discusses the results of a series of two concrete panel tests subjected to diagonal tension, which were conducted to evaluate the potential and the concept of the combined application of fibre optical strain sensing on the reinforcing bars and digital image correlation on the concrete surface. These tests were used to inform the instrumentation setup to be used in the Large Universal Shell Element Tester LUSET at ETH Zurich [46] consisting in 2 m side length and 350 mm

Concluding remarks

This publication shows that the combined application of distributed fibre optical and digital image correlation measurements has a great potential to increase knowledge in many fields of structural concrete, such as bond and cracking behaviour or load carrying and deformation capacity of elements with low amounts of transverse reinforcement. Both measuring systems can document the phenomenological effects continuously, comprehensively and at a very detailed level, both inside and on the surface

CRediT authorship contribution statement

Jaime Mata-Falcón: Conceptualization, Methodology, Investigation, Writing - original draft, Supervision. Severin Haefliger: Conceptualization, Methodology, Investigation, Writing - original draft. Minu Lee: Software, Investigation, Writing - original draft, Visualization. Tena Galkovski: Software, Investigation, Writing - review & editing, Visualization. Nicola Gehri: . : Software, Investigation.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

Part of the experiments described in Section 3.2 were carried out as a part of the Master’s Project of Nathalie Reckinger and Caroline Wirion at ETH Zurich, whose contributions are greatly acknowledged. The authors would also like to thank the staff of the structural engineering laboratory at ETH Zurich for their support, Walter Kaufmann for the proofreading and the providers of the digital image correlation system Isi-sys GmbH and Correlated Solutions Inc. for their support and suggestions for

References (56)

  • F.J. Vecchio et al.

    The Modified Compression Field Theory for Reinforced Concrete Elements Subjected to Shear

    ACI Journal

    (1986)
  • W. Kaufmann et al.

    Compression field analysis of FRC based on the cracked membrane model

    ACI Struct J

    (2019)
  • T. Markic et al.

    Strength and deformation capacity of tension and flexural RC members containing steel fibres. ASCE

    J Struct Eng

    (2020;in)
  • P. Valeri et al.
  • Kaufmann W, Mata-Falcón J, Beck A. Future directions for research on shear in structural concrete. fib Bulletin 85:...
  • J.C. Walraven et al.

    Theory and Experiments on the Mechanical Behaviour of Cracks in Plain and Reinforced Concrete Subjected to Shear Loading

    HERON

    (1981)
  • R. Eligehausen

    Local bond stress-slip relationships of deformed bars under generalized excitations: experimental results and analytical model

    (1983)
  • M. Alvarez
    (1998)
  • J.C. Walraven

    Fundamental Analysis of Aggregate Interlock

    J Struct Divis, ASCE

    (1981)
  • A. Beck et al.

    Cracked Membrane Model with Fixed, Interlocked Cracks: Numerical Implementation and Validation

    J Struct Eng, ASCE

    (2020)
  • V.F. Bignell et al.

    A new photoelastic material for use in problems concerning reinforced concrete

    Mag Concr Res

    (1963)
  • P.B. Morice et al.

    The design and use of a demountable mechanical strain gauge for concrete structures

    Mag Concr Res

    (1953)
  • S. Campana et al.

    Analysis of shear-transfer actions on one-way RC members based on measured cracking pattern and failure kinematics

    Mag Concr Res

    (2013)
  • M.A. Sutton et al.

    Image Correlation for Shape, Motion and Deformation Measurements: Basic Concepts, Theory and Applications

    (2009)
  • P. Reu

    Introduction to Digital Image Correlation: Best Practices and Applications

    Exp Tech

    (2012)
  • Schreier HW. Accuracy Improvements and Meaningful Error Metrics for Stereo System Calibration. Extended Abstract,...
  • Huber P, Huber T, Kollegger J. Investigation of the shear behavior of RC beams on the basis of measured crack...
  • D. Samiec

    Verteilte faseroptische Temperatur- und Dehnungsmessung mit sehr hoher Ortsauflösung

    Photonik

    (2011)
  • Cited by (0)

    View full text