Use of Digital Image Correlation and Machine Learning for the Optimal Strain Placement in a Full Scale Composite Tidal Turbine Blade

Aug 5, 2026

Large-structure testing and health monitoring require extracting maximum information from a specimen while limiting the number of sensing devices. This work presents a data-driven strategy that combines Digital Image Correlation (DIC) measurements with machine learning algorithms to determine the optimal placement of single-point strain gauges. The approach uses DIC-derived strain maps to select measurement locations that preserve the sparsity structure while replacing many gauges with a minimal set. Results are demonstrated on full-scale composite tidal turbine blade experiments.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages3

Reading Time8 min

AccessFree to download

Abstract

Introduction

Test Data and Setup

Blade data

Method and Results

Digital Image Correlation

Edinburgh Research Explorer Use of Digital Image Correlation and Machine Learning for the Optimal Strain Placement in a Full Scale Composite Tidal Turbine Blade Citation for published version: McLoughlin, J, Munko, M, Valdivia Camacho, MA, Cuthill, F & Lopez Dubon, S 2023, Use of Digital Image Correlation and Machine Learning for the Optimal Strain Placement in a Full Scale Composite Tidal Turbine Blade. in BSSM 17th International Conference on Advances in Experimental Mechanics. British Society for Strain Measurement, 17th International Conference on Advances in Experimental Mechanics, Glasgow, United Kingdom, 31/08/23 . <[https://www.bssm.org/media/qjyhimyp/compensation-of-apparent-strain-data](https://www.bssm.org/media/qjyhimyp/compensation-of-apparent-strain-data)due-to-temperature-gradients-in-a-full-scale-mechanical-test-of-a-composite-tidal-turbine-blade. pdf> Link: Link to publication record in Edinburgh Research Explorer Published In: BSSM 17th International Conference on Advances in Experimental Mechanics General rights Copyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Take down policy The University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorer content complies with UK legislation. If you believe that the public display of this file breaches copyright please [contact openaccess@ed.ac.uk](contact openaccess@ed.ac.uk) providing details, and we will remove access to the work immediately and investigate your claim. Download date: 13. Nov. 2023 Use of Digital Image Correlation and Machine Learning for the Optimal Strain Placement in a Full Scale Composite Tidal Turbine Blade J. McLoughlin1, M.J. Munko 1, M.A. Valdivia Camacho1, F. Cuthill 1, S. Lopez Dubon1a 1 School of Engineering, The University of Edinburgh, The King’s Buildings, Mayfield Road, Edinburgh EH9 3JL, Scotland Corresponding [author](author a: Sergio.LDubon@ed.ac.uk)[ a](author a: Sergio.LDubon@ed.ac.uk)[: Sergio.LDubon@ed.ac.uk](author a: Sergio.LDubon@ed.ac.uk) Abstract. The testing and health monitoring of large structures represents many challenges , how to get as much information as possible from the specimen under analysis with a limited number of sensors. In this work, we use a data-driven approach using information from a Digital Image Correlation (DIC) system mixed with machine learning algorithms (MLA) to decide the optimal location of single-point strain gauges. Introduction Nowadays, technologies such as DIC allow for capturing vast information that reveals the deformation and stresses of small and large structures on almost all their surface. Compared with traditional strain gauges, it offers a huge benefit. Nevertheless , DIC requires specific setup and environmental conditions (i.e. , light) that hinder their implementation outside of controlled environments (i. e. , laboratories) , and then the use of stain gauges is necessary. Still, the optimal number of measurement points and their location remains open. To address this , we used the information from DIC in an MLA to select the points that keep most of the sparsity contained in the stain map revealed by the DIC. Test Data and Setup The data was collected at FastBlade , a testing centre that allows for testing large slender structures (2-14 meters) under either static or fatigue loads, as shown in Figure 1. The facility utilises a unique Digital Displacement® Pumps system, which incorporates regenerative pumping and digital displacement hydraulics, enabling accelerated testing (up to 1Hz) with high loads (up to 1MN) [1] . The DIC images were collected and processed using MatchID® [2] software for a clamping test, two static and one fatigue test. The test was performed using

FAQs

What is the goal of combining DIC and machine learning in this work?

To use DIC information together with machine learning to choose the optimal locations for a small number of single-point strain gauges, preserving the most informative structure of the strain map.

Why are strain gauges still needed even though DIC provides dense deformation information?

DIC requires specific setup and environmental conditions such as suitable lighting, which limits reliable use outside controlled laboratory environments, making strain gauges necessary.

How was the DIC system set up during the blade tests?

The setup used four cameras in stereo pairs, with two cameras covering the region between the blade root and the first actuator and the other two covering the region between the first and second actuators.

Aug 5, 2026

Upload by Lucas Martin