Novel Strain-based NDE for Online Inspection and Prognostics of Composite Sub-structures with Manufacturing Induced Defects

Home > Research > Feasibility Studies

Novel Strain-based NDE for Online Inspection and Prognostics of Composite Sub-structures with Manufacturing Induced Defects

Host Institution: University of Southampton

Lead Investigator: Daniel Bull, Ian Sinclair, Janice Barton, Ole Thomsen


To save time and reduce wastage it is essential that an inspection technology is developed that can be used to intervene at the manufacturing stage and provide high fidelity data for model based prognostic capability (beyond simply sizing defects) to further inform the decision process of ‘accept’, ‘rework’, ‘repair’ or ‘scrap’. To this end a novel inspection procedure for cured composite components needed to be developed, and the overarching aim of the project is a system deployed alongside current inspection approaches in the production environment. This procedure has the ability to simultaneously measure the strain and a stress measure in the vicinity of a manufacturing defect or intrinsic subsurface artefact (including the variability of fibre volume fraction and fibre orientation, not just the geometry and size), and provide high fidelity data to inform model-based prognostics, and define how a given defect will evolve under service load.

The objectives of the of the feasibility study were:

  • To establish the underlying physical principles, and to demonstrate the viability of the experimental methodology based on combining thermoelastic stress analysis (TSA) and digital image correlation (DIC) and collecting data simultaneously.
  • Demonstrate that the data necessary for the model based prognosis system could be obtained by demonstrating the viability of the approach at a sub-structural level.


The feasibility study is completed. It has been fully demonstrated that DIC and TSA can be used simultaneously to collect data from composite components. The approach has been demonstrated on a high value carbon fibre aircraft component and shown that the results can be linked to the findings of high fidelity prognostic models, with defect geometry defined by X-ray CT.

The approach has also been demonstrated on materials typical of those used in high volume manufacturing made from carbon fibre/epoxy discontinuous compression moulded preforms. The work has linked observations from X-ray CT to the mechanical response and that the combination of DIC and TSA for assessment of the material shows great promise. In particular it is possible to predict the localised stiffness variations, linked to local variability of fibre volume fraction and fibre orientation, throughout the preform. This indicates that the manufacturing control process could be directly informed/updated using the technique. It has also been demonstrated that the technique can be protable (i.e. no need for a test machine to load the components) by exciting the component briefly at its resonant frequency.

The PhD study is focusing on developing a low cost infra-red camera for TSA. The work has revealed in the first 9 months what the scientific challenges are in developing the technology further, and provided the first steps in developing signal processing routines that will address the challenges.

Fully integrated DIC and TSA: the key challenge was in collecting the data simultaneously. This was done by using lock-in processing for the DIC. Although the technique was developed at University of Southampton previously, the key challenge was deploying this on a realistic component.

The strain measurements from the DIC and the thermoelastic response, which is dependent on the local fibre orientations and fibre volume fractions have provided strong indication of identification of stiffness distribution and resin-rich volumes. Initial demonstration of vibration based technique on panels.

Evidence of Impact

The technology developed in the feasibility study is new and the next steps will provide practical demonstrators based on a realistic high volume manufactured and high value manufactured components loaded either naturally or by excitation at their resonant frequency. Now feasibility has been demonstrated, a strong consortium with a focus on manufacturing of composite components both high value and high volume will be established to seek further funding from the Future Composites Manufacturing Research Hub, to develop the study into a Core Project.

Copyright © 2017 CIMComp