Resin transfer monitoring using capacitive sensors

Tools

McInnes, Martin and Gomes, Rylan and Mohseni, Ehsan and Pierce, S. Gareth and Dobie, Gordon and Zhang, Dayi and MacLeod, Charles N. and Munro, Gavin and O'Brien-O'Reilly, Janet and O'Hare, Tom (2023) Resin transfer monitoring using capacitive sensors. In: 50th Annual Review of Progress in Quantitative Nondestructive Evaluation, 2023-07-24 - 2023-07-27, Sheraton Austin Hotel at the Capitol,.

[thumbnail of McInnes-etal-QNDE-2023-Resin-transfer-monitoring]
Preview
 Text. Filename: McInnes-etal-QNDE-2023-Resin-transfer-monitoring.pdf
Accepted Author Manuscript
License: Strathprints license 1.0
Download (1MB)| Preview

Abstract

Resin Transfer Moulding (RTM) process is widely used in the manufacturing of composite materials. Stacks of carbon fabrics are preformed and moulded, and then placed under vacuum to transfer the resin into the mould. The samples are often cured at room or high temperatures depending on the process and requirements. Defects such as porosity and dry spots can form if the carbon fabrics are insufficiently impregnated by the resin. This brings in the need for resin transfer monitoring particularly for high value manufactured aerospace components. This research proposes a Non-Destructive Evaluation (NDE) sensor technology based on capacitive sensing. The sensor uses electric field distribution formed between two electrodes to measure the changes in electrical permittivity of the test material. Owing to the difference between air and resin electrical permittivity’s, the method can be deployed to monitor the resin flow front in the RTM process in the manufacturing of aerospace grade composites. A series of co-planar capacitive sensors with three different surface areas (5x5mm, 10x10mm & 25x25mm) and four separation distances (1mm, 5mm, 10mm & 25mm) of electrodes were designed and manufactured. The performance of these sensors was examined for vertical and horizontal flow where all the experiments substituted resin (relative permittivity - 3.6) for water (relative permittivity -80) for initial feasibility study. For vertical flow sensitivity studies, two experiments were carried out. In the first experiment the sensor was fixed on the top of a container and water level was changed incrementally in steps of 2mm to change the distance of water to sensor from 140mm to 70mm and the resulting changes in the sensor's impedance were recorded across a frequency range of 10kHz – 1MHz. The second experiment involved monitoring rising water levels with the addition of a non-crimp triaxial (0/45/135) carbon fabric sheet to understand the effectiveness of the sensing technology beyond carbon fabric conductive layers. The result of the study was compared to COMSOL Multiphysics Finite Element Analysis package simulations for validations, and to observe the distribution of electric field as the water level rises. According to the impedance changes, observed both experimentally and in COMSOL, across a large frequency range, sensor (25x25mm, 10mm separation) found to be the most sensitive with peak sensitivity at 324 kHz. The sensor was able to sense the water surface when 70mm away from the sensor and a good match was found between the simulation and experiment. However, a significant reduction in the detection level was observed, to only 2mm away from the sensor, when one layer of triaxial carbon fabric sheet was added to the set up. Vertical flow studies were followed by implementing a series of capacitive sensors to monitor horizontal resin flow at room temperature. Initially a single sensor (25x25mm, 10mm separation) was placed on a custom mould set up and water was injected from one side for the flow front to traverse along the sensor’s longitudinal axis (the symmetry axis cutting both drive/sense plates in half). The set up without any carbon fabric showed high sensitivity of 4.4pF to the horizontal flow showing gradual increase of capacitance the water passes along the sensing electrode. In view of this success, a second experiment was conducted where 5 layers of biaxial carbon fabric (+45/-45) were stacked in a vacuum (-28mmHg) bagging mould. Four capacitive sensors with a new configuration (with a drive electrode surrounded by sensing electrode) and a new geometry (elongated stripe with length of 70mm), were placed on top of the vacuum bag. A custom LabVIEW VI was developed to monitor the progress of resin front flow across each of these sensors and to plot 2D map of the resin coverage at each instance of the infusion process. This demonstrator showed that the horizontal flow of resin can be monitored in 2D in real time. Dry spots were intentionally created into some of the samples to verify the sensors’ sensitivity to such defects during the infusion.

ORCID iDs

McInnes, Martin, Gomes, Rylan, Mohseni, Ehsan ORCID logoORCID: https://orcid.org/0000-0002-0819-6592, Pierce, S. Gareth ORCID logoORCID: https://orcid.org/0000-0003-0312-8766, Dobie, Gordon ORCID logoORCID: https://orcid.org/0000-0003-3972-5917, Zhang, Dayi ORCID logoORCID: https://orcid.org/0000-0003-4611-4161, MacLeod, Charles N. ORCID logoORCID: https://orcid.org/0000-0003-4364-9769, Munro, Gavin, O'Brien-O'Reilly, Janet and O'Hare, Tom;
CORE (COnnecting REpositories)