During the curing process of epoxy resin, volume shrinkage and differences in thermal expansion coefficients between the resin and the curing mold lead to residual stress accumulation, which further induces residual strain on the mold surface. This residual strain degrades component performance, affecting their service life and reliability. Conventional strain monitoring methods, such as resistance strain gauges, exhibit limitations including large temperature drift, susceptibility to electromagnetic interference, and inability to operate stably in harsh environments, thus being unsuitable for long-term monitoring. Therefore, it is crucial to develop a reliable and high-performance method for real-time monitoring of the annular surface strain of epoxy resin curing molds, providing a basis for optimizing the curing process and improving component quality.

To address the above issues, fiber Bragg grating (FBG) sensors were employed for real-time strain monitoring. The FBG temperature sensor was encapsulated using a sleeve. The FBG with a grating length of 7 mm was placed within a 304 stainless steel sleeve, with both ends fixed using 704 silicone rubber (Fig.1). For the FBG strain sensor, the entire grating area and the optical fibers at both ends of the grating were bonded to a substrate using EPO-TEK 353ND adhesive, and the substrate was then bonded to the surface under test using 946 adhesive (Fig.2). QSn6.5-0.1 phosphor bronze was selected as the substrate material. Software simulation results showed a uniform strain distribution within the substrate (Fig.4~Fig.6), verifying the reliability of the FBG strain sensor. The epoxy resin curing mold consisted of a silicone rubber base plate, a small cylinder , and a large cylinder(Fig.7). The FBG strain sensor was attached perpendicular to the cylinder axis, and the FBG temperature sensor was attached parallel to the axis, with their centers symmetric about the cylinder axis (Fig.8). The epoxy resin system used was bisphenol F epoxy resin and diethyltoluene diamine at a mass ratio of 5:1. The strain monitoring system included a self-developed fiber grating demodulation system and a ZT-XL-150 programmable constant temperature and humidity test chamber (Fig.9). Four temperature control processes were designed for the experiments (Fig.10), with three repeated tests conducted under each process. The sensor performance (strain and temperature response characteristics) was tested using a cantilever beam and a constant temperature and humidity chamber, respectively.

Performance tests showed that the FBG strain sensor had a strain measurement error within ±5 με and a full scale (FS) measurement accuracy of ±4.69%. The strain sensitivity coefficient of the FBG strain sensor ranged from 0.76 pm/με to 0.82 pm/με, demonstrating high linearity (Fig.11). The FBG temperature sensor encapsulated within the stainless steel sleeve exhibited a strain sensitivity coefficient ranging from 0.0073 pm/με to 0.0094 pm/με, effectively avoiding strain interference (Fig.11). The temperature sensitivity coefficient of the FBG strain sensor ranged from 14.0 pm/°C to 16.5 pm/°C, also exhibiting high linearity (Fig.12). Strain monitoring during the curing process revealed distinct strain characteristics at different stages (Fig.13). During the heating stage, the mold surface initially showed no strain. Then, the positive strain increased (due to thermal expansion of resin molecules), reaching a maximum of 38 με ± 3 με at 75 °C, and it decreased to negative strain (due to chemical shrinkage). During the holding stage at 95 °C, the strain stabilized at -40 με ± 3 με. During the natural cooling stage, a strain maximum of 20 με ± 4 με occurred (due to volume expansion of released reaction products), and finally stabilized at -133 με ± 3 με after cooling down to 20 °C. Post-treatment processes reduced residual surface strain, with effectiveness varying depending on the parameters (Fig.14). The most significant post-treatment effect was observed at 90 °C for 48 h, where the surface residual strain decreased by 41.8%, followed by 70 °C for 48 h (8.3%) and 70 °C for 24 h (3.1%).

The FBG sensor designed in this study can accurately monitor the annular surface strain of epoxy resin curing molds in real time, exhibiting excellent linear responses to both strain and temperature. The FBG temperature sensor effectively achieves temperature compensation, ensuring monitoring accuracy. The optimization of the epoxy resin curing process based on FBG strain monitoring, particularly the post-treatment at 90 °C for 48 h, significantly reduces the surface residual strain. This study provides a reference for strain monitoring on curved surfaces under different temperature environments and offers a feasible solution for improving the long-term performance of epoxy resin components.