In the past few decades, there has been much attention paid towards carbon fiber reinforced plastics (CFRPs) as a structural material for various applications, such as aerospace, transport, infrastructure, wind energy and so on, because of superior mechanical, electrical, thermal, and chemical characteristics compared to steel or metal. However, they are practically used as a complicated architecture including laminates of different ply orientations and woven fabrics, due to the improvement in an anisotropic mechanical strength. Thus, when the CFRP laminates was used in real products with a complicated architecture, it is difficult to investigate and predict the stress/strain distributions and rupture models. The development of in-situ evaluation techniques capable of providing a stress/strain distribution has been strongly demanded. Recently, we have reported that the mechanoluminescence (ML) techniques using composite films consisting of an ML material and organic compounds can be useful technique for visualizing the dynamic equivalent strain distributions in complicated architectures. The ML is the luminescence phenomenon induced by mechanical actions, such as compression, tension, friction, or torsion. Especially, the ML intensity is proportional to the strain energy in the elastic region. In the present study, therefore, we examined the feasibility of a non-destructive testing (NDT) for carbon fiber reinforced plastic (CFRP) laminates by using a mechanoluminescence (ML) technique. Specifically, we investigated to visualize the relative strain distribution of the CFRP laminates with bidirectional fiber bundles (twill woven) by using the fabricated ML sensor consisting of a mixture of SrAl2O4:Eu (SAOE) ML material, which shows the highest ML intensity among various materials reported, and epoxy resin attached to the surface of CFRP laminates. For comparison, the CFRP with unidirectional fiber bundles was also investigated. An ML sensor was fabricated by using a composite of SAOE powder and epoxy resin. The SAOE powder was prepared by using a solid-state reaction method. SrCO3, α-Al2O3, Eu2O3, and small amount of Ho2O3 powders were mixed in an agate mortar with ethanol solution. The obtained mixture powder was calcined at 800ºC for 1 h in air, and then sintered at 1350ºC for 4 h in reducing atmosphere (H2+Ar). The sensing characteristics of the ML sensors painted on the surface of the CFRP laminates were evaluated under a tensile testing in a dark room at room temperature. The commercial strain gage was also applied on the opposite side of CFRP laminates with a commercial adhesive. Tensile tests were performed with a material testing machine. The ML images of the ML sensor and longitudinal strains during the tensile tests were recorded by using a CCD camera and a uniaxial strain gage connected a data logger, respectively. Both devices were connected to a personal computer to simultaneously obtain the data. For quantitative and reproducible measurement, the ML sensor was once irradiated by blue light-emitting diode (LED) for 1 min and kept under dark condition for 5 min. From recording the ML images in the tensile tests, the ML sensor attached to the CFRP laminates with unidirectional fiber bundles gave uniform ML intensity distribution over the whole of ML sensor, and the ML intensity increased with increasing the tensile load, indicating that the whole of CFRP laminate deforms uniformly by the applying tensile loads. On the other hand, the CFRP laminates with bidirectional fiber bundles exhibited a twill weave ML intensity distribution similar to the original twill weave pattern. Specifically, the ML intensities on the areas of a longitudinal fiber bundle (tensile direction) were found to be higher than those on transverse fiber bundle areas. This indicates that a longitudinal fiber bundle deform much larger than the transverse ones. Therefore, it is concluded that ML sensor could directly visualize the load share between fiber bundles with different directions in CFRP laminate with high spatial resolution. Based on the obtained results in this study, we concluded that the ML sensor could visualize a relative strain distribution of CFRP laminates with both unidirectional and bidirectional fiber bundles. Additionally, the ML sensor could detect a tiny strain difference at the interface between longitudinal fiber bundles even in the elastic deformation region. Thus, it is considered that the ML sensor could be a possible candidate for a dynamic strain distribution technique of the CFRP products.
Read full abstract