Chang Sun Hong

Simultaneous sensing of the strain and points of failure in composite beams with an embedded fiber optic Michelson sensor

Abstract A fiber optic Michelson sensor was embedded in composite beams to sense the internal strain and points of failure of the composite structures. The bending deformation and matrix cracking were investigated by four-point bending tests of cross-ply composite beams with the embedded fiber optic sensor. The failure points of composite beams were detected by using both a PZT sensor and a fiber optic sensor in order to investigate the fiber optic failure signals. The failure due to matrix cracks in a composite beam was confirmed by the edge replica method. The digital processing of the fiber optic signal was carried out to determine the strains and failure points of composite beams. The failure points were observed from the processed failure signal by high-pass filtering. The initial failure strain of the composite beam was measured and processed from the fiber optic strain signal after low-pass filtering.

Development of fibre optic ingress/egress methods for smart composite structures

The fibre optic ingress/egress is one of the important issues for the application of fibre optic sensors to real structures because the optical fibre is so fragile at the ingress/egress point. Little attention has been paid to methods of practically accessing embedded optical fibres in the literature of sensor activities. In this paper, we propose the novel out-of-plane ingress/egress method to improve the limit of the existing in-plane ingress/egress method. For the safety of the embedded optical fibre, we developed a component to protect the optical fibre at the ingress/egress point and a preservation method for an optical fibre against the composite curing environment of high temperature and pressure. We also developed an optical connection fixture to integrate the connection part and a composite structure. Some tests were performed to confirm the safety and survivability of the embedded optical fibre using the devised ingress/egress method.

A digital signal processing algorithm for structural strain measurement by a 3 × 3 passive demodulated fiber optic interferometric sensor

A digital signal processing algorithm, the moving reference line crossing count method (MRLCCM), was developed to implement the structural strain measurement by a passive demodulated fiber optic Michelson interferometric sensor. The fiber optic Michelson sensor, which is constructed of a 3 × 3 fiber optic coupler, can give information about the magnitude and direction of the strain of structures. The beating, drifting and noise, which are caused by the longitudinal strain and the lateral strain of the fiber, bring about the counting error of the phase differences which are directly converted to the structural strain. The developed algorithm is based on the reference line crossing count method and resets the reference line during the presence of the signal drifting. The accuracy of the strain calculation was confirmed by the various simulated fiber optic signals with signal beating, drifting and noise. A passive demodulated 3 × 3 fiber optic Michelson interferometric sensor was bonded on the cantilevered aluminum beam for experimental strain sensing. The capability of real-time processing was verified by the real fiber optic signals.

Mechanical strength characteristics of fiber bragg gratings considering fabrication process and reflectivity

Application fields of fiber Bragg gratings (FBGs) have gradually expanded to include primary structures as well as secondary structures and specimens. For this reason, verification for FBG reliability, such as signal characteristics and mechanical strength, has become increasingly important. In this study, the mechanical strength characteristics of FBGs of various grating lengths are investigated considering their fabrication process and reflectivity. Tension tests show that the mechanical strength of optical fibers decreases about 50% just by jacket stripping and that the amount of decrease is dependent on the stripping methods. The mechanical strength of the stripped optical fibers with gratings formed in the core decreases 55% regardless of grating lengths or reflectivity. However, the width of strength distribution increases relative to increases in reflectivity.

Stochastic Finite Element Method for Laminated Composite Structures

The stochastic finite element method based on second moment method is developed for the probabilistic analysis of laminated composite structures in which physical properties have uncertainties and variability. The material properties, fiber angles, laminate thickness, and curvatures are modeled as random parameters. The first order shear deformable shell element is employed in application. Numerical results are presented for three example problems. The validity and accuracy of the method are verified through comparison of the present results with those obtained by the Monte Carlo simulation. Statistics of strains and stresses are derived very accurately by considering the randomness of the strain-displacement relation and principal material directions. The probability distribution of structural responses of composite structures can be assumed to be normal distribution, and therefore, can be evaluated quantitatively.

Structural Analysis and Strain Monitoring of the Filament Wound Motor Case

Filament wound structures such as pressure tanks, pipes and motor cases of rockets are widely used in aerospace applications. The determination of a proper winding angle and thickness is very important to decrease manufacturing difficulties and to increase structural efficiency. In this study, possible winding angles considering the slippage between a fiber and a mandrel surface are calculated using the semi-geodesic path equation. In addition, finite element analyses using ABAQUS are performed to predict the behavior of filament wound structures considering continuous change of the winding angle and thickness at the dome part due to fiber built-up near the metallic boss. Water-pressuring tests of a third stage motor case are performed to verify the analysis procedure. The strain gages are attached on the surface in the fiber direction. Progressive failure analysis predicted the burst pressure a little bit higher than the experiment and the weakest region of the motor case very well. The effect of reinforcement is also studied to increase its performance.

Buckling and Postbuckling Behavior of Stiffened Composite Panels Loaded in Compression

A modified arc length method combined with the progressive failure model was used in the nonlinear finite element method. The finite element program was applied to stiffened composite panels under axial compression, and buckling and postbuckling behavior of the panels was characterized. The web and the lower part of the cap of stiffener were favricated by the continuous plies of the skin for cocuring the stiffened panel

Strain and failure sensing by the fiber optic Michelson sensor embedded in composite beam

The bending deformation and matrix cracking were investigated by conducting a four-point bending test for a cross-ply composite beam with an embedded fiber optic Michelson sensor. The fiber optic Michelson interferometric sensor was constructed and embedded in the composite beam. The failure of composite beam, due to the matrix cracking, was successfully detected by the fiber optic sensor and the matrix crack in the composite beam was confirmed by an edge replica method. The characteristics of the failure signals from the fiber optic sensor were studied. The strain and failure signals of the composite beam were separated by digital filtering of the signal from the fiber optic sensor. The failure was obviously detectable by the failure signal filtered from the optical signal.

Bimodal bound of system reliability for random composite structures

A procedure of system reliability analysis using the stochastic finite element method for composite structures is presented. Composite structure is considered as a random structure. The material properties, fiber angles, laminate thickness, and curvatures are modeled as random parameters. The curvature means initial curvature as a result of manufacturing error for the case of a flat plate. The random thickness and fiber angles affect the straindisplacement relation and material directions. In the stochastic finite element formulation, these randomnesses are considered to obtain accurate statistical responses. In reliability analysis, composite laminates are assumed to be multicomponent series system, and the correlations between possible failure modes of different components are considered to obtain bimodal bounds of failure probability. The method is verified through comparison of the results with those obtained by the Monte Carlo simulation for two example problems. The developed method is efficient and accurate enough to be applied to engineering problems and can assess the risk of structures quantitatively and logically.