MAGNETIC PRESSURE STRESSING OF LAP JOINTS - MODELING AND EXPERIMENTAL-VERIFICATION

Citation
Dj. Mayton et al., MAGNETIC PRESSURE STRESSING OF LAP JOINTS - MODELING AND EXPERIMENTAL-VERIFICATION, Journal of nondestructive evaluation, 16(1), 1997, pp. 11-20
Citations number
13
Categorie Soggetti
Materials Science, Characterization & Testing
ISSN journal
01959298
Volume
16
Issue
1
Year of publication
1997
Pages
11 - 20
Database
ISI
SICI code
0195-9298(1997)16:1<11:MPSOLJ>2.0.ZU;2-5
Abstract
A novel electromagnetic stressing/optical detection method has been de veloped in response to the need for better nondestructive evaluation t echniques for the detection of disbonds in aging aircraft lap joints. This technique uses magnetic pressure to pull the top surface of a thi n conductive bonded sheet and senses the out-of-plane displacement of the surface with an optical lever fiber bundle detector. This method o f inspection has the advantages of being noncontacting, relatively ine xpensive, and because it pulls on the top surface, is a promising cand idate for the detection of ''kissing'' disbonds-a condition in which t here is no material missing from the joint, but the bond has failed. A series of three models was developed and implemented to simulate syst em performance from the driving circuitry to the measured response of the sample. Using a computer model of the driving circuit, component v alue variations could be analyzed to optimize the current through the electromagnetic coil as a function of time. An analytical pressure mod el was developed to predict pressure on the sample as a function of ti me for a given current waveform input. The predicted pressure was then used as the driving function in. finite element structural model whic h predicted displacement of the sample surface. Laboratory experiments were conducted on simple bonded and unbonded samples, and the mio cas es exhibited large differences in amplitude, resonant frequency, and d amping. Test results compared favorably to the predicted displacement data. The close correspondence between measured and predicted results indicates that the models are useful not only in the system design but also as a means to predict performance.