Share Email Print

Proceedings Paper

Predicting the geometry and location of defects in adhesive and spot-welded lap joints using steady-state thermographic techniques
Author(s): Daniel Turler; Ernest Orlando
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Development of nondestructive evaluation (NDE) methods for spot-welded and adhesive-bonded sheet metal joints is essential for widespread use of lightweight materials and new construction techniques in automotive applications. An important objective of research in progress is development of NDE methods to identify and characterize critical flaws in welded and adhesive-bonded joints. We used steady-state heat- flow and thermographic imaging techniques to test welded and adhesive-bonded lap joints in steel and aluminum samples and in adhesive-bonded composite panels and to identify defective spot welds. The resulting surface-temperature maps or thermograms were used to detect voids and areas where the adhesive was not bonded. To better characterize defects in welds and adhesive layers, algorithms have been developed to post process temperature data, producing more accurate definition of the geometry and location of defects than in previous images. Classic heat-transfer theory was used to calculate the heat-flux equilibrium for each individual pixel on the thermograms. Convective and radiative surface heat- transfer coefficients were applied to compensate for the heat exchange between the sample and the environment. This post processing permits us to determine the locations of spot welds and the sizes of the weld nuggets in welded joints, and to clearly image voids in adhesive layers between joints. The effectiveness of the image-processing algorithms was investigated using data from laboratory experiments on test specimens with flaws of known size and location. In addition, the images of the defects produced with the new method were compared to results of two-dimensional heat transfer simulations through the same samples. The simulations were also used to determine boundary conditions for post-processing of images.

Paper Details

Date Published: 19 March 1999
PDF: 9 pages
Proc. SPIE 3700, Thermosense XXI, (19 March 1999); doi: 10.1117/12.342333
Show Author Affiliations
Daniel Turler, Lawrence Berkeley National Lab. (United States)
Ernest Orlando, Lawrence Berkeley National Lab. (United States)

Published in SPIE Proceedings Vol. 3700:
Thermosense XXI
Dennis H. LeMieux; John R. Snell, Editor(s)

© SPIE. Terms of Use
Back to Top