Instrumentation & Measurement Magazine 24-4 - 59

Infrared Thermal Detection of
Ice Defects inside Honeycomb
Sandwich Skin
Qingying Li, Baolei Li, Hairong Xu, and Tian Bai
I
nfrared thermal wave detection, scanning and recording
of temperature responses in the form of digital
signals is considered a powerful supplement to traditional
nondestructive testing (NDT) methods. It offers a range
of advantages, such as a large testing area, a wide range of applications,
a fast testing speed, the fact that there is no need to
make contact with the tested objects, and the possibility of presenting
test results in a pictorial form [1], [2]. In relation to the
interests of this paper, it has been applied to detect ice in the
field of aeronautics [3], in marine structures [4] and in wind
turbine blades [5].
The quality of infrared thermography is affected by a number
of factors, including infrared thermal excitation, and the
effectiveness of approaches such as thermographic signal
reconstruction (TSR) and thermal image processing. In its application,
an external heat source excites samples to highlight
the target, which improves the detection accuracy and ensures
the detection efficiency. This approach is suitable because there
is usually no obvious natural temperature difference between
targets and their background when testing samples. However,
the targets may not be detected if the temperature difference
cannot be monitored effectively for nonuniform heat distribution.
Another issue is that the original infrared thermal
sequence data acquired by the infrared camera represents the
temperature change of the sample over time. This is because
the transmission rate changes according to a sample's different
thermophysical properties when thermal excitation is applied
and accompanied by the release of heat waves. The signal data
therefore needs to be reconstructed, and because it is noisy, it
occupies a greater amount of storage space, and the contrast
between the target and the background can be insignificant [6].
A further concern is that after processing the sequence data,
the edge of the infrared thermal images may remain blurred
and polluted by noise, making a detected defective area too
large or too small. There is therefore a need for more effective
image processing methods that can improve upon the quantitative
calculation of a target range.
In this study, three kinds of thermal excitation devices
were assessed to determine the one best suited to generating
June 2021
a uniform distribution for high-quality infrared thermography.
To do this, samples containing ice accretion defects inside
a honeycomb sandwich skin were constructed to be detected
by an infrared thermography system. Ice accretion not only
affects the skin's performance [7] but can also interfere with
external detection, so this factor needs attention when undertaking
ice damage detection. A TSR technique using an
optimum order polynomial was used to facilitate a qualitative
judgment of the ice defects. In addition, the defective area was
quantitatively resolved using an image processing approach
that combined an adaptive Canny edge detection algorithm
and a region growing technique.
Thermal Excitation Devices
Common infrared excitation methods include photothermal
radiation excitation, ultrasonic excitation and electrical excitation.
However, the non-uniformity caused by the location
of the photothermal radiation excitation device can interfere
with signals and make testing more difficult. The instability
of the contact between ultrasonic excitation devices and the
samples being tested can also affect the measurement results
[8], and electromagnetic thermal excitation is only effective for
conductive objects [9]. In view of the structural characteristics
of honeycomb sandwich skin and the nature of ice accretion
Fig. 1. Infrared thermal excitation devices. (a) Hot air gun; (b) 2 × 2 matrix
array of xenon flash lamps.
IEEE Instrumentation & Measurement Magazine
1094-6969/21/$25.00©2021IEEE
59
Instrumentation & Measurement Magazine 24-4

Table of Contents for the Digital Edition of Instrumentation & Measurement Magazine 24-4
