FIELD OF INVENTION:
The present invention relates to a system and method for non destructive characterization of solids and composites particularly for detection of defects using thermography.
PRIOR ART:
Non-Destructive Testing (NDT) plays an important role in performing quality assurance as well as in materials characterization. There are several methods that are available in non-destructive testing and evaluation of materials such as Thermal Wave Imaging (TWI), Radiographic Testing (RT), optical testing such as Holographic Non-Destructive Testing (HNDT), Ultrasonic Testing (UT) and Eddy current Testing (ET). TWI rates one of the key techniques available. This is because TWI is a non-destructive, non-contact and whole field technique for defect detection capabilities. TWI can be potentially useful for defect detection in wide variety of materials such as metals, composites and semiconductors. This is because all solids conduct heat.
The principle of thermal NDT involves mapping the surface temperature of the sample for detecting the surface and subsurface features (ex. Voids, disbands, cracks etc.) of the
object. TNDT can be broadly divided into active and passive methods. In passive techniques, defect shows up if its inherent temperature is different from its surroundings. However detection of defects below the surface of the test sample, especially if they are deep, is difficult with passive thermography, and active thermography is preferred. In active thermography an external stimulus is applied to view the deeper defects with higher contrast.
Presently, two different approaches are possible in active thermography : Pulse thermography (PT) and lock-in (modulated) thermography (LT). In PT, the examined material is warmed up with a short duration energy pulse (light, eddy current, or ultrasonic pulse) and the thermal response is recorded. The resultant sequence of infrared images indicates defects in the material at different depths. The method requires high-power heat sources and has the additional drawback of being sensitive to surface in-homogeneities. In contrast, lock-in thermography uses mono-frequency sinusoidal thermal excitation, and information about the phase and magnitude of the reflected thermal wave is derived from the recorded images. The phase angle has the advantage of being insensitive to local variations of illumination or of surface emissivity. However the mono-frequency excitation in LT can form standing wave patterns, leading to misinterpretations. The time taken.
to take measurements over a span of frequencies, can also be very large in LT.
OBJECTS OF THE INNOVATION :
1) One object of the present invention is to have found a method of defect detection which does not require high powered heat sources.
2) Another object of the present invention is to have found a method of defect detection which is not sensitive to surface hetrogenities
3) Yet another object of the present invention is to have a method of detection which is less prone to misinterpretation of results.
4) Another object is to have a method which takes lesser or minimum amount of time for analyzing and interpreting the data for having good information about the defects.
5) Yet another object of the present invention is to allow for the transmission of a low- peak-power, long-duration modulated wave inorder to provide detection range and resolution comparable or better to that achieved by short duration, high peak-power pulse methods.
6) The primary objective of this study was to present a new technique for defect detection based on frequency modulated thermal wave imaging.
7) Yet another object of this invention is to yields a finer range resolution than can be achieved with conventional thermographic methods.
8) Another object of the present invention is to be able to probe deeper in solids and composites for defect detection therin
BRIEF DESCRIPTION OF THE FIGURES
FIG. 1 Frequency-modulated (chirp) heating incident on sample surface.
Fig. 2 Fast Fourier Transform (FFT) of the waveform in Fig. 1
Fig. 3 FMTWI simulations showing the frequency response of reflected thermal waves from different depths. The expected attenuation of higher frequency components with increasing depth is clearly illustrated.
Fig. 4 Experimental arrangement for FMTWI .using Pulse compression for characterization of solids and composites.
Fig. 5 Typical temperature profile on the sample surface
Fig. 6 Principle of pulse compression.
Fig. 7 Temperature profile on the sample surface
after removing the average Increase in temperature.
Fig. 8 Compressed pulse obtained by the matched filter
DESCRIPTION OF THE INVENTION;
The present invention relates to a system and method for non- destructive characterization in solids. More preferably the present system relates to detection of surface defects in solids or composite materials.
The present invention uses frequency modulated thermal waves. In this method which may be called 'Frequency-Modulated Thermal Wave Imaging (FMTWI)', the surface heating is not at a single frequency (as in LT), or at all frequencies (as in PT), but in a limited range of frequencies. The frequency range is decided by the sample
characteristics such as thermal diffusivity, density, thermal conductivity and thickness/defect depth.
FMTWI experiments were carried out on a mild-steel sample using a CEDIP IR system. A frequency modulated signal of about 169 sec duration and linear frequency variation from 0.04 Hz to 0.09 Hz, generated from a signal source is used to drive the heat sources via a source control unit. The resultant temperature change over the sample surface was temporally captured in ALTAIR software.
Measurements were made over one frequency-modulated cycle (0.04 to 0.09 Hz in 169 seconds)^ It was found that FMTWI can scan the entire sample thickness by utilizing thermal waves, whose diffusion length changes with time (eq.(2)), depending on the appropriate frequency modulated surface heating.
In FMTWI the frequency varies with time causing variable depth probing. In this regard it can be compared to Pulse Phase thermography in which a comparatively much wider range of frequencies are probed, simultaneously.
The frequency dependent thermal diffusion length characteristics of the present invention determines spatial
resolution_This is in contrast to a fixed test frequency (LT) where the thermal diffusion length gets fixed and limits the depth resolution of the test. Further as compared to PT, considerably less peak power is required from the heat sources.
The present invention uses pulse compression technique which allows for the transmission of a low peak-power, long-duration modulated wave. This provides detection range and resolution comparable or better to that achieved by short duration, high peak-power pulse methods.
In the present invention the pulse compression technique has been applied for the first time to thermal nondestructive characterization. This results in widening of the bandwidth of the transmitted pulse by modulating it in either frequency or phase which yields a finer range resolution than can be achieved with conventional thermographic methods.
Figure 5 shows the principle of pulse compression technique. The received waveform reflected from the target is processed using some variant of a filter matched to the transmitted signal, compressing the long pulse to a duration 1/B where B is the bandwidth of the transmitted waveform.
Let s(t) be the reflected thermal wave from a defect. It can be considered to be the same as the transmitted wave except for a delay and a reduction in amplitude. In a matched filter, the time response of the filter is matched to the transmitted waveform, i.e. its impulse response is the time reverse. Therefore, if s(t) is a linear up-chirp waveform (frequency increasing with time), then its matched filter will be a linear down chirp (frequency decreasing with time). When the signal s(t) goes through the matched filter, the output waveform (figure 5), is a narrow correlation peak called a compressed pulse, and can be represented as the convolution of the s(t) with h(t) as follows
Under the present method , the depth of the defect can be conceptually estimated in two ways; by measuring the magnitude of the reflected thermal wave from the defect, and secondly by measuring the time delay of the reflected thermal wave with respect to a reference. Estimation from the former can be misleading because of the presence of surface in-homogeneities and non-uniform heating of the sample surface. Therefore a correlation approach based on the complete reflected thermal signal has been considered.
For the experimental data, the reflected thermal wave from 0.1 mm depth is taken as the reference. The image
sequence was captured for the same duration of 169 seconds but at a frame rate of 20 Hz. Taking the sequence of pixel intensity values at a defect location throughout the movie provides a temporal thermal profile. Before doing correlation between the reference thermal profile and the thermal profile at the defect regions, the average increase in temperature (dc part) has been removed by linear curve fitting. Correlations of the reflected thermal waves from various depths, with respect to the chosen reference (signal from 0.1 mm defect), were obtained. The time delays for various defects are calculated from the observed time shifts of their correlation peaks. Figure 6(b) shows the correlation peaks obtained for 1mm, 2.1mm and 4.2 mm defects with respect to the reference. It may be noted that since the experimental image sequence was captured at 20 Hz frame rate, while the simulation was carried out at 108 Hz sampling frequency, causes the difference in their x-axes (figure 6(a) and 6(b)).
For the purpose of easy comparison, the time shifts in seconds of peaks due to defects at 2.0 and 4.1 mm depths, are given in Table 1. The agreement between prediction and experiment is good. The time shift for the 1 mm defect was easily discernible in simulations, but is not clearly visible in experimental correlation peaks because of the low image sequence capturing rates (20 Hz). The concept of defect detection by shift in correlation peaks for FMTWI technique
is thus introduced and illustrated. A rigorous quantitative analysis of the relationship between shifts in the correlation peaks as a function of defect depth is underway.
detection.
It may be concluded the comparative tests provede conclusively that FMTWI is able to probe deeper into the composites sample.
It was further concluded that the capability to detect deeper defects using the present hardware (camera and pulse excitation source) can be achieved by going in for Pulse Phase Thermography.
Further characterizing new composite samples with the developed FMTWI technique can be applied to some of the samples viz. (i) GFRP samples, (2) CFRP for depth/diameter studies, (3) Bi-directional CFRP samples and so on.
A need was also felt for appropriate models to predict thermal response of composite samples to various stimuli, which would help to optimize the inspection parameters for composite materials.
with respect to the incident thermal wave, at a sampling frequency of 108 Hz.
EXAMPLE 1
Figure 6(a) shows the correlated peaks obtained by MATLAB Simulink simulation, for the mild steel sample under test. Only 51.53 s of the central portion of the entire 338 s (= 2 X 169 s) correlated waveform is shown along with an inset zooming in on the correlation/compressed pulse peaks for various defects. In this simulation the correlation is carried out for the reflected thermal waves
EXAMPLE 2
CFRP composite sample 250 mm x 150 mm x 4.5 mm 0/90 configuration, 30 layers each of 1.15 thickness, 15 numbers of Teflon inserts (15mm x 15mm x 50 microns), with the first kept below layer 2, and thereafter below 4, 6, 8,..., 30 layers.
Results;
Pulse hermography: Presently ableto detect the 7 defects
upto 14 layers (i.e. 2.1 mm)
FMTWI: Able to detect all 15 defects upto 30 layers (i.e. 4.5 mm)
Conclusion: The comparative tests provede conclusively that FMTWI is able to probe deeper into the composites sample.

WE CLAIM
1) A system for non-destructive characterization of solids and composites
more preferably a system for defect detection in solids and composites comprising
of
one or more than one sources of thermal waves, means for driving the heat source, means for generating frequency or phase modulated chirp signals, means for controlling the level of thermal radiation.
a thermographic camera (1) for receiving the thermal radiation returning from the solid or composite Central processing facility (2) means for processing analyzing the received waveforms wherein the received waveform is filter matched with a variant of transmitted signal compressing the long pulse to a duration 1/B where B is the bandwidth of the transmitted waveform (Fig. - 4)
2) A system as claimed in Claim 1 wherein the said processing facility (2) produces a fourier transform of the incident and the returned waves which is performed on the entire duration of the frequency or phase modulator signals (Fig. - 2).
3) A system as claimed in Clam 1 and 2 wherein the fourier transform produces a spectrum with equal energies at all frequencies (Fig. -3).
4) A system as claimed in Claim 1 wherein the representation obtained by the said means clearly shows the attenuation of the higher frequency components with increase in depth of the solid
5) A system as claimed in claim 1 wherein the camera and the said means for processing and analyzing the returned signals enable deeper probing and defect detection in composites.
6) A method for non-destructive characterization particularly a method for
detection of defects in solids comprising the steps of-
using one or more heat sources driven by a function generator
through a source control unit
coating the sample surface with a candle root
capturing the temporal profile of the sample through a CCD camera
preferably an IR camera
curve fitting of the said temperature temporal profiles to remove the
mean increase in temperature
filter matching preferably through correlation method between the
temperature profiles of the defective & non-defective regions in
order to obtain a compressed pulse (Fig. - 5).
measuring the magnitude of the reflected thermal wave from the
defect
measuring the time delay of the reflected thermal waves with respect
to a reference
7) A system for thermal non-destructive characterization of solids and composites substantially as herein described with reference to the accompanying figures.
8) A method for thermal non-destructive characterization of solids and composites substantially as herein described with reference to the accompanying figures.