Effect of coupling conditions on ultrasonic echo parameters

Transcription

1 J. Pure Appl. Ultrason. 27 (2005) pp Effect of coupling conditions on ultrasonic echo parameters ASHOK KUMAR, NIDHI GUPTA, REETA GUPTA and YUDHISTHER KUMAR Ultrasonic Standards, National Physical Laboratory, New Delhi In the pulse echo method of ultrasonic non-destructive testing, the signal reflected from opposite face is studied. An analysis of this signal reveals important information about the characteristics of the material through which ultrasonic waves have propagated. Any change in a material characteristic will change signal parameters to varying extent. However, signal parameters can also get affected by the conditions of couplant used between the transducer and material. In the present paper, effects of coupling conditions on signal (echo) parameters have been studied. A detailed analysis is given on various echo parameters such as peak frequency, amplitude, their ratios, etc. for first and second back wall echoes. INTRODUCTION Testing and evaluation of materials is becoming more and more reliable with the availability of better equipment and introduction of new techniques of non-destructive testing 1-2. Ultrasonics has taken a leading role in this direction. Though ultrasonic non-destructive testing has been in use for several decades, research and development work is still active with large scope in this area 3-5. One such work relates to the utilization of hitherto un-utilised information. When ultrasonic pulse travels in any medium, the pulse continuously gets modified. The changes depend upon the characteristics of the medium besides the distance traveled through the medium 6. Since the pulse modifications bear the signatures of material characteristics, it can be said that several of the material characteristics can be evaluated by properly analyzing the pulse and decipher the maximum possible information carried by the pulse. The ultrasonic pulse picked up for such analysis is usually one of the back wall echoes received by the transmitting transducer in receiving mode. The back wall echoes are a result of to and fro multiple reflections between opposite walls of the material where one of the walls is the surface on which ultrasonic transducer is mounted. Ultrasonic transducer is actually coupled to the surface of the material under test through a coupling agent (or couplant). The couplant eliminates the air gap between the transducer and material and provides efficient transmission of ultrasonic waves from the transducer into the material and vice-versa. The thickness of the couplant is kept as small as possible, yet it is sufficient enough to modify the pulse characteristics. The characteristics of ultrasonic pulse are thus modified by the characteristics of the material as well as couplant. In order to evaluate the characteristics of material, it therefore becomes imperative to find out how the pulse characteristics are affected by the couplant. A simplistic approach would be to find which of the parameters of pulse are least affected by the type and conditions of couplant. There are two parameters of an ultrasonic pulse, which are measured in time domain with 70 J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005)

2 fairly good accuracy. One of these is the time of travel. The effect of couplant on the time of travel of ultrasonic waves measured has already been very well studied 7. The other parameter is signal strength or the peak height of half cycle having maximum amplitude, called echo height. The echo height depends upon the coupling conditions very appreciably. In fact, slight variation in pressure on transducer, or the way the transducer is held and scanned by the user, changes the echo height significantly. The appreciable effect of couplant on both of the time domain parameters renders these parameters unsuitable to study small changes in material characteristics. Evaluation of bulk parameters of material or large variation in material characteristics can, of course, be satisfactorily studied by these parameters. In addition to the time domain parameters, the ultrasonic pulse, like any other electrical pulse, can also be characterised by parameters in frequency domain. A study on grain size has shown that some of the frequency parameters are not affected by the coupling conditions 8. It is the objective of this paper to study the effect of coupling conditions on some of the frequency domain parameters with a view to identifying those parameters which are least or never affected by the couplant. Experimental The experimental setup is shown in Fig. 1. A contact type broadband ultrasonic transducer of nominal frequency 5 MHz and 25 mm in diameter was placed over a metallic block. The block of thickness 25.4 mm is made of aluminium alloy 7075 with finely polished surfaces. The transducer is excited by fast rise time, short duration, negative spikes of about 600 V, generated by the ultrasonic flaw detector. The echoes are converted into electrical signal by the transducer and then amplified by the receiving system of flaw detector. The flaw detector sends the signal to a high-speed digitiser. The digitiser used is a dual channel analog to digital converter (ADC) with 8-bit vertical resolution. It normally operates in real time sampling mode where sampling rate is a maximum of 100 MS/s. It can also operate in Random Interleaved Sampling (RIS) mode where the sampling rate can be chosen between 200MS/s and 2.5 GS/s. Such a higher sampling rate is possible only in recurrent waveforms. Since ultrasonic pulse echo testing produces repetitive wave trains, RIS mode becomes a useful function to increase the sampling rate (or reduce the sampling interval down to 0.4 ns) and hence increased sensitivity. The digitiser also provides a built-in 50 ohm terminator which provide accurate voltage measurement at frequencies above 1 MHz (common in ultrasonic testing) due to minimization of transmission line reflections. The signal acquired and sampled by ADC is displayed and controlled on the computer screen by NI software, Scope-SFP, Soft Front Panel developed for high speed Digitisers. The signal is further processed for signal analysis by LabView 7.0 software. Observations From the multiple reflections within the aluminium sample, first two echoes (1BWE and 2BWE) are extracted and processed further. Higher order echoes are not considered, for several other phenomena such as beam spreading start contributing to pulse characteristics which is unwanted. Hence it is better to restrict only to first two echoes and search for the appropriate parameters from these two echoes. In order to study the effect of coupling conditions on these echoes, two different types of couplants were used. One of the couplants used is lubricating oil and the other is grease. These couplants were used in two different conditions. In one of the two cases the transducer was just wriggled over the couplant and left. Observations were taken when the echoes became steady. In another case, a weight was placed over the transducer until the echo height reached a maximum under the pressure of a weight. The ratio of echo heights of 1BWE with and without weight pressure ranges from 7 db to 14 db. For the sake of studying the effect of digitization on the observations, two different sampling rates were selected, one giving sampling interval of 2 ns and the other of 10 ns. All cases studied are summarised in Table 1. RESULTS AND DISCUSSION Fig. 2a and 2b show first two echoes for J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005) 71

3 Table 1. Various cases studied Couplant Conditions Sampling Interval 2 ns Sampling Interval 10 ns Oil without pressure Case 1 Case 5 Oil with pressure Case 2 Case 6 Grease without pressure Case 3 Case 7 Grease with pressure Case 4 Case TEST BLOCK Fig. 1. Schematic diagram of the experimental setup (a) Fig. 2. Waveform and FFT for Case 1: (a) Waveform of 1 st BWE, Waveform of 2 nd BWE, Waveform of 1 st BWE s FFT, Waveform of 2 nd BWE s FFT 72 J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005)

4 Case 1. Fig. 2c shows FFT of first BWE and Fig. 2d FFT of second BWE for coupling conditions of Case 1. Similarly Fig. 3-9 show echoes, FFT of 1BWE and FFT of 2BWE for coupling conditions of Cases 2 to 8. The frequency spectrum can be characterized by several parameters. These parameters may or may not be affected by coupling conditions. An analysis of the effect of coupling conditions on the parameter of a frequency spectrum is given below. Peak Frequency The frequency at which the peak is observed is given in Table 2 for all the coupling conditions and for both 1BWE and 2BWE. Amplitude at Peak Frequency The amplitude at peak frequency is given in Table 3 for all the coupling conditions and for both 1BWE and 2BWE. Amplitude at Other Frequencies Since the peak of the echo is usually flat, it may be better to use a frequency lower or higher than the peak frequency where there is a deep straight line slope to get better sensitivity. The frequency can be arbitrarily chosen. The amplitude of both first back wall as well as second back wall echo is measured at the same frequency on both the sides of peak. The amplitude at lower frequency is given in Table 4 for all the coupling conditions and for both 1BWE and 2BWE. Table 5 gives the amplitude at higher frequency. Ratio of Amplitudes While the amplitude of the echoes may show variation due to coupling conditions, the ratio of 1BWE to 2BWE at a particular frequency may not show any variation. This ratio at peak frequency is given in Table 6 for all the coupling conditions, while the ratios at other frequencies are given in Table 7 and 8. Bandwidth The -6 db bandwidth for both 1BWE and 2BWE for all the coupling conditions is given in Table 9. Lower and Upper Frequencies The lower and upper frequencies, Fl and Fp, are for both 1BWE and 2BWE for all the coupling conditions are given in Table 10 and Table 11, respectively. A study of peak frequency from Table 2 shows that the variation is quite large from 4 MHz to 6 MHz. This is because the frequency spectrum shows more than one peak. The peak frequency is 4 MHz if the first peak is higher than the second peak; otherwise it is at 6 MHz. It is always possible to select either of the two peaks for any coupling conditions. But the first peak also shows variation from 3.67 to 4 MHz and similarly second peak shows variation from 5.33 to 6.33MHz. This means that a shift in peak frequency due to coupling condition cannot be ruled out. This is true for both first as well as second back wall echo. Amplitudes at peak frequency also show considerable variations (Table 3). For first BWE, it varies from to 0.04 and for the second BWE, the variation is from to Though some variation was expected, but a change by an order of magnitude is on higher side. Also, Cases 2, 4, 6 and 8 are almost similar conditions due to high pressure. A variation even in these cases in first BWE from 0.02 to 0.04 and in second BWE from 0.01 to 0.03 shows that this parameter can have some arbitrary value due to coupling conditions. Since significant variation is observed in peak frequency, it was thought worthwhile to analyse the behaviour at other frequencies too. The frequency spectrum of all the echoes shows that 3 MHz on lower side and 8 MHz on upper side may be chosen for further studies. This is because the frequency spectrum is linear with steep slope at these frequencies compared to nearly flattop at peak frequency. The amplitudes at these frequencies for first and second BWE are given in Tables 4 and 5. The variation at 3 MHz in first BWE is from to 0.02 and in second BWE from to The variation at 8 MHz in first and second BWE is from to These variations are very large. It was expected that at higher pressures, the variation will not be much, but the figures of Cases 2 and 4 at 8 MHz and also those of Cases 6 and 8 are against the expectations. It is easy to state that decrease in first BWE J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005) 73

5 (a) Fig. 3. Waveform and FFT for Case 2: (a) Waveform of 1 st BWE, Waveform of 2 nd BWE, (c ) Waveform of 1 st BWE s FFT, Waveform of 2 nd BWE s FFT. (a) Fig.4. Waveform and FFT for Case 3: (a) Waveform of 1 st BWE, Waveform of 2 nd BWE, (c ) Waveform of 1 st BWE s FFT, Waveform of 2 nd BWE s FFT. 74 J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005)

6 (a) Fig. 5. Waveform and FFT for Case 4: (a) Waveform of 1 st BWE, Waveform of 2 nd BWE, Waveform of 1 st BWE s FFT, Waveform of 2 nd BWE s FFT. (a) Fig.6. Waveform and FFT for Case 5: (a) Waveform of 1 st BWE, Waveform of 2 nd BWE, Waveform of 1 st BWE s FFT, Waveform of 2 nd BWE s FFT. J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005) 75

7 (a) Fig. 7. Waveform and FFT for Case 6: (a) Waveform of 1 st BWE, Waveform of 2 nd BWE, Waveform of 1 st BWE s FFT, Waveform of 2 nd BWE s FFT. (a) Fig.8. Waveform and FFT for Case 7: (a) Waveform of 1 st BWE, Waveform of 2 nd BWE, Waveform of 1 st BWE s FFT, Waveform of 2 nd BWE s FFT. 76 J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005)

8 (a) Fig. 9. Waveform and FFT for Case 8: (a) Waveform of 1 st BWE, Waveform of 2 nd BWE, Waveform of 1 st BWE s FFT, Waveform of 2 nd BWE s FFT. Table 2. Peak Frequencies (MHz) for Various coupling conditions Echo No. Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 1BWE BWE Table 3. Amplitude at Peak Frequencies, Ap for Various coupling conditions Echo No. Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 1BWE BWE Table 4. Amplitude at Lower Frequency (3 MHz), Af1 for Various coupling conditions Echo No. Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 1BWE BWE Table 5. Amplitude at Higher Frequency (8 MHz), Af2 for Various coupling conditions Echo No. Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 1BWE BWE J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005) 77

9 Table 6. Ratio of Amplitude at Peak Frequencies, Rp for Various coupling conditions Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case Table 7. Ratio of Amplitude at Lower Frequencies, Rf1 for Various coupling conditions Frequency Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 MHz Table 8. Ratio of Amplitude at Higher Frequencies, Rf2 for Various coupling conditions Frequency Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 MHz Table 9. Bandwidth of 1 BWE and 2 BWE in MHz Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 1BWE BWE Table 10. Lower frequencies Fl of 1BWE and 2BWE in MHz Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 1BWE BWE Table 11. Upper frequencies Fu of 1BWE and 2BWE in MHz Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 Case 7 Case 8 1BWE BWE due to poor coupling should follow a decrease in second BWE to the same extent, and hence the ratio of 1BWE to 2BWE at the same frequency should not show much of variation. Keeping this in mind, ratios at peak frequencies, 3MHz and 8 MHz have been calculated and given in Tables 6, 7, and 8. The variation in ratios at peak frequencies (Table 6) falls between 1.13 and 2.13, at 3 MHz between 1.20 and 2.95 and at 8 MHz between 1.16 and The variation is much less than the parameters discussed earlier where it was 10 times, compared to the variation in ratios where it is only 2 times. The values for Case 2 and 4 at 3 MHz and 8 MHz, however, show much less variation. Another parameter that has been analysed is bandwidth at -6 db along with the lower and upper frequencies (defined as frequencies at which peak value is half). The variation in bandwidth is from 4.20 to 4.80 (15%) leaving aside the outliers of case 6 and 2BWE of case 2. In lower frequency, the variation in 1BWE is from 3.02 to 3.56 (18%) and in 2BWE it is from 3.03 to 3.38 (10%). In higher frequency, the variation in 1BWE is from 7.57 to 7.87 (5%) and in 2BWE, it is from 7.53 to 7.87 (5%). These variations are much less than the variations in earlier parameters. CONCLUDING REMARKS The least affected parameters by coupling conditions are lower frequency, upper frequency and bandwidth. If the coupling is reasonably good, these parameters are almost equal to each other. 78 J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005)

10 Since the minimum variation is observed in bandwidth, lower frequency and upper frequency, these parameters can be taken as the appropriate choice for studying material characteristics. REFERENCES 1. Ashok Kumar, Basant Kumar and Yudhisther Kumar, Evaluating the quality of ultrasonic calibration blocks, Material Evaluation (USA) 55 (1997) Ashok Kumar, Development, characterization and applications of ultrasonic transducers for NDT, Insight - Non-Destructive Testing and Condition Monitoring 45 (2003) V.N. Bindal, Ashok Kumar, Yudhisther Kumar and Jagdish La1, Intensity of ultrasonic beam reflected from solid-solid interface as a function of incident angle, Acoustics Letters 10 (1987) Yudhisther Kumar, Ashok Kumar and Basant Kumar, Effect on ultrasonic propagation in metal rods due to contact with liquid, Acta. Acustica 83 (1997) Ashok Kumar and Yudhisther Kumar, Ultrasonic velocity measurement in cylindrical rod material, J. Pure Appl. Ultrason. 20 (1998) J. Krautkramer, and H. Krautkramer, Ultrasonic Testing of Materials (Springer Verlag, 3 rd Edition) (1983). 7. Ashok Kumar, Correction factor due to couplant in ultrasonic thickness measurement, Insight (UK) 38 (1996) Anish Kumar, T. Jayakumar, P. Palanichamy, and Baldev Raj, Influence of grain size on ultrasonic spectral parameters in AISI type 316 stainless steel, Scripta Materialia. 40 (1999) J. Pure Appl. Ultrason. Vol. 27 No. 2 & 3 (2005) 79