| NDT.net - October 2002, Vol. 7 No.10 |
Digitizing and representation of an A-scan using a graphic display leads to a distortion of the echo shape in comparison to an A-scan on an analogue display (CRT). In addition the A-scan refresh rate is limited by the used screen technology.
New techniques of digitization including interpolation methods, of signal processing, as well as a new way of displaying the signals on a digital screen (“Smart View”or “EchoMax”) in modern Ultrasonic Flaw Detectors are presented, with the aim of reaching the display quality of an ana- logue screen. The operator shall get the feeling of using an analogue instrument without missing the advantages of the modern digital technology.
Still today, analogue instruments show sometimes a better performance on the screen compared to digital instruments, especially with high speed scanning, large dimensions of the test objects and subsequent high gain settings. With the critical applications like forging and casting inspec- tion, the new instrument technology was tested. The results in comparison with the analogue technique will be discussed.
Applications in modern ultrasonic inspection require an optimized visualization of ultrasonic signals and special instrument characteristics to comply with today’s materials, inspection stan- dards and testing techniques, especially with respect to welds, large components like generator shafts or ultrasonically difficult materials and high speed testing, e.g. tubes.
Digitizing and representation of an A-scan using a digital display leads to a change of the echo shape in comparison to the analog CRT display. Modern digital Ultrasonic Flaw Detectors should have the display quality of an analog screen with the advantage of visualizing defects, which are only hit by one ultrasonic shot.
2.1 Analog ultrasonic instruments
With highly sophisticated applications today analog instruments are still used. The most impor-
tant advantage compared to digital flaw detectors is related to the way signals are displayed:
2.2. Digital ultrasonic instruments
Since about 1980 when digital flaw detectors were introduced on to the market their features
have improved continuously. Due to the flat screen like LCD or EL display, user friendly on-
screen functions and the instrument memory, these instruments have many advantages compared
to heavy and bulky analog instruments. However, today still applications are seen, where digital
Flaw Detectors are hardly used: Large work pieces with low attenuation, such as generator shafts,
requiring low pulse repetition frequencies to avoid phantom echoes, and all application at higher
scanning speeds, especially mechanized scanners which require a high pulse repetition frequency.
3.1. Principle of digitization
The analog signal, Fig. 1.a) is divided into equally wide time intervals t, Fig. 1.b). The number of
time intervals per second is called sample rate f. In each interval t =1/f the signal amplitude is
measured, and steps (amplitude resolution) are used to detect the amplitude value, Fig. 1.c). Fi-
nally the signal is described by a series of time/amplitude pairs, Fig. 1.d) which now will be fur-
ther processed. The result in Fig. 1.d) however can be very coarse, due to the sample rate and low
amplitude resolution. Digitization is most easily performed the lower these values are, because a
lot of further data processing is required until the signal can be displayed on the screen. In order
to reach an acceptable signal reconstruction, the sample rate should be approximately 10 times
higher than the ultrasound signal frequency. The amplitude resolution requires 8-10 bit, those are
256 to 1024 steps with 200 to 1000 time intervals.
Fig 1: a) Original b) Time raster c) Amplitude raster d) Digitized signal.
|
3.2. Optimizing techniques
For probe frequencies of 25 MHz, a sample rate of 250 MHz will be required. Electronic compo-
nents for these sample rates are available on the market, but they are expensive and require
power, not suitable for battery driven instruments. But even with lower sampling rates sufficient
good signal digitization can be achieved using multi-phase digitization or using interpolation
techniques (upsampling).
3.2.1. Multi-phase digitization
This technique is based on ultrasonic
signals which do not change signifi-
cantly from one shot to the next shot
(corresponding to one transmit/receive
cycle). This is the case when the pulse
repetition frequency (PRF) is high,
and the scanning speed is low. This
condition is fulfilled with a maximum
scanning speed of 100 mm (4 inches)
per second and a PRF of 1000 Hz. In
this case an ultrasonic pulse is trans-
mitted and received every 0.1 mm.
With multi-phase digitization the sig-
nal is processed in n sequential shots,
whereby the digitizing intervals are shifted by a time distance of
t/n, Fig. 2. The resulting sample
frequency therefore increases to t/n Hz and a complete A-scan will be ready after n cycles or n
phases.
Fig 2: n-phase digitization.
|
3.2.2. Interpolation
Using modern fast signal processing
devices it is now possible to calcu-
late additional data points between
two existing measured points in real
time, Fig. 3. Different ways of calcu-
lating (linear interpolation, best fit,
etc.) can be applied. The advantage
of this technique is that the digitized
signal having sufficient resolution is
available with each ultrasonic cycle.
Even the factor of upsampling (num-
ber of calculated values between two
measured points) is flexible thus
achieving a high signal reproducibility at a comparatively low sampling frequency.
Fig 3: Upsampling by linear interpolation.
|
4.1. Persistence and intensity control of the A-scan
Fig 4: a) Signal on a CRT b) „sparkling“ signal on a digital screen.
|
4.2. Intelligent data display
As described above, a
huge amount of data is
generated with the dig-
itization of ultrasonic
signals. The screen has
a fixed number of pix-
els, and is refreshed
with a certain maxi-
mum frequency, typi-
cally 60 Hz. Therefore
the existing raw data are compressed and only a reduced amount is further processed for the dis-
play. It becomes very important now to select certain A-scans from this data in order to display
the most relevant echoes from the work piece, namely the defect echoes.
The "SmartView" function in the USN 60 and the “EchoMax” function in the VIS system ensure
that, during data reduction for the A-scan, only these will be processed for display which contain
the most relevant echoes, Fig. 5. This guarantees that even at a high scanning speed no defect
echoes stay undetected. For example: If every 100th A-scan would be displayed out of a number
of 6000 A-scans per second (giving 60 A-scans per second), the amplitude of the flaw echo
would not appear constant, because the A-scan randomly processed for the display may come
from the edges of the flaw, or from the middle part, or may not hit the defect at all. “SmartView”
and “EchoMax” compare all 100 A-scans and sends the A-scan with the highest echo to the dis-
play. “SmartView” uses the highest signal in the gate to store the complete corresponding A-scan
and “EchoMax” stores the maximum of the complete A-scan from shot to shot.
Only these techniques guarantee a high flaw delectability at an even high scanning speed, either
with manual inspection or automatic testing systems.
Fig 5: Intelligent data display.
|
5.1. Forging testing
Large forgings, e.g. generator shafts, undergo a
100% ultrasonic inspection, either manually,
Fig. 6, or automatically on specific installa-
tions. The instrument setup requires the range
calibration and sensitivity setting to be made
according to given standards, using defined
calibration blocks. Due to the fact that very
small defects have to be detected, the instru-
ment gain is set to a very high value causing
increased noise indications on the screen. How-
ever, the smallest defects should still exceed
the “grass” by at least 6 dB.
Fig 6: Manual inspection of a generator shaft.
|
Initial use of the instrument on a large genera- tor shaft was very successful. Ultrasonic in- spection could be performed at the same scan- ning dynamic as with analog systems. Even at the very low PRF (25 Hz), which is necessary to avoid phantom echoes, the USN 60 per- formed as good as the analog USIP 11. Fig. 7 shows a group indication in a test shaft for a round robin trial from a distance of about 500 mm. The USN 60 fulfils the demands in forg- ing testing with respect to the dynamic per- formance of the test as well as the high sensi- tivity with sufficient signal-to-noise ratio in order to detect small defects at long distances.
Fig 7: Comparison: USIP 11 versus USN 60.
|
5.2. Tube testing
Fig 8: Once hit defect echo displayed on VIS-System.
|
Now, having the “EchoMax”-function including an “Envelope”-function – this stores each peak amplitude - in a digital system, allows the operator to see echoes on the display even if they are hit only by one ultrasonic pulse, so the adjustment of a system using small test defects gets much easier.
Using the VIS system in “EchoMax”-mode in Fig. 8 a longitudinal, outside flaw on a tube is shown, at a pulse repetition rate of 10 kHz it is stroked once.
The latest digital Ultrasonic Flaw Detectors, like the USN 60 or the VIS system, fully satisfy the demands in testing applications where the visualization of signals during dynamic scanning at high speed and the detection of smallest defects at long distances are required. A large screen with analog performance fully competes with the characteristic of an analog CRT, thus replacing old analog instruments. The display functions like persistence, intensity variation ("Sparkle") and the “SmartView” or “EchoMax” allow an optimal matching of the instrument characteristics to the various ultrasonic applications.
| © NDT.net - info@ndt.net | |Top| |