·Table of Contents
·Methods and Instrumentation
Novel Solid Contact Ultrasonic Couplants Based on Hydrophilic PolymersS. Bourne - Sonatest PLC 'United Kingdom; M. Newborough, D. Highgate - Cranfield University 'UK
The feasibility of employing hydrophilic polymers as novel, solid contact ultrasonic couplants for this application has been investigated. Hydrophilic materials absorb large quantities of water, and become soft and flexible, yet they maintain hydration under mechanical pressure. Many of their ultrasonic properties have been explored, including acoustic impedance, ability to transmit high frequencies and the coupling efficiency to a smooth steel surface. This work has culminated in a prototype dry coupled wheel probe capable of working at > 5 MHz in pulse echo mode and providing excellent resolution while eliminating the risk of test piece contamination.
The type of couplant used depends on the application and the nature of the test piece. Ultrasonic applications can be broadly broken down into two main categories, manual testing and automated testing.
Manual testing encompasses thickness metering, condition monitoring and weld testing and would typically involve an operator taking point measurements with a portable ultrasonic unit.
Historically, oil was one of the most commonly used couplants for such tests. Nowadays, there are extensive ranges of oil and water based gel-type couplants commercially available.
One of the most commonly used automated techniques is immersion testing whereby the probe(s) is scanned over the test piece via computer control. This involves submerging the test piece and probe(s) in a tank of water. Hence, a continual water path is established between the probe and test piece, along which the ultrasound propagates. Thus, water is the coupling media. This technique maintains the most constant level of coupling achievable, providing consistent sensitivity.
Often, the presence of such couplants is inconvenient as it places many limitations on the testing that may be performed. In the case of manual testing, the application and removal of the couplant is time consuming and therefore becomes an expensive component of the testing process. Another common problem is loss of couplant from the probe face, which causes the signal amplitude to vary making interpretation difficult. This is most likely to occur when a degree of scanning is required of the probe and often masks smaller, important signal fluctuations. This problem is most pronounced when the test piece surface is rough and scanning speed is high. Perhaps a more universal limitation is the issue of contamination. This is of vital importance for both manual and automated testing as many materials that require ultrasonic inspection are sensitive to liquid couplants. Examples include, open cell structures, porous materials and many metals and composites which suffer unacceptable contamination through prolonged contact with liquid couplants. Obviously, this is most critical for components that require immersion testing.
Many attempts have been made to overcome these limitations in the form of non-contact techniques. These include air coupled ultrasound, laser generation and detection systems and electromagnetic acoustic transducers (EMATS). These are summarised by Billson and Hutchinsm  and Drinkwater and Cawley [2-3]. However, due to the limitations of these approaches, the vast majority of cases make use of liquid coupling media.
Another possible way to overcome the need for a liquid couplant is the employment of a solid coupling media. Transducers of this type commonly use a soft rubber between the transducer and test piece, which conforms around surface undulations of the test surface when in pressed contact. There are two common designs for such a transducer; a static probe whereby a rubber tip is applied to the face of a conventional contact transducer, and a wheel probe typically consisting of a transducer crystal at the hub of the wheel which is free to roll over the test piece, coupling being provided by a rubber tyre. The wheel probe design is very convenient for the scanning of large areas of the test piece. Ideally, such a device would emulate an immersion test enabling use with existing equipment and encouraging ready interpretation of the signals generated.
Solid coupled transducers have been available for some time  but their usefulness has been hampered by the materials selection. For example, natural rubber has been commonly used for solid coupling applications and although it has the ability to conform to an uneven test surface reasonably well, it imparts high attenuation to ultrasound, often forcing undesirable designs to be employed. For example, in an effort to improve signal to noise, the transducers were low frequency (typically below 1-2 MHz) and often left undamped. This resulted in a long ringing signal, which reduced the resolution to such an extent that pulse echo operation was not possible.
More recently, Billson and Hutchins  described the use of a static, solid coupled longitudinal wave probe operating at a centre frequency of 5 MHz using a new low loss, synthetic rubber coupling medium. Drinkwater and Cawley  reported a similar static, solid coupled longitudinal wave probe operating at a centre frequency of 7 MHz using another low loss rubber. Both devices worked in pulse echo mode. Such new coupling materials gave scope for higher frequency wheel probes such as the one described by Drinkwater and Cawley [2-3] which operated at a centre frequency of 3.8 MHz. This device was capable of operating in pulse echo mode and showed great promise for use in generating C-scan data without risk of test piece contamination. The main breakthrough here was the new solid coupling material. This enabled a shift in design philosophy away from that of low frequency, un-damped, through transmission devices which used a thin layer of highly attenuative solid couplant, towards higher frequency, highly damped, pulse echo devices where the solid couplant was used as a delay line.
This paper will introduce hydrophilic polymers to the field of ultrasonic NDT. In particular, the ultrasonic properties of a range of such materials will be discussed with a view to the development of a new solid coupled wheel probe. Finally, results will be presented demonstrating the suitability of a prototype wheel probe to a number of applications.
Hydrophilic polymers are available in both homopolymer and copolymer forms. Homopolymers are single molecular species and are restricted to relatively low water uptake. Such a material is typified by HEMA, which is limited to absorbing 38% water by wet weight. Hydrophilic copolymers are essentially made up of two monomer constituents - hydrophilic and hydrophobic. The hydrophobic part (e.g. PMMA) provides the long-term structure of the final material whereas the hydrophilic part provides hydration sites (e.g. OH or N). It is to these sites that water bonds ionically. In addition, a small amount of free water enters some tiny voids opened upon expansion of the polymer. The amount of water absorbed by a hydrophilic copolymer is dictated by the ratio of hydrophilic to hydrophobic components.
Complete hydration is reached after a period of time depending on the final water uptake capacity of the polymer together with the thickness of the polymer. Water advances through the polymer as a front leaving two clearly defined regions (hydrated and dehydrated) . The water acts as a plasticiser causing the polymer to expand and become flexible. As the polymer is cross-linked it does not dissolve when hydrated. The cross-links also guarantee a predictable expansion ratio.
An asset to hydrophilic polymer technology is the associated chemical flexibility. As the quantity of the hydrophilic component of a copolymer may be easily controlled, the ultimate water content on full hydration may be accurately defined. Similarly, the hydrophobic (or backbone) component can comprise of a number of different polymers providing alternative mechanical properties for the final hydrated material. Conventional hydrophilic polymers have low notch-tear strength. Suitable choice of monomer system and polymer fabrication method can dramatically improve tear strength and therefore the longevity in arduous environments [9&10].
Conventional hydrophilic polymers have been used extensively in medicine and pharmaceutics owing to the body's readiness to accept them as a result of their high water contents. Indeed, such materials may be more commonly recognised as those used for soft contact lenses . This project marks successful progression and transfer of this technology to the NDT arena and has resulted in patent application covering this wide area of application.
Where WH is the hydrated weight of the polymer and WD is the dehydrated weight of the polymer.
|Sample||Density (kg m-3)||Ultrasonic Velocity (ms-1)||Acoustic Impedance (Nsm-3)x106|
|Table 1: Acoustic impedance of range of hydrated hydrophilic polymers.|
Acoustic impedance is the product of the density of a material and the velocity at which ultrasound passes through it. Consider two materials in pressed contact, achieving a theoretical 100% surface contact at the interface. If ultrasound is propagated perpendicularly to the interface, a portion of the ultrasound will be reflected back into the first material and a portion transmitted across the interface into the second material. The amounts of ultrasound transmitted and reflected are dictated by the acoustic impedances of the two materials. The smaller the mismatch, the greater the portion of energy transmitted across the interface. Table 1 lists the acoustic impedance of a range of hydrated hydrophilic polymers together with distilled water and natural rubber. As one would expect, the higher the water content of the polymer, the closer the values are to those of pure water. It is also noteworthy that the acoustic impedance of all of the polymers evaluated in this study demonstrated acoustic impedance's very close to that of water. This implies similar transmission of ultrasound to that achieved through immersion testing, assuming good surface contact could be achieved. Equation 2 defines the coefficient of transmission in terms of the acoustic impedance's of the first material Z1 and the second material Z2 .
As alluded to earlier, the ability of a solid coupling material to transmit a wide range of frequencies is of major importance to give scope for the broadest range of applications. Figure 1 shows the frequency spectra of identical ultrasonic pulses generated from a 10 MHz transducer, having passed through the same thickness of hydrophilic polymers B, C and D. What figure 1 proves is that the higher the water content of the polymer, the higher the amplitude of the returning signal and the higher the centre frequency.
|Fig 1: Frequency Spectra of a 10MHz pulse transmitted through different hydrophilic polymers of the same thickness.|
|Fig 2: Attenuation vs frequency. Comparison between polymer D, Perspex and a conventional solid couplant rubber.|
Subtracting the frequency spectrum of a hydrophilic polymer from that of aluminium (negligible attenuation reference) and dividing it by the ultrasonic travel length, provides a guide for the attenuation with respect to frequency. Figure 2 illustrates this for polymer D, Perspex and one commonly used solid couplant rubber. It is clear that the attenuation in the rubber increases rapidly with increasing frequency. The attenuation in the Perspex is seen to increase slightly with increasing frequency, whereas attenuation in polymer D does not. In fact, the centre frequency of the ultrasonic pulse having travelled through polymer D was higher than that having passed through the Perspex. This explains why the attenuation of polymer D drops below that of Perspex at the higher frequencies.
|Fig 3: Representation of how coupling efficiency tests were conducted.|
The procedure was repeated six times for each polymer to monitor reproducibility. The first round of tests was conducted with the interface between the hydrophilic polymer and the test piece completely dry. (Note that the interface between the probe and hydrophilic polymer was coupled with a drop of water to emulate perfect contact.) Figures 4 and 5 illustrate typical data achieved for loading and unloading cycles. Note that echo amplitude data is given in terms of the gain required to increase the signal to full screen height on the ultrasonic instrument. Therefore, a low gain corresponds to an echo of high amplitude and vice versa.
|Fig 4: Frequency of pulse echo signal during one loading and unloading cycle. Coupling via polymer B with dry interface with test piece. (5MHz)|
|Fig 5: Amplitude of pulse echo signal during one loading and unloading cycle. Coupling via polymer B with dry interface with test piece. (5MHz)|
The higher the contact pressure, the higher the frequency of the steel backwall echo. As expected, the level of coupling increased with increasing contact pressure enabling transfer of some of the higher frequency components. Similarly, echo amplitude increases with increasing contact pressure, again suggesting increased transfer of ultrasound. It was observed that the amplitude of the back wall echo remained high once pressure had been removed. It is believed that this is due to the polymer 'sticking' to the test piece like a suction cup.
The results achieved from the different hydrophilic polymers followed consistent trends - the higher the water content, the higher the frequency and amplitude of the steel backwall echo. This may be visualised from figures 6 and 7 where data has been pooled from all four hydrophilic polymers.
|Fig 6: Frequency of steel back wall echo through different hydrophilic polymers during unloading cycle. (5MHz)|
|Fig 7: Amplitude of steel back wall echo through different hydrophilic polymers during unloading cycle. (5MHz)|
From the scatter data, trend lines may be calculated, from which, intercept values with the Y-axis may be determined. This yields an indication of the frequency and amplitude at a theoretical zero pressure for rapid comparison. Likewise, the gradients of the lines provide information concerning the change in amplitude and frequency per unit pressure.
|MHz/kgf||MHz @0kgf||dB/ kgf||dB @0kgf|
|Table 2: Trend data from different hydrophilic polymers acting as dry contact ultrasonic couplant during unloading cycle.|
In addition to the tests being performed with the interface between the polymer and the test piece completely dry, similar data was acquired with the interface saturated with water. This guarantees that surface contact is 100%. This data is represented as solid lines on figures 6 and 7. As these 'wet' lines are within or very close to the 'dry' scatter data, it is evident that similar levels of amplitude and frequency are attainable whether the polymer/steel interface is either wet or dry.
|Fig 8: Amplitude comparison between loading and unloading cycles for polymer B. (5MHz)|
The solid green line represents the data obtained when the interface between the polymer and the steel was wet. The dry loading data appear to converge on the unloading and wet data from approximately 1000 - 2000g onwards. This indicates that the efficiency of ultrasound transfer will not continue to increase at the same rate with further increases in contact pressure, representing an optimum contact pressure for practical use. Similar contact pressure threshold estimates can be made from looking at the data from the other polymers tested. Table 3 lists this data in terms of the applied force divided by the area of hydrophilic polymer in contact with the test piece. While this is a practical measure, it is appreciated that the actual area of contact between the hydrophilic polymer and the test piece may deviate from this very slightly owing to surface finish and the nature of the test.
|Hydrophilic Polymer||Pressure Threshold (kg cm-2)|
|Table 3: Approximate contact pressure threshold of a range of hydrophilic polymers.|
It was noted that at very low contact pressures, signal amplitude could be seen to increase with time while the pressure was held constant. This was felt to be a result of the hydrophilic polymer continuing to conform to the test piece surface, further improving the level of coupling. This was compensated for in these tests as measurements were taken soon after the pressure was applied. This is important for practical application as hydrophilic polymers used as solid couplant on wheel probe devices will be required to continually conform to a test piece surface as the tyre rotates. This situation mimics that achieved in this series of tests where measurements were taken rapidly. In the case of manual 'static' testing where the probe may be in contact with the test piece for a few moments before the measurement is taken, a reduced level of contact pressure may be required.
|Fig 9: Model of prototype wheel probe.|
As the water path and tyre thickness are used as a delay, transducer dead zone effects are minimised making inspection very close to the test piece surface possible as would be the case with conventional immersion testing.
The first prototype employed a single crystal 5 MHz, focussed, medium damped transducer, which has proven most effective for a range of applications. The probe itself is interchangeable increasing the flexibility of the device. Current trials are underway with a highly damped 10 MHz transducer.
|Fig 10: Example of prototype wheel probe being used both manually and as part of a scanning system.|
Figure 11a shows typical echoes received on a Sonatest Masterscan 335 flaw detector from the wheel probe when it is not coupled to a test piece. The gain has been increased to 53 dB so that the repeat echoes can be seen. The first (left most) peak is the transmitter pulse, the second is the reflection from the water/tyre interface and the third is the reflection from the outer surface of the tyre. The other echoes are multiple reflections from these interfaces. To demonstrate the excellent transmission of ultrasound between the hydrophilic polymer tyre and water, the probe was dipped into water while the settings on the flaw detector remained unchanged. Figure 11b shows the screen of the flaw detector when the tyre is in contact with water. The amplitude of the reflection from the outer surface of the tyre drops significantly as energy is transmitted into the water. Also of note is the disappearance of the multiple reflections.
|Fig 11: a) Signals from uncoupled wheel probe. b) Signals from wheel probe dipped into water. In both cases the gain was set to 53dB and the screen delay was set to zero.|
For normal application the flaw detector would be set up appropriately so that the reflection from the outer surface of the tyre was at the left of the screen and the range appropriate for the thickness of the test piece to be inspected. The black trace in Figure 12 shows the flaw detector screen when set up for the inspection of a 6mm thick Perspex plate. The superimposed blue trace is the signal from the Perspex.
The initial interface echo between the tyre and the Perspex (left most black) can be seen to reduce in size and move slightly to the left when the probe was placed in contact with the Perspex sample. This is due to the slight compression of the tyre, reducing the ultrasonic travel time. The reduction in amplitude is caused by a reduction in the amount of energy reflected from the outer interface of the probe as energy is transmitted into the Perspex. Of particular note is that the interface echo is lower in amplitude than the Perspex back wall echo. This is evidence of the excellent transmition achieved via this arrangement and gives scope for defect detection very close to the test piece surface. Figure 13 shows the signal received from a 2 mm flat bottomed hole, 2 mm below the surface in the Perspex sample. Even though the gain is only set to 53 dB, a repeat echo from the defect is clearly visible.
|Fig 12: Black - trace from wheel probe with screen set-up appropriate for the inspection of Perspex plate. Blue - trace achieved when probe placed in contact with Perspex. Gain remained constant at 53dB.||Fig 13: Signal from a 2mm flat bottomed hole 2mm from the surface of a Perspex test piece.||Fig 14: Inspection of corroded steel plate. a) - Standard immersion test. b) - Wheel probe with 2 drops of water. c) - Wheel probe operating completely dry.|
Using the wheel probe as part of a scanning system, data may be represented in C-scan form. To date a number of materials have been examined this way including metals, composites and ceramics. Of particular interest was how such data acquired with the wheel probe compared to that acquired through conventional immersion testing. In order to gauge this a steel plate was immersion tested with the miniaturised immersion probe used within the wheel probe. The result is shown in figure 14a. The severe corrosion on the rear of the plate can be clearly seen. Figure 14b shows the test repeated but using the wheel probe. In this instance two drops of water aided coupling. With this arrangement the defects were still detected with clarity matching that of the immersion test. Figure 14c shows the results after the test was repeated completely dry, coupling being provided solely by the hydrophilic tyre. In all three cases the defects on the underside of the plate are clearly visible.
In order to demonstrate the time of flight mode of operation, a steel wedge has been scanned. The minimum thickness was 2.5 mm and the maximum 13 mm. The uniformity of the scan reflects the consistency of the measurements obtained.
|Fig 15: Time of flight inspection of a steel step wedge.|
|© AIPnD , created by NDT.net|||Home| |Top||