Table of Contents ECNDT '98
Session: Materials Characterization
Non-Destructive Testing of Green Ceramic Materials.H.-D. Tietz, M. Dietz, L. Bühling, B. May - Westsächsische Hochschule Zwickau, Germany.
Corresponding Author Contact:
|TABLE OF CONTENTS|
It would be advantageous to detect defects or inhomogeneousities in the early process of production to reduce the production costs. The main point of this work is the optimization of different non-destructive testing methods for the testing of green ceramic bodies. The non-destructive characterization methods, specially the ultrasonic testing, the measurement of the natural frequency and the measurement of the Ultrasonic-Contact-Impedance (UCI), are very interesting for measuring the different material properties quickly and with low costs.
The problem of ultrasonic testing is to couple the ultrasonic impulse into the testing material. Because of the destruction of the green ceramic, the conventional coupling possibilities with water or other liquids are not applicable. An alternative way can be presented only by the dry coupling method. In this paper, different kinds of ultrasonic dry coupling will be tested for their sensitivity and lateral resolution to detect material mismatches. With dry coupling methods it is very difficult to analyse the lateral resolution of a real defect in green ceramic bodies. Appropriate devices that allow a water immersion technique for the green ceramic testing will be demonstrated. Another possibility of ultrasonic dry coupling with air-coupled ultrasonic transducers will be supposed.
In the first test series, the possibilities of the contact coupling with conventional transducers for the testing of green ceramics were examined. Measurements were made at different specimens with contact coupling transducers in pulse-echo testing method. The nominal frequencies of the transducers were 1 MHz and 2 MHz. The ultrasonic system HILLSCAN 3000 of the firm Dr.W.Hillger, and a device USIP 12 of the firm Krautkrämer were used.
Fig 1: A-scan of a blind hole with a diameter of 2 mm in a depth of 28 mm in a 33 mm thick SiC green ceramic, pulse echo testing method, signal amplification 81dB, broadband transducer 1 MHz
Figure 2 and 3 show A-scans of model defects in SiC green ceramic bodies. A green ceramic step bar and a cylindrical ring with different blind holes were test specimens. The blind holes were produced in various depths and with various diameters. The hole with a diameter of 2 mm in a depth of 28 mm SiC-ceramic was the smallest hole repeatedly detectable.
The velocity vl of the longitudinal waves in green ceramic was measured in the pulse-echo testing method to describe the smallest differences in density which can measured in ultrasonic testing. A 1 MHz broadband transducer from Panametrics was used. The smallest differences in density tested amounted to 0,017 g/cm³ for the SiC-ceramics (fig. 4), and to 0,023 g/cm³ for alumina-ceramics.
Fig 2: Determination of density differences in SiC- bending bars
In ultrasonic contact technique it is difficult to realize an area scan because of the different coupling conditions on every point of the scanned area. It is only possible to take a test at certain points. For this reason, two devices were constructed for testing green ceramics in immersion technique with a waterbag. (fig. 3)
Fig 3: Immersion technique device for the ultrasonic testing of green ceramic bodies
The first device was constructed with a coupling pad of the firm Sonotec GmbH Halle. This coupling pad is a solid silicone coupling gel with acoustical properties of water. The advantage of the material is to osculate to the surface of the sample.
The coupling pad forms with two clamps a bag which can be filled with water.
The bag of the second device was produced from PE-foil. For a better coupling, an adhesive tape was placed between bag and sample.
The devices filled with water make a testing of the green ceramic bodies in immersion technique possible without the direct contact between ceramics and liquid. So the green ceramic could not be destroyed by the liquid, and the coupling conditions are equal over the scanned area.
The best coupling for the system sample-device-transducer was produced by the PE-foil device. In the case of the silicone pad waterbag it is supposed that the air gap between waterbag and surface of the sample, and the damping by the silicon pad are too high to get an usable ultrasonic signal from the back surface of the sample. This device and the testing equipment will be optimized in the future work.
Fig 4: C-Scan of bending bars with different densities
Fig 5: Embedded Teflon disks simulate an area of cracks in a SiC-sample. The amplitude of the flaw echo was rated.
Figures 6 to 8 show scanned pictures of green ceramics with different densities and model defects. These samples were scanned in immersion technique with the PE-waterbag. The transducer was a focused transducer for immersion technique with a nominal frequency of 2.25 MHz. The samples were green ceramic bodies with different synthesized defects which were produced during the pressing process.
Defects with an area of more than 10 mm in diameter, like Teflon disks, small Styrofoam sticks, or plastic tubes with a diameter of 1mm could be detected with high resolution. The amplitude of the flaw echo was evaluated (fig. 5). Areas (Ø 10mm) with small defects like 2 mm metal or wood splits, and prepressed material could be detected with the evaluation of the back surface echo. The SiC-samples had a thickness of 15 mm. For this thickness, single defects smaller than 5 mm were not detectable in green ceramic bodies.
Other possibilities to minimize the air gap between the waterbag and the green ceramic sample are examined. In this way the insertion of ultrasonic energy into the green ceramic material could be increased.
Fig 6: Amplitude response of a 5 mm SiC bending bulk produced with two air coupled capacitance transducers in ultrasonic through-transmission testing method
Air-coupled ultrasonic waves produced with capacitance transducers were used for studies in the non-destructive green ceramic testing. The capacitance transducer works with the technical principle of a capacitor and microphone.
The first result of the tests with the capacitance transducers is shown in fig. 6.
The materials of the samples were acrylic glass and SiC green ceramics. The transducers were used in a distance of 3 mm to the front and back surface of the sample and in the ultrasonic through-transmission testing method.
The frequency response of the received ultrasonic signals showed the maximum of the amplitudes between 300 to 800 kHz. For this reason, the air-coupled ultrasonic testing is less suitable for the detection of defects at present, but it presents the possibility of an area scan to determine the allocation of the density in materials with lower acoustic impedance. For the time being, the material thickness that can be tested with air-coupled ultrasonic waves amount to 10 mm for Perspex and silicon nitride, and to 5 mm for silicon carbide. The received amplitude in SiC had a signal-noise-ratio of only 6 dB.
Future studies will research into the possibilities of the air-coupled ultrasonic technique. The aim is to scan the full volume of the green ceramic bodies in through-transmission method.
The ultrasonic hardness testing method developed by Kleesattel /3/ allowed the determination of the hardness by measuring the increase in resonance frequency of a mechanical resonator, brought into contact with the test specimen. By finding the increase in the resonant frequency of the resonator, the size of the contact surface AC could be determined if the elastic constant is known. "UCI method" (Ultrasonic Contact Impedance) is the denomination of this procedure.
The UCI method is mainly suitable for the small load, and macro hardness tests. It is necessary to transmit the test load onto the indenter without dampening the oscillation. Therefore, a mechanical resonator whose standing wave shows at least one vibration node K is used. The resonance equation is approximately represented by the formula: /5/
EP/I Young's modulus of the specimen/ indenter rs density of the swinging stick As cross-sectional area of the swinging stick Ak contact surface f resonance frequency if Ak> 0 fI no-load resonance frequency µP/I Poisson's ratio of the test surface/ Indenter
The described method of measurement is only applicable under specific conditions.
Test specimens may be neither too light nor too thin, and the reactance of the specimens should be at least two orders of magnitude larger than the contact reactance.
At commercially available probes a Vickers indenter is located on the top of the sensor rod. As a result of the penetration of the indenter with a test load the hardness and the Young's modulus influence the increase of the frequency. The conventional UCI hardness measuring is not recommended for porous components.
From the theoretical pre-considerations for the UCI hardness measuring it is clear that the change of the operating frequency, in the case of known Young's modulus of the swinging stick, is dependent on the elastic constants of the specimen, and on the contact surface. If a possibility to keep the area of contact surface swinging stick/ specimen constant is created the change of the operating frequency in the case of contact is only dependent on the elastic constants of the specimen, and of the swinging stick. The contact surface must be kept as small as possible to prevent a compaction of material. A compaction falsifies the Young's modulus. Simultaneously a compromise must be made with the technical possibilities for the production of an extremely hard probe.
Fig 7: Schematic diagram of the UCI-indenter for the green ceramic examination
The UCI-measuring method could be modified for the test of green ceramic bodies. With the aim of a special indenter it is possible to determine the distribution of the elastic characteristic at the surface or over the cross-section of the green ceramic bodies. Therefore, a probe of a high-strength steel alloy was produced for our investigations.
The small cylindrical form of the probe, as seen in figure 7, reduces friction losses, that would result from the contact bar surface/ test specimen. The indenter should penetrate only insignificantly into the specimen. However, the contact indenter/ specimen must occur via the entire contact surface of the indenters. The indenter for our UCI hardness measuring instrument was produced with a surface of
If the device constant of theswinging stick is
and the effective Young's modulus of the indenter
Fig 8: Variations of the Young's modulus of a green ceramic ring measured by the UCI- method
the equation for the calculation of the contact surfaces
will yield the relationship for the determination of the Young's modulus from the contact impedance:
For the investigations, an UCI hardness measuring instrument with special swinging stick of the BAQ firm from Braunschweig was available. Within the framework of the study, this devicewas used for the first time for measurements wich helped to alter and modify it forour applications
Well reproducible variations of the Young's modulus could be measured over the cross-section of a SiC-ring (Fig. 8). The distance to the outer rim of the ring was least in the case of the test series 6, and largest with the test series 9. Differences in elastic qualities caused by test specimen production could be detected. Further investigations will have to discover unwanted variations of elastic qualities over the cross-section of the specimen with this procedure, and to minimize effects caused by unfavourable positions of the measurement points.
The testing of components by taking advantage of their ability to perform a temporally variable and mostly sinusoidal oscillation after an impulse stimulation, is frequently used above all for a good-bad-examination.
Elastic longitudinal, transversal and torsioned oscillations can be generated in materials by defined stimulation. If the generated wavelength corresponds to specific dimensions of the specimen, resonance effects will occur and particularly large oscillation amplitudes are observed. These natural frequencies allow to determine the Young's modulus provided geometry and the density of the test bar are known.
The natural frequency of the test piece can be determined from the recorded oscillation. In the case of known weight and form factor, the calculation of the elastic modulies is possible with use of this frequency.
According to Spinner and Tefft /4/ the theory of Pickett is recommended /5/ for the flexural stimulation. For a fundamental oscillation n=1, the Young's modulus of a prismatic test specimen is according to Pickett:
T it is a correction factor represented by the equation (8) according to /4/.
Fig 9: Young's modulus of different green ceramics, that in the eigenfrequency measurement determines, stimulation: flexural and longitudinal
The results of these calculations show a good repeatability of the measured frequency for test specimens whose geometry ratio length/ thickness is larger than 6. The measurement of the natural frequencies was carried out with an oscillation analyzer Grindo Sonic as a amplification unit. A digital oscillator with FFT function was used for the precise evaluation of the oscillation signals recorded with a microphone.
The Young's moduli were calculated according to the calculation models of Spinner and Tefft /4/ with the computer programme EMOD.
For distinction testing of green ceramic bodies the work with the natural frequency measurement is favourable. However, it is not possible to conclude a distribution of areas with different densities in the same specimen. Therefore, ultrasonic measurement is the only possibility for testing the full specimen volume, and the UCI method for random tests over the cross-section.