·Table of Contents
·Materials Characterization and testing
New Electric Discharge Frequency Method of Non-Destructive Testing of Dielectric MaterialsV.V. Kozharinov
Technical Expert Society "TUV Nord Baltik" Non-Destructive Laboratory , 23 Klijanu str., Riga, LV-1012, Latvia.
Institute of Electronics of the Belarus National Academy of Sciences, Minsk.
Institute of Physics-Organic Chemistry of the Belarus National Academy of Sciences, Minsk.
In the given article it is offered to use a new electric discharge frequency method for a non-destructive testing of quality of dielectric coatings and the results of its applying for determination of thickness, locus and expansion of local defects, quality estimation of electrohardness of dielectric coatings and permeability of dielectric membranes are introduced.
|Fig 1: The measurement scheme: 1- needle electrodes; 2- investigated plant; 3- flat welding rods; 4- high-voltage power sources; 5- precision resistors; 6- frequency meter.|
The measurement scheme is shown in Figure 1 A needle electrode 1 made of NiCr wire of 2 mm diameter with sharpening angle of 60 degrees and rounding radius of 10 mm can be moved by means of a micrometric screw relative to a plane electrode 3 upon which the specimen 2 is placed. The needle electrode is connected to an output of a high-voltage power supply 4. The plane electrode is earthed through a precision resistor 5 on which the pulsation frequency of the gas-dicharge leak current is measured by means of a frequency meter 6. The whole measurement set (except for power supply and frequency meter) was placed in a climastat capable of setting and maintaining constant parameters of gas environment (temperature, relative humidity). Dielectric materials were studied both in the form of substrates and in the form of thin-film layers coated with thin-film conductive layers.
Pulsed nature of the discharge current is due to periodic discontinuation of electronic flow in the microdischarge channel mainly caused by accumulation of adsorbed charge on the surface of dielectric-coated object electrode. The presence of such surface charge imposes a limitation on the maximum permissible value of the charge carried through the channel of a unit microdischarge. Generally, this value is determined by dielectric properties of the coating and characteristics of the power supply circuit. The frequency of the discharge current pulsations is determined by the dynamics of the surface charge which is testingled by properties of the surface and gas influencing the movement of the charge over the surface (surface discharge, charge spreading due to surface and bulk conductivity). This dynamics is conditioned physical - chemical propertiess of a dielectric surface and environment. Environment influences on storage and moving of a charge on a surface. The dependence between oscillations frequency of an electric current in a discharge and mean value of this current has linear nature with a high scale accurancy . It allows at analysis of influencing of physical -geometrical parameters of a discharge gap on the gases discharge performances, of small currents measuring (nA) to approximate by frequency measurings. Thus process of measurings is essentially simplified and their reliability is increased.
The spacial resolution of the method is defined by the microdischarge channel diameter and does not exceed 0.1 mm for the discharge gap less than 1 mm for the electrode geometry representing needle electrode and plane cathode coated with dielectric. It is also testingled by the effective area of interaction between the discharge and dielectric coating whose physical properties influence charge transfer processes in the discharge zone (no more than 0.01 mm2) and is equal to 0.1 mm .
In this article the examples of tendered method usage for an estimation of dielectric coatings thickness, phase composition and different dielectric properties of anodic alumina (AA), permeability of dielectric membranes on its basis are given.
AA was obtained at an anodic oxidation of aluminum A99 in 3% solution of ethanedioic acid at galvanic-static condition . Dielectric thickness was instituted by time of anodization. The thickness concerning dense (so-called barrier) AA stratum was varied re anodization in insolubilizing electrolyte. The separation of AA from aluminum was carried out under the standard method.
For obtaining of AA substrates of definite dielectric properties, AA annealing was carried out at temperatures down to 1200° C. It allowed to receive alongside with source amorphous and polycrystallic samples from g up to a modifications . The electrical link to AA substrates was reshaped at evaporation of a thin-film nickeliferous welding rod by thickness 0,1 microns. The evaporation was conducted on the part of a barrier layer.
Serial samples of AA dielectric membranes was prepared by removal of a barrier layer and variation of AA treating conditions for obtaining of pores different diameter and spacing intervals between them .
On Fig. 2 the dependences of leakage current pulsations frequency f of a discharge from thickness d of AA substrate are shown. Magnitude of a discharge gap were various for samples with thickness of 0,4 microns barrier layer. Such stable dependence weren't obtained for samples with barrier layer thickness of 0,1 microns because of instability of gas-discharge process and wide scatter of leakage current pulsations frequency of a discharge for each measuring point. The electric voltage given on a needle electrode was 1.6 kV.
|Fig 2: The dependences of leakage current pulsations frequency f of a discharge from thickness d of AA substrate for different magnitudes of a discharge gap: ? - 350 microns, C - 400 microns.||Fig 3: The dependences of leakage current pulsations frequency f in a discharge from of films thickness d of electrical-insulating epoxy coating for different magnitudes of voltage difference between electrodes: B - 3.25 kV, C - 3.70 kV, D - 4.26 kV.|
The dependences of leakage current pulsations frequency f in a discharge from of films thickness d of electrical-insulating epoxy coating for different magnitudes of voltage difference between electrods are presented on Fig. 3. The films were put on aluminium laminas and differed on thickness. The testing of coatings thickness was carried out with the help of the detecting instrument of thickness with fidelity ± 1 microns. The voltage ranged 3-6 kV on a needle electrode. The air gap between a needle-like and flat electrods with dielectric coating was saved by a constant and was 8 mms.
Experiment on study of dependence of leakage current pulsations frequency of a discharge from dielectric coating thickness demonstrated. For each dielectric coating depending on its dielectric properties, state of a surface and structural features the area of stable existence of gases discharge is available. So, for lectrical-insulating epoxy coating with rather good dielectric properties the stable discharge exists at a voltage on a needle electrode superior 3.5 kV, and a discharge gap magnitude of 5-10 mms (the high bound of a discharge gap magnitude depends on the appended voltage).
For AA having the worse dielectric property and porous surface, the voltage on a needle electrode was 1-3 kV (depending on coating thickness and a barrier layer thickness). The magnitude of air gap is comparable to thickness of AA film and was 0,1-1 mms.The inaccuracy on leakage current pulsations frequency in a discharge also depends on physical characteristicss of coating. So, for epoxy coatings the inaccuracy on frequency for certain conditions did not exceed 1 %. For AA with its porous structure at 0.4 microns thickness of a barrier layer the inaccuracy was about 10 %, and at 0.1 microns thickness of a barrier layer reached magnitude of a mean rate of oscillations frequency.
Thus, it is visible, that the leakage current pulsations frequency in a discharge is responsive to variation of physical characteristics of dielectric coating (in particular, thickness), that allows to monitor each parameter of dielectric coating for given conditions, used the calibration curves obtained on standard samples. The example for it can be served by thickness profile of electrical-insulating coating obtained by approximating frequency measurement of a leakage current in a discharge at scanning of needle electrode on an investigated sample surface, shown on Figure 4.
|Fig 4: Frequency f of leakage current pulsations of a discharge and profile of thickness d of epoxy coating at scan of a lamina on coordinate x.||Fig 5: Dependence of leakage current frequency on magnitude of a discharge gap (by a gases stable discharge).|
For an estimation of influencing of phase composition and bound with it of AA different dielectric properties the dependence of leakage current frequency on magnitude of a discharge gap, described by a stable gases discharge, was investigated. Obtained curves for applicable samples are shown on Figure 5.
In case of amorphous AA at annealing temperatures from source up to T < 420° C rather large oxide dielectric losses are instituted by presence at AA matrix bulk of molecular water and OH-groups, on which the water molecules can be adsorbed . All other impurities presenting in essential amounts in AA, on dielectric losses, do not render noticeable influencing. The composite dependence of leakage current frequency on magnitude of a discharge gap described, for example, maxima is supervised at 120¸ 130 microns and stable gases discharge in range 0 <d < 180 microns (figure 5, curve 1).
In a spacing 420-820° C - e " ~ 0,01. For such amorphous AA, having high dielectric qualities, the large leakage current frequency is characteristic (f » 200¸250Hz), with a maxima at ~80 microns and stability of a gases discharge at 0 < d < 120 microns (figure 5, curve 2).
In case of polycrystalline AA (annealing at T > 840° C), the dielectric losses are determined by surface OH groups. In air, the entire surface of such oxide is covered with OH groups here H2O molecules are adsorbed in amounts depending on the environmental humidity and temperature. The corona discharge is stable for the discharge gap from 80 to 280 mm. However, for d £ 80 mm there occurs breakdown of the discharge gap (Figure 5, curve 3). In this case, the frequency range of the leak current pulsations does not go beyond 0 ¸ 15 Hz.
Thus, knowing these standard dependences (Figure 5), qualities other AA samples, obtained otherwise is possible to value. At last, one more possible range of application of this method can be installation of a correlation between a gastightness of dielectric membranes on the basis of AA and their specific conductivity.
The dependence of frequency (f) of leakage current pulsations of gases discharge from voltage (U) between welding electrods for two types of dielectric membranes on a AA basis with different pores was investigated. The permeability of first sample was n1 =0,835 sm/sec/atm. The membrane of the second type before invetigation was specially sustained in solution for anodization for cloning microtdefects in AA pattern. This permeability was anomalously high n2 =10,32 sm/sec/atm. The measurings were conducted in climastat at fixed parameters of a gaseous fluid (temperature, damp, air pressure) at identical to both samples of discharge gap h magnitudes).
This experimental results are reduced in a Fig. 6. Dependences f (U) for samples with permeability n1 = 0,835 sm/sec/atm (curve C) and n2 =10,32 sm/sec/atm (curve B), obtained at magnitude of a discharge gap h=0,4 mm.
|Fig 6: Dependences of frequency of current pulsations of leakage of a discharge on the appended stress for membranes with permeability variations at h=0,4 mm.|
Tangents of leans angle of linear parts of curves B and C, proportional specific conductivities of samples, approximately equal to a ratio of permeabilities n1 /n2 =0,081. The discrepancy is interquartile is conditioned by averaging on the different spaces for both methods.
The obtained results allow also to do a conclusion about a possibility of local testing of uniformity of texture and permeability of ceramic membranes by electrical discharge frequency method. The space of testing zone thus will not exceed 10-2MM2. Such localization is impossible for traditional method of non-destructive testings of permeability.
The method allows to meter dielectric coatings thickness, substructure, films as well at oneway access to testing object, that is necessary at diagnostic of extended objects.
|© AIPnD , created by NDT.net|||Home| |Top||