Defects thickness

In order to confirm the capability of acoustic emission to detect the evolution of a defect during a pneumatic test, a vessel with a lack of thickness due to the action of corrosion was pressurized pneumatically until burst. 

Radiography is less effective for detecting arbitrarily oriented planar defects (such as tight cracks or lack of fusion) in thick metal sections unless the likely location and orientation of cracking is known in advance. 

In order to reply to some particular cases of NDT such that the detection two close defects and the fine thickness measure, we have developed techniques adapted to this type of problem. 

This system has been applied also on a weld containing real defects. It consists of two sheets in ordinary steel, of equal thickness (35 mm), welded with a chamfer in X. 

Several types of Eddy current probes were used with the SQUID system and the commercial system as well. High inductance wire wound probes with a ferritie eore and low induetance planar thick frhn coils were applied. The wire wound probe is the commonly used probe for high resolution conventional testing. The low inductance planar cod is more suited to be apphed in combination with the SQUID system. It is well adapted for surfaee defects and shallow defects. 

The application of fundamentally new ECT (Russia patent Jf 2063025) has made it possible to provide high-efficiency defect control accompanied by detecting both small surface defects and more rough under-surface defects under non-magnetic metal layer of 7 mm thick, or surface defects under protection coatings, dye, corrosion, hermetic and other type of layer of 10 mm thick. 

The defects detection under the layer of any dielectric (varnish - colour, heat - insulation coating, sealant etc.) to 10 mm in thickness. 

Detection of corrosion and defects under a layer of protective coatings, rust and foulings to 8-10 mm in thickness. 

There are difficulties of detecting defects in axial canal because of solid sediment layer of 1. 2 mm thick on the canal surface. When using known defectoscope devices a preliminary labor-intensive mechanical treatment of the axial canal surface is needed. The experience of application of different methods of rotor axial canal control in multifunction automatic device ROTOR - K has pointed to the fact that the most effective method is eddy current one [1]. All the dangerous cracks were just detected by the eddy current method, the part of the cracks were not... 

If we suppose that the presence of a long axial emerging rectangular defect doesn t change the thickness e of the ring where the induction in the tube is not equal to zero but only its circumferential length, and also that it is it does not change the local induction in amplitude and phase, then the presence of the defect increases only the of the piece where the induction is not equal to zero, this increment is equal to 2he. 

It can be observed from the Figure 1 that the sensitivity of I.I. system is quite low at lower thicknesses and improves as the thicknesses increase. Further the sensitivity is low in case of as observed images compared to processed images. This can be attributed to the quantum fluctuations in the number of photons received and also to the electronic and screen noise. Integration of the images reduces this noise by a factor of N where N is the number of frames. Another observation of interest from the experiment was that if the orientation of the wires was horizontal there was a decrease in the observed sensitivity. It can be observed from the contrast response curves that the response for defect detection is better in magnified modes compared to normal mode of the II tube. Further, it can be observed that the vertical resolution is better compared to horizontal which is in line with prediction by the sensitivity curves. 

In this figure one can clearly see an image of a weld defective area with a superimposed groove defectometer Fe2 4 mm thick. The depth of a minimum groove was 0,5 mm. One can clearly see a groove 0,2 mm deep, which corresponds to sensitivity less than 2% Both incomplete root penetration and metal weld beads can be clearly seen in the image. 

The volume of defects is calculated using intensity evaluation. Considering the polychromatic radiation of microfocus X-ray tubes the X-ray beam is represented by an energy dependent intensity distribution Io(E). The intensity Ip behind a sample of thickness s is given by integrating the absorption law over all energies ... 

For detection of defect dimensions defectometers were used made of examined sample material allowing to reveal defects of 0.25-1% x-rayed steel layer thickness in range of 100-500mm thickness at 11 MeV. 

Because of the large thickness x-rayed by high-current betatron (more than 1500 ram on plastic) the picture of defect situated near the front wall of the sample is too increased what makes difficult to detect in practice real dimensions of the defect. But there is a possibility of stereosurvey due to existence of two radiation beams simultaneously generated by betatron. 

General hydrodynamic theory for liquid penetrant testing (PT) has been worked out in [1], Basic principles of the theory were described in details in [2,3], This theory enables, for example, to calculate the minimum crack s width that can be detected by prescribed product family (penetrant, excess penetrant remover and developer), when dry powder is used as the developer. One needs for that such characteristics as surface tension of penetrant a and some characteristics of developer s layer, thickness h, effective radius of pores and porosity TI. One more characteristic is the residual depth of defect s filling with penetrant before the application of a developer. The methods for experimental determination of these characteristics were worked out in [4]. 

Here a - surface tension pa - atmospheric pressure 9 - contact angle of crack s wall wetting by penetrant n - coefficient, characterizing residual filling of defect s hollow by a penetrant before developer s application IT and h - porosity and thickness of developer s layer respectively W - minimum width of crack s indication, which can be registered visually or with the use of special optical system. The peculiarity of the case Re < H is that the whole penetrant volume is extracted by a developer. As a result the whole penetrant s volume, which was trapped during the stage of penetrant application, imbibes developer s layer and forms an indication of a defect. 

One can see from the formulas (1) and (2) that PT sensitivity strongly depends on the thickness of a developer s layer. But during liquid s penetration into developer s layer the powder particles are sinking and more tightly packing each other. It results in decrease of layer thickness h Physical meaning of the influence of this process upon defect s detection is obvious as follows. 

As far as the real reduced thickness h of imbibed zone becomes smaller than h due to the sedimentation of the particles, the volume of the penetrant, which is extracting from the defect by the developer, is enough to form the indication (or so called trace ) of larger square on the outer developer s surface. It means that one can reveal defects of smaller sizes and PT sensitivity becomes higher. 

For defect sizing by TOED, use of L waves involves a penalty in resolution of almost a factor of two at a given frequency because of difference in velocity as compared to shear waves and use of SV waves runs into difficulties because of the mode conversion problems. Further, problems due to couplant thickness variations, surface roughness affects, beam skewing and distortion problems in anisotropic welds can also be expected. On the contrary, SH waves are not affected... 

---