Enhancing the Quality of Heavy Wall Austenitic Stainless Steel Welds of Nuclear Reactors through Ultrasonic Examination

Transcription

1 National Seminar & Exhibition on Non-Destructive Evaluation, NDE 2014, Pune, December 4-6, 2014 (NDE-India 2014) Vol.20 No.6 (June 2015) - The e-journal of Nondestructive Testing - ISSN Enhancing the Quality of Heavy Wall Austenitic Stainless Steel Welds of Nuclear Reactors through Ultrasonic Examination Yokesh Ravichandran, Mitesh Panchal, Suresh Jakkula, Niyant J Mehta and Pemmaraju Raghavendra NDE Department, Heavy Engineering Division, Larsen and Toubro Limited, Hazira, Surat, India Niyant.Mehta@larsentoubro.com Abstract. Fabrication of Heavy wall stainless steel (SS) vessels is still a challenge for the Heavy Fabrication Industries, as quality check methods of the weld is still under development through the study on its Anisotropic Metallurgical structure. Volumetric examinations such as Radiographic Examination (RE) and Ultrasonic Examination (UE) are having their inherent limitations. This paper explains how UE can be a suitable volumetric method for high thickness SS 304Lwelds inspite its inherent property of getting scattered and attenuated due to the coarse dendritic structure of the weld. The high thickness and coarse dendritic structure of the weld made UE very tough. Especially Austenitic SS weld metal has very big grain boundaries compared to its base metal causing beam skewing. Common UE probes of regular use are very high prone to this scattering and spurious indications were produced making the interpretation difficult. Special longitudinal angle beam ultrasonic probes were custom-made in collaboration with APplus RTD to enhance the penetration power and to address the attenuation and scattering concerns. Zone wise inspection technique was adopted with focusing the sound intensity at the different depths of interest. A qualification block of 200mm thick SS 304L material was made by SAW weld process, for the validation of the UE. Different types of reflectors were introduced in the fusion zone and in the weld volume at different depths of the block. Further implementation of the above measures successfully enhanced the reliability of the test in terms of repeatability and characterization of indications observed. Keywords: Anisotropic, Austenitic, Longitudinal Angle beam. High thickness. Introduction: Austenitic stainless steels are an important class of engineering materials which are generally used for applications where corrosion resistance, high strength and creep resistance at elevated temperatures are required. They are now been widely used as the structural material in nuclear industry. Fabrication of Heavy wall austenitic stainless steel (SS) vessels is a challenge for the Heavy Fabrication Industries, as quality check methods of the weld is still under development through the study on its Anisotropic Metallurgical structure. Ultrasonic examination is one among the commonly used NDE method for the weld inspection. Although stainless steels are considered to be having good weldability, the coarse, dendritic grain structure of the weld material, formed during the welding process, is extreme and unpredictably anisotropic. During welding of austenitic material columnar crystalline structure of weld material cannot be avoided until today because it cannot be transformed by recrystallization into a fine grain isotropic one [1]. Columnar grains as well as anisotropic and heterogeneous structure make their

2 ultrasonic testing difficult. Higher grain scattering noise and beam skewing in thick welds lead to poor signal to noise ratio and distortion of ultrasonic beam [2]. The degree of attenuation is so high that even with high gain setting in conventional ultrasonic equipment; it is difficult to obtain the back wall echo from a 75mm thick plate [3]. The problems become more acute as the weld thickness increases. As the thickness of weld increases, the reduced heat transfer rate due to the low thermal conductivity of the austenitic stainless steel results in much coarser grain structure. The present study aims at developing an ultrasonic methodology for inspection of high thickness austenitic stainless steel welds with the classic A-scan presentation. The study is performed on a 200mm thick block of SS 304L material made by SAW weld process. Longitudinal wave mode was used for inspection and the reason for choosing the same is briefed in the paper. Special longitudinal angle beam ultrasonic probes were custom-made in collaboration with Applus RTD to enhance the penetration power and to address attenuation by scattering. Zone wise inspection technique was adopted with focusing the sound intensity at the different depths of interest. Grain structure of Austenitic stains less steel welds: The grain structure of austenitic stainless steel weld differs drastically from that of ferritic steel weld. In the initial stage they both will have a columnar grain structure. In multi pass welding process deposition of subsequent weld beads will re-melt the earlier passes and forms a columnar grain structure. In case of ferritic welds as there is a phase transformation from austenite to ferrite these columnar grain structure will get destroyed but in the austenitic welds as there is no phase transformation these are not getting destroyed. Fig. 1 Micrographs of different zones of the weld joint

3 The thermal gradient in the weld pass will be the source for this columnar grain structure. The direction of crystal growth can be predicted to an extent. The thermal gradient is determined by the solid metal which is in contact with the weld metal. For example if a vertical plate is welded with horizontal passes. Heat flows out of the bead deposited in two directions and therefore crystal growth is at an angle to horizontal. [4] The microstructure can vary from region to region in a weld (Fusion zone, HAZ, Base metal). Microstructure has a profound effect on weld properties also. From micrographs shown in figure 1 the weld metal microstructure shows the dendritic structure of ferrite in Austenite matrix, whereas base metal microstructure has islands of ferrite in austenite matrix. The weld metal clearly shows dendritic structure. Scattering and beam skewing effect of Ultrasonic sound wave due to the coarse grain structure: A reflected wave is said to be homogenous like transmitted, only when the observed patterns are easy to interpret with less signal to noise ratio. Testing the Austenitic stainless steel weld is a difficult and challenging task still today because of the anomalies in the sound wave received. [5] In some type of welds such as SAW and FCAW, not even the back wall echo is observed from the weld metal that is transmitted normal to thickness. Sensitivity and resolution of the ultrasonic wave can be achieved with higher frequencies. But the higher frequency sound waves finds very difficult to penetrate higher thickness. Penetration or maximum depth range of an ultrasonic wave from which useful indications can be detected, reduces as the frequency increases. This effect is most pronounced in the coarse grain structure material or weld metal. Scattering and beam skewing are major problems which are faced while testing a coarse grained weld metal. The large grain boundaries plays vital role in scattering and attenuating the sound wave. The angle between grain flow directions with respect to the transmitted beam makes the beam skewing. The orientation of the grains axis and the direction of sound transmitted will play a vital role in beam skewing. The attenuation due to scattering will be less compared to beam skewing. A cylinder is machined out from a horizontal welded plate having grain axis along the diameter. Two probes were used, one as transmitter and one as receiver probe. The receiver probe is moved in increasing and decreasing angle along the diameter and the sound wave transmitted is found to be received with less noise and high energy when transmitted sound wave makes a 45º to the grain axis. When the sound wave is transmitted in line with the grain axis maximum beam skewing was observed and when it is approaching higher angle with grain axis minimum beam divergence was observed [4]. This leads to the point that should be considered while selecting the angle of refraction for test in stainless steel welds. There will be less beam skewing as the normal of the transmitted wave makes 45º or 0º to the grain axis. Zonal discrimination technique: Testing a 200mm thick austenitic stainless steel (SS 304L) weld, with usual probes is a difficult task due to the effect in its character by the grain structure as mentioned above. A zonal discrimination technique with simple A-scan presentation was adopted. The huge thickness was divided into different zones as suitable for inspection. The schematic of zonal discrimination done is as shown in the figure 2.

4 In addition, to compensate for loss of energy due to scattering and beam skewing, we used the probes which will focus the ultrasonic wave to the particular depth of interest. Longitudinal wave mode was used for the testing. The probe s refraction angle was selected not only considering the perpendicularity to the weld edge preparation and maintaining less travel distance for reducing the attenuation due to distance of travel but also considering the angle with the grain axis to have a control on beam skewing. Creeper wave probes for depth up to 3-4 mm. Top zone up to 20mm from surface 20mm to 60 mm from top with FD 30 probes 60 to 120 mm from top with FD 70 probes 200 mm Fig. 2 Zonal discrimination approach. The probes used were tailor made for focusing to different depths such as 10 mm, 30mm and 70mm. These probes were specially made along with APplus RTD. These are twin crystal probes with narrow band width. A 2 MHz frequency was chosen for the test in order to take care of penetration. Actual Test Procedure: The important challenge before testing this huge thickness is to simulate the welds to be tested with that of a mock up block that will represent more or less similar grain structure of the weld.so that the sensitivity calibration part shall be done to the satisfactory level. The best way is to weld a block with same parameters which is used in the weld to be tested. The testing was carried out as per ASME sec VIII division I acceptance standards. ASME section V used for calibration block making [6]. As shown in the figure 3, the SDH s were drilled at one of the fusion line of the weld i.e. in the WEP. A 200mm thick mock up block was made for the calibration and qualification of the test. The SDH s of 8mm diameter were made at six different depth s H1-10mm, H2-25mm, H3-50mm, H4-75mm, H5-100mm, & H6-125mm from the surface. In addition to this, notches of dimension 1.5(w)x1.5(h)x25(l)mm,1.5(w)x3(h)x25(l)mm &1.5(w)x5(h)x25(l)mm were made at three different depths.the qualification of the calibration was justified by studying the response from these notches.

5 H1 H3 H2 H4 H5 200 mm H6 Fig. 3 Mockup block sketch The scan plan was prepared in a manner to catch all the holes to their respective zone of inspection passing through the weld, so that the actual attenuation shall be simulated to an extent while calibration itself. The weld is flush grinded to ensure the test with zero degree also. The signal to noise ratio was also found to be less as shown in figure 4&5. A normal raster scanning was performed with A-scan presentation. Scanning was done from both the faces to ensure complete coverage including the blind zones near to the surface. Results and discussion: The response from the notches was used to study the actual attenuation felt in the weld metal. Even though the ASME recommends for an 8mm diameter hole, the minimum sensitivity is to be established. The sensitivity achieved is 3mm which was ensured with the notch made in the block. The detection alone not only provides a path for the acceptance or rejection of an indication, there are some important parameters of the indication that too should be studied. One among them is characterization and another is dimension of the indication. For characterization we performed the regular way of using the orbital scan, checking the observed indication with other angles having a difference in angle of at least 15º and plotting it with a 1:1sketch. For dimensions of the indication we used the 6dB drop technique. The noise due to the grain structure influence was reduced to a satisfactory level that enables us to easily interpret the observed signals. The figures 4 & 5 shows A-scan displays captured after plotting the DAC curve. The signals are observed with less noise so that it is easy to interpret the weld. 70º probe with focal depth 10mm. 45º probe with focal depth 30mm. Fig. 4

6 70º with focal depth 30mm. 70º with focal depth 70mm. Fig. 5 Conclusion: This work successfully demonstrates an inspection methodology for high thickness austenitic stainless steel welds. Though the proposed inspection methodology requires access from both sides of the weld, the reliability in detection and repetition of observed signals surpasses the time taken for examination. The main hurdle of receiving the useful signal against the strong scattering and beam skewing effect were overcame by changing the probe s parameter selection by considering the axis of the dendritic grain structure also. The special purpose twin crystal longitudinal probes were found effective for high thickness inspection. By applying zonal discrimination approach, the complications involved in very high thickness inspection were narrowed down. References [1] J.F. Shackelford: Introduction to material science for engineers. Upper Saddle River, New Jersey, [2] Anish Kumar, B. Sasi et al: Non-destructive evaluation of Austenitic Stainless Steel Welds. Advanced Materials Research Vol. 794 (2013). [3] R. Subbaratnam, M. Palaniappan et al, Ultrasonic Examination of Heavy Austenitic Stainless Steel Welds - our experience, WCNDT 96- New Delhi. [4] Ultrasonic Inspection of Austenitic Welds J. R. TOMLINSON, A. R. WAGG, M. J. WHITTLE U. D. T. Applications Centre CEGB, Manchester United Kingdom. [5] ASM Metals Handbook Vol.17. [6] AMSE BPVC section VIII & V Edition