PSI - Issue 14

Altaf Ali et al. / Procedia Structural Integrity 14 (2019) 273–281

276

4

Author name / Structural Integrity Procedia 00 (2018) 000–000

angular position and then three shots, at 185 kV X-rays for 35 seconds each, were taken at corresponding positions for full coverage of the weld. ASTM 1A 6 type penetrameter is used as Image Quality Indicator (IQI), Halmshaw and Kowol (2001). A schematic of the setup and actual experimental setup are shown in Fig. 2 (a) and (b) respectively.

Table 1. Sample preparation details Sample ID Expected defect

Method of preparation

By applying few powder particles at the time of welding

A-1 A-2 A-3 A-4

Inclusion

LOP/Root Pocket

Welded at lesser current

Porosity

By applying some grease at the time of welding

Incomplete Weld

By stopping welding before full circumferential coverage

LOP type of defects at various currents

FCW

Not any defect

Welding at full current

TFCW HCW QWC

LOP LOP LOP

Welding at three fourth value of full current Welding at half value of full current Welding at quarter value of full current

Fig. 2. (a) Schematic representation of geometrical settings for experimental radiography setup; (b) Experimental setup for digital radiography

3. SNR Analysis for LOP Before exploring ultrasonic testing, it is important to know about the nature of defect specially LOP and smallest defect detectable in radiography so that a better comparison can be made between the two (RT and UT). For defect detection in radiography, SNR & CNR play important roles, Ewert et al. (2005). However, quality of present radiographs was interpreted using wire type IQI where smallest diameter visible wire was checked. But the IQI region may have different SNR in comparison to the weld region having the defect. To have an idea of smallest defect detectable in weld region, smallest detectable details of both the regions should be compared based on CNR value of both the regions. However, scattering effect can be ignored here for comparison as two regions of the same radiograph are being analysed and they should have similar scattering. So final equation for comparison is So, SNR of 51 radiographs taken earlier was measured in the two regions of radiograph using YXLON Image 3500 software. The distribution is shown in Fig. 3 (a) & (b) below. SNR in weld zone is having average of 1236.6, standard deviation of 333.77, maximum value of 1759.4 and 543.2 is its minimum value observed. On the other hand, SNR in the IQI zone is having average of 1839.84, standard deviation of 129.53, maximum value 2000 and minimum value 1645. To give sufficient flexibility to RT in comparison to UT, let � � � = avg. - std. dev. = 1839.843-129.529 = 1710.314 and � � � = avg. + std. dev. = 1236.6 + 333.76 = 1570.36. and the threshold in the w w p P d ) ( )   SNR d ) ( ) (    SNR ( (1)