4.5 Test Results of austenitic Welds, Dissimilar Metal Welds and Claddings

In testing austenitic welds, dissimilar metal welds and claddings, overlay and backlay structures it is usually possible to detect inclusions and crack shaped reflectors in the probe's near surface zone. Not only the weld should be tested, at least the heat affected zone left and right from the weld needs to be tested as well.

Shear-Wave Probes and Neighbor Echo 2 Probes are well suited for testing the heat affected zone because it the detection of the probe's far surface cracks is particularly important. A problem occurs if the only feasible access for testing of weld is from one side. Therefore ultrasonic testing must be performed penetrating through the weld, but this is only possible in the few cases that present only the usual (vertical polarized) Shear-Waves (qT_v). If testing with a longitudinal-wave is done, the penetration through the weld is possible, but the detection of cracks at the probe's far surface shows insufficient sensitivity since for longitudinal waves neither corner effect nor Neighbor Echo 2 exists, or is less clearly indicated.

A-scan 1 A-scan 2 A-scan 3 A-scan 4 Fig 4.15: Test of claddings and austenitic welds with TH - Waves, EMAT- Array [104] Fig 4.16: Miscellaneous weld types Fig 4.17: Reference body with buffer and cladding I; Probe: KWU SEL 45° 1,5 MHz Reference flaw: Side Drilled Hole d=3mm Fig 4.18: Reference with buffer and cladding II; Probe: KWU SEL 45° 1, 5 MHz Reference flaw: Side Drilled Hole d=3mm

These drawbacks should not be present if horizontal polarized shear-waves (T_H or S_H) are used. EMAT probes were used in laboratory work and this was confirmed in the first practical applications [104], Fig. 4.15.

Tests and validation of different testing methods are described as follows:

4.5.1 Austenitic - and Dissimilar Metal Welds and Claddings

Fig. 4.16 shows a selection of dissimilar metal welds for which practical testing is needed. Fig 4.17 and 4.18 verify, that testing can be successful for this type of structure also. The use of the 45°-SEL- Probe is applicable only for detection of defects located deep inside. At the probe's near surface zone creeping wave probes are applied with good results - 70°SEL- Probes as well as ADEPT- Probes. On the other hand, Fig. 4.19 impressively demonstrates the drawbacks of the testing techniques: Longitudinal waves couldn't produce test effects for small near surface defects, the usual (vertical polarized) shear waves couldn't penetrate through the grain. [128].

In Fig 4.20 and 4.21 shows how particular cracks within the weld can be detected by use of a mode conversion probe applied through the parent material. Note the neighbor Echo 2 indications, these are significant for depth cracks.

4.5.2 Overlay- and Backlay

These weld repair methods are widely applied only in the USA.

With overlay- welding (Bandage) the integrity of cracked piping is guaranteed for a limited period. The repaired parts are replaced after a specified time.

The backlay- structure is constructed similarly to the overlay. The backlay is added onto the weld at the pipe's backwall surface; the backlay thickness is removed so that the original pipe contour remains. The backlay prevents attacks of stress corrosion cracking, which can occur in unstabilized pipe material (e.g. AISI 304 / DIN 1.430). ( such materials are not used for piping in German nuclear power plants).

Fig 4.19: Notch detection by use of different probes; Notch depth = 1 mm; Notch length = 50 mm
Notch	Sound direction	S/N ratio				Chen& Lehmannn SET/ 2 MHz	RTD SET45/ 1,5MHz
		MWB 45N4	MWB 45N4	MWB 45N4	MWB 45N4
1	Through Parent material	24	26	16	14	18	16
2	1	22 0	26 0	12 0	12 0	14 0	16 0
3	1 2	0 0	0 0	0 0	0 0	0 0	0 0
4	1 2	0 0	0 0	0 0	0 0	0 0	0 0
5	Through Parent material	2	2	2	2	2	0
Fig 4.20: Reference body with cracking; Probe WSY 70-2, Mode Conversion Fig4.21: Crack detection in austenitic welds; Probe WSY 70-2 Fig 4.22: Detection of depth cracks under the Overlay with LLT-Probe.

4.5.2.1 Overlay

The state of the art is approximately as follows: Only depth cracks are detectable under the overlay, those reach app. unto the upper quarter of the wall under an overlay- weld. It is also possible to detect manufacturing faults in the overlay. For the detection of depth cracks, 60°- Longitudinal probes with 2 MHz recommended [33].

Experiments using the LLT-Technique produces surprisingly good results for depth cracks, since the ultrasound wave must penetrate through the overlay structure with a 60° propagating shear wave, Fig. 4.22.

Cracks which reach into the overlay weld can be detected with probes such as creep wave probes, 70°- SEL- probe and the ADEPT- probe.

4.5.2.2 Backlay

Backlay testing is described with one example. Piping can be renovated using a backlay weld and tested with NDT (with an Ultrasonic Pipe Crawler). Before tooling and welding the welds to be renovated need volumetric inspection according to ASME Code. After weld renovation the repaired welds need another volumetric inspection. Since the number of probe fixtures is limited, the number of probes to be used should be minimized.

These requirements are best met with a 3 MHz- ADEPT- 60°- Probe. The performance of a combination of two generating probes, operating in opposing directions, in a (25 x 25 x 25) mm big case is illustrated in Table 4.1.

Table 4.1: Summary of results displayed in Fig. 4.23 to 4.25
Reference Flaw	Signal/Noise - Ratio	Fig.
Backlay- and Surface testing with a D-ADEPT- Probe. 3 mm surface notch	20 dB	Fig 4.23:
Testing of the probe far surface with a D-ADEPT-Probe. 2 mm probe far surface notch (wall thickness = 25 mm)	20 dB	Fig 4.24:
Volume test with D-ADEPT- Probe. 2 mm side drilled hole ultrasonically tested through a austenitic weld	20 dB	Fig 4.25:
Volume test with D-ADEPT- Probe. 2 mm side drilled hole ultrasonically tested through a austenitic parent material	30 dB	Fig 4.25:
Backlay- and Surface testing with a D-ADEPT- Probe (1 x 0,5) mm Notch ultrasonically tested through a 4 mm thick backlay	20 dB	Fig 4.23:

4.5.3 Narrow Slot Welds

Fig 4.26: Comparison: conventional weld / narrow slot weld

The state of the art for welding of austenitic piping is the WIG- Impulse- Narrow Slot Weld Method [153]. For this method the amount of applied weld material is much lower than for other weld methods. Extremely small seams can be done very well.

For wall thickness of, e.g., 80 mm app. 60 % of the welding time is saved, since only 1/3 of the weld turns are necessary. Fig. 4.26 shows the volume for an app. 80 mm thick austenitic weld of a narrow slot welds versus a 'conventional' weld.

For that weld type full UT incl. Tandem Testing in addition to X-ray testing was required. The volume test could be performed with almost no problems by using SEL- Probes. Classical Tandem Testing failed due to high interface reflection.

The LLT- Technique and the neighbor echo 1- Method were compared at two 4 mm and one 8 mm tandem flaws. The reference flaws all lay at 40 mm depth and in different positions to the weld; the 4 mm flaw was in the parent material and at the weld flange, the 8 mm flaw was at the probe far weld flange (s. Fig. 4.27). Results were achieved as follows:

4.5.3.1 Classical Tandem Testing

• In the parent material a 4 mm circular reflector was detected with 18 dB S/N ratio.
• The interface (parent material - weld flange) reflected app. 10 dB - better than the 4 mm circular reflector!!!
• The through transmission of the weld is clearly not possible.
4.5.3.2 LLT-Technique
• In the parent material a 4 mm circular reflector was detected with app. 12 dB S/N ratio.
• The interface reflected app. 4 dB less than the 4 mm circular reflector
• The through transmission of the weld is clearly not possible.
4.5.3.3 Normal Probe - Mode Conversion Technique (Neighbor Echo 1)
• In the parent material a 4 mm circular reflector was detected with 8 dB S/N ratio.
• The interface reflected app. 4 dB less than the 4 mm circular reflector
• The through transmission of the weld with its aim to detect the 8 mm circular reflector, shows app. 12 dB S/N ratio in relation to the interface indications.

The present testing task was later performed with the help of the LLT- Technique applied to the parent material and weld flange testing using the Neighbor Echo 1- Method in conjunction with the direct longitudinal wave for the weld volume [170].
Fig 4.27: Comparison of Tandem- and equivalent Tandem technique; reference body wall thickness: 80 mm

<< ..... < ..... >