4.6 Test Results Analysis and Crack depth Determination

The state of the art is to evaluate the UT results by comparing the echo amplitude with that of a known reference flaw amplitude. That only allows us to determine the defect position and length, whereas the description of the true defect geometry using the amplitude method is not possible. Neither the type of defect, nor its reason for being, nor its depth distribution can be obtained from the amplitude result, so that further judgments must necessarily be very conservative. That means indications which are small in relation to the critical reference flaw should not be considered, Fig. 4.28 [63].

Fig 4.28: Principles of flaw detection and evaluation with ultrasonic testing Registration of Reflectors: Echo amplitude exceeds the reference amplitude
Testing method Test scope Test sensibility	<=	Standard Testing Guideline Defect catalog	=>	Flaw detection
Quantification of Reflectors: Quantification of reference amplitudes
Spot (A) Length (d) Number of defects	<=	Standard Testing Guideline	=>	Flaw Detection

Nevertheless, the required level of safety can be achieved with this method and cracks, and their length can be detected, but it will mean increased testing time and costs. It is also very possible that unnecessary repairs are done.

Reasons for the difficulty of crack depth evaluation with ultrasonics include the variety of defect geometry and configurations which exert very complex influences on the ultrasonic result. There are many well known measurement methods for crack depth determination [66, 93], none is valid taken alone. Therefore, crack depth must be measured manually during result analysis. The work can be standardized if the required crack type is known, however the probability of detection also relies on inspector qualification. One example of such standards is the currently accepted general guideline for depth determination of Intergranular Stress Corrosion Cracking [33].

Crack depth measurements, which are verified by means of results analysis, can be very labor intensive. This means that some ultrasonic test methods need to be applied, which are possibly supported by other nondestructive testing methods. Possible defect geometry can be predetermined by use of construction or process documents of the component to be tested, Fig. 4.29. In general, knowledge of theoretical defect configuration, its effect on ultrasound, and the specific ultrasound propagation in anisotropy material are preconditions for the development and adaptation of the testing technique.

Fig 4.29: Process in case of crack suspected result. Fig 4.30: Contour projection Fig 4.31: Determination of Reflector location

4.6.1 Test Result Analysis

During the evaluation of ultrasonic test results, in the case of indications, they occurred as a result of:

• Cracks
• Inclusions
• Object indications
• other reflectors?

Before answering these questions it is first important to determine exactly the location of the indication in relation to the weld. Test documentation is not conclusive, so results analysis as described here is necessary.

For the present indication of austenitic piping evaluation is needed. The results analysis is divided in two parts:

• the determination of the components' contour projection and
• the depicting of sound beams to the reflector in the contour projection.

The contour projection is performed by use of a comb reprint of the outer surface. At the same positions of the comb reprint the wall thickness was measured with a 0°-Probe. As Fig. 4.30 shows, it was possible to determine the contour projection.

Then, using an angle beam probe (56° SET 1,5 MHz), the parameters 'probe position' and 'time of flight to the reflector' were determined and depicted in the contour projection, Fig. 4.31.

This time of flight analysis determines the indication location at the probe's far surface flange (indication of the echo after through transmission of the weld root). Crack-sensitive reflectors show another behavior; finally the indications can be interpreted as root indications. For confirmation of these results a neighbor echo 1 Probe was used and its result couldn't prove the indication of a crack either.

4.6.2 Crack Depth Determination

After a results analysis has confirmed the existence of cracks, the cracks can be measured. The important results should be obtained using defect edge reconstruction methods and planar sensitive methods, (e.g., with Neighbor Echo 1- Method).

Fig 4.32: lack of side wall fusion defect (surface breaking) Fig 4.33: fatigue crack

One element of this strategy was to perform the test with an adequately high redundance, since there are problems with defect edge reconstruction methods because of the low diffraction amplitude. As examples of this, Fig. 4.32 and 4.33 show a lack of side wall fusion defect (surface breaking) and a fatigue crack.

In cases of difficult defect edge determination, e.g., for intergranular propagation, all possibilities should be used to verify the crack tip measurement.

Methods which use the crack 's plane as reflector are applicable, however the application of diagrams for evaluation of crack depth is inconclusive because the geometry of the reference reflector, (e.g., notch), strongly relies on the crack geometry. Nevertheless, with those measurements and diagrams it is possible to obtain important qualitative results for the determination of crack depth, Fig. 4.34.
Fig 4.34: Evaluation diagrams; Wall thickness for all diagrams: 20 mm.

Fig 4.35: Crack Tip Detection Techniques Shadow technique (SE) Phase Evaluation (A-scan display - Radio Frequency (RF) ) DELTA-Technique (SE) ADEPT 60 (integr. Tandem technique) Topic: Focused Longitudinal Wave

Some crack tip detection techniques are shown in Fig. 4.35. For evaluation of small cracks, up to 6 mm depth, high damped probes must be used. Those probes make it possible, with their high axial resolution, to clearly separate the crack root indication from the crack tip indication. Such extremely short pulses also make it possible to evaluate the phase of the crack tip echoes, Fig. 4.35.

Therefore, the sound beam width must be big enough to reflect at the crack root and crack tip at the same time, so that both echoes are displayed in the A-screen and the tip echo can be relatively evaluated to the root echo during probe movement.

The detection of crack tips is uncertain. In some cases, evaluable amplitudes are indicated, but for others no indication is visible. An important parameter seems to be pressure of the crack. [169].

4.6.3 Detection of Natural Cracks

Figs 4.36 and 4.37 present the crack detection and crack depth determination of disassembles piping, which contains Intergranular Stress Corrosion Cracking (IGSCC²).

The crack location is interesting; it lies directly at the counterbore and caused the failure of the Neighbor Echo 2- Method (see Fig. 4.7). However, when the 48°-Probe and 60°-Probe is used the crack is clearly detected. Fig. 4.37 illustrates methods for crack depth determination and evaluation.

Fig. 4.38 documents crack detection for non-assembled piping. The radiation and the near inaccessibility of the weld were strong hindering circumstances. The ultrasonic result was radiographically confirmed and the weld was renovated with an overlay.

Fig 4.36: Crack Detection Fig 4.37: Crack Tip Detection Fig 4.38: Documentation of crack detection (Examination Contour)

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