WRAP_THESIS_Charlesworth_201100082

							    This is in contrast to the results from experiment 9, section 7.4.5, where no consistent

    responses were acquired from any of the serrations or defects other that notch E.

        The variation in the active wedge angle due to errors in the design and modelling of the

    wedge were more problematic using the convex wedge, as seen in Figure 7-22 and Figure

    7-25. The angle required at the centre of the wedge should be 36° compared to the selected

    law parameter of 30° used for the scan. The angles required toward the outer limits of the

    wedge were in the region of 26° and the sector scans were skewed as a result. However the

    error did not prevent positive detection of the defects and was only significant towards the

    ends near notches B and G which are covered by a complementary inspection from the

    platform, see section 4.4.5.


7.5 Conclusions

        The aim of this chapter was to experimentally validate the function of the continuous

    wedge in producing the appropriate skew deflection for any position around the extent of the

    aerofoil. In that respect it has been shown that the wedge performed as designed and

    produced positive detection of all the defects in range of the scan; importantly no skewing of

    the phased array probe was required to maximise responses from notches in the reference

    blade.

        It is noted that the experiments were in no way designed to validate the sizing capabilities

    of the technique; instead they were able to show that the relative sensitivity to the EDM

    notches at various positions around the blade root was improved. It can be seen however that

    the axial extent of defects could be measured with reasonable accuracy and that the 5mm

    long EDM notches resulted in 6 dB drop of sizes within 1mm. The through wall extent of

    0.5 mm deep notches would be difficult to accurately measure and would be estimated based

    on an amplitude relative to reference notches, however defects in excess of 1.5 mm might be

    assessed using tip diffraction measurement techniques to a high level of accuracy.

                                                                220
  Thesis – Ultrasonic phased array testing in the power generation industry – Novel wedge development   November 2011
      The results of experiments 5 and 6 confirmed that the skew deflection parameters had

  been met, whilst taking account of the wedge attenuation from the results of experiment 7,

  the sensitivity to each notch compared favourably to the baseline results attained in

  experiments 1 and 2 (section 5.3.4).

      Experiments 8 and 9 clearly illustrated how enacting encoded line scans of the aerofoil

  using traditional wedges led to inconsistency of coupling, poor sensitivity, and in the case of

  the convex wedge no detection of all but one notch in the concave root. Experiments 10 and

  11 went on to prove that the continuous wedges facilitated accurate and repeatable line

  scans, detecting all notches in range of the aerofoil scans. A significant feature of the

  recorded line scans was the repeatable and continuous responses from the root serrations,

  further confirming that skew deflection was taking place and the appropriate normalisation

  of the ultrasonic beam was achieved.

      The experiments carried out in this chapter did however illustrate a fundamental error that

  had been made in the design and modelling of the continuous wedges. It was found that the

  active refraction angle, designed to produce refraction of 36° along the full axial extent of

  the wedge, had not taken enough account of the variation in the slope of the aerofoil relative

  to the platform. The zero degree datum used to set the active angle during the modelling was

  taken through the vertical centre line of the blade; this was accurate in the centre of the

  blade aerofoil but was inaccurate towards the ends. Figure 7-26 and Figure 7-27 illustrate

  the slope of tangents taken at the incident point for different sections around the convex

  aerofoil and shows that the 0° datum from which the active 36° angle was measured was up

  to 8° out. However, despite this error the prototype wedges performed as designed in all

  other respects and simple changes to the model would rectify the issue.




                                                              221
Thesis – Ultrasonic phased array testing in the power generation industry – Novel wedge development   November 2011