This procedure can be used as a general procedure to calibrate the ect MAD 8D eddy current system to inspect non-ferrous heat exchanger tubing for internal pits using the vertical volts method. The vertical volts method is considered to be the best method for measuring internal pits and is especially useful when the internal surface of the tube has thousands of pits with overlapping signals.
This procedure was written using the example of calibrating to inspect a 1" x .049 inch wall 90/10 copper nickel tube.
The vertical volts method uses lower frequencies than are used for phase analysis. The technique relies on having some knowledge of the pit size and relies on the assumptions that the pits to be measured are created by one mechanism and have the same depth to diameter ratios for any given depth of pit.
In the case of this example, a tube as described above had been removed from service. Pits were visibly evident on the inside surface of the tube. Eddy current testing was used to locate the deepest pit available close enough to the end of a tube that it could be measured mechanically. The pit was measured mechanically and was found to be .004 inch deep or eight per cent of wall. The diameter of the pit was estimated at one quarter inch diameter. As a calibration standard, a one quarter inch diameter (same diameter as the pit) 100 per cent hole was drilled through the wall of the tube. This pit is used for the 100 per cent of wall reference. (In an actual application, it will be necessary to use a hole the diameter of the pits to be measured as the calibration defect in the calibration tube.)
For background information on this technique, please refer to the AUTOMATIC SIZING OF INTERNAL PITS IN HEAT EXCHANGER TUBING USING EDDY CURRENT Report, the EDDY CURRENT PIT SIZING TECHNIQUES FOR COPPER ALLOY TUBING Report, the EDDY CURRENT PIT SIZING: REVISITED Report, the FOURTH EPRI BALANCE OF PLANT HEAT EXCHANGER NDE SYMPOSIUM Report, and ELECTROMAGNETIC NDE GUIDE FOR BALANCE OF PLANT HEAT EXCHANGERS, Revision 1, Section 3.1-1, Recommended NDE Procedures for Non-Ferromagnetic Tubing, 90/10 Copper Nickel Prime Surface Tubing published by the EPRI NDE Center.
Follow the sections for probe selection, cable connections, and selecting frequencies from the ect MAD 8D Calibration Procedure.
In this example, the wall thickness of the tube is thinner than in the example given in the ect MAD 8D Calibration Procedures. The frequencies selected in this section are: Channel 1, 64 kHz; Channel 2, 32 kHz; Channel 3, 64 kHz; Channel 4, 16 kHz.
Follow the procedure given in the ect MAD 8D Calibration Procedure for frequency channel phase adjustment.
FREQUENCY CHANNEL GAIN ADJUSTMENT
As discussed earlier, the defects to be measured in this example are very large, with a diameter of about one quarter inch; therefore, a comparatively low gain setting is required than in the general calibration procedure. Set each of the differential channels 1, 2, and 4 to .25 volts per division and adjust the gain controls so that the amplitude of the signal from the through wall hole is one volt peak-to-peak minimum. Note that due to the clearance between the outside diameter of the probe and the inside diameter of the tube, the distance between the probe coils and the defect will vary each time the probe is pulled past the defect. For this reason, the amplitude of the signal will vary somewhat.
When the through wall hole signal has been adjusted to one volt, it will be four divisions, or one half the screen size. This meets the requirements of the ASME Code.
For the absolute channel, Channel 3, adjust the 20 per cent I.D. groove in the ASME Calibration Tube to be five or six volts in amplitude. A high volts per division screen sensitivity (low volts per division number) may be used to meet the requirements of the ASME Code if this is a requirement of the inspection; however, it is more useful to use a low screen sensitivity so that large defects will remain on the screen in the Absolute Channel.
The vertical volts method relies on using relatively low frequencies; therefore, select Mixer 1 to mix the signals from Channels 2 and 4, the 32 kHz and 16 kHz differential channels. Follow the procedure given in Mixer Adjustment in the ect MAD 8D Calibration Procedure to cancel the signal from a support plate and to adjust the mixer output phase control.
MIXER GAIN CONTROL ADJUSTMENT
Set the mixer volts per division to two volts per division and adjust the mixer gain control so that the signal from the one quarter inch hole in the calibration tube almost, but not quite, fills the screen.
From the Analysis Menu, type <C> to enter the Calibration Table Menu. Select Table 1 and clear it by typing a <C> followed by a <Y>. Type the letter <P> one, two, or three times as necessary to set the parameter to vertical volts.
Return to the Analyze Menu by typing <A>. Note that the first line of the menu reads ANALYZE: V MAG indicating that the analyzer is in vertical volts mode. Pull the probe through the calibration tube to obtain a signal from the one quarter inch hole. Use the Analyze function to measure this defect in Mixer 1. Record the reading for the one quarter inch hole. In this example, the value was five volts. It may be necessary to set the Don't Analyze Less Than to zero per cent and the volts to zero in the Setup Menu in order to make this measurement.
Move to the Calibration Table Menu by typing <C>. Enter the values of 0, 25, 50, 75, and 100 as the top four lines in the per cent wall column. In column M1 (for Mixer 1), enter 0 for zero per cent wall and the value measured for the one quarter inch through wall hole for the one hundred per cent value (In this example, five volts). Note that as volts are recorded in this table to a resolution of .1 volts, it is necessary to type <5> <0> <Enter> to enter the value of five volts in the M1 column.
The Table is now set up for straight line linear interpolation of defect sizes with the values entered; however, in real life, this function is a curve and not a straight line. The best way to determine this curve is to build a calibration tube that has 25, 50, and 75 per cent I.D. pits with a diameter of the actual pits that are to be measured (In this example, one quarter inch). As this is difficult to do, you may follow this procedure to approximate this curve as follows: Multiply the value for 100 per cent by 60 per cent and enter this value for the 75 per cent wall loss. In this example, that value is three volts. Multiply the value for 100 per cent by 40 per cent and enter this value for the 50 per cent wall loss. In this example, that works out to two volts. Multiply the value for 100 per cent by 20 per cent and enter this value for 25 per cent wall loss. In this example, that works out to one volt.
Collect data in the usual manner. Set the AutoMAD features as desired.
In the Analyze Menu, make sure that Table 1 is selected and that Mixer 1 is selected for analysis. Analyze the data in the usual manner.
In this example, the eight per cent (.004 inch deep) pit that was physically measured close to the end of the tube measures as eight per cent wall loss. The deepest pit in the sample of tube that was tested measures eighteen per cent wall loss.