RFA Application Guide

252
INTRODUCTION
This document presents how to use an RFA probe with Magnifi 5.1 on an Ectane 3 instrument.

EQUIPMENT
The RFA probes use a 41-pin connector that can be connected to an Ectane 3 with the
‘’E64RNM’’ or the ‘’E128RNM ‘’ option.
Magnifi acquisition and analysis software version 5.1 or greater must be used. A license allowing
for tubing array techniques is required.
Possible probe diameters are offered from 11.0 mm to 23.0 mm in 0.5 mm increments. It is
recommended to use the largest diameter possible. This will decrease the lift-off and improve
performance. You can refer to the table below for the recommended diameters depending on
the size of the tube to be inspected.
The center frequency of the probe is 300 Hz and it should be used with frequencies ranging
from 50 Hz to 2 kHz.
From the following table, the best probe for your application can be selected.

The RFA calibration tube used in this document includes the following machined defects and
features:
• External grooves, 20%, 40% and 60% of wall loss
• Flat bottom hole, 60% of wall loss
• Hole, 100% of wall loss
• A support plate

253
RFA PROBE DESCRIPTION & CHANNELS
The RFA probe includes 2 drivers at each end of the probe head. The driver on the cable side is
named the Lead since it is the first to pass under the defects and features when the probe is
pulled through the tube. The second driver is called the Trail.
The probe includes 2 circumferential bobbin coils that create absolute and differential channels,
referred to as “conventional bobbin channels”, similar to what is obtained with an RFT probe.
The absolute and differential signals from the bobbin coils will produce Strip charts, Lissajous
and Voltage planes.
Next to these 2 coils is the array portion of the probe. It consists of 2 rows of sensors. The number
of sensors per row depends on the diameter of the probe. The second row on the cable side is
used to create the absolute array channels. The differential array channels are obtained by
subtracting the signals of row 1 from row 2. As with conventional channels, the absolute array
is used to analyze large volumetric defects, such as taper defects for example, and the
differential array is used to analyze smaller defects such as pits. The channels of the array
component of the probe are combined to produce C-Scan imaging.

Note that neither the conventional coils nor the array part is perfectly centered between the 2
drivers.

254
PROJECT AND INSPECTION FILES
In this section, we will create a folder structure that will manage the saving location of your
setup, data, and report. This management is operated through the creation of a Project.
Magnifi suggests two levels of files. The first file level is the Project. It is meant to include
different inspections that are related in some way. For instance, it can include all the inspections
done at a refinery shutdown and be named Refinery_Shutdown_May_2022. The second file level
is the Inspection folder. Inspection folders are saved in the project file. An inspection folder can
include the data specific to the inspection of a tube bundle with a specific technology and could
be named CS_075x0.083_RFA, for instance. This inspection folder groups the setup, the tube
list, the data files and the Magnifi report.

PROJECT
(ex: Refinery 04-17)

INSPECTION 1
INSPECTION 2 INSPECTION X
(ex: Bundle No A1D3FX)

... ...
SETUP

TUBE
DATA
LIST

REPORT

255
When you open Magnifi, the first page displayed is called the Backstage.

To create or open a project, click on Open/Add Project in the Backstage. You can select an
existing project/folder, or you can create a new folder.
5. Create a folder by right-clicking on the location where you want to add your project
file. Select New, Folder and enter the chosen name. You can then select the newly
created folder and click on Select Folder.

256
 6. Click on Add/Open Inspection in the backstage, then click on Add and enter the
name of your inspection.

7. Click Select. This will define the location where the setup(s) and data will be saved.

RFA SETUP WIZARD

In this section, we will show how to create a setup using the RFA Setup Wizard in Magnifi.
Before going further, you can select the appropriate measurement unit. To do so, click on
Preferences. In the System tab, you can change the measurement units from metric to imperial
and vice versa. For the metric units, you can choose to use meters, centimeters, or milimeters
in the General tab. For imperial units, inches must be used. When finished, click on Preferences
again to go back to the General window of the Backstage menu.

257
To create a new setup, use the RFA Setup Wizard process. Go to the Frontstage by clicking on
the blue triangle at the top left corner of the Backstage. Then, click on the New button under
the Setup tab and select the RFA Setup option in the drop-down menu.

PROBE SELECTION
The first page shown by the RFA Setup Wizard is the Probe Selection page.
In this window, the probe used for the inspection can be selected, the interaction between the
bobbin and array channels can be set, and the display and settings for the absolute channel
can be changed.

258
CALIBRATION OF C-SCAN IN CAL. TUBE
In the software, the channels generated from the bobbin coils and the C-scans produced by the
array are coupled, making the use of the RFA probe very similar to the RFT probe. Adjustments
made to the conventional channels, such as the normalization of a support plate signal, are
automatically propagated to the C-scans, allowing for easier analysis with a limited number of
operations in the software.
The Calibration of C-scans option offers 2 possibilities:

• Automatic: With this option, the probe is used the same way as an RFT probe would be
used. The initial calibration of the conventional channels made in the calibration tube is
automatically applied to the C-scans. Calibration of the C-scans independently from
the channels is not possible.

259
 • Manual: With this option, the C-scans must be calibrated independently from the
conventional bobbin channels for the initial calibration in the calibration tube. It must
be understood that once the initial calibration has been performed, any subsequent
change to the rotation or gain made to the conventional bobbin channels to adjust the
nominal point, for instance, will be automatically applied to the C-scans as well. This is
true for both options.

These options will be explained in more detail in the calibration section of this document.

REPRESENTATION FOR ABSOLUTE CHANNELS (ABS)

The Representation for absolute channels (ABS) drop-down menu allows for the selection of
either a standard Lissajous plot or a Voltage plane for the bobbin absolute channels. Note that
this option does not affect the representation for the Absolute C-scans, which always use a
Lissajous representation. The Voltage plane tool is explained in the appendix.

PROBE SETUPS
In the Probe Setups section of the window, select the probe model used for inspection. The
numbers inside the parentheses correspond to the range of probe’s diameter covered with a
specific setup. Refer to this information to find the appropriate model. For example, if an
18.0 mm mid-frequency probe is used, the PRBT-RFA-DDSA-(180 to 190)MF setup should be
selected.

COMPONENT DEFINITION
The second page shown by the Setup Wizard is the Component Definition page.

260
Click on the Material field to open the drop-down menu. Select the material of the tube to be
inspected.
If the inspected material is not on the list, go to Backstage and open the Material Database
tab. To add a new material to the list, click on Add. A new line will appear in the list. Give the
material a relevant name and set the material resistivity and permeability to its theoretical
value. The ultrasound velocity is used to set IRIS parameters only, and doesn’t need to be set if
an IRIS inspection is not performed on this material. It is possible to add optional notes about
the material to specify information such as its application or composition.

261
Also in the Component Definition page, adjust the tube wall thickness, outside diameter and
length by entering the values in the designated fields. All the information filled in on this page
will be used by the software when generating the analysis report.
The tube length defined here is also used as a default suggestion when enabling the automatic
detection of landmarks.
Click Next when everything is set correctly.

SCAN DEFINITION
The Scan Definition window is used to configure the data acquisition rate of the instrument,
the typical probe pulling speed, and the acquisition size.

262
The acquisition rate corresponds to the number of acquisition points, or samples, taken per
second. By default, the acquisition rate is set to 500 Hz for RFA, but depending on the drive
frequency used to inspect, the actual acquisition rate may need to be changed.

The recommended probe pulling speed for RFA depends on the excitation frequency, as defined
in the Data Definition page of the Wizard, and is shown in the table below:

Excitation Frequency (Hz) Maximum pulling speed

Minimum Maximum Imperial (in/s) Metric (mm/s)
20 100 4 100

100 200 6 150

200 350 8 200

350 500 10 250

500 1000 12 300

1000 - 20 500

263
The density, or axial resolution, is the number of samples per unit of distance. This field is non-
editable as it is a direct result of the acquisition rate and pulling speed combined. Use this
information to confirm that both parameters are defined adequately.
A resolution of 2 points per millimeter is typically targeted. Given that the probe speed value is
based on the excitation frequency, the acquisition rate can be edited to obtain the desired axial
resolution.
For instance, if the pulling speed was set to 300 mm/s, then the acquisition rate should be set
to 600 Hz to reach a resolution of 2 points per millimeter.
Note that if a probe pusher-puller mechanism is not used, the pulling speed will not be constant.
Therefore, it is recommended to target a lower pulling speed to be able to reach the axial
resolution target.
These parameters also serve another purpose. When analyzing a data file where landmarks are
not enabled or defined yet, the axial position of the cursor in a strip chart is defined assuming
a constant pulling speed. Therefore, if the cursor is located at exactly 4.0 seconds, and a probe
speed of 300 mm/s was used, then Magnifi will indicate a position of 1200 mm. This information
can be used to axially locate features and defects in the tube.

The last parameter that can be edited on this page is the Scan Size or scan duration, in units of
time. The non-editable field below shows the same parameter in units of distance instead. The
scan duration should be set to allow for acquisition of slightly longer than the tube length. This
additional buffer time will ensure that the full tube length is captured even if there are some
pulling speed variations and to provide enough time to insert the probe in the tube and
completely pull it out in the air. For instance, if 30 seconds is needed to exactly capture the full
9 m length of tube, then a 40 second acquisition time could be set.
It is important to note that the acquisition size should not be set more than what is needed, as
this could result in software memory issues.
Click Next when finished.

DATA DEFINITION
The Data Definition window is used to set the hardware gain, excitation frequencies and drive
voltages (or amplitude). It is important to set these parameters correctly before acquiring the
data, since they are driven by the instrument and cannot be modified during the analysis.

264
The hardware gain is applied to both the bobbin and the array portions of the probe.
Two frequencies are always enabled with RFA setups. The table in this window can be used to
edit the frequencies and drive voltages values. The sum of their amplitudes cannot exceed 10 V.

Selecting an appropriate frequency can be aided with the following equation:

40.5 ⋅ 𝜌
𝑓= ,
𝜇𝑟 ⋅ 𝑡 2
where 𝑓 is given in kHz, 𝜌 is the tube’s resistivity in 𝜇Ω ⋅ cm , 𝜇𝑟 is the relative permeability, and
𝑡 is the thickness in millimeters. It is important to make sure the frequency selected is within
the probe limits found in the table above. Another point of consideration when selecting the
operating frequency is the frequency of the power transmission lines and outlets in your global
region of inspection. It is recommended to avoid the harmonics of the outlet frequency to
prevent noise issues. The figures below show the frequency ranges for 60 Hz and 50 Hz.

265
 BW = 8 Hz BW = 10 Hz BW = 15 Hz BW = 20 Hz BW = 25 Hz
SPD = 0.1 m/s SPD = 0.15 m/s SPD = 0.22 m/s SPD = 0.3 m/s SPD = 0.35 m/s

30 50 75 85 135 165 195 225 255 290 310 355 365 500

0 60 120 180 240 300 360 420 480 540 600

Frequency (Hz)

Frequency range to avoid Frequency range allowed

Recommended frequencies for power outlet at 60 Hz

BW = 8 Hz BW = 10 Hz BW = 15 Hz BW = 20 Hz BW = 25 Hz
SPD = 0.1 m/s SPD = 0.15 m/s SPD = 0.22 m/s SPD = 0.3 m/s SPD = 0.35 m/s

30 40 60 85 115 135 165 205 215 240 260 295 305 500

0 50 100 150 200 250 300 350 400 450 500 550 600
Frequency (Hz)

Frequency range to avoid Frequency range allowed

Recommended frequencies for power outlet at 50 Hz

By default, the RFA probe operates as a dual-driver probe. However, it is possible to turn one
driver off to use it as a single-driver probe. To do so, select the Lead or the Trail option in the
Enabled Drivers drop-down menu. The Lead corresponds to the driver that is close to the cable;
this is the first driver to reach a feature when the probe is pulled. The Trail is the driver closest
to the tip of the probe.
If you choose to use the probe in single-driver mode, the Setup Wizard will automatically divide
the drive voltage values in half for both frequencies, and the amplitudes will no longer be
editable. A warning will also appear, indicating that these values were adjusted to generate
approximately the same driver output.

266
This is done such that the driver signal has approximately the same amplitude, regardless of
whether dual-driver or single-driver mode is used. An example is given below, where the
amplitudes of the driver signals have roughly the same amplitude (shown with the red dotted
line) for both mode settings. Additionally, it can be noted that the amplitude of the receiver
signal, when the defect passes right under the receiving coil, has decreased by half as expected.

267
Should you need to change the drive voltage in single-driver mode, select the option Both, edit
the voltage value and then select the option Lead or Trail again.
Click on Next when the desired parameters are entered.

DATA PROCESSING
The Data Processing section is where post-processing software filters can be activated on the
channels of the RFA probe.
The hardware parameters that were defined in the Data Definition page of the Wizard cannot
be modified once the acquisition is completed. Therefore, it is crucial to validate those values
before performing data acquisition. In the case of the Data Processing page, the available
parameters allow you to enable and edit post-processing filters. Therefore, it is always possible
to edit these parameters to fine tune the analysis after the data acquisition if needed.
The first page of the Data Processing section presents the list of available C-scans and their
filters.

The following table describes each of the C-scans:

268
 Name Description
Frequency 1, Absolute Array, Large Signal. Dedicated C-scan that
F1ABSA-LargeSignal should be configured (filters values and color palette range) to
optimize the analysis of large indications such as taper defects
Frequency 1, Absolute Array, Trail row of sensors (farthest from the
F1ABSA
cable)
F1DIFA Frequency 1, Differential Array
Frequency 2, Absolute Array, Large Signal. Dedicated C-scan that
F2ABSA-LargeSignal should be configured (filters values and color palette range) to
optimize the analysis of large indications such as taper defects
Frequency 1, Absolute Array, Trail row of sensors (farthest from the
F2ABSA
cable)
F2DIFA Frequency 2, Differential Array

Three types of filters are available. The low-pass filter eliminates part of the signal that is above
a certain frequency. As an example, it is useful when a defect signal has a lower frequency
content than the background noise. In this case, using a filter will remove part of the noise
without removing the defect’s signal. This may help with data analysis. However, a cut-off
frequency that is too high won’t remove much noise, and a cut-off frequency that is too low
will filter out the defect signals.
The median high-pass filter is used to filter out low frequency noise or drift such as lift-off
variations of the probe within the tube, or changes in material, geometry, or thickness. As a
rule-of-thumb, the window size of the high-pass median filter should be set to at least three
times the axial length of the longest flaw that may be encountered. This rule-of-thumb is
specific to what you are looking for with the channel you are using. If, for example, you plan to
use the differential channel to detect holes, you can adjust this setting to about three times
the length of the largest hole you are looking for with this channel. The window size should be
adjusted with care. If it is set too small, it could filter out defects of interest. Note that the size
of the filter is set in samples and not in mm. For example, if you have set your parameters to
have 2 samples/mm and you want to enter a length of 300mm, then you should use a value of
600 samples.
Pilger noise 1 is yet another type of noise that can be observed on tubing data that can
negatively alter the amplitude and phase of the signal. The Periodic filter can be enabled to

1
Pilger noise is caused by the cold pilgering process in which the diameter and wall-thickness of a tube
are reduced to form seamless metal tubing. Pilger noise is typically seen as spiral-shaped variations on
a C-scan. Because of the helical shape of pilger noise in the tube, the noise is somewhat periodical, but
not constant.
269
reduce the effects of this phenomenon. Similar to the other filters, there is always a trade-off
between the noise reduction level and the side-effects of the filter on the indication signals
themselves when adjusting the parameters of the filter. It is not recommended to apply this
filter by default, but it can be used on tube data that contains a strong pilgering signal.
More details about the adequate configuration of this filter can be found in Appendix 2.

The second page of the Data Processing section presents the list of available bobbin channels
and their filters.

The following table describes each of the channels:

Name Description
F1ABS Frequency 1, Absolute Bobbin
F1DIF Frequency 1, Differential Bobbin
F2ABS Frequency 2, Absolute Bobbin
F2DIF Frequency 2, Differential Bobbin

270
Low-Pass and Median High-Pass filters are available for these channels. The comments made
for the C-scan filters above can also be applied with the channel filter.
Data should be examined in its filtered and unfiltered states. It is important to keep in mind
that the filters can distort the phase and change the amplitude of the signals.
Click Next when finished.

INDICATION CODES
The Indication Codes page is used to define the entries that can be added to the report when
analyzing the data.

When an indication is added, its abbreviation (code) is shown in the code pane, next to the
data.
You can modify the default indication code list by changing the parameter in the table. New
indications can be defined by clicking on the Add button.
Click Finish to complete the RFA setup wizard process.

271
SETUP MODIFICATIONS
Some parameters or preferences may need to be modified after completing the Setup Wizard
process. To modify a parameter previously entered, go to the Setup tab in the Frontstage and
click on the button associated with the parameter you want to change.

This will open one of the windows previously described, where you can change the values you
originally entered. If applicable, go through the process by clicking on Next, and then click on
the Finish button. This will apply the modification to the setup.

It is important to note that the parameters defined in the Probe Selection window cannot be
changed after the initial generation of the setup. If you wish to change the probe model, the
Calibration of C-scans option, or the representation for the Absolute channels, you must
create a completely new setup.

Advanced setup modifications are also possible in the Advanced Setup section on page 310.

DATA LIST
Magnifi will save a file for each inspected tube. The file names are defined by creating the list
of tubes.
This list can be created in the Data List section of the General tab of the Backstage.

Four options are available to set the filename format:

1. Free format:
Each file has a custom name. Can also be defined from the Data tab of the Front
Stage.

272
 2. Prefix:
The file name includes a defined prefix followed by a sequential number.
3. Row, Col:
Row number, Column number. Mostly used for tubing inspections.
4. Zone, Row, Col:
Zone number, Row number, Column number. Mostly used for tubing inspections.
Click on the Create New List button. The displayed window will be different depending on the
chosen filename format.
For the Row, Col option, enter the number of rows, the starting row number and row interval.
You can then enter the number of tubes per row in the Nb. of Columns fields of the table. Click
on Create to generate the list of tubes. You can also use the Create & New button to add
another set of tubes to your list.

The same principles apply to the other file formats, except for the Free format option for which
the file name(s) needs to be entered manually in the Data window of the Frontstage.
273
The tube list can also be imported from another project using the Import List button. The tube
list file can be found in the Inspection folder. It is also possible to import a list created in the
Tubepro software (available from Eddyfi).

PERFORMING AN ACQUISITION
1. If you are in the Backstage, move to the Frontstage by clicking on the Start/Resume
button.
2. Click on the Connect button under the Instrument tab. This will open the Instrument
Connection window. Click on the line showing the instrument to which you want to
connect and then click Connect.

Note: Your Ectane firmware version may not match the version of Magnifi you are using. If this
is the case, a red and white X icon will be shown in the Status field of the Instrument connection
window. To send a matching version to your Ectane, click on the Config. button and then click
the Send firmware to the instrument button of the Instrument configuration window.

274
Two (2) acquisition modes are available in Magnifi: Setup Mode and Acquisition Mode. Setup
Mode is used to scan your calibration tube and make the necessary adjustments on your setup
without saving the data automatically. Acquisition Mode is used for the inspection. In this mode,
the software automatically saves the acquired data using file names based on the tube list.
Follow the steps below to perform the acquisition of the calibration standard:
1. Enable Setup Mode by toggling on the Setup Mode button under the Home tab. This
mode is active when the button is grayed.

2. Plug the RFA probe into the Ectane 41-pin connector.

3. Bring the probe head in a clear area in your standard and balance the probe by
clicking on the Null button under the Home tab, or press F6 on your keyboard.
4. Push the probe head all the way through the tube so that it exits the tube.
5. Start the acquisition by clicking on the Acquire button under the Home tab, or by
pressing F2 on your keyboard.
6. Pull the probe at the recommended speed.
7. When the probe exits the tube, press the Stop button, or again F2 on your keyboard.

The image below shows a typical RFA scan of a calibration tube. No calibration has been
performed yet.

275
Note that a red zone in the code pane means that the raw signal from at least one of the
channels is saturated. This is usually the case when your probe is not inside the tube.

VISUALIZING THE DATA

LAYOUT
By default, the setup generated with the RFA Wizard offers a layout that includes the following
elements:
• 2 strip charts and Lissajous/Voltage plane corresponding to the bobbin (ABS, DIF) coils,
similar to the conventional RFT setups generated with the Generic Wizard

276
277
 • 2 C-scans and 2 Lissajous corresponding to the array coils

The tools found in the Layout ribbon can also be used to adjust the display settings to suit your
preferences. Use the Replace, Remove or Split button to adjust the current display setting.

DISPLAYED CHANNEL
There are four buttons at the upper left corner of the Strip charts and Lissajous windows. These
buttons are used to adjust the display settings for a channel in the specified window.

1. Links Strip charts and Lissajous to the same channel. For instance, if a Lissajous and a
Strip chart are both set to 1, setting the Lissajous to DIF-F1 will also set the associated
Strip chart to this channel.

278
 2. Clicking on the corner with the black triangle gives the list of available channels. Click
on the desired channel to select it. Right-click or Left-click on this button to switch to
the following or previous channel in the list.

3. Same principle as 2., but for the type of channel only (absolute or differential).
4. Same principle as 2., but for the frequencies only.

STRIP CHARTS AXIS ORIENTATION

The Strip charts are projections of the Lissajous on the vertical and horizontal axis. To switch
from one axis to another, click on the box showing an arrow at the upper right corner of the
Strip chart.

ZOOMING
BOBBIN CHANNELS
On the strip chart, hold the right button of your mouse and drag on the zone of interest to
zoom in this section.

279
To zoom out, right-click on the Strip chart and select Previous Zoom or Full Extent.

280
ARRAY CHANNELS
On the C-scan, hold the right button of your mouse and drag on the zone of interest to zoom
in on that section. You can drag your mouse either horizontally or vertically on the C-scan.
To zoom out, right-click on the C-scan and select Zoom to Previous, Zoom to Content or Full
Extent.

ADJUSTING THE CURSOR LENGTH

On the Strip chart, the cursor is composed of 3 lines. The dotted lines represent the limit of your
cursor, and the solid line is the center of the selection.
Only the section bound by the dotted lines will be shown in the Lissajous.
Adjusting the cursor length is easily done using the mouse. Clicking and dragging on a dotted
line with the left mouse button will adjust the 2 dotted lines symmetrically. To adjust only one
dotted line, do the same operation with the right mouse button.

281
ADJUSTING THE SCALE
The scale of a window (Lissajous or Strip chart) can be modified by clicking on the scale button
with the left button (decrease scale) or the right button (increase scale) of your mouse.

Another way to modify the scale is to click and drag (up or down) the right button of your
mouse on the scale button.

MEASUREMENT METHOD
The buttons at the lower left corner of the Lissajous windows are used to select the
measurement method.

1. Clicking on the corner with the black triangle gives a list of measurement methods for
the amplitude of the signal. Click on the desired method to select it. Right- or Left-click
on this button to select the next or the previous method in the list.
2. Same as 1., but for the phase measurement.
3. Rotate the measured phase by 180°. This option can be used if the software doesn’t
measure the phase with the correct orientation.
4. Take the same two points in time to take the measurement in the other Lissajous.
5. Allow a manual measurement of the signal. Click and drag the left button of your mouse
to draw a vector on your Lissajous.

LISSAJOUS ROTATION AND PANNING

The signal in a Lissajous can be rotated by holding CRTL on the keyboard while holding the left
button of your mouse and dragging it around the rotation axis. Note that this operation cannot
be performed on the raw channels since these channels have no gain or rotation applied, by

282
definition. Rotating the signal will affect your calibration, so if you perform this operation, make
sure to recalibrate afterwards.
The origin point can also be moved by holding the left button of your mouse on the Lissajous
background and by dragging it in the desired direction.

DATA CENTERING
To center the data in the different windows, put your cursor at a point where you want the data
to be centered and press on the space bar on the keyboard.

C-SCAN DISPLAY ADJUSTMENT

Manual adjustments can be made to the C-scans to improve their display for the detection of
defects.

One way of changing the display of a C-scan is by changing the color palette type and/or
adjusting the scale of the color palette. To do so, double-click on the color palette ribbon at the
top of the C-scan.

Another way of adjusting the display of the C-scan is to manually rotate the signal in the
associated Lissajous. Below shows a C-scan with a pit signal circled in red. As shown by the

283
arrow in the upper right corner of the C-scan pane, the C-scan shows the vertical components
of the Lissajous.

On the Lissajous (below, left), the noise signal sits at approximately 60°. Manually rotating the
Lissajous can be done by clicking and dragging the left-mouse button on the field highlighted
in red. To minimize the noise signal, the Lissajous can be rotated such that the noise is horizontal.

After rotation (below, right), the pit signal becomes more prominent on the C-scan and the
noise in the nominal sections of the tube (green) has been reduced.

Caution: If you use sizing curves with the array part of the probe, it is necessary to readjust the
manual rotation to zero so that the sizing curves give the expected values.

SETUP ADJUSTMENT

FREQUENCY VALIDATION
The injection frequency (set in the Data definition section) will influence the phase angle
separation between flaws of different depths. It can also affect the signal-to-noise ratio of
certain flaws.

284
If this frequency is too low, the phase angles of your flaws will be close to each other. Therefore,
it will be difficult to build accurate sizing curves because a small variation of the phase will lead
to a significant change of depth estimate.
Multiple methods can be used to validate that the injection frequency was correctly set. Here’s
two of them:
• There is approximately 1° of phase change per 1% of wall loss between two grooves of
different depths. The two grooves must not be too close in terms of depth.
• There is approximately 65° of phase separation between a through-wall hole and a 4 x
FBH 20% depth.
To adjust the injection frequency:
8. Click on the Data button under the Setup tab.

9. To increase the phase angle spread between your flaws, increase the injection
frequency. Decrease the injection frequency to lower the phase angle spread.
10. Click on Next and then click on Finish.
11. Rescan your calibration tube.
12. Verify that your injection frequency is correct.
13. Redo steps 1 to 5 until you have an acceptable frequency.
NOMINAL PONT VARIATION TOLERANCE
When renormalizing the support plane signal in the Voltage plane, the angle correction applied
relative to the nominal point is displayed in the bottom left corner of the view. Magnifi offers
the possibility to display this correction value in red if it exceeds an acceptable value.

285
A common practice is to set the threshold angle correction value at +/- 30°. To change this
default value, click on the Voltage Plane window to highlight its borders and change the
Tolerance field under the Current View tab.

Outside this value, the inspection frequency should be adjusted. This is described in more detail
in Appendix 1 of this document.

CALIBRATION
Calibration is used to set the amplitude(s) and phase(s) of a channel using the selected
measurement method. Given the options selected in the Probe Selection step of the Setup
Wizard, the calibration of the bobbin channels can be performed with either the Lissajous or
Voltage Plane, and the array channels can be independently calibrated if the Manual link type
was chosen. All methods of calibration are shown below.

LISSAJOUS
1. On either the conventional absolute (ABS_F1 or ABS_F2) or differential (DIF_F1 or
DIF_F2) bobbin channel, position your cursor on the defect or feature corresponding to
the calibration reference. In the picture below, the through-hole is selected. Make sure
that the measurement vector includes the ends of the signal.

286
 Absolute Lissajous before calibration Differential Lissajous before calibration

2. Click the System button under the Calibration tab.

3. Select the Channel radio button.
4. Select the item that corresponds to your calibration reference from the list.
5. Then press the green arrow. This action tells the system that the previous
measurement matches the selected reference.
6. Click Calibrate. This will apply the gain and rotation necessary to bring the reference
signal to the desired value.

If you have calibrated the hole to 1V, 90°, the signal after calibration should look like the
images below:

287
 Absolute Lissajous after calibration. Differential Lissajous after calibration.

VOLTAGE PLANE

14. To adjust the signal, the user must manually draw a vector on the voltage plane from
the nominal point of the calibration tube signal to the point of full attenuation (the
effect of a completely blocked signal). Its length must initially be set to twice the length
of the support plate, in the case of a Dual-Driver probe. To draw this vector, click on the
manual measurement button "M", click on the operation point of the signal, hold the
left button of the mouse and drag it to the desired location. Once the vector is correctly
drawn, click on the normalize button (blue compass). Magnifi will put the origin of the
vector to the theoretical nominal point. It will apply a gain of 1/(vector amplitude) and
a phase rotation of –(vector phase). If, for instance, the drawn vector has an amplitude
of 0.8 and a phase angle of 10°, the system will apply a gain of 1.25 (1/0.8) and a rotation
of -10° to the system.

288
 15. The signal may not be perfectly adjusted after the previous steps. You can adjust the
signal again by drawing a new vector and by clicking on the normalization button until
the signal is set correctly. In the examples below, the calibration tubes included two 360°
grooves and we can see that both grooves reach the theoretical 360° coverage line. This
can be a relevant indication to determine if a voltage plane is calibrated correctly.

289
 16. Save your calibration by clicking on the green Save Nominal button.

ARRAY CHANNEL – AUTOMATIC LINK TYPE

With the Automatic link type option selected during setup, the settings made on the bobbin
channels will be copied to the array channels; meaning the gain and rotation applied to the
bobbin channels (absolute and differential) will be also copied to the array channels. Therefore,
no action needs to be taken to calibrate the array channels. The example below shows how the
array channels get calibrated side-by-side with the conventional bobbin channels.

290
Above, calibration has not been applied to either channel. Below, the calibration was applied
to the conventional bobbin channels which has been automatically pushed to the C-scan
channels.

291
ARRAY CHANNEL – MANUAL LINK TYPE

With the Manual link type option selected during setup, you must manually calibrate the array
channels according to the following steps.
1. On an array channel, position your cursor on the defect or feature corresponding to your
reference for your calibration. In the image below, the through-hole is selected on the
array.
2. Make sure that the angle of the measurement vectors is oriented correctly on the array
channels. The vector should include the signal ends on the differential. On the absolute,
it should start from the nominal point and go to the signal end.

3. Press the System button under the Calibration tab.

4. Select the C-Scan radio button.

292
 5. Select the item that corresponds to your calibration reference from the list.
6. Then click the green arrow. This action tells the system that the previous measurement
matches the selected reference.

7. Click Calibrate. This will apply the gain and rotation necessary to bring the measured
reference signal to the desired value

293
If you have calibrated the hole to 1V and 90°, the signal after calibration should look like the
image below:

CALIBRATION RESET
A calibration reset will reset changes made to the software gain and rotation, while preserving
the sizing curves that have been previously created. To perform a soft reset, follow the steps
below:
1. Under the Calibration ribbon, select the System menu. For both the Channel and C-
scan radio buttons, click the “Reset” button to remove the stored information.

294
 2. If you have chosen the Voltage Plane representation for your setup, select the voltage
plane in Magnifi. Under the Current View ribbon, click on the “Reset Nominal” button.
Note that you must perform this step individually for both frequencies. To do so, change
the displayed frequency from the Voltage Plane view and then click on the Reset
Nominal button again.

3. The final step to completely remove calibration information from the current setup is
to reset the manual rotation applied to the C-scan Lissajous. To do so, click on the
rotation button, hold the left button and drag the mouse up or down until the display
value is 0. Do this for all array channels.

After performing the steps above, the calibration reset has been completed.
295
SIZING CURVES CALIBRATION
The sizing curves allow the size of a defect to be estimated based on the calibration point signals
(amplitude or phase) obtained with your calibration standard. Magnifi will give you the
interpolated flaw size based on the sizing curves generated.
The default RFA setups come with predefined curves for bobbin and array channels.
To calibrate your sizing curves:
1. Go to Calibration tab and click on the System icon.

2. In the Calibration window,

a. If the sizing curve to calibrate corresponds to bobbin channels (ABS, DIF): Select
Channel and Sizing Curves
b. If the sizing curve to calibrate corresponds to array channels (C-scan): Select C-
scan and Sizing Curves

3. Place your cursor on the part of the signal that corresponds to an item in the list (on the
strip chart or the C-scan).
Note: The system calibrates the signal with a measurement method that is defined in the Sizing
Curves Configuration page (under the Advanced tab). Therefore, it is recommended to read
the data with the same method as the system when you select your calibration points. This will
allow you to see what the system is using to make its calibration.

296
 4. Select the associated signal name in the list.
5. Click on the green arrow.
6. Redo the previous steps for all the other indications in the list.
7. Click on the Calibrate button.

When one or more points were not measured correctly, Magnifi generates an error message to
indicate the user that the sizing curve(s) could not be built. Here are some typical examples of
problems that could arise during sizing curve calibration.
In the left image below, the blue measurement arrow is pointing downward towards the
nominal point, instead of pointing up. In this case, the measured phase is rotated by 180˚ from
the correct value. To correct the situation, the swap orientation button (green circling arrows)
can be clicked before making the measurement during the calibration process.
In the image on the right, the arrow is pointing up towards the end of the defect signal, which
is correct. However, the arrow starts quite far from the nominal point, which will cause an error
in the measurement. In this case, we suggest adjusting the cursors until the measurement is
taken from a point near the base of the defect signal.

297
To readjust your sizing curves (e.g., if error messages were obtained during calibration), go the
Calibration tab, and click on View Curves.

The sizing curve window will appear. Each sizing curve will be listed in the drop-down menu. If
an invalid notification is present on the curve, it means that either the sizing points have not
been entered yet or that Magnifi failed to create the curve. To adjust the sizing curve manually,
enter the values in the table.

298
To validate that a curve works properly, bring the measurement cursor over a flaw’s signal from
your calibration tube and verify the value shown in the lower right corner of the Lissajous. It is
recommended to use the same measurement method as the one that was used to build the
sizing curve. To change the measurement method, click on the icons at the bottom left of the
Lissajous.

299
If the default size does not appear, it means that you are not displaying a channel where a
sizing curve has been defined.
Built sizing curves can give you access to the phase-depth sizing overlays on the Lissajous and
the Voltage planes. To activate this option, click on a window, go to the Current View tab and
check the Depth-V. checkbox.

LANDMARK CALIBRATION
This section describes how to calibrate your landmarks.
1. Go to the calibration tab and click on the Landmarks button.

300
 2. Calibrate the landmarks shown in the Landmarks window the same way the sizing
curve(s) points were calibrated. You can use the Land channel to do so.

The positions of TSI and TSO are described in the Landmark window in the Setup tab (TSI is on
the far-side tube sheet, that is, the first one encountered by the probe while pulling it during
data acquisition; TSO is on the near-side tube sheet).

Once the landmarks are calibrated properly, the system should be able to recognize them
automatically and the name of the landmark will appear in the Code view with a blue triangle.

Note: In order to calibrate the default Exit landmark, data that includes the probe exit at the
far end of the tube is needed.

FAST NORMALIZATION SETTINGS

Material properties, such as relative permeability, can change from one tube to another in the
same tube bundle. This will cause a shift of the nominal point on the voltage planes and a
rotation of the channels using a Lissajous. Consequently, the sizing curves calibrated with the
reference tube will no longer be valid for these tubes. More information on this phenomenon
can be found in Appendix 1 of this document.

301
If the angular movement of the signals is not too great (typically less than 30°), a common
practice is to renormalize the absolute channels displayed on the Voltage plane (if this
representation was chosen) and recalibrate the support plate signal to the reference obtained
with the calibration tube for the differential channel.
As these operations are often performed several times during an inspection, Magnifi offers quick
keys to facilitate the analysis.
To configure them, go the RFT tab and click on the Settings Shortcuts button. The displayed
window will show the channels using voltage planes and standard scale process.

The keyboard shortcut next to the channels can be customized to normalize the associated
channel. To set the reference values for channels using a Lissajous, select the reference
indication signal with the cursors in your calibration tube data, then click on the Measure
Reference button. The reference used in this case is normally the support plate.

SAVING YOUR SETUP

Once all your setup adjustments are done, you can save your setup by clicking on the Save
Setup button under the Home tab. The displayed window will allow you to give an appropriate
name to your setup and to save it to the desired location. The save location is, by default, your
inspection sub-folder. Note that when data are saved, the setup is also saved with it.

302
You can also save the data of your calibration standard by clicking on the Save As button under
the Home tab.

PERFORMING AN INSPECTION
Once you have used the Setup Wizard to create a setup, calibrated the data from your
calibration standard, and created your sizing curves, you are now ready to perform an
inspection.
Perform a preliminary scan of the tube bundle using the frequency estimated from the equation
on page 265. Be sure to renormalize the signal again to ensure the selected frequency is
appropriate by checking if the angle correction is within the set tolerance.
Renormalization and recalibration during the data-taking process is simplified with the shortcut
keys you defined earlier. To renormalize the channels using the voltage plane process, select a
sound area in the inspected tube and enter the keyboard shortcut you selected earlier for the
fast normalization settings.
To recalibrate the channels using the standard scale process, select the reference (normally a
support plate) and enter the keyboard shortcut. The system will adjust the selected reference
to the value set in the RFT Normalization Settings window (Amplitude and Angle). Note that
these operations should be performed using the conventional channels, not the array channels.
The gains and rotations made on the conventional channels will also be applied to the array
channels.
Whenever any hardware settings are changed, be sure to re-null the probe in the calibration
tube before performing another acquisition.

DATA LIST MANAGEMENT

For each acquisition, Magnifi can automatically save a file using the file name defined previously
in the Data List section of this document. To do so, checkmark the Automatic File Recording
option that can be found under the General tab of the Preference menu of the Backstage.

303
The list of tubes is also shown in the Data window.

A common practice is to rescan your calibration tube and balance the probe periodically. You
can save this new calibration tube data by adding a new tube in your tube list (999 001 for
example). Or you can exit out of acquisition mode, scan your tube, and click on the Save Cal
As… button under the Calibration tab to save your data. You can then go back to Acquisition
Mode to continue your inspection and to automatically save the acquired tubes in your bundle.
Once a tube has been scanned, the "play" icon next to the file name will be replaced with a
checkmark icon.
You can rescan a tube by selecting its name in the list and by clicking on the Rescan button in
the Home tab.
A tube name can be changed by right-clicking on its name and by selecting the Rename option.

LOADING A FILE
17. First disconnect your computer from the Ectane by clicking on the Disconnect button
under the Home tab (optional step).
18. You can load a file by double-clicking on the file name in the Data window. It can also
be done by selecting the file in the list and by clicking on the Load button under the

304
 Home tab. Note that double-clicking on a tube when you are still connected to an
instrument will initiate data acquisition.
19. You can open the next or the previous file in the list by clicking on the Previous or Next
button of the Home tab.
The data files can be filtered by using the Filter button of the Data window.

REPORT

INDICATIONS
The two Indication buttons in the lower-right corner of the Lissajous view can be used to add
an entry to the report. These two buttons indicate the code that is associated with the defect
being entered. Both buttons are identical, but can be set to identify different flaws.
To add an indication to the data:
20. Select the defect signal in the strip chart and adjust the cursor so that the signal in the
Lissajous includes only the defect signal.
21. Then, click on the red triangle in the corner of the Indication button to select the type
of defect to enter.
22. Click on the defect button to add an entry to the report.

305
Report entries can also be added using keyboard shortcuts. The SMART Setup section of this
manual explains how to configure them.
Another type of report entry, called “feature”, can also be added to a tube to indicate, for
instance, that it has no flaw or that it is plugged. A feature is an indication that is general to
the tube, it is not localized like defects, and therefore has no measurement or size associated
with it. To add a feature to a tube:
23. Load the file
24. Click on the appropriate indication button available under the Analysis tab

306
REPORT TABLE
To access the list of indications entered:
25. Make sure that the Report option is selected under the Layout tab.

26. Click on the report tab at the bottom of the screen to make the list visible

Entries in the report can be modified by changing the value in the table. You can also delete an
entry be clicking on the X next to it.

REPORT GENERATION
Magnifi can automatically generate a full report with the report table.
To generate this report:
27. Go to the Backstage by clicking on the arrow at the upper left corner of the
Frontstage.
28. Click on the Generate Report button under the Report section of the General tab.
29. Choose your preferences and enter the required parameters. The # of tubes in
component is used to show the percentage of tube in each category.

307
 30. Click Finish to generate the report.
This will create a PDF report that will show information such as the list of indications in your
bundle and a report summary with a pie chart.

308
The report logo can be modified by clicking on the Select Company Logo that can be found
under the System tab of the Backstage.

The report table file in the Inspection folder can also be imported into other reporting software
such as Tubepro.

309
ADVANCED SETTINGS AND OPERATIONS

ADVANCED SETUP MODIFICATIONS

The RFA Setup Wizard offers possibilities to generate your setup, but some options are not
directly accessible with the Wizard. To access these modifications, it is necessary to use
advanced menus.

MODIFICATION OF THE DEFAULT SCALE PARAMETERS FOR CALIBRATION

By default, if the absolute channels are represented with a Lissajous when generating setups
with the RFA Wizard, and the calibration of the absolute channels is performed on the Support
Plate indication signal by bringing it to 0.5V and 0°. For the differential channels, the calibration
puts the signal of the Through-Wall Hole indication at 1V and 90°.
These settings are the same for both the bobbin and array channels.
If needed, these calibration parameters (reference point, amplitude or phase value and
measurement method) can all be customized using the Advanced Channels and C-scans
menus.
Reminder: If the conventional absolute channels are represented with the Voltage Plane, the
calibration uses the Voltage Plane process. Therefore, it can’t be customized.
Bobbin Channels:
1. To modify the calibration parameters for the bobbin channels: Click on Process Ch.
under the Advanced menu.

2. In the Data Processing page, select the page (ABS (Scale) or DIF) that you want to
modify.

310
 3. Double-click on the Scale icon for the channel that requires modification.
4. For both the Amplitude and Phase, you can change the targeted value, the reference
point and the measurement method. Note that different reference points can be used
to calibrate the phase and the amplitude independently.

311
Array Channels:
If the calibration mode for C-scans is set to Manual, the C-scans must be calibrated
independently from the bobbin channels in the calibration tube. In that case, you might also
want to modify the calibration parameters for the array channels:
1. Click on C-scans under the Advanced menu.

2. In the C-scan Configuration page, select the C-scan that requires modifications.

3. In the Process section associated with the selected C-scan, select the Scale process, then
click Edit to open the Scaling Setup page.

312
 5. For both the Amplitude and Phase, you can change the targeted value, the reference
point and the measurement method. Note that different reference points can be used
to calibrate the phase and the amplitude independently.

SIZING CURVE MODIFICATIONS

In the Sizing Curves Configuration window (under the Advanced tab), you can either create a
new curve by clicking on Add, or modify the general parameters of predefined curves such as

313
measurement type, channel source, etc. By default, all curves are measured using phase angle.
However, they can be changed to amplitude.

For every sizing curve seen under the General section, there will be a specific page in the
Definitions section to configure the curve properties. By default, the Average Peak (MP)
measurement method is set for the absolute channels, and the Peak to Peak (PP) option is used
for the differential channels.

314
The plot of sizing curves for each individual channel can be seen by clicking on the Curve tab.
From there, you can click Select... to see or modify the calibration points for each curve.

Note: When you modify an existing curve or add a new curve, you must recalibrate/calibrate all
sizing curves (channels and C-scans).

315
RFA LINKS EDITOR

The RFA Links Editor menu can be found under the RFT ribbon. The drop-down menu available
at the top of this window shows the selected mode of calibration for the C-scans in the
calibration tube, as initially defined during the Setup Wizard.

In the Links table, the left column shows the conventional bobbin channels, and the right
column shows the C-scans associated to them. When a channel in this list is calibrated, or its
software gain or rotation is modified, the same changes are automatically applied to the C-
scans linked to it.

It is highly recommended to leave this menu as it is. However, it is accessible to allow for
advanced customization of an RFA setups. For instance, should a new C-scan be added to a
setup, this menu would allow to link it to the correct bobbin channel.

LOSS OF COMPATIBILITY WITH THE RFA WIZARD

All the setup modifications covered until this point are considered as being a typical setup
creation workflow for RFA technology. If additional customization of the setup is required for
specific purposes, you should be aware that it might result in a loss of compatibility with the
dedicated RFA Wizard.
Modifications made to the following menus would break compatibility:
• Advanced Scan Definition menu

316
 • Channels and C-scans Advanced Process menus: Modifications to processes other than
the Scale (as shown previously)

• Instrument Configuration menu

Important: If you need to make modifications to the above menus, you should stop using the
RFA Wizard and only use the Advanced menus for all the adjustments to the setup, including
the frequency selection, scan length, etc.

SMART SETUP TOOLS

Additional functions, which facilitate data analysis, can be set up under the SMART Setup tab.

LOGBOOK
The first button on the left side under the SMART Setup tab allows you to enable or disable the
logbook.

317
The logbook stores information about the analysis in addition to allowing the following
functions: batch screening, seeing results at a glance from the data list pane, and saving
manual modifications made to the landmarks during the analysis.
The logbook stores the following information: landmarks (type, position, dimension, including
manual modifications), indications (AI and conventional, label, position, dimension, etc.),
indication grouping, and data quality validation results. When enabled, an additional column
is added to the data file list to allow seeing results and to filter, sort, or analyze according to
these results.

The following table shows the possible icons and their meaning:

Icon Description Meaning

No logbook entry associated with this file
(empty) Empty
because it has not been screened yet.
Logbook entry associated with the file
contains information such as landmark
detection or regions.
Dashed outline circle,
The absence of additional information on the
without a number
icon reveals that no indication detection
engine or DQV tests were enabled when the
logbook was written.

318
 Logbook entries associated with the file
contains information such as landmark
detection as well as detected indications. The
Dashed outline circle, with
number displayed inside the circle shows the
a number
quantity of potential indications detected.
No DQV tests were enabled when the logbook
was written.
Dashed outline circle, with The green check mark indicates that the DQV
a green check mark test(s) passed on this file.
Dashed outline circle, with The red cross indicates that at least one DQV
a red cross test(s) failed on this file.
The logbook entries associated with this file
Dashed outline circle, with
indicates that potential indications have been
a number, with a green
detected by the detection engine and that the
check mark
DQV test(s) passed.
The logbook entries associated with this file
Dashed outline circle, with suggest that potential indications have been
a number, with a red cross detected by the detection engine and that at
least one DQV test(s) failed.
The AI detection tools identified an unusual
situation. In this case, the level of confidence
Orange warning triangle in the detection findings could be very low.
Therefore, users should pay special attention
when reviewing the detection results.

The logbook is saved within the inspection folder. Its data is written when a file is screened using
the dedicated buttons in the Analysis ribbon. It is also automatically written when a file is loaded
while the logbook is enabled, and no logbook entry existed for this file yet.
It is worth mentioning that the logbook also includes defects and landmarks, even if these have
been detected with detection tools that do not use AI (conventional landmark detection and
indication boxes). Also note that the automatic indication detection tools are not available if
you use the acquisition license (ACQ).

LANDMARKS
The Landmarks button is used to setup the parameters for landmark detection with
conventional tools.

319
The Landmark Detection window is used to configure the automatic detection of features such
as tube sheets and support plates. Landmarks are not mandatory and don’t need to be set to
have functional setup. They can, however, give a reference for the axial position in a tube. They
can also be used by the software to trigger automatic acquisition sequences.

Three landmarks are created if you click on the Suggest button. The channel selected to detect
these landmarks should be the lowest frequency absolute channel. The R_ before the channel
stands for Raw. This is the signal of the channel without software filter, rotation, or software
gain applied.
You can base the position of your landmarks either on the location where you start to record
(usually the tube entrance, opposite from the operator position), or the place where you stop
recording (usually the operator side).

320
In the above example, the Exit landmark is detected when the probe exits the tube at its outer
end. It has a negative position because this event happens before entering the tube. It is
enabled only when doing the acquisition.
TSI is the first tube sheet encountered when the probe is pulled. TSO is the last tube sheet
encountered at the end of the acquisition.
Landmark detection can be set up manually by describing the shape, component, and voltage
threshold that will trigger the detection. The Shape describes the appearance of the signal when
the landmark is reached. If a differential signal shape is chosen, the Peak-to-Peak distance
(“P2P”) will also be needed. The Direction is the projection axis (horizontal or vertical) of the
Lissajous signal that will be taken to trigger the landmark. The Threshold is the voltage
amplitude threshold.
Landmarks can be calibrated on real signals; in this case, there is no need to change these
parameters as they will be automatically measured by the software. It’s important to set the
landmark positions as accurately as possible. If the position is not accurately set, the software
might prevent their automatic detection since it won’t be at an expected position.
The Type field is a name that associates the calibration point with the landmark. If landmarks
share the same Type, they will be calibrated at the same time using the same point and process.
To associate two landmarks with the same Type, their signal must be the same. If support plates
of the same geometry are present in a bundle, they can share the same Type. In the above
example, TS1 and TS2 don’t share the same Type because one is triggered when the probe goes
inside the tube and the other is triggered when the probe exits the tube.
The Detection Engine drop-down menu can be set to Legacy or Extended. With the Legacy
mode, all the landmarks need to be entered with the correct position. The system will look for
the exact number of landmarks entered at positions close to those entered in the table. With
the Extended mode, the system will look for a number of landmarks between the Qty max and
the Qty min. With this mode, the exact number of support plates doesn't need to be constant
or known.

REGIONS
The Regions option allows the user to define inclusion/exclusion zones based on Landmarks. This
is used to categorize the data to process correctly. For example, they can be useful for
restricting the detection of channel saturation to an area within the tube or specify an
automatic analysis type for a given region (e.g., under the tube sheets).

321
If you are using the conventional detection tools, you can set up the regions according to the
automatic analysis tools you will use with the Indications and DQV buttons (described later).

INDICATIONS
The Indications module allows the user to define various zones or even waveform types to trigger
a certain detection threshold.

These detection zones can be applied to conventional channels and C-scan channels. The Label
is the name that will appear in the Code view.

322
There are five zone types:

Absolute pie
Differential pie
Centered rectangle
Rectangle
Over or under a horizontal or vertical threshold

And seven waveform types:

Up
Down
Down-up
Any direction
Up-down
Down-up-up
Up-down-up-down

Each zone or waveform can be linked to a defined region, and thresholds can be set. The zone
types also include the boundary setting to prevent signal oscillation from triggering unnecessary
detections.
Detection zone parameters can also be set by clicking on the Lissajous of the used channel and
activating the Edit Alarm option on the Current View tab.
The detected indication can be found in the Indication window, which can be activated in the
Layout tab.

Clicking on an indication in the pane instantly positions the cursor on the selected indication in
the Strip chart view. If you click on an indication in the code view (described later), the
associated indication is highlighted in the Indications pane.

323
DATA QUALITY VALIDATION (DQV)
Data Quality Validation allows defining tests to validate the quality of the data.

Five test types are available. To add a test, click on the Add button on the DQV page and select
the test type with the drop-down menu. Parameters associated with this test can be changed
in the Details section of the DQV window.

324
 • Dead Channel: Verify that channels are reading a signal whose amplitude is above a
defined threshold. An amplitude that remains below the threshold throughout the scan
indicates a dead channel.
• Landmark Detected: Verify the presence of all the selected landmark names in the
Details section. To validate full tube-lengths of straight bundles for ECT bobbin data
using AI detection, select this option and check the TSI and TSO landmarks. When AI
setup for landmarks is completed and the TWH calibrated, the DQV test will
automatically use the AI results to perform the test.
• Hardware Saturation: Verify if signal saturation appears on raw channels. The merge
field allows grouping a zone together in which the test has failed. This function avoids
having many notifications if the test fails several times in one area. This test can be
limited to a specific region, for example, the areas inside the tube.
• Channel Saturation: Verify if signal saturation appears on calibrated channels. The
merge field allows grouping a zone together in which the test has failed. This function
avoids having many notifications if the test fails several times in one area. This test can
be limited to a specific region.
• C-scan Saturation: Verify if signal saturation appears on C-scans. The merge field allows
grouping a zone together in which the test has failed. This function avoids having many
notifications if the test fails several times in one area. This test can be limited to a
specific region.
The results of the DQV tests can be found in the DQV window, which can be activated in the
Layout tab.

325
GROUPING
The grouping tool can help the analyst find critical indications for review as it categorizes and
groups the automatically detected indications.

The Grouping menu offers a default configuration with three predefined categories:

• Support Plates
• Critical
• Less Critical

326
The Details section of each group shows the parameters defining the conditions that must be
met for a detected indication to be classified in this group. The user can modify all parameters.
By default, the Support Plates group is configured with the mix channel with a Peak-to-Peak
(PP) measurement method, and Max/Min threshold values are set for amplitude and phase.
For this group, the criteria are limited to indications detected under the SP regions. As a result,
the indications detected under support plates for which all conditions are true will fall into this
category.
Critical and Less Critical are two other groups predefined in the Grouping window. Both groups
are set on DIF-F2 channel, but with different Amplitude/Phase measurement settings. In both
cases, the grouping criteria is restricted to In Tube and Air to Air (for calibration tubes) regions,
which covers the full-tube length data.
It should be noted that each automatically detected indication can only be categorized in a
single group. Also, the grouping operation is executed according to the order of the groups
defined in the Groups list. Therefore, if a given indication meets all the criteria of the first group,
it will be categorized in this group. If not, it will be passed to the next group. If it meets all the
criteria of the second group, it will be categorized into this one. If not, it will again be passed to
the next group, and so on. If an indication does not meet the criteria of any group, it will be
placed in the Unclassified group. This implies that every single indication will be categorized,
and none can appear simultaneously in more than one group.
New groups can be created with custom parameters.

327
It is also possible to remove a group or use the up/down arrow keys to change the group priority
within the list.

In the Front stage interface, the Grouping results will be displayed in the Indications pane.

328
REPORT SHORTCUTS
Keyboard shortcuts can be used to make entries in the report. By opening the shortcut window,
it is possible to select a channel with a measurement method and to associate an indication
with it.

For example, we could use the differential channel F2 with a Peak-to-Peak measurement
method to enter pits in the report.

To select a key, press the Shortcut field and press the desired key on your keyboard. It is possible
to use a single key or a combination of keys such as ALT+D, for example.
When a report entry is made with the keyboard shortcut, the abbreviation of the type of
indication displays momentarily on the corresponding Lissajous.

329
DATA SCREENING
Data screening set up with the SMART Setup tools can be performed by the two buttons on the
left under the Analysis tab.

It is possible to run analysis on a selection of files or all files. The analysis is typically performed
on all files when the setup and calibration are adequate. The results are recorded in the logbook
and can be viewed in the Data window.

LIMITATIONS
Some operations, processes, and options are not available for the RFA probe and configuration
in Magnifi. This section lists the limitations of the workflow in Magnifi.
• RFA setups can only have two frequencies. More frequencies cannot be added.
• The Absolute C-scan does not have a Voltage plane representation, only a Lissajous
representation.
• Contrary to conventional RFT setups, it is not possible to link the Absolute and
Differential conventional channels together with RFA setups when the Voltage plane
representation is used.
• The RFA Wizard cannot generate encoder-based setups.
• If the Periodic Noise filter has been activated, it may take a few seconds to complete
processing. This is normal - the process can take up to more than a minute for very long
scans.
• The setup wizard does not suggest excitation frequencies based on the tube
characteristics. This parameter must be determined by the user.

330
APPENDIX 1 – VOLTAGE PLANE
The following two equations describe the through-wall transmission behavior of RFT:
The phase lag equation shows that the signal is not only influenced by the tube properties, but

B = B0e − = B0e
− d f 0  r
 = d f0 r
also depends on the frequency (f) at which the tube is inspected. Given a tube of constant

B = Magnetic field strength at a depth  = Phase lag

corresponding to phase lag ()  = 0 r = Permeability
B0 = Initial magnetic field strength d = Thickness
 = Phase lag f = Frequency
 = Conductivity
permeability and conductivity, and using a fixed inspection frequency, the phase lag will be
proportional to the thickness of the tube.
The magnetic field strength (B) will depend on this phase lag. For an infinite theoretical
thickness, the resulting magnetic field will be null. The explanation being that the field is
completely blocked and that nothing is caught by the receiver(s).
The voltage plane is a graphical representation of these formulas. The curves built on a voltage
plane show the effect of the thickness variation on the signal. There are four curves that
represent a flaw coverage of 90°, 180°, 270° and 360°. The point crossing the horizontal axis is
the nominal point. This is the signal obtained when the probe is in a sound area of the tube. It
has an amplitude of 1V and a phase of 0°.

331
When you first scan your calibration tube, this point doesn’t match the nominal point of your
theoretical curve. You must calibrate the signal to match your probe signal to the theoretical
curve(s). This operation is called normalization.
A bundle of tube may include tubes that have different properties. The permeability, for
example, may be different for some tubes. This will affect the nominal point position and the
built sizing curves won’t be applicable to this tube.
If the nominal point has not moved too far from the theoretical nominal point, the common
practice is to renormalize and to size the flaw with the built sizing curves.
If the nominal point has moved far from the theoretical value, it means that the tube properties
are significantly different from the calibration tube. Therefore, the phase lag for a given flaw
will also be significantly different and this will generate an error in the sizing that must be
considered. In this case, it is recommended to change the drive frequency to put the nominal
point back to the renormalization zone.

332
A common practice is to set the threshold angle at which the inspection frequency needs to be
changed to 30°.

333
APPENDIX 2 – PERIODIC NOISE FILTER
The Periodic noise filter addresses the presence of cyclic or periodic noise in C-scans caused by
the tube manufacturing process.
The pilger process in tubing manufacturing may create a cyclic permeability change along the
length of tubes. This in turn may cause noise that appears like visible diagonal stripes in C-
scans. In most cases, this noise is low and will not prevent the analysis of the C-scan signal but
in the instances where the noise is stronger, the Periodic noise filter may help to reduce it.

Pilger noise
By default, the filter is not enabled. All filters affect the signal to which they are applied to a
greater or lesser extent, and improper use can lead to the creation of artifacts or can negatively
affect the signal. It is recommended to use this filter only on data that has pilger noise that
interferes with the analysis.
It is possible to enable the filter and to set most parameters within the “C-scans” step of the
RFA setup wizard.

334
Checking the box on the desired C-scan line under the Periodic Noise column enables the filter.
The other parameters are described below.
It is also possible to enable the filter and access the filter parameters by right-clicking on the
desired C-scan and selecting the C-scan configuration menu. Double-clicking on Periodic Noise
Filter brings up the Periodic Noise Filter Configuration.

335
The "Enable filter" option allows you to activate or deactivate the filter.
The correct signal type, either "absolute" or "differential", must be selected according to the
type of C-scan to which it is applied.
The Window Size is the axial direction zone size, in data samples, that is considered by the filter.
The size of the window can be determined by measuring the axial length of a small defect, in
samples, such as the through-hole in the calibration tube. For RFA probes, this value is then
divided by 5 for absolute C-scans, and by 8 for differential C-scans.
Target defect size in samples
Window SizeAbsolute =
5

Target defect size in samples

Window Size Differential =
8

To measure the axial length of a defect, place the beginning and the end of the cursor at the
defect’s axial extremities and read the value displayed in the ΔY (smpl) field.
In the example below, on an absolute C-scan, the complete target defect signal is 185 samples
long. Therefore, a Window Size value of 185/5 = 37 samples could be used.

The Noise Threshold parameter allows the filter to process only noise that comes from a periodic
source that needs to be reduced. Therefore, this threshold is set at an amplitude, in volts,

336
between the observed periodic noise level and other baseline noise. The absolute peak
measurement method can be used to read the noise level.
Baseline noise can be measured when the probe is stationary in the calibration tube in an area
free of defects. The signal coming from the pilger noise can be measured in a fault-free zone.
If, for instance, the periodic noise has an approximate amplitude of 0.100 V and the baseline
noise is about 0.020 V, the Noise Threshold must be set between these two values.
The High Intensity checkbox does a more intense version of the filter. In the case of very high
periodic noise, it will reduce the noise further, but may create signal discontinuities.

The example above (left) shows a relatively high noise level. On the right, the result after
applying the filter is shown. On the unfiltered C-scan impedance plane (top), the cyclic noise
(seen as overlapping ovals at the base of the defects) affects the signal to noise ratio and
enlarges the defects such that the phase is difficult to measure. On the filtered C-scan
impedance plane (bottom), the signal is cleaner, and the phases are easy to measure since
the signals are narrow.

The Periodic Noise Filter does several operations in a row, and certain design constraints must
be highlighted to understand how it behaves:

337
 • As previously mentioned, low noise levels usually do not necessitate filtering. It might
prove difficult to set a Noise threshold that would be consistently below the periodic
noise and above other noises. If this value is not set correctly or the periodic noise is
too low, the filter will not work correctly.
• The filter must be used on signals that are around the zero mark (nulled / high-pass
median-filtered signals).
• The filter uses parts of the signal that only include pilger noise to determine its
parameters. If there are very large defects in a zone, the software might not be able to
find an adequate pilger noise signal and it will not be correctly filtered.
• Similarly, filtering might be inefficient at the beginning and end of the scan because
there is not enough data to measure the pilger noise parameters.
• In the C-scan configuration window (seen above), each filter in the list sequentially
processes the data
• The filter receives the signals obtained from the processes that precede it (see the C-
Scan configuration window above). This means that any signal amplitude used to
select a Noise Threshold must be measured on the signal that the filter receives. Since
the periodic noise filter is last in the C-scan processing chain, measure the signal with
the filter disabled to determine the correct value.
• Prior processes like median filters or other filters might in some circumstances
generate artifacts (straight lines, data at zero) that will in turn affect the periodic
noise filtering (seen below).

338