Investigation of High Intensity UV

Lights on FPI and MPI

September 24, 2014

John Lee David Piotrowski

Technical Specialist Principal Engineer
ASNT Level III – MT, PT ASNT Level III – UT, ET, PT
DAL Level III – MT, PT, IR

Brad Loggins Scott Talbott

Lead NDT Inspector NDT Inspector
DAL level II – MT, PT DAL Level II – MT, PT, RT

Delta TechOps | September 24, 2014 |

 Purpose

• High Intensity UV lights available.

• Standards/Specs not fully caught up.
• ASTM E 2297 update in 2010; 2011 ASTM E07.03 Subcommittee.
− ASTM E 1417, ASTM E 1444 updates in 2011/12 reference ASTM E 2297.
− Good changes => Emission Spectrum, Visible light.
• Warning note in ASTM E 2297 6.5.2 precipitated technical
concerns on high intensity UV lights.
(Warning—When a high intensity UV-A lamp (light sources that produce light intensity
greater than 10,000 μW/cm2 at 38.1 cm (15 in.)) is used for inspection, care must be
exercised to prevent the UV fading of indications and that the excess blue light that is
produced, does not mask blue/white indications.)

• 2012 A4A NDT Forum – Sherwin paper, “UV-A LED’s: Are they all
the same?”
• 2006 ISU Studies (Lopez), Magnaflux papers (Geis) => Emission
spectrum, Visible light.

How do we interpret the warning statement?

Delta TechOps | September 24, 2014 | 2
 Purpose

• Pratt requires high intensity ‘torch’ inspection on critical rotating

parts (in excess of 10,000 μW/cm2).
• Boeing: PODs for FPI/MPI all done with Mercury Vapor lights
(max= 6000 uW/cm2); Need to do POD work for higher intensities.
• Other OEMs = No response at all.
• Delta recently conducted a study to alleviate internal concerns.
― ‘Apparent crack lengths’ from IN718 panels (internal POD study).
• Measured fatigue cracks at varying intensities with LED and Mercury Vapor lights.
― UV fade experiment with TAM panels, IN 718 panels (fatigue cracks), and Al-
quench crack blocks.
• Conclusions: 1) LED UV lights produced better contrast; 2) Fading
due to extreme UV exposure did occur, but was beyond standard
practice.
• Recommend ‘independent industry study’ or ‘no further changes’.
− Revision to ASTM E 2297, ASTM E 1444, ASTM E 1417 => No interpretation!

Delta performed experiments to satisfy concerns

Delta TechOps | September 24, 2014 | 3
 FPI: Procedure
Equipment needed:
•TAM panels
•Two anodized Al quench-crack blocks
•6 FPI POD study panels
•Mercury vapor light source (Gould Bass CR2000)
•UV 'torch'/flashlight (Spectroline Opti-Lux 365 UV Light)
•UV light meters (DLM-1000, and XRP-3000 Accu-MAX)

Procedure:
•Process all panels with post-emulsifiable, Ultra-high sensitivity = ZL-37.

Apparent Crack Length

•For the 6 FPI POD study panels, measure the 'apparent crack length'
using the mercury vapor light at 1000 uW/cm2 (at the surface), then
move the light back and measure at 1500 uW/cm2, then 3000 uW/cm2,
then 6000 uW/cm2;
•Then repeat with LED UV 'torch'/flashlight at 1000, 1500, 3000, 6000,
10,000, and 15,000 uW/cm2.

Experiment to address glare and UV fade concerns

Delta TechOps | September 24, 2014 | 4
 FPI: Procedure
UV Fading
•For the anodized Al quench-crack blocks and the TAM panel, perform
regular UV fade test with Mercury Vapor UV and LED UV Lights;
− Place high intensity UV light source directly on the surface (1/2” offset);
examine the part every 5 minutes until fading is noticed.

•Extended UV Fade on TAM panel, POD cracks.

− Every 15 minutes for 6+ hours with both MV and LED.

Regular & Extreme UV fade studies

Delta TechOps | September 24, 2014 | 5
 FPI: Apparent Crack length – Mercury Vapor

Table 1. ‘Apparent Crack length’ (inches ) versus UV intensity using

Gould Bass CR2000 Mercury Vapor UV Light.

Actual Crack 1000 uW/cm2 1500 uW/cm2 3000 uW/cm2 5700 uW/cm2
length/Intensity
0.378 0.381 0.382 0.382 0.382
0.340 0.338 0.338 0.339 0.340
0.156 0.145 0.145 0.146 0.148
0.135 0.132 0.132 0.132 0.133
0.122 0.118 0.119 0.121 0.121
0.106 0.103 0.102 0.103 0.103
0.099 0.096 0.097 0.097 0.098
0.073 0.068 0.068 0.070 0.070
0.040 0.037 0.038 0.038 0.038
0.024 0.023 0.024 0.024 0.024
0.022 0.020 0.021 0.021 0.021
0.013 0.011 0.011 0.012 0.012
0.010 0.008 0.008 0.008 0.008
0.010 0.009 0.010 0.010 0.010

Increasing UV intensity results in equivalent or better detectability

via ‘apparent crack lengths’. This was true for both mercury vapor
UV lights and high-intensity LED UV light.
Delta TechOps | September 24, 2014 | 6
 FPI: Apparent Crack length – LED
Table 2. ‘Apparent Crack length’ (inches ) versus UV intensity using portable Spectroline
Opti-Lux 365 UV Light (battery powered flashlight/torch).
Actual Crack 1000 1500 3000 6000 10000 15000
length/Intensity uW/cm2 uW/cm2 uW/cm2 uW/cm2 uW/cm uW/cm2
2

0.378 0.381 0.381 0.382 0.383 0.383 0.383

0.340 0.339 0.340 0.341 0.341 0.341 0.341
0.156 0.148 0.148 0.149 0.149 0.149 0.150
0.135 0.132 0.132 0.132 0.133 0.133 0.134
0.122 0.120 0.120 0.120 0.120 0.120 0.121
0.106 0.103 0.102 0.103 0.103 0.103 0.104
0.099 0.098 0.098 0.098 0.098 0.098 0.099
0.073 0.068 0.068 0.070 0.070 0.070 0.071
0.040 0.037 0.038 0.038 0.038 0.038 0.039
0.024 0.023 0.024 0.024 0.024 0.024 0.024
0.022 0.020 0.021 0.021 0.021 0.021 0.021
0.013 0.011 0.011 0.012 0.012 0.012 0.012
0.010 0.008 0.008 0.008 0.008 0.008 0.009
0.010 0.009 0.010 0.010 0.010 0.010 0.010

At the same intensity level, LED lighting provided a more-

accurate representation of the crack length. This was true across
the board, and did not vary even at smaller crack lengths.
Delta TechOps | September 24, 2014 | 7
 FPI Apparent Crack Length: MV vs LED

MV MV
1000 uW/cm2 3000 uW/cm2
0.118” 0.121”

LED LED
1000 uW/cm2 6000 uW/cm2
0.120” 0.120”

Delta TechOps | September 24, 2014 | 8

 FPI: High Intensity LED UV-A Background

LED
15,000 uW/cm2
0.120”

Increasing the UV intensity for LED caused an increase in the background. However,
the contrast was enhanced (i.e., signal-to-noise ratio) at increasing UV levels, causing
a brighter indication, resulting in better overall detectability.

Delta TechOps | September 24, 2014 | 9

 FPI: UV Fade – Al Quench Crack Blocks

Al quench crack blocks

~40,000 uW/cm2 exposure.

Start
5 min

10 min

Fading of indications due to intense

UV light does occur, but at times so
lengthy, and intensities so great, it is
not representative of FPI at Delta.
Typically, any area of the part is only
exposed to UV light for seconds.
Delta TechOps | September 24, 2014 | 10
 FPI: UV Fade – TAM Panel
MV MV
3000 uW/cm2 3000 uW/cm2
Before Before

Surface exposed to Mercury Vapor light at 15” for 6 hours (3000 uW/cm2).
MV MV
3000 uW/cm2 3000 uW/cm2
After After

Fading does occur with extreme conditions, but still detected.

Delta TechOps | September 24, 2014 | 11
 FPI: UV Fade – TAM Panel

LED
11000 uW/cm2
Before

Surface exposed to LED light at 15” for 6 hours (11,000 uW/cm2).

LED
11000 uW/cm2
After

Even with extreme conditions, very minimal fading.

Delta TechOps | September 24, 2014 | 12
 FPI: UV Fade – Fatigue Cracks on IN Panels

Table 3. Cracks lengths of fatigue cracks (IN 718 POD Panels) before and after extreme UV exposure.
Fatigue Crack length(in) - After 10 minute exposure to After 10 minute exposure to
Actual 40,000 uW/cm2 UV light, 40,000 uW/cm2 UV light,
measured with LED light Mercury Vapor light

0.099 0.098 0.096

0.122 0.121 0.119

Representative pictures of a crack after 10 minutes of extreme UV

exposure, and then examined with a mercury vapor light (left), and LED
light (right). The 0.099” crack measured 0.096” with mercury vapor light,
and 0.098” with LED.
Which one is better?
Delta TechOps | September 24, 2014 | 13
 UV Fade – Fatigue Cracks on IN718 Panels

•The 0.99” long crack is barely visible with the Mercury Vapor light at 15
inches, and only measures 0.096”.

•The LED light produced a brighter indication, resulting in a more

accurate measure of 0.098”.

•The background is indeed higher with the LED light versus the Mercury
Vapor light. However, the contrast (i.e., signal-to-noise ratio) is much
better with the LED lighting compared to the Mercury Vapor due to the
much brighter indication.

•This would positively affect the detectability of the crack (using LED light
instead of Mercury Vapor).

Delta has therefore allowed the use of these lights

with only a “do not dwell” caution added.

Delta TechOps | September 24, 2014 | 14

 Magnetic Particle Inspection Studies
Equipment needed:
• Flawtech panels.
• Mercury vapor light source (Gould Bass CR2000).
• Micro-gas discharge light source (Spectroline Maxima).
• UV 'torch'/flashlight (Spectroline Opti-Lux 365 UV Light).
• UV light meters (DLM-1000, and XRP-3000 Accu-MAX).
• Steel bolts = crack sizes of 0.034”, 0.079” 0.095”, 0.100”, 0.159”, 0.190” 0.282”.
• Ketos Ring.

Similar studies for MPI – Fluid chemistry.

Delta TechOps | September 24, 2014 | 15
 MPI: Procedure

UV Fading
• For the Flawtech panels, KETOS ring and bolt with 0.034” crack and bolt
with 0.282”, perform UV fade test;
– Place source such that UV reading is off-scale at the surface; examine the part
every 5 minutes. Repeat study for each light source type (micro-gas discharge,
mercury vapor, LED).

• For the KETOS ring and bolt with 0.034” crack and bolt with 0.282” crack,
perform ‘extended’ UV fade test;
– Place source 15” away from surface; take UV reading; examine the part every
15 minutes. Repeat study for each light source type.

Apparent Crack Length

• For the Flawtech panels (0.200" long cracks in different configurations),
and bolts.
– Measure the 'apparent crack length' using the mercury vapor light at 1000
uW/cm2 (at the surface), then move the light back and measure at 1500
uW/cm2, then 3000 uw/cm2, then 6000 uW/cm2 (or max as you can get);
Then repeat with UV 'torch'/flashlight and micro-gas discharge light at 1000,
1500, 3000, 6000, 10,000, and 15,000 uW/cm2.

Delta TechOps | September 24, 2014 | 16

 MPI: Apparent Crack Length Results

• No major difference in ‘apparent’ crack length was observed in using either the
mercury vapor, micro-gas discharge lamp, or the LED light.

• Similar to the FPI study, at the same intensity level, LED lighting provided a
more-accurate representation of the crack length.

• As intensity was increased, a slight increase in ‘apparent’ crack size was noted,
also across all crack lengths. Smaller sizes not impacted disproportionally.

• Increasing the UV intensity for LED caused an increase in the background.

However, the contrast was enhanced (i.e., signal-to-noise ratio) at increasing UV
levels, causing a brighter indication, resulting in better overall detectability.
Actual Crack 1000 1500 3000 6000 10000 15000
• Panels also provided length/Intensity uW/ uW/ uW/ uW/ uW/ uW/
a glare example. cm2 cm2 cm2 cm2 cm2 cm2
0.034 0.032 0.032 0.033 0.033 0.034 0.034
• Bolts had tricky 0.079 0.077 0.077 0.077 0.078 0.078 0.079
cracks. 0.095 0.096 0.097 0.097 0.098 0.098 0.098
0.100 0.104 0.101 0.102 0.103 0.103 0.104
0.159 0.155 0.156 0.156 0.157 0.158 0.158
0.190 0.187 0.188 0.188 0.189 0.191 0.191
0.282 0.283 0.282 0.282 0.283 0.283 0.284

LED lighting provided a more-accurate representation of the crack length.

Delta TechOps | September 24, 2014 | 17
 MPI: UV Fade Results
• Concluded to be little to no change after the extreme UV exposure, within the
measurement error.
• Measurements were taken to see if the ‘apparent crack length’ had changed as a
result of UV fading. There was no change in length and it was concluded UV
Fading was negligible.
• Same conclusions for the Mercury Vapor UV light and the micro-gas discharge UV
lamps.
• The KETOS ring also showed no difference in holes detected before and after the
UV fade.
• Some fading of the steel bolt cracking was noted after about 20 minutes (apparent
crack length), but this timeframe was deemed extreme (as well as the extreme
exposure).
• After 6 hours, slight fading was noticed (apparent crack length) for each light
source type, with the LED showing the greatest amount of fade.
• However, this fading was still deemed to be negligible as the crack lengths did not
change, though the apparent brightness did vary. This extreme measure was
deemed not representative of the process at Delta (i.e., typically the part is under
UV light on for a few seconds).

Similar result to FPI.

Delta TechOps | September 24, 2014 | 18
 MPI: UV Fade Results

BEFORE AND AFTER (20 MINUTES)

PHOTOS OF THE CRACK AFTER
EXPOSURE TO ~40,000 uW/cm2 UV
INTENSITY WITH LED LIGHT.

Negligible difference after extreme UV exposure.

Delta TechOps | September 24, 2014 | 19
 Conclusions - FPI
• Increasing UV intensity results in equivalent or better detectability via ‘apparent
crack lengths’. This was true for both mercury vapor UV lights and high-intensity
LED UV light.
• At the same intensity level, LED lighting provided a more-accurate representation
of the crack length. This was true across the board, and did not vary even at
smaller crack lengths.
• Fading of indications due to intense UV light does occur, but at times so lengthy,
and intensities so great, it is not representative of FPI at Delta. Typically, any area
of the part is only exposed to UV light for seconds.
• Increasing the UV intensity for LED caused an increase in the background.
However, the contrast was enhanced (i.e., signal-to-noise ratio) at increasing UV
levels, causing a brighter indication, resulting in better overall detectability.
• UV Fade experiment on the TAM panel using post-emulsifiable (Class 2, Level 4)
penetrant produced some fading after 20 minutes, but the 5th star crack (smallest
crack) was still visible even after 90 minutes.
• The quench crack blocks did show fading after 10 minutes of exposure to extreme
UV intensity. Therefore, it is recommended that a caution note be added to avoid
dwelling and limit total exposure of any one area to 5 minutes or less.

High intensity UV lights allowed at Delta.

Delta TechOps | September 24, 2014 | 20
 Conclusions - MPI

• Increasing UV intensity results in equivalent or better detectability via

‘apparent crack length’. This was true for both mercury vapor UV lights,
micro-gas discharge lights, and high-intensity LED UV light.

• At the same intensity level, LED lighting provided a more-accurate

representation of the crack length.

• Fading of indications due to intense UV light does occur, but at times so

lengthy, and intensities so great, it is not representative of MPI at Delta.
Typically, any area of the part is only exposed to UV light for seconds.

• Increasing the UV intensity for LED caused an increase in the

background. However, the contrast was enhanced (i.e., signal-to-noise
ratio) at increasing UV levels, causing a brighter indication, resulting in
better overall detectability.

• UV Fade experiment produced negligible fading after 20 minutes, when

the test was concluded. Typically, any area of the part is only exposed to
UV light for seconds.

Don’t put unnecessary burdens on industry w/o data.

Delta TechOps | September 24, 2014 | 21