Service Investigation

Accident / Incident Report Report No.: 22-093

P&WC 8114 (2021-02)

Webb Flying Services Inc.

Air Tractor AT-602, Registration N5007R
Allport Arkansas
Accident Date: 1 June 2022
PT6A-60AG, Serial No. RG0035

Written By:
Service Investigation Department

Approved By:
Manager, Service Investigation Department

Date of Issue: 7 December 2023

Civil Aviation Authorities Identification: CEN22LA228

TSB Identification: A22F0163
P&WC File No.: 22-093
RFA No.: 23SIE00164
Sale order: 232542

LOCATION U.S. REGULATION

EAR ITAR
U.S. P-ECCN 9E991 P-USML NSR
EXPORT CONTROL CLASSIFICATION
OUTSIDE U.S. ECCN 9E991 USML NSR
LOCAL REGULATION
EIPA (ECL) DPA (CG)
CANADA NSR No
Date of marking 21 November 2023
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 2 of 42

Table of Contents

Page No.
I ANALYSIS

1.0 ACCIDENT SYNOPSIS 3.

2.0 SUMMARY OF FINDINGS AND DISCUSSION 3.

3.0 CONCLUSION 4.

II FACTUAL INFORMATION

1.0 INVESTIGATION PARTICIPANTS 4.

2.0 ENGINE HISTORY 5.

3.0 ENGINE EXAMINATION 5.

4.0 CHEMICAL ANALYSIS 34.

5.0 ACCESSORIES ANALYSIS 42.

.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 3 of 42

I ANALYSIS

1.0 ACCIDENT SYNOPSIS:

1.1 On June 1st, 2022, about 1700 central daylight time, an Air Tractor AT-602 airplane, N5007R,
sustained substantial damage when it was involved in an accident near Allport, Arkansas. The pilot
sustained minor injury. The pilot reported he was on the 35th flight of the day and the airplane had
been performing all day with no issues noted. The airplane was refueled two flights prior, and at
the time of the accident, each wing fuel tank was about half full of fuel. The airplane’s 6,500 lb.
hopper was loaded with 3,600 lbs. of urea. For the accident flight, the pilot intended to fly to a rice
field near Humnoke, Arkansas, to apply fertilizer. Immediately after the takeoff from the private
airstrip, the pilot heard a “loud pop” noise from the turboprop engine and he observed flames emit
from the left side of the airplane. The pilot confirmed that a loss of engine power occurred, he
made sure the flaps were down, and he began to scan the area to perform a forced landing. During
the forced landing sequence, the pilot reported that it felt like there was a momentary “slight
recovery” of engine power, but then the engine ceased producing power. The pilot retarded the
throttle and landed to a flat field consisting of grass, dirt, and mud. During the landing, the airplane
nosed over and came to rest inverted. The pilot was able to egress from the airplane without further
incident.

2.0 SUMMARY OF FINDINGS AND DISCUSSION:

2.1 The engine exhibited contamination of the fuel filter and various fuel wetted accessories. The
composition of many of the particles was consistent with fertilizer type material. Although bench
test of the fuel control unit (FCU), the flow divider / dump valve and fuel nozzles were acceptable,
P&WC experience has shown that fuel contamination can lead to engine operability issues.

2.2 Prior to disassembly, the bleed off valve (BOV), the unit was tested in accordance with (IAW) the
P&WC overhaul manual. Test results revealed the Px cavity would not pressurize. Disassembly of
the bleed valve revealed that the diaphragm was torn and the BOV piston was not indexed properly.
An analysis of a clear gummy fluid found on the piston retainer revealed that the composition was
fluorosilicone as opposed to a silicone based sealing compound specified by P&WC overhaul
manual. A review of the available maintenance logs indicated that, at TSN of 9,816.5 in November
2017, installed the overhauled BOV serial number 2L039. There were no
indications of any pre-impact distress found on any of the remaining examined components.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 4 of 42

2.3 The engine gaspath components exhibited heavy soot accumulation. The compressor and
compressor turbine had a uniform coating of soot, while the power turbines had localized soot
accumulation. This suggests that the compressor and compressor turbine section likely continued
to run after the power section ceased rotation.

2.4 Rotational rubbing marks were observed on the compressor turbine (CT) disc assembly and the 1st
stage power turbine (PT) disc assembly with their adjacent static components.

3.0 CONCLUSION:

3.1 The engine exhibited rotational contact signatures to the internal components that are characteristic
of the engine producing power at the time of impact. The exact power level could not be
determined.

II FACTUAL INFORMATION

1.0 INVESTIGATION PARTICIPANTS:

The powerplant investigation was performed on 10 to 11 January 2022 at the Pratt & Whitney
Canada (P&WC) Service Investigation Facility at Bridgeport West Virginia. The following
individuals participated in the investigation as representatives of their respective organizations:

Role: Air Safety Investigator

Organization: NTSB

Role: Air Safety Investigator

Organization: NTSB

Role: Air Safety Investigator

Organization: Hartzell Propellers

Role: Service Investigation

Organization: P&WC

Role: Service Investigation

Organization: P&WC
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 5 of 42

2.0 ENGINE HISTORY:

Engine Model: PT6A-60AG, Serial No.: RG0035

Logbook Available: Yes: ☒ *No ☐

*The engine logbooks were not available at the time of the investigation.

Time Since New (TSN): 11,429.55 (Estimated).

Cycles Since New (CSN): UNK

Time Since Overhaul (TSO): UNK

Cycles Since Overhaul (CSO): UNK

3.0 ENGINE EXAMINATION:

All positional references are in relation to view from aft looking forward. Upstream and
downstream references are in relation to gas path flow from the compressor inlet to the turbine
exhaust.

Front Aft
Figure 1: Engine Reference layout of large PT6A. Note the PT6A-60AG has three stages of axial
compressor rotors as opposed to four in this illustration.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 6 of 42

3.1 External Condition:

3.1.1 Photo #1 shows the aircraft accident wreckage.

3.1.2 The engine and propeller were received assembled and secured in a wooden container
(Photo #2). Along with the propeller, several airframe components to include the starter
generator, air conditioner compressor, overspeed governor, tach generator, firewall,
exhaust stacks and several external lines were received (Photos #3 - #7). Both the gas
generator and the power section data plates identified the module serial numbers as
RG0035, (Photo #8).

3.1.3 The compressor and compressor turbine (NG) module were capable of manual rotation
with no anomalies noted. The power section (NF) module was stiff and binding.

Photo #1: Courtesy of NTSB.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 7 of 42

Photo #2

Photo #3: Front section of the engine and propeller.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 8 of 42

Photo #4: Front right and front left-hand position.

Photo #5: Mid right-hand side and mid left-hand side.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 9 of 42

Photo #6: Rear right hand side and rear left hand side.

Photo #7: Rear section of the engine.

Photo #8: Gas generator and power section data plates.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 10 of 42

3.1.4 The power section and gas generator were partially separated at the “C” flange. The exhaust
duct along with the oil pressure and scavenge tubes exhibited compressional impact
damage (Photos #9 and #10).

Photo #9: Separation noted at the “C” flange.

Photo #10: Compressional impact noted to the exhaust duct and the oil tubes.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 11 of 42

3.2 Power Control Linkage and Reversing Linkage:

3.2.1 The FCU input rod was secured to the FCU arm. The FCU input rod was disconnected at
the cam box as received at P&WC, (Photo #11 red arrow). The FCU arm moved freely.
The cam box was secured with no pre impact damage noted.

Photo #11: General view of the rear linkage.

3.2.2 The reversing cable was free to move within the guide tube. The propeller governor reset
arm, reversing linkages, along with the beta valve and speed control lever moved freely
(Photo #12).

Photo #12: General view of the front linkage.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 12 of 42

3.3 Pneumatic Lines:

3.3.1 Compressor Discharge Air (P3) and P3 Filter: The P3 airline connectors were torqued
and secured with safety wire, (Photo #13). After removal of the P3 lines (Photo #14),
compressed shop air was applied confirming no blockage. The P3 filter was discolored, the
P3 filter bowl and penny valve exhibited what appeared to be corrosion deposits, (Photo
#15) likely a result of the marshy environment the engine had been exposed during and
after the incident.

Photo #13: P3 tubes attachment points.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 13 of 42

Photo #14: P3 tubes post removal.

Photo #15: P3 filter bowl, filter and penny valve.

3.3.2 Power Turbine Control (Py): The PY line was not received with the engine.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 14 of 42

3.4 Chip Detector(s) and Filters:

3.4.1 Chip Detector(s): Both the accessory gearbox (AGB) and reduction gearbox (RGB)
magnetic chip detectors (MCD) were free of metallic debris (Photo #16).

Photo #16: RGB and AGB MCD.

3.4.2 Oil Filter: Visual examination of the oil filter and strained oil samples revealed particles
of non-metallic flakes consistent with the appearance of varnish (Photo #17).

Photo #17: Oil filter and particles post straining of residual oil.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 15 of 42

3.4.3 Fuel Filter: The fuel filter exhibited sediment particles (Photo #18). The fuel filter was
sent to P&WC chemical laboratory for analysis reference section 4.0.

Photo #18: Fuel filter.

3.5 Disassembly Observations:

3.5.1 Compressor Turbine (CT) Section:

The engine was separated at the “C” flange. The CT blades and CT vane exhibited heavy soot
deposits. The CT bolt and cup washer were secured. Scoring was noted on the CT disk inner bolt
bore region. The CT blades and CT shroud segments exhibited heavy scoring (Photos #19 - #23).

Photo #19: General view of the hot section.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 16 of 42

Photo #20: CT disk inner diameter with scoring.

Photo #21: Downstream and upstream view of the CT disk after removal.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 17 of 42

Photo #22: Close up of the CT blade tips.

Photo #23: CT vane assembly and scored CT shroud segments.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 18 of 42

3.5.2 Combustion Section:

Both the inner and outer combustion chamber (CC), liners exhibited heavy soot accumulation.
(Photo #24).

Photo #24: Inner and outer CC liner.

3.5.3 Power Turbine (PT) Section:

The PT gas path components exhibited varying degrees of soot accumulation, most notable in a
local region of the vanes and turbines. The 1st stage PT vane ring had scoring on the inner baffle
at approximately the 12 o’clock position from contact with the CT disk inner bore (Photo #25).
Scoring was noted on the 1st stage PT vane inner shroud area from approximately the 12 o’clock
position to the 4 o’clock position from contact with the 1st stage PT blades (Photo #26 and #27).
The PT shroud exhibited scoring at approximately 12 o’clock position to the 4 o’clock position
from contact with the 2nd stage PT blades (Photos #28 and #29). The PT shaft along with the #3
and #4 bearing housing were distorted which led to binding of the power section. Both the #3 and
#4 bearings were intact, were oil wetted and did not exhibit any obvious visual anomalies (Photos
#30 and #31).
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 19 of 42

1st stage PT vane

ring with scoring
on the baffle.

Photo #25: General view power section / PT stator housing.

Scoring contact
from 1st stage PT
blades.

Photo #26: 1st stage PT stator, downstream position.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 20 of 42

Close up, scored blade

tips.

Photo #27: 1st stage PT disk. Upstream view. Notice heavy soot accumulation (Ref arrows).

Photo #28: PT shroud exhibiting scoring contact and soot accumulation (Red arrows).
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 21 of 42

Close up, scored blade

tips.

Photo #29: 2nd stage PT disk assembly upstream view. Notice heavy soot accumulation (Red arrows).

Photo #30: Rear RGB housing and #3 and #4 bearing housing.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 22 of 42

Photo #31: PT shaft along with the #3 and #4 bearings.

3.5.4 Reduction Gearbox (RGB):

Post removal of the RGB from the power section, both the forward and rear RGB gear systems
rotated freely. The front and rear RGB housings were separated at the “A” flange for visual
examination which did not reveal any anomalies (Photos #32 and #33).

Photo #32: Forward RGB housing and gear system.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 23 of 42

Photo #33: Rear RGB housing and gear system.

3.5.5 Compressor Section:

The compressor section including the gas generator diffuser area were covered in a uniform coating
of soot. Several 1st stage compressor rotor airfoils exhibited leading edge nicks (Photos #34 - #39).

Photo #34: Compressor assembly post removal from the gas generator.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 24 of 42

Photo #35: 1st stage compressor rotor and stator.

Photo #36: 2nd and 3rd stage compressor rotors.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 25 of 42

Photo #37: Compressor impeller.

Photo #38: Compressor stator / impeller housing assembly downstream view.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 26 of 42

Photo #39: Gas generator case and diffuser pipes downstream view.

3.5.6 Accessory Gearbox (AGB):

The AGB gear train could be manually rotated with no binding noted. The AGB was separated
from the inlet case (Photos #40 and #41). Visual examination of the AGB post removal did not
reveal any anomalies. The AGB was not disassembled as part of this investigation.

Photo #40: View of the inlet post AGB removal.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 27 of 42

Photo # 41: View of the AGB post removal. Photo courtesy of NTSB.

3.3 Controls and Accessories:

3.3.1 Ignition System:

The ignition cables were secured to the exciter box along with the igniters and safety wires. Visual
examination of the ignition components did not reveal any anomalies (Photos #42 - #44).

Photo #42: Ignition cables secured to the exciter box.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 28 of 42

Photo #43: Left and right side ignition cable secured to the igniters.

Photo #44: Igniters.

3.3.2 Fuel System:

Fuel Oil Heat Exchanger (FOHE): The fuel oil heat exchanger was properly mounted and
secured, (Photo #45). A sample of residual fuel was collected from the FOHE. The sample
exhibited sediment particles. The sample was sent to P&WC chemical laboratory for analysis
reference section 4.0.

Fuel Pump: The fuel pump was mounted and secured to its respective accessory gearbox pad
(Photo #46). Residual fuel was collected from the filter bowl. The sample was cloudy and
contained sediment particles. Sediment particles were also noted on the pump outlet port (Photos
#47 & #48). The sample was sent to P&WC chemical laboratory for analysis reference
section 4.0.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 29 of 42

Fuel Control Unit (FCU): The fuel control unit was mounted and secured to the fuel pump pad
(Photo #46). A sample of residual fuel was collected from the FCU. The sample exhibited sediment
particles. The sample was sent to P&WC chemical laboratory for analysis reference section 4.0.
The FCU was sent to P&WC Engine Controls and Accessories (ECA) for function test, reference
Appendix 1.

Flow Divider Valve (FDV): The flow divider was mounted and secured on the nozzle inlet pad
(Photo #49). The flow divider was sent to P&WC ECA for function test, reference Appendix 1.

Fuel Nozzles: The fuel nozzles were mounted and secured on their respective gas generator case
pads. All the fuel nozzles exhibited heavy soot accumulation on the sheaths and nozzles (Reference
photo#50) The fuel nozzles were sent to P&WC ECA for flow test reference Appendix 1.

Photo #45: FOHE.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 30 of 42

Fuel pump

FCU

Photo #46: Fuel pump.

Photo #47: Fuel filter bowl and residual fuel.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 31 of 42

Photo #48: Fuel pump outlet port with sediment red oval.

FDV

Photo #49
Close up of fuel nozzle with
sheath installed. All nozzles
appeared in similar condition.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 32 of 42

Photo #50

3.3.3 Air System:

Compressor Bleed Valve (BOV): The compressor bleed valve and the orifice were secured
(Photo #51). Removal of the bleed revealed that a section of the diaphragm was protruding out of
the BOV piston (Photo #52). The BOV piston could be manually actuated and was not stiff or
binding. The bleed valve was sent to P&WC ECA for function test and investigation (See
Appendix 1).

Photo #51: BOV installed.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 33 of 42

Photo #52: BOV with protruding diaphragm.

3.3.4 Oil System:

Propeller Governor (CSU): The CSU was secured on its respective mounting pad with no
obvious damage (Photo #53). The CSU was sent to P&WC ECA for function test
(See Appendix A).

Overspeed Governor (OSG): The OSG was secured to its’ respective mounting pad with no
obvious damage (Photo #53).

CSU

OSG

Photo #53: View of the propeller CSU and OSG.

Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 34 of 42

4.0 CHEMICAL ANALYSIS:

4.1 In addition to the fuel filter, fuel samples that were collected from the FCU, the fuel filter bowl
and the FOHE, illustrated in photo #54, the NTSB also provided a fuel sample taken from the
aircraft strainer, photos #55 and #56 for analysis.

A B
Photo #54: Fuel filter and fuel samples taken from the engine.

A B
Photo #55: Fuel sample as received by P&WC Service Investigation in West Virginia (A) and sample
extracted and provided to P&WC Chemical laboratory (B).
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 35 of 42

A B C
Photo #56: Test sample received at P&WC Chemical laboratory (A & B). View of portion of the sample
transferred to a glass vial for analysis (C).

4.2 The fuel samples were filtered through a 1.2μm filter patch and the composition of the residue was
evaluated by Automatic Particle Analysis (APA) excluding organic, fluorinated, and carbon rich
particles. The fuels were also evaluated by gas chromatography (GC) to determine if there were
other contaminations present. The fuel filter element was ultrasonically cleaned in pre-filtered
heptane and the solvent was filtered through 1.2μm filter patches to collect the solid residue. One
filter patch was analyzed by APA. Some particles were manually checked by scanning electron
microscopy (SEM).

4.2.1 Fuel Filter Sample: The solid contamination present on the fuel filter was 56.9mg. The filter patch
contained fine brown particles and fibrous material (Photo #57). By APA, the particles were
identified as the following.

• Calcium Rich particles.

• Traces of potassium rich particles.
• Traces of sulfur rich particles (often present with calcium).
• Traces of silicon rich particles
• Traces of aluminum rich particles.
• Traces of phosphorus particles.
• Iron based particles.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 36 of 42

A B
Photo #57: Filter patch after filtration (A). Microscopic image of the filter patch (B).

4.2.2 FCU fuel sample: The filtered volume of the FCU fuel was 51mL which contained 115.7mg/L of
solid contaminations. By GC, the fuel was similar in composition as the filter patch contained fine
brown particles and fibrous material (Photo #58). By APA, the particles were identified as the
following.

• Calcium rich particles.

• Sulfur rich particles.
• Silicon rich particles
• Traces of aluminum rich particles.
• Traces of potassium rich particles.
• Traces of phosphorus rich particles.
• Iron based particles.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 37 of 42

A B
Photo #58: Filter patch after filtration (A). Microscopic image of the filter patch (B).

4.2.3 Fuel filter bowl sample: The filtered volume of the filter bowl fuel was 25mL which contained
1444mg/L of solid contaminations. The filter patch contained fine brown particles and fibrous
material (Photo #59). By APA, the particles were identified as the following.

• Calcium rich particles – often found with phosphorous or sulfur.

• Sulfur rich particles – often found with calcium.
• Silicon rich particles
• Traces of aluminum rich particles.
• Traces of potassium rich particles.
• Iron based particles.
• A spherical particle that was rich in potassium and sulfur, with traces of carbon, oxygen,
sodium, phosphorous and aluminum was also detected (Photo 60).
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 38 of 42

A B
Photo #59: Filter patch after filtration (A). Microscopic image of the filter patch (B).

B
A

Photo #60: SEM image of the spherical particle rich in potassium and sulfur (A). SEM spectra of the
particle (B).

4.2.4 Sample from the FOHE: The filtered volume of the FOHE fuel was 4mL which contained 25mg/L
of solid contaminations. The filter patch contained fine brown particles (Photo #61). By APA, the
particles were identified as the following.

• Aluminum rich particles – often found with zinc.

• Silicon rich particles
• Calcium rich particles.
• Traces of zinc rich particles.
• Iron based particles.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 39 of 42

A B
Photo #61: Filter patch after filtration. Microscopic image of the filter patch (B).

4.2.5 Sample provided by the NTSB: The solid debris removed from the fuel sample consist to an
accumulation of white to orange brittle thin films, aggregates of brown particles with a rubbery
texture, white to yellow powder and traces of metallic particles and fiber (Photo #62).

• The white to orange thin films as well as the white to yellow powder are rich in phosphorous,
aluminum, potassium, calcium and iron in presence of oxygen with traces of manganese
(Photo #63).

• The brown particles with a rubbery texture are identified by FTIR as a polysulfide compound
containing calcium carbonate (Photo #64).

• There were also traces of metallic particles identified as cadmium, aluminum alloy and steel.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 40 of 42

Photo #62: View of the solid residue collected after filtration of the sample.

A B
Photo #63: Close up of the white to orange thin films (A). SEM spectrum of the white to orange thin
films (B). Such composition could suggest compounds in the form of phosphate salts and oxidative
species.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 41 of 42

Photo #64: Brown particles.

4.2.6 The chemical analyses of the fuel filter and various fuel wetted accessories showed that the
composition of many of the particles was consistent with fertilizer type material.
Service Investigation
Accident / Incident Report Report No.: 22-093
P&WC 8114 (2021-02) Page: 42 of 42

5.0 ACCESSORIES ANALYSIS

(A) P&WC Accessories Accident Report 23ECN00012 Appendix 1

REFERENCE “A”

This document is subject to the restriction contained on the cover page.

Accessories Investigation Report No.: 23ECN00012

Webb Flying Service Inc.

Air Tractor AT-602, N5007R
Allport, Arkansas, U.S.A.
1 June 2022
PT6A-60AG, RG0035

Written By:
Air Safety Investigator, Engine Controls Accessories Technical Support

Approved By:
Manager, Service Investigation Department

Date of Issue: 2 November 2023

Distribution:

Civil Aviation Authorities Identification No.: CEN22LA228

TSB Identification No.: A22F0163
Red File No.: 22-093
Accessories RFA No.: 23ECN00012
Service investigation RFA No.: 22SIE00164

This report is in support to the engine review.

LOCATION U.S. REGULATION

EAR ITAR
U.S. P-ECCN 9E991 P-USML NSR
EXPORT CONTROL CLASSIFICATION
OUTSIDE U.S. ECCN NSR USML NSR
LOCAL REGULATION
EIPA (ECL) DPA (CG)
CANADA NSR No
Date of marking 20 September 2023
Page 1 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

1. Accident synopsis:

On June 1st, 2022, about 1700 central daylight time, an Air Tractor AT-602 airplane, N5007R, sustained
substantial damage when it was involved in an accident near Allport, Arkansas. The pilot sustained minor
injury. The pilot reported he was on the 35th flight of the day and the airplane had been performing all day
with no issues noted. The airplane was refueled two flights prior, and at the time of the accident, each
wing fuel tank was about half full. The airplane’s 6,500 lbs. hopper was loaded with 3,600 lbs. of urea. At
the time of event, the pilot intended to fly to a rice field near Humnoke, Arkansas, to apply fertilizer.
Immediately after takeoff from the private airstrip, the pilot heard a “loud pop” noise emitted from the
turboprop engine and he observed flames exiting from the left side of the airplane. The pilot confirmed
that a loss of engine power occurred and made sure the flaps were down before he began to scan the area
to perform a forced landing. During the forced landing sequence, the pilot reported that it felt like there
was a “slight recovery” of engine power, but then the engine ceased producing power. The pilot retarded
the throttle and landed to a flat field consisting of grass, dirt, and mud. During the landing, the airplane
nosed over and came to rest inverted. The pilot was able to exit the airplane without further incident.

2. Summary of findings and discussion:

The investigation encompassed the visual examination and functional testing of the fuel control unit
(FCU), the flow divider and dump valve (FDV), the fuel manifold adapters (fuel nozzles), the compressor
bleed valve (BOV) and the propellor governor (CSU). The CSU was sent to the supplier for
the functional testing. The FCU, the FDV and the CSU were found unremarkable.

The fuel nozzle stems and nozzle tips were covered with soot. The fuel nozzle test results were in a normal
in-service condition typical of fuel nozzles returned from service.

The BOV diaphragm was torn. The BOV could not be pressurized during the functional test due to the
torn diaphragm. Disassembly showed the tear was ranging from the piston to retainer clamping interface
to approximately the end of the diaphragm rolling surface. Part of the fracture surface near the retainer
showed a darkened coloration while the remainder of the fracture surface appeared fresh, suggesting the
tear initiated gradually near the retainer and expanded rapidly.

Evaluation of the diaphragm at the material laboratory showed that the reenforcing fabric on the P2.5 side
of the diaphragm was locally frayed, which may have led to the inability to sustain the mechanical force
associated with the Px air pressure. Fraying of the reenforcing fabric was also observed in an area 180°
away from the tear, possibly illustrating an early stage of the phenomenon. With the loss of support from
the fabric, the orange elastomer layer (fluorosilicone rubber) stretched, delaminated from the edge of the
disk-shaped reenforcing fabric on the Px side and tore. The square shape of the tear was the result of the
fracture running along the main axis of the weaving of the fabric. The fraying of the fabric occurred on
the P2.5 side of the diaphragm, which is forced by the Px air pressure against the side of the piston. The
diaphragm rolls up and down as the bleed valve opens and closes. This movement implies some friction
between the rolling diaphragm surface and the piston surface, thus a possibility of wearing the orange
elastomer top layer and exposing the subsurface reenforcing fabric.

Page 2 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

Small traces of clear gummy fluid found on the bleed valve piston retainer was analyzed at the chemical
laboratory and assessed to be of a different composition than the specified in the P&WC
overhaul manual (OHM) for the piston and retainer interface with the diaphragm. The analyzed fluid was
determined by Fourier Transform Infrared (FTIR) to be composed mainly of fluorosilicone, while
is a silicone-based material.

Ink from identification markings on the diaphragm was observed to have transferred to the piston surface.
These transferred markings were located 120 degrees from the as-removed position. The clocking of the
piston “X” mark to the diaphragm and BOV housing was also offset by 120 degrees. This suggests the
bleed valve may have been assembled reusing the standard replacement diaphragm while incorrectly
positioning the piston.

The transferred marking “0816” was identified with the diaphragm manufacturer
to represent a manufacturing date of August 2016. The manufacturer confirmed 137 units manufactured
in 3Q16 were delivered and that the location of the marking on the diaphragm may vary. It is also a
possibility that the ink transfer on the piston originated from another diaphragm manufactured in August
2016 with a clocking of the marking 120 degrees offset.

The possible re-use of the standard replacement diaphragm, use of an improper lubricant or a combination
of both may have contributed to the tearing of the diaphragm.

3. Conclusion:

The torn compressor bleed valve diaphragm may have led to a power reduction from full power. There
were no defects or damage evident on the FCU, FDV, fuel nozzles and CSU that would have prevented
normal operation prior to the event.

4. Investigation results:

Conducted at: Pratt & Whitney Canada

Initiated on: 15 March 2023

In attendance:

Representing Title
Transportation Safety Board Regional Senior Investigator
(TSB) of Canada
Woodward Inc. Staff Mechanical Engineer, R&D and
Systems. Product Integrity Coordinator
Pratt & Whitney Canada Air Safety Investigator

Page 3 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

Hardware investigated:

Accessory P&WC Manufacturer Manufacturer S/N TSN/TSO

P/N P/N
Compressor bleed valve 3117032-01 N/A P&WC 2L039 Unk
Fuel control unit N/A 8061-333 12014921 Unk
Flow divider 3019906 25536-4 L0275 Unk
Inlet fuel nozzle 3100453-01 N/A P&WC Not legible Unk
Primary fuel nozzles 3100456-01 N/A P&WC Multiple Unk
Secondary fuel nozzles 3100458-01 N/A P&WC Multiple Unk
Propellor governor N/A 8210-310 11906859 Unk

Accessory name: Compressor bleed valve (BOV)

Note: This unit serial number (S/N) was not the one installed on the engine at initial engine delivery.

External condition and attachments:

The external surfaces of the BOV were dirty. The piston conical face was covered with a black film of
deposit similar to soot. A section of the diaphragm of approximately one inch was visible between the
piston upper lip and the housing. The P3 port packing was present, and the port was unremarkable. The
piston was moving freely. The lockwire was present from the diagnostic plug to the secondary orifice.

Test results:

An adhesive strip was adhered to the exterior circumference of the bleed valve prior testing, to capture
any potential debris being ejected during the test. The BOV test in accordance with the P&WC OHM
3034343 was attempted, however, it was not possible to pressurize the Px cavity. No debris were observed
on the adhesive strip after the test.

Disassembly:

Prior to disassembly, the location of the tear in the diaphragm was marked on the housing and on the
piston (Figure 1, Figure 2, Figure 3 and Figure 4). The orifice and metering plug were removed and found
unremarkable. A translucent fluid was observed on the piston retainer, in the vicinity of the tear in the
diaphragm. The retainer was sent to the chemical laboratory to determine the nature of the fluid.

The piston showed ink transfer from the diaphragm identification marking. The ink transfer from the
diaphragm to the piston face were located 120 degrees from the as-removed position. The piston clocking
(“X” mark) to the diaphragm and housing transfer port was also rotated by 120 degrees (Figure 5).

The diaphragm was torn from the edge of the retainer to piston surface to approximately the end of the
rolling surface (Figure 6, Figure 7). To verify the integrity of the diaphragm, the diaphragm was torn by

Page 4 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

hand in an inconspicuous area, in the area clamping between the housing and the cover. The diaphragm
was easily torn by hand. Some level of debonding between the diaphragm fabric and rubber was observed
under magnification in the vicinity of the diaphragm piston bolt holes (Figure 8).

Droplets of clear gummy fluid were observed on the diaphragm Px side, in an area matching the location
of the clear fluid observed on the piston retainer. Two metal chips were observed loosely attached by the
gummy clear fluid on the diaphragm surface away from the tear (Figure 9). The fracture surface near the
diaphragm retainer showed accumulation of debris, while the remainder of the fracture surfaces were
generally free of debris (Figure 10, Figure 11). The debris outlining the retainer on the diaphragm was
uniform around the circumference, except in the area of the tear, where the debris did not align with the
retainer outline (Figure 11). The reenforcing fabric on the P2.5 side was observed exposed and frayed
above the orange elastomer, near the tear of the diaphragm. Rows of strands loop were disengaged from
their neighboring strands in the waiving in the frayed area (Figure 12). The tear was observed propagating
in the diaphragm material along the main axis of the reenforcing fabric on the P2.5 side. Delamination of
the Px side disk-shaped reenforcing fabric as well as stretched orange elastomer were observed at the tear
in the diaphragm (Figure 13).

The opposite side of the diaphragm surface resting on the bleed valve piston, 180° from the tear, showed
the fabric was exposed on top of the orange elastomer, with the strands frayed and the orange elastomer
worn in this area (Figure 14). The size of the frayed and worn elastomer was approximately the same size
as the tear.

Figure 1 Figure 2

Page 5 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

Figure 3 Figure 4

Tear location Diaphragm ink transfer

“X” mark (reproduced

from side of piston) to be
aligned to transfer port.

Figure 5

Page 6 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

Approximate tear length

BOV open position

Figure 6

Piston in the closed position

Figure 7 Figure 8

Figure 9

Page 7 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

Debris accumulation

Figure 10 Figure 11

P2.5 side

Px side

Figure 12 Figure 13

Page 8 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

Opposite side of tear

Figure 14

Chemical laboratory findings:

The bleed valve piston retainer was sent to the chemical laboratory in its seal bag for analysis of the clear
fluid. The analysis of the clear viscous fluid by Fourier Transform Infrared (FTIR) detected a
fluorosilicone composition.

Accessory name: Fuel control unit (FCU)

Note: This unit S/N was the one installed on the engine at initial engine delivery.

External condition and attachments:

The external surfaces were generally clean. The condition and throttle levers were rotating by hand on
their full range. The throttle shaft lever was contacting the housing at the maximum position (Figure 15).
The condition lever was returning to the unloading valve cap under its own spring tension. The input shaft
was rotating by hand. There was a light deposit of soil in the cavity of the bellows adjustment nut. The
identification plate was marked OH 11/30/16. The anti-tamper paint was missing on one of the backplate
screw. The lockwire seal was missing on the dead band adjustment screw.

Figure 15

Page 9 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

Test results:

The FCU was tested in accordance with the CMM 73-20-08 and the following deviations were
observed.

The pressure and internal leakage test, test point (TP) 1.1 was 40 psid below the CMM minimum limit.
The condition lever and shut down test TP 3.1a low idle was 32 RPM above the CMM maximum limit
and the TP3.2a high idle was 265 RPM below the CMM minimum limit. The power lever test TP 4.1
maximum reverse was 285 RPM below the CMM minimum limit and the TP 4.2 maximum forward was
63 RPM above the CMM maximum limit. The TP 4.6a dead band was 10 degrees above the CMM
maximum limit and the maximum forward power lever stop angle was 4 degrees below the CMM
minimum limit. The deviations observed on test points 3 and 4 are permissible field adjustments. The
governor droop schedule TP 5.2 was 16 pounds per hour (PPH) below the CMM minimum limit. The start
and acceleration schedule TP 9.2 was 2 PPH above the CMM maximum limit. The TP 9.3 was 2 PPH, 3
PPH and 2 PPH above the CMM maximum limit. The TP 9.4 was 5 PPH above the CMM maximum limit
at each of the three power lever settings. The TP 9.5 was 2 PPH above the CMM maximum limit at each
of the three power lever settings. The TP 9.6 was 3 PPH, 2 PPH and 2 PPH above the CMM maximum
limit. The TP 9.7 was 2 PPH above the CMM maximum limit.

The FCU was calibrated using the field adjustments and the dead band adjustment with the following
results.

The condition lever and shut down test TP 3.1a low idle, the TP 3.2a high idle, the power lever test TP
4.1 maximum reverse, the TP 4.2 maximum forward, the TP 4.6a and the governor droop schedule TP 5.2
were returned within the CMM limits.

Disassembly:

Based on the test results, it was agreed not to disassemble the FCU.

Accessory name: Flow divider valve (FDV)

Note: This unit S/N was not the one installed on the engine at initial engine delivery.

External condition and attachments:

There was an tag on the external surface. The external surfaces were soiled. The inlet and dump
fittings as well as the primary and secondary outlet ports were unremarkable. The lockwire was present
and in good condition.

Test results:

The FDV was satisfactorily tested in accordance with the CMM 73-10-01.

Page 10 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

Disassembly:

Based on the test results, it was agreed not to disassemble the FDV.

Accessory name: Fuel nozzle (fuel manifold adapter) set

Note: It was not possible to determine if these nozzles were on the engine at initial engine delivery.

External condition and attachments:

The external surfaces were soiled and the fuel nozzle stems and nozzle tips were covered by a dark
substance similar to soot (Figure 16). The tab washers were in good condition and correctly installed.

Typical
condition

Stem

Tip

Figure 16

Test results:

The fuel nozzles were tested in accordance with the P&WC overhaul manual (OHM) 3034343 and the
following deviations were noted.

The #14 position fuel nozzle spray angle was 1 degree above the OHM maximum limit. All fuel nozzles
showed some level of streakiness, characteristic of fuel nozzles returned from service.

Page 11 of 12
Engine Controls Accessories Technical Support
Accessories Investigation

Accessory name: Propellor governor (CSU)

Note: This unit S/N was not the one installed on the engine at initial engine delivery.

External condition and attachments:

The external surfaces were covered with soil. The reset arm was rotating by hand on its full travel and was
returning to the original position on its own spring tension. The speed lever was rotating by hand up to the
feathering valve piston but the piston was not moving inwards. The piston external diameter showed
presence of corrosion. The speed lever was not returning to the maximum stop screw by its own spring
tension. The electrical connector showed deposits on the contacts. The contact pins appeared straight. The
interface seal was not touching the contact pins outside diameters. There was an label present on the
external surface. The eccentric adjustment screw slot was approximately 20 degrees clockwise from the
datum mark on the retaining plate. The drive shaft and pilot valve were moving freely by hand. The beta
valve was moving by hand and was returning to its original position by its own spring tension. Debris was
present in the Py fitting air passage. The anti tamper paint was present at the right locations. The lockwire
seal was not on the pressure regulator valve plug. The remaining lockwires were in good condition and at
the right locations. The speed adjusting lever was rotated approximately 180 degrees from the
manufacturer’s specifications.

Test results:

The CSU was sent to the manufacturer for testing. The following is an extract from the
manufacturer’s report:

“As received testing showed all test points were within new limits, except the following.
The speed setting maximum RPM was [1 RPM below the CMM minimum limit.]
The control lever angle was [165 degrees below the CMM minimum limit.]
The speed setting lever travel was [1 degree above the CMM maximum limit.]

The control lever angle was significantly out of limits however, the lever angle is often adjusted in the
field for engine rigging and did not have an impact on the unit performance. Although there were a few
test points out of limits, the unit showed full functionality. Due to these findings, there was no evidence
of a pre-existing condition that would have precluded normal governor operation.”

Page 12 of 12