Shopping List

AGE
1. Executive summary
The following 5 operators: Viva, Jet, Han, East and Az have reported hydraulic leakage from the Reservoir of EV
Torque Link Damper P/N 10-450701-000, at the interface with the TLD body.
Since January 2019 until March 2020, 63 TLD Reservoir leakage occurences have been reported to Sf LS. 57%
of these cases were reported by Viva Air Colombia.
The leakages occur on A320 only (CEO and NEO) within 5 to 13 month after the aircraft delivery.
In order to understand the leakage cause, the following units/parts have been investigated:
- 2 complete TLDs retrieved from Viva Air Colombia after leakage report (S/N MG4419 and MG4672),
- 3 lockwires retrieved from TLD units fitted on VVC A/C and removed after a leakage report,
- 1 Reservoir coming from Jet Smart
1 more unit from VVC will be investigated to confirm the laboratory report findings.
This report presents the root cause analysis performed to understand the reason of the TLD Reservoir leakages.
Despite all actions carried out in agreement between Ailus and Sf LS, no clear root cause linked to the TLD itself
has been identified.
Laboratory investigations brought evidences that leakages are most probably linked to a or several external
parameters which would lead to untightening of the reservoir, therefore leading to a leakage.
These external parameters are suspected to be related to :

- The evidence of Collins brake configuration correlation with each aircraft TLD failure (PN 2-1759-1)
- Unusual landing, take-off and taxiing operation impacting the Torque Link Damper assemblies.
-
Page 1/52
This document is the property of Sf Landing Systems and shall not be reproduced or communicated without its prior authorization
Ce document est la propriété de Sf Landing Systems et ne peut être reproduit ou communiqué sans son autorisation
Table of contents
1. Executive summary 1
2. Abbreviations 3
3. References 3
4. Identification and description 3
5. Assembly desciption 4
6. Maintenance 7
7. Event description 8
8. Leakage cases 10
9. Investigation 16
9.1. Investigation of leaking units 16
9.1.1. In-service history 16
9.1.2. MRO troubleshooting report 17
9.1.3. Laboratory Investigations on Torque Link Dampers 18
9.2. Laboratory Investigations on lockwire 22
9.3. Laboratory Investigations of broken Reservoir 23
9.4. Breakaway Torque Test 26
9.5. Conclusion of the investigation: 26
10. Root Cause Analysis 27
10.1. Material 29
10.2. Manufacturing 33
10.3. Design 37
10.4. Installation 39
10.5. Environment 41
10.6. Maintenance 48
11. Airworthiness Assessment 50
12. Conclusions & Recommendations 50
Page 2/52
2. Abbreviations
A/C - Aircraft
AZU - Az Airline
CES - East Airline
CSN - Cycles Since New
EV - Evolution
FC - Flight Cycles
FSE - Field Service Engineer
FH - Flight hours
HNA - Hai
ID - Inner Diameter
JAT - Jet
MLG - Main Landing Gear
MRO - Maintenance, Repair and Operation
NFF - No Fault Found
OD - Outer Diameter
P/N - Part number
PSE - Product Support Engineer
S/A - Sub-assembly
TA - Technical Adaptation
TLD - Torque Link Damper P/N 10-450701-000
VVC - Viva Air Colombia Airline
3. References
[1] MAF1-0005-2167 : Introduction of EV MLG’s to meet the NEO loading requirements

[2] Lab report 19/4078 : Materials Laboratory Investigation Report A320 MLG Torque Link Dampers
[3] TE-GL-00155593: Ailus Single Aisle Family Torque Link Damper Cap End Leakages
[4] Lab report 20/4004: A320 MLG Torque Link Damper Lockwire MSN8716
[5] Lab report 20/4016: A320 MLG Torque Link Damper Lockwire MSN8994
[6] NAS 1613: Packing, preformed, ethylene propylene rubber
[7] Lab report 19/4105: A320 MLG Torque Link Damper Unit MSN8493
[8] EDES2-0005-2559: Tightening torque of Torque Link Damper
[9] SDF/B83-02/N/09/1234 17th January 1991 : A320 MLG Investigations to determine the volume of oil
contained in a fully charged Torque Link Damper and the fluid loss following loss of self contained
reservoir
[10] BAE/FIL/DYN-R1/A320/68/90 22nd December 1990 : A320 MLG Torque Link Damper Effect of loss of
reservoir on damper performance
[11] SA32RP1263861: MLG Torque Link Damper removal evaluation
4. Identification and description

a. Unit: Torque Link Damper
b. Part number: 10-450701-000
c. Aircraft type affected : A320 CEO and NEO
d. Customer: VVC, JAT, HNA, AZU, CES
e. Aircraft MSN affected Refer to table in appendix 1
5. Assembly desciption
The Torque Link Damper (TLD) was introduced on Single Aisle MLG as a solution for gear vibrations
phenomenon absorbtion during touchdown. During A320 flight test and early service experience significant Main
Landing Gear vibration events were reported at touchdown. These vibrations did, on occasion, result in minor
damage to airframe and to gear [11].
The TLD consists in a lateral shock absorber composed of a spring-centered, two-way unit containing its own
reservoir of hydraulic fluid (Figure 1). The function of the TLD is to absorb vibrations from the Torque Links to
which it is attached. As the Torque Links assembly is not handed, the TLD is attached on the left side of the
Torque Link assembly on both the left and right Main Landing Gear (MLG) (Figure 2).
FIGURE 1 : TORQUE LINK DAMPER UNIT FIGURE 2 : TORQUE LINK DAMPER FITTED ON THE
MLG
The TLD includes a Housing subassembly, a Cap subassembly and a Reservoir subassembly. It also has a check
valve allowing hydraulic fluid going into the Damper when it is filled and pressurized.
The Housing and Cap subassemblies make a chamber for a piston. A piston rod extends equally on the two sides
of the piston. On each side of the piston, an equal number of disk springs keep the piston at the center of the
chamber (Figure 4).
A Reservoir piston, a Spring, a Cage and a Barrel are the primary components that make the Reservoir Sub-
assembly. Its function is to contain a sufficient quantity of hydraulic fluid for the correct operation of the Damper.
When the Damper is filled and pressurized with hydraulic fluid, the reservoir piston moves through the Barrel. At
the same time the cage moves to compress the spring. When the cage moves to the correct position along the
Barrel, the word FULL shows. If the amount of hydraulic fluid in the Damper is less than the necessary amount,
the word REFILL shows.
FULL LINE
REFILL LINE
FIGURE 3: TLD FILLED AND PRESSURIZED
VALVE
RESERVOIR
SPRING
HOUSING
BARREL (from RESERVOIR

VALVE
S/A)
Seal
FIGURE 4: CROSS SECTION OF THE TORQUE LINK DAMPER
The Reservoir Barrel which is made of steel is screwed thanks to an Adaptor into the Damper Housing
Subassembly, which is made of aluminium, even though they have different thread dimensions.
The adaptor is made of steel and its outer diameter and inner diameter thread dimensions match respectively with
the threads of the Housing and Reservoir subassembly.
The Adaptor sits at the bottom of the Housing subassembly. After it is assembled, it forms a groove where sits a
static O-ring seal which ensures the sealing of the assembly (Figure 5).
FIGURE 5: ADAPTOR ASSEMBLED INTO HOUSING

FIGURE 6: LOCKWIRE
S/A
The Reservoir and Housing are maintained in rotation through a lockwire, as shown in Figure 6. The function of
the lockwire is to prevent the unthightening between the Reservoir and the Housing.
This Reservoir/Housing joint assembly has not changed since the Torque Link Damper was introduced in the
MLG Single Aisle design, before the A320 EV introduction.
Ref [1] describes a design change that was introduced for EV variant. Depth of the the holes on the Housing S/A
interfacing with an attachment bracket has been increased to alleviate binding problem with bolts.
This minor design change has no influence on the Reservoir/Housing assembly and sealing function.
6. Maintenance
The only in-service maintenance actions performed on the TLD should be the replenishment of the TLD as per
AMM task 12-12-32-611-006-A. During this maintenance operation the top and bottom valve must be untightened
and their associated lockwire must be detached. The lockwire attaching the Reservoir and Housing S/A is not
removed during this operation.
Other AMM tasks are related to the TLD assembly and deals with troubleshooting operations :
- Inspection of the Torque-Link Damper-Assemblies as per AMM task 32-11-27-210-001-A
- Functionnal check of MLG Torque Link Damper Fluid Level as per AMM 32-11-11-200-001-A
The Removal and installation of the TLD are described in the following documents:
- AMM 32-11-29-000-001-A and AMM 32-11-29-400-001-A
7. Event description
The following 5 operators: Viva, Jet, Hai, East and Az have reported hydraulic leakage from the Reservoir of EV
Torque Link Damper P/N 10-450701-000, at the interface with the TLD Housing.
The leakages occured on A320 only (CEO and NEO).
From January 2019 until March 2020, operators reported 63 TLD leakage occurences.57% of these cases were
reported by Viva Air Colombia alone.
The leaks were discovered during walk-around inspection. The number of FC to failure for the first leak on each
A/C varies between 530 – 2573 which is around 5 to 13 month after the aircraft delivery.
At the time of the event, all affected a/c were fitted with Goodrich brakes P/N 2-1759-1.
Operators have reported :

1. The Torque Link Damper is empty and hydraulic fluid is found on the floor (Figure 7)
2. The lockwire that attaches the Reservoir and the Housing sub assembly is broken (Figure 8)
3. The Reservoir is untightened and can be rotated by hand
VVC, JAT and AZU airlines are based in South America. HNA and CES are based in China.
Fluid leaking from the

Reservoir/Housing
interface
FIGURE 7: RESERVOIR EMPTY (AT REFILL LINE) AND HYDRAULIC FLUID ON THE FLOOR
FIGURE 8: LOCKWIRE IS BROKEN
8. Leakage cases
From January 2019. to March 2020, operators reported 63 TLD leakage occurences, they are all listed in
appendix A.
Following information can be deduced from the raw data:
1. The most concerned operator is Viva Air Colombia with 57% of the reported cases (see Figure 9)
2. All MSN concerned are recent, they have less than 13 months in service when the first event is reported
(Figure 10)
3. The TLD leakage can occur either on LH or RH side (see Figure 11)
4. The TLD leakage can occur several times on the same MSN and same Gear side even if it is fitted with
different Torque Link Dampers (see Figure 11, Figure 12 and Figure 13)
5. The reported leaking Torque Link Dampers were either:
o fitted on the gear since aircraft delivery, units were coming from the supplier Middlesex
o brand new spare units coming from the stock and supplied by the supplier Middlesex
o repaired units from Queretaro or Singapore MRO stock.
6. The concerned MLGs where assembled in both Bidos and Gloucester (see Table 1)
7. Leakages occur randomly along the year (Figure 14)
HNA
11%
AZU
2%
VVC
CES JAT
24% CES
VVC AZU
57% HNA
JAT
6%
FIGURE 9: REPARTITION OF LEAKAGE CASES AMONG OPERATORS

14
13
12
10.4 10.6 10.6
10
9.2
8.9 8.9
8.3 8.4 8.2
7.7 7.8 7.6 7.7 7.9
8 7.6 7.6
7.3
6.7 6.7 6.6 6.6
6.3
5.9
6
5.3 5.4
5.1
Age at the first event (month)
FIGURE 10: AGE OF THE AIRCRAFT IN MONTH WHEN FIRST EVENT IS REPORTED
Where built MSN Airline Airline
Glos MSN07811 TIANJIN AIRLINES HNA

Bidos MSN08122 BEIJING CAPITAL AIRLINES HNA
Glos MSN08148 EASTERN AIRLINES CES
Glos MSN08204 BEIJING CAPITAL AIRLINES HNA
Glos MSN08475 LUCKY AIR HNA
Glos MSN08479 VIVA AIR VVC
Bidos MSN08493 JETSMART JAT
Glos MSN08539 JETSMART JAT
Bidos MSN08994 VIVA AIR VVC
TABLE 1: BUILDING LOCATION OF THE GEAR
Cumulative CSN per TLD before leakage - Viva Air Colombia
3500
3000
2500 CSN 7th TLD

CSN 6th TLD
CSN 5th TLD
2000 CSN 4th TLD
CSN 3rd TLD
CSN CSN 2nd TLD
CSN 1st TLD
1500
1000
500
FIGURE 11: VIVA AIR LEAKAGE CASES (LH/RH SPLIT)
Cumulative CSN per TLD before leakage - China Eastern

2500
2000
CSN 5th leak

1500 CSN 4th leak
CSN 3rd leak
CSN CSN 2nd leak
1000 CSN 1st leak
500
0
MSN MSN MSN MSN MSN MSN
8148 8301 8406 8449 8488 8544
FIGURE 12: CES LEAKAGE CASES
Cumulative CSN per TLD before leakage - HNA

1200
1000
800
CSN 4th leak
CSN 3rd leak
600 CSN 2nd leak
CSN CSN 1st leak
400
200
0
MSN MSN MSN MSN MSN
7811 8122 8204 8290 8475
FIGURE 13: HNA LEAKAGE CASES
Leakage per month in 2019 and 2020

8
7
6
5
4 Viva Air
3 CES
2
1
0
y y h il ay ne ly st r r r er
u ar uar arc Apr M Ju Ju ugu mbe obe be mb
n r M e t m e
Ja Feb A pt Oc ove Dec
Se N
FIGURE 14: DISTRIBUTION OF LEAKAGES PER MONTH IN 2019 AND 2020 – VVC AND CES
Sf Landing Systems retrieved two units from Viva Air Colombia for further investigation.
Both MG4419 and MG4672 TLD’s and their accompanying seals were sent to the Materials and Processes
Laboratory, Gloucester. Seals from a non-leaking TLD unit MG3111 were also provided for investigation.
Broken lockwires from units MG4419 and MG4672 could not be retrieved therefore lockwires from units showing
the similar failure mode as the investigated unit were retrieved to investigate their breaking mode.
9. Investigation
Several investigations were conducted in order to understand the failure mode and to determine the root cause of
the event. The synthesis of the investigations are described in the following sections.
9.1. Investigation of leaking units

9.1.1. In-service history
Below is presented the background of the two leaking TLDs from Viva Air Colombia retrieved for investigation.
MG4419:
The MG4419 TLD was assembled in August 2018 and fitted on MSN08657 MLG RH in January 2019. The TLD
was found at a “low level” on the 15th of July, the 26th of July (when a leak was also reported) and the 6th of
August, and was replenished according to AMM 12-12-32-611-006-A on each occasion. The unit was finally
replaced on 20/09/2019, having undergone 1426 cycles since new (CSN) and 1384 hours since new.
FIGURE 15: MG4419 IN SERVICE HISTORY
MG4672:
The MG4672 TLD was assembled in October 2018 and replaced MG4159 on MSN08485 MLG RH on the 15th of
July 2019, following multiple reports of refilling MG4159. A leak was then discovered on MG4672 on the 16th of
August 2019 and was replenished according to AMM 12-12-32-611-006-A. MG4672 was then replaced on the
22nd of October 2019, having undergone 105 CSN and 127 hours since new.
FIGURE 16: MG4672 IN SERVICE HISTORY
VVC report including video and photo evidence confirmed:

1. The Torque Link Damper is empty and hydraulic fluid is found on the floor (Figure 7)
2. The lockwire that attaches the reservoir and the housing sub assembly is broken (Figure 8)
3. The reservoir is untigntened and can be rotated only by hand
9.1.2. MRO troubleshooting report
The units MG4419 from MSN08657 and MG4672 from MSN08485 were sent to Sf Gloucester MRO to identify the
leakage path and to understand the failure mode.
However, these units were “serviced” by VVC. In fact, even though MG4419 and MG4672 were reported leaking
through the Reservoir/Housing interface with the lockwire broken and the Reservoir being loose, they were
received with the reservoir tightened and a new lockwire installed.
The lockwire diameter of TLD unit MG4419 did not comply with the diameter requirements (ø<0.8mm).
Even if some evidence could have been lost because of this servicing, the two units were troubleshot and the
report is attached in appendix B.
The two units passed all the function test and did not show any leak, and they could be considered serviceable
(NFF). As a result, the leakage path could not be confirmed.
The troubleshooting demonstrated that as long as the Reservoir is screwed into the Housing there is enough seal
squeeze to ensure the sealing of the unit.
On unit MG4672, the reservoir was untightened by 45 then 180° while it was still pressurized with hydraulic fluid to
check if it would leak but the unit did not show any leakage.
Finally these two units were disassembled and sent to Sf LS Laboratory for further investigations
9.1.3. Laboratory Investigations on Torque Link Dampers
The Lab report 19/4078 [2] presents the lab investigations performed on the units MG4419 and MG4672. The lab
investigations focused on the following parts:
Part title Part number CMM32-12-25

Barrel 201419605 1-70
Cage 201419604 1-80
Spring 860013017 1-100
Adaptor 201419614 2-330
Damper piston 201160693 2-210
Damper housing 55-4507001-00 2-350
O-ring NAS1611-329A 1-120
O-ring NAS1611-211A 1-130
O-ring NAS1611-222A 2-130
O-ring NAS1611-242A 2-140
O-ring NAS1611-222A 2-310
TABLE 2: MG4419 AND MG4672 INSPECTED PARTS
FIGURE 17: INSPECTED TORQUE LINK DAMPER COMPONENTS
The main conclusions of the lab investigation are the following:
9.1.3.1. Detailed visual inspection

Scratches were observed on the cage, barrel and spring components of both TLD’s.
On the barrel, these scratches were located from the full line up to the the flange as illustrated in Figure 18. In this
particular location, the Barrel interfaces only with the Spring. The fact that the scratches are visible above the full
line suggest they were caused by the Spring when it was compressed, therefore, when the reservoir was full with
fluid.
Circular scratches pattern were observed on the Barrel flange. This circular scratch pattern illustrated in Figure 20
was not observed on the TLD which did not experience TLD Reservoir leakages.
These scratches suggest axial and rotationnal movements between the Spring and the Barrel.
They were also observed on the units from other operator such as CES and HNA that experienced Reservoir
leakages.
FULL Line FULL Line
FIGURE 18: BARREL OD CONDITIONS MG4672 VIVA (LEFT) MG4178 CES (RIGHT)
FIGURE 19: BARREL FLANGE CONDITIONS MG4672 VIVA (LEFT) MG4178 CES (RIGHT)
FIGURE 20: BARREL FLANGE CONDITIONS- CIRCULAR SCRATCH PATTERN
The thread of the Barrel, Adaptor and Housing were found in good condition. Some light wear on the housing
thread were noted but it is considered to be normal and would not cause the TLD leakage.
FIGURE 21: MG4419 BARREL, ADAPTOR AND HOUSING THREAD
The MG4419 and MG4672 Cage had circular scratch patterns (similar to the Barrel flange scratches) around the
whole inner shoulder surface and light circumferential scratches around the inner diameter (ID).
FIGURE 22: MG4419 AND MG4672 CAGE HAD CIRCULAR SCRATCH PATTERNS
Both the MG4419 and MG4672 Springs were scratched on both end faces with the scratches having a similar
circular pattern to those seen on the Barrel and Cage flange. This suggests a combination of rotational and radial
movement in contact with the Barrel flange and Cage flange. Axially oriented scratches were also present on the
ID’s of both springs suggesting contact with the barrel OD during loading / unloading of the spring.
FIGURE 23: MG4419 SPRING SCRATCHED ON END FACE (LEFT) AND ID (RIGHT)
A permanent deformation was observed in the lockwire holes of the Barrel and Housing where the lockwire was
assembled. The deformation was probably caused by the lockwire when it was loaded because of the Reservoir
untightening. The orange arrow in Figure 25 indicates the untightening direction and the blue arrow indicates the
orientation of the loading (Figure 24, Figure 25, Figure 26)
The opposed lockwire holes from the Barrel and Housing where the lockwire was not attached are in good
condition.
The fracture faces observed on the broken lockwire confirms the conjecture. In fact the lockwire broke because of
an overload see section 9.2).
FIGURE 24: LOCKWIRE ASSEMBLY
FIGURE 25: PERMANENT DEFORMATION OF BARREL LOCKWIRE HOLE (LEFT)
FIGURE 26: PERMANENT DEFORMATION OF HOUSING LOCKWIRE HOLE (LEFT)

9.1.3.2. Seals inspection
The seals located in the Housing/Reservoir interface showed light wear but were still in good condition to ensure
the sealing.
FIGURE 27: HOUSING/RESERVOIR INTERFACE SEAL CONDITION
Minor swelling was observed in the 2 returned units. In fact, the inner diameters (ID’s) of all measured seals were
greater than specification; including the seals from the non-leaking MG3111 TLD unit.
The hardness of both TLD’s NAS1611-222A (2-310) damper housing seals were below specification, while the
equivalent MG3111 seal was within specification.
These dimensions deviation and hardness decrease were expected after exposure to hydraulic fluids and
temperature. Dimensions and Hardness data are provided in appendix.
9.1.3.3. Hydraulic fluid

The MG4419 and MG4672 hydraulic fluids were identified as Skydrol LD-4 low density phosphate ester hydraulic
fluid and Skydrol 500B-4 high density phosphate ester hydraulic fluid respectively.
The fluid properties in terms of water content, viscosity and acidity were within specification.
The Reservoir leakage occurred regardeless of the fluid used to operate the TLD.
9.1.3.4. Dimension
Adaptor, Housing, Barrel, Spring and damper Piston dimensions of TLD units MG4419 and MG4672 are
presented in appendix alongside data from units MG3432 and MG3972 that did not experience Reservoir
Leakage.
Dimensions in locations pertinent to the investigation were within specification.
The majority of the Adaptor’s dimensions met the drawing requirements, however the MG4419 major thread
diameter exceeded specification when measured in a specific orientation but the average dimension is still in
compliance with drawing requirements.
Some dimensions were measured out of limits for units that experienced Reservoir Leakage as well as those
which did not and these deviations are very small and could be due to the measurement tool precision. Therefore
these deviation are considered not critical and would not cause the leakages (see dimension 10 and 17 in
appendix C).
9.1.3.5. Material and Hardness

The hardness of all tested metallic components was within specification.
The electrical conductivity of both damper Housings which are made from Aluminium exceeded the minimum
requirements.
The microstructures of the Barrel, Housing and Adaptor were consistent with their respective materials and heat
treatments.
9.2. Laboratory Investigations on lockwire

The Lab reports 20/4004 [4] and 20/4016 [5] present the lab investigation performed on the lockwire coming from
units MG1007 from MSN08716 and MG5242 from MSN08994.
Unit MG1007 and associated lockwire operated 7FC after being fitted on the aircraft. And MG5242 performed
1717FC since new.
The dimensions of the lockwire comply with the requirements (ø0.8 ± 0.04mm).
The chemical composition of the wire was consistent with 304 stainless steel grade, which is in compliance with
PCS-7610.
The examination of the lockwires fracture faces show it was evident that the wire had been subjected to high
stresses and fractured due to overload. Substantial elongation was seen, as well as micro-voids on the fracture
face which are typical features of ductile overload.
The diameter of the fractured wire was measured at approximately 0.81 mm away from the fracture surfaces and
approximately 0.65 mm at one fracture end indicating necking, which is typical of ductile overload.
The scanning electron microscope (SEM) examination found dimples on the fracture surfaces, typical of ductile
overload without any particular features at the initiation areas. The orientation of the dimples suggested that the
overload mechanism was due to shear stress (see Figure 28).
FIGURE 28: EXTRACT FROM REF [4] – LOCKWIRE FRACTURE FACE
9.3. Laboratory Investigations of broken Reservoir
On June 9th 2019, operator JetSmart (JAT) reported unit MG4170 was “inoperative” and the lockwire was “loose”.
JAT informed the lockwire was replaced, however, no evidence of the lockwire condition was communicated
therefore it was not confirmed the lockwire was loose or broken.
On June 10th, JAT replenished the TLD, suggesting it was empty from hydraulic fluid.
Later on, on June 20th 2019, JAT reported the Reservoir from unit MG4170 detached and was found on the
runway after take-off at airport of Calama (CJC - Chile). A/C continued the flight to Santiago SCL and landed
without problem.
It was suspected the Reservoir detachment event is related to the Reservoir leakage and the subject MG4170
was investigated. Only the reservoir from TLD unit MG4170 was retrieved and the outcome of lab investigation is
presented in Ref [7].
FIGURE 29: RESERVOIR FROM TLD UNIT MG4170 AT RECEPTION
The visual inspection conducted on the Barrel showed similar findings as the 2 other units investigated from VVC:
- Scratches on the Barrel OD: beyond the “REFILL” oil level line suggesting that the damper unit operated
at low fluid levels
- Scratches on the Barrel flanges: suggesting that the spring rotated and fretted against the barrel flange.
- Plastic deformation in the lockwire holes: suggesting the area was loaded due to the Reservoir
untightening.
FIGURE 30: MG4170 BARREL DAMAGE
The Lab report Ref [7] indicates the barrel fracture initiated due to repetitive loading. The fracture initiated at the
175° location, from the radius between the flange and the first thread.
FIGURE 31: PHOTOGRAPH OF THE FRACTURE SURFACE
Only a very small number of striations were seen over the 0.120 mm of propagation, indicating high mean stress
from the onset in this area. Multiple associated cracks were evident in the surrounding areas of the initiation area
which indicates that substantial stress levels were applied
The final fracture occurred when the crack propagated by overload to the first thread root at the 0° location where
the final ductile fracture occurred.
The Barrel hardness and microstructure showed the material was within the specification requirement for 17-4PH
in the H900 condition as specified in the drawing.
FIGURE 32: SEM PHOTOGRAPHS

9.4. Breakaway Torque Test
Sf LS measured the torque required to break the lockwire when the Reservoir is only tightened by hand (not
torque tightened) to have an idea of the amount of force necessary to break it. Results of this test is presented in
Table 3.
# Breakaway
Locking TLD S/N Source Comments
Torque [Nm]
1 As per test procedure, reservoir
0.79mm lock
23,1 MG4093 Middlesex only tighten by hand so no residual
wire
torque in the thread.
0.79mm lock
wire
0.79mm lock
wire
0.79mm lock
wire
TABLE 3: TORQUE REQUIRED TO BREAK THE LOCK WIRE
9.5. Conclusion of the investigation:

The trouble shooting investigation performed on the returned units showed the Torque Link Damper passed the
functional test and no leakage was identified. The TLD passes the required CMM tests after the Reservoir has
been tightened back.
Thanks to the evidence found on the parts, the failure scenario was identified to be the following:
1) Loss of the torque preload caused by unknown external factors
2) Reservoir untightening driven by external factors (outside of the TLD)
3) Lockwire breaking
4) Reservoir continues to untighten
5) Hydraulic leakage when seal squeeze is lost
The lab investigation performed on the TLD did not reveal any significant findings incriminating the Torque Link
Damper itself.
A root cause analysis has been conducted to establish the actions in order to find the potential causes and
contributors to the TLD leakage. This root cause analysis in presented in the following section.
10. Root Cause Analysis
Ailus and Sf LS followed the PPS/MFT methodology to determine the root cause of the event. Several potential
contributor to the event were raised and grouped into the following 6 main headers:
A. Material
B. Manufacturing
C. Design
D. Installation/assembly Maintenance
E. Environment
F. Maintenance
Each potential cause has been evaluated based on the numerical analysis, tests results, Sf LS background and
history, in-service feedback, engineering judgement and whatever is appropriate to state on the implication in the
Torque Link Damper leakage issue.
For each potential cause we assessed the likelihood of the contributor in order to categorize and prioritize the
investigations.
We categorized them as:
- Likely contributor
- Potential contributor
- Unlikely contributor
The outcome of this analysis is presented in the Ishikawa diagram on :

Sf Landing Systems Report Number: TE-GL-00197158
CAGE CODE K0654 Issue / édition : 1
Safran Landing Systems Tech Event Reference: TE-VE-00197158

CAGE CODE K0654
Issue / édition: DRAFT
A. Material B. Manufacturing C. Design
A1– Shrinkage or swelling of the seal (1-

130) leading to reservoir leakage B1 – Incorrect lockwire installation B10 – Disk spring is non conforming C1 – No torque value specified for the reservoir assembly
A2 – Non-conformity lockwire material or B2– Seal groove not manufactured to the drawing B11 – Surface roughness in the sealing and C2 – Non-optimised locking design
hardness diameter of housing
B3 – Change of lockwire specification or supplier C3 – Contact between the lockwire and the TLD
A3– Use of incorrect hydraulic fluid or use of mix B12 – Variability in tightening the reservoir assembly resulting in chaffing
fluid B4 – Change of seal specification or supplier
B13 – Sharp edges on the lockwire housing/reservoir
B5 – Wrong storage conditions of assembled unit holes that cut the lockwire C4– Design of the seal groove doesn’t provide
A4 – Non conformity of seals material or hardness
adequate the sealing
A5 – Non conformity of adaptor material or hardness B6 – Incorrect dimensions of thread diameter B14 –Manufacturing non conformity of the damper piston
or length area of barrel C5– Stack up between lockwire and
A6 – Non conformity of barrel material or hardness
B15 – Manufacturing non conformances (concession) of reservoir cage
B7 – Machining of the thread area on TLD housing
PN at the interfaces are not accepted
A7 – Non conformity of housing material or hardness is non conform
B8 – Machining of external or internal thread B16 – Manufacturing changes

Likely contributor of the adaptor is non conform
B17 – Testing equipment non conform (not
Potential contributor calibrated, not clean etc)
B9 – Housing dimensions is non conform
Unlikely contributor
Lockwire broken,
reservoir leakage
D6 – Lack of training of the workforce at manufacturer E14 – Approach and landing condition
E7 – Damage to the TLD during operation (FOD)
D5 – Reservoir sub assemblies not fitted properly E13 – Altitude of operator’s base
E6 – High level of vibration due to runway condition
D4 – The adaptor is tightened incorrectly at supplier E5 – Abnormal level of vibration due to the wheels/tyres E12 – TLD environment
temperature
E4 – High/Abnormal level of vibration due to the brakes E11 – TLD duty cycle
D3 – Lockwire not installed as PCS 7610 at MRO
or operator E10 – Braking practices of operator in operation
E3 – Contamination of the hydraulic fluid
F2 – Damage to the TLD during
D2 – Damage to the TLD during Installation of the maintenance
TLD on the Gear E2 – Use of a non-adapted type of hydraulic fluid E9 – The spring 1-100 rotates and drive
the untightening of the barrel 1-70
F1 – Incorrect servicing procedure by
E1 - Abnormal events on the main landing gear i.e.
D1 - Incorrect installation procedure operators
hard landing, off runways, undamped free fall E8 – Taxiing condition
D. Installation/Assembly E. Operation/Environment F. Maintenance
FIGURE 33: A320 MLG TORQUE LINK DAMPER RESERVOIR LEAKAGES ISHIKAWA FISHBONE DIAGRAM
Page 28/52
10.1. Material
10.1.1. Shrinkage or swelling of the seal (1-130) leading to reservoir leakage:
Lab investigation reports presented in Ref [2] shows minor swelling was observed in 2 returned units.
In fact, the inner diameters (ID’s) of all measured seals were greater than specification; including the seals from
the non-leaking MG3111 TLD unit.
These dimensions deviations were expected after exposure to hydraulic fluids and temperature.
In fact NAS1613 standard define the swell increase and hardness decrease to be the following after exposure to
hydraulic fluid and temperature see Ref [6] and extracted data in Figure 34 and Figure 35.
FIGURE 34: VOLUMETRIC SWELL OF ETHYLENE PROPYLENE RUBBER SEALS AFTER EXPOSURE TO
HYDRAULIC FLUID AND TEMPERATURE
FIGURE 35: HARDNESS DECREASE OF ETHYLENE PROPYLENE RUBBER SEALS AFTER EXPOSURE TO
HYDRAULIC FLUID AND TEMPERATURE
As a result, the seals are behaving as expected.

In addition, the swell of the static seal NAS1611-211A (1-130) (located at the Reservoir/Housing interface)
increases the seal squeeze and enhances the sealing arrangement of the assembly.
The static seal NAS1611-211A (1-130) was found in good condition and discard the seal as a potential contributor
to the leakage (see Ref [2]).
Plus two different type of Hydraulic fluids were found in the investigated units. The fluid influences the seal
behavior. The units were reported leaking when filled with a mix of High and Low Density type 4 fluid (refer to
chapter 10.1.3).
FIGURE 36: RESERVOIR/HOUSING INTERFACE SEAL (1-130)

Conclusion:
10.1.2. Non-conforming lockwire material or hardness

Lab investigation reports presented in Ref [4] and [5] show the lockwire samples retrieved on in-service units are
consistent with stainless steel according PCS-7610 (Z2CN18-10, AISI 304).
Middlesex lockwire CoC confirmed the lockwire used is Stainless steel in accordance with AS44725-2 and
PCS7610. (refer to appendix D).
VVC confirmed lockwire presented on Figure 37 was used to ensure the locking of the reservoir. It is in
compliance with specification PCS-7610.
Lab investigation reports presented in Ref [7] indicates the lockwire material is Ni based alloy instead of stainless
steel. However this lockwire was not the one originally fitted on the TLD, Ailus reports shows it was assembled by
operator JAT.
Conclusion:
FIGURE 37: LOCKWIRE USED BY VVC Ø0.8MM, 302/304 STAINLESS STEEL
10.1.3. Use of incorrect hydraulic fluid or use of fluid mix

Ref [2] shows the result of the fluid analysis of the two units inspected. One unit was filled with LD Type 4 and the
other with 500B4.
The supplier uses 500B4 only, in line with the drawing and CMM requirements expectations.
CMM includes 500B4. AMM includes 5 different fluids. Operators have been requested to confirm what fluid they
use. Viva Air stated they use Hyjet Type 4.
Conclusion:
If fluid type was a contributor to the root cause, this would occur via degradation of the seal. The evidence
available from returned leaking units shows that the seals are not significantly affected (only minor swelling also
typical of non-leaking units). Additionally, units only leak when they are unscrewed, they do not leak when
reservoir is screwed in, therefore the issue cannot be related to the seal.
10.1.4. Non-conformity of seal material or hardness

Lab investigation reports presented in Ref [2] shows Minor swelling was observed in 2 returned units along side
hardness decrease.
The hardness of both TLD’s NAS1611-222A (2-310) damper housing seals were below specification, while the
equivalent MG3111 seal was still within specification.
These dimensions deviation and hardness decrease were expected after exposure to hydraulic fluids and
temperature.
The wear observed on seals 2-130 and 2-310 were probably caused by the contact with the damper piston. This
level of wear has not caused any leakage during the troubleshooting and no cap end leakage was reported in
service.
The seals inspection and the troubleshooting conducted at MRO confirmed the leakage path to be different from
the previous TLD investigation that is described in Ref [3].
Conclusion:
The seals are behaving as expected. No non-conformity was identified on returned units.
10.1.5. Non-conformity of Adaptor material or hardness

Lab investigation reports presented in Ref [2] shows the hardness of the Adaptor is within specification.
The microstructures of the Adaptor is consistent with 17-4 PH material and its associated heat treatments.
Conclusion
10.1.6. Non-conformity of Barrel material or hardness

Lab investigation reports presented in Ref [2] shows the hardness of the Barrel is within specification.
The microstructures of the Barrel is consistent with the 17-4 PH material and its associated heat treatments.
Conclusion :
10.1.7. Non-conformity of Housing material or hardness

Lab investigation reports presented in Ref [2] shows the hardness of the Housing is within specification.
The microstructures had acceptable grain size distributions and dispersions of intermetallic precipitates; typical for
this 7000 series aluminium alloy in the T74 condition
Conclusion:
10.1.8. Material Conclusion
Material lines of enquiry are not a contributor to the root cause.

10.2. Manufacturing
10.2.1. Incorrect lockwire installation
Investigations conducted at Viva Air and Middlesex did not reveal any non-compliance to lockwire assembly in
Middlesex and Viva Air.
FIGURE 38: LOCKWIRE INSPECTION OF TLD FROM STOCK (LEFT) AND ON SITE AT MEDELLIN
AIRPORT (RIGHT)
Conclusion:
10.2.2. Seal grove not manufactured to drawing

Laboratory investigation reports presented in Ref [2] show a minor deviation on the radius dimension of the
Housing found in one unit. However, it has been assessed that this deviation does not contribute to the root cause
since it was also observed on unit MG3432 and MG3972 which did not report reservoir leakage. The deviation is
8 microns and it does not detrimentally affect the seal squeeze and the sealing arrangement in this area.
Third and final unit remains to be inspected for completeness. However enough information is available to
discount this route.
Conclusion:
10.2.3. Change of lockwire specification or supplier

Supplier Middlesex confirmed no change in specification or supply has been introduced for lockwire.
Conclusion:
10.2.4. Change of seal specification or supplier

Supplier Middlesex confirmed no change to the seal specification or supply chain has been introduced. In fact the
evidence provided by Middlesex, and presented in appendix shows the seal were sourced from International
seal/Freudenberg in 2017, 2018 and 2019. This period correspond to the assembly date of the reported leaking
TLD. The seals provided come from different batches therefore a seal batch issue is not contributing to the
Reservoir leakage.
Conclusion:
10.2.5. Wrong storage conditions of assembled unit

The TLD units are manufactured and then stored in purpose built storage box. Within a couple of days after
delivery to SLS they are assembled to gears which are then delivered to Ailus. No evidence has been found to
support contribution to the root cause.
Conclusion:
10.2.6. Incorrect dimensions of thread diameter of length area of Barrel

Lab investigation reports presented in Ref [2] shows the thread dimensions and length are in compliance with
specification. No non conformity was identified with Middlesex quality/ manufacturing processes.
Conclusion:
10.2.7. Machining of the thread area on TLD housing is not conforming

Lab investigation reports presented in Ref [2] shows the thread dimensions are in compliance with specification.
No non conformity was identified with Middlesex quality/ manufacturing processes.
Conclusion:
10.2.8. Machining of external or internal thread of the adaptor is non-

conforming
Lab investigation reports presented in Ref [2] shows the thread dimensions are in compliance with specification.
Dimension 10 was measured out of drawing requirements on the leaking on non-leaking unit and therefore this
deviation is not considered a contributor to the Root Cause. No non conformity was identified with Middlesex
quality/ manufacturing processes.
Conclusion:
10.2.9. Housing dimensions is not conforming

Lab investigation reports presented in Ref [2] show minor deviation was identified (dimension 17 and 24) with one
of the returned units. However this has been determined not to be a contributor to the root cause since it is
observed on units that have and have not experienced Reservoir leakage. Middlesex quality/manufacturing
processes did not reveal any non-conformity.
One more unit will be investigated for completeness.
Conclusion:
10.2.10. Disk spring is non conforming

Disk springs (item 190 and 220 from Figure 17) about do not contribute to this failure mode due to design
considerations. Visual inspection conducted showed normal wear caused by in-service operation. No cracks of
defect were detected during the inspection.
Conclusion:
10.2.11. Surface roughness in the sealing and diameter of the housing

Lab investigation reports presented in Ref [2] shows no issue was identified with returned units. In addition,
Middlesex quality/ manufacturing processes did not reveal any non conformity.
Conclusion:
10.2.12. Variability in tightening the reservoir assembly

The contribution of this factor can be downgraded to “unlikely contributor”. In fact the reported Leaking TLD are
coming from different sources : Supplier Middlesex, MRO Queretaro, MRO Singapore. Even though no torque
tightening value was defined in the drawing the different assembly practices from the three different location
reveal that no matter the torque applied, the TLD was reported leaking.
In order to improve the assembly process and reduce the Reservoir variability, an improvement action has been
taken to define the Reservoir torque tightening value (see Ref [8]).
Conclusion:
10.2.13. Sharp edges on the lockwire housing/reservoir holes that cut the
lockwire
Lab investigation reports presented in Ref [4] and [5] show the locking wire failure to be a ductile tensile failure not
a first order shear failure. No initiation site was found. Therefore this failure mode is unlikely and can be ruled out.
Conclusion:
10.2.14. Manufacturing non conformity of the Damper piston
No non-conformity was found on the 2 units inspected (see Ref [2]). The measurement resolution and capability
unable to confidently discern minor measurement results).
Middlesex quality/manufacturing processes reports do not indicate any history of non-conformity.
Conclusion:
10.2.15. Manufacturing non-conformance of PN at the interface are not

accepted
Same as para 10.2.14 above
10.2.16. Manufacturing changes

Middlesex’s report reveals no major manufacturing changes that would contribute to the root cause. Middlesex
have introduced tooling changes in its manufacturing process:
1. Nut runner (auto socket wrench) as of june 2019 to speed up operations (operation 50)
2. New test rig (serial number e11860) has been available since jan 2019
The leaking TLD units were built in 2017, 2018 and 2019, therefore there is no correlation with the tooling
changes introduced.
Conclusion:
10.2.17. Testing equipment non conform (not calibrated, not clean, etc)
Middlesex acceptance test is a mature quality control measure to screen non-conforming parts prior to delivery
therefore not considered a contributor to the root cause.
Conclusion:
10.2.18. Manufacturing conclusion

Any potential root cause related to manufacturing has not been found during this investigation.
The variability of the torque tightening is not considered a contributor (data shows TLD leaking coming from
supplier and also MRO), but needs to be understood.
10.3. Design
10.3.1. No torque value specified for the reservoir thread
TLD drawing does not indicate the Reservoir Torque tightening value. As an improvement action , in order to
standardize assembly process at supplier and MRO and to support operators, Reservoir torque tightening value
is proposed through EDES2-0005-2559 (see Ref [8]).
The TLD units manufactured with the assembly torque defined in Ref [8]) are not yet produced since the EDES is
not approved yet.
In order to support operators, a TA for re-torqueing of units has been provided to Viva Air and Jetsmart Operators.
No feedback to date however it is considered that A/C won't have flown since delivery due to COVID-19.
Conclusion:
The lack of torque specification may be a contributing factor (which makes the "symptoms" more likely to
occur) but not the main contributor to the unscrewing of the reservoir (i.e. the actual root cause). There needs to
be another key element which leads to unscrewing.
10.3.2. Non optimized locking design

The design feature has not changed since the introduction of the TLD, and it is considered mature with no history
of systematic faliures in service.
Conclusion:
10.3.3. Contact between the lockwire and the TLD resulting in chaffing
Lab investigation reports presented in Ref [4] and [5] show the locking wire did not have any indications of
chaffing as a cause of failure. Also no systemic failures is noted in fleet history.
Conclusion:
10.3.4. Design of the seal groove

The design feature has not changed since the introduction of the TLD. There has been no leakage through this
interface in the fleet history.
In addition, in service reports and MRO troubleshooting shows the TLD units do not leak when the reservoir is
tightened, they only leak when the Reservoir is unscrewed, so the failure mode does not put sealing arrangement
capability into question.
Conclusion:
10.3.5. Stack up between lockwire and reservoir cage

Design has proven 30 years of in-service experience with no systemic failure. Also, any sudden change to
tolerance stack up (e.g. sudden production at extreme of tolerances) potentially contributing to the root cause is
not compatible with the issue consistently occurring with few operators.
Conclusion:
10.3.6. Design conclusion

The investigation has revealed the failure mechanism is not related to the design. The lack of torque specification
is not considered an unlikely contributor but an aggravating feature
10.4. Installation
10.4.1. Incorrect installation procedure
Failed units come from both Gloucester and Bidos as shown in Table 1. Also, replacement in service is done via
AMM which applies to entire fleet. In addition, Sf field service engineers confirmed installation performed by Viva
Air line maintenance engineers is adequate during on site inspection. Additionally, early and recurrent would be
more widely observed as a consequence of installation issues but this is not experienced.
Conclusion:
10.4.2. Damage to the TLD due to installation of the TLD on the gear
No evidence of damage was found to date of sealing components and assemblies.
Conclusion:
10.4.3. Lockwire installation

Investigations did not reveal abnormal installation of the lockwire during visit to Viva Air, at Gloucester MRO and
at Middlesex.
Gloucester MRO uses safe-T-cable rather than lockwire, no failure has been experienced with safe-T-cable.
Conclusion:
10.4.4. Adaptor tightened incorrectly

The manufacturing process at supplier remained stable, no change occurred in the adaptor assembly process.
Plus the leaking TLD population includes repaired unit from MRO.
The adaptor assembly was found to be in compliance with requirement on returned unit MG4419 and MG4672,
refer to Appendix.
Conclusion:
10.4.5. Reservoir sub-assemblies not fitted properly

Investigations did not reveal evidence of inappropriate workmanship in installation in service and at Middlesex.
Plus the leaking TLD population also includes repaired unit from MRO. Inappropriate workmanship coming from
different sources make the contributor unlikely.
Conclusion:
10.4.6. Lack of training of workforce

The TLD assembly is carried out by few and adequately trained Middlesex employees. No evidence requiring
further investigation. Also, no batching trends identified that could be associated to a specific Middlesex
employee.
Conclusion:
10.4.7. Conclusion Installation

All the lines of enquiry reveal the installation is not a contributor to the root cause.
10.5. Environment
10.5.1. Abnormal events
Ailus Technical Request & In Service Safety Occurrence Review provided no evidence of abnormal occurrences
on impacted aircrafts.
Conclusion:
10.5.2. Use of non-adapted type of hydraulic fluid

Use of wrong hydraulic fluid cannot induce unscrewing of the Reservoir. Also, leakage occurences occur also with
500B4 which is the fluid used by Sf (see Ref [2]).
Conclusion:
10.5.3. Contamination of hydraulic fluid

Lab investigation reports presented in Ref [2] show the hydraulic fluid properties were checked and did not reveal
any findings.
Conclusion:
10.5.4. High/Abnormal level of vibration due to the brakes

The brakes fitted on all impacted aircraft are Collins brake 2-1759-1 (see below Table 4).
It is common knowledge that vibration could lead to untightening of theaded assemblies. The brakes are the
closest source of vibration to the TLD.
Ailus/SF MFT Team acknowledges there is a correlation between TLD leakages and type of brake installed but
link with a potential root cause is not yet defined. No vibration information has been collected so far. It has been
agreed to discuss the topic with experts to assess the contribution of the brake vibration.
Conclusion:
Potential contributor
Airlin Aircraft Brake Brake

MSN Country Engine Type Tyre supplier Tyre PN
e model supplier PN
MSN0791 320- Goodric BRIDGESTON
2-1759-1 APR06755
8 AZU BRAZIL 251N LEAP-1A26 h E
2-1759-1 APR06755
8 CES CHINA 251N LEAP-1A26 h E
2-1759-1 APR06755
2-1759-1 APR06755
2-1759-1 APR06755
2-1759-1 APR06755
2-1759-1 APR06755
MSN0847 COLOMBI Goodric
2-1759-1 MICHELIN M01103-02
9 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
1 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
5 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
9 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
1 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-03
7 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-04
6 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
6 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
1 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
0 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
4 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
0 VVC A 320-214 CFM56-5B4/3 h
2-1759-1 MICHELIN M01103-02
4 VVC A 320-214 CFM56-5B4/3 h
MSN0829 320- Goodric
2-1759-1 MICHELIN M01103-02
0 HNA CHINA 251N LEAP-1A26 h
2-1759-1 MICHELIN M01103-02
1 HNA CHINA 271N PW1127G-JM h
2-1759-1 MICHELIN M01103-02
2-1759-1 MICHELIN M01103-02
2-1759-1 MICHELIN M01103-02
MSN0849 JAT CHILE 320-232 V2527-A5 Goodric 2-1759-1 Goodrich 467Q02-3
3 h
MSN0853 Goodric
2-1759-1 Goodrich 467Q02-3
9 JAT CHILE 320-232 V2527-A5 h
TABLE 4: LEAKING MSN AND ASSOCIATED CONFIGURATION (ENGINE, BRAKE, TYRE)
10.5.5. Abnormal level of vibration due to the wheels/tyres

No trend has been identified in terms of wheel and tyre manufacturers involved (see above Table 4).
No evidence of unbalanced wheels for affected aircrafts.
Conclusion:
10.5.6. High level of vibration due to runway condition

This contributor is being investigated by Ailus as part of the In Service Arcane Study. However the likelihood of
this factor has been downgraded to unlikely contributor since the leakages are occurring in two different region
(South America and China).
Conclusion:
10.5.7. Damage to the TLD during operation (FOD)

No evidence of impact has been found on returned TLD.
Conclusion:
10.5.8. Taxiing condition

Linked to para 10.5.6 above
Conclusion:
10.5.9. The spring 1-100 rotates and drives the untightening of the Barrel
The sctaches on the Barrel and Cage flange (Figure 39) are oriented towards the anticlockwise direction. These
scratches are on the whole circumference of the flange. There is no mechanical stop to prevent the rotation of the
Cage, therefore compressed Spring would not drive the Barrel untightening but would rotate the Cage.
This finding proves the TLD is reacting to an external force. This finding is considered to be an effect of the root
cause and not the root cause itself.
FIGURE 39: BARREL FLANGE SCRATCHES ORIENTATION

Conclusion:
10.5.10. Braking practices at operator in operation

Viva Air braking practice feedback has been reviewed by Ailus flight operations and confirmed to be compliant
with Ailus protocols.
Conclusion:
10.5.11. TLD duty cycle

The TLD has been originally introduced to dampen vibration generated at Langing Gear level during aircraft
operations defined at this time by Ailus. Since the introduction no comparable cases have been reported. More
in-service events would have been reported if the design intent was not met.
No modification of these operations have been reported by Ailus. Considering the 30 years of in-service
experience and the absence of operation modification it is considered the TLD design remain relevant and would
not be a contributor.
Conclusion:
10.5.12. TLD environment (temperature, humidity)

Temperature in Medellin where Viva Air is based varies between 17 to 23°C during the year while the leakage
events are randomly spread throughout the year. There is no correlation between environment temperature and
leakage events.
Humidity in Medellin varies between 61 to 72% during the year while the leakage events are randomly spread
throughout the year. There is no correlation between environment humidity and leakage events.
FIGURE 40: AVERAGE TEMPERATURES BY MONTH IN MEDELLÍN AT OLAYA HERRERA AIRPORT –

1981-2010, SOURCE: INSTITUTO HIDROLOGIA METEROROLOGIA Y ESTUDIOS AMBIENTALES
FIGURE 41: AVERAGE HUMIDITY BY MONTH IN MEDELLÍN AT OLAYA HERRERA AIRPORT – 1981-2010,
SOURCE: INSTITUTO HIDROLOGIA METEROROLOGIA Y ESTUDIOS AMBIENTALES
8 40
7 35
6 30
Leakage events at VVC
5 25
4 20 Temperature °C
3 15
2 10 Viva Air
Average min Temperature
1 5
Average max
0 0 Temperature
1 2 3 4 5 6 7 8 9 10 11 12
FIGURE 42: LEAKAGE DISTRIBUTION AND TEMPERATURE PER MONTH – VIVA AIR
8 100
90
7
80
6
70
Leakage events at VVC
5
60
4 50 Viva Air
Humidity %
Humitidy (%)
40
3
30
2
20
1
10
0 0
1 2 3 4 5 6 7 8 9 10 11 12
FIGURE 43: LEAKAGE DISTRIBUTION AND HUMIDITY PER MONTH – VIVA AIR
Conclusion:
10.5.13. Altitude of operator’s base

An analysis of VVC maintenance record performed on the TLD was conducted to determine a potental correlation
between the altitude and the leakage events. It shows the majority of the TLD leakage events for MSN8481 were
detected in Lima Perou (airport altitude 34m). While MSN8485 has the majority of the TLD leakage events
detected in Bogota and Medellin (airport altitude 2648m and 2142m).
An extract is provided of the table is presented below and it shows that despite the altitude of both base is
different, several TLD were replaced due to reservoir leakage. The TLD was replaced twice on LH on three times
on RH MLG MSN8481 in Lima airport altitude 34m. While, for MSN8485 TLD was replaced once on LH on three
times on RH MLG in Bogota and Medellin airport (altitude 2648m and 2142m).
There is no obvious correlation between the base altitude and the leakage events. Therefore , The contribution of
this factor can be downgraded to “unlikely contributor”. A deeper study that will be launched and it will include
more variable such as the airport visited. The purpose will be to understand runway and approaches of impacted
aircraft and compare against same data for unaffected aircraft.
Leg Airport
A/C
MSN Station Date ACTION Sid Altitude Country
Reg
e (m)
HK- TLD
N8481 LIM 26/04/2019 LH 34 Perou
5286 replacement
HK- TLD
N8481 LIM 03/06/2019 LH 34 Perou
5286 replacement
HK- TLD
N8481 LIM 11/09/2019 RH 34 Perou
5286 replacement
HK- TLD
N8481 LIM 21/09/2019 RH 34 Perou
5286 replacement
HK- TLD
N8481 LIM 27/10/2019 RH 34 Perou
5286 replacement
TABLE 5: MSN8481 TLD REPLACEMENT AND LOCATION
Airport
A/C Leg
MSN Station Date ACTION Altitude Country
Reg Side
(m)
TLD
HK-
N8485 BOG 18/06/2019 replacemen LH 2648 Colombia
5276
t
TLD
HK-
N8485 MDE 31/07/2019 replacemen RH 2142 Colombia
5276
t
TLD
HK-
N8485 BOG 22/09/2019 replacemen RH 2648 Colombia
5276
t
TLD
HK-
N8485 MDE 18/10/2019 replacemen RH 2142 Colombia
5276
t
TABLE 6: MSN8481 TLD REPLACEMENT AND LOCATION
Conclusion:
10.5.14. Approach and landing conditions

The contribution of this factor has not been assessed yet. Study will be launched by Ailus to:
- Map out airports visited by affected a/c
- Highlight whether these airports have particular risk factors
Conclusion:
Potential contributor
10.5.15. Environment conclusion

Sf and Ailus cross functionnal team has identified potential contributors to the root cause. In fact, an external
factor within the TLD environment is suspected to act on the TLD and cause the loss of preload and untightening
of the reservoir.
The most likely contributor and closest source of vibration are the Collins brake 2-1759-1. In addition they are
fitted on all the a/c experiencing TLD Reservoir leakage. However Ailus did not report any particular vibration
issue coming this brakes from other operator or other variants. Therefore it is suspected an other contributor such
as the approach and landing conditions is also acting with the brakes.
10.6. Maintenance
10.6.1. Incorrect servicing procedure by operators
The AMM procedures are common to worldwide fleet. Sf PSE and FSE witnessed correct application of AMM
procedures by Viva Air maintenance team. It is considered that servicing of the damper (including overfilling)
cannot cause the failure mode of reservoir untightening.
Conclusion:
10.6.2. Damage to the TLD during maintenance

No evidence of damages were found on the returned units. Plus, the correct and methodological application of the
AMM procedure witnessesed onsite ruled out this contributor.
Conclusion:
10.6.3. Maintenance conclusion

In the line of enquiry related to the maintenance, investigations did not reveal any contributor.
11. Airworthiness Assessment
Prior to the subject leakages, assessment and testing of the damper was performed to determine both the volume
of fluid remaining in the cylinder if the reservoir was not fitted (Ref [9]), and the amount of further fluid loss that
would occur during subsequent operation of the damper (Ref.[10]). The testing showed that despite the loss of
fluid from the damper during operation, it would still function as required with the fluid remaining in the housing.
Therefore, the loss of fluid from the reservoir is not considered as an airworthiness risk.
In addition, Ailus MMEL 32-11-01 allows operation for 7 FC with deflated Torque Link Damper.
12. Conclusions & Recommendations
The MRO investigations performed on the returned units showed the Torque Link Damper passed the functional
test and no leakage was identified during the test. The TLD passed the required CMM tests after the Reservoir
has been tightened back considering the TLD has been serviced by Viva Air before they were sent to Sf for
investigation.
The lab investigation performed on the TLD did not reveal any significant findings incriminating the Torque Link
Damper itself.
Nevertheless, the failure scenario was identified thanks to the evidence found on the parts:
1. Loss of the torque preload caused by unkown external factor(s)
2. Reservoir untightening driven by external factors (outside of the TLD)
3. Lockwire breaking by ductile overload
4. Reservoir continues to untighten
5. Hydraulic leakage when seal squeeze is lost
The outcomes of the Lab investigation has proven an external force is acting on the Torque Link Damper and
drives the untightening of the reservoir.
The TLD Reservoir leakage occurred only on A320 aircrafts.

The TLD Reservoir leakage is reported by 5 airlines on 27 A320 A/C out of 3000 which represent less than 1% of
A320 A/C population.
The population of leaking TLDs so far is 63 out of 6000 in service which represent 1% of the TLD population.
In addition the current TLD design has been in-service for more than 30 years without any Reservoir leakage
event reported until recently.
The manufacturing of the TLD remained stable and no non conformity was found on the TLD
The review of the Ishikawa headlines and associated elements, the evidence and delivered actions arising from
the investigation reveals:
- All the elements associated with the Manufacturing, Material, Installation and Maintenance headers of the
Ishikawa diagram were assessed and it is concluded there are no elements contributing to the root cause.
- All the elements associated with the Design header of the Ishikawa diagram were assessed and it was
concluded there are no elements contributing to the root cause. However there is an outstanding
improvement action associated to the torque value not being specified on the Torque Link Damper
assembly drawing that is recommended to be closed as part of the continued investigation. The purpose
of this action is to ensure the stability of the assembly process.
- All the elements associated with the Environmental header of the Ishikawa diagram were assessed and it
was concluded that the root cause is contained in the environmental external factors acting upon the
Torque Link Damper assemblies installed on the Main Landing Gears. The suspected root cause
associated to the TLD leakages would be the exposition to unusual vibrations.
The origin of the vibrations are suspected to be related to two major Environmental contributor :
o The evidence of Collins brake configuration correlation with each aircraft TLD failure
o Unusual landing, take-off and taxiing operation impacting the Torque Link Damper assemblies.

Shopping List

Uploaded by

Copyright:

Available Formats

Shopping List

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Shopping List

Uploaded by

Copyright:

Available Formats

AGE

These external parameters are suspected to be related to :

[1] MAF1-0005-2167 : Introduction of EV MLG’s to meet the NEO loading requirements

4. Identification and description

FIGURE 3: TLD FILLED AND PRESSURIZED

BARREL (from RESERVOIR

FIGURE 5: ADAPTOR ASSEMBLED INTO HOUSING

Operators have reported :

Fluid leaking from the

FIGURE 9: REPARTITION OF LEAKAGE CASES AMONG OPERATORS

10.4 10.6 10.6

Age at the first event (month)

Glos MSN07811 TIANJIN AIRLINES HNA

2500 CSN 7th TLD

FIGURE 11: VIVA AIR LEAKAGE CASES (LH/RH SPLIT)

Cumulative CSN per TLD before leakage - China Eastern

CSN 5th leak

Cumulative CSN per TLD before leakage - HNA

FIGURE 13: HNA LEAKAGE CASES

Leakage per month in 2019 and 2020

9.1. Investigation of leaking units

FIGURE 15: MG4419 IN SERVICE HISTORY

VVC report including video and photo evidence confirmed:

9.1.2. MRO troubleshooting report

Part title Part number CMM32-12-25

FIGURE 17: INSPECTED TORQUE LINK DAMPER COMPONENTS

The main conclusions of the lab investigation are the following:

9.1.3.1. Detailed visual inspection

FULL Line FULL Line

FIGURE 20: BARREL FLANGE CONDITIONS- CIRCULAR SCRATCH PATTERN

FIGURE 25: PERMANENT DEFORMATION OF BARREL LOCKWIRE HOLE (LEFT)

FIGURE 26: PERMANENT DEFORMATION OF HOUSING LOCKWIRE HOLE (LEFT)

FIGURE 27: HOUSING/RESERVOIR INTERFACE SEAL CONDITION

9.1.3.3. Hydraulic fluid

9.1.3.5. Material and Hardness

9.2. Laboratory Investigations on lockwire

FIGURE 28: EXTRACT FROM REF [4] – LOCKWIRE FRACTURE FACE

9.3. Laboratory Investigations of broken Reservoir

FIGURE 30: MG4170 BARREL DAMAGE

FIGURE 32: SEM PHOTOGRAPHS

9.5. Conclusion of the investigation:

The outcome of this analysis is presented in the Ishikawa diagram on :

Safran Landing Systems Tech Event Reference: TE-VE-00197158

A. Material B. Manufacturing C. Design

A1– Shrinkage or swelling of the seal (1-

B8 – Machining of external or internal thread B16 – Manufacturing changes

D. Installation/Assembly E. Operation/Environment F. Maintenance

As a result, the seals are behaving as expected.

FIGURE 36: RESERVOIR/HOUSING INTERFACE SEAL (1-130)

10.1.2. Non-conforming lockwire material or hardness

FIGURE 37: LOCKWIRE USED BY VVC Ø0.8MM, 302/304 STAINLESS STEEL

10.1.3. Use of incorrect hydraulic fluid or use of fluid mix

10.1.4. Non-conformity of seal material or hardness

10.1.5. Non-conformity of Adaptor material or hardness

10.1.6. Non-conformity of Barrel material or hardness

10.1.7. Non-conformity of Housing material or hardness

10.1.8. Material Conclusion

Material lines of enquiry are not a contributor to the root cause.

10.2.2. Seal grove not manufactured to drawing

10.2.3. Change of lockwire specification or supplier