Content uploaded by Peter Spanoudakis
Author content
All content in this area was uploaded by Peter Spanoudakis on Oct 28, 2019
Content may be subject to copyright.
Page 1
ESMATS18_MTG-CCM_PAPER_submission_Final_with footer.docx 18th ESMATS 18-20 Sept. 2019, Munich, Germany
CORNER CUBE MECHANISM (CCM) FOR MTG IRS: QUALIFICATION AND
ACCEPTANCE TEST CAMPAIGN RESULTS AND LESSONS LEARNED
Peter Spanoudakis, Philippe Schwab, Lionel Kiener, Gérald Perruchoud, Hervé Saudan,
Mathias Gumy, Yves-Julien Regamey
Centre Suisse d'Electronique et de Microtechnique (CSEM), Jaquet-Droz 1, CH-2002 Neuchâtel, Switzerland
Email: peter.spanoudakis@csem.ch; +41 32 720 5443
ABSTRACT
The Corner Cube Mechanism (CCM) of the Infra-Red
Sounder (IRS) for the Meteosat Third Generation
(MTG) satellites completed an extensive qualification
test campaign demonstrating the high-precision
mechanism met stringent requirements for operation in
the harsh environment of space at geostationary orbit.
The Qualification Test Program was performed both at
component/sub-system level and at mechanism level.
The FM (Flight Model) Acceptance test campaign was
hindered by two major setbacks. The first was
encountered just prior to the random vibration tests on
the PFM (Proto-Flight Model), where it was discovered
that a number of critical bolts had loosened despite the
fact that they were locked by adhesive. The second with
the discovery of the FM2 Optical Switch (OS) ruler
broken inside the mechanism while the final sequence
of the bolt replacement procedure.
The PFM and FM2 performance tests confirmed the
previous EQM (Engineering Qualification Model) test
results. The maximum lateral deviation for a stroke of
10mm was measured <0.4µm compared to the 2µm
specification.
1 INTRODUCTION
Results on the subsequent qualification and acceptance
test campaign at mechanism level are reported here
which were performed on the EQM and the PFM. The
results of the component and sub-system level
qualifications were published during the 16th ESMATS
in Sept. 2015 [R1].
The main performance requirements for the CCM are:
A linear trajectory generation along the delay line
translation axis over a functional stroke of ±5 mm
(±9 mm calibration on ground)
Maximum lateral deviation of the corner-cube
apex of less than 1 µm from a true straight line
Constant linear motion of 1 mm/s and speed
stability requirements during dwell time
(instantaneous speed error <0.25mm/s and RMS
speed value <0.06mm/s)
Highly limited dynamic exported forces from the
mechanism to the optical bench in a micro-
vibration environment
The critical qualification tests performed on the EQM
were the environmental random vibration and shock
tests followed by the crucial micro-vibration tests. The
remaining qualifications tests for thermal balance and
thermal cycling were performed on the PFM.
Figure 1. PFM and FM2 with dummy CC mirrors
Both the EQM and PFM survived the vibration and
shock tests but these events were not the only technical
and programmatic challenges that were faced and
resolved. With the hurdle of the qualification tests
behind the team, the acceptance level tests of the FMs
were considered as a repetition and less critical.
The acceptance test campaign of the two flight model
CCMs started in mid-2017 following the successful
qualification test campaign on the EQM. The FM
acceptance test campaign was comprised of a series of
critical tests: performance tests, random vibration tests,
thermal vacuum cycling (TVC), performance tests after
environmental tests and micro-vibration tests. The test
campaign was covered by successes and setbacks but, at
the end, solutions were found and implemented.
The TVC part of the Proto-Flight Model (PFM) tests
completed the qualification of the CCM since these tests
were not performed on the EQM. A delta-QR
(Qualification Review) was held successfully validating
the overall test program for final delivery of the PFM to
Thales Alenia Space (TAS-F) for integration of the
CCM at interferometer level.
2 VOICE-COIL MAGNETS
At component level, the voice-coil actuator NdFeB
magnets (VACODYM633AP) went through a
qualification programme with their Aluminum IVD
coating as reported in [R1] to address a major issue of
Nickel coating delamination. However, this new
Page 2
ESMATS18_MTG-CCM_PAPER_submission_Final_with footer.docx 18th ESMATS 18-20 Sept. 2019, Munich, Germany
combination was not selected by the MTG team
following long duration thermal humidity tests where
some samples showed divergent results with respect to
coating thickness discontinuities and cracks. EDX
(Energy Dispersive X-Ray) analysis revealed the
presence of oxygen on the top surface of one magnet
and underneath the IVD Aluminium coating which
could lead to oxidation or corrosion conditions of the
magnet. In view of the results, it was concluded that the
IVD magnet coating process needed further
improvement and could not be considered as fully
qualified [R1].
A new qualification campaign ensued with the change
to Sm2Co17 magnets (Recoma32S from Arnold
Magnetic) without a coating. Thanks to the recent
SmCo alloy performance improvements, the impact to
the CCM was negligible and the motorisation margins
were maintained.
Figure 2. FM Voice-coil actuator stator from Cedrat
3 EQM QUALIFICATION TEST CAMPAIGN
The Qualification Test Program was performed both at
component/sub-system level and at mechanism level.
One of the critical components at sub-system level is the
voice-coil actuator supplied by Cedrat Technologies (F).
A lengthy qualification program was undertaken with
the supplier to qualify the new magnets and coatings to
meet specific MTG requirements for long term storage
conditions of 15 years. Typical tests included thermal
vacuum cycling (100 cycles, -40°/+80°C), thermal
humidity tests (95%RH, 1bar, 45°C, 240h) and epoxy
resin adherence to validate the robustness of the various
processes used.
The Optical Switch was also qualified at component
level by Codechamp (COD) where the process consisted
of performance tests, random vibration (16grms), shock
tests (350g), thermal cycling (1 cycle +45/-25°C with
8 cycles, -15°/+35°C) and performance tests.
3.1 Performance tests
The EQM was identical to the Flight Model version and
was used to qualify the CCM to MTG specifications.
Performance level tests with EQM were performed such
as:
Trajectory generation and motion control
Lateral deviation of the corner cube from a true
straight line
Dynamic exported forces
The objective of the test campaign was to validate the
design, the manufacturing and assembly processes of a
mechanism that is as representative as possible to the
final flight version.
Figure 3. EQM mounted on performance test bench
with interferometer in ISO5 (Class 100) cleanroom
conditions.
The critical performance parameters measured during
these tests were the mobile mirror lateral shifts and the
speed stability.
The maximum short term (15 min.) lateral deviation
parabolic shift for the on-ground calibration stroke of
18mm (±9mm) was extremely low and measured at
±2nm in Z and ±4nm in Y directions compared to the
±0.5µm specification for a functional stroke of ±5mm.
3.2 Mechanical vibration tests
Mechanical vibration tests simulate the extreme noise
and vibration environment generated during the launch
phase. The delicate mechanism is in a launch locked
configuration in order to ensure that it will survive the
vibration loads. The EQM survived the random profile
vibration tests and shock tests in all three directions.
Following the environmental tests, the performance
tests were repeated and a close inspection of the launch
locking device critical surfaces were made.
Figure 4. EQM instrumented on vibration shaker table
Page 3
ESMATS18_MTG-CCM_PAPER_submission_Final_with footer.docx 18th ESMATS 18-20 Sept. 2019, Munich, Germany
3.4 LLD contact pads
One of the inspection areas following the vibration tests
was to verify the state of the LLD contact pads. The
contact surfaces between the WC (Tungsten carbide)
coated Titanium pads and the stiffening columns
maintain the mobile stage in its neutral position without
sliding.
Figure 5. LLD in locked configuration
12
34
xx
xx
xx
X
Z
Y
Figure 6. LLD Pads localization on CCM assembly
Figure 7. LLD WC pad #3 (+X/+Z)
before/after vibration tests
Figure 8. LLD Stiffening column #3 (+X/+Z)
before/after vibration tests
Observing the clamping surfaces of both WC pads and
stiffening columns, it can be deduced that no mobile
stage sliding was induced by the shocks or vibrations
tests. Considering the mating materials, the pitting
appearing after environmental tests is as expected and
results from the double flight locking of the LLD.
3.4 EQM Micro-vibration tests
The EQM micro-vibration tests were performed on a
dedicated micro-vibration test bench developed by
TAS-F to measure the exported forces of the CCM
while in operation and to inject a simulated spacecraft
disturbance noise profile. The results with the EQM
showed non-conformances to the specifications which
required significant exchanges and clarifications
between CSEM-TAS-OHB-ESA to correctly understand
and interpret the impact in performance.
While the mechanism is displacing the corner cube at a
speed of 1mm/sec, the exported forces to the instrument
during the reversal stroke were measured at 7.2mN.
Some of the major results discovered were high
amplification of CCM resonance modes from induced
perturbations from the CCS (cryo-cooler system)
harmonics. With the injected micro-vibration
disturbance profile, two speed stability parameters were
measured:
• Absolute speed error during dwell time measured at
0.73mm/sec (spec: 0.25mm/sec)
• Standard deviation of speed error during dwell time
measured at 0.29mm/sec (spec: 0.06mm/sec)
Even though these values were out of specification, the
results were expected since they are directly
proportional to the injected disturbance levels. The
injected disturbance is a sum of various satellite sub-
system contributions mainly cryo-coolers and reaction
wheels which at the time of the tests were being
reviewed at satellite level to determine budgets and
margins.
The most critical frequencies affecting the CCM
performance were 173Hz (Y-direction, lateral shift
impact) excited by the 3rd CCS harmonic, 230Hz (X-
direction, lateral shift impact) excited by the 4th CCS
harmonic and 330 Hz (X-direction Intermediate Stage,
speed stability impact). The solution investigated to
minimise the impact of the CCM resonance modes on
performance was the implementation of mobile mass
tuning to shift critical frequencies outside the
disturbance bands. Analyses were performed to
determine the impact of the additional masses based on
EQM measurements and verified on the PFM.
A compromise had to be found to avoid excessive mass
on the mobile stage which would have an effect of
additional stresses during vibration tests and a sufficient
frequency shift away from the CCS harmonics. The
optimum configuration that was tested on the PFM was
an intermediate stage mass of +10g and a mobile mass
of +50g. The addition of 10g on the intermediate stage
reduced the 320Hz mode to 313Hz shifting it away from
the CCS disturbance zone identified at 330Hz.
Page 4
ESMATS18_MTG-CCM_PAPER_submission_Final_with footer.docx 18th ESMATS 18-20 Sept. 2019, Munich, Germany
Figure 9. Location of balancing mass added to
intermediate stage
Figure 10. Location of mobile stage balancing mass
240 260 280 300 320 340 360
-145
-140
-135
-130
-125
-120
Hz: 3 16.4
dB: -123.3
Magnitude (dB)
Hz: 319.1
dB: -122.3
Hz: 313.5
dB: -122.3
Hz: 310.5
dB: -123.4
Hz: 3 08. 1
dB: -123.4
Hz: 253.4
dB: -120.9
Hz: 254.4
dB: -123.6
Hz: 24 6.8
dB: -126.7 Hz: 2 47. 8
dB: -128.1
TF from VC force N to CC position m
Freq uency (Hz)
0g
5g
10g
15g
20g
Figure 11. CCM PFM transfer function as a function of
mass tuning on intermediate stage
Figure 12. CCM PFM transfer function as a function of
mass tuning on mobile stage
100 150 200 250 300 350 400 450 500 550 600
-100
-80
-60
-40
X: 229
Y: -36.28
TF from acc in X to spd, perturbation X
Frequency Hz
Magnitude dB
X: 228.5
Y: -24.82
PFM +30g (max in -5 to 5 mm CC pos)
PFM +50g (max in -5 to 5 mm CC pos)
Pert.
100 150 200 250 300 350 400 450 500 550 600
0
0.02
0.04
0.06
0.08
0.1
Perturbation m/s
2
Oveloading of BM
tends to increase
response in the area
Figure 13. Example of CCM frequency shift to the left
and outside the CCS disturbance zone
4 CRITICAL BOLTS LOOSENED
The FM test campaign was hindered from a number of
critical anomalies both at sub-system and mechanism
level. A major setback was encountered just prior to the
start of the random vibration tests on the PFM. A
planned inspection of the CCM was carried out on the
shaker table and it was discovered that a number of
critical bolts had loosened despite the fact that they
were locked by adhesive. For the PFM on the CSL
vibration shaker, 9 bolts were identified and on the FM2
still at CSEM, 10 bolts were found loosened.
Figure 14. One of the locations where bolts were
loosened
Since some of the identified bolts were previously
removed to install the additional mobile masses for
frequency tuning, it was believed that heating the bolt
heads to soften the EC2216 for removal was the culprit.
A full inspection found that the majority of the loosened
bolts had been exposed directly or in proximity to
temperatures around 80 to 100°C by the hot air gun.
Figure 15. Typical assembly configuration of guide
membranes (left) and representative test setup for
investigation (right)
Page 5
ESMATS18_MTG-CCM_PAPER_submission_Final_with footer.docx 18th ESMATS 18-20 Sept. 2019, Munich, Germany
A short list of possible root causes and contributors
(out of 23 identified) are listed below:
Lack of adhesion of glue on screw heads
Glue deterioration induced by local heating during
glue removal
Thermal expansion coefficient mismatching
Wrong tightening torque
Screws touching the bottom of the tapped holes
APS Nuflon-N friction coefficient too low
APS Nuflon-N coating w.r.t screws reversibility
influenced by temperature
An investigation was carried out to determine the source
of this major non-conformity. Submitting test samples
(Fig. 15 & 16) to 80°C heating in a climatic chamber
produced “spontaneous” screw untightening with screw
head rotation (see Fig. 17) of 40-50°. The untightening
torque was verified and found to be 0Nm to 40% of the
initial tightening torque. The EC2216 epoxy had little to
no effect in blocking the bolts.
Figure 16. R-Sat bolts with antifriction coating
Figure 17. Screw rotation clearly visible from
subsequent tests with test setup
The six-month investigation concluded that the APS
Nuflon-N antifriction coating (varnish charged with
PTFE) applied on the bolt threads (and under the head)
is affected by a temperature rise (local heating or
thermal cycling). The coating thickness was between 4-
8 µm providing a tested coefficient of friction of 0.05 to
0.09. The friction coefficient of the Nuflon-N coating
dropped below 0.04 due to a temperature increase
(during bolt exchange procedure) creating a reversibility
condition which allowed the bolst to loosen.
The implementation of a corrective action consisted of
the replacement of all mechanism bolts using the IASI
heritage coating and without disassembly of the CCM.
All the R-SAT screws had their Nuflon coating removed
as part of the process and recoated with the IASI
sprayed PTFE coating. The TEFLISS-2 PTFE spray
coating (Fig. 18) was tested to determine optimum
parameters for coating thickness and tightening torque.
The final friction coefficient range for the coating was
µ=0.17 to 0.24. The tightening torque for a M5 screw
was set to 9.6Nm to provide a value of 70% of yield
during tightening.
Screw + washe r
Force
measurement
sensor
Guiding
membrane
Exchangeable
insert with Heli-
Coil
Figure 18. Test setup to measure preload through
tightening torque and to determine friction coefficient
Figure 19. TEFLISS-2 PTFE coated bolt (left) and wear
after 1st tightening
Various tests were performed to validate the process and
measure the screw untightening torque. Samples were
tested at room temperature (RT) for comparison with
subsequent tests. Samples were heated to 80°C with and
without the EC2216 epoxy. Results showed that the
heating of samples increased the untightening torque
from 5 to 20% when compared to the RT data.
The samples were finally tested in a thermal chamber
with a cycling profile (-15 to +45°C) similar to the TVC
tests and again followed by measurements of the screw
untightening torque (Fig. 19).
All the bolts (110 per CCM) were replaced sequentially
without compromising the alignment and assembly
tolerance of the mechanism (Fig. 20). Thanks to the
design (and by chance), all bolts were accessible
without the need of disassembling critical components.
Figure 20. Part of the bolt replacement GSE during
PFM refurbishment
Page 6
ESMATS18_MTG-CCM_PAPER_submission_Final_with footer.docx 18th ESMATS 18-20 Sept. 2019, Munich, Germany
5 FM2 OS ruler broken inside the mechanism
Another setback was encountered during the acceptance
test campaign with the discovery of the FM2 Optical
Switch (OS) ruler broken inside the mechanism while
the final sequence of the bolt replacement procedure
was being carried out.
Figure 21. FM2 OS broken ruler and its support
Figure 22. FM2 OS broken ruler during inspection
1mm
gap
Inspection windows
CC displacement
Figure 23. Location of ruler and inspection windows
A Non-Conformance Review Board (NRB) was held
and following the inspection and expertise of the glass
ruler by Schott, a root cause analysis indicated a failure
due to a flaw on the glass surface and possible contact
with an endoscope during an inspection step.
Figure 24. Details of crack origin and propagation
Observations were provided by Schott following their
expertise of the broken ruler indicating that a material
flaw was present. Lots of small chipping were observed
along the length of the glass edges. Flaws such as these
on the glass surface most often originate during the
grinding process to remove the sharp edges. The pit
marks along the chamfer are typical damages and
problematic since they are potential fracture origins.
The Schott interpretation of the failure was:
The breakage was caused by the combination of a
flaw and an applied tensile stress
The flaw is located on the edge of the ground glass
surface facing the metal frame
It is highly probable that the flaw comes from the
grinding process itself and is not introduced by
additional mechanical contact after grinding
A flaw on the surface of a brittle material
represents a weak spot & diminishes its strength
and facilitates breakage
The shape of the broken glass part indicates an
introduced tensile stress by bending
The course of failure is therefore:
The glass part was bent away from the metal
frame, thus introducing a tensile stress into the
grinded glass edge which leads in combination
with the present flaws to the fracture of the
specimen
The general conclusion and most plausible failure
scenario considered was a combination of individual
parameters that led to the optical switch glass ruler to
break. If a surface defect existed from the start
(following the grinding process), the Codechamp
component level vibration tests (two vibration tests)
could have propagated a fracture along the glued
surface. This fracture would not have been visible either
due to the glue or small crack size. Subsequent
inspections by CSEM with an endoscope through the
inspection window could have made contact with the
ruler and further propagated the crack. The various
manipulations during the bolt exchange procedure could
have led to total failure on a ruler that was already
fragile.
Following the investigation, Codechamp’s supplier
implemented corrective actions by polishing instead of
grinding the glass edges. The glass ruler was replaced
and the OS hardware retested (vibration and TVC) at
component level before being re-integrated in the CCM.
Additional inspections were made on the PFM to verify
that its ruler was intact and in good condition.
6 FM PERFORMANCE TESTS
Following the various corrective actions, the PFM and
FM2 acceptance tests were performed during the first
half of 2018. The PFM and FM2 performance tests
confirmed the previous EQM test results. The maximum
Page 7
ESMATS18_MTG-CCM_PAPER_submission_Final_with footer.docx 18th ESMATS 18-20 Sept. 2019, Munich, Germany
lateral deviations for both long term (e.g. LLD release)
and short term for a functional stroke of 10mm (±5mm)
was measured after the environmental vibration tests
and given in the table below.
Table 1. FM Lateral shift test results
Lateral deviations (PFM) Spec Y-dir Z-dir
Constant shift (Long term) ±20µm 0.6µm 8µm
Linear shift (Long term) ±2µm 0.1µm 0.4µm
Parabolic shift (Short term) ±0.5µm 3nm 2nm
Lateral deviations (FM2) Spec Y-dir Z-dir
Constant shift (Long term) ±20µm 1.3µm 9µm
Linear shift (Long term) ±2µm 0µm 0.4µm
Parabolic shift (Short term) ±0.5µm 3nm 4nm
Following the EQM micro-vibration tests,
improvements were identified to shift mechanism
frequencies away from the input disturbances generated
by the cryo-coolers which affected speed stability and
lateral shift. Local mass tuning implemented on the
mobile stages of the CCM improved performance in
these areas.
The test setup at TAS-F, Cannes premises was slightly
improved with respect to the EQM tests by placing the
Z-shaker directly underneath the platform and taking
into account the accelerometer mass in the balancing
mass used for the mobile stage.
Figure 25. PFM during micro-vibration tests at TAS-F
With the injected micro-vibration disturbance profile,
the two main speed stability parameters were measured:
The absolute value of speed error during dwell
measured at 0.54mm/sec (EQM 0.73 mm/sec)
(spec: 0.25mm/sec)
The standard deviation of speed error during dwell
measured at 0.12mm/sec (EQM 0.29 mm/sec)
(spec: 0.06mm/sec)
Both models were delivered to TAS-F, Cannes at the
end of 2018 for integration in the Interferometer
Assembly and further testing.
7 LESSONS LEARNED
Some of the main lessons learned over the course of this
project are provided below from the project manager’s
point of view:
The collaboration and commitment of the suppliers
to the CCM project allowed for fruitful discussions
and identification of solutions when problems
arose (both minor and major).
Having two people following a subcontractor is
important for redundancy.
Do not underestimate a recurring development if
manufacturers and materials need to be changed
The support from company management to
organize resources attributed to the project during
critical phases was crucial.
The support from the client to propose and back
solutions without finger pointing keeps a positive
spirit during the investigation process when
unplanned events arise.
Just because items have been identified in a risk
register or a FMECA (Failure Modes, Effects and
Criticality Analysis) table does not mean all
potential problems have been identified. The loose
bolts or broken ruler could never have been
envisioned in any stage of the project.
The flight spare kit that was manufactured in
parallel to the FM hardware was important when
the bolt replacement process started. All critical
hardware was available in case anything was
damaged during the refurbishment.
Progress Meetings with the client and participation
from suppliers and sub-contractors keeps everyone
in the loop and identifies responsibilities.
8 CONCLUSION
The CCM project started with the design and
qualification of new magnets and coatings for the voice-
coil motor following the discovery of Nickel
delamination on the IASI (Infrared Atmospheric
Sounding Interferometer) CCM due to NdFeB magnet
corrosion. An extensive development and test campaign
was carried out to finally implement a solution using
Sm2Co17 magnets with no coating.
Following the EQM test campaign, the CCM
development was on track and with significant margin
for delivery of the flight hardware. A major setback was
encountered with the FMs which consisted of critical
bolts found untightened prior to the start of vibration
tests. The investigation concluded with the
implementation of a heritage PTFE sprayed coating
used on the IASI CCMs. The various problems
encountered were successfully resolved with the open
discussions during the non-conformance review boards
among participants from CSEM, TAS-F, OHB and
Page 8
ESMATS18_MTG-CCM_PAPER_submission_Final_with footer.docx 18th ESMATS 18-20 Sept. 2019, Munich, Germany
ESA. Frequent discussions with the customer and sub-
contractors at technical level was maintained throughout
the duration of the project.
The PFM and FM2 performance tests confirmed the
previous EQM test results. The maximum lateral
deviation for a stroke of 10mm was measured <0.4µm
compared to the 2µm specification. Both flight models
have been delivered to TAS and are one of first flight
hardware items delivered in the frame of the MTG IRS
programme.
Figure 26. CCM FM2
ACKNOWLEDGMENTS
The authors would like to thank:
Thales Alenia Space (Cannes, F) for their continued
support and co-operation over the course of this
project (G. Luciano and his colleagues).
Subcontractors and partners for their contribution &
involvement throughout the development effort:
o Cedrat Technologies SA for the development of
the voice coil motor (A. Guignabert)
o Ruag Space System Nyon (CH) for manufacture
& assembly of mechanisms (S. Liberatoscioli)
o CODECHAMP for the optical switches
(P. Vuillemard and her colleagues)
o ARCOFIL SA (St-Imier, CH) for their
manufacturing (WEDM) capabilities on the
critical flexure components notably the driving
lever and membranes
o Almatech (CH) for the FEM simulations
The European Space Agency (ESTEC).
CSEM thanks them for their support.
REFERENCES
1. Spanoudakis, P., Schwab, P., L. Kiener, et. al.
(2015). Development challenges of utilizing a
corner cube mechanism design with successful
IASI flight heritage for the infrared sounder (IRS)
on MTG. Proceedings of the 16th European Space
Mechanisms & Tribology Symposium, Bilbao,
Spain, 23-25 Sept. 2015 (ESA SP-737)
2. Falkner, M., Blythe, P., et. al. (2019). Meteosat
Third Generation (MTG) Payload Mechanism
Developments: Lessons Learnt and Observations
from ESA Perspective, Proceedings of the 19th
European Space Mechanisms & Tribology
Symposium, Munich, Germany, 18-20 Sept. 2019
Key words: Mechanism, linear scan, high-precision,
flexures, interferometer, earth observation,
