Akatsuki Venus Climate Orbiter Options

[nprev]

True! Mea culpa.

But still: The ship was designed to do it, and presumably Akatsuki was not.

[djellison]

Magellan wasn't designed for it either

[pandaneko]

(page 35)

5. Summary of this 3rd report

1. We have looked into the causes of CV-F closure and gained following conclusions.


• We need to take into consideration the transmission of the oxidiser across the valve seal
• CV-F can be closed by salt formation as a result of chemical reaction of fuel and oxidiser

2. By looking into the effects OME suffered from CV-F closure we have obtained following conclusions for re-imsertion.

• Thruster nozzle is very likely to have been damaged during VOI-1
• We need to conduct ground tests and experiments to see if re-ignition of the damaged thruster nozzled engine is possible and test firing in orbit, and depending on the outcome we will have to carry out preparations for re-insertion.

３．We must come up with a viable operational plan for orbit changes.

• OME test firings ---> early September 2011
• Orbit change at nearest approach to sun, depending on the outcome of test firings ----> November 2011

(end of page 35)

[pandaneko]

(I am feeling a little uneasy because I remember saying at the start that this report was a 48 page report. I do not remember now where this number came from, but I could not have said it without a firm basis...

Anyway, this JAXA 3rd report came to an abrupt end (to me) at page 35 and what is left is a set of appendices and the last page is 61, yes, 61, I just had a look. What follows is the contents list. I will carry on translating as these seem more detailed and we should be able to finish the rest well before OME test firings.)

Appendices

A. FTA results presented at 2nd Investigation Meeting

A.1 Looking into the causes of Akatsuki mulfunction
A.2 Akatsuki FTA

B.Looking into the causes of CVF-closure

B.1 Discussing cause candidates relating to CV-F design and manufacturing
B.2 Discussing cause candidates relating to dynamic behaviour of the valve
B.3 Discussing cause candidates relating to excessive insertion of the valve
B.4 Discussing cause candidates relating to wear and tear
B.5 Evaluation of the speed of propellant flow
B.6 Estimation of the propellant (fuel) flow amount inside Akatsuki's propulsive sysytem
B.7 Investigation of the past mulfunctions relating to oxidiser flow
B.8 Examples of gas supply piping system with 2 liquid propulsion system on board for prolonged flight in space

C. Effects suffered by OME

C.1 Current understanding of the status of OME through analyses
C.2 Breakage surface observation of the damaged burner
C.3 History of acceleration and angular velocity during the latter half of VOI-1
C.4 Propulsion characteristics of the damaged burner
C.5 Trying to find ways to reduce re-ignition impacts
C.6 Discussing the possibility of continued firing of OME
C.7 Discussing ways of jettisoning oxidiser

(end of appendices page)

(My translation will start from B. Here, suffice to say that the very end of A talks about doing another FTA for P3 pressure drop as an additional analysis and finding that its cause was also judged to be CV-F closure. P)

[pandaneko]

(page 43, start of appendix section (I seem to have bad keyboard contact...)

B.1 Cause candidates for CV-F relating to design and manufacturing

We have had talks with manufacturers regarding design, manufacturing and materials used including factory visits and obtained following information



E‐1) Material incompatibility used for the seals

We checked manufacturing and inspection records and confirmed that the seals had been manufactured with propellant compatible materials and according to design specifications

E‐4) Sliding area mulfanction due to incompatible materials (&, P)
E‐10) Bad manufacturing of the sliding parts

From the records (manufacturing and inspection) such as material verification records and surface treatment records we confirmed propellant compatibility of the materials used

E‐6) Bad design/manufacturing of the clearance of the sliding portion (&, P)
E‐7) Bad alignment of the valve and its moving section

Records inspection confirmed that valves had been manufactured according to specifications


E‐5) Clearance changes due to inapproapriate fixing method

We investigated the valve body deformation due to inapproapriate torque and this confirmed that the deformation was sufficiently small.

From all above we confirmed CV-F design/manufacturing information and are satisfied that above candidates are sufficiently innocent.

(end of page 43)

[pandaneko]

(page 44)

B.2 Cause candidates relating to the dynamic behaviour of the valve

１．Possibility of the valve sliding vibrationally due to the resonance of the regulator and the piping system

E‐12) Sliding products gritting due to an unexpected number of sliding motions

We conducted dynamic characteristics simulation and tests assuming fuel tank pressurisation. As a result, we confirmed that speedy (?, quickly reached?) vibrations between regulators themselves, between regulator and CV-F, and vibration of components did not occur.

２．Possibility that CV-F itself reached resonance over a certain working range, with mechanical parts breaking, falling away, and leading to gritting

E‐12) Sliding products gritting due to an unexpected number of sliding motions
E‐14) Mechanical parts breaking, dropping (or falling) off, and gritting due to an unexpected number of valve workings
E‐15) Falling off of coil springs

We conducted experiments and analyses where we changed the pressure both up and down streams of CV-F, covering the whole range of pressure status obtained from the flight history. As a result, we confirmed that CV-F closure, chatterings, frettings and other unexpected vibrations were not detected.

３．Possibility that the resonance due to transient response to the rising tank pressure in orbit might have affected mechanical parts

E‐14) Mechanical parts breaking, falling off, and gritting due to an unexpected number of valve workings
E‐15) Falling off of coil springs

Based on the telemetry data of the rapid tank pressure rise in orbit we conducted experiments controling the upstream pressure of CV-F, and for the down stream we conducted experiments using same piping diameter, same length, and same flow (?) volume.

As a result, we confirmed that the spread of cracking pressure and reseat pressure was contained within 0.002MPa and not leading to the closure.

(There is a simple picture of the valve test)

Top left caption on the picture reads "Valve working tests"

An arrow points down to the picture from "CV-F (spare CV-F)"
An arrow points up to the picture from "Acceleration sensor: Data obtained : acceleration during valve working (vibration monitor)and valve capability trend after load imprinting"

From these analyses and tests of the dynamic behaviour of the valve we are satisfied that above candidates are sufficiently innocent.

(end of page 44)

[pandaneko]

(page 45)

B.3 Cause candidates relating to excessive valve insertion

E‐3) Excessive valve insertion due to long term reverse pressure imprint

We conducted a 21 day test simulating the 35 day long in-orbit reverse pressure state by applying the pressure 1.5 times that in orbit

The cracking pressure (that pressure which changes CV-F status from close to open) and the re-seat pressure (that pressure which changes CV-F status from open to close) before and after the test are shown in the following graph. (Note: see section 1.2 for CV-F valving actions)

From the test we confirmed that there was no excessive valve insertion (valve moving beyond the regular close position into further close state with too much seal pressing) and we are satisfied that above candidate is sufficiently innocent.

(end of the main text of this page 45)

(the graph is simple with the vertical showing the % change after AT. Horizontal is the quantum time line and there are 9 of them from left to right. They are;)

1. AT (whatever this is, P)
2. Pre-assembly inspection
3. before valve action test
4. after valve action test
5. before dynamic characteristics simulation test

6. after 1st in-orbit tank pressure rise simulation
7. after 2nd in-orbit tank pressure rise simulation
8. after 3rd in-orbit tank pressure rise simulation
9. after 21 days of reverse pressure imprint

(end of page 45)

[pandaneko]

(page 46)

B.4 Cause candidates relating to wear and tear

Most of the ground tests for checking the health of the valve were conducted in helium gas environment. However, in real situation the environment is the mixture of helium and propellant vapours. We therefore set out to see how this difference might have led to increased friction and wears.

E‐8) wears and surface roughing due to sliding motion
E‐9) surface corrosion due to incompatible materials used for sliding mechanism

We conducted fuel environment friction tests. These tests use a pin-on-disc setup and we observe static/dynamic friction coefficients, amount of wear, wear particles by subjecting the test pieces (which have been immersed either in helium or propellant vapour environment) to sliding motion.

As a result we confirmed that there was no worsening by the propellant vapours. Rather, we observed that we obtain markedly larger results from the tests under helium gas environment. (surprise! helium should be less sticky!!!, P)

E‐11) Propellant vapour environment producing products which grit

Measuring the disc wear amount

The depth of the grooves left on the test pieces in the wake of sliding motion under helium environment was 5 to 10 microns. In contrast to this the maximum depth unde propellant environment was 4 microns and fuel environment did not contribute to the worsening.

Products under propellant fuel vapour environment

Upon sliding tests we dis see a minute amount of metal-metal friction induced metal particles, but we did not find any chemically produced products reacting with fuel or others

(there is a diagram and captions are;)

Fuel environment friction test
Pin
Disc

＜data obtained＞
Friction coefficient, wear amount, wear particles

As a result of these tests we did not observe any phenomena where fuel environment gave adverse effects to friction issues and we are satisfied that above candidates are sufficiently innocent.

(end of page 46)

[pandaneko]

(page 47)

B.5 Evaluating the propellant flow speed

We are now going to show two different evaluation methods, 1) method employed at the time of design and 2) method introduced as the transmission model based on our new insight into the mulfunctions

1) Evaluation at design time: by leak model

1-1) From the helium leak speed we assumed a viscous flow through an equivalent orifice and estimated the equivalent orifice diameter. In helium leak measurements we only have one component pressurising the downstream of CV-F and the upstream is vaccum, leading to adoption of viscous flow hypothesis.


FORMULA for viscous flow: (I cannot type it out. Please refer to the original page, P) where

Q(Flow,mass)： Flow mass speed ( mg/s)
ρ： density at the average pressure of up and down streams (g/m3)
μ：viscocity coeeficient （Pa・s）
d, L： bore diam. and length (mm)
ΔP： differential pressure (MPa)

1-2) We evaluated the propellant flow speed assuming difusion (or dispersion?, P) in a ficticious orifice due to the differential pressure within the two component system (helium + propellant).

In estimating the propellant flow speed we have a configuration where against a constant one atmospheric pressure (of helium and propellants) we have a saturated propellant vapour pressure downstream of CV-F and no propellant concentration upstream.

We therefore evaluated the propellant leak speed as due to propellant diffusion by the in-orifice differential pressure.

FORMULA for diffusion: (I cannot type it out. Please refer to the original page, P) where

Q(Leak,mass)： leak speed (mg/s)
Dgass： mutual dispersion constant of gas （m2/s）
RT： gas constant （8 3 J/mol K）×temp.（K）
ΔP： differential pressure（＝saturated vapour pressure） (MPa)

1-3) Evaluation result (for Q(Leak,mass))

(this is a simple table and I will narrate its contents as follows, P)

For fuel Q(Leak,mass) is 2x10-10 mg/s @0.0014MPa

For oxidiser Q(Leak,mass) is 2x10-8 mg/s @0.1MPa

(-10 and -8, to the power, naturally!, P)

(See section 2.2.1, item A on the table "Comparison of measured values and model values of propellant leak speed")

In all above we assumed all of the flow was due to leakage, ignoring transmission

(end of page 47)

[pandaneko]

One addition to page 45's graph:

I simply forgot to translate those captions on the graph. The red line is the re-seat pressure and the blue line the cracking pressure. Apologies!

P

[pandaneko]

(page 48)

(There are two tables here, one of them is simple and I will type it all out, but the second table will be translated after the main texts, P)

B.5 Evaluation of the propellant flow speed (continuation)

2) Evaluation at time of re-evaluation based on transmission model

(I now realise that my translation of "transmission" may be wrong, because the suffix to the formula below has "Per" and it probably means perforation...?, but I will carry on with "transmission", P)

2-1) Actual measurements and literature survey of transmission coefficient of the sealing material (polymeric)

(Here below is the table to be translated later on)


Transmission coefficient of gaseous molecules in polymeric materials is proportional to the product of degree of solvance (am afraid I do not know the right word, P) (itself proportional to the partial pressure of the gas in contact) and the dispersion coefficient.

Of these, dispersion coefficient has positive proportional relationship with molecular weight, but degree of solvance depends critically on molecular type (such as polarity, or polarisation). It is for this reason that helium has a larger transmission coefficient than fuel.

Also, the absolute value of the transmission coefficient is critically dependent on the crystaline structure of the polymeric material (itself dependent on production process) and the measured values may differ from the values given in literatures.

(here below, transmission coefficient, Per, is defined as; )


Per = S times Dsolid

Per： transmission coefficient (m2/sMPa)
S： degree of solvancy (1/MPa)
Dsolid： dispersion coefficient inside solids (m2/s)

2-2) Evaluation of transmission speed through the valve

Based on the transmission formula we evaluated the propellant flow speed by estimating the sealing material's geometric parameters (A/t) using helium flow speed

(here again, am unable to type out the forlura, P)

transmission formula

Q(Per, mass) ： transmission speed (mg/s)
A, t : sealing material's contact area with gas (mxm, m)
ΔP： differential partial pressure (= saturated vapour pressure) (MPa)
P： avareage partial pressure （＝(1/2) saturated vapour pressure） （MPa）

2-3) result of evaluation (of Q(Per, mass))

This is a table, but I will spell it out as;

Q(Per, mass) = 1x10-10 mg/s @ 0.0014MPa for fuel
Q(Per, mass) = 3x10-5 mg/s @ 0.1MPa for oxidiser

Note that item (B, bee) of the table in section 2.2.1 "Comparison of propellant flow speeds, model values and measures values" is based on the assumption that all the flows both up and down streams of the valve originate from transmission through the sealing material.

(end of main texts of page 48)

(here below the table mentioned above) (it is simple and carries the values (measured and literature) of Per, S, and D solid for oxidiser vapour and fuel vapour and helium)

(In this matrix the top entry in the 1st column is oxidiser vapour, next down is fuel vapour and at the bottom is helium)

(1st character string from the left in the 1st row is "measued value" and the next string to the right of it is "literature")

(the rest are all alphanumeric and please refer to the table on page 48)

Source: Polymer Handbook, 3rd ed., J. Brandrup and E.H. Immergut, John Wiley & Sons, 1989

(end of page 48)

[pandaneko]

(page 49)

B.6 Estimating the amount of propellant flow in Akatsuki propulsive system (fuel)

(hereafter is a schematic of propellant supply system)

(Leftmost box: fuel tank where diaphgram is located
section to its right: section D, between fuel tank and CV-F just before the step in the middle
small box just to the right of this step is a valve, then further to its right is CV-O, then finally oxidiserl tank, the box at the right end
between oxidiser tank and CV-O is section A
between CV-O and the simple valve is section B
between the simple valve and CV-F is section C to which helium from above flows)

Using the fuel flow speed across the valve and diaphgram which we measued we esimated the amount of propellant vapour crossing in the gas supply piping system. We found that the amount of fuel vapour moving across CV-F is almost zero and most stayed within section D (downstream of CV-F)

Note: Volume of D is the initial volume of empty space and does not include the increase as a result of fuel consumption

(hereafter there are 6 graphs in 2 rows and 3 columns)

(1st row from thre left are sections D, C, and B in this order and the same order goes to the second row as well)
(horizontal is the time line for 6000 hours, and the vertical for the 1st row is the amount of existing fuel in mg and the vertical for the 2nd row is fuel's partial pressure in MPa)

(red solid line is the measued value and red dotted is the pre-flight analysis)


Note: Measued values and pre-flight anaysis values are almost in agreenment in all sections

(end of page 49)

[pandaneko]

(page 50)

B.7 Looking at past examples of mulfunctions relating to oxidiser movement

Given Akatsuki's mulfunctions we conducted a survey of mulfunctions (minor mulfunctions included) during ground tests and in flight of overseas probes. We noted that there was no example of lost mission arising from salt formation. However, we also noted that there were a few cases of crystal formation of salts involving MMH (mono methyl hydrazine) as fuel.

 Example of severe mulfunction leading to loss of probe - reflected during design stage (direct translation, meaning unknown, P) -

• Mars Observer （launched in September 1992, approached Mars in August 1993, fuel : MMH and oxidiser ：NTO※)

It is estimated that during its flight the oxidiser condensed and liquefied at a cold portion of the gas supply system flew to the fuel side when the pyro valve was opened and led to explosive reaction

 Examples of mulfunction in orbit (mission was completed)

• Viking‐1 (August 1975 launch, Mars arrival June 1976, fuel : MMH oxidiser ：NTO※)

Regulator valve internal leak was observed during flight. Salt formation was assumed to be its cause.

• Intelsat‐603 (Launch 1991 fuel ：MMH oxidiser ：NTO※)

During its 1sy manuever an internal leak of regulator valve occurred. During 2nd and 3rd manuevers fuel side CV-O (or something similar, P) closure was observed. Salt formation was estimated to be its cause due to its long mission period.

 Example during ground tests

• Marienr‐9 (May 1971 launch, arriving at Mars in November, fuel ：MMH oxidiser ：NTO※)

CV closed during ground test burnt. Crystal formation of iron nitrate was confirmed upon dis-assembly of the system. Its cause was put to swelling of TFE.

※NTO is an abbreviation of N2O4. In actual use NO is often added to reduce possibility of metal corrosion. In the case of Akatsuki 3% of NO was added to N2H4 (forming MON-3).

(end of page 50)

[machi]

"Examples of mulfunction in orbit (mission was completed)"

Interesting! I didn't know this.

[Paolo]

in fact, it was the problem with the pressure regulator on Viking which prompted controllers to delay pressurizing the tanks on Mars Observer to just before orbit insertion, and the system (borrowed from telecom and meteo sats) was not designed for that...