Nozomi in perspective, Revisiting the causes of failure Options

[pandaneko]

http://www.mext.go.jp/b_menu/shingi/uchuu/...ts/04061101.pdf

Above pdf file will be translated for aspiring students in aeronautics, control engineering etc. so that in future lay people like me will be able to enjoy planetary scenes and events without worrying about failures.

The overall title is "Looking into the causes of failure and trying to find the right measures to take for the future with respect to the 18th scientific satellite (PLANET-B ) not inserted into Mars orbit as planned" and it is dated 21 May 2004.

This file is very much detailed at 1.1 megabytes and the number of pages is about 40, I think. In addition, I will be translating 3 more files after this particular file. They will be;

1. ISAS file with views and comments on the failure
2. Another ISAS file, a newsletter written out in a series of 4 individual letters.
3. JAXA file, which is a press release and it is a very concise document with just sufficient details.

Re concise link making I tried a few times, but I simply failed and all the links will be fully pasted out as required.

Pandaneko

[pandaneko]

http://www.mext.go.jp/b_menu/shingi/uchuu/...ts/04061101.pdf

above for ease of reference

Paolo, I am so glad my translation is being of help with your work! What was strange about yesterday's translation, I had a look at its original. My conclusion is the number 23 towards the end should have been 22, as simply I was translating the bulk of page 22. Having said that I am still feeling very uneasy because I did not type this number 23 myself. I was simply overwriting the copy pasted from the original. Anyway,

page 23

As for ② of 2) it might be thought of arising from the FET (field effect transistor, I think, P) switch failure for DC/AC conversion or IC failure for controling the pulse widths in the primary system (PWM). However, these two possibilities can be discarded because voltages are present on the secondary side through the transformer inside CI-PSU.

As for ③ of 2) components (except those imported from overseas) which could have caused short circuiting on their own singly include 24 ceramic condensors, 7 resisters, and unprotected 43 ICs. It is possible that any one of these could have caused short circuiting mulfunction, or alternatively, short circuiting inside the imported components could be assumed to have caused the mulfunction we have been looking at.

From all these above based on FTA we may summarise the failure causes whose responsibility cannot be ruled out completely as shown in the following (2

(2) Failure causes of short circuiting mulfunction

1) Influence of high energy particles arising from solar flare

a) Deteriolation effects by total dosage

Deteriolation by total dosage is often seen as the cause of solar cell deteriolation by the cumulative effects of high energy particles and appears as an increase in power consumption.

It is without doubt that NOZOMI encountered a very rare and massive groups of high energy particles. However, as of the peak flux on 22 April 2002 plus or minus a few days there was no significant power consumption increase which suggested above mentioned deteriolation (see schematic III-2-3).

In fact, the cumulative dose estimated from the lower portion of schematic III-1-7 suggests that it was similar to NOZOMI's design value (10krad equivalent assuming 1mm Al thickness) was within the tolerance limit. For this reason we may discard, as unlikely, the possibility of the total dosage leading to deteriolation which caused the mulfunction.


Failure by a single event upset

For this to be the cause following two conditions must be satisfied at the same time to explain the mulfunction of this time

① Unexpected switching over took place by a single event upset (note 11)


② Devices which could be switched on by above ① had already caused short circuting, or caused short circuiting following it

Of these, one possibility with NOZOMI is the INS-SA which is used only at the time of launch. Since it is used only at launch time components after the relays are all meant for commercial uses.

However, this particular device had been turned off after launch and it is confirmed that it stayed that way until the day when we lost signals on 24 April 2002. The possibility of this particular relay device being induced to be turned ON, as estimated from the solar proton minitor's count number (graph III-1-6), is 1,000 times higher on 22 April compared with the signal loss date of 24 April.

Furthermore, all this cannot explain the mulfunction of this time unless a short cuircuiting had already taken place in the system after the relay by the time the relay was turned ON, or alternatively a short circuiting did take place within a few hours of being turned ON.


- 24 -

(See, 24 above, it is happening again! this should be 23. It has been there all the time until I noticed it now, I think)

(note 11) single event upset (this ref is on page 24)

Bit flipping by passgae of high enery particles through ICs

end of page 23

P