Jump to content

Talk:Ariane flight V88

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
(Redirected from Talk:Cluster (spacecraft))

Origin of error

[edit]

Apparently this error was from using the "with Ada.Unchecked_Conversion;" feature... in which Ada allows one to loosen the strict type safety. — Preceding unsigned comment added by 131.120.202.60 (talk) 16:39, 28 November 2005

No "pragma Suppress" which switches off range checks (just as bad). --Krischik T 07:48, 12 March 2007 (UTC)[reply]

Grammar/Spelling

[edit]

I was reading some text and noticied "Since the back-up inertial system was already inoperative, correct guidance and attitude information", is attitude the appropriate word here or should it be altitude? I was about to change it but noticed the same spelling elsewhere in the document, not sure if this is some sort of flight or aerospace term. — Preceding unsigned comment added by 131.183.20.212 (talk) 16:52, 22 August 2006

Your assumption is correct, the text should not be changed. 'Attitude' is an aerospace term and denotes the angular rotations of a vehicle with respect to its flight path. Traditionally, the angles are given as 'roll, pitch and yaw'. Uffe 15:13, 8 December 2006 (UTC)[reply]

Disputed

[edit]
  • Only recently have Ariane 5 launches become as reliable as those of the predecessor model.

This doesn't seem right. 3 failures in 107 launches is a 2.8% failure rate; 4 in 36 launches to date is 11.1% (or generously, counting only complete failures, 5.5%). Possibly the statement is based on an Arianespace calculation, but it isn't clear, and it should in that case be cited as a claim ("Arianespace now estimates ...") rather than a fact. --Dhartung | Talk 09:05, 6 May 2007 (UTC)[reply]

In March 2007 issue of ASCEND, there was an interesting article about the improvement of the reliability of the western launchers after past failures (which includes Ariane 5 variants). Their main unit of measurement is the the differential rate of failures (delta # of failures/delta # of launches) rather than the integral one (# of failures/# of launches). I think this is qualified to back up the original wording of the sentence. I propose adding this article as a citation. 195.115.92.134 14:24, 20 July 2007 (UTC)[reply]

"Original source code"

[edit]

The link to the German document supposedly containing the original source code: The code given there appears to be written to explain the problem. IMHO it is not even an excerpt from the original, but entirely made up. Production code looks different, more cryptic vsriable and function names for example. —Preceding unsigned comment added by 91.97.63.136 (talk) 17:43, 1 April 2008 (UTC)[reply]


The code looks perfectly authentic. In the space industry, reliability is the important issue (!), and therefore names are required to be both meaningful and unique.Pe holmdahl (talk) 08:17, 2 March 2010 (UTC)[reply]


According to this presentation by Jean-Jacques Levy (who was part of the team who searched for the source of the problem), the actual source code that caused the problem was as follows.

L_M_BV_32 := TBD.T_ENTIER_32S ((1.0/C_M_LSB_BV) * G_M_INFO_DERIVE(T_ALG.E_BV));

if L_M_BV_32 > 32767 then
	P_M_DERIVE(T_ALG.E_BV) := 16#7FFF#;
elsif L_M_BV_32 < -32768 then
	P_M_DERIVE(T_ALG.E_BV) := 16#8000#;
else
	P_M_DERIVE(T_ALG.E_BV) := UC_16S_EN_16NS(TDB.T_ENTIER_16S(L_M_BV_32));
end if;

P_M_DERIVE(T_ALG.E_BH) := UC_16S_EN_16NS (TDB.T_ENTIER_16S ((1.0/C_M_LSB_BH) * G_M_INFO_DERIVE(T_ALG.E_BH)));

The tricky line is the last one, where the conversion from 64 bits to 16 bits unsigned is not protected. The code before is protected by testing before the assignment if the number is too big.

Schmitt Alan (talk) 13:13, 19 May 2011 (UTC)[reply]


Reason of Failure

[edit]

"The greater horizontal acceleration caused a data conversion from a 64-bit floating point number to a 16-bit signed integer value to overflow and cause a hardware exception."

That is the enough answer.

--78.173.33.239 (talk) 11:51, 17 November 2012 (UTC) -- zinzinzibidi.com[reply]

Title

[edit]

This article seems to be mostly about Ariane 5 rather than the satellites. Should the title be changed? bobrayner (talk) 19:59, 4 June 2013 (UTC)[reply]

  • No, the article should be expanded to cover the satellites. Per an earlier discussion, an individual launch is not notable enough to warrant an article, however a mission, regardless of its outcome or whether it is repeated, is notable. --W. D. Graham 20:24, 4 June 2013 (UTC)[reply]
    • Multiple reliable sources covered this launch; that's what enables us to have so much sourced content about the launch in the article. In what way do you feel it fails the GNG?
    • If you link to an essay that you wrote, calling it "an earlier discussion", but in reality it isn't a policy or a guideline, isn't a discussion, wasn't edited by anybody else, and doesn't mention this launch at all... that's rather unhelpful. bobrayner (talk) 21:48, 4 June 2013 (UTC)[reply]
      • That page outlines the outcome of the discussion that was held on the project's talk page. Multiple reliable sources cover most launches, however we hold that a launch is just a small part of a longer mission which is more notable, and as such the two should be covered together in an article about the mission. --W. D. Graham 00:41, 5 June 2013 (UTC)[reply]


List of launch failure

[edit]

The list in Launch failure section have letters of the original document where they come but not mention about this is in the article so it is a little confuse to see unordered letters in the list. Should it be rewritten as 'a)', 'b)', etc. or added info about its precedence?--PhoneixS (talk) 10:02, 9 January 2014 (UTC)[reply]

Confusing Title

[edit]

This title only made sense because I was redirected here from "Ariane 5 Flight 501". Do people really refer to this mission as the Cluster Spacecraft? It might have made more sense to call it the Cluster Satellites, although I think this article is really about the failed launch, and I do understand that a satellite is apparently not a satellite until it actually achieves orbit. Note that the caption below the picture does say "satellites." Robertwharvey (talk) 21:04, 15 May 2014 (UTC)[reply]

  • I tend to agree, though there's an older discussion of the title that appears to justify the current one. Admittedly, the payload didn't last to be satellites, but the article IS about the Cluster Spacecraft (satellite) Mission, even though the mission failed. I think the best option is to (also) have a separate entry on the failed launch (moving most of the info on the failure from here to that entry), and have a disambiguation line at the beginning of this entry pointing to "Ariane 5 Flight 501 failure." This is surely a significant event in studies of complex systems failures, as much as the mission itself is significant. OTOH I'm not a proficient Wikipedia editor, and don't know how to justify such a thing.Benbradley (talk) 02:14, 30 September 2019 (UTC)[reply]

The paragraphs h) i) etc.

[edit]

--Rainald62 (talk) 15:24, 14 June 2014 (UTC)[reply]

Summary: This bug has nothing to do with assertions

[edit]

In the summary it is said that the cause of the bug was turning off assertions. It has nothing to do with the problem, and I would never run software with enabled assertions during a rocket launch. It would make sense to say that this problem could be caused by ENABLING assertions, but never by turning them off. Assertions make sure the software WILL fail as it should, so that they tell developers about their mistakes during testing. They can never prevent crashing, only good code does that.

The launch, which took place on Tuesday, 4 June 1996, ended in failure due to an error in the software design caused by assertions having been turned off, (...)

If I don't see any complaints, I will rewrite that as I see fit:

The launch, which took place on Tuesday, 4 June 1996, ended in failure due to an error in the software design caused by a type conversion lacking range checking, (...)

Bertoche (talk) 19:21, 3 September 2014 (UTC)[reply]

I was wondering about that myself. I agree with your assessment and I believe this statement should be removed. In any other case, the term "assertions" should be explained properly Plmr (talk) 22:11, 2 December 2014 (UTC)[reply]

source code moved from article

[edit]

This content needs rework before it is suitable for inclusion. A "dump" of source code isn't really encyclopedic material, and there is no explanation or context given to help the reader make sense of it. 99.249.15.40 (talk) 09:38, 10 November 2014 (UTC)[reply]

Arithmetic overflow

[edit]

According to a presentation by Jean-Jacques Levy (who was part of the team who searched for the source of the problem), the source code in Ada that caused the problem was as follows.[1]

L_M_BV_32 := TBD.T_ENTIER_32S ((1.0/C_M_LSB_BV) * G_M_INFO_DERIVE(T_ALG.E_BV));

if L_M_BV_32 > 32767 then
    P_M_DERIVE(T_ALG.E_BV) := 16#7FFF#;
elsif L_M_BV_32 < -32768 then
    P_M_DERIVE(T_ALG.E_BV) := 16#8000#;
else
    P_M_DERIVE(T_ALG.E_BV) := UC_16S_EN_16NS(TDB.T_ENTIER_16S(L_M_BV_32));
end if;

P_M_DERIVE(T_ALG.E_BH) := 
  UC_16S_EN_16NS (TDB.T_ENTIER_16S ((1.0/C_M_LSB_BH) * G_M_INFO_DERIVE(T_ALG.E_BH)));

The last line (shown here as two lines of text) caused the overflow, where the conversion from the 64 bit float to 16 bit unsigned integer is not protected. The correct code would have been:

L_M_BV_32 := TBD.T_ENTIER_32S ((1.0/C_M_LSB_BV) * G_M_INFO_DERIVE(T_ALG.E_BV));

if L_M_BV_32 > 32767 then
    P_M_DERIVE(T_ALG.E_BV) := 16#7FFF#;
elsif L_M_BV_32 < -32768 then
    P_M_DERIVE(T_ALG.E_BV) := 16#8000#;
else
    P_M_DERIVE(T_ALG.E_BV) := UC_16S_EN_16NS(TDB.T_ENTIER_16S(L_M_BV_32));
end if;

L_M_BH_32 := TBD.T_ENTIER_32S ((1.0/C_M_LSB_BH) * G_M_INFO_DERIVE(T_ALG.E_BH));

if L_M_BH_32 > 32767 then
    P_M_DERIVE(T_ALG.E_BH) := 16#7FFF#;
elsif L_M_BH_32 < -32768 then
    P_M_DERIVE(T_ALG.E_BH) := 16#8000#;
else
    P_M_DERIVE(T_ALG.E_BH) := UC_16S_EN_16NS(TDB.T_ENTIER_16S(L_M_BH_32));
end if;

in other words, the same overflow check should have been present for the horizontal part of the calculation (E_BH) as was already present for the vertical part of the calculation (E_BV).

References

Self destruct?

[edit]

Did the self-destruct function (designed by UK?) destroy the rocket, or was destruction deliberately commanded from the contol centre? All the pieces had to be recovered by the French Navy from a swamp? 20.133.0.13 (talk) 11:56, 3 November 2015 (UTC)[reply]

Image "Recovered support strut of the satellite structure" ?

[edit]

Quite clearly (at least for everybody being familiar with the Cluster satellite system design) the image titled "Recovered support strut of the satellite structure" is wrong because it does not show such item but a recovered ELECTRONIC BOX instead (together with some part of the honeycomb platform where it was attached to). Probably the error is resulting from the fact that the description of 2 pieces of recovered Cluster parts (support strut and electronic box, respectively), as shown in ESA/ESTEC exhibition area, was exchanged by mistake and obviously not corrected in time when this picture was taken. This mistake, i.e. to understand what a support strut should really look like, could be verified e.g. by surfing to http://sci.esa.int/cluster/31381-3d-model/ and then to have a closer look at the tank support structure in the "full size flash video" being offered.

In conclusion: The image title should be corrected to "Recovered electronic box", maybe adding "(not support strut as indicated on the picture)"

95.33.193.176 (talk) 14:07, 21 February 2016 (UTC)[reply]

section needs more references

[edit]

Most of the "Launch failure" section has no footnotes. Please add them. Thanks. howcheng {chat} 18:32, 2 June 2017 (UTC)[reply]