RE: Way OT : Any maths bods around?

["christopher purves" <CHRIS_PURVES@xxxxxxx>]

Yep - just checked it with an ex. stats professor who works for me
(really).
He says most people assume exponential - resetting distributions for such
things. You integrate the distro between 0 and 10 hours where the distro
has
a mtbf of 100,000 hours. Of course there are 4 engines and you only need
one! In fact, they glide fairly well also.

Chris



----Original Message Follows----
From: "christopher purves" <CHRIS_PURVES@xxxxxxx>
Reply-To: ukha_d@xxxxxxx
To: ukha_d@xxxxxxx
Subject: RE: [ukha_d] Way OT : Any maths bods around?
Date: Mon, 28 Feb 2005 14:08:44 +0000


I would imagine you need to know the underlying distribution for this then
integrate between 0 and 10 hours.

Chris




----Original Message Follows----
From: "Hawes,Timothy Edward (GEG)" <haweste@xxxxxxx>
Reply-To: ukha_d@xxxxxxx
To: <ukha_d@xxxxxxx>
Subject: RE: [ukha_d] Way OT : Any maths bods around?
Date: Mon, 28 Feb 2005 13:56:47 -0000


1 in 10,000  ?  (100,000 MTBF / 10 hrs)

It's probably more complicated than that, e.g. what happens if it's
already flown for 99,989 hrs and you're flight is the last 10 before
it's service ? I guess they're serviced at 50% of their quoted MTBF
though and a 747 can still fly with one engine lost ?

Not planning a trip are you ?
;-)

Cheers,

Tim H.

> -----Original Message-----
> From: Ben McCormack
>
> OK this is way OT but I have got a query concerning maths and stats.
>
> Right here goes
>
> An engine on a 747 has a Mean Time Between Failure of 100,000 Hours.
>
> So is there a simple calculation for the probability that an
> engine will fail in a 10 Hour flight?