Sender: <marv@lancaironline.net> (Marvin Kaye)
To: lml
Date: Mon, 04 Nov 2002 11:40:12 -0500
Message-ID: <redirect-1852285@logan.com>
Mime-Version: 1.0
In-Reply-To: <redirect-1851640@logan.com>
References: <redirect-1851640@logan.com>
From: "Lorn H. Olsen" <lorn@dynacomm.ws>
Subject: Re: [LML] Gas Springs
Content-Type: text/plain; charset="us-ascii" ; format="flowed"

>  >I wonder if using two gas cylinders of 60# each might double the chances
>that your emergency gear-down extension would fail?<
>
>The primary failure mode of the gas spring is "leak down" of the pressure.
>I have never heard of one failing any other way.  If you crank in all the
>statistical data into the equations, the probability of failure would
>definitely not be doubled.

If the probability of failure is f(120) on the 120# strut and f(60) 
on the 60# strut then the probability of failure on the combination 
strut would be 1-((1-f(60))*(1-f(60))).

I think that the failure rate f(60) would be less than the failure 
rate f(120) because of the higher pressure. Lets assume, for example 
purposes, that the failure rate for the 60# strut is .01 and that the 
failure rate for the 120# strut is .02.

If we look at this as a probability problem, then the probability of 
failure of the 2 sets of struts is:

   1)  pf(120) = f(120) = .02
   2)  pf(60) = 1-((1-f(60))*(1-f(60))) = 1-.99*.99 = .0199

The example above shows that the 2 strut system would fail less if 
the probability of failure of each of the 2 60# struts is 1/2 of the 
120# strut.
.
.
.
Actually, after thinking more about it, I don't think that this is 
really a probability problem. I think that the struts most likely 
leak in proportion to their internal pressure. Who has the pressure 
formulas handy? Is the pressure proportional to the square of the PSI?
-- 
Lorn H. 'Feathers' Olsen, MAA, DynaComm, Corp.
248-478-4301, mailto:lorn@dynacomm.ws
LNC2, O-320-D1F, N31161, Y47, SE Michigan