X-Virus-Scanned: clean according to Sophos on Logan.com Return-Path: Received: from ms-smtp-04.southeast.rr.com ([24.25.9.103] verified) by logan.com (CommuniGate Pro SMTP 5.0.9) with ESMTP id 1126895 for flyrotary@lancaironline.net; Wed, 24 May 2006 20:43:39 -0400 Received-SPF: pass receiver=logan.com; client-ip=24.25.9.103; envelope-from=eanderson@carolina.rr.com Received: from edward2 (cpe-024-074-111-186.carolina.res.rr.com [24.74.111.186]) by ms-smtp-04.southeast.rr.com (8.13.6/8.13.6) with SMTP id k4P0grhn010202 for ; Wed, 24 May 2006 20:42:55 -0400 (EDT) Message-ID: <001201c67f94$43412cf0$2402a8c0@edward2> From: "Ed Anderson" To: "Rotary motors in aircraft" References: Subject: Re: [FlyRotary] Safety margin assessment Date: Wed, 24 May 2006 20:43:38 -0400 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_000F_01C67F72.BBE341B0" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2900.2869 X-MIMEOLE: Produced By Microsoft MimeOLE V6.00.2900.2869 X-Virus-Scanned: Symantec AntiVirus Scan Engine This is a multi-part message in MIME format. ------=_NextPart_000_000F_01C67F72.BBE341B0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable MessageWhy did I have a sneaky suspicion that there would not be any = quick and easy solution to this challenge {:>). Makes sense - without a = statistical base of sufficient size of failures and their causes, = numbers don't mean much. It appears the next best approach is to make = available the practices and approaches that appear to work with a high = degree of success. But, again, experimenters want to experiment, so I = don't have high hope that the situation is going to change much. =20 Ed ----- Original Message -----=20 From: Al Gietzen=20 To: Rotary motors in aircraft=20 Sent: Wednesday, May 24, 2006 6:01 PM Subject: [FlyRotary] Safety margin assessment I agree that other than a seat-of-the-pants feel that my subsystem's = are robust (in the sense of avoiding failure), I have no quantitative = measure of my system's "safety Margin".=20 =20 Would certainly be interesting for everyone to take the 0-10 exam on = their system and report the results of the assessment. Perhaps we might = find those above a 5 and give them some careful study. The problem, of course, is that with a one-of-a-kind system 'guessing' = at a number from 1 - 10 is not much different than 'guessing' that your = system is OK in regard to a given condition. The only way to get a = numerical rating is to have a large number of systems tested to failure = for a number of different variables. Al G ------=_NextPart_000_000F_01C67F72.BBE341B0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Message
Why did I have a sneaky suspicion that there = would not be=20 any quick and easy solution to this challenge {:>).  Makes sense = -=20 without a statistical base of sufficient size of failures and their = causes,=20 numbers don't mean much.  It appears the next best approach is to = make=20 available the practices and approaches that appear to work with a high = degree of=20 success.  But, again, experimenters want to experiment, so I don't = have=20 high hope that the situation is going to change much.  =
 
 
Ed
----- Original Message -----
From:=20 Al = Gietzen=20
Sent: Wednesday, May 24, 2006 = 6:01=20 PM
Subject: [FlyRotary] Safety = margin=20 assessment

 

I agree=20 that other than a seat-of-the-pants feel that my subsystem's are = robust (in=20 the sense of avoiding failure), I have no quantitative measure of my = system's=20 "safety Margin". 

  

 Would=20 certainly be interesting for everyone to take the 0-10 exam on their = system=20 and report the results of the assessment.  Perhaps we might find = those=20 above a 5 and give them some careful study.

 

The = problem, of=20 course, is that with a one-of-a-kind system =91guessing=92 at a number = from 1 =96 10=20 is not much  different than =91guessing=92 that your system is OK = in regard=20 to a given condition.  The only way to get a numerical rating is = to have=20 a large number of systems tested to failure for a number of different=20 variables.

 

Al=20 G

------=_NextPart_000_000F_01C67F72.BBE341B0--