View Single Post
Old Wed, Jun-20-2018, 06:07:07 PM   #30
nowanker
"He's no wanker!"
 
Join Date: Sep 2012
Posts: 900
Reputation: 0 nowanker is on a distinguished road
Location: Mountain View, SF bay area

United States




Default Re: A good example of why my engine internals have remained stock

Quote:
Originally Posted by Volke View Post
In general for hardware, unless it's aerospace or safety critical, you don't design to or test to worst case conditions. You come up with a composite application profile that represents how your average customer will use your product and design/test to that. Also, the design metrics usually allow for a certain amount of failures before the target life. For example, when building a new engine, BMW may decide they want a B10 life of 100,000 miles and a B50 life of 150,000 miles. This means that they are okay with 10% of the population failing before 100,000 miles and 50% before 150,000 miles.

EDIT: On top of B10 lifes, they may also say they only want a 90% confidence interval when determining how many samples they need to test and for how long to verify that the design meets the requirements.
I'm sure every manufacturer has a room full of actuarials (right term?) crunching data to determine the most profitable Manufacturing Cost/Warranty Repair Cost ratios.
Marketing probably adds their input on Acceptable Failure Rate/Brand Tarnishment ratio.
That paper (what I could understand of it!) was some impressive engineering, and without a doubt, BMW can make excellent engines. But just because the engineers have done their homework on how to make the best part is no guaranty at all that those parts are going into production.
And unforseen things do happen (M-Clunk? Sudden Exploding S54 Syndrome?), Then I guess corporate decides whether to own up and either make repairs, blame the owner, or just tap-dance around it.
Jump to top nowanker is offline   Reply With Quote