Rebreather World - View Single Post - Split from Cells thread

Genesis · 23rd November 2006, 20:22

Faceless, did you read the link you posted before you posted it?

Off one of the links in that wiki is the key to this:

Quote:

For engineering purposes, reliability is defined as:
the probability that a system will perform its intended function during a specified period of time under stated conditions.

and
First, reliability is a probability. This means that there is always some chance for failure. Reliability engineering is concerned with meeting the specified probability of success, at a specified statistical confidence level.

The problem that quoting a bare "MTBF" or "MTBCF" presents is that the uncertainty in the quoted number is undefined, and unfortunately, the uncertainty in most electronic component MTBFs, which make up the data from which a systemic value is computed, tends to be quite high.

Adding to this is the usual failure mode for electronic devices, more commonly known as a "bathtub"; that is, it is expected that a significant percentage of failures that will ever occur will happen early on, while the remainder (statistically speaking) will happen somewhat close to the expected service life of the device. This does not mean you can't have a failure any time during the totality of service life, but it does speak to the probability.

This is not limited to rebreathers. Just recently I saw a presentation on marine engines where the manufacturer was claiming "10,000 hours before overhaul" as the expected service life. That sounds pretty good! But - ask them to put that in writing in the form of a warranty, and suddenly you get presented with a warranty that specifies 3 years with no more than 400 hours of operation annually! That's 1200 hours if you use the boat exactly 400 hours every year. In reality since most recreational vessels spend one hundred engine hours underway a year, the warranty is good for three hundred hours - or three percent of the claimed "service life".

Now why is this important? Only because Alex has spent his life on this forum, including right here in this thread, slagging off every other eCCR design on the market.

Well, if indeed the DeepLife design is so vastly superior that I should not expect one I purchase to fail in a way that will force me off the loop for 2.6 billion hours, then by God, DeepLife should be willing to advocate that I dive that unit alpine! After all, the odds of me drawing the short straw in the gene pool via the unit are roughly equal to that of being struck by an asteroid.

If the computed value, however, is just that - a computed number, and there's no actual belief that the unit will in fact continue to function properly from the date of its construction until well past the time of my demise (from other causes, natch) then throwing this figure around is being done for marketing or political (e.g. to claim complience with an EN directive) purposes, and is not actually a statement of expected reliability.

My question is simple - is it smoke and mirrors like the Marine Engine folks, or is it real?

If its real then its worth a very significant premium in the marketplace in terms of cost. But if its smoke and mirrors, well, then you're really just paying for bragging rights, aren't you?

Anyway, I've said it here on this thread enough times that I believe thinking people can understand. Those who wish to argue for purchasing products based on marketing-speak are welcome to do so - just don't pollute the discussion by claiming that you're doing based on scientific evidence.

You're not.

23rd November 2006, 20:22	#39 (permalink)
Genesis Bubbless Box of Death Current Rebreather/s: Home Build Other Rebreather/s: Home Build Join Date: Oct 2005 Location: Sunny Florida Posts: 1,395	Re: Cell Linearity Faceless, did you read the link you posted before you posted it? Off one of the links in that wiki is the key to this: Quote: For engineering purposes, reliability is defined as: the probability that a system will perform its intended function during a specified period of time under stated conditions. and First, reliability is a probability. This means that there is always some chance for failure. Reliability engineering is concerned with meeting the specified probability of success, at a specified statistical confidence level. The problem that quoting a bare "MTBF" or "MTBCF" presents is that the uncertainty in the quoted number is undefined, and unfortunately, the uncertainty in most electronic component MTBFs, which make up the data from which a systemic value is computed, tends to be quite high. Adding to this is the usual failure mode for electronic devices, more commonly known as a "bathtub"; that is, it is expected that a significant percentage of failures that will ever occur will happen early on, while the remainder (statistically speaking) will happen somewhat close to the expected service life of the device. This does not mean you can't have a failure any time during the totality of service life, but it does speak to the probability. This is not limited to rebreathers. Just recently I saw a presentation on marine engines where the manufacturer was claiming "10,000 hours before overhaul" as the expected service life. That sounds pretty good! But - ask them to put that in writing in the form of a warranty, and suddenly you get presented with a warranty that specifies 3 years with no more than 400 hours of operation annually! That's 1200 hours if you use the boat exactly 400 hours every year. In reality since most recreational vessels spend one hundred engine hours underway a year, the warranty is good for three hundred hours - or three percent of the claimed "service life". Now why is this important? Only because Alex has spent his life on this forum, including right here in this thread, slagging off every other eCCR design on the market. Well, if indeed the DeepLife design is so vastly superior that I should not expect one I purchase to fail in a way that will force me off the loop for 2.6 billion hours, then by God, DeepLife should be willing to advocate that I dive that unit alpine! After all, the odds of me drawing the short straw in the gene pool via the unit are roughly equal to that of being struck by an asteroid. If the computed value, however, is just that - a computed number, and there's no actual belief that the unit will in fact continue to function properly from the date of its construction until well past the time of my demise (from other causes, natch) then throwing this figure around is being done for marketing or political (e.g. to claim complience with an EN directive) purposes, and is not actually a statement of expected reliability. My question is simple - is it smoke and mirrors like the Marine Engine folks, or is it real? If its real then its worth a very significant premium in the marketplace in terms of cost. But if its smoke and mirrors, well, then you're really just paying for bragging rights, aren't you? Anyway, I've said it here on this thread enough times that I believe thinking people can understand. Those who wish to argue for purchasing products based on marketing-speak are welcome to do so - just don't pollute the discussion by claiming that you're doing based on scientific evidence. You're not. __________________ "A venturesome minority will always be eager to get off on their own, and no obstacles should be placed in their path; let them take risks for Godsake, let them get lost, sunburnt, stranded, drowned, eaten by bears, buried alive under avalanches - that is the right and privilege of any free American." http://www.denninger.net http://www.diversunion.org/liability.htm - Fix the Diving Cert racket
(Offline)