|
Home > Archive > Alternative Power sources > December 2005 > ?MTBF - Big and Heavy versus Modern and light ...?
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
?MTBF - Big and Heavy versus Modern and light ...?
|
|
| philkryder 2005-12-30, 10:21 pm |
| I recently read a posting that might fall under the category of common
wisdom.
It read along the lines that in mechanical devices Heavy low power to
weight ratio devices would last better than lighter high power to
weight ratio devices.
I'm sure that there was an unspoken caviat along the lines of "other
things being equal."
But, so often other things aren't equal.
For example some of the worst running and most unreliable autos I can
remember date from the early 1970s when Detroit manufacturers were
having trouble competing on both cost and function with Japanese
imports.
The American cars certainly had weight on their side.
And, they certainly had a low power to weight ratio - many had very
poor performance despite massive cast iron engines and 400 plus cubic
inch displacements.
A second example is a generator we are currently running (or not).
It is a heavy duty, moderate speed unit based on a teledyne comercial
V4 wisconsin.
It is heavy, side-valve, low power to weight, but plagued with a poor
magneto design that leaks oil into the breaker point area.
It doesn't "fail" - (no broken rods or cracked heads) but it fails to
run due to sub-system (ignition) failure.
Finally, many of the "heavy duty" units that I've looked at contain
many more PARTS than the cheaper, lighter units.
For example, a Cummins/Onan GNAB has twice as many cylinders, valves,
pistons, rings, tappets as a Generac 7500EXL, but lower continuous
output.
And it has a water pump, thermostat, radiator, cap, and hoses which all
have some probability of failure.
The GNAC with smaller peak output has 3 cylinders and THREE TIMES the
number of most parts.
When does "part count" start to count for more than "weight"?
When does more advanced technology become more important than brute
strength and weight when considering longevity.
And is there any substitute for "empirical" data.
Your thoughts appreciated.
Phil
| |
| Robert Morein 2005-12-31, 2:21 am |
|
"philkryder" <alt.google@Kryder.com> wrote in message
news:1135993262.115422.205190@z14g2000cwz.googlegroups.com...
>I recently read a posting that might fall under the category of common
> wisdom.
>
> It read along the lines that in mechanical devices Heavy low power to
> weight ratio devices would last better than lighter high power to
> weight ratio devices.
>
It is a factor. No one factor can be said to dominate, and there are obvious
exceptions which force tightening of the comparison. One of these is the
Texas Instruments DLP processor. The tiny micromirrors flap 60 or so times
per second, yet, according to TI, they have never been observed to break.
Some of the oldest surviving machines date from the 19th century, which, due
to their slow speed AND constant lubrication by human attendants, survive to
the present day.
Jet turbines run hotter and "faster" than reciprocating engines, yet are far
more reliable.
So it is necessary to equalize every possible factor; design, capability,
materials used. If this is done, it does appear to be true. But at any time,
an innovation is possible that, when added to the lightweight design, could
conceivably make it outlast the heavyweight design without the innovation.
| |
| Bughunter 2005-12-31, 9:21 am |
|
"philkryder" <alt.google@Kryder.com> wrote in message
news:1135993262.115422.205190@z14g2000cwz.googlegroups.com...
>I recently read a posting that might fall under the category of common
> wisdom.
>
> It read along the lines that in mechanical devices Heavy low power to
> weight ratio devices would last better than lighter high power to
> weight ratio devices.
>
Some of this effect may be due to marketing influences rather than
physics. Heavier design. and low-power-weight is often applied to
stationary units, used more for industrial and prime power applications
where weight is not a primary concern, but reliability and longevity
are more of an issue.
High-power-weight is emphasised on portable units, that are often
used sporatically, by the homeowner market. Who cares if it only runs for
1000 hours if you only use it for 100? Put youself in the position of
the marketeer and consider what design parameters you might
specify for the homeowner or even the "contractor" market. "If it's too
heavy
to load in my pickup truck, it will never get used, no matter how long
in can last."
You want to look at what market a generator is designed for, and not
necessarily
how much it weighs.
> I'm sure that there was an unspoken caviat along the lines of "other
> things being equal."
>
> But, so often other things aren't equal.
>
Most of the time, they are not equal. The common wisdom might get you
looking
in the right direction, but you have to examine the details.
> For example some of the worst running and most unreliable autos I can
> remember date from the early 1970s when Detroit manufacturers were
> having trouble competing on both cost and function with Japanese
> imports.
>
> The American cars certainly had weight on their side.
> And, they certainly had a low power to weight ratio - many had very
> poor performance despite massive cast iron engines and 400 plus cubic
> inch displacements.
>
Weight is too simplistic a parameter. The automotive market is an entirely
different
beast, driven by so many other factors, reliability and longevity being only
a couple of several. If you 57 Chevy still ran, would you still be driving
it?
Probably not, because it didn't come with electric windows, airbags or
satellite radio. A 57 Chevy was never designed to run for 300000 miles and
rarely did.
> A second example is a generator we are currently running (or not).
>
> It is a heavy duty, moderate speed unit based on a teledyne comercial
> V4 wisconsin.
> It is heavy, side-valve, low power to weight, but plagued with a poor
> magneto design that leaks oil into the breaker point area.
> It doesn't "fail" - (no broken rods or cracked heads) but it fails to
> run due to sub-system (ignition) failure.
>
A chain is only as strong as it's weakest link. If you put crappy tires on
a Mecerdes Benz, you get a crappy car. Still, replacing points or a magneto
is easier to do than replacing a crankshaft. The Wisconsin engine is what, a
1940's design? You think somebody would have solved this problem by now
with a simple redesign.
There is this notion of MTBF, and MTBF with repair. Ease and cost of repair
is a consideration. The military has a classifiation (H7) where repair is
considered infeasible. It defines end of life.
> Finally, many of the "heavy duty" units that I've looked at contain
> many more PARTS than the cheaper, lighter units.
>
> For example, a Cummins/Onan GNAB has twice as many cylinders, valves,
> pistons, rings, tappets as a Generac 7500EXL, but lower continuous
> output.
> And it has a water pump, thermostat, radiator, cap, and hoses which all
> have some probability of failure.
> The GNAC with smaller peak output has 3 cylinders and THREE TIMES the
> number of most parts.
>
> When does "part count" start to count for more than "weight"?
>
Parts count and complexity "always" counts. Weight is not really a factor,
but more
often a side effect of more rugged design. High mass may contrbute to the
ability
of an engine to wisthstand higher force (e.g. diesel needs to be more
ruggedly built
to withstand higher compression ratios), and to disipate heat.
Heat is a major factor in reliability. Which lasts longer, the lightbulb in
your refrigerator
of the lightbulb in your oven?
Contemporary automotive engines rarely fail because of failures in the main
engine iron, (bearings, crankshafts, pistons). Failure occurs more often in
the
vastly increased complezity of supporting items, computers, fuel injectors,
vacume hoses, etc., where the parts count has soared in recent years.
Diagnosis can be a bear, because of the shere numbers of parts.
> When does more advanced technology become more important than brute
> strength and weight when considering longevity.
>
It depends on how advanced the applied technology. Materials science has
made dramatic
advances in the last 50 years. It is not uncommon for an automobile engine
(shortblock) to
get well in excess of 200000 miles these days. Clearly, something has
changed. Better
design, and better materials. How often are ring jobs done these days.
Rarely. It is usually
the supporting subsystems that fail, like fuel injectors, computers, wiring,
vacume systems
and ten thousand other anicilliary parts.
But, better design and materials must be used to have any effect. It is not
clear that
they are being applied in many low cost homeowner generator designs where
low
cost drives the market. The engines in many are not much more then lawnmower
engines.
The antiquated Petter designs are often praised here. Better materials, no.
Advanced design, no.
Simple design and low parts count, yes. Slow speed, yes. Nostalga factor,
yes.
Pleasant exhaust note, yes. Repairable, yes, by a guy with not much more
than a hammer and
a cresent wrench. Power to weight, miserable. Quality of manufacture
(replicas), dismal.
Yet, they are attractive to a small segment of the market that only needs
relatively low power
and is willing to do the ocasional major overhaul because it requires only
basic mechanical skills
and not a mechanical engineering department. I doubt you wil find these in
use in
hospitals, except in third world countries.
> And is there any substitute for "empirical" data.
>
No. Unfortunately, getting empirical MTBF data has always been difficult. It
is rarely
provided by the manufacturer for mechanical assemblies unless maybe you are
looking at military spec items. Ancidotal is about all you are going to get.
> Your thoughts appreciated.
> Phil
>
| |
|
| "philkryder" <alt.google@Kryder.com> wrote in message
news:1135993262.115422.205190@z14g2000cwz.googlegroups.com...
:I recently read a posting that might fall under the category of
common
: wisdom.
Common wisdom usually turns out to be anecdotal experience from
the few when more than the few seem to have the same experience.
It's pretty much human nature, really.
:
: It read along the lines that in mechanical devices Heavy low
power to
: weight ratio devices would last better than lighter high power
to
: weight ratio devices.
Weight has little to do with MTBF or any of the supporting
parameters other than it comes along in order to get the numbers
needed. Weight is often a result, not a cause. And, it's not
really an indicator because there are as many factors that go in
the opposite direction too where lighter increases longevity.
Weight is never a factor in any MTBF that I'm aware of, although
mass does occasionally come into the picture. Mass is not
weight.
:
: I'm sure that there was an unspoken caviat along the lines of
"other
: things being equal."
:
: But, so often other things aren't equal.
:
: For example some of the worst running and most unreliable autos
I can
: remember date from the early 1970s when Detroit manufacturers
were
: having trouble competing on both cost and function with
Japanese
: imports.
Bad analogy, IMO, but I see what you're trying to say.
Calculating MTBF is a science and a rather complex one in its
implementation. MTBF also is not an indication of the length of
life of a product. MTBF is signigicant only statistically very,
very seldom is such that it can be used to predict the lifetime
span of any multi-part assembly. An MTBF composite of say 1,000
hours on a genset does NOT indicate a life of 1,000 hours for
that product. What it does say is that some component of that
genset will statistically, not WILL, require repair after 1,000
hours when it is used in accordance with design parameters. The
part that fails after 1,000 hours may also fail at ten hours of
10,000 hours - so it's not even a predictor of when a repair will
be needed in reality.
Every single part within an item has its own MTBF. Every
component of an item has an MTBF that is calculated on the MTBFs
of the single items. Then, the MTBF must be weighted, in order
to bring real world variables into the picture. And still it's
only a statistical estimate based on engineering knowledge and
formulae.
So, the value of MTBF is in comparison of collections of
parts, not an estimation of the lifetime of that product. Once
out of the laboratory, MTBF means almost nothing about the
lifetime of any single product and isn't even accurate for the
entire collection of all products built. It also includes of
course, the application and environment of the product.
As a result, MTBF has little value to anyone except the
marketing people who might like to add another "spec" to their
list.
That is NOT to say that a higher MTBF isn't a good thing; but
it's not dependable in indicating one thing better than another
in general. It's only when MTBF falls below expectations that it
would be useful and very few people would know how to apply it.
:
: The American cars certainly had weight on their side.
: And, they certainly had a low power to weight ratio - many had
very
: poor performance despite massive cast iron engines and 400 plus
cubic
: inch displacements.
:
: A second example is a generator we are currently running (or
not).
:
: It is a heavy duty, moderate speed unit based on a teledyne
comercial
: V4 wisconsin.
: It is heavy, side-valve, low power to weight, but plagued with
a poor
: magneto design that leaks oil into the breaker point area.
: It doesn't "fail" - (no broken rods or cracked heads) but it
fails to
: run due to sub-system (ignition) failure.
In general, parts count is the worst enemy of MTBF, of course.
But keep in mind that "failure" in the sense of MTBF does not
mean failure of the complete product. Replacing a worn out
muffler may be necessary, but it's a repair, not an end of life
event.
Then you have things like the MTBF of a piece of wire inside a
the throttle casing, an MTBF for the throttle casing flexing,
etc., and finally the weighting of the application of the pair
(rubbing, bend redius, pressure at points, etc.). So both parts
together can have an extremely long MTBF of many years, but one
improper radius can increase the wear and decrease the usability
fo the pair by thousands of magnitudes; thus, the weighting
factors, which is another science on its own. "Weighting" here
has nothign to do with weight; it's a statistical figure used to
add real world effects to an application.
:
: Finally, many of the "heavy duty" units that I've looked at
contain
: many more PARTS than the cheaper, lighter units.
:
: For example, a Cummins/Onan GNAB has twice as many cylinders,
valves,
: pistons, rings, tappets as a Generac 7500EXL, but lower
continuous
: output.
: And it has a water pump, thermostat, radiator, cap, and hoses
which all
: have some probability of failure.
: The GNAC with smaller peak output has 3 cylinders and THREE
TIMES the
: number of most parts.
:
: When does "part count" start to count for more than "weight"?
Parts count always counts more than weight. Weight is only
significant in moving parts when the weighting figures come into
play for friction and slippage, stress, etc.
:
: When does more advanced technology become more important than
brute
: strength and weight when considering longevity.
:
: And is there any substitute for "empirical" data.
And there really isn't any substitute for empirical data. That's
why anecdotal experience gets to be so important; it's often a
lot more reliable. Only trouble is, by the time enough
empirical data can be obtained for it to be reliable, the product
is usually long out of production <g> adn not much good for
purchase recommandations.
If there are many engineers here reading this, I'm aware of
the flack I'm going to get, BTW; I'm also an engineer.
The final thing that makes MTBF less than desirable unless the
difference is many, many magnitudes, is when you have replaceable
parts. When you look at a marketed MTBF, did the figures include
replaceable parts? Like spark plugs, filters, linkages, tune,
etc.? You would also probably find it interesting that MTBF can
vary substantially just with elevation above sea level and
components in the environment.
About the only MTBF the public is likelyh to see will come from
marketeers who have carefully picked thru the engineering
documentation for the numbers they want. An MTBF on a Honda
isn't very likely to have much comparison to say one on a
Kawasaki because you don't know where or how it was calculated.
:
: Your thoughts appreciated.
: Phil
:
Them's my thoughts today anyway; they might be different
tomorrow. My apologies to any statistical engineers here; other
than as nostalgia and pleasure I don't intend to do much
debating - I provided what was asked for here.
Regards,
Pop
| |
| daestrom 2005-12-31, 11:21 am |
|
"philkryder" <alt.google@Kryder.com> wrote in message
news:1135993262.115422.205190@z14g2000cwz.googlegroups.com...
>I recently read a posting that might fall under the category of common
> wisdom.
>
> It read along the lines that in mechanical devices Heavy low power to
> weight ratio devices would last better than lighter high power to
> weight ratio devices.
>
> I'm sure that there was an unspoken caviat along the lines of "other
> things being equal."
>
> But, so often other things aren't equal.
>
<snip>
You raise some interesting points. Reliability is a complicated function of
the number of parts, how each part is designed and used, and how technology
changes.
A common engineering practice in the old days was to design things with a
safety margin of two or more. Partly, this was done because calculations
and testing were limited to just a few decimal places (remember how
accurately one can read a slide-rule, or how many places were in your tables
of logs/trig). So parts *were* a lot heavier/stronger than they really
needed to be, and consequently, seldom failed.
As engineers learned to/pressured to save material/weight, they 'sharpened
their pencils', and started reducing the safety margins on components. But
then we all learned that absolute breaking strength wasn't the only
criteria. With larger stresses in these new, thinner/lighter components,
fatigue became an issue that wasn't very well understood before. And the
learning curve gets another 'hill' in it.
In today's society of consumerism, there are many who feel if it's cheap
enough, it doesn't matter so much that it doesn't last as long. Sort of,
"better to have a cheap tin 'pot' that burns out in a couple years, then to
not have any 'pot' at all." Hence the emphasis has sometimes shifted to
light/cheap/quick versus stronger/expensive/durable.
But even older stuff, designed with a large safety margin and long life
expectancy can still be a maintanence nightmare. Like your ignition system
example. Not a problem because it's heavy, or older, just a bad design
problem in the first place. I think there were just as many crappy designs
80 years ago as there are today, we just don't hear about them because
they've been consigned to the trash heap.
The good designs (then and now) have a lot of thought put into maintenance
and repair. And parts/surfaces that are expected to wear are designed for
long life and yet can still be replaced.
Some of the best ones I've ever seen were ones that a 'shadetree mechanic'
could tear down and reassemble in an afternoon, with only a handful of
tools. Part of what made Ford's first cars so popular was that a
*blacksmith* could re-line the bearings and any radio shop could fix the
ignition system with some magnet wire.
Some of the worst require all sorts of special pullers, fittings, tools and
require you to disassemble 3/4 of the thing just to *get* at the one part
that is faulty (we've all heard those stories about having to remove
alternator, power steering, A/C and an engine mount to replace a spark plug
;-)
And yes, different technologies bears a lot on it. A turbine, with forced
lubrication journal bearings can run 'forever' compared to a small gas
engine with splash lubrication. Not as much vibration means lower peak
stresses. Good lubrication means no startup wear. Gas engines can be
pretty finicky about their fuel, especially in cold weather. Diesels on the
other hand, can burn almost anything if you can get it to flow.
daestrom
| |
| nicksanspam@ece.villanova.edu 2005-12-31, 12:21 pm |
| Pop <nobody@devnull.spamcop.net> wrote:
>Calculating MTBF is a science and a rather complex one in its
>implementation. MTBF also is not an indication of the length of
>life of a product. MTBF is signigicant only statistically very,
>very seldom is such that it can be used to predict the lifetime
>span of any multi-part assembly. An MTBF composite of say 1,000
>hours on a genset does NOT indicate a life of 1,000 hours for
>that product. What it does say is that some component of that
>genset will statistically, not WILL, require repair after 1,000
>hours when it is used in accordance with design parameters. The
>part that fails after 1,000 hours may also fail at ten hours of
>10,000 hours - so it's not even a predictor of when a repair will
>be needed in reality.
> Every single part within an item has its own MTBF. Every
>component of an item has an MTBF that is calculated on the MTBFs
>of the single items. Then, the MTBF must be weighted, in order
>to bring real world variables into the picture. And still it's
>only a statistical estimate based on engineering knowledge and
>formulae.
> So, the value of MTBF is in comparison of collections of
>parts, not an estimation of the lifetime of that product...
Generators have a small number of parts compared to computers and
Cruise missles, and they wear out, so this kind of MTBF thinking with
failure rates and exponential distributions seems less appropriate
than wearout models with Weibull distributions.
Nick
| |
|
|
<nicksanspam@ece.villanova.edu> wrote in message
news:dp67j7$9mc@acadia.ece.villanova.edu...
: Pop <nobody@devnull.spamcop.net> wrote:
:
: >Calculating MTBF is a science and a rather complex one in its
....
:
: Generators have a small number of parts compared to computers
and
: Cruise missles, and they wear out, so this kind of MTBF
thinking with
: failure rates and exponential distributions seems less
appropriate
: than wearout models with Weibull distributions.
:
: Nick
:
Possibly, except I think I disagree with the "simplicity"
inference. MTBF was the topic however, and I addressed MTBF.
Mainly because in reality, it isn't that meaningful nor are the
wearout models for a lot of very similar reasons, actually. I do
think Weibull is a little easier to "prove out" in some ways
(assuming you have the higher math ability), but it's still
statistical in nature. Plus, I doubt very much whether anyone
here is going to have more than a passing knowledge of
what/who/why Weibull is. MTBF is still the phonified preference
of too many marketing types and it always seems to be misused
when it's presented to the public.
I'm about to get in over my head so I'm not likely to have
much more to say in this direction, but ... good points, at
least.
Regards,
Pop
| |
| Charles Foot 2005-12-31, 5:21 pm |
| daestrom wrote:
> "philkryder" <alt.google@Kryder.com> wrote in message
> news:1135993262.115422.205190@z14g2000cwz.googlegroups.com...
>
>
>
> <snip>
>
> You raise some interesting points. Reliability is a complicated function of
> the number of parts, how each part is designed and used, and how technology
> changes.
>
> A common engineering practice in the old days was to design things with a
> safety margin of two or more. Partly, this was done because calculations
> and testing were limited to just a few decimal places (remember how
> accurately one can read a slide-rule, or how many places were in your tables
> of logs/trig). So parts *were* a lot heavier/stronger than they really
> needed to be, and consequently, seldom failed.
>
> As engineers learned to/pressured to save material/weight, they 'sharpened
> their pencils', and started reducing the safety margins on components. But
> then we all learned that absolute breaking strength wasn't the only
> criteria. With larger stresses in these new, thinner/lighter components,
> fatigue became an issue that wasn't very well understood before. And the
> learning curve gets another 'hill' in it.
>
> In today's society of consumerism, there are many who feel if it's cheap
> enough, it doesn't matter so much that it doesn't last as long. Sort of,
> "better to have a cheap tin 'pot' that burns out in a couple years, then to
> not have any 'pot' at all." Hence the emphasis has sometimes shifted to
> light/cheap/quick versus stronger/expensive/durable.
>
> But even older stuff, designed with a large safety margin and long life
> expectancy can still be a maintanence nightmare. Like your ignition system
> example. Not a problem because it's heavy, or older, just a bad design
> problem in the first place. I think there were just as many crappy designs
> 80 years ago as there are today, we just don't hear about them because
> they've been consigned to the trash heap.
>
> The good designs (then and now) have a lot of thought put into maintenance
> and repair. And parts/surfaces that are expected to wear are designed for
> long life and yet can still be replaced.
>
> Some of the best ones I've ever seen were ones that a 'shadetree mechanic'
> could tear down and reassemble in an afternoon, with only a handful of
> tools. Part of what made Ford's first cars so popular was that a
> *blacksmith* could re-line the bearings and any radio shop could fix the
> ignition system with some magnet wire.
>
> Some of the worst require all sorts of special pullers, fittings, tools and
> require you to disassemble 3/4 of the thing just to *get* at the one part
> that is faulty (we've all heard those stories about having to remove
> alternator, power steering, A/C and an engine mount to replace a spark plug
> ;-)
>
> And yes, different technologies bears a lot on it. A turbine, with forced
> lubrication journal bearings can run 'forever' compared to a small gas
> engine with splash lubrication. Not as much vibration means lower peak
> stresses. Good lubrication means no startup wear. Gas engines can be
> pretty finicky about their fuel, especially in cold weather. Diesels on the
> other hand, can burn almost anything if you can get it to flow.
>
> daestrom
>
>
The engine in our power system is a Lister 6-1... 6hp, single cylinder.
Stands about 4' tall, weighs 860kg, runs at 650 rpm. Built in 1935. Not
sure about its MTBF, but I do know that over the years it has had 1 new
bearing, 1 new oil seal and 3 new head gaskets. It currently runs around
4 hours a day.
Reliable, basic, very well designed and built, and VERY easy to work on.
Last time I did anything to it was last year when I replaced the bearing
and oil seal, so I don't expect to have to take a spanner to it for
another decade or so.....
| |
|
|
"philkryder" <alt.google@Kryder.com> wrote in message
news:1135993262.115422.205190@z14g2000cwz.googlegroups.com...
> I recently read a posting that might fall under the category of common
> wisdom.
>
> It read along the lines that in mechanical devices Heavy low power to
> weight ratio devices would last better than lighter high power to
> weight ratio devices.
>
> I'm sure that there was an unspoken caviat along the lines of "other
> things being equal."
>
> But, so often other things aren't equal.
>
> For example some of the worst running and most unreliable autos I can
> remember date from the early 1970s when Detroit manufacturers were
> having trouble competing on both cost and function with Japanese
> imports.
>
> The American cars certainly had weight on their side.
> And, they certainly had a low power to weight ratio - many had very
> poor performance despite massive cast iron engines and 400 plus cubic
> inch displacements.
>
> A second example is a generator we are currently running (or not).
>
> It is a heavy duty, moderate speed unit based on a teledyne comercial
> V4 wisconsin.
> It is heavy, side-valve, low power to weight, but plagued with a poor
> magneto design that leaks oil into the breaker point area.
> It doesn't "fail" - (no broken rods or cracked heads) but it fails to
> run due to sub-system (ignition) failure.
>
> Finally, many of the "heavy duty" units that I've looked at contain
> many more PARTS than the cheaper, lighter units.
>
> For example, a Cummins/Onan GNAB has twice as many cylinders, valves,
> pistons, rings, tappets as a Generac 7500EXL, but lower continuous
> output.
> And it has a water pump, thermostat, radiator, cap, and hoses which all
> have some probability of failure.
> The GNAC with smaller peak output has 3 cylinders and THREE TIMES the
> number of most parts.
>
> When does "part count" start to count for more than "weight"?
>
> When does more advanced technology become more important than brute
> strength and weight when considering longevity.
>
> And is there any substitute for "empirical" data.
>
> Your thoughts appreciated.
> Phil
We replaced a 400 hp hammer mill motor. The old one worked fine. The new
one did the job, but you could hear it straining on the harder materials.
There was about 450 pounds difference from the old to the new. New was
lighter. New was sold as drawing less amps. It was really close so the
"electrical savings" was mute. To bad we could not get the old motor
rewound. Some new guy in the front office makes all of the decisions now. I
expect there will be several of us losing our jobs soon. Oh well, looking
for a job when I found this one.
| |
| Bughunter 2005-12-31, 7:21 pm |
|
"daestrom" <daestrom@NO_SPAM_HEREtwcny.rr.com> wrote in message
news:O2xtf.47170$XJ5.16200@twister.nyroc.rr.com...
>
> "philkryder" <alt.google@Kryder.com> wrote in message
> news:1135993262.115422.205190@z14g2000cwz.googlegroups.com...
>
> <snip>
>
> You raise some interesting points. Reliability is a complicated function
> of the number of parts, how each part is designed and used, and how
> technology changes.
>
> A common engineering practice in the old days was to design things with a
> safety margin of two or more. Partly, this was done because calculations
> and testing were limited to just a few decimal places (remember how
> accurately one can read a slide-rule, or how many places were in your
> tables of logs/trig). So parts *were* a lot heavier/stronger than they
> really needed to be, and consequently, seldom failed.
>
> As engineers learned to/pressured to save material/weight, they 'sharpened
> their pencils', and started reducing the safety margins on components.
> But then we all learned that absolute breaking strength wasn't the only
> criteria. With larger stresses in these new, thinner/lighter components,
> fatigue became an issue that wasn't very well understood before. And the
> learning curve gets another 'hill' in it.
>
> In today's society of consumerism, there are many who feel if it's cheap
> enough, it doesn't matter so much that it doesn't last as long. Sort of,
> "better to have a cheap tin 'pot' that burns out in a couple years, then
> to not have any 'pot' at all." Hence the emphasis has sometimes shifted
> to light/cheap/quick versus stronger/expensive/durable.
>
> But even older stuff, designed with a large safety margin and long life
> expectancy can still be a maintanence nightmare. Like your ignition
> system example. Not a problem because it's heavy, or older, just a bad
> design problem in the first place. I think there were just as many crappy
> designs 80 years ago as there are today, we just don't hear about them
> because they've been consigned to the trash heap.
>
> The good designs (then and now) have a lot of thought put into maintenance
> and repair. And parts/surfaces that are expected to wear are designed for
> long life and yet can still be replaced.
>
> Some of the best ones I've ever seen were ones that a 'shadetree mechanic'
> could tear down and reassemble in an afternoon, with only a handful of
> tools. Part of what made Ford's first cars so popular was that a
> *blacksmith* could re-line the bearings and any radio shop could fix the
> ignition system with some magnet wire.
>
> Some of the worst require all sorts of special pullers, fittings, tools
> and require you to disassemble 3/4 of the thing just to *get* at the one
> part that is faulty (we've all heard those stories about having to remove
> alternator, power steering, A/C and an engine mount to replace a spark
> plug ;-)
>
> And yes, different technologies bears a lot on it. A turbine, with forced
> lubrication journal bearings can run 'forever' compared to a small gas
> engine with splash lubrication. Not as much vibration means lower peak
> stresses. Good lubrication means no startup wear. Gas engines can be
> pretty finicky about their fuel, especially in cold weather. Diesels on
> the other hand, can burn almost anything if you can get it to flow.
>
> daestrom
>
>
When you mention old designs done with slide-rules, overbuilt margins of
safety that lasted
forever, the DC3 comes to mind. There are undoubtedly a couple still flying
out there.
| |
|
|
|
|
|
|
|
