Fwd: [Diy_efi] Re: Burning Aluminum etc.

[Adam Wade]

--- John Deakin <jdeakin at advancedpilot.com> wrote:

> I'm flying a little blind here, with no idea who any
> of you are, or what the genesis of this
> "conversation" is!

Whoops!  As we discussed, I posted our email thread to
the list, and discussion was thus spawned.  There was
a question (not yet in your email, since I have yet to
answer your last in response) regarding what happened
to pistons that caused the holes they sometimes
acquired (whether it was vaporizing, burning, or being
hammered away).  Phil I mentioned in the first email I
sent you rgd. gary's thread.

> I speak only of aircraft engines, almost always
> air-cooled, driving constant-speed props, with each
> cylinder having two widely-separated spark plugs.

I understand the redundancy factor and all, but man,
it sure cuts down on the operating margin between
"won't run" and detonation something fierce to have
two widely-separated plugs.  :(

> We have an enormous amount of data from the world's
> finest engine test stand in Ada OK, at the "General
> Aviation Modifications, Inc" test stand that has
> over 100 sensors on the present engine, the big TIO-
> 540 J2BD engine that powers the Piper Navajo,
> arguably the most "difficult" engine in the general
> aviation fleet.

Wowzers.  I am salivating at the thought of having
access to such a rig, or even the data taken with it
under varying conditions.  You make it actually
tempting to return to the midwest!  :D

> The data we see corresponds very closely to what is
> in the older textbooks from the older engine
> manufacturers (Pratt & Whitney, Curtiss-Wright,
> etc.) and some older engineering data from the "big
> two" today, Teledyne-Continental, and Textron
> Lycoming, as well as the magnificent two-volume set
> from Charles Fayette Taylor, "The Internal
> Combustion Engine in Theory and Practice."

I have had CFT's two-volume set for quite a long time,
and it truly is a comprehensive and invaluable
resource for anyone wanting to understand ICEs.  I
refer to both volumes on a nearly daily basis.

> The discussion of whether the aluminum burns or
> melts, and talk of weighing them is simply silly.
> Either way, the lost material is going out the 
> exhaust port,

That's something I'd question.  How much ends up in
the oil pan?  How much goes out the breather (and if
the breather doesn't attach to the airbox, or a main
seal blows, then it's not a closed loop).  I'm
definitely interested in where the metal goes and how
it gets there.  Further, many of the the ways to
combat the damage differ with different methods of
removing the metal, so from an engineering standpoint,
it definitely helps to know what is actually going on.

> I believe the violence of pre-ignition (or heavy 
> detonation) is creating hot spots that are intensely
> hot, and the aluminum simply melts and blows out the
> exhaust.

Looks like things are pointing solidly in that
direction, from the conversation so far today.

> Once the smooth surface is breached, further
> "events" will take advantage of the rough spots, and
> they will continue to be eaten away.

I would think that this might not be the case. 
Increasing average piston temps would increase the
likelihood of detonation, whether preignition
continued or not, but that would be "global", not
localized in the pits.  Perhaps Phil can weigh in on
this one as well.

> Detonation can be demonstrated and measured, and it
> is NOT the instant "explosion" of all the remaining
> fuel and air so commonly believed!  It begins
> with "light" detonation, tiny explosions of local
> pockets of mixture that are "perfect" to self-ignite
> under the local pressure and temperature
> conditions.  These cause shock waves that can easily
> be measured with two pressure transducers (one in
> each spark plug cavity), and these bounce back 
> and forth within the combustion chamber at the local
> speed of sound.

Check pages 40 and 41 of CFT's book, volume 2.  There
he discusses "autoignition" as being the cause of
detonation, and while the "flash-over" is not always
100% instantaneous, it is nearly so in almost all
conditions.  Since the volume of the combustion
chamber changes only negligibly during the burn, it
can be said to be "instantaneous" for practical
purposes.  Combustion always takes SOME amount of
time, but if it happens so quickly that nothing else
in the system changes for the duration of the event,
it is essentially "instantaneous".  Check out the
time-lapse photo montage on pg. 47; note how there is
essentially no combustion showing in the frame
immediately before total "flashover".  That mixture
just went from "not burning" to "all burning" in less
than 0.000125 sec.  That's 125 microseconds.  All the
standard modifiers for burn rate are still in play,
only on a much finer scale.  Proportionally the
factors are still the same, but the actual change in
burn rate is much, much smaller.  Look at the slopes
shown in the graph on page 55; two of three of those
pressure curves go STRAIGHT up.  The last is not quite
straight, likely due to movement of the piston rapidly
increasing the volume of the combustion chamber, and
the minimal heat energy remaining to be unleashed in
the much smaller amount of "end gas".

> "Light" detonation is harmless, and can be
> beneficial, for it cleans deposits off the piston
> head marvellously.

It takes very little to cross that threshold and move
into heavy detonation, which can cause nearly
instantaneous damage to the engine.  It's a very, very
touchy thing to try and stay in that exceedingly thin
"border area"; I would not rely on it, myself!

> An aircraft engine could probably run at 60 to 90%\
> of rated power for its entire life (usually
> something under 2,000 hours, or about 500,000 miles)
> in "Light" detonation 100% of the time, provided the
> cylinder heads are kept from getting too hot. 
> Generally, 400F is a good upper limit, though
> factories allow 460F and more.

Yep, I remember your diagram with the yield point for
aluminum based on temperature.  Has your experience
been, then, that one can reliably manage to balance an
engine's many factors right on that threshold of
detonation, without "going over" and damaging the
engine?

> We have one aircraft engine that has about 2 hours
> of "heavy" detonation in short bursts, terminated
> before the CHT could rise too much.  Probably 200
> hours of "light," 50 of "medium" along with that,
> and when the engine was finally torn down, no
> problems were found.

I'd be interested to know what the overheating of the
pistons had done to their structure.  Obviously you
allowed them to cool by restricting the extent of the
heavy detonation, which is why the engine lasted "so
long".

> It is pre-ignition that is the monster in all this,
> and we demonstrate and record that too, in every
> seminar we do.  Primary cause in aircraft engines 
> is probably a "helicoil tang," a small sliver of
> metal from the helicoil (spark plug insert) that
> projects into the cylinder.  A cracked spark plug 
> ceramic will prevent heat transfer from the spark
> plug tip to the cylinder, and that is probably a
> major cause of pre-ignition.  We've seen no evidence

> that "deposits" cause pre-ignition, though that is
> widely believed.

Something has to superheat the carbon to get it hot
enough to be an ignition source.  I'd wager that
something else would have to introduce substantial
heat to get to that point, thus carbon deposits would
be a "contributing factor" in a catastrophic cascade
failure of a destroyed engine.  But without other
events to initiate things, the carbon would sit there
without incident indefinitely (unless the compression
ratio raised enough from the presence of the deposits
to compression-ignite the mixture).
 
>>> Detonation is on the other hand an extremely
>>> violent and unforgiving condition. It is the
>>> result of spontaneous combustion of the air fuel
>>> mixture. There is no flame front to be seen as
>>> it occurs everywhere at the same time.

CFT's writings back the above part almost 100%.

>>> The pressure rise traces are vertical

This as well, from his own graphs.

> No.  Absolutely not correct.  I believed this myself
> until a few years ago, and it a very, very common
> misconception, even among engineers.  Detonation 
> shows on the pressure trace first as little
> "jaggies" superimposed on a normal trace.

I believe CFT indicated that the huge instantaneous
spike in pressure was too sudden to show up on
pressure traces, and the "spikes" were in fact shock
waves ricocheting around the combustion chamber, and
varying based on the local speed of sound through the
different "gas pockets" in the cylinder, with varying
amounts and types of combustion having occurred in the
already-burned, partially-normally burned, and
auto-combusting areas of the cylinder's contents.

> In these aircraft engines, normal combustion is
> around 800 PSI max, with a few engines seeing 1,000
> PSI at full power.  Heavy detonation may drive that
> up a few hundred PSI.

CFT indicates that the actual peak pressure is well in
excess of the figures you mention, which makes sense,
since he claims the pressure transducer cannot react
quickly enough to capture the huge and sudden spike at
the point of auto-ignition.  The fact that much more
radical and complete combustion happens much closer to
the auto-ignition point is what makes "heavy"
detonation more damaging.  Note that it's not the
SPREAD of combustion that is at issue; it's how
quickly the already burning mixture gives up all its
energy and finished combusting.  Or that is how I
understand it from CFT's writings, anyway.

| 82 Honda CX500 Turbo (Cassandra)  90 Kwak Zephyr 550 (Daphne) |
| "It was like an emergency ward after a great catastrophe; it  |
|   didn't matter what race or class the victims belonged to.   |
|  They were all given the same miracle drug, which was coffee. |
|   The catastrophe in this case, of course, was that the sun   |
|     had come up again."                    -Kurt Vonnegut     |
| M/C Fuel Inj. Hndbk. @ Amazon.com -  http://tinyurl.com/6o3ze |


		
__________________________________ 
Do you Yahoo!? 
Plan great trips with Yahoo! Travel: Now over 17,000 guides!
http://travel.yahoo.com/p-travelguide