Barry Gardner wrote:

Paul,

Here are some perspectives on some of the turbo issues that have
been kicking around the list the last couple days.

You wrote: "I assume the divided inlet is a dual scroll. I am not
familiar with the term 'divided inlet'."

These are different. Let me quote you Corky Bell on this (which
he calls split-inlet):
"A split-inlet exhaust housing permits the exhaust pulses to be
grouped (or separated) by cylinder [or in our case, by rotor] all
the way to the turbine. The merit of doing this is in keeping the
individual package of energy, an exhaust putt, intact and
unmolested by other putts all the way to the turbine. This can
give the turbine a little better kick to get [or keep] it
moving...[he then proceeds to talk about V-8 exhaust energy, then
shifts to 4 cylinder engines...] A four-cylinder, by comparison,
which see only one putt every 180 degrees of crank rotation,
needs all the energy it can get from each pulse. Keeping them
separate and undisturbed will therefore pay some dividends."

I think it is safe to say that a two-rotor motor would benefit
from the same concentration of exhaust energy as a four-cylinder
piston engine. It also means that your manifold needs to be
divided by rotor all the way to the turbo. The HKS manifold is
set up this way.

"Dual scroll" was Mazda's proprietary way of trying to increase
low-end turbo power. You know how you can stick your thumb
partially over the end of a water hose and transform a high
volume of low pressure water to become a lower volume of high
pressure water? That's what the dual scroll Mazda turbo does. It
has a little flapper that concentrates the exhaust energy at the
turbo to get it to spin faster at relatively low exhaust volumes.
That way the turbo comes on line earlier. You are correct to
think that this makes no difference for aircraft applications.
Mazda dropped it themselves in later models, though the twin
turbo setup of the 93+ models was another way to accomplish the
same thing, i.e., develop turbo boost at lower RPM ranges, thus
making the motor feel like it has more torque.

I ran across another paragraph in Corky Bell's book that is also
of interest to the "do-it-yourself" crowd concerning manifolds.
In discussing manifolds, he writes about the various materials
from which manifolds can be made: "All stainless materials have a
very high coefficient of thermal expansion; thus, the design,
style, and fit of a stainless manifold must account for this
unusual property. For example, a stainless header flange drilled
perfectly for an exhaust bolt pattern with .3125-inch-diameter
bolt holes attached to a cylinder head with .3125-inch-diameter
bolts will shear half the bolts on the first warm-up cycle.
Larger-than-normal bolt holes are therefore necessary." All you
guys making those perfect-fitting lexan templates are hereby put
on notice!

Concerning coking of oil bearings versus water-cooled bearings,
it is true that shutting down an oil-cooled turbo quickly after
running it a 100,000 rpm tends to do bad things to it. And it is
true that the water-cooled bearings as in a Mazda seem to
eliminate that problem. However, you should remember that rotory
engines used in aircraft use already have a natural pattern of
use that should pretty well minimize the probability of that
happening: it's called landing and taxi.

Cars have potential for this problem because one can race into
one's destination, low gear, high rpm, and then reach for the
key. Aircraft are almost always throttled way back prior to
landing and then taxied at relatively low rpm all the way to the
pumps, tie-down, or hanger.

It would be nice to have water-cooled bearings but I don't know
how much effort or expense I'd go to to get it. Extra money,
extra plumbing, slight extra failure mode of the plumbing, slight
benefit in bearing life. The only scenario where the motor could
be making high rpms and then stop is an accident and frankly, I'd
have other concerns at that time beside worrying about the
bearings in my turbo. I'm open to someone making a case but...

Lastly, on intercoolers, you and Jeff are certainly correct that
the rear baffle on the factor air-to-air intercooler would need
to come off for use in aircraft. However, you might be interested
to know that it seems to work in the car! Engineering theory
aside, the proof in design is found in reliable empirical
testing.

There was a guy on the rotary Mazda list who believed, as you do,
that the rear baffle was an "obvious" bad idea. He first tested
the intercooler effectiveness by putting in dual thermocouples
before and after the intercooler and tested for temperature drop.
Then he drilled out those little pop rivets, reinstalled his
thermocouples, and re-tested, hoping to be vindicated. The
result: that silly looking baffle seems to work. He sheepishly
found his pop rivet tool and put it back on. It may have
something to do with the local pressure under the hood but, for
whatever reason, letting the air find its own way out was less
effective than directing it with the baffle *in that particular
location.*

A bunch of people have commented that afixing the intercooler
atop a very hot engine is probably more problematic than the
design of the baffle. A more effective intercooler is out in
front but that increases one's plumbing length, which increases
turbo lag.

Those Mazda engineers aren't complete idiots, you know. They did
bring us the engine and did a lot of neat stuff with it! And
maybe some of their other work was good too (though my nylon oil
lines were brittle and cracked, too, and I wonder....).

Barry Gardner
Wheaton, IL

Good post Barry. Thanks! Nothing like real world experience to keep us engineers
in line. In this case perhaps the intercooler baffle is the lesser of the two 
evils. Perhaps the radiant energy from the engine is affecting 
the intercooler when the baffle is not there. I have never been inside
the engine compartment with the hood closed :-)

Paul Lamar
 
The Aircraft Rotary Engine Newsletter.              Powered by Linux.

http://home.earthlink.net/~rotaryeng/            http://www.linux.org