|
You make some excellent points, Al, even though I
don't agree with viewing every failure as a system failure. An operator
failure can screw even the best of system designs - so I believe we need to
segregate failures into those two categories. To attempt to make some
system fool proof (even if possible) can add to complexity, weight and
cost. So I personally think you are forced to make compromises in any
design - just a personal opinion, of course.
However, I understand your point - approach it from
a systems viewpoint to minimize( so much as feasible) as many factors that can
cause a failure as we can - because you'll certainly never redesign the operator
{:>).
You've presented a lot of good points over the years,
Al. Have you considered producing a guideline pamphlet with an overall
approach to some of these areas?
Also, I know you're not poking at me, Al. Just a
starting point for the discussion which hopefully will benefit somebody on the
list.
Ed
.
----- Original Message -----
Sent: Friday, October 05, 2007 4:50
PM
Subject: [FlyRotary] Re: Coolant Water
Pressure
Occasionally, when experimenting, you encounter a
solution that has far reaching positive effects. This dynamic air bleed is one
of those. Pretend we have 100 planes with dynamic bleed, and 100 without.
We would find greater incidence of overheated engines in the ones without.
More compression seal failures, more pinging, more flights with low coolant
level, etc etc. Your description of effects you experienced makes this
clear....along with all my years of measuring effects, testing
concepts.
I think it's hard to appreciate just how
significant this is, because it's easy to get by, or just react to the problem
by taking it thru extra cooling cycles. Since we don't measure how effective a
solution is, we never realize a slight change can dramatically improve safety.
So "ok" solutions get transferred from plane to plane, when a "knock your
socks off" solution gets overlooked.
On my first engine install, I did a bunch of
things that were dumb in hind sight. I had a custom radiator made with the
radiator hose location 2" below the top of radiator. Hello! That meant I had
2" of air in radiator. So I had them install fitting on top that allowed me to
add 1/4" hose up to filler neck. Dynamic air bleed. With my new engine
install, I have no extra hoses anywhere. It's guaranteed to work, and I can
prove it all out before ever flying.
I replied to your post, but my focus is on the
guy that's designing his cooling system now and in the future. If he just
pretends coolant doesn't flow....and considers how air will exit each
component....he'll have safer system. Ed, you do much better job than I do of
combining theoretical and measurement. I really like your btu calculations
leading to diagnosing cause for high temps. That's good stuff. Valuable. To
some extent my post takes advantage of your open mindedness. Sorry,
I don't intend to pick on you. Just trying to save a crash, or overheat, or
whatever.
If you view every failure as a system failure
(not operator), you will find significant solutions. Good systems are
insensitive to operator errors. The dynamic bleed is a good example. I add
coolant, it fills right up. No trying to coerce coolant. No topping it off.
Less prone to operator goof ups. Most aircraft crashes have operator error as
major component. Same is true for manufacturing businesses. If you make most
of the systems insensitive to operator, then he is allowed to focus on those
few items that don't have system solutions. Operator makes fewer mistakes.
That's one of the key items I discovered as QA manager. It's what I've done on
my plane.
regards
----- Original Message -----
Sent: Friday, October 05, 2007 10:48
AM
Subject: [FlyRotary] Re: Coolant Water
Pressure
You are absolutely correct, Al. Just did not
think, gas/air compresses, liquids do not (to any appreciable
amount).
If by failure, you mean my system would not
have supported flight in that condition (air in coolant) , then you are
quite correct, on the other hand, one of the reasons to test (as you have
often pointed out) is to discover problems before flight inorder to preclude
failures in the air. In this case, I discovered my configuration
requires a bit more effort to remove the air. Once that is done, the
system is operating within its intended parameters. So does that make
the overheating a system failure or a failure to configure the system
to the proper operating parameters.
Even in the automobile, the rotary is known for
trapping air and requiring burping although many can get away without it
because most autos don't operate anywhere near WOT for more than a few
seconds, so the effects of trapped air (depending on amount) may
not be noticed.
I think there is a difference, for
instance you could design the perfect coolant system but fail to put
sufficient coolant into the system resulting in overheating - is that a
system failure or a operational failure? or perhaps more accurately -
operator failure?
I am always impressed by your meticulous
attention to detail and systematic approach. I could certainly
have benefit from your knowledge/approach back 10 years ago in my
initial design{:>). But at that time, there appear to be bigger
issues - such as trying to understand how a rotary engine
really worked - sorted out. Air flow and cooling were just
vague notions back then and I just assume they would naturally fall into
place {:>)
I still have a photo of my first oil cooler
installation to remind me of how ignorant of those matters, I was back
then . It had a "Plenum" that conformed to the area of the core - and
its wall stood 2" way from the core face at every point. Then I had a
2 1/2" dia hose piping air to one corner. Is there small wonder that
my first flight was limited to once around the pattern due to oil
temps{:>). Here is a photo - the oil cooler plenum and inlet are
the brown colored box on the left side of the engine (facing the engine)
close to the firewall with the large back hose (that one is 5" in dia as one
of my several early attempts to address the oil temp problems).
The radiator plenums were only slightly better.
No question, knowing what I now know, I would
have done some things different. As you know, Al, I
may sometimes take issue with your characterization, but not your approach
and insight. A valuable contribution to say the least.
Best Regards
Ed
----- Original Message -----
Sent: Friday, October 05, 2007 12:49
PM
Subject: [FlyRotary] Re: Coolant
Water Pressure
>at 22 psi the air would likely occupy even more
space
The opposite is true. Air space reduces
substantially when you increase pressure. Think of coolant as a
solid, and air as a giant spring. When you add btu's to coolant, it
immediately responds by expanding. This causes the air molecules to
compress...a lot.
There are some fun exhibits at the science museum I
volunteer at we use to demonstrate air/ water compression. An eye dropper
inside a pop bottle. When you squeeze the pop bottle, the eye dropper
plummets to the bottom. When you let go, dropper rises to surface. The
pressure increase when you squeeze bottle causes air in dropper to
compress(less air volume). This allows water to displace that
air....dropper is now heavier and falls to bottom.
It's really ironic. You can design a system that
seems to work fine. You fly with that for years with no failures( well I
guess you could call the pinging and high temps a failure). But a small
change can greatly reduce your risk. There's a lot of value to changing
your air bleed design to one that dynamically removes air. No shrader
valve, no repeated cooling cycles to remove air. By "dynamic" I mean that
it automatically removes air from the system. No muss no fuss.
As you design system, just pretend their is air at
the top of each component. Then find simple way to allow that air to move
to higher component in system. So, my radiator is lowest point in system.
If I place my radiator tube near the top of radiator, then all air
naturally leaves rad and flows to engine. Next I look at highest coolant
passage in engine. In my case I had to drill and tap a little 1/4 npt into
the coolant manifold, then run tube from there up to highest point in
system. Suddenly I've got a system that automatically removes all air
immediately. Try as I can, I can no longer trap air anywhere. If I develop
compression leak that pumps air into system, it has much less effect,
because it rises out of the coolant flow. When I drain and refill, it all
immediately and rapidly fills, I can get every drop back into the
system.
Dynamic air bleed is a safety advantage, easy
to accomplish. Each of these safety advantages adds up mathematically.
Likewise, contrary to popular theory, leaving two
cups of air under cap increases safety. You can prove that to yourself
with simple experiment I described earlier.
-al wick
----- Original Message -----
Sent: Friday, October 05, 2007 9:07
AM
Subject: [FlyRotary] Re: Coolant
Water Pressure
Yes, at 22 psi the air would likely occupy even
more space, but since I do my runup with the cap off or loose, there is
no pressure during the process. So while I have never measured it
and it probably varies from one time to the next, there appears to be
approx 1/3 of the top part of each core which has air on the initial
fire up of the engine - after a complete drain and refill of the coolant
system.
I must admit that the first several times of
draining and refilling coolant, not being as knowledgeable as I am now,
I almost cooked the engine, because I assumed that when the header tank
was full - the engine had all the coolant it could take. Rapidly
climbing coolant temps and pinging of hot engine cooling off soon make
it clear that just because the header tank was full didn't mean a whole
lot. Of course, I noticed after each run up that the coolant level
in the header tank would decrease permitting me to put more coolant
in. That finally made me realize what the problem was - would have
been very nice to have this list around back then {:>)
After burping the system there is still small
amount of air left, but the overflow tank set gradually removes the
remaining air over a couple of flights. Then the hydraulic "lock"
phenomena starts with initial pressure of 21-22 psi immediately on
engine start, dropping off quickly to zero and then gradually
climbing back to 5 - 7 psi as the coolant heats
up.
But, other than having to "clear" the air out with
a couple/three run ups to 5000 rpm, it works just fine and has since
97. I occasionally toy with the idea of putting in simple small
air bleed on the top of each - but, like I said, it works fine and other
things to do {:>)
Sounds like your approach will avoid my burping
problem. However, Lynn has mentioned that even in the car
installation it often takes burping the engine to get the air
out.
Ed
----- Original Message -----
Sent: Friday, October 05, 2007
11:12 AM
Subject: [FlyRotary] Re: Problem?
[FlyRotary] Re: Coolant Water Pressure
Hi Ed,
for sure I saw your installation before (numerous times...), but
I do not recall your exact pluming.
Your description below sounds like inlet and outlet are facing
down.
At 22psi it should even be more like 1/2 the radiator with air
:)
Anyway, I assume waterflow is radical enough to strip the air out
in 3 trials.
My system will have a bottom inlet and a top outlet. If it
doesn't fit the outlet may exit the bottom of the tank but will have
an internal standpipe - this way there is next to no space where air
can get trapped, just a small bubble atop the standpipe, won't be big
enough to cause any cooling detriment.
I still see BMW motorcycle oil-coolers mounted this way. Don't
know the exact make-up today, but the earliest ones where simple
single pass bottom feed bottom exit (cheapest solution and
esthetically least disturbing), a big problem to purge. 1/4 was
useless because of trapped air...
Furthermore, if the pump had a little leak or just a long time
between runs would drain the oil fro mthe cooler and at start-up you
had a fresh load of cold air inthe cooler! As it heats up the
air-bubble expands and reduces cooler volume even more...
Best Regards,
TJ
snipped..
In my case, if I do a complete
drain and refill of the system, on the first run up the core's tanks
will be hot approx 2/3 of the way up and then they are much
cooler - indicating that the remaining 1/3 of my core is filled with
air. It generally takes me 3 runups reaching 5000 rpm
before I can touch the core tanks and find them hot all
the way from top to bottom. So depending on your radiator set
up that might be something you can quickly check.
snipped....
--
Homepage: http://www.flyrotary.com/
Archive and
UnSub:
http://mail.lancaironline.net:81/lists/flyrotary/List.html
|