|
|
First, let me say that I'm far from
being an authority on this subject.
The idea of coolant (oil, water, air, etc) moving too quickly
through a heat exchanger comes up often. People who's opinion I
trust (trained engineers) say that slowing flow does not improve
efficiency. What I've been told is that yes, you may see higher
delta T across the cooler with lower flow, but that's not a true
and complete picture of what's happening. My understanding, based
on what I've read & been told, is that the best heat exchange
occurs with the max temperature difference between the media
(oil>air, water>air, etc). If you slow the flow through the
exchanger, then yes, you will see a bigger delta T across the
exchanger, but that means that a lot of the oil (in this case) in
the exchanger has already been cooled 'early' in the flow, so
effectively, part of the exchanger is operating at a much lower
temperature difference with the air, and therefore, its efficiency
is reduced. So it follows that higher flow, keeping the entire
exchanger hotter (lower delta T) actually improves efficiency.
Yes, it's counter-intuitive (at least for me). But supposedly, the
most BTUs get removed from the system when the entire exchanger is
kept at close to the same temp across its face.
There's obviously a point of diminishing returns, where you're
actually adding heat by overpressurizing the flow path trying to
speed up flow, but I doubt we're there yet. :-)
Perhaps a real engineer could step in and clarify.
On 6/23/2018 9:35 PM, Andrew Martin andrew@martinag.com.au wrote:
Lynn, my setup is pretty much stock where most
oil should pass through cooler direct to rear iron ocv, only
oil that enters oil gallery is filtered, pressure, temp &
redrive oil taken from a block after filter,
But the cooler issue is a bit more incidious in
that without a pressure gauge at pump outlet there is no
indication of the restriction. I have no problem with having
“some” restriction in the cooler but as it builds markedly
with increased flow at rpm, Oil delta t drops as oil flow is
too fast through the core to cool the oil, and when front
cover relief opens at high rpm due to the restriction, only
part of pump output is getting cooled and temps rise more.
Setrab, Fluidyne etc do claim low pressure drop
but I have struggled to find at what flow rates, Adding smaller
coolers in parallel is an option but the data is still needed to
choose the correct sizes that allows all oil to pass through a
cooler without pressure drop and have just enough surface area
to transfer heat to air.
My test showed 140psi pump output 80psi at back
iron, I still dont know what my front cover relief is set at, as
140 was max pressure of gauge I had. But front cover relief
valve should never operate in normal operation as it is a safety
valve for the pump,front cover & cooler only.
Only engine that is diferent is 2009+ renesis as
that has only one valve in the system & diferent oil flow
design to the rest of the mazda rotaries.
Andrew
A restrictive cooler would (might) show a higher oil
pressure than the control valve will allow if
measured before the cooler. Because the stock relief
valve is at the end of the system. So the stock
valve might allow for 80 PSI, but never open if the
full 80 PSI never gets to it so as to activate.
Racers measure oil pressure where the oil enters the
engine. Usually in an aluminum block that replaces
the stock oil filter stand.
What
do the bearings see, is the information you want. We
raced for years with 80 PSI entering the engine.
And
that was turning the engine to 9,000 RPM on each
shift. Oil coolers are constructed of many sharp
edged tubes . Pushing oil or any liquid or gas into
the end of a sharp edged tube is nearly impossible.
So many more tubes than you would calculate
necessary are used in order to overcome the sharp
tube flow problem. So,
if the stock relief were set at 79 PSI
(stock on early engines) you would want to see 79
PSI on you oil pressure Gage as taken out of that
aluminum block. Mistral calculated the cooler size
required on the test Piper. The plane would overheat
the oil while still within sight of the airport.
The
were also using aircraft oil in the engine. 20-50 if
I remember correctly. So, flow got worse as the oil
heated up.
The
racer had an external oil pump with one pressure
section (adjustable up to any pressure you might
want) and two scavenge sections. The scavenge
sections returned oil and air to a storage tank
through a set of bug screen filters and two Setrab
44 row coolers in series. The pressure section
pulled from the tank and pressurized oil went
through two K&N oil filters in parallel and then
through a single 44 row Setrab cooler. So, we ran
100 PSI at the engine. Shifting at 9,700 RPM. 250
HP. Oil is Red Line 20-50 racing synthetic. A
common choice for rotary racing. Not a single oil
related failure in 30 years. Oil coolers (and
filters) in parallel reduce flow resistance. Coolers
and filters in series increase flow resistance.
Racing oils collect heat and give it up more quickly
than do conventional oils.So any cooler performs a
bit better with a synthetic.
Lynn
E. Hanover
Any
question, any time.
In a message dated 6/23/2018 4:59:30 AM Eastern
Standard Time, flyrotary@lancaironline.net
writes:
Just got around to plumbing in
mechanical gauge before cooler to see whats really
happening with my oil flows, wish I’d done it
years ago! Learnt so much in a couple of minutes
on things that I have wasted so much time second
guessing. my second attempt oil cooler did work
better than the original mazda cooler, but was
atrocious overall, Pressure drop was about 60psi
at 1400 prop rpm. No wonder I cant cool the oil,
bugger all is going through it, just enough to
give me about 80psi oil pressure.
Ended up bypassing cooler all
together to confirm it is the cooler that is
problem not lines or anything else, well what a
diference pressures constant at 78psi at all rpm’s
Trouble is no cooler manufacturer
here seems to have charts of flow & pressure
drop on their coolers, very frustrating especially
as prices seem to range between $100-900 for
similar sizes, so makes it very hard to select
correct one.
Andrew
--
Regards Andrew Martin Martin Ag
--
Regards
Andrew Martin
Martin Ag
|
|