Date: Thu, 1 Nov 2001 00:11:47 EST
Reply-To: FrankGRUN@AOL.COM
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: Frank Grunthaner <FrankGRUN@AOL.COM>
Subject: Re: On Engine Efficiency, Comparing I4s,
WBs and Subies (not short)
Content-Type: text/plain; charset="US-ASCII"
n a message dated 10/31/01 6:22:28 PM, warren8@EARTHLINK.NET writes:
<< Frank,
Your post is very interesting, although I need to read your explanation of
BMEP more carefully as I must admit, my non-engineering type brain doesn't
quite comprehend the significance of the single number for BMEP and just what
it compares.
Does this mean the most heat derived from a given amount of fuel (the same
fuel??.)
The brake mean effective pressure reduces to the maximum pressure exerted on
the piston (although actually this reduces to the thrust on the crankshaft
rod journal). This takes into account the heat generated from burning fuel,
the coupling of that heat into the air charge (remember, heat transferred
anywhere but to the air charge is waste heat), the density of that charge
(volumetric efficiency of filling - involves flow resonances from the intake
and exhaust fluid circuits) and the frictional losses involved in translating
the piston/rod assembly. Good engineering will pack the cylinder with an
extremely turbulent charge at a high density - even a pressure slightly above
ambient and will slosh finely atomized microdroplets of fuel uniformly
throughout the gas. This charge will have a minimal amount of the previously
burnt charge mixed in from incomplete exhaustion of the cylinder. When the
flame front is initiated it will travel at nearly sonic velocities to ignite
the fuel and thereby release the thermal energy to heat the gas. This heating
and the flame front propagation dynamics as well as the turbulent flow all
contribute to putting as much heat into the gas as uniformly as possible. If
the gas were moving in laminar flow (smooth) large temperature gradients
could be established literally insulating parts of the charge from the heat
generated. Anyway, the flame initiation and burnrate must be carefully
synchronized so that the point of maximum pressure occurs just beyond Top
Dead Center, so the piston has the optimal initial acceleration at the best
point along the crankshaft rotation profile. It must also be timed to
complete the burn well before the exhaust valve opens. As far as piston
frictional losses are concerned, the current ring technology (total seal or
equivalent) offers a nearly hermetic seal with minimal friction for piston
speeds encountered with modern engines at around 7500 rpm. The next issue is
heat loss to the head and to the piston crown (just wasted). Here again air
flow dynamics is crucial. Hot surfaces conduct heat better than cold ones do,
so minimizing the surface area of the hot exhaust valve face while maximizing
gas flow at early opening angles, helps engineers justify all those calculus
and analytical geometry courses, not to mention partial differential
equations. I often think this gives the torque advantage to the 2 valve per
cylinder, non-crossflow head. Of course, eventually, the cross-flow design
will pump air with less loss. This means that the torque won't fall off so
rapidly at higher rpm, therefore higher horsepower. No five valve per
cylinder tractors that I know of!
Anyway maximum pressure in the heat engine wins the prize. Notice that newer
engine designs from a given manufacturer lead to higher pressures with each
new revision. But the turbo's pull ahead. More gas charge. Higher velocity
flow and volumetric efficiency, more turbulence, more bang. Plus in the best
sense of the Carnot cycle for the heat engine, turbo's add an energy
withdrawal stage by extracting work from the otherwise wasted heat in the
exhaust. Remember the as the gas does work by pressing the piston downward,
the gas is cooled (expanded). In the ideal case, the gas would be cooled to
ambient. In the real case the exiting gas at high efficiency load (most
efficient A/F ratio) is right around 740 degrees centigrade. These
temperatures drop below 300 C after exiting the turbo. So highest efficiency
has high compression ratio, 2 valves per cylinder, full reversion gas
kinetics (not cross flow), high pressure fuel injectors (minimum size fuel
droplets), sequential port fuel injection, multifire spark, variable valve
timing with rpm, and a variable flow turbocharger with intercooler and
conductivity sensed ignition timing. All this will approach a plasma inside
the chamber, so diamond or hafnium nitride surfaces would help.
Oh well, its a simple number to compare engine efficiencies. All that is
required is that the displacement is known (easy!) and that the torque maxima
numbers be real (more difficult if numbers are issued by sales types.
This number lets you assess the engineering quality of a candidate motor. The
torque curve is essential to understanding how that motor will perform under
your conditions. Flat torque curve from 1500 to 5500 rpm is good. Peaked
curve that drops to 50% of peak value at 1000 rpm above and below that peak
is bad (in my opinion). Bell shaped curves are the norm for HP vs. RPM but
tell a tale of obsolescence in torque output. The fuel efficiency maps are
the next part of the puzzle showing how broad the engineering optimum is in
terms of load vs. rpm. The final piece is the part throttle efficiency which
is a map of fuel efficiency vs. manifold vacuum. At least for self confident
German manufacturers, all this information is a matter of record in the
engineering literature.
I did notice some mistakes, however, in your representations of the various
Subaru 2.5L engines that, in the interest of posting accurate information to
the list, you might wish to correct. Not sure if this moves their various
placement up or down in your chart ....but regardless... they should be
accurate.
Warren, I try to be as accurate as I can with the information available to
me. All of the data I used here relating to the Subaru group was taken from a
summary note (posted to or from you) that appeared recently on the
SubaruVanagon list. I will update the numbers as soon as possible. Note that
the rpm for maximum torque is irrelevant to generating the brake mean
effective pressure comparison. Only the maximum net torque and the
displacement are important. I'm not really familiar with the Subaru engine
range and most of the data in the literature seems to apply to the 1.8L and
smaller engines.
According to the info I'm able to find so far,....ONLY the 96 model year
Subaru 2.5L develops its max torque at 2800 rpm. This engine, a DOHC
version requiring premium fuel was replaced in 97 with a SOHC version using
regular fuel which develops its peak torque higher in the rev range.
According to Subaru, the 97-98 2.5L (SOHC) produced its maximum torque of
162ft.lbs at 4000rpm (rather than at 2800 rpm).
In 1999 Subaru introduced their so-called "Phase II" engines (2.2L and 2.5L)
which they touted as having much improved torque curves throughout the rpm
range over previous years. The published info on the 1999-2001 2.5L is
166ft.lbs at 4000rpm, which doesn't look much different to me, but I assume
the torque curve for these engines is flatter with higher torque kicking in
at lower rpms. I have not been able to find any charts to verify their
claims.
Unfortunately, performance graphs are extremely hard to find. Those of us
with Subaru engines are not trying to keep any secrets, as you suggest.....I
just can't find any graphs. The only one I have been able to find is the
following one, which I believe to be for the 97-98 2.5L.
I never suggested that this data was being secreted by the SubaruVanagon
community, but rather I have been astonished at the attitude of the Subaru
USA and Fuji Heavy Industries about their information. All the major
manufacturers (and quite a few minor ones) participate in the engineering
literature, or will provide standard performance results to back their claims
(e.g. Ford, GM, Chrysler, VW, SAAB, BMW, Volvo, Renault, Citroen, Porsche,
Audi, Fiat, Toyota, Mazda, Mitsubishi and Nissan just to quote a few in my
files). In the past, Ford and GM refused to publish such functional data and
it developed that the numbers and ratings were a sales and marketing
fabrication. More recently a more legitimate attitude has developed were
companies tout their engineering accomplishments. Until I see otherwise, I
will conclude that the Subaru have highly peaked torque curves indicating
resonance tuning and poor broad band performance at low rpm. I have driven
several Subaru vehicles (all N/A) and have not been impressed with their
performance. Had to keep them up on the cam to hustle. I've only ridden in
two SubaruVanagon conversions. Both were 2.2L engines and both were dead off
the line. Just no lower end torque. The came alive over 4000. Hence the
reason I thought the gear ratio discussion would be appropriate. I received
16 messages stating that the Subaru engine had so much torque, any
transmission would be fine. I'll try to confine my posts to the Vanagon list
henceforth.
http://www.protekperformance.com/rv7/engine/dyno_graph.htm
The source is unverified. This graph (if correct) does show a rather
attractive flat torgue line however ( for a normally aspirated gasoline
engine) with torque at only 1200rpm reaching 140ft.lbs and maxing at 162 at
4000rpm, .......making most of the torque available in everyday driving
situations.
I'll check it out.
Attached is the latest version of my best attempts to compile info on Subaru
engines from 1990 to the present models. The file is in Excel spreadsheet
format.
Very helpful.
I don't know what impact these figures have on your BMEP chart but I think
the info, to be useful to the members of vanagon.com, should be accurate.
I have no vested interest in pushing Subaru engines over any other choice. I
made my decision based on my own unscientific research using the "seat of my
pants". Which (as you say "nonlinear tail" (scientific instrument??) had
not yet "desperately sensed the positive improvement stimulated by....a
......stunning infusion of cash "
I continue to suggest that, in addition to charts and data, prospective
converters do the same "scientific investigation" and drive one of each of
the conversions being considered.....then make a choice. There are enough
converted vans out there now that most should be able to do this. For me the
smoothness of the engine, near silent idle, and easy reving nature sold me
and my "nonlinear tail". Both of us were quite happy with our experience of
moving down the road quite briskly (at all speeds) in a brick.
As far as I am concerned, the best engine for the Vanagon is a diesel and a
TDi at that. Unfortunately the 1.6 N/A is too underpowered for a 5300 pound
vehicle. Of course, after I got started with the RV engine I wanted more
power and better mountain performance. The SVX and the 2.5 are interesting,
but not enough to deal with Kennedy. I had looked into other conversions when
I did the RV engine including Audi, Porsche (924 and 944), Volvo and Isuzu.
Tried to talk to the Kennedy group. Always brushed off. Tried calling back
over their clutch and pressure plate capability. Eleven tries in 12 months.
Say Vanagon Diesel and you are in the round file. Certainly agree about the
drive and test approach. Best I've driven is the 1.9 TD IDI with a 4.83 rear.
Also, regarding your posts to the SubaruVanagon list,........... you were not
(as you stated in your post), ......."caught and excised by Warren as an
unwelcome alien"......... I made a general post to the list...not to you
specifically....that it is the stated policy of the list that the purpose of
the list is to discuss the placement of Subaru engines into Vanagons. All
parties receive this notice upon signing up for the list, that.....
discussions of other engines or vehicles will be allowed ...for a short
while, ....but will eventually be referred to other lists which were created
for discussions of those engines and vehicles and where the members will be
interested in the info.
Sorry, you were just exposed to my version of humorless hyperbole. This is
what passes for my sick sense of humor. I lurk just to follow and learn about
interesting and novel solutions!
As the moderator of the SubaruVanagon list, it is my goal to keep the list
interesting and useful to the majority of the approximately 575 members.
Sincerely,
Warren Chapman
Moderator
SubaruVanagon Group
>>