Date: Wed, 20 Jun 2001 05:24:17 EDT
Reply-To: FrankGRUN@AOL.COM
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: Frank Grunthaner <FrankGRUN@AOL.COM>
Subject: My Turbo Audi 2.0 Project (not short)
Content-Type: text/plain; charset="US-ASCII"
I'm actually rather excited about this (smile). Been working on this idea for
several years. As I have often stated, I converted the diesel to the 1.8 L 8V
GTi engine from a '90 GTi a very long time ago. It has done every thing I
have asked of it save for two problem areas: lack of Moxie in the mountains,
and inadequate humph when running at 75 and faced with the need to pass.
After reading the reports of an extended mid-life crisis when Brian (?-sorry)
squeezed in a Subaru 6, I got more serious. I had entertained the possibility
of a 1.8 turbo, but it wouldn't fit without engine cover mods. Likewise the
VR6 and the ABA. Of course, the 1.8T, the TDi and the VR6 were all eliminated
because of financial reasons. I must be very careful around SWMBO. Hiding
cash, milk money, recycled bottle returns etc, and its hard to score a few K
that way.
Anyway, the 3A has always had my eye. Cheaper in the national yards since
they don't move compared to the ABA (just an Audi 4 you know). As you are
aware, the 3A has the same block height as the RV digifant 1.8 L, but an
honest torque enhancement. In looking at the engine internals, I discovered
that all 3A's have a forged crank, and only the early ABA's (up to '95) are
forged. The later ones are cast. Also, the 3A has piston cooling oil
squirters, while this appeared on the 96 and later ABA's. For a while, I was
going to swap over the head from my digifant, use the same simple engine
management system, etc, etc. But then I got my hands on the 3A engine and
discovered that there were several real differences as compared to the ABA
and the Digifant heads (always planned for a 8V engine and a non-crossflow
design for more torque). As opposed to my digifant head, the fuel injector
ports are in the intake manifold and not in the head. David Marshall alerted
me to the fact that this is a temperature bonus for the fuel injectors. Of
course he remains fixated on these arcane CIS systems, so I'm sure he didn't
appreciate some of the more subtle ramifications of this!
Anyway, taking off the Audi manifold and comparing it to the Digifant
version, the Audi intake ports were much larger with what appeared to be a
cleaner flow pattern. After some research, a bit of engineering and a few
hours on the lathe, I have built a set of modified injector ports to let me
use the Bosch (Digifant and LH-Jetronic ) injectors on the Audi manifold. To
do this, I also had to cut back the Audi CIS-E injector holder to unshroud
the new injectors. The net result is that the injectors now spray into the
intake port with no shrouded blockage! Remember, these injectors can be
placed from 1 to 4 inches back from the intake port. I've tested the injector
spray pattern at 35 psi and 50 psi, and the cone just fills the intake port
right before the valve guide. Good!
So far then, I am able to use the Audi manifold, reasonable injectors, add
the diesel pan, etc, etc. etc. But two other bonus points. I can use the more
efficient Audi crankcase ventilation system, and the throttle port is now on
the flywheel side of the engine. Well, I have the pan off and the rod shank
looks more robust that the RV, but can't compare it to the ABA, as I haven't
done an internal on one of those. Went to take it to the balancing shop and
was told to go home. The MOIC (Machinist Owner in Charge) told me that these
were already nearly perfect from the factory. He was willing to take my
money, but said that the balancing would only be effective for continuous
operation over 7200 rpm! We parted friends and I gave him the head for
dress-up work. To be reclaimed on Monday.
At this point its time to go to the turbo part of the tale. For years I had
toyed with the idea of building a low boost, high compression turbo.
Originally, I wanted to take the exhaust, intake and turbo from a 1.6L TD
Quantum engine, adjust the waste gate to 4 psi max, and pickup both power and
torque. As to my motives 1) There is my story of finishing my son's Porsche
944 turbo, and finding myself crushed back into the seat and accelerating to
more than 120 at 2 in the morning on the Foothill (210) freeway! Later we
would take it to 142 on the Palmdale corridor. Bet there were lots of donut
stains on CHP uniforms that morning! Oh yeh, the point: the car was drivable
at low rpm but when boost came on, it turned on like a rocket. 2). Rode in a
friends Saab 900 turbo in 1989. I owned the 16 V 2.0L 9000 at the time. Wow.
You would swear (sorry) that that thing had a classic V8 underhood. No turbo
turn-on just hard hauling from the bottom to the top of the band. 3. Rode in
a well tuned rabbit with the 8.5 CR engine, then in a rabbit with 10:1 CR.
The difference was low speed torque. Gobs more in the high CR car.
The typical approach of the turboheads is to cut the compression to around 8
to 8.5 to 1, then boost to power (12 to 15 psi), or till the sweet smell of
an aluminum plasma permeates the passenger compartment. So, I planned stock
high compression for low end torque and modest boost for genteel acceleration
as appropriate.
Then I began to search over the issue of how much boost was reliably achieved
by hobbyist turbocharged VW 2.0L motors. The answer is that with knock sensor
engines and premium fuel, the consensus was 8 psi. 8 psi! More than I needed
and at the stock 10.5:1 compression ratio. Now this was for a set of ABA
engines. There is one significant difference between the 3A and the ABA
engine. Rod length. Lots of discussion (complete with the strident pounding
of hairy chests) as to which is better. Well, it is said in the turbo
literature that the most important figure of merit on a turbocharged engine
is the rod - to - bore ratio. Ideal is supposed to be 1.75. The ratio for the
ABA is 1.66, while that for the 3A is 1.72. So the ABA with longer rods
(same displacement) is a bit more of a torquer, while the 3A will have a
better top end.
Unfortunately, the diesel turbo is not the correct choice and could be
readily damaged in the gas engine application. The diesel has a low
temperature exhaust (remember efficiency) and the gas engine is considerably
hotter. The increased energy (flow and Temperature) can drive the turbo into
a surge condition. Possible self destruction or at least manifold cracking
(temperature runaway).
Then I discovered the Saab turbo story in technical detail. The Saab 900
turbo engines through 1990 are 2.0 L with very similar torque and power
curves for the NA 8 V cars as compared to the VAG 2.0L motors. Therefore my
first thought was off to the boneyard and pickup a properly sized turbo for
peanuts. Actually came to $100. Then while stripping this '86 900T 16V motor,
I realized that the injectors were the same type as the digifant, the fuel
pressure regulator, etc., etc. Took em all. Then studying the Saab approach,
I realized that this was just what I was after.
Saab uses a high compression engine with very good engine management (Bosch
hot wire LH-Jetronic) and an independent system to control the turbo! Instead
of the classic approach of setting up the wastegate to dump at a particular
pressure, or setting it up for manual boost control (invites the Aluminum
plasma together with loud metallic noises), Saab electronically controls
boost with a separate analog computer. The waste gate is set up for a
baseline dump (if the full control is malfunctioning) at 4 to 5 psi. Then the
Saab APC computer monitors engine rpm and a knock sensor. The APC (Automatic
Performance Control) controls the wastegate operation through a solenoid and
applies the maximum boost just shy of detonation. For the Saabs, this value
is set at 10 psi. In other words, this system keeps the boost pressure below
detonation levels always. Back I went for the APC box, the LH-Jetronic
computer, ignition control computer, all switches, senders, fuel fittings,
ignition goodies, etc.
So, the plan: Audi 3A as described. Digifant fuel rail to feed Saab injectors
in the formerly CIS-E intake manifold. Adjustable fuel pressure regulator
with vacuum and boost referenced to dial in the proper pressure. Custom
exhaust manifold with T-3 flange pointing down at the flywheel end of motor
(bigger duplicate of the quantum TD manifold with proper flange). Liquid to
air intercooler between turbo output and the intake (recirculating pump, VW
Golf coolant reservoir mounted at frame on wheel well, with Mazda RX-7 oil
cooler mounted below current radiator acting as heat exchanger, liquid level
sensor in coolant tank. Recovery tank in engine compartment and temp sender
in circuit hose line). Saab hot wire air mass meter input to turbo, fed by
cooled K&N intake system in driver's side vent - previously described). Full
Saab engine control electronics. Remote oil filter and Mazda RX-7 oil cooler,
with MECCA Accusump oil lubricator - previously described. Diesel pan, mount,
as before.
Only major stumbling block is duplicating the Saab vacuum advance only Hall
triggered distributor.
So, to summarize the summary. This is planned to be a 2.0 turbo that fits
under the engine cover. All components are western European. Built for high
reliability. Anticipated power : 160 hp at 5500 rpm and 188 LB-ft at 3000
rpm. This is probably the limit for the clutch and transmission package in
the Vanagon.
Good Night,
Frank Grunthaner
|