Date: Wed, 6 Sep 2000 02:33:18 EDT
Reply-To: FrankGRUN@AOL.COM
Sender: Vanagon Mailing List <vanagon@gerry.vanagon.com>
From: Frank Grunthaner <FrankGRUN@AOL.COM>
Subject: Engine and Transmission Swap BS: was Thinking Swap ...(Now Full)
Content-Type: text/plain; charset="US-ASCII"
Just have to weigh in over some of the absurd and misleading comments being
made here over the relative merits of the "built" WBX vs. the inline 4 engine
in the course of Lilley's (sp?) latest apostolic crusade over the engine swap
issue. Ordinarily this would just drive me to another Augustinerbrau
synchronized with the delete key and a localized snarl, but this has pushed
me over the edge. The list is full of new members and many with minimal
mechanical expertise and I am concerned that these comments will go
unrebutted. I see that many other list members have also waded in but I had
to offer my two centimes as well.
First: as several listmembers have been pointing out, this "built" WBX engine
is simply an experiment. I will admit that blueprinting an engine (been
there, done that) can significantly enhance engine efficiency by reducing
frictional losses (here significant translates as a 1 to 5% change).
Deburring, radiusing, shot peening and balancing can all effect reliability.
These effects are nearly impossible to quantify without serious statistical
evidence (numbers!). This has (along with ceramic coatings for thermal
control) often been claimed as the racers advantage (blueprinted engines).
Yet, as Heath and others have pointed out, these results are only validated
by high stress accelerated testing (running the engine full stress, ass
against the wall) under service conditions never seen by a street engine.
Many small dollar racing enthusiasts follow carefully the latest engine
building tricks promulgated by Smokey Yurnik and his colleagues, and much of
the results of their work decorate the trash bins of race tracks all over the
country. The concepts are straightforward, but the devil is in the details. I
have several friends who lost their vehicles, their savings and their wives
over this intoxicating attempt to outengineer the next guy and make the
uniquely competitive machine.
But back to the point: ONE INSTANCE OF ANYTHING IS AT BEST AN EXPERIMENT, AT
WORST AN OBSERVATION. IT IS NOT REPRESENTATIVE OR PREDICTABLE. Science and
modest family economics demand a statistical burden of proof for claims of
longevity. Should Robert's engine last for 200,000 miles without an oil
change and give steadily increasing gas mileage, IT WOULD STILL REMAIN AN
ANECDOTE! To have confidence in its viability, it must be tested, it must be
measured and it must be reproduced! This requires deep pockets on the part of
the volunteer experimentalists.
MY ADVICE: Do this to play, do this to learn but do this with the full
understanding that the real world reliability of this system could be worse
than a nominal WBX rebuild.
Second: This old barstool tale that engineers are not permitted to build in
reliability is complete bullshit. Heath said it well when he noted that
reliability and longevity sell cars. Many makes (Rolls, Toyota and others)
literally owe their market to a reputation for outstanding reliability. I am
speaking of the first Silver Ghost here. Other makes have gone to the brink
of ruin over poor reliability (Hyundai and Chrysler, the Roots group,
Renault, etc.). This anecdote sits right up there with the 50 mpg carburetor
that Detroit wouldn't let the inventor build, an on and on!
Unfortunately this misinformation flows readily in politics and religion. In
my opinion, when VW bought NSU and got the K70 engine (ancestor of the
existing inline 4) it was an act of near desperation in attempt to save the
company. The air-cooled engine had run its course, and a water-cooled design
was sorely needed.
Back to the point: An engineer will build the best engine he can within his
initial resource limitations. Those resource limitations will be determined
by the volume of the market, the sophistication of the machine tools, the
foundry process in short, the technology at his disposal and the amortization
base over which he can spread his cost. "The engineer" of which we speak is
probably the chief engineer of a very large team. Over the years, of
production, a stunning wealth of reliability data will be obtained from the wa
rranty records and continued development. VW has a history of adding very
sophisticated materials and air flow technology to their mass market engines.
Often these enhancements result in the addition of dollars to the cost of the
engine (would be hundreds at the Porsche production scale). The I-4 engine
has demonstrated its remarkable robustness over a period of more than 30
years of continuous development.
Having said all of the above, it is also obvious that a large market exists
for performance add-on hardware. Why? Are the small time tuners like Berg or
Darryl Vittone and others more intelligent than the VW equivalent? Do they
have an english-speaking intellectual advantage? Of course not. The VW
engineers are also bound by the statistical variation of process and the
range of operating conditions (read emissions control, fuel quality,
operating temperatures, maintenance tradition) which define the mass market.
Most US tuners start by offering hardware that brings US vehicles up to
European standards, then they press the margin (chip manufacturers) and focus
on a few configurations. The results push the engineering compromises more to
performance and further away from longevity.
You pays your money and you take your licks. The average performance addict
is blissfully unaware of the placebo effect! (Translation: In all medical
drug scientific studies, the patient study pool is broken into two groups.
One will be given the new drug that is the subject of the experiment, the
other will be given an identical looking pill with no active ingredient
(placebo). Generally, the placebo group shows a cure rate at least as good as
the new drug. Think pre-viagra study groups. The human mind is a powerful
agent but a lousy sensor. After you pay your money, bleed to do the install,
then put it on the road, it always feels stronger. (Been there, done that).
No substitute for dyno tests. Read the enthusiast mags that test exhaust
headers! Often hard pressed to show 2-3 hp gains (with a dyno error of 2-3
hp!).
Third: THIS CONTINUING OLD WIVES TALE THAT HIGH RPM's ARE BAD FOR ENGINE
LIFE. BULLSHIT. A traditional low volumetric efficiency undercooled American
V8 had drastically reduced lifetime when run at 4000 rpm and above for
extended periods. In the seventies, I was consulting with Ford over
metallurgical wear issues in the design of piston rings working in cast iron
blocks. The data readily showed the relationship of surface abrasion (wear)
as a function of piston speed. For reference technology (designs developed in
the fifties), the wear curve steadily climbed from 1000 rpm, then took off at
a much higher rate at 3700 rpm. This threshold behavior was a function of the
oil temperature and the oil base stock for a given ring/block combination.
However the then current seventies technology showed a plateau. That is, the
wear first increased from 1000 to 2000 rpm, then stayed constant to about
3850 rpm, then increased at a slow rate to 4700 rpm whereupon it again took
off. So no lifetime difference operating the engine in a band from 2000 to
3850. New materials, different oil control ring technology and ring thickness
made it possible to expand that flat wear band to about 4000 rpm (so same
wear from 2000 to 6000 rpm) - all this in the seventies. Then in the
eighties, Honda and Phillips (independently) developed a ring technology that
literally reduced wear as the piston surface velocity went up. Of course at
high rpm the curve eventually reversed itself and wear again increased. In
modern engines (post the eighties with EFI and tight emission controls), the
engine/piston/rod/ring/bearing compromise is optimized for the anticipated
operating range of the engine. European engines have been traditionally
designed for higher operating speeds and small engines are routinely operated
between 4500 and 6500 for 250,000 to 300,000 km without the need for a bottom
end tear down. A properly designed engine will deliver its maximum efficiency
at the designed power band. For the VW I4 this band covers a very broad range
(the +/- 10% points off from maximum efficiency are 3200 and 5800 rpm as I
recall - Its all in the SAE Journal!). So actual fuel economy (fuel consumed
per unit time per power developed) is a function of the vehicle drag more
than rpm. Engine longevity is assured by operating the engine with minimal
bearing surface loads (not lugging!!!).
Summary: Running your engine (I4 statement) at 3400 (60 mph) or 4500 (60 mph)
has no bearing on engine lifetime, and only a modest impact on fuel economy.
Remember ... torque at the road wheel is what counts in accelerating this
mass or propelling it up the hill. That torque is given by the product of the
engine torque and the multiplication ratio of the gear ratios and final drive
ratio. If the maximum torque output of the engine is the same at 3400 rpm and
4500 rpm, the guy running 4500 rpm at 60 in his vanagon will have 30% more
torque available than is 3400 rpm counterpoint. This can be expended for
acceleration (pass the Miata), used to defeat the hurricane winds you are
driving into, or is resolved in a smaller throttle opening (less fuel
consumed, hint). This discussion begs the issue of frictional driveline
losses which should be on the order of the variance of the reported high MPG
event vs. the actual value found at the pump!
So: the only loss in sticking with the standard Vanagon diesel transmission
in a CA smog approved swap is the noise. Insulate, and muffle, add a 600 watt
stereo and a little Mozart and all is fine.
Swap issues:
1. I-4 swaps are many. Quality varies, but all have a certain factory
heritage. They are different in the implementing details and are therefore
mostly experimental, just as is Robert's. The major issue in the conversion
stems from the engine choice. Personally I find the installation of a CIS
engine to be a neolithic approach. The efficiency of electronic engine
management with an O2 sensor feedback loop is essential. CIS-E would be the
minimum, 8V 1.8L digifant preferred because of the ease of implementation.
Use the dual downdraft A2 manifold. Use the 8V 2.0 L if the cash is
available since it uses Motronic control. For westys this is reasonable with
Marshall's hinge mod to clear the raised deck required by the cross-flow
head. Remember to carefully design the muffler system (or go to any good
muffler shop and ask for the longest path they can readily come up with).
Quite is key. NOTE: in CA, putting in any replacement engine from the same
year or later with original smog equipment is smog legal so long as the
engine was sold in CA.
Now the Tiico engine - SA package is overpriced by about $1K in my opinion,
but offers the possibility of a reproducible commercial package. How
commercially viable will be established by the way in which warranty service
is handled. The first rod through the block from whatever for example. Only
warranty service will show that the importer has the capital to mass market,
even on this small scale. The other issue will be parts support to mechanics
doing the install. Business questions. The only other issue is the lack of O2
sensor feedback. Until this is resolved you couldn't give me one. Living in
CA, I'd be busted fer sure. Unfortunately, after one is added, the importer
would have to go through a CARB exemption or certification process. Expensive
and time consuming. Not doing this will blow off the strong smog control
states. Recall, the standards you must meet in an engine swap (CA) are those
in effect for the engine at its date of manufacture. So 82 vanagon with 81 I4
engine - illegal. 82 vanagon with 84 I4 engine must meet 84 smog standards.
82 vanagon with new Tiico/SA engine must meet 2000 emission standards. Ouch.
As to the Subaru swap, the big advantage is flat and Japanese. I'd go for the
latest 6 cylinder version I could find. Here in LA you can find
direct-from-Japan importers of used Japanese engines. They take orders.
Example: Subaru large displacement 6 with electronics - $600. Wait 6 weeks.
Accept or reject. Then rebuild and go. Use the diesel trans to get the torque
curve in the right range and go. No superstitious German brand loyalty here.
I also like Sapporo draft. Good engines (late years) good control technology,
but experimental and dependent on Kennedy's adapters, CARB, etc, etc.
Summary on swaps:
1. Modest money, minimum risk, rebuild the WBX with a good engine rebuilder.
2. Eyes open, more money, modest risk, go Tiico.
3. Less money, more risk, more time, any of the DIY I-4 conversions.
4. Similar money to 2, more risk, Smog viable, long term stable, go Subaru
5. Much more money, much more risk, go evangelical with Lilley WBX approach
looking for enhanced reliability.
6. Rapid boat to hell, install 5 cylinder Audi turbo with Vanagon project.
Me: I'm making a CAD model of the 2.8 L Porsche 944 turbo mounted to the
Tiptronic automatic. Just have to close a Florida land deal.
Sorry for the length. Just got overwrought. Evidence: two erroneous partial
posts from hitting the enter key on my powerbook.
Frank Grunthaner
|