[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [New Search]

Re: Thermostat/Oil Cooler (long... but maybe useful...)


Hey Jim- 

> > Knowing the type of "testing" Gene Berg typically used, I can just about 
> > guarantee that those numbers hold no significance.  They were probably just 
> > pulled out of someone's.......... 
> 
> At least Gene wrote up his testing (in his own, very unique, style.) 
> You've offered us a lot of rather unique thinking of your own, but 
> backed it up with very little. 

I'm just starting out... so far, I have some articles on ignition timing out 
there and, as soon as VWT gets their act together, I'll be sending off a set of 
articles on A/C systems for the aircooled VW. 

The problem I have with Gene Berg is this: too much preaching, not enough 
engineering.  He came about at the right time and spoke loudly to kids that 
didn't know anything about how to hot rod these things.  Some things he said 
were quite accurate, but many are quite wrong.  I'm an engineer, but have no 
desire to preach.  I'll only talk about what I actually know and shutup 
otherwise :-)  It may not make as many people take interest in hi-po products 
that I produce for a lot of money that still break, but oh well.  Hell, I've 
got my hands tied up trying to produce ONE product for the VW aftermarket.  
(I'm finishing development on a piston/cylinder/stud set for the T4 at a 
101.6mm bore - the cylinders are nikasil-plated aluminum and the studs prevent 
compression leaks :-) 


> Can you offer anything that can back 
> up some of your claims? 

I learned all about the VW cooling systems by taking them apart and analyzing 
them myself.  The reason that I don't know about specific 25-36hp details 
(although from what little I've seen I'm *pretty* sure they're all very similar 
in design to the non-fresh air 40hp) is because I haven't taken them apart 
myself :-).  As for the 356 one, I've just seen them and read about them - they 
look just like the non-fresh air T1 style.  I've never taken apart a 911 
cooling system myself, but know their basic design second-hand from other 
engineers I know who like to toy with these things. 

> While I can't compete with you on your 
> apparent knowledge of the various VW and Porsche cooling 
> systems, I'm skeptical about SOME of your claims, particularly 
> that the air flow to the type 4 oil cooler is shut off as the engine 
> warms up. I'll have to look thru my manuals to see if I can find any 
> evidence of that. 

Forget the manuals, take apart an engine.  On that specific point regarding the 
T4, I am 100% absolutely positive that I'm correct.  I can describe the 
differences between the T3 (with which everyone here is pretty familiar) and T4 
setups it in more detail if you like: 

First off, one correction to what I stated earlier: all T4 used the 80-85C (or 
is it 85-90C?  I really don't remember offhand... I'll check next time I'm near 
my parts and I'll check one).  Until 1974, the T1 used 65-70C, but afterward 
they went to 80-85C.  I *think* that the T3 came with the 65-70C variety from 
the factory, but on this point I could be wrong. 

Anyway, the bellows thermostat for the T3 and T4 are of identical dimensions 
except that the T3 is vertical and is threaded on top whereas the the T4 is 
mostly horizontal (slightly angled) with a cable coming out of the top instead 
of threads.  This cable goes backward (toward the shroud) and turns to become 
vertical by way of a ~2" wheel bolted to the case between the shroud and the #2 
cylinder.  This is how the thermostat expansion motion is transmitted to the 
flaps, as opposed to the solid linkages attached to the right-hand head on the 
T3 engine. 

The T3 has two similar flaps: one on each side of the engine.  They pivot on a 
rod attached to the tops of the flaps.  When nearly vertical, they almost block 
the air (the hole in the right-hand flaps lets a bit go by... whether or not 
this is by design to allow a little air to facilitate the thermostat or by 
economics to make the linkage easier to attach is debatable).  As the bellows 
thermostat expands, the solid linkage pulls the right-hand flap toward the 
front (i.e. toward the cylinders) of the engine, which then turns the rod 
between the two flaps and moves the left-hand flap in a identical fashion. 

The T4 has two dissimilar flaps.  They are still attached by a rod similar in 
design to the T3 rod, but instead of the fancy (and hard to adjust!) linkage 
like the T3, the simple cable connects to its own small arm.  The adjustment of 
the cable on this small arm is easily adjustable from the engine compartment 
without removing much at all (maybe a hose at most is in the way... clearly an 
upgrade from the T3 :-).  The rod's position is spring-loaded into an "open" 
position and is pulled back by a cold thermostat, kinda like the rest of the 
flap-based cooling systems.  The right-hand flap moves just like the right-hand 
flap of the T3 engine: hinged at the top and it moves toward the front of the 
engine to open (although it doesn't have any holes in it - the only hole is 
designed to be plugged by a rubber stopper, hinting that the T3 hole is more 
likely by economics and not design :-).  The left-hand flap is a hole 'nuther 
ball game.  First off, the oil cooler is located between the fan shroud and the 
#4 cylinder horizontally with the mounting surface on a vertical plane, NOT 
above the #3 and #4 cylinders horizontally with the mounting surface on a 
horizontal plane like the T3.  The left flap is hinged at the bottom, but is 
not hinged on the rod.  The rod connects via an arm on the end of the rod to 
the flap to move it dissimilarly from the right-hand side.  When mostly 
vertical (again, actually on an angle), it blocks the air.  Then, as it opens, 
the flap again pivots toward the front of the engine.  Partially open, it lets 
a bit of air into the ductwork, allowing air to go to the cyls/heads and the 
oil cooler.  When fully open, the left-hand flap is nearly horizontal and 
completely blocks the oil cooler cooling air inlet.  The flap and the ductwork 
around the oil cooler fit together like puzzle pieces, blocking it completely. 


> > And, this makes sense - the clearance between the pistons and cylinders is 
> > larger when cold than hot.  [BTW, this is one downfall of forged (Aluminum 
> > 2618... or is it 2816?  I don't remember...) pistons as opposed to the new, 
> > cast hypereutectic (Aluminum 390) pistons - the new ones run with a 
> > considerably smaller clearance due to lower thermal expansion.] 
> 
> If you actually check your numbers I think you'll find that the actual 
> coefficients of expansion of different alum alloys don't differ by any 
> significant amount. Of course we can debate what constitutes a 
> "significant" difference, but we have to agree that the difference 
> between iron alloys and alum alloys is something like a factor of 3. 
> So I'm gonna claim tha a 10% difference in alum alloys is 
> insignificant; calculate what difference this will make in cylinder 
> wall clearance and we'll see that we can't even measure it. 

Sorry Jim, but that is entirely untrue.  The big difference is that 390 
Aluminum is a hypereutectic alloy.  I'm not sure what you know about material 
science, but please forgive me (and skip down a bit...) if I insult your 
intelligence with the following explanation.  The eutectic composition of an 
alloy of two materials is defined by the point when as much of the minor 
constituent is totally dissolved in the major constituent as possible.  An 
analog is salt in water.  When you put in a little bit of salt, it gets 
dissolved.  That is like a hypoeutectic alloy.  When you put in lots of salt, 
pretty soon it can't all dissolve and some falls to the bottom of the glass as 
salt crystals.  That is like a hypereutectic alloy.  The point when the maximum 
amount of salt as possible was in there and fully dissolved is like the 
eutectic composition.  In the case of aluminum, many times the main alloying 
agent is silicon.  It's the analog of carbon in iron for steel. 

Generally, the eutectic composition of silicon in aluminum (it depends slightly 
on other minor constitents, but not much) is about 12% by weight.  So, if you 
have 12% of silicon in your alloy, it's all dissolved in the aluminum matrix.  
However, 390 aluminum has 16% silicon.  So, primary silicon crystals appear in 
the material matrix.  This DRASTICALLY changes the properties of the metal.  It 
becomes much stronger and resilient to wear, and pretty brittle (<1% elongation 
at room temperature).  Additionally, its thermal expansion is greatly reduced. 

Modern technology is replacing the old technology forged aluminum 2618 pistons 
with cast aluminum 390 pistons both at racing and OEM levels.  First of all, 
the tooling used for casting 390 along with the different properties allows 
less metal to be used than for forging 2618 without any degredation in 
strength, so the pistons weigh less.  Second of all, the reduced thermal 
expansion means that there can be a smaller piston to cylinder clearance, 
resulting in less piston and ring wear when cold (less banging around).  
Groovy, eh? 


> > And, the additional function of the T4 makes sense - if the heads get too 
> > hot, the heads will crack!  This is contrary to what you said that Berg 
> > preached - heats crack when they are too HOT, NOT when too COLD! 
> 
> Here I'm confused. I thought there was agreement among all of us, 
> Berg included, that too hot was bad. 

I thought that the post I responded to said the opposite... I thought it said 
that too long a warmup period made cylinders crack more easily.  Perhaps it was 
a misunderstanding my either myself (sorry!) or the poster, I dunno. 


> > I mean, c'mon, why did they continue to use the magnesium-alloy 
> > case even when they knew that they didn't last?  Sim ple - getting 
> > magnesium from sea water was dirt cheap! 
> 
> This surprised me, but I have verified that this is basically the way 
> that magnesium is produced in the US. We should not confuse this 
> to mean that it is cheap. It's an electrolytic process, much like the 
> production of aluminum and uses huge amounts of electricity. I 
> think you will find that magnesium is more expensive than 
> aluminum and that aluminum is more expensive than iron or steel. 

Per pound, that pricing is true.  Then again, the magnesium case weighs much 
less. 


> So perhaps we should ask ourselves why they choose such an 
> expensive material. 

Interestingly enough, in the very beginning (1930s), one of the reasons that 
the engine and transaxle cases were magnesium alloys was political.  The KDF 
was the pride of Nazi technology of the era and they wanted to boast it and use 
it as a propaganda piece to glorify Nazi Germany.  The hubcaps bore Nazi 
symbols.  Heck, even the name was an acronym for a propaganda phrase.  Anyway, 
getting metals from sea water was new technology during this time and Germany 
was one of the countries that really pushed it.  Hitler claimed that metals 
from magnesium to gold could be extracted from sea water in pretty good 
quantities.  While the gold claim brought about nothing more than the Nazis 
building fake gold "factories" to interest foreigners and investors (that, for 
the most part, believed him!), the magnesium claim was true.  The KDF was to 
showcase this new Nazi technology... :-) 

Why they stuck with it could be due to many reasons.  One is simple: the 
equipment to extract the metal from sea water is already made and the tooling 
was already set up.  Why change?  Bauxite is primarily in places like 
Australia, Russia, Brazil, Guinea, and Jamaica.  Especially with large amounts 
in Brazil, one must wonder why they didn't convert to aluminum eventually.  
Well, the reason is simple: the vast majority of the equipment to actually 
refine bauxite into aluminum is in the US :-) 

In the early 1960's when they went to the 40hp-style case, they probably kept 
the same metal because a cost-benefit analysis showed that it was cheapest to 
keep using it.  Why establish new tooling?  Plus, although the metal is pricier 
per pound, it also weighs much less.  I have a T1 and T4 case around here... 
perhaps I should weigh them for fun to see the difference.  But, although 
weight might have been a concern in the beginning, I doubt they cared much 
about adding 20lb to the bug once they started adding safety features like 
heavy bumpers and supports and the like... 

But all that is speculation.  I am curious to know exactly what their reasoning 
is... 

Take care, 
Shad 

> 
> - 
> ******************************* 
> Jim Adney, jadney@vwtype3.org 
> Madison, Wisconsin, USA 
> ******************************* 
> 
> ------------------------------------------------------------------- 
> List info at http://www.vwtype3.org/list or mailto:help@vwtype3.org 
> 
> 


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [New Search]