Leaded Fuel vs. Unleaded
In his "Compleat Idiot" book John Muir admonishes us strongly to NEVER use unleaded fuel in a VW engine. Of course now we can't buy anything BUT unleaded fuel in the U.S. Volkswagen has stated that all of their heads from about 1966 on have been made with hardened valves, seats, and valve guides. For this reason the aircooled VW engine does not need lead (or lead replacement) to protect the engine from Valve Seat Recession (VSR) and so the octane number is the only consideration. If your heads have been rebuilt, then the machinist probably used the good silicon-bronze or aluminum-bronze valve guides. Also, the valves need no rubber valves seals (as they point spring-end-slightly-down to drain excess oil off), and so get plenty of splash oil on the spring end, which is enough for the valve guides.
Unleaded fuel should be just fine for the VW engine. The octane number of unleaded IS sufficient, provided you are using the "standard" VW compression. Rob uses Shell unleaded, without any problems, but the formulation is different between different oil companies, and so it might be worth trying different brand of unleaded, just to test for any differences. Dave in the U.S. found that the cheap "supermarket" brand of fuel (like Woolworths sell here I guess) performed very badly in his VW, but Chevron of the same octane worked fine. The formulation is different between different oil companies, and so it might be worth trying different brand of unleaded, just to test for any differences.
At a compression ratio of 7.5:1 you can use 91 Octane. At 8:1 it needs 93 octane, and at 8.5:1 you need 95 octane - premium unleaded.
Note: Compression ratios for some common engines are as follows: 6.0:1 for the 36hp engines, 6.6:1 for the 40hp engines, 7.3:1 for the 1300s, and 7.5:1 for most 1500 and 1600s sold in Australia.
The Wisdom of Rob Boardman
As the fuel is ignited and the flame front races across the cylinder, it increases the temperature and the pressure of the remaining unburned mixture - the final combustion temperature reaches about 2000 degrees. So the still unburned mixture is put under extreme heat and pressure, and WILL start to decompose into various sub molecular components as the flame front approaches. Some of these easily auto-ignite, and if this happens, you end up with the remaining mixture igniting before the flame front reaches it, resulting in a opposing super-rapid pressure wave in the cylinder. THIS is what produces the hammer blow to the piston instead of the nice steady push, and causes the distinctive harsh clicking sound of detonation.
Various additives interfere with the decomposition process in the still unburned mixture, and allows the flame front to race across the cylinder normally to produce a smooth burning process. TEL (Tetra Ethyl Lead) was the most effective compound found, and had the added advantage that in old cast-iron head engines, it helped protect the valves seats and valve guides from excessive wear - it prolonged cylinder head life.
In the USA, MTBE (Methyl Tertiary-Butyl Ether) and other oxygenates have replaced TEL, and do a similar job of interfering with the production of auto-igniting compounds (enhancing the octane rating), but has other undesirable side effects, and does not provide and valve seat protection for older cast iron head engines.
The only thing which changes the flame speed significantly is the presence of burnt exhaust products from the previous power stroke - a higher concentration of burned gases slows the flame speed, and a low concentration increases the flame speed.
Now look at the engine design. I'll use a 1600cc engine for the numbers here. If the VW engine was 100% efficient in filling a cylinder with fresh mixture, each cylinder would have 396cc of mixture (1584cc). But the cylinder head volume is still 53cc, so even with 100% efficiency, there will still be 53cc of burnt mixture from the previous cycle mixed into the fresh charge, since the piston can't sweep out the head volume of burned gases. This is as fresh as the mixture can get. But the VW engine is only about 85% efficient at best in filling the cylinders (no car engine is 100% unless it's super/turbo charged) so that means instead of 396cc of mixture, the best you can do is the equivalent of about 336cc, still with 53cc of burned mixture. Now throttle the engine.... this will further reduce the amount of fresh mixture, whilst still leaving at least 53cc of burned mixture in there.
See what is happening? At various rpm (different airflow efficiencies) and different throttle settings, the proportion of burned/fresh mixture is changing enormously.
When the flame speed is high, we need less advance (since it will take less time to burn) and when the flame speed is low (a high proportion of burned mixture), the flame speed is low, so we need more advance. The aim is always to have all the fuel burnt at the point where the piston starts it's descent, so a maximum push is applied to the piston.
And when rpm is high we need more advance, and when rpm is low we need less advance.
Fortunately, the vacuum distributor (and the combined centrifugal/advance of 74+ distributors) does a reasonable job of following these two conflicting needs. The vacuum port on solex carburettors is placed just UNDER the main venturi, close to where the throttle plate passes as it opens. So imagine the engine idling. The throttle plate is nearly closed, so there is a low airspeed through the main venturi - not much vacuum there or just under it (above the throttle plate). The vacuum port sees very little vacuum, and the idle advance setting prevails (7.5-10 degrees BTDC on most models).
Now open the throttle a little, so the edge of the throttle plate passes by the vacuum port. This creates a mini-venturi with very high air speed, which creates a lot of vacuum, so you get a shot of advance to help speed up the engine (this effect is entirely missing with the Bosch 009 distributor, which is what causes the "009 flat spot"). Since in a part-throttle condition you still have a high proportion of burned gases for a low flame speed, this high advance also meets the advance condition needed to deal with that too.
Now open the throttle right up. The throttle plate moves away from the vacuum port (no mini venturi) and so the MAIN venturi is providing the vacuum effect, but since the airspeed hasn't yet increased much yet (engine hasn't yet increased rpm), the vacuum signal is lower than part throttle, so the advance is reduced a little. Perfect for a fresher mixture (lower proportion of burned gases with an open throttle remember?). Now the engine rpm starts to catch up with the open throttle, so the airspeed through the main venturi increases, vacuum increases, and the advance increases progressively, which is just what you want for the increasing rpm, since the crankshaft is rotating in less time so you need more advance to get that fresh charge completely burned at the right moment.
So the vacuum distributors allow for high advance at high rpm/open throttle, where rpm is the dominant factor; and also allows high advance at part throttle/medium rpm, where the proportions of burnt/fresh mixture is the predominant consideration.
Therefore, the advance curve has less to do with the leaded/unleaded issue (fuel type); but rather it accounts for induction efficiency and rpm range. If you alter the efficiency of the engine (larger valves, different cam) your ideal advance curve changes too - nothing to do with the fuel you are using.
But use a completely different type of fuel, and your argument holds true - LPGas DOES have a lower flame speed than petrol, and if running LPG only, the spark can be advanced 3-4 degrees (and ideally the advance curve should be steeper too) to account for this. If using dual fuel though, you can only advance about 2 degrees or so, since when on petrol the total advance could be too high. Cars with ECUs are great in this situation - they can have a separate advance curve for LPG mapped into the ECU, and so you get an advance curve matched to either fuel. (I have an old Falcon with dual fuel, and have been running LPG cars for about 8 years now, but since the Falcon doesn't have an ECU, the best I can do is 2 degrees additional advance).
Getting back to the petrol though - Avgas has a very low volatility compared to road fuels (since you don't want the lighter fractions evapourating at low pressure as the aircraft climbs higher), yet similar advance conditions apply. But in the aviation engine, a centrifugal advance system is often used since they operate at relatively constant rpm, quite unlike the car. When the VW engine is used as a generator (the Aus army uses them), a centrifugal distributor is used because it's running at constant rpm. So the engine design and use has more to do with the advance, rather than any variation in the liquid fuel.
* * * *