Dear Paul,

I can't understand this effect :

"The suction piston will drop before the induction stroke starts. When induction begins, both the inertia of this piston and the damper will slow the rate it rises. Under these conditions, the carburettor will deliver a richer mixture"

In my understanding, the piston is actually lower than it sould be, so the annulus around the needle is smaller and the mixture should be weaker.

Can you explain this please ?

Laurent.

Laurent,

I am afraid this is another case of reality not matching expectations. There is a second effect with a lower piston height.

The pressure in the choke is reduced which means more petrol enters the airstream for a given annulus. In practice the lower air pressure wins over the reduction in size of the annulus so the mixture gets richer.

This is better explained on page 99 /100 of the book where it describes the effect of different springs.

Paul

An interesting observation.

My son has a Lotus Europa fitted with a 1,470 cc (1.5 L) Renault A1K I4 engine that has been tuned. It is fitted with two, twin choke Weber carburettors. He recently took it on a rolling road. Look at the full throttle Lamda measurements (the pink line) it matches that predicted for the XPAG. I am not sure what fuel he was using.

The mixture (thick red line) is much richer until the point where maximum torque is reached when it returns to the correct setting (0.9). This is exactly the same effect as was predicted for the XPAG. The main difference is that for the Lotus, these are actual measurements as opposed to calculated measurements reported in the book.

The rolling road data for the Lotus is very interesting for two reasons:

1. This engine has been race-tuned with a gas flowed cylinder head. This reduces turbulence, increasing Cyclic Variability, in turn this makes the enrichment effect worse. This is particularly obvious at 4000 rpm where the mixture is very rich with a lambda of 0.7. It is unlikely this effect is due to the carburettors as they have been tuned.
2. This engine is fitted with fixed jet Weber carburettors. It is easy to understand why back pressure pulses could cause a variable jet carburettor to deliver a richer mixture. It is less obvious what would cause fixed jet carburettors to exhibit the same behaviour. It is possible the back pressure in the auxiliary venturi reduce the effect of the emulsion tube resulting in a richer mixture. What this data shows is that fixed jet carburettors also appear to suffer from the same enrichment problems as variable jet ones.

Some Lotus owners have suggest that a better exhaust manifold, which increases the scavenging effect, will help. Yes it probably will. This will reduce the back pressure pulses from the late combusting cycles.

Another option would be to increase the ignition advance between 3000 and 4000. This would allow more time for the mixture to burn, decreasing the number of late combusting cycles. However, this will cause the early combusting cycles to burn too early potentially resulting in engine damage. If you look at the Ign Advance section of this site, it is interesting to compare the advance curves for the Triumphs compared to the MGs of a similar age.

Those for the Triumphs are typically more advanced than the MGs over the 1000 to 3000 rpm range. The head design of the Triumphs is such that it could reduce turbulence over this rpm range compared to the standard MGB.

Adding acetone also may help. This thread and the article it references show that acetone reduce the levels of CO and unburned hydrocarbons. As was found at Manchester this suggests it may reduce the degree of cyclic variability.

I had a comment from my son:

"Reprofiled ignition curve with more advance where it was rich and flat spot has gone, it is now epic! Runs up from 3k-7k sweet as a sewing machine and much much smoother. It is a peach of an engine."

This demonstrates the enrichment effect was almost certainly caused by back pressure pulses in the inlet manifold caused by cyclic variability. Advancing the ignition meant that fewer cycled occurred "early", reducing the back pressure pulse and effect on the carburetors.

Gas flowing inlet.

A comment regarding gas flowing inlet and negative consequences.

What I have learnt is to shape the inlet smooth and polish to mirror surface. Then you sand blast it to add "roughness" but still maintaining

the smooth shape. A bit like the dimples on a golf ball.

Best of two worlds!

I am afraid sandblasting the inlet track will not increase turbulence. It causes the boundary layer or air to “stick” to the surface. This effectively acts as a lubricant, reducing the turbulence in the air and allowing it to flow more easily.

One classic example are the skins of sharks. Counter to what one may think, they are rough for the same reason.

The discontinuities in un-ported manifolds are far more effective at increasing turbulence and mixing.

It may not increase turbulence which was not the idea in first place, but it will lessen the fuel particles tendency to stick to the wall and form larger drops. That is creating a more uniform mixture. At least that is what I've been told. Also makes fluid dynamic sense.

Yes you are probably correct. As I said the thin boundary layer of static air acts like a "lubricant" with a layer of laminar flow between it and the air entering the engine. This will make it harder for small droplets to reach the wall of the inlet tract and stick. However, I am not sure if it will stop larger droplets reaching the wall.

If these do make contact with the wall, the roughening will increase the surface area and their tendency to stick.

Perhaps a topic for further research