Problems with Modern Fuel?

Last year a student project at Mechanical Aerospace and Civil Engineering Department (MACE) of Manchester University was sponsored to mount an XPAG engine onto a dynamometer and use this to investigate the running problems many owners experience with modern fuels. Unfortunately, timescales meant this work was not completed; however, there was time to run limited tests.

When running the tests, during the part load tests, the relative float height of the suction pistons in the two carburettors were different, indicating they were no longer in balance. What was striking was the degree of out-of-balance depended on which fuel was being used.

A part load test – engine running at 3000 rpm and 3⁄4 throttle using Esso 95 octane.

At the time we had no means of measuring the float height of the suction pistons, so part way through the tests I started taking photographs with the intention of making measurements from these.

In practice this was not possible; however, to my surprise, when you blow up the photographs and digitally enhance them, you can see something very interesting that may explain the problems we are experiencing with modern fuel.

With SU carburettors, the float height of the suction pistons is a measure of the volume of air entering the engine; the front (right hand) carburettor feeding cylinders 1 and 2 and the rear cylinders 3 and 4.

While the carburettors were balanced at full throttle, there is no guarantee they will remain balanced at part throttle. The quality of the XPAG cylinder head and manifold castings is poor with differences in both the size and shape of the cylinder ports. The cylinder head of the engine under test has not been gas flowed nor have the inlet and exhaust manifolds been matched to the cylinder head. As a result, changes in turbulence in the inlet and exhaust tracts between cylinders 1 & 2 and 3 & 4 could alter the gas flow and hence the balance at different throttle settings. What is surprising is that the balance also changed when using the different fuels.

The two photographs, below, were taken with the engine running at ¾ throttle and 3000rpm. The top picture using E10 bought in France, the bottom Esso 95 octane. You can clearly see the height of the piston on the left hand side is higher than that on the right in both cases. This indicates more gas is flowing into cylinders 3 & 4 in comparison with cylinders 1 & 2. However, the piston of the right hand carburettor is noticeably lower in the bottom picture (Esso) than in the top (French E10).

With the same throttle setting and rpm in each case, the most likely explanation is that the increased difference in balance must be due to the fuel. It was the only thing that had changed.

How can this be?

On a standard XPAG engine, the inlet valve starts to open 11o before TDC and the exhaust closes 24o after TDC. For 35o both inlet and exhaust valves are open. This valve overlap allows the rush of exhaust gasses out of the cylinder to start the induction of the next air/petrol charge. Differences in combustion profiles between the cylinders, especially late combustion, will affect the pressure in the cylinder when the inlet valve opens and hence the volume of gas flowing through the carburettor. In the worst case, exhaust gasses could flow back into the inlet manifold. The dispersion of the fuel and combustion process depends heavily on turbulence. On the test engine, it is very likely there is a difference in turbulence between cylinders 1 & 2 and 3 & 4 and hence it is possible that the combustion process and back pressures could also be different between the two pairs of cylinders.

If this explanation is correct, the differences in piston heights is a direct comparison of the average combustion in cylinders 1 & 2 versus 3 & 4 and suggests E10 is burning better than the Esso 95 as the pistons are higher (more air flowing into the engine) and the carburettors are better balanced.

This is where the enhanced photographs tell a very interesting story.

The photo below shows the blown up sections of the rear carburettor from the pictures previously shown. They clearly show the fuel leaving the jet, entering the airstream and passing over the top of the butterfly. The fuel has been artificially coloured red to make it clearer.

Two features are obvious. The first is that the piston is floating higher with the E10 than the Esso indicating a lower back pressure and hence better combustion. Secondly, and more importantly, the dispersion of the E10 into the airstream is considerably better than the Esso.

Even though the atomisation of the E10 is not good, the Esso is considerably worse looking like there is virtually no atomisation or dispersion of the fuel as it leaves the jet.

Poor dispersion of the fuel in the carburettor, will lead to poor mixing and vaporisation in the cylinder which in turn will lead to slow combustion, higher back pressure during the overlap of the inlet and exhaust valves and lower air flow through the carburettor, consistent with the observations. It would also explain why not all cars suffer from the problems to the same degree. Variations in the inlet manifold will produce different levels of turbulence in the same way that cylinders 1 & 2 are affected more than cylinders 3 & 4 in the test engine. It is likely the carburettors performed well at full throttle as there was sufficient air flow through them to better atomise and disperse the fuel. The loss of balance only showed itself on the ½ and ¾ throttle tests.

Is this the problem with modern fuel and is there any other evidence to support this? Previous tests and observations lend support to this hypothesis.

Firstly, the acceleration tests and exhaust temperature tests performed by David Heath and myself suggested the engine needed to be significantly advanced (“Modern Fuel On Trial”). Poor mixing at the time the spark plug fires will slow the initial growth of the flame front requiring the engine to be advanced in order to run properly.

Secondly, my two-tone plugs. After fitting long nosed plugs, I noticed that one side of each plug facing the inlet valve, looked like new. This is caused by the fuel entering the cylinder “washing” the plug clean; another indication of poor dispersion. The following picture shows the brown half of plugs 1 and 3 and the clean half of plugs 2 and 4 counting from the left.

Perhaps this finding is not surprising. Modern fuel has a higher specific gravity than “classic” fuel, it is denser and it consists of a greater fraction of heavy distillates. All these factors make it less likely to break up when it leaves the jet. The challenge now is to confirm these observations and evaluate possible solutions.

Paul Ireland