Apologies for the delay in producing this newsletter! With examinations and uncertainty in what was happening at Manchester, I was waiting for some concrete news before writing.
By the end of their project, the students had just managed to get the engine running and two of them, Victor and Ahmed, were offered one month internships to continue the testing after their examinations. In the event, broken crankshaft bolts and a limit on technician time resulted in only two hours of tests. However, these were sufficient to give an insight into our problems. Since then, I have been negotiating with the MACE department at Manchester to agree a way forward.
The tests were run by Victor, one of the students, Prof. Yates, Dave Houghton (Technician) and me. The results surprised us all and suggest a more detailed phase of testing is required to answer the questions and find solutions.
We tested three fuels: Avgas, the 1960 – 70’s equivalent as a control fuel, Esso 95 octane from a local garage and Supermarket E10 obtained in France. Running over a range of engine revs for each fuel, we measured torque, air fuel ratio, exhaust temperature and investigated changes in ignition timing. Tests were performed with a wide open throttle and part throttle producing about 80% of the maximum torque.
On the full throttle tests not only did the engine run almost exactly the same on the three fuels, we found the standard advance curve produced by the Distributor Doctor’s distributor produced the optimum power. What differences we did see could easily be explained by the physical differences in the fuel and the “enleanment” effect of ethanol in the E10.
The part throttle tests gave a different story, which is not obvious when you look at the torque curves. To achieve these values we had to WEAKEN the mixture of the Avgas by 2 flats on the carburettor relative to the full throttle setting and RICHEN it by 6 – 7 flats from the Avgas setting for the Esso and E10.
In other words, fuels that behaved identically on full throttle ran differently on part throttle – modern fuels running significantly weaker than Avgas, our control fuel.
This effect is very worrying. Have your car tuned on a rolling road, at full throttle, then drive it away on part throttle and it will run weak and hot! The problems we are experiencing with modern fuel are real and elusive.
There are two possible explanations, evaporation in the carburettor, changing the mixture or differences in combustion. Unfortunately, the measurements we took were insufficient to give the answers, hence the need for further test.
Paul Ireland
Ed’s Note: Very worrying indeed! In discussion with Paul I have said that lack of money should not be a bar to further research. Further research is potentially costly but the funding must be found. Whilst I have pledged further financial support on behalf of readers there are other funding avenues which need to be explored. You can be assured that Paul is ‘on the case!’
The variance is possibly due to the inefficiency of the SU carburetor at anything over 3/4 throttle as the fuel tends to stay in droplet form rather than atomizing. That’s why we typically see SUs sized for larger airflow than needed. Avgas is an interesting control fuel, but it has protection in it to maintain its vapor pressure characteristics at altitude.
The leaning effect can be measured by knowing how many threads per inch are in the SU H jet adjusting nut- 26 per inch, and then knowing which station you have now jumped up to. This means one flat is .0064 of adjustment in the jet height, two flats are .0128 and six flats are .0385 inches.
Normal idle tuning is usually started with the jet .070 below the bridge, and ends up at about .063 below the bridge, or one flat up from the initial setting.
These numbers can be used to determine a new needle profile required by examining the profile of the existing needle and comparing to other needles available.
While not a foolproof method, and one that is dependent on the type of fuel for the color reading, shutting down the engine at speed and putting it in neutral (or clutch in) and then examining the spark plugs can tell you quite a bit about whether you are running rich or lean compared to idle mixture setting.
David,
Thanks for the comments. If your explanation is totally correct, I would have expected to see the leaning effect with the modern fuels occuring at both 3/4 AND Full throttle. However, your comments about “fuel tends to stay in droplet form rather than atomizing” certainly aligns with my thoughts.
My personal view is that modern fuels are not atomising properly producing larger droplets that take longer to evaporate in the cylinder. Hence the reason that advancing the engine appears to mittigate the problem. I think we do not see the leaning problem at full throttle because the compressive heating of the charge is greater than at part throttle. Your comments about the carburettor suggest that at low throttle settings (less than 3/4) the carburettors should produce better atomised fuel that will evaporate better and engine will run with the correct mixture. As I said in the report, we need more tests.
Unfortunately, a different needle will not resolve the problem. It is the airflow through the carburettor that determines the needle height in the jet, not the throttle setting. Crudely, one could imagine that the airflow into an engine running at 2000rpm – full throttle is the same as that for the same engine running at 3000rpm – 2/3 throttle. Giving identical needle positions in each case. However, the full throttle setting runs fine, the 2/3 throttle setting runs weak.
Good point about the mass flow being constant at the two conditions, however I would not expect the first condition to be one to design to because it could easily be mitigated by changing down in gear. I am rarely below 2800 RPM on my car.
What advance are you running? I use a modified distributor with a more aggressive advance curve, and use either a static timing of 8 degrees BTDC, or 12 degrees BTDC at 1000 RPM. Jeff Schlemmer set it up for me. Of course I run points on a high lift distributor cam.
The point I made about the mass flow being the same was simply to point out the position on the needle would be the same between the two conditions. Therefore if the carburettor was delivering the correct air / fuel mixture at full load at this point, it should also be delivering the correct air / fuel mixture at the lower throttle setting. In the latter case we measured the engine as running weak. The only explanations for this at the moment are either the correct air / fuel mixture is being delivered to the cylinder but it is not burning correctly, or something is affecting the carburation such as evaporation of the fuel in the jet due to higher heat transfer from the exhaust.
Your timing, i.e. a static advance of 12 – 14 degrees at tickover (800 rpm) was the advance I used prior to these tests. Interestingly, when we first ran the engine it had been set at 30 degrees advanced static and it appeared to run OK with no signs of pinking, certainly up to 2500 rpm. Originally, my advance curve was very aggressive, with the springs that were in the car when I bought it, however, I replaced these with a new set that gave a flatter curve and found the engine ran better, particularly above 3500 rpm.
On the test run, we used the “standard” advance curve recommended by the Distributor Doctor that basically reaches 30 degrees at 2000 rpm. The distributor was set at 0 degrees static.
I have attached a poor graph of these three advance curves – click to view it full size:
The point I made about the mass flow being the same was simply to point out the position on the needle would be the same between the two conditions. Therefore if the carburettor was delivering the correct air / fuel mixture at full load at this point, it should also be delivering the correct air / fuel mixture at the lower throttle setting. In the latter case we measured the engine as running weak. The only explanations for this at the moment are either the correct air / fuel mixture is being delivered to the cylinder but it is not burning correctly, or something is affecting the carburation such as evaporation of the fuel in the jet due to higher heat transfer from the exhaust.
Your timing, i.e. a static advance of 12 – 14 degrees at tickover (800 rpm) was the advance I used prior to these tests. Interestingly, when we first ran the engine it had been set at 30 degrees advanced static and it appeared to run OK with no signs of pinking, certainly up to 2500 rpm. Originally, my advance curve was very aggressive, with the springs that were in the car when I bought it, however, I replaced these with a new set that gave a flatter curve and found the engine ran better, particularly above 3500 rpm.
On the test run, we used the “standard” advance curve recommended by the Distributor Doctor that basically reaches 30 degrees at 2000 rpm. The distributor was set at 0 degrees static.
I have attached a poor graph of these three advance curves – click to view it full size: