Walker’s World May 04

Workshop Web: Sequential testing, compression ratios, pressure regulator mods.

Sorry for the delay in publishing the latest Workshop editorial, - pressure of work!

You will not be too surprised to hear that not everything goes smoothly in the Emerald workshop. Take the Metro we tried to map this month. It came in under its own steam and that’s always a good start. In theory all we had to do was strap it down and map the fuel and ignition. With hundreds of K-Series engines under my belt that isn’t normally much of a problem.
I started by checking the cam timing but soon found a problem with the cam belt tension. Basically it was too slack and there was no adjustment left on the pulley adjusting slot. I’ve seen this before on heads that are heavily skimmed. In one extreme case I had to remove the locking bolt and cut the slot right out to allow the final tension to be reached.
On this car even that trick did not work – I had to cut a big piece off the top of the adjuster to stop it lifting so high that it ran into the cam pulley! Having sorted the cam belt I set to on mapping but soon ran into trouble. The car would not take any sensible amount of ignition advance without pinking. I checked the ECU reference but everything was set up okay. A compression check was called for. My gauge isn’t wonderful and I couldn’t warranty the readings but I do know what it normally shows on fast road or race engines.
Cylinder number one was checked first, ECU disconnected, all plugs removed and throttle wide open: Wow! - The gauge reached previously unexplored territory. The most I have seen before is 240psi but this Metro sent the needle spinning past the 275psi mark on the first few pumps. I reckon it had enough compression to run on diesel fuel.
The owner had collected the parts for the engine over a period of three years and was absolutely gutted that he had to take the head off and lower the compression before the mapping could be completed.
The moral of the story is that when you build an engine you have to measure things: like combustion chamber volume. Take nothing for granted.

Pressure test

Yet another K-Series was giving me a problem. This one was making too much power for the available injector flow. It was marginal and by checking the injectors on the ASNU rig I knew that an increase of fuel pressure from the stock 3 bar to 4 bar would give us enough fuel flow to sort it all out. But how do you do that without buying an adjustable regulator? We needed an instant fix.
Sitting in the corner of the workshop is a Caterham SLR that we are storing for a friend. This runs a Caterham 4 bar regulator as stock. I decided to nick the regulator as a temporary fix for the customer’s car and replace it later with a new regulator from Caterham – or wherever I cold get it at the best price. All I had to do was note the part number of both regulators (3 bar and 4 bar) and sort it all out later on.
The fly in the ointment was that both regulators had the same part number. A quick look revealed the trick – the SLR regulator had been squeezed up to compress the spring and increase the pressure. Obviously this is not an exact science so the SLR got its own regulator back and I looked at tweaking the Elise regulator that I was working on.
I had to turn up a little tool which could hold the regulator and by applying compressed air I could read off the pressure. Two sockets and a vice supplied the compression and it only took a few tries before I had a 4 bar regulator. I’ve done a couple of K-Series regulators since and I suspect this is going to be a pretty standard mod for K-Series engines when they come in for mapping.

Sequential injection

This is strictly a development project at the moment but will be available later this year. We have done a couple of engines on fully sequential injection: a Rover V8 and a VVC Rover K. This is my first experience of sequential injection and I wanted to know if the stories I had been given were correct. Basically this was that sequential injection was all about emissions and fuel consumption and not worth bothering with on a competition engine.
With batch fired injectors some fuel always hangs around in the inlet port waiting for the valve to open. With a sequential system the fuel is only injected during the valve open period and also timed so that none of the fuel follows any air escaping out the exhaust port during the overlap period (all valves open around TDC at the top of the exhaust stroke). That isn’t as simple as it sounds.
With a soft, or standard cam, the valve open period is relatively short so you do not have very long to get the fuel injected. When the fuel demand stretches the injection period past the valve open period you end up with the same result as batch firing: fuel hanging around the inlet port waiting for the valve to open again. This means you need a lot of fuel delivery in a short time period if you want to operate truly sequentially.
You can achieve this with bigger flowing injectors, higher fuel pressure, or a combination of both. The down side is that on idle and light throttle the injector pulse gets very short and it’s more difficult to control the injector accurately. In brief: big injectors are great for big fuel flow and bad for small fuel requirements.
The picture gets better when you run bigger cams with longer duration. Having the valve open longer gives you more time to get the fuel into the cylinder before the valve shuts again. The down side is that bigger cams need more overlap in order to work; therefore on idle and light throttle you need good control of the fuel timing in order to inject at the right moment to avoid spilling your fuel out of the exhaust. In theory you have to match your flow rate to the cam duration, not just the overall fuelling requirement. The general consensus in the aftermarket is that none of this added complication is worth the bother on a tuned engine.

Practical tests

We carried out some testing with a development Emerald ECU to see if sequential was worth the effort. The first engine was a TT Supercharged K-Series from a 111S. Given the added complication of altering the cam duration from the lap top it might not have been the best starting place!
Our initial impressions are that the engine feels no different from a batch fired one. However it looks like sequential allows you to get away with slightly weaker mixtures without misfiring. Altering the time of the injection period makes a big difference to fuel consumption. Having set up the fuelling to 14.2 to 1 AFR we then found that changing the injection timing made the mixture richer by one whole ratio: to 13.2 to 1. You could then back the fuelling off to 14.2 and have the same power and mixture ratio (lighter throttle setting in this instance) as before but using less fuel.

Emissions

We do not have a full emissions test facility here (who does?) but we do have an MoT gas analyser. This car was running almost no CO but hydrocarbons (HC is unburned fuel) were at 200ppm. You could hear the light misfire on idle which was pumping unburned fuel down the exhaust. By altering the injection timing the misfire disappeared and the HC came down to single figures.
To be fair we have not done a back-to-back test on batch firing and it may be that the badly timed sequential injection was causing the problem in the first place - but it is certainly interesting. The same applies to power output: we need to do back-to-back tests to see if the sequential injection is worth power or not but that will have to wait until we get the engine dyno installed.
As part of our sequential investigation we did a one-off ECU for a distributorless V8 with full sequential injection. We have not come to any conclusion other than it works in terms of fuel economy and emissions but the owner reports that the engine is more “responsive” than when previously batch firing the injectors. Again, the engine Dyno is the place for these investigations and once the Superflow is up and running we will make time for more in depth testing.
Earlier I mentioned the VVC control for the Elise 111S. This will be available with the next software release along with a lot of other added feature like map switching. Expected release date is middle of May.