While working with the modern fuel system, we attempted to simulate 28 fuel nozzles that spray in the combustion chamber, by using Valves. Simplified description of the system is as follows:
- Big fuel line capable of taking the entire fuel flow (fuel flow per 1 psi set correctly)
- Line splits into 28 smaller lines, each capable of taking 1/28th of the total flow per 1 psi
- Each of the 28 lines has a valve that simulates a nozzle
- After the 28 valves, the lines meet again into one big line that can take the entire fuel flow. That line connects to the engine.
The first problem is that we couldn’t connect the 28 lines directly to the engine. Instead we had to merge them into one, but that is not serious. The big issue appears when we started closing fuel valves.
We tried closing half the valves and expected that the engine would only receive half the fuel flow, as the manifold pipes were thinner and the flow per 1psi was reduced to 1/28th per pipe. Yet the engine still gets all of the fuel it wants.
Any suggestions on how to fix it?
The tests conducted during this approach were for starters taking the incoming line containing the full volume of the fuel flow and leading that to 28 lines dividing the available volume and fuel flow at 1 psi into 28 the result was that the full volume of the fuel flow still only passes through the 1st line.
Second experiment was to “shrink” the volume and fuel flow at 1 psi of the 1rst line (the one out of the 28 that the fuel only passes through) in dimensions so small that realistically the fuel could not pass through it and would be forced to at least go to the 2nd of the 28 lines. The result of this was the same with the full amount of fuel only passing through the 1rst line
I conducted a similar experiment, but looking for a drop in fuel flow due to pressure from pumps missing, which did not occur upon setting realistic values. Only after setting the fuel flow at 1 psi to ridiculously tiny amount I saw some changes. However, even after that behaviour of the fuelsystem never looked right, with heavy surges and drops in RPM occuring. Now we’re just shutting flow off completely when the pump isn’t running.