Twin Turbo 928 fixed and back out there terrorizing the streets!
#496
First, my plan was to make a fuel system that I would never have to worry about or tweak because it had enough capacity. (What I actually made was a fuel system that I have had to worry about and tweak all the time, but never mind... ;-) )
Second, and more seriously, it has been my goal from day one to create a fuel system and controller that will dynamically only provide slightly more fuel than what is needed. I believe that system will combine the best features of a traditional return line system with a mechanical regulator and a modern flat-head system with electronic pump control - that is, the best fuel system, like, ever.
In a week or two we'll know whether were there.
#497
A fuel system update:
The Fuelab EFPR is installed. One can just connect/disconnect a wire to have it control the fuel pump or just work as a dumb mechanical regulator.
When the EFPR is not controlling the pump, the regulator holds the pressure steady, so the mechanical regulator components works. The pump sounds loud as before. The car runs the same as with the stock regulator if the adjustable pressure is set to be the same as the stock regulator pressure. This is all as expected.
When the EFPR is controlling the pump, the pressure is also steady. After the start up, the pump can't be heard at idle. As the engine is revved, the fuel pressure stays steady. The pump just runs a little bit faster. Haven't measured fuel temperatures yet, but logic dictates they will be more stable than in a completely stock S4 since the system recirculates less fuel. This is also as expected (or, more accurately, as hoped).
The remaining potential failure points that need to be tested and ruled out are:
(1) When the EFPR made of aluminum heats up in the engine compartment, will thermal expansion of the case alter the mechanical fuel pressure set point? This can be a problem with shiny "billet" regulators.
(2) If one cruises at 5500 rpm with light load, and then floors the throttle, will the pump respond quickly enough to the fuel demand to keep the pressure up? That's not a common operating mode, but I want to make sure that the controller and pump are fast enough to respond to even absurd changes in fuel demand without a flat spot.
In other words, so far the Fuelab products seem to work as advertised. This is rare in the aftermarket performance part industry.
The Fuelab EFPR is installed. One can just connect/disconnect a wire to have it control the fuel pump or just work as a dumb mechanical regulator.
When the EFPR is not controlling the pump, the regulator holds the pressure steady, so the mechanical regulator components works. The pump sounds loud as before. The car runs the same as with the stock regulator if the adjustable pressure is set to be the same as the stock regulator pressure. This is all as expected.
When the EFPR is controlling the pump, the pressure is also steady. After the start up, the pump can't be heard at idle. As the engine is revved, the fuel pressure stays steady. The pump just runs a little bit faster. Haven't measured fuel temperatures yet, but logic dictates they will be more stable than in a completely stock S4 since the system recirculates less fuel. This is also as expected (or, more accurately, as hoped).
The remaining potential failure points that need to be tested and ruled out are:
(1) When the EFPR made of aluminum heats up in the engine compartment, will thermal expansion of the case alter the mechanical fuel pressure set point? This can be a problem with shiny "billet" regulators.
(2) If one cruises at 5500 rpm with light load, and then floors the throttle, will the pump respond quickly enough to the fuel demand to keep the pressure up? That's not a common operating mode, but I want to make sure that the controller and pump are fast enough to respond to even absurd changes in fuel demand without a flat spot.
In other words, so far the Fuelab products seem to work as advertised. This is rare in the aftermarket performance part industry.
Last edited by ptuomov; 01-23-2013 at 08:46 PM.
#499
Today is the day.
John has the car and my permission to blow it up on the dyno. The first serious attempt to destroy the engine will take place today. I think he'll succeed now that the intake tract has been fortified to take over 30 psi. But I've been optimistic in the past, and, despite the high expectations, the car has left the dyno under it's own power.
#502
I hope so.
Some photos below. No engine internals or fluids visible yet.
The most recent situation report from John: "We're at 517 whp now on just under 10psi. That was on many back to back pulls...car was plenty hot. Turbo speed was 102,000 rpm. Exhaust manifold pressure was 19 psig. No pressure in the tailpipe...not even 1 psi."
This is just a warm-up to optimize the fuel and ignition maps for lower loads. I've caned the engine with 20 psi and it's just said "Thank you, Sir, may I have another one, Sir!" One of the advantages of taking it slow while not having to worry about the engine is that one can really develop the knock-constrained optimum tune for every cell very effectively. This is because one can let the tuning algorithm explore the high-knock probability region more freely. Once the map optimization is complete, the wastegate will be bolted shut and then we really see who's the boss.
Some photos below. No engine internals or fluids visible yet.
The most recent situation report from John: "We're at 517 whp now on just under 10psi. That was on many back to back pulls...car was plenty hot. Turbo speed was 102,000 rpm. Exhaust manifold pressure was 19 psig. No pressure in the tailpipe...not even 1 psi."
This is just a warm-up to optimize the fuel and ignition maps for lower loads. I've caned the engine with 20 psi and it's just said "Thank you, Sir, may I have another one, Sir!" One of the advantages of taking it slow while not having to worry about the engine is that one can really develop the knock-constrained optimum tune for every cell very effectively. This is because one can let the tuning algorithm explore the high-knock probability region more freely. Once the map optimization is complete, the wastegate will be bolted shut and then we really see who's the boss.
#503
This is cool!
My clients don't like it when I break their prototypes or production samples. I explain to them the only real way to know is to test to yield.
I suggested to a particular client to hook his product to the hitch of his truck, and give at a good pull. He had a strange look on his face.
My clients don't like it when I break their prototypes or production samples. I explain to them the only real way to know is to test to yield.
I suggested to a particular client to hook his product to the hitch of his truck, and give at a good pull. He had a strange look on his face.
#504
WOW This is going to be really definitive, just how much boost you can pump into a stock 32V 928 engine.
You know what, I reckon it is not just the 928 engine that will be under the pump, I wouldn't be surprised to find that the entire turbo system also reaches beyond it's point of efficiency, ie turbo overspeed, uncontrolled inlet temps etc.
I'm sure you have an idea where that will occur and what sort of numbers to expect should/when it gets there.
I expect that should anything fail within the engine, it will be the head gasket, but under the pressure that it fails at it will take the bore with it, and blow up the radiator !
You know what, I reckon it is not just the 928 engine that will be under the pump, I wouldn't be surprised to find that the entire turbo system also reaches beyond it's point of efficiency, ie turbo overspeed, uncontrolled inlet temps etc.
I'm sure you have an idea where that will occur and what sort of numbers to expect should/when it gets there.
I expect that should anything fail within the engine, it will be the head gasket, but under the pressure that it fails at it will take the bore with it, and blow up the radiator !
#506
Longevity at any power level is the more interesting metric.
Top fuel and funny car dragsters put down what? 6000 hp? But they only run for 8 seconds between rebuilds.
Given the conditons, I will be mildly surprised if the long block fails before stuff attached to it fails. Once in the "real world" though, it'll be different and I too bet on head gaskets.
Regardless, this is an awesome experiment!
Top fuel and funny car dragsters put down what? 6000 hp? But they only run for 8 seconds between rebuilds.
Given the conditons, I will be mildly surprised if the long block fails before stuff attached to it fails. Once in the "real world" though, it'll be different and I too bet on head gaskets.
Regardless, this is an awesome experiment!
#507
You know what, I reckon it is not just the 928 engine that will be under the pump, I wouldn't be surprised to find that the entire turbo system also reaches beyond it's point of efficiency, ie turbo overspeed, uncontrolled inlet temps etc.
I'm sure you have an idea where that will occur and what sort of numbers to expect should/when it gets there.
#510
Not really, as he is not optimized for what the stated test - which is pure air flow, fuel, and the mechanical strength of the engine package. Whether or not the engine will blow up when he detonates it with a bad spark or a lean run will prove nothing other than that means nothing. Detonation will blow up an engine at low power as well. A true test is how much power will it hold without detonating.