I've been thinking lately about boost control as well as the 4000rpm cliff that our beloved engines suffer from. Â I'm just throwing this out there to get some opinions. Â My logic goes like this...
Torque is directly related to cylinder pressure. Â I believe this is prety much an unarguable fact. Â Following that enters the concept of volumetric efficiency. Â I have read (and it seems reasonable) that the engines VE is directly related to the torque curve. Â Essentially that an engines torque peek will be at the engines VE peek.
Assuming that there is some validity to the first set of assumptions... Â here's another... Â VE is essentially the engines ability to move the air efficiently. Â This is affected by both intake efficiency (abilty to completely fill the cylinders with air and fuel) and exhaust efficiency (ability to expel the exhaust).
Moving on for any that are still following... Â As I understand it, detonation is most likely to occur at the engines VE/Torque peek when cylinder pressures reach there highest point.
So here's my theory. Â If intake efficiency is a contributing factor to the torque fall-off @ 4000rpm because the cylinders are unable to fill effectively, then increasing boost after this point would allow increased VE past this point. Â Of course, if the major problem is on the exhaust/turbine side of the engine, my idea obviously won't help.
My vision of how this could be achieved is through a "smart" boost controler. Â I have seen commercial EBCs that adjust the max boost based on vehicle speed, but none that adjust PSI based on RPM. Â It's a relatively simple task to use a microcontroller programmed to operate a solenoid to control the wastegate actuator and also be able to set different boost levels for various RPMs. Â I am going to build one of these to test in the future, but I am curious what others think of the idea. Â I admit, I may be really stretching some of my assumptions here, but I'm going to build one anyway for fun and to test my theory. Â I expect to add knock sensing also to help me avert disaster and possibly make my EBC as smart as modern ECUs and control boost based completely on learned limits. Â It would also not be much of a stretch to incorporate intake temperature or other sensor data into the equation.
As I said, I will be building a prototype or at least something like it in the coming months, but I would value any feedback from any here that might have some sound ideas, even if they're critical.
Thanks and sorry for the long post
By the way, there is no plan to produce this as a commercial item, so I'm not looking to make money from anybody's input here, just test some new ideas. Â However if it proves to be effective, I will certainly share the design plans. Â I expect that the parts will be ~$100 to ~$150 for a prototype, but if it works, a more economical unit could likely be produced for $100 or less.
TurboBitten