Throttle response EFI: TBI vs. MPFI on 800 hp BBC

[NewbVetteGuy]

https://www.enginelabs.com/news/dyno-co ... big-block/

I get that throttle body injection will result in more intake charge cooling and a small boost in torque from that, especially if timing is optimized, but the difference between Holley throttle body injection and port injection in this test is so large and they claim a night-and-day difference in throttle response with throttle body injection... WHY?

Better atomization improves throttle response, right?
Is this just better intake cooling plus better atomization for the throttle body injection?



This leads me back to beg the question of whether injecting fuel, or water+meth at the throttle body on a dry intake would restore the power differential between port EFI and throttle body injection. (I have found ZERO tests where this was attempted despite expending as much google-fu as I can muster.)

-Obviously dual injectors or moving the injectors higher in the port increases cooling, so why not go all the way to throttle body / plenum injection for cooling especially with all the modern intake designs that have such even distribution?

-What type of setup would be ideal for throttle body or "plenum injection" as a test? A "9th fuel injector" just after the throttle body? ZEX nitrous style perimeter plate?



Adam

[NewbVetteGuy]

That's not a small torque gain for the throttle body injection....



Adam

[David Redszus]

Throttle response is often a debatable issue with little supporting evidence.

So I pulled up some engine data from a port injected Porsche running at Road America.
Two sets of data are shown. In a 2:30 lap, coming off a corner, the data does not repeat consistently.

Time.Sec...Throttle %.....Lambda........MAP psi
.535.........3.3..............0.76...........-2.2
.742........82.0..............1.01...........-1.6
0.207response time

.417.........8.5...............0.80...........-1.2
.659........81.6...............0.95...........-2.2
.242response time

What we can observe from the graph (not shown) is that any sudden change in throttle
results in some measurable time delay. And that neither Lambda nor MAP transition
smoothly. Instead, they tend to overshoot and undershoot with each step change.

While Lambda values will vacillate between lean and rich, it does take some time for the mixture to
return to a stable Lambda value.

In both of the above cases, the fuel response time was quicker than the MAP response time.
The engine rpm response was slower due to inertia forces.

It is not very difficult to use data loggers to gain an insight into engine reality.
Perhaps we should do a lot more.

[ptuomov]

I think the original article is mostly fake news.

If I wanted a normally aspirated engine with fixed cam timing that cruises well, makes a lot of power, and transitions between the two well, then the best solution is obvious. Independent throttle bodies with throttle plates relatively close to the intake valve with port injection with the appropriate inductor spray pattern. Fueling computed both from the throttle position sensor signal and either the MAF sensor feeding the air box or from a MAP sensor connected to all the Intake ports.

The only reasons why I’d run a carb-looking throttle body injection system on top of a plenum manifold is because of the looks or because of the rules.

[turbo camino]

Simplest way to 'improve throttle response' is to use a lighter return spring. Which is a pretty good indicator that 'throttle response' is something that happens inside your head, not inside your engine.

[Orr89rocz]

Steve morris did a similar test using holley sniper throttlebody and port efi on same intake and used the sniper tb on the port injection test, just disconnected the tbi injectors.

Port made more power everywhere above 4200 to 7500 ish. It was a 733 hp deal. Tbi was down alot in the 5000-6000 range however the peaks were about the same for hp

He even tested a distributor vs coil packs on those same engines and coil packs made more low end power but lost some up top after peak power, likely due to inductive delay or just needed different timing commanded.


I never see many high performance tbi setups. Only port efi. But higher up the runner you place the injectors the more power you make but you will lose low speed control and drivability. Best combo is two injector sets, one in port, second in plenum or in front of throttle really, for wot activation

Theres a good chance more tuning with the mpfi would improve response

[ptuomov]

Simplest way to 'improve throttle response' is to use a lighter return spring. Which is a pretty good indicator that 'throttle response' is something that happens inside your head, not inside your engine.

Or take out a muffler or two...

As was noted earlier, your fuel system can impact throttle response. However, if your lambda stays at the desired range despite the throttle position and position change, the fuel system is doing its job.

Then, with turbo systems you do have some lag, which you can try to beginner to a minimum. Camshaft overlap usually not an issue.

For a normally aspirated high power car with a lot of camshaft overlap the issue is whether the car cruises nicely without surging and bogging at low rpm and low throttle angle. ITBs often do while plenum manifold systems sometimes don’t.