Injector selection information and handy calculator
01-23-2004, 08:57 PM
#1
Registered User
Thread Starter
iTrader: (7)
Join Date: Dec 2002
Location: Sacramento CA
Posts: 10,029
Car Info: 02 Impreza WRX sedan
Injector selection information and handy calculator
RC Engineering, known for its legendary Sorcerer drag bike of the '70s, has broadened its services over the years to including custom fuel injectors and blueprinting of your existing injectors.
RC Engineering has a page that contains a formula to calculate the proper fuel injector flow rate for any engine output, although if you don't have access to an engine dyno, some of the variables may have to be estimated. There is also a handy online calculator.
--
0==WW==0
"…axles of evil…" - george w. bush
RC Engineering has a page that contains a formula to calculate the proper fuel injector flow rate for any engine output, although if you don't have access to an engine dyno, some of the variables may have to be estimated. There is also a handy online calculator.
--
0==WW==0
"…axles of evil…" - george w. bush
01-25-2004, 03:03 PM
#4
Registered User
Thread Starter
iTrader: (7)
Join Date: Dec 2002
Location: Sacramento CA
Posts: 10,029
Car Info: 02 Impreza WRX sedan
I've been online checking out businesses that recondition/blueprint OEM injectors because I know that most factory injectors can vary by several percent in flow rate, even though they have the same rating. After many miles of use, the flow difference worsens as heat soak builds up varnish on the injector seat.
I learned about OEM injector variance many years ago when Mopar offered replacement injectors for my old GLHS Omni. The flow rating was the same as what was already on the engine, but they were sold as a matched set of injectors, meaning the variance was less than 3%, IIRC. I still have these injectors!
If the flow variance is too great between any of the injectors, the ECU will sense only what the 0² sensor measures downstream of all four exhaust ports on a subie. What this means is, if one injector is clogged with varnish deposits and fuel flow is decreased by 10%, the ECU will see a lean condition and enrichen the mixture by 10% using the other three injectors. Richening the mixture will not truly compensate for the one lean cylinder, and in fact may make matters worse by causing those other three cylinders to run too rich, killing power and increasing emissions.
If the flow variance of OEM WRX and STi injectors are significant, making sure they all perform equally well will probably show measurable gains if back-to-back comparisons can be made following a good dyno tune, before and after the injector swap.
EDIT: I just read the About page on RC Engineering's site - it's a great read! Guys like Russ Collins don't limit themselves to what everyone else is doing. The quest for more horsepower means trying something different if you want that last attainable 1% advantage over your competitors.
--
0==WW==0
"…axles of evil…" - george w. bush
I learned about OEM injector variance many years ago when Mopar offered replacement injectors for my old GLHS Omni. The flow rating was the same as what was already on the engine, but they were sold as a matched set of injectors, meaning the variance was less than 3%, IIRC. I still have these injectors!
If the flow variance is too great between any of the injectors, the ECU will sense only what the 0² sensor measures downstream of all four exhaust ports on a subie. What this means is, if one injector is clogged with varnish deposits and fuel flow is decreased by 10%, the ECU will see a lean condition and enrichen the mixture by 10% using the other three injectors. Richening the mixture will not truly compensate for the one lean cylinder, and in fact may make matters worse by causing those other three cylinders to run too rich, killing power and increasing emissions.
If the flow variance of OEM WRX and STi injectors are significant, making sure they all perform equally well will probably show measurable gains if back-to-back comparisons can be made following a good dyno tune, before and after the injector swap.
EDIT: I just read the About page on RC Engineering's site - it's a great read! Guys like Russ Collins don't limit themselves to what everyone else is doing. The quest for more horsepower means trying something different if you want that last attainable 1% advantage over your competitors.
--
0==WW==0
"…axles of evil…" - george w. bush
Last edited by Wingless Wonder; 01-25-2004 at 03:24 PM.
02-23-2004, 03:44 PM
#5
VIP Member
iTrader: (2)
Join Date: Apr 2003
Location: pompous douchebag
Posts: 9,351
Car Info: $200,000 sports car
so, quick question on the calculators at the bottom of the page. what value do you use for system fuel pressure @ fuel rail do we use on a wrx? if you use 43psi, BSFC of .60, and duty cycle of .8, then it looks like STI pinks are barely able to support 280 hp at the crank.
04-12-2004, 01:31 PM
#6
Registered User
Join Date: Jun 2003
Location: New York
Posts: 173
Car Info: 2002 WRX wagon
I sent RCeng a note asking about costs to spec injectors for my 02 WRX and got an almost immediate reply. I also asked for advice on which brands are least fouling: awaiting a response. I am forced to use 93 octane 10% ethanol right now, and I am not too pleased about that. (There is no 91 octane in this area, that I know of, and I don't want to drop to 89 when the book says 91.)
05-02-2006, 12:10 PM
#7
Registered User
Thread Starter
iTrader: (7)
Join Date: Dec 2002
Location: Sacramento CA
Posts: 10,029
Car Info: 02 Impreza WRX sedan
Originally Posted by RussB
if you use 43psi, BSFC of .60, and duty cycle of .8, then it looks like STI pinks are barely able to support 280 hp at the crank.
Letter to Dave Coleman's "Dear Dave" column in the June 2006 issue of Sport Compact Car:
Originally Posted by Terry Culp, Columbus OH
While reading the letter on injector sizing in "Dear Dave" in the Feb. '06 issue, I decided to plug in some numbers and see what crunched out. A stock WRX has 420cc/min injectors and is rated at 227 hp. According to the foumula, a stocker should have 447cc/min injectors (assuming a 0.6 BSFC and 80 percent duty). STI injectors are about 565cc/min and the formula says that they need to be 591cc/min. Am I suffering from a bad case of "rectal/cranial inversion," or is your formula inaccurate?
Originally Posted by Dave Coleman
Ok, you caught me. That formula comes from RC Engineering and it has been the fallback of lazy keyboard jockeys like myself for years, Instead of straining my brain to really think things through, I just fell back on the old standby. For those of you who don't carry your SCC back issues around at all times like your Binkey, here's the formula again:
((HP x BSFC)/(# of injectors x duty cycle)) x 10.5 = injector flow in cc/min.
The formula is right, but the constants I suggested are extremely conservative.
Russ Collins of RC Engineering suggests using 0.80 for duty cycle because injectors can't be trusted to operate consistently above an 80 percent duty cycle, often sticking open, or "going static" when pushed farther. Shiv Pathak of Vishnu Perfromance has done more factory-ECU Subaru tuning than anyone else in the country, and he claims even the factory calbration on a WRX pushes the injectors over 90 percent duty cycle. Furthermore, he claims he doesn't see the injectors go static until at least 96 percent duty cycle. So who's right?
Both of them.
Duty cycle is not the true measure of an injector's limitation; minimum closed time is. If you look at a one-second interval, an injector that is open for the first 800 milliseconds and closed for the next 200 is running at an 80 percent duty cycle. Another injector that opens for 3.2 milliseconds, then closes for 0.8 milliseconds and repeats that 250 times in a row is running at the same 80 percent duty cycle, but is going through a whole different kind of hell.
It takes a finite amount of time to actually pull the injector pintle closed, and at somewhere around 0.8 milliseconds it simply isn't possible to get it closed and back open again in time. This is the minimum closed time.
Collins comes from the world of motorcycles, where peak torque can come at 15,000 rpm, and batch-fire fuel injection systems that fire the injectors every revolution are not the ancient history that they are in the four-wheel world. 15,000 rpm is one revolution every 4 milliseconds. At an 80% duty cycle, the injector is open for 3.2 of those milliseconds and closed for 0.8. In Russ Collins' world, 80 percent is the limit.
On planet Pathak, peak torque is closer to 6000 rpm. That's one revolution every 10 milliseconds. A modern automotive fuel injection system is sequential, meaning it fires each injector in sync with that cylinder's intake stroke. That means the injector has two revolutions, or 20 milliseconds to do its cycle. The limit of 0.8 milliseconds is only 4 percent of that 20-millisecond window, so the duty cycle can be 96 percent before the injector goes static.
Now, before someone tries to fact check me again, realize that the 0.8 millisecond limit is a pretty rough number. The actual minimum closed time depends on the injector, the fuel pressure, the voltage used, the phase of the moon, and the type of injector driver in the ECU.
((HP x BSFC)/(# of injectors x duty cycle)) x 10.5 = injector flow in cc/min.
The formula is right, but the constants I suggested are extremely conservative.
Russ Collins of RC Engineering suggests using 0.80 for duty cycle because injectors can't be trusted to operate consistently above an 80 percent duty cycle, often sticking open, or "going static" when pushed farther. Shiv Pathak of Vishnu Perfromance has done more factory-ECU Subaru tuning than anyone else in the country, and he claims even the factory calbration on a WRX pushes the injectors over 90 percent duty cycle. Furthermore, he claims he doesn't see the injectors go static until at least 96 percent duty cycle. So who's right?
Both of them.
Duty cycle is not the true measure of an injector's limitation; minimum closed time is. If you look at a one-second interval, an injector that is open for the first 800 milliseconds and closed for the next 200 is running at an 80 percent duty cycle. Another injector that opens for 3.2 milliseconds, then closes for 0.8 milliseconds and repeats that 250 times in a row is running at the same 80 percent duty cycle, but is going through a whole different kind of hell.
It takes a finite amount of time to actually pull the injector pintle closed, and at somewhere around 0.8 milliseconds it simply isn't possible to get it closed and back open again in time. This is the minimum closed time.
Collins comes from the world of motorcycles, where peak torque can come at 15,000 rpm, and batch-fire fuel injection systems that fire the injectors every revolution are not the ancient history that they are in the four-wheel world. 15,000 rpm is one revolution every 4 milliseconds. At an 80% duty cycle, the injector is open for 3.2 of those milliseconds and closed for 0.8. In Russ Collins' world, 80 percent is the limit.
On planet Pathak, peak torque is closer to 6000 rpm. That's one revolution every 10 milliseconds. A modern automotive fuel injection system is sequential, meaning it fires each injector in sync with that cylinder's intake stroke. That means the injector has two revolutions, or 20 milliseconds to do its cycle. The limit of 0.8 milliseconds is only 4 percent of that 20-millisecond window, so the duty cycle can be 96 percent before the injector goes static.
Now, before someone tries to fact check me again, realize that the 0.8 millisecond limit is a pretty rough number. The actual minimum closed time depends on the injector, the fuel pressure, the voltage used, the phase of the moon, and the type of injector driver in the ECU.
0==WW==0
"…axles of evil…" - george w. bush
Thread
Thread Starter
Forum
Replies
Last Post
