# How to determine injector size for custom engine?

I've got a 2.5 L engine here for which I do not know the performance specs (yet), because it's customized.

I need to select injectors for this build, but there seems to be only one equation on the interweb to get the injector size, and this takes the engine's power in account, which I don't have.

My idea was to calculate the maximum mass airflow that the engine will displace, and then calculate how much fuel gets mixed with that at the richest AFR I will ever maintain (12.05). Then dividing by max. duty cycle should give me the injector size. I took 100% VE at WOT.

This is the result:

``````Specific air mass:  1.27 kg/m3
Specific fuel mass: 0.75 kg/L
Volume air flow:    6000rpm * 2.5L * 1/2
= 7500 L/min = 7.5 m3/min
= 0.125 m3/sec (because 4 stroke)
Mass air flow:      0.125 * 1.27 = 0.16 kg/sec
Mass fuel flow:     0.16 / 12.05 = 0.013 kg/sec
Volume fuelflow:    0.013 / 0.75 = 0.018 L/sec = 1062 cc/min
Duty-Cycle 0.8:     1062 / 0.8 = 1328 cc/min
``````

This looks a bit too high to me... VE and AFR could be a bit lower but it'd still be too large.

Is there any other method to determine injector size?

Cheers

• One slight issue in your calculation is that you (hypothetically) filled the cylinders totally with air, rather than with an air-fuel mixture. Therefore, your mass fuelflow is slightly high. This shouldn't make a huge difference, though. Oct 19 '16 at 13:10
• How many cylinders?
– Zaid
Oct 19 '16 at 13:20
• @Zaid dang, i completely forgot that this is the flow that 4 injectors have to deliver... The devil is in the details i guess.. then it would mean the injector size has to be 1328/4 = 332cc/min. That looks a bit better already! But the question still remains, is this an accurate calculation of the required injector size?
– Bart
Oct 19 '16 at 13:26
• Bart, when you say custom engine... are you putting a turbo on this? If so your VE estimate is off. Also, the density value to be assumed would be different
– Zaid
Oct 19 '16 at 13:40
• BTW, excellent question with obviously a great amount of research. Oct 19 '16 at 13:54

Assuming this is for a naturally-aspirated application, your calculations are reasonable.

I think you just missed dividing the value obtained by the number of cylinders.

Usually 2.5 L engines have 4 cylinders and (subsequently) 4 injectors.

So

``````1328 cc/min / 4 = 332 cc/min
``````

You would select the next biggest injector size available (although 330 cc/min injectors would work just fine here)

• I can't believe I sat here for 15 minutes going through his calculations and missed that... :) Oct 19 '16 at 13:23
• @anonymous2 it happens to the best of us :)
– Zaid
Oct 19 '16 at 13:25

Looks good except the Number of Cylinders (or more specifically, injectors) part.

In the case of a EJ257 Subaru motor (ironically 2.5 liters, 4 cylinders) a set of Deatschwerks 750cc units will get you well into 500+ horsepower with room to spare (90% max IDC).

And keep in mind this is a turbo setup, with probably a large turbo, and a VE over 100%.

I ilke your math and perservation of units. One thing that is very educational about this kind of exercise is to truly realize what IDC "Injector Duty Cycle" truly means. In your example, a very conservative max IDC of .8 is used. That means the injector is injecting 80% of the time. Of ALL time.

So? There's a common misconception that injectors only inject when the intake valve is open.

For "fun" (yes, I realize I will never date a girl or reproduce... count your blessings) take a nice streetable intake cam with a 270 degree duration, and determine the timing window you can inject in with intake valve open (ok to assume 270) and what flow you would need to accomplish this at say, 7500rpm. Remember it's still an Otto cycle.

Wasn't that fun?? [cough]

Injectors on performance engines, especially turbo cars, are on almost continuously at high loads. Doesn't seem intuitive to me, but it's the truth. With the right tuning and intake airflow, the fuel doesn't even condense out on cold valves like it used to with CIS and bank injection systems. The details of the flow and aerosol physics invovled at that level are far beyond my comprehension.

On edit:

It appears my ramblings may have lead the OP astray. The screenshot below is for a turbo car with a target desired 500hp at flywheel. It also describes the most extreme WOT condition. However, the link is useful as it does the math the OP is already quite adept at, in an easy plug-and-play. Note that "Normally Aspirated" (non-turbo) is a button selection. I picked what was appropriate for a Subaru EJ257 (which I know and love) but I did not mean to imply the screenshot was an answer to the original question. By all means, knowing the underlying math is a far greater skill than depending on an online calculator.

• Thanks for your response, but what are you trying to say? Do i have to calculate my injector size with 90% duty? I don't have a turbo, ram air effect or fancy camshaft(yet) or any of these cool performance increasing stuff. Still, 729cc is more than 2 times as high as in my calculation, how do you ever get to use that much fuel with a 2.5L engine and 500hp? A VE of 300%? (which doesn't sound credible to me)
– Bart
Oct 20 '16 at 8:30
• Maybe i get your point, i'd have to calculate the timespan that the injector is allowed to be open, and the amount of fuel that i want to get in that timespan. Then you get the required injector flow, which may be higher than in my calculations. Is that right?
– Bart
Oct 20 '16 at 8:32
• No, no - don't get me wrong... I started to ramble as I often do. You can use any max IDC you desire. 80% is conservative and good. 90% is an unwritten bubbe-miese which allows a margin for odd temperature/air density/humidity conditions. The fact is you don't want to see an IDC of 100%, because the slightest condition change that asks more fuel will leave you lean... and as you can imagine, that can be very bad. As the all-knowing @Zaid mentioned, go one step higher. I don't think "too big" injectors are ever a problem, as long as the latency and injection is properly mapped. Oct 20 '16 at 9:13
• The thing about "injection while intake valve is open" was a personal Quixotic windmill tilt.. . You already have done your homework and have good answers. I only want to drive home the point (as I do with my students) that 80% IDC means open 8 counts and resting two, with NO regard for camshaft or intake valve(s) position. Oct 20 '16 at 9:16
• @Bart I think Steve is emphasizing the fact that port injection isn't limited to the duration of the intake stroke alone. He's not proposing that you change the calculation
– Zaid
Oct 20 '16 at 9:40