Couldn't find the simple answer on a search:

I need the mathematical equation for aiflow in (lb/hr) using normal inputs. I will be logging this with my ECU+ so I need to know in terms like (MAS HZ) etc. I want to see if my boost taper is really causing a decreased flow or if my efficiency just somehow went up.

And while I'm making posts... would someone link me to where I can find the GM BCS at its best price?

 

Load= (roughly) hz/rpm*852

g/rev= load/95

lbs min= g/rev*rpm/454

Peak power is the point where 2 byte airflow stops "climbing". In a log you can see where the jumps cease moving in 10-20hz increments to something like 5-8hz, revert for a second than flatline....generally anyway.

 

Thank you sir, I knew you'd come through for me!

 

I mean obviously its better to log 2 byte load, but since I am unfamiliar with ECU+ I thought I'd give the shorthand form as well. Its not 100% but its better than nothing. You'd also to have 2 byte airflow to make sure that this stayed constant but at that point I'd assume you'd already have the load.

 

rockauto.com has a good deal on the GM bcs and connector harness as well.

There is a post with part numbers somewhere

 

I have 2 byte load defined for the ECU+ so is there a more accurate way to define airflow with load as a variable?

...Yeah I ended up ordering from RA... but I don't think their prices are close to unbeatable. Ah well. Thank you.

 

Do you mean airflow hz in relation to load where its not quite so "out there" as with my previous post?

 

lol, no idea what you said in that last post. What I'm saying is that I can currently log 2-byte load and I can define custom math channels so if you can use 2-byte load as a variable to make a more accurate (lb/min) airflow reading then could you please detail that equation? Thanks man.

 

Solve from the other direction-

Fuel mass X AFR = mass flow of air.


You can't get mass flow from the factory volume airflow sensor by just using Hz, temp and baro must be factored in. There's a damn good reason they stuck a baro sensor and an air temp sensor right next to the flow sensing element.

 

Yes, a mass airflow (lbs/hr) calculation needs to know air volume, air temperature, and air pressure. PV=nRT stuff..............

 

Try this-

850cc/min injector static flow rate @3 bar


1cc water = 1g

SG of Stoddard solvent is .79

850cc/min = 671.5 g/min of Stoddard solvent

SG of pump fuels is ~.75

(671.5*.75)/.79=637.5 g/min

637.5*4 (injectors) = 2550/454= 5.616 lb/min of fuel at 100% DC

We will ignore latency. We will ignore the pressure difference between tested inj capacity pressure (3bar) and normal pressure drops across the injectors tips as the difference in fluid viscosities sorta makes up for it. Injector output will be fairly linear per % DC once you get off the low side where the large latency + actual open time differential exists.

If you are converting to AFR from Lambda you need to factor in stoich, keep in mind the 14.7 number bandied about does not count for modern pump fuels, its closer to 14.4-14.5. Or less.

75% DC on log with a good wideband at 11.5AFR

5.616*.75*11.5=48.4 lb/min air

If you want to make it accurate you need to factor in fuel SG, fuel stoich (wideband sensors all deal in La and convert to AFR in software), what the pressure is, the injector test fluid SG and pressure, and latency if known but the fudge method above is pretty damn close to a calibrated airflow sensor.

 

Nice, some of the better applied science that we've seen here in awhile. I would have immediately gotten tripped up going from water, to stoddard solvent, and would have probably picked the wrong SG of the average E10 pump gas (at least on the east coast thats what it usually is).

 

It wasn't close to my calibrated MAF data until I realized injectors are not tested with gasoline. It has lots of holes in it, it can be accuratized, but its reasonably close. Its a lot closer than assuming all air passing the factory MAF is at STP. Look on any logs from any ECU that reports airflow at how close the airflow and inj DC curves are when they are scaled similar.

About the only thing I remember from 6th grade science class having to do with metric stuff is the definition of a gram is 1 cc of water, and SG is what something is compared to water volume to volume.

Fuel properties can be had from Big Oil websites.

 

Yeah I like the fact that you are basically using a couple knowns (injector PW and fuel SG) and then utilizing the wideband AFR as your measurement. When guys run a variety of stock maf, stock maf with different intake, GM MAF, or MAP setup doing it your way eliminates the ambiguity. It comes down to what size are your injectors, what is your duty cycle, what is the SG of your fuel, and what is your AFR.

In evoscan we can do all that math (including subtracting out the latency) and log lbs/min realtime. It wouldn't be as accurate until the car starts flowing some fuel, but that is where the data is useful anyways. Knowing what your lbs/min are *approximately* to see where you are on your compressor map etc.

 

For those running on stock evo 550's, stock fuel pressure, pump gas with a specific gravity of 0.75, here is the formula that I got working on my car in evoscan.

[InjDutyCycle]*[WB02LC1]*0.0363436

Notice that I took my constant for lbs/min of fuel flow at 100% duty cycle and divided it by 100. I did this because evoscan calculates duty cycle in xx% instead of 0.xx %. Also I am ignoring injector latency for now. I am using WB02LC1 for my calculations instead of the external wideband because I don't like hooking up my LC1 via the serial port, adjust the formula as necessary to fit your car. As always logging duty cycle via 2byte rpm and enabling logger superspeed will provide a smoother graph trace.

EDIT:
This is an approximation. It is good for seeing where you are at on a compressor flow map etc. Making quotes like "I move 75lbs/min of air" is assinine at best.