Converting Karman to AirFlow

[DSM90AWD]

Is there anyway to convert Karman Hz to AirFlow?

I assume there is a table based on Hz, RPM... that would produce an airflow figure the ECU would use to lookup fuel requirements.. and this table would be different for 1G and 2G MASs.

I'm using a TMO datalogger on a 1G so can record most of the available variables.. but would like to measure airflow s/a the DSM-Link.

 

[kpt4321]

Mathematicially, it's not that bad of a calculation. However, the problem is, we're missing some of the information that would be necessary for a calibration.

The karman Hz value has a direct relationship with the velocity of the airflow through a "pipe," or rather the MAF in our case. The problem is, all of the formulas I can find are the karman Hz relative to the velocity of a pipe with a round cross-section. I don't know if there is a thermodynamics major out there (who isn't a freshman like me ) who can help us out?

Once you have the velocity there, you can find the volume flow by multiplying by the cross-sectional area of the MAF. From there, you can correct for temperature and pressure using the temp and pressure stuff.

The thing is, the sensor was calibrated by Mitsu with hours of testing, so I think that in the end you'd be way off. Sorry!

 

[DSM90AWD]

I'm curious as to how the DSM-Link would calculate it then as they are also using a Karman MAS and believe all the same parameters are avail on OBDI and OBDII ECUs

Another piece that would have to be accounted for was if your are correcting airflow via VPC/S-AFC, run a modified MAS or have an additional air tube on the intake.. as the Hz figures coming from the ECU may/may not be accurate.

I guess as an alternative you could calculate amount of fuel used (Inj DC, RPM, Size Inj, Fuel Pressure) and back into AirFlow.

 

[kpt4321]

Right, if you are hacking the airflow signal, than it's not accurate. That is the whole point of tuning with a SAFC/AFR/VPC.

DSMLink is able to view it because the 2g diagnostics software has a spot for the ECU to output lb/min. The ECU does have to figure this out for its own fuel calculations, but the problem is that we don't know exactly what constants and such to use. DSMLink just pulls the number out of the ECU after it has done the work.

If you know your injector flow rate, fuel pressure, injector pulsewidth (or duty cycle), dead time, and A/F ratio, you can get a very good estimate of airflow.

 

[DSM90AWD]

Originally posted by kpt4321
DSMLink is able to view it because the 2g diagnostics software has a spot for the ECU to output lb/min. The ECU does have to figure this out for its own fuel calculations, but the problem is that we don't know exactly what constants and such to use. DSMLink just pulls the number out of the ECU after it has done the work.

I've seen people quote AirFlow #s from the DSM-Link that don't seem to make sense (i.e. 46lbs/min from a 16G). I wonder if it is estimating like it does for the AFR (i.e. minus a W-B you'd just be guessing).

 

[LightningGSX]

Originally posted by kpt4321
Mathematicially, it's not that bad of a calculation. However, the problem is, we're missing some of the information that would be necessary for a calibration.

The karman Hz value has a direct relationship with the velocity of the airflow through a "pipe," or rather the MAF in our case. The problem is, all of the formulas I can find are the karman Hz relative to the velocity of a pipe with a round cross-section. I don't know if there is a thermodynamics major out there (who isn't a freshman like me ) who can help us out?

Once you have the velocity there, you can find the volume flow by multiplying by the cross-sectional area of the MAF. From there, you can correct for temperature and pressure using the temp and pressure stuff.

The thing is, the sensor was calibrated by Mitsu with hours of testing, so I think that in the end you'd be way off. Sorry!

KPT, I think you are confused about Von Karman Vortex theory and how our MAS sensors work.Its not airflow through a "pipe", but rather airflow around a "pipe".The shape of our sensors has nothing to do with the vortex street, the street is produced by a vortex generator(the so called "cylinder" in von karmans theory) inside the MAS..

 

[Hal]

Acutally, you are missing quite a bit.

Karman count is only one of 3 variables used to determine the air mass.

The ecu also takes into account intake temp and barometric pressure.

With these 3 items, you can derive an air mass number, which is what the ecu uses.

Dsmlink reports 100% accurate air flow numbers (grams per rev for example).

The data reported is the actual data used by the ecu.

If it's off, it's because the MAF is jacked up, hacked, or one of the three sensor signals has been altered by some other device (MAFT as an example).

Hal

 

[LightningGSX]

Originally posted by Hal
Acutally, you are missing quite a bit.

Karman count is only one of 3 variables used to determine the air mass.

The ecu also takes into account intake temp and barometric pressure.

With these 3 items, you can derive an air mass number, which is what the ecu uses.

Dsmlink reports 100% accurate air flow numbers (grams per rev for example).

The data reported is the actual data used by the ecu.

If it's off, it's because the MAF is jacked up, hacked, or one of the three sensor signals has been altered by some other device (MAFT as an example).

Hal

Isn't that exactly what KPT said? You are missing a variable though, the area of the sensor.

 

[Hal]

oops, I missed his post, my mistake.

Hal

 

[kpt4321]

Originally posted by LightningGSX
KPT, I think you are confused about Von Karman Vortex theory and how our MAS sensors work.Its not airflow through a "pipe", but rather airflow around a "pipe".The shape of our sensors has nothing to do with the vortex street, the street is produced by a vortex generator(the so called "cylinder" in von karmans theory) inside the MAS..

We're both right.

They form around a pipe (or other obstruction, really), within a pipe (or a square flow area, in the case of our MAF).

 

[LightningGSX]

Technically not true, for vortex shedding to happen inside of a "pipe", you still need an obstruction inside that "pipe".Not sure what you mean by square flow area, but there still has to be a physical obstruction to produce a vortex street.Air flowing through just a pipe(round or square) will not produce a vortex street.

 

[nine5raptor]

Originally posted by LightningGSX
Not sure what you mean by square flow area, but there still has to be a physical obstruction to produce a vortex street.

The sharp leading edge and subsequent flat plane of the metering section is what causes the vortices to form.

 

[DSM90AWD]

Originally posted by Hal
Acutally, you are missing quite a bit.

Karman count is only one of 3 variables used to determine the air mass.

The ecu also takes into account intake temp and barometric pressure.

With these 3 items, you can derive an air mass number, which is what the ecu uses.

Dsmlink reports 100% accurate air flow numbers (grams per rev for example).

The data reported is the actual data used by the ecu.

If it's off, it's because the MAF is jacked up, hacked, or one of the three sensor signals has been altered by some other device (MAFT as an example).

Hal

Thanks for the clairification. I was wondering if DSM-link was actually referencing the air/fuel tables combined with it's own ability for adjustments. I hope someone from the DSM-ECU list can find something comperable for the 1Gs

However.. from what I remember.. the MAS only reports Karman to a certain limit (1550Hz?).. as does the air/fuel tables in the ECU. So I assume DSM-link extrapolates info after a certain point.

My confusion also about logged DSM-Link airflow#s and actual dyno results: 41.9 = 297WHP(~330FWHP) on a PTE3251 for instance (stock 2G MAF BTW). I always thought the correlation btw airflow and HP was ~1:10

 

[Defiant]

None of this stuff really works, y'know. It's really just mass hypnosis.

My god, but humans come up with bizarre stuff, don't we?

 

[kpt4321]

Originally posted by DSM90AWD
However.. from what I remember.. the MAS only reports Karman to a certain limit (1550Hz?).. as does the air/fuel tables in the ECU. So I assume DSM-link extrapolates info after a certain point.

No. The 1g datalogger is only able to output Hz values up to 1606 Hz, because the ECU has a limited amount of space to save the diagnostics information. The MAF is able to read above 2000Hz for a 1g, and more than that for a 2g. The ECU can read basically any Hz value.

 

[Bohrn]

ahhhh, THAT makes sense Kyle!!!

I often wondered about how that was possible, ECU reads hz, but the logger can't show it. NOW I know...

a day something is learned is a good day.

 

[kpt4321]

Originally posted by Bohrn
I often wondered about how that was possible, ECU reads hz, but the logger can't show it. NOW I know...

I'm confused. What do you mean by this?

The logger does show Hz...

 

[Bohrn]

oh, cmon

The logger does show Hz of course, but it only shows it to 1606.

I always thought that the ECU could only go that high, but I have seen higher than that on the SAFC readout.

What you said put it all togeather, the ECU doesnt have enough space for all that output.

 

[LightningGSX]

Ram registers that hold variables are only 8 bits wide(0-255), your logger reads thats variable and multiplies it by 6.29(I believe).So 255 X 6.29= 1603.95.

 

[timloomis]

holy crap,and i thought calculus in high school was crazy!

 

[Jehu]

Originally posted by DSM90AWD
My confusion also about logged DSM-Link airflow#s and actual dyno results: 41.9 = 297WHP(~330FWHP) on a PTE3251 for instance (stock 2G MAF BTW). I always thought the correlation btw airflow and HP was ~1:10

The 1:10 correlation is between flywheel HP and airflow. Which is within the ball park in your case assuming 15% drivetrain losses. The accuracy of the 1:10 correlation is also dependent on your A/F ratio, timing, i.e. general state of tune, so it's a very rough guide at best.

 

[tmizer]

I don't have any logs to look at, but maybe the following would work:

1. Find correlation betweek Hz and V, velocity.

2. Use the following equation to get mass flow rate:

m.dot = 0.07647*(P/Po)*(To/T)*A.tb*V.tb

Read P off your boost gauge (add to 14.7)
Get T (out of your a**, ie -make an assumption, add to 460)
Po and To are 14.7 and 519R (59+460)
A.tb is throttle body area (60mm I.D. = 0.0304 ft^2
V.tb is the mysterious correlation between Hz and ft/sec

My numbers read a little "high" for some basic dart throwing. One could make 2 points on the Hz vs. V.tb curve: the first being 0,0 the second being a stock 14b pushing 405 cfm thru a 60 mm throttle body (get ft/sec), and this would be the rightmost point on the curve corresponding to some 2000 Hz equaling to about 220 ft/sec.

The "easiest" way would of course be to ask someone who races alot and datalogs....

-T

 

[crankbender]

dead thread but here yall go. This will give you a good idea about the cross sectional area you need to use (your conversion to a circular cross section KPT ).

http://www.engineeringtoolbox.com/21_458.html

 

[nine5raptor]

Thats a whole lot of craziness to end up with the same thing

 

[crankbender]

Nah there are established values of karman frequency knowing the flow area. The reason they are all done for circular flow areas is because for hydraulic flow (fluids including air) you almost always convert using the equations on that page.

So there is no mystery between flow and Hz it is known. Additionally you know the pressure (atmospheric) and temp (temp at intake you can guess). So the calculations to get mass flow are fairly simple.