Converting Karman to AirFlow

[kpt4321]

Crank is the friggen man.

I'm taking so many thermo classes next year, it's going to freaking rock.

 

[crankbender]

Actually all that stuff is fluid dynamics stuff not thermo

Remember gasses are fluids when in motion

 

[nine5raptor]

Fluid mech was one of my favorite classes.


I actually understood most of the stuff in that one

Right now, System Dynamics has me thoroughly confused

 

[crankbender]

DUDE is it the controls systems version??? if so that class sucks butt!!!

Dynamics is pretty cool though by itself.

 

[nine5raptor]

You got it.

The order:
Statics (was ok - got tired of trusses)
Dynamics (actually liked)
System Dynamics (currently not understanding)
Controls and Feedback (coming up next)

So systems is a precursor to controls.

 

[crankbender]

Controls and feedback sucks IMHO....it is just so stupid.

 

[tmizer]

Okay, I wanted to get a little more info before replying, and short of finding my old 1G MAF, here's a rapidly condensed set of calculations I came up with a little help from a technician I work with who's familar with the vortex shedding meters we use at work.

The following assumes an approximately maxed out 14b pushing 30 lb/min and causing a 2000 Hz signal thru the MAF. I'm assuming the bluff body in the MAF has a diameter of 0.15 in. This was done in order to get the Strouhal number to around 0.19 ~ 0.20 which is about what a circular cross section has for a Reynolds number in th 10^5 to 10^6 range (I get 235000 in this 60mm throttle body bore). Square section bluff body's have Strouhal numbers in the 0.12 range. I'm not sure where my 1G MAF is anymore since I switched over to a MAFT. Maybe someone could look and reply.

You must be logged in to view this image or video.

Hydraulic diameter for a circular duct is simply the diameter of the duct... 4*A/P, do the math, not sure why Crank said it was so special.... consequently, the area used is simply the area of the throttle body bore. One could split RCH's on the butterfly plate and rod, but Mach number is very low, around 0.1 in this case. Getting the Strouhal number correct is a 1st order effect. Area changing by 5 or 10 % is more of a 2nd or 3rd order effect. E.g., change the Strouhal number to 0.12, and mass flow changes from 30 to 49 lb/min. Change effective area by 10% (downwards), and 30 goes to, duh, 29.7....

Good luck in fluids, but pay attention to your instruments & measurements class as well (it's more fun anyway... unbelievably, I did an isobaric subwoofer enclosure for my semester project, what fun college was).
-T

 

[tmizer]

...and yes, controls does SUC*!

 

[crankbender]

Are you just guessing at those Hz and stuff.

The intent was to take the Hz and knowing the Dh of the maf and Hz using some known formulas you could get flow velocity directly.....

Couple that with the pressure and temp you can get mass flow rates without any guessing.

 

[DSM90AWD]

wow.. that formula made my eyes bleed

Now if you could package it into an "Excel for Dummies" format that could allow for changes for temp, Barometer and MAS (1G, 2G, GM MAF).. and be reasonably accurate.. It'd be worth a $1 from me

 

[tmizer]

"Stuff"? Is that the technical term? The terminology is SWAG.

Swag descriptions:
1G MAF's peg at 2000 Hz apparently, so I took a "swag" at a stock 1G MAF bluff body Dia = 0.15 inches (a "swag" because the Strouhal number needs to be <= 0.20). I also took a swag at manifold inlet temperature (is 620R not reasonable for stock SMIC?) for a pressure ration of 2 (that would be 14.67 psig or 29.34 psia). Not shown is also the assumption that effective area of the MAF at the bluff body axial location is equal to throttle body bore area.

Plug-and-chug into the above equation to calculate flow rate thru the throttle body with the above equation equaling 30 lbs/min. A reasonable number for a 14b?
-T

 

[tmizer]

Only a buck?! Geez, just when me and a few other engineers keep thinking of ideas we can sell on Ebay to get us out of cubicle hell, we get slammed down with reality. Oh well, it's worth about what the napkin it was wrote on is worth!
-T

 

[Bohrn]

wow nice work.

our mafs have a square upper section.

 

[DSM90AWD]

Originally posted by tmizer
Only a buck?! Geez, just when me and a few other engineers keep thinking of ideas we can sell on Ebay to get us out of cubicle hell, we get slammed down with reality

.. you and me both

digging thru my attic.. 1G MAS' metering area is a rectangular 1 7/8 x 1 1/8.

Since there is a section around the MAS' metering area that bypasses air.. would using the same equation above still be valid.. does the Hz the ECU produces correct for it?

 

[tmizer]

DSM90AWD, your dimensions of the rectangular MAF metering inlet would be added with the bypass and all hacked areas if hacked. That's because of the basic mass flow equation of m.dot = rho*velocity*area. In the previous equation way above, everything to the right of A.tb is density. To the left of A.tb is velocity as determined by the late herr Karman's equation. Multiplying his velocity by total area nets volumetric flow rate, then multiplying by density will finally get mass flow rate. And no, the ECU has this area as a constant. IE - if you hack another bypass inlet, the ECU won't magically know (unless maybe very indirectly thru the closed loop O2 trim table?)

The kicker now, is looking at the MAF inlet, I see an *effective* area issue. I assumed A.tb, throttle body area, would be "close enough for government work", to MAF inlet area. It certainly looks a little smaller than double your dimensions (metered plus bypass). If you pull a flow straightener (honeycomb) out and "fold" it back down, you get an idea of just the blockage part. Throw in the bluff body, the bypass middle wall, add in some honeycomb losses, it's enough to pull your hair out. The simplifying part of this all is the wealth of information from people with logs on relatively known equipment. I'd throw in a k.loss factor next to A.tb, set Dia.bb = 0.35 inches (I measured), and tweak k.loss to match calculated flow rate to measured logged flow rate. Obviously it's a *positve* number between 0 and 1...! When I get to work tomorrow I'm going to ask our technician about this crazy serrated body, flat leading edge, eliptical wedge of a bluff body (this would effect the Strouhal number). As is, however, the equation presented appears to work. Tweak to fit.
-T

 

[tmizer]

Here's the tweak:

You must be logged in to view this image or video.

Which correlates well with this:

14b compressor map

I brought everthing back to the MAF, temperature, static pressure, and updated the bluff body dimension. Strouhal number for a long bluff body is 0.2 in one of our books, and doesn't have a Reynolds number dependency.
-T

 

[nine5raptor]

Finally, a real 14b map!

I now know I am at ~69% efficiency as I am flowing ~30lb/min (0.19 m^3/sec) at a PR of 2.3 (20 psi) which is spinning that compressor at ~150,000 rpm.

Sorry, it's the little things in life that excite me

 

[dsmchips]

As a point of reference, the factory MAS is calibrated to read liters of air, which is converts to grams per revolution or grams per second. At 1000 Hz:
1G MAS = 90 liters per second
GVR-4 MAS = 94 liters per second
2G MAS = 119.5 liters per second
3G #482 MAS = 150 liters per second
EVO 7/8 MAS = 150 liters per second
As others said, though, you would need to correct these values for temperature and barometric pressure to get true air mass.

 

[DSM90AWD]

Oldie but a goodie

So relating this now "known" to my orig question.. if I datalog 1000 Hz, 1ATM Pressure and 80deg Air Temp on my 2G MAS (DSMChips 1G EPROM compensated) what would be the formula to derive volume of air in lb/min?

From what I understand the Hz relationship to Airflow is not linear so would assume some additional compensation is performed inside the ECU?

 

[tmizer]

Not much of a Mustang EEC guy, but another engineer here is, and that nifty TwEECer(?) box can change the shape of their MAF "curve"...

...but "academically", for your question, after adjusting the Strouhal number slightly (0.20 to 0.1976) to move the ~120 L/hr result the current equation gets, to 119.5 L/hr (per Dsmchips' calibration information which I'm assuming is at STP), you would then get at 80F, 1atm:

You must be logged in to view this image or video.

 

[sweet97]

Way above my head. My Hz reading max is 2520Hz. (Keydiver does a 2xKarman with 1g chips) I willhave to ask a friend where he gothis info but hesaid he found something on DSMtalk I believe that calculated the 2520Hz to approx. 30lbs/min. I will see if he wilshare. Mark

 

[sweet97]

*BALLPARK* Guy with dsmlink and MAFT gets simultaneous readings forHz and lbs/min. woks out to .51lbs/min=32Hz. I rn a 60-1 and get a 1260Hz reading which is half due to the dsmchip 2xKarman so I am getting 2520Hz.
2520/32=78.75. 78.75x.51=40.xxlbs/min. If you thin a 60-1 at 7000 rpm's with my mods could flow40lbs/min then this simple formula is close. .51 lbs/min=32Hz. Mark

 

[1_bar_goes_far]

How would you tweak the formulas for a 1g mas that just has the lower hc removed? I would think that would raise the airflow per hz only a little bit, but I would like to know. Thanks guys for all that work on those formulas

 

[sweet97]

The formula would not change. You need something to read the airflow like a Scanmaster. Then plug in the Hz you read. problem is that I also needed a dsmchip with the 2xKarman feature which allows the Hz to be read half what it is so I can double it for the true reading. We cando intake and exhaust work but it will coe down to head work and cams to really flow more. Mark

 

[tmizer]

Removing honeycombs is professionally frowned upon as the correct technical term is "flow straightener" not honeycomb.

But in the cheap DSM world pre-2004(?) it was a fairly common occurrence. I highly suggest replacing with a MAFT and GM MAF ASAP. Then one can worry more about more important things like.... "knock" and wideband O2, and "knock"...and "knock"... did I mention "knock"?

The above equations really are more of an academic exercise. To use a simple linear relation of simply x Hz = y lbm/min is not very encompassing given typical density changes (and thus mass flow rate) due to varying pressures and temperatures our turbocharged IC engines make pushing enough air for the MAF to register everything from 0 to 2500+ Hz. It might work "ok" on a car witha big FMIC in Alaska in February, but for a SMIC-equiped car in Florida in July, it probably won't match most of the range.