Air flow...just some numbers

MACKJ · **Posted:** Mon Sep 25, 2006 10:58 pm

Ranger1 wrote…
Quote “Readings from the obdii port sitting in the driveway, 85F outside temperature:

Engine Speed------------------------------------------------1410 rpm
Comprehensive Component Monitoring Status--------Complete
EGR Monitoring status--------------------------------------Complete
Engine Coolant Temperature------------------------------187F
Fuel System Monitoring Status----------------------------Complete
Intake Air Temperature-------------------------------------176F
Manifold Absolute Pressure(MAP)-------------------------31.6 in Hg
Mass Air Flow Rate-------------------------------------------4.515 lb/min
Misfire Monitoring Status -----------------------------------Complete
OBD Requirements--------------------------------------------OBDII FED
Vehicle Speed--------------------------------------------------0 mph

Interesting data is Mass Air Flow at idle, hot, EGR on 1.125 lb/min
------------------------Mass Air Flow at idle, hot, EGR off, 2.25 lb/min
EGR on at idle appears to cut fresh air flow by 50%.

1st set of readings at 1410 rpm... Last set of MAF readings at 762 rpm, added data points to show MAF readings when egr shuts off after 60 seconds of idle.” Unquote.

My quest… Can I calculate an approximate air flow requirement for the CRD engine using some of this data and therefore question the need of expensive aftermarket air intakes?
Some assumptions and data I have used are as follows;

Peak power of 120kW is gained at 3800 rpm
Boost pressure at peak power is approx 20 psi
Ambient air temp is 70 degrees F
Air intake temp (post CAC) is 170 degrees F (from Ranger1 above)
Engine idle is 750 RPM
Volumetric Efficiency estimated at 85% (0.85)

(Volumetric efficiency is derived from calculating losses in the form of restrictions in intake ports/valves, back pressure from some exhaust gases remaining in the cylinder etc. The figure was derived from “google” research)

The Ideal Gas Law is used to calculate mass air flow.
Absolute pressure is gauge (boost) pressure plus 1 atmosphere (+14.7psi)
Absolute temperature is in degrees Rankine (deg R) which is air intake temperature plus 460. Therefore Absolute temp is 170 + 460=630 degree R.

Engine Volumetric Flow for a four stroke engine can be expressed as follows;

Air Volume (CFM) = Engine (RPM) x (Engine displacement) divided by 3456

Engine displacement is in cubic inches. Thus 2.8 litres is equivalent to 170.8 cubic inches.

Air Volume at peak power= 3800 x 170.8 / 3456 = 190 CFM

Air Volume at idle= 750 x 170.8 / 3456 = 37 CFM

Mass Air Flow (MAF) can be expressed as follows;

Mass air flow (lbs/min) = Air Volume (CFM) x Absolute pressure (psi) x 29 x Volumetric Eff / (10.73 x Absolute Temp)

MAF at peak power = 190 x (20 psi boost + 14.7) x 29 x 0.85 / (10.73 x 630)= 24 lbs/min

MAF at idle = 37 x ( 0 psi boost + 14.7) x 29 x 0.85 / (10.73 x 630)= 2 lbs/ min

The above calculated MAF at idle is very close to the number Ranger1 published earlier (2.25 lbs/min). Therefore it is a reasonable assumption that my figures closely reflect real world numbers.?!
I know that environmental factors (ambient temp etc) and the dynamics happening in our little engine will be more complex than the parameters I have assumed but hey, I am an engineer and I like to see numbers somehow validate what I have always thought.
Working backwards from the above MAF calculations it seems reasonable that the maximum air flow for the CRD is in the order of 190 to 200 CFM, which means …the OEM air filter (at 298 CFM) has acceptable air flow for our requirements and we don’t have to spend big bucks on fancy air intake systems!

DarbyWalters · **Posted:** Mon Sep 25, 2006 11:31 pm

The one problem with the above is that the VE of a turbo is usually in the range of 115% to 130%. A normally aspirated engine is 70%-85%. A normally aspirated "race engine" is 90% or better up to about 105%.

Even with those numbers the stock air filter is adequate. Now a well designed snorkel and actually add "boost" to a normally aspirated system .5-1.5 pounds just like a fully funtional "ram air" system.

Kewl stuff !!!

MACKJ · **Posted:** Wed Sep 27, 2006 2:52 am

DarbyWalters wrote:

The one problem with the above is that the VE of a turbo is usually in the range of 115% to 130%. A normally aspirated engine is 70%-85%. A normally aspirated "race engine" is 90% or better up to about 105%.

Even with those numbers the stock air filter is adequate. Now a well designed snorkel and actually add "boost" to a normally aspirated system .5-1.5 pounds just like a fully funtional "ram air" system.

Kewl stuff !!!

Firstly, I am no expert on this…far from it...and I apologise in advance for the following verbal diarrhoea…
I have just read some technical papers on the very subject. Yes, I have seen high (upwards of 150% !!) volumetric efficiency figures also mentioned for turbo charged diesel engines.
However, being conservative by nature and training, I wanted to find some real numbers that may reflect the volumetric efficiency of a Variable Geometry Turbine (VGT) equipped engine like ours.
By definition…“the intake system restricts the amount of air that can be inducted into the cylinder during one cycle. The volumetric efficiency is the parameter describing the effectiveness of the induction process. The induction process is defined as all events taking place between the inlet-valve opening (IVO) and the inlet-valve closing (IVC).”

Many volumetric efficiency models available today are developed with some assumptions on the pressures in the air path system. New turbos like the VGT affect pressures in such a way that these assumptions must be re-evaluated.
As you know the gas flows and dynamics happening in an engine at any point in time are very complex as there are a lot of variables at work.
In particular I have read a couple of papers on the subject that describe the non-linear aspects of a turbocharger, made even more complex by the variable geometry turbo and exhaust gas recirculation (EGR) which both introduce feedback loops from the exhaust to the intake manifold. Also in a diesel engine back flow of exhaust gas occurs when the intake and the exhaust valves are open at the same time and exhaust gas can flow from the exhaust port to the inlet port (called valve overlap).
“If the exhaust gas density is greater then inlet gas density gas will flow from the exhaust manifold back into the cylinder and push out gas to the inlet manifold. During the intake stroke a combination of the fresh mixture and the exhaust gas that was pushed back into inlet manifold by the back flow is inducted into the combustion chamber. Therefore the mass of the exhaust gas will take room from the fresh mixture. Thus back flow will decrease the volumetric efficiency.
If the exhaust gas density is smaller than the inlet gas density gas will flow from the inlet manifold to the combustion chamber and push out gas through the exhaust port.
Assuming that the gas pushed out contains exhaust gas infers that the back flow in this case will increase the volumetric efficiency.
Large overlap is often used in turbo charged diesel engines. Flow from the valve overlap is used to reduce the turbine temperature”.

Low to medium turbine speeds (say between 10,000 to 100,000 rpm) represent those conditions that most of us would experience. Most manufacturer (turbo) maps don’t show this region of low air flows yet they are very important due to the fact that they regularly correspond to the region of emission cycle testing.
At these speeds the effect of heat transfer between the turbine on the exhaust side to the compressor on the inlet side greatly reduces the compressor efficiency. In addition at these speeds there are pulsating flow conditions in the engine (back flow etc) that affect volumetric efficiency.
One paper I have read specifically addressed modelling and measurement of volumetric efficiency of a diesel engine with a VGT.
A number of theorems were proposed and real data was collected to validate the formulae.
All data collected was at a constant engine speed. The turbo engines were setup in test cells and the sample data included steps in load, VGT position and both load and VGT position.
A number of mathematical models were tested against real data from two engines and the best results came from the Overlap and Residual Gas Model (ORM). This model uses the relationship between the gas densities at the exhaust and inlet together with the ambient (air) gas density and the air to fuel ratio.
The ORM was shown to be stable and was the only model that could be shown to extrapolate other VGT positions reasonably well.
The volumetric efficiency of these engines was close to an average 85% at medium rpm’s. Another paper showed a smaller VGT turbo engine (with similar outputs to our CRD) that had quite low mass air flows like I have calculated between turbine speeds of 10,000 to 100,000 rpm.
I have not seen real numbers at other engine speeds and they may well have somewhat higher volumetric efficiencies at much higher compressor speeds.
I welcome someone to supply the data. This could/would be more relevant for people that subject their rigs to high, consistent loads.
That said, my calculations were simplified and I was only really looking to see if some calcs mirrored some real data (from Ranger1) and to check if the DC engineers might be way off the mark with respect to the air intake design. I don’t think they are.
The only real way to tell is to spend heaps of time and money to do the tests on the CRD! Hopefully my ramblings will justify to others not to waste money on aftermarket intake systems.
NB: I acknowledge my sources, no plagiarisms intended.

Jeger · **Posted:** Wed Sep 27, 2006 5:23 am

If valve overlap is intended to lower turbine temps, wouldnt that indicate that cool intake air would have to be flowing across the cylinder, out the exhaust valve. If it is actually flowing backwards, how could it cool down the turbine?

Now what I would really like to know is, does the CRD have any valve overlap, and if so, how much?

DarbyWalters · **Posted:** Wed Sep 27, 2006 7:15 am

The real factor is that you should be looking for VE at max output. At different rpm's it is true that VE could be 85% ... or ...70% .. or 90%. For practical purposes with "forced induction", VE at max engine output will most likely be greater than 100%...it is just the nature of the beast. If you took a 2.8L nonturbo Diesel and measured hp at max output, it would probably have a hp rating of 35 to 35% less hp. I would suspect that you could find numbers on two valve 2.8L to 3.0L nonturbo engines that fall into the 120 hp number range or less. Now add two more valves per hole plus a turbo and intercooler that can operate at ~20psi and you get numbers of ~160hp. Engine size has not changed but the VE has changed drastically. Either way, the CRD air filter arrangement is sufficient for the CRDs output in general.

Now you have to think why would we even bother changing anything. Well colder more dense air will increase efficincy...aka...snorkel or similiar device. Increasing filtering area will also "speed up" exchange...it does matter in that less restriction creates a better flow and when a filter starts becoming filled up with particles (getting dirty), a larger filter will allow efficiency to remain higher.

As you stated above, most of us will be at more sedate levels of performance ( turbine speeds of 10,000 to whatever ) at a more constant rate of speed...but you want to measure VE at max output under load to attain the true VE. This is going to be for the times that we are at WOT or towing a heavier load, ect.

DarbyWalters · **Posted:** Wed Sep 27, 2006 7:52 am

Jeger wrote:

If valve overlap is intended to lower turbine temps, wouldnt that indicate that cool intake air would have to be flowing across the cylinder, out the exhaust valve. If it is actually flowing backwards, how could it cool down the turbine?

Now what I would really like to know is, does the CRD have any valve overlap, and if so, how much?

Valve overlap is not engineered to cool down the turbine temps...but overlap will allow cooler air-fuel into the exhaust steam that will be routed back to the turbo thru the normal process...not really a design factor to cool a turbo by any means. Now, fuel itself is a cooling agent is the combustion cycle. Some people, I used to, would install a Miller-Woods injector just after the intercooler and before the turbo, that would inject fuel into the intake stream of returning air charge. This would act like a second intercooler in my Mazda Turbo and added quite a kick. Heck, it might even work here...hummm.

All overlap is (controlled basically by your cam), is the time between the complete closing of the exhaust valve and the start of the opening of the intake valve. You need some overlap (or it is beneficial) to help evacuate the spent gases from the cpmpression cycle. Without some overlap your engine could not expel 100% of the expent gases. With NO overlap you would have an engine that develops torque at a very low rpm but would become very inefficient at higher RPMs. There would not be enough "time" to expel spent gases at higher RPMs. With Too Much overlap at lower RPM your O2 sensor would be reading a fuel rich environment and would deliver less and less fuel to your system...choking it down. The low pressure caused by the high overlap produces a scavenging effect that would be sucking and expeling large amount of unburned air/fuel mixture. However at higher RPM the intake charge will be greater from the same scavenging effect and produce hp. So you need to look for the proper combination for your application. RV cams usually reduce overlap to increase low end torque, whereas high redline cars use a cam that increases overlap to a point. Of course, varialable cam setups get the benefits of both.

So basically when your reduce backpressure, you increase overlap and when you increase backpressure you reduce overlap. Each comes at a cost. You can compensate for this effect by increasing intake air charge. This will have the effect of increasing backpressure ( more air to expel during exhaust stroke). This works until you have reached the limits of your setup. To increase efficiency you now need to increase the expulsion of gases...hence the exhaust upgrade. This is the basic reason that increasing our air supply and then adding the proper exhaust will increase our benefits.

Jeger · **Posted:** Wed Sep 27, 2006 8:12 am

Lots of good info. I feel like I have had my edumacation for the day

I understand more fully what variable valve timing can do. And now I am wondering where are the VVT diesels? And are these Mazda Rotary engines you used to race? I was glad to see the newer rx-8 with the rotary in it. I have always wondered what a rotary diesel would be like.

MACKJ · **Posted:** Wed Sep 27, 2006 9:46 am

DarbyWalters wrote:

The real factor is that you should be looking for VE at max output. At different rpm's it is true that VE could be 85% ... or ...70% .. or 90%. For practical purposes with "forced induction", VE at max engine output will most likely be greater than 100%...it is just the nature of the beast. If you took a 2.8L nonturbo Diesel and measured hp at max output, it would probably have a hp rating of 35 to 35% less hp. I would suspect that you could find numbers on two valve 2.8L to 3.0L nonturbo engines that fall into the 120 hp number range or less. Now add two more valves per hole plus a turbo and intercooler that can operate at ~20psi and you get numbers of ~160hp. Engine size has not changed but the VE has changed drastically. Either way, the CRD air filter arrangement is sufficient for the CRDs output in general.

Now you have to think why would we even bother changing anything. Well colder more dense air will increase efficincy...aka...snorkel or similiar device. Increasing filtering area will also "speed up" exchange...it does matter in that less restriction creates a better flow and when a filter starts becoming filled up with particles (getting dirty), a larger filter will allow efficiency to remain higher.

As you stated above, most of us will be at more sedate levels of performance ( turbine speeds of 10,000 to whatever ) at a more constant rate of speed...but you want to measure VE at max output under load to attain the true VE. This is going to be for the times that we are at WOT or towing a heavier load, ect.

Darby,
I basically agree with what you are saying, particularly wrt the benefits of a better air filter, and I admit I am a novice when it comes to understanding the complexities of the fluid and thermal dynamics happening here. However... I have found another paper published and commissioned by Daimler Chrysler that discusses intake modelling and validation for a 2.2 liter, 16 valve, four cylinder diesel engine with common rail direct injection fitted in a Mercedes E-class. The engine is equipped with both Exhaust Gas Recirculation (EGR) and Variable Nozzle Turbine turbocharger (VNT) or VGT by another name. The maximum VE is shown on various maps as no greater than 92 to 95% for a speed range between 1000 to 5000 rpm, certainly nothing ever got above unity (100%).
I rest my case, and I mean that in the nicest possible way.

DarbyWalters · **Posted:** Wed Sep 27, 2006 10:27 am

Jeger wrote:

Lots of good info. I feel like I have had my edumacation for the day

I understand more fully what variable valve timing can do. And now I am wondering where are the VVT diesels? And are these Mazda Rotary engines you used to race? I was glad to see the newer rx-8 with the rotary in it. I have always wondered what a rotary diesel would be like.

All sorts of Mazdas but mostly rotaries. The biggest problem with a diesel rotary would be the compression ratio. A stock rotary had around 9:1...when you add a turbo it can reach even more under full boost near the apex seals ( 3 on each rotor ). I used to run up to 20#s of boost with special carbon apex seal and one off apex springs...but it would still be just on the edge. If you went to a diesel with standard 17.5:1 and then add a turbo...I don't think it could take it. With the rotary it builds a higher compression ratio "in the corners" (no better way to put it) and would either flaten the apex springs (shaped like a leaf spring and sits under the apex seal...kinda makes the apex seals float like a boat on water) or destroy the carbon apex seal (acts like the ring on a piston and rides continuously along the rotor housing). I am not saying that it could not be done but there would be some challenges.

DarbyWalters · **Posted:** Wed Sep 27, 2006 10:35 am

MACKJ wrote:

Darby,
I basically agree with what you are saying, particularly wrt the benefits of a better air filter, and I admit I am a novice when it comes to understanding the complexities of the fluid and thermal dynamics happening here. However... I have found another paper published and commissioned by Daimler Chrysler that discusses intake modelling and validation for a 2.2 liter, 16 valve, four cylinder diesel engine with common rail direct injection fitted in a Mercedes E-class. The engine is equipped with both Exhaust Gas Recirculation (EGR) and Variable Nozzle Turbine turbocharger (VNT) or VGT by another name. The maximum VE is shown on various maps as no greater than 92 to 95% for a speed range between 1000 to 5000 rpm, certainly nothing ever got above unity (100%).
I rest my case, and I mean that in the nicest possible way.

You are basically saying the same thing as I am (we are just a little at odds at how we are arriving at VE numbers)...the VE for the engine that you pointed at was nearing 100%. Now, put a higher flow exhaust , a better intake system and maybe a module and the VE will more than likely surpass the 100% benchmark. The real point is that stock our air filter setup is adequate. Once you start the "mod" process, better intake air supply (snorkel), better fuel management (DIesel Module) and a less restrictive exhaust will increase the VE,hp,torque to the levels some are looking for. Now I can rest also...snooze

MACKJ · **Posted:** Wed Sep 27, 2006 5:47 pm

DarbyWalters wrote:

MACKJ wrote:

Darby,
I basically agree with what you are saying, particularly wrt the benefits of a better air filter, and I admit I am a novice when it comes to understanding the complexities of the fluid and thermal dynamics happening here. However... I have found another paper published and commissioned by Daimler Chrysler that discusses intake modelling and validation for a 2.2 liter, 16 valve, four cylinder diesel engine with common rail direct injection fitted in a Mercedes E-class. The engine is equipped with both Exhaust Gas Recirculation (EGR) and Variable Nozzle Turbine turbocharger (VNT) or VGT by another name. The maximum VE is shown on various maps as no greater than 92 to 95% for a speed range between 1000 to 5000 rpm, certainly nothing ever got above unity (100%).
I rest my case, and I mean that in the nicest possible way.

You are basically saying the same thing as I am (we are just a little at odds at how we are arriving at VE numbers)...the VE for the engine that you pointed at was nearing 100%. Now, put a higher flow exhaust , a better intake system and maybe a module and the VE will more than likely surpass the 100% benchmark. The real point is that stock our air filter setup is adequate. Once you start the "mod" process, better intake air supply (snorkel), better fuel management (DIesel Module) and a less restrictive exhaust will increase the VE,hp,torque to the levels some are looking for. Now I can rest also...snooze

Totally agree!!

Pablo · **Joined:** Thu Feb 02, 2006 10:06 pm **Posts:** 1201

Jeger wrote:

Lots of good info. I feel like I have had my edumacation for the day

I understand more fully what variable valve timing can do. And now I am wondering where are the VVT diesels? And are these Mazda Rotary engines you used to race? I was glad to see the newer rx-8 with the rotary in it. I have always wondered what a rotary diesel would be like.

I wanna see that little Lotus with a Rotary in it :twisted:

DarbyWalters · **Posted:** Thu Sep 28, 2006 12:20 am

I used to race a Lotus Europa with a rotary...auto-x...a wild ride

Always wanted an Elan tho

ATXKJ · **Posted:** Thu Sep 28, 2006 10:05 am

Off the original subject - but I remember seeing an article about a Rotary Diesel many years ago - used stacked Rotors - a larger one that was a gear driven supercharger and a smaller one that was the combustion chamber. European design - I don't know if it ever went into production.
For the original subject - most Mass Air Flow sensors are hot wire type - i.e. you have a small wire in the air flow that you force a current and measure the voltage (or vice versa) and based on the power used you calculate heat transfer, based on the heat transfer and the geometry of the intake you calculate the air mass - (all very easy with a microprocessor) - however there are some assumptions on the heat capacity of the air - it's a function of the humidity- which isn't well controlled (especially in Texas) so you could easily get the variation between Ranger1's readings and Mackj's calculation based on current air conditions vs the original assumptions. (add exhaust pipe Oxygen sensors with a feedback loop and your engine will run great even if your calculations are off)

DarbyWalters · **Posted:** Thu Sep 28, 2006 11:44 am

The more I thought about it...the rotary (kept normally aspirated) would work great for a diesel. It could be kept very small...reason...each rotor has 3 three combustion chambers so with two rotors you would have the equivalent of a 6 cylinder within the size of a 80 cubic inch engine. The equivalent of a 176 cubic inch ( with a factor of 2.2 ).

You could even have dual glow plugs for each rotor...hummm...no valvetrain to clog up...

Only problem is how to build up enough compression to fire off the diesel fuel

Jeger · **Posted:** Thu Sep 28, 2006 11:55 am

DarbyWalters wrote:

The more I thought about it...the rotary (kept normally aspirated) would work great for a diesel. It could be kept very small...reason...each rotor has 3 three combustion chambers so with two rotors you would have the equivalent of a 6 cylinder within the size of a 80 cubic inch engine. The equivalent of a 176 cubic inch ( with a factor of 2.2 ).

You could even have dual glow plugs for each rotor...hummm...no valvetrain to clog up...

I think there would be a lot of advantages to a rotary diesel, it would have to be a very smooth running engine with a high redline as well.

DarbyWalters · **Posted:** Thu Sep 28, 2006 12:02 pm

Yep...the trick is if the engine rpm registers at 9000 rpm the internals are only turning 3000 rpm (1 rotation of a rotor gives 3 cyclinder firings)...that is the main reason they are so smooth. If you couldn't get the compression up enuff...a mixture of Gas/Diesel or Boi/Ethanol would work at a lower compression rate.

BCool · **Posted:** Thu Sep 28, 2006 12:17 pm

http://www.angellabsllc.com/index.html

what do you guys think about this engine, looks interesting to me.

DarbyWalters · **Posted:** Thu Sep 28, 2006 12:56 pm

Looks like some kind of radial design (like old prop planes)...don't know if I trust a lab that quote the 80 cubic inch Mazda Rotary as a "4 cylinder"...lol

Jeger · **Posted:** Thu Sep 28, 2006 1:34 pm

I read about that thing some time ago. I think it might work for a steam type engine, or maybe as a sterling cycle engine, if they can get it sealed up completely, but I think it will self destruct if ran on a high energy fuel.

Seems they are afraid to run it on anything but air.

Air flow...just some numbers

Who is online