Using Maths channel to Tune for Volumetric efficiency

[Hugh]

Am still at the start of my Syvecs tuning journey, working out how to get to grips with how to best tune my engine.

Due to the potentially unusual temperatures created in and around the engine bay, I am not a huge fan of Dyno tuning although have to agree the Dyno is a safer method, so wanted to use the datalogging and Maths channel potential of the Syvecs system to capture whether the correct steps were being taken to set the engine up correctly.

I can see scenarios where it would be possible to poorly set the cam timing, or set too high a boost yet optimise the fuel and ignition to give a false impression that the maximum power had been extracted from the engine. Similarly ambient air pressure needs to be accounted for.

My hope is that optimising the engines Volumetric efficiency to maximise the mass of oxygen the engine is pumping is an important step on the path to maximising power.

Given that the volume of air is significantly affected by temperature, it is important to remove the effect of the ambient air temperature from the calculations since intercooler efficiency improves greatly on a cold day.

There may be additional factors on air temperature that are not captured with the following routine, but I have chosen to calculate Volumetric Efficiency using the theoretical mass of air consumed by a 100% efficient non-aspirated engine compared to the actual throughput of the engine to account for ambient pressure and temperature.

Tuning for maximum Volumetrical efficiency will hopefully allow the maximum possible power output once AFR and Ignition timing have been optimised.

Step 1 was to use the Maths channels in SView to calculate the airflow mass in lbs of air through the turbo engine. There is a thread here that describes the steps to create a Maths channel and also calculate the mass of air the engine is consuming viewtopic.php?f=4&t=843

The VE formula being used is:- Mass of air being consumed / calculated mass that a 100% efficient non-turbo engine would consume.

The one parameter that I do not appear to have access to in the Syvecs ECU is the ambient air temperature. This information is available on the R35 GTR dashboard, perhaps someone can advise if there is a way to bring this parameter into the Syvecs logs through the CAN system?

Calculating the maximum theoretical air mass of the non-aspirated engine is a function of engine capacity x RPM x barometric air pressure x correction for ambient air temperature.

Density for a given temperature can be expressed as:- Pressure / Gas constant * temperature °K
Where pressure is Pascals (millibar * 100)
Gas constant, J/(kg*degK) = 287.05 for dry air
T = temperature °K = deg C + 273.15

The density of 15°C air at sea level is therefore = 101325 / (287.05 * (15 + 273.15)) = 1.2250 kg/m3

Air density in a Syvecs Maths channel will in my case need ambient temperature 15°C added in manually, so will look like this:- bap*100/(287.05*(15 + 273.15))

The 100% VE normally aspirated calculation for the mass of dry air in lbs per minute will therefore be RPM*(engine capacity in litres/2/1000)*density*conversion from kg to lb, as follows:- rpm*3.8/2/1000*bap*100/(287.05*(15 + 273.15))*2.20462

Calculating the actual VE of the engine is now straightforward using the earlier Air flow per turbo calculation.

Need to remove the 2 that was at the end of the formula since we now need engine airflow rather than turbo airflow. 14.7*lam2*sqrt(12.47267*relFp1/2999)*fuelDutyPri1/100*6

Combining into a % VE, it looks like this. (14.7*lam2*sqrt(12.47267*relFp1/2999)*fuelDutyPri1/100*6)/( rpm*3.8/2/1000*bap*100/(287.05*(15 + 273.15))*2.20462)*100

Not done any work with it yet, but here is what the VE log looks like on a short pull, it shows that the engine VE peaked at theoretical 320% mass at 4200RPM.

Going to play around with the cam timing first to see what I can find out.

Hopefully the logic stacks up, let me know if I got it wrong / needs adjusting.

[RICE RACING]

I do Ve but as its intended taking off the mass fraction to do with MAP over Ambient pressure, this then gives you a proper Ve (as it should be IMHO).
Eg: turbo or NA you can compare apples to apples.
Anyway looking at it that way on various engines I get a raft of readings racing types ~125%, road 100% and so forth, can be thrown off it you dont use relative fuel pressure correction in the ECU, for obvious reasons.

Anyway I use this system to generate full main injection tables, not sure if I ever posted it here, but its complex, well above SVIEW, needs to be done in spread sheet, however once set up I can generate a whole new main injection table in a key press for any hard changes say to fuel pressure, specifications, hardware changes like injectors and so forth. If I get time later I will put up some screen shots for you. OR pm me and can email to you etc.

I do this to keep my sanity and be more 'productive' in time taken to map an engine correctly rather than an endless series of suck it and see iterations everyone else seems to do......

[Hugh]

Thanks RICE RACING,

Can see how you have applied VE logic to your tuning, hadn't really considered whether I could use VE to define the fuel table.

Agree with your comments on the need for fuel pressure to be accounted for, I have relfp1 accounted for in the turbo air flow calculations, the formula to apply is pretty simple, but not one that I had used much before.

Your comment about avoiding suck it and see operations was behind my interest in creating a system to evaluate engine set-up changes that will be performed on different days, with the likelihood of differing environmental effects masking the adjustments.

I am not completely certain that the effects of air temperature and pressure on a normally aspirated engine will apply as a fixed ratio to a turbo engine, so may make the above VE calculation no more effective than comparing the response to turbo airflow on a given day. Will need to log under different environmental conditions without any setup changes to answer that question.

[RICE RACING]

Blue cells are input variables
Yellow cells are calculated outputs.
I put down a few of the main pages, the fuel table map auto generates.
I apply this system to all LR/Syvecs and other ECU as well, makes life easy as

With this method and a few iterations I normally get the fuelMltCll to within a tolerance of 3% or lower across all operation ranges.

[Mark_r33]

Very interesting thread!

How will air wasted through valve overlap effect these calculations? Thinking mis-adjusted cam timing (perhaps purposely) on a variable cam system?

I bet an R35 would lend itself well to one of these for road tuning; http://www.dynomitedynamometer.com/rota ... t-dyno.htm

[RICE RACING]

Very interesting thread!

How will air wasted through valve overlap effect these calculations? Thinking mis-adjusted cam timing (perhaps purposely) on a variable cam system?

I bet an R35 would lend itself well to one of these for road tuning; http://www.dynomitedynamometer.com/rota ... t-dyno.htm

This waste is accounted for in BSFC {brake specific fuel consumption}
The power in the brake part is derived from a load cell (assuming you can hold if measuring in chassis application) that can be from a conventional dyno or from vehicular analysis (working out load, F=M*A) with something like a VBOX3i + IMU and some preliminary data collection of accurate vehicle mass, frontal area, worked out Cd and rolling resistance from a coast down test, obviously few more variables to account for in real world test but in many ways it can be validated against measured performance *the bullshit stops when flag drops* (separate topic).

So the air/fuel ratio part is what it is, the BSFC is tied in to the power measure and the Ve is what it is, how much 'air' is utilized effectively is another factor be that wasted through the example you gave or inefficient engine (common real reasons, mechanical *low CR to operate on poor fuel?* or calibration *high excess fuel ratio + reduced ign spark timing to cover poor fuel for given boost on turbo application as seen in road cars* reasons) then you have the plethora of other factors, cam timing as you mentioned, excessive tip/map ratio, excessive ccp (rooted engine), ............ you name it all will show up as poor BSFC g/hp/hr obviously if you start looking into that the fuel types that are low in energy to mass will show high values *high in this case is bad!* eg, alcohol based but you get the idea.

[stevieturbo]

Very interesting thread!

How will air wasted through valve overlap effect these calculations? Thinking mis-adjusted cam timing (perhaps purposely) on a variable cam system?

I bet an R35 would lend itself well to one of these for road tuning; http://www.dynomitedynamometer.com/rota ... t-dyno.htm

AEM sold one of those for a very short time.

No idea whether they were just rubbish or no real market for them though, but they werent on sale for long.