building E30 328 24v-turbo or 335i 12v holset turbo

[FormerMember]

There're number of people who are running stock M30 from E34 (with 9:1 CR) with 0,9-1,2 bar boost without problems. The engine can handle that no problems, but the main and most important single thing is the ECU. It can blow well built engine in no time, as also it can keep stock engine running surprisingly high hp / boost for long long time..

[Gunni]

It´s not the ECU you should be considering but the TUNING of any ecu fitted.
That´s 100times more imporant then the ecu.

[GeoffBob]

M30 engine 8:1CR or 9:1CR will both be fine with Holset HX40. I do not recommend you try the 9:1 with an HX52 or similar (GT4088R) turbo. As Gunni says, go with the HX40. Else be prepared to pour a lot of money into upgrading all your peripherals in order to support/tolerate/survive an HX52 or GT4088R.

[GeoffBob]

I recently took a M30 dyno chart and reworked the VE´s backwards.

From a NA or FI M30?

It has very poor VE to begin with at top, it´s all about the low end

2k = 88%
4k = 91%
5k = 81%
6k = 54%

Your VE at 5k & 6k look low to me, even for a 2V/cyl engine. How did you account for the temperature (and hence density) of the inducted air if from a FI dynograph?

[Gunni]

Both actually.

They are terribly low, this engine is all about the low end.
my calculator allows me to adjust the air temps all I want,
I set them to 80F.
Which is very normal on the cars I have tuned.

Just the other day I was reviewing a 1.5bar boost run from a B25 and it showed a 7C° step from
normal driving to 1.5bar boost at the top of the rev range. From 27C° to 34C°


This map is with a B34 the american version I think.

[GeoffBob]

Gunni,

Anything placed inline with the inlet or exhaust will modify the VE of an engine (since this effects the ability of an engine to breathe). For example, the VE of a FI engine will appear to drop as the turbine chokes the flow of exhaust gasses. This is why I think your posted VE values (at 5k and 6k) for the M30 engine are too low.

Similarly any object in the inlet, including the throttle and the design of the inlet headers, will contribute to the engines VE.

It is possible to calculate the VE of an engine using the equation:


Equation 1.

where the two bracketed terms at the top of this equation are simply the mass air and fuel flow into the engine (in units of kg per second). From left to right, the bottom of the equation is the density of the inducted air (in units of kg per cubic metre), Vd is the volume of the engine (in units of cubic metres) and N is the engine speed (in units of rotations per second).

It is possible to populate this equation with values from a dynograpgh provided the dyno session included a datalog of the AFR, MAP, and temperature of the inducted air.

From first principles the density of the inducted air is calculated using the equation:


Equation 2.

where P is the absolute (not boost) pressure measured in the plenum (in units of pascals), na is the molar mass of air (na=0.029 kg per mol), R is the universal gas constant (R=8.3144 Joule per Kelvin.mol) and T is the temperature of the inducted air (in units of Kelvin). FYI, 0'C=273kelvin. For a 3.5 litre engine, Vd=0.0035 cubic metres.

Putting equation 2 into equation 1:


Equation 3.

Looking at equation 3 you will see that the VE of an engine is effectively a function of the temperature and pressure of the inducted air. To calculate the VE of an engine under test you must have a log of the temperature and pressure of the air inducted into the engine (typically measured at a point slightly after the throttle) or you will not be able to correctly solve equation 3 (or any equation that determines engine VE). An estimate of the air temperature will not be sufficient due to the sensitivity (direct relation) of the equation to temperature. In other words, a slight inaccuracy in your estimation of T will have a noticeable effect upon your calculation of your engines VE.

Finally, the mass fuel flow may be determined from the engines power output (recorded during the dyno session) using the equation:


Equation 4.

where W is the engine power output (in units of Watts) measured at the flywheel, and for a reasonably tuned engine running on 95 octane fuel the brake specific fuel consumption (bsfc) is 0.000,000,0833 kg per Joule. The mass air flow through the engine is simply the mass fuel flow (as above) multiplied by the recorded AFR.

For an accurate measure of the VE of an engine, however, the mass air and fuel flow through an engine should be measured directly (with a MAF and MFF sensor (or MAF and AFR sensor) respectively) rather than calculating from the recorded engine flywheel power. This is particularly important since the power measured at the flywheel does not include a measure of the power the engine is delivering to the alternator, water-pump, bearing friction etc.

Hence, so long as you can fit a rpm, temperature, pressure, AFR, MAF (or MFF) and AFR sensors to an engine (connected to a data logger unit), you have a mechanism to measure the VE of any engine simply by taking it for a drive and logging the respective values as a function of engine speed.

Oh, and add a speed sensor input to your data logger (and providing you know the mass of the vehicle under test) and you have a means to determine the cars bhp (and hence torque) as a function of rpm, purely by taking car for a drive.

[Jhonno]

Would working out the outright VE of the engine only work for values when the throttle is fully open? Or am I missing a trick?

[Gunni]

The problem with the B34 american version is not it´s inability to breathe. But it´s terrible tuning for 87octane ´80s fuel in which I based the numbers on.

I agree with you GeoffBob that those are low numbers, but they are based on low recordings of B34 versions and rather not recordings but BMW published numbers.
B35´s have it a littler better.

The VE of an engine will not change given that the pressure differential between the exhaust and intake are the same at any intake pressure. Or to further say that as backpressure goes up, VE goes down or visa versa.

The HP assumption I use is that given any 10lbs of air consumed(I think it was at 80F) you should get 100rwhp.
It´s just an assumption as the variables that dictate precisely the VE and the HP output are many and I haven´t bothered to incorporate all of them into my little calculator.

I even added a BSFC adjuster and adjustable drivetrain %´s to have fun with.

Side note:
I have seen a B34 or B35 I´m not sure that had tubular manifold, HX40 and maintained the VE alot higher(6k and dropping slower after that) and thus produced alot better numbers, but similar sized compressor with a log manifold on a B34 only maintained VE to 5k but then fall flat on it´s face.

Johnno.

Throttle angle is by no means a variable in VE calculations apart from the pressure adjustments after it it makes,
You should see 100kpa or WOT pressure at about 60-75% throttle on most cars.

[Jhonno]

Ah ok.. It was touched upon as a subject, but never went into detail fully at uni. Cheers

[GeoffBob]

Gents,

Anything, and I really do mean anything, that affects the ability of an engine to breathe will affect its VE. This includes inlet and exhaust valve size, inlet and exhaust shape and dimensions, air box and resonator, even right up to the air filter and it's location in the engine bay! Throwing out the catalytic converter and restrictive mufflers is the most well known example of how to improve an engines VE.

VE is, by its fundamental definition (see "Internal Combustion Engines" by Ferguson and Kirkpatrick), the ratio of the actual volume of air drawn into the engine (for the purpose of combustion) in the presence of all actual restrictions on the engine, compared to the ideal volume of air drawn into the engine if those restrictions were not there.

Now I know that this definition sounds a little silly to start out with, since the volume of air drawn into an engine is always the same (regardless of restrictions) since it is defined by the capacity (volume) of the engine. However, the actual and ideal volumes used to determine an engines VE are first normalised to standard temperature and pressure (STP) and therefore make good sense.

It is, however, much easier in practice to calculate the VE by the ratio of the actual mass of air inducted (with flow restrictions) to the ideal mass of air inducted (without flow restrictions), where the ideal mass inducted into a cylinder is simply the volume of a cylinder multiplied by the density of air at STP.

In all calculations of VE every effort is usually made to exclude the effect of extraneous factors that may differ from car to car, so as to be able to compare apples with apples (the throttle is always assumed wide open), but the truth is this is very difficult to do and not always very honest. It's a bit like selling someone a race car and saying "Sir, this car only weighs 500kg", and then in the fine print saying "un-sprung mass excluded from weight calculation".

Gunni, It is interesting that you raise the issue of the American engine. I have never personally worked on one of these, but I do know that the fundamental difference between the two was the diameter of the inlet runners! This would definitely explain your low values of VE at 5k and 6k.

[Gunni]

Compression was also lower and tuning alot more held back.

Working back VE from HP is just a long collection of assumptions.
So it´s only going to ever be that close.

I do however have logs where I can calculate the air flow in lbs/min@standard pressure and temp or adjusted for on some of the engines I have done and then compared to the hp ratings I have gotten. And then you throw in the various dyno results and it´s again just estimations.
I did this to calculate the BSFC of my 2.8 engine which was 0.48 in the end.
But that´s tainted by dyno variation and assumed %´s drivetrain losses on that dyno.
So that´s again not accurate enough.

[GeoffBob]

Gunni, interesting discussion is this. In response to some of your comments:

The problem with the B34 american version is not it´s inability to breathe. But it´s terrible tuning for 87octane ´80s fuel in which I based the numbers on.

As in my last post, I believe that this probably has more to do with different design inlet runners, but, if you are working out your mass air and fuel flow from HP numbers then I agree that tuning will have an effect. Of course, the correct way to determine VE would be to measure the mass air and fuel flow directly. I mentioned this in my first post because I know that you are in the tuning game (and a small PC operated data acquisition unit is cheap these days). Think of the possibilities Gunni

I agree with you GeoffBob that those are low numbers, but they are based on low recordings of B34 versions and rather not recordings but BMW published numbers.
B35´s have it a littler better.

OK, accepted. Because of your work I figured there was probably a turbocharger in there somewhere.

The VE of an engine will not change given that the pressure differential between the exhaust and intake are the same at any intake pressure.

Agreed, but keep in mind that in the real world any change to the air intake or exhaust systems can have a marked effect upon the pressure differential between the inlet and exhaust sides of the combustion chamber. In reality you desire to compare the VE of one engine with the VE of another engine so as to choose the best engine (complete system) for your application. Unless all you seek to do is optimise the design of the valves you would not compare the VE of one combustion chamber with another.

Or to further say that as backpressure goes up, VE goes down or visa versa.

Not so. As back pressure goes up the mass of air inducted into an engine will drop. Thus the turbo and exhaust have a very real effect upon engine VE.

The HP assumption I use is that given any 10lbs of air consumed(I think it was at 80F) you should get 100rwhp.
It´s just an assumption as the variables that dictate precisely the VE and the HP output are many and I haven´t bothered to incorporate all of them into my little calculator.

I even added a BSFC adjuster and adjustable drivetrain %´s to have fun with.

I understand

Side note:
I have seen a B34 or B35 I´m not sure that had tubular manifold, HX40 and maintained the VE alot higher(6k and dropping slower after that) and thus produced alot better numbers, but similar sized compressor with a log manifold on a B34 only maintained VE to 5k but then fall flat on it´s face.

This makes sense (but do you mean similar sized compressor or similar sized turbine?). I'm afraid I don't like log manifolds with turbos, but I guess they have their place (and who am I to stop people having fun with them). In general, the larger the turbine the lower the backpressure (see the turbine maps on the Garett site as an example) and thus the less the effect upon an engines VE.

[GeoffBob]

Compression was also lower and tuning alot more held back.

Aah, now that is interesting! Purely for interest sake only, the one factor that does not have any effect upon engine VE is the compression ratio. It does, however effect the thermal efficiency of an engine (and thus the bsfc), and therefore will effect the HP numbers the engine can push. May I suggest that in your calculations you adjust your value of bsfc to account for changes in the CR.

Working back VE from HP is just a long collection of assumptions.
So it´s only going to ever be that close.

Fully agreed Gunni. Not my favorite way of calculating VE, but it has its place. Consider what I suggested to you about measuring the mass air and fuel flow directly. It's a great way to measure engine VE, can help diagnose problems, and will put you one up on other tuners in your game.

I did this to calculate the BSFC of my 2.8 engine which was 0.48 in the end.

Understood, but what are the units of that number? Sadly, bsfc is, in reality, a variable simply because bsfc is a direct function of your engines thermal efficiency (TE) which is itself a function of rpm, temperature and spark angle. I also work with a "typical" value of bsfc for my calculations, but it works only as guide to aid me. Like yourself (I expect) I never treat it as gospel.

[Gunni]

Or to further say that as backpressure goes up, VE goes down or visa versa.

Not so. As back pressure goes up the mass of air inducted into an engine will drop. Thus the turbo and exhaust have a very real effect upon engine VE..

But that´s what we are both saying
mass inducted dropping is loss of VE
And I´m using VE as a simplication on
EVE or effective VE based on intake pressure.
So if at 200kpa you have calculated 200% VE.
Then 200kpa / 2 = 100% disregarding intake pressure.
I always use this when tuning as VE drops as boost goes up,
then the numbers for fuel start dropping.

Side note:
I have seen a B34 or B35 I´m not sure that had tubular manifold, HX40 and maintained the VE alot higher(6k and dropping slower after that) and thus produced alot better numbers, but similar sized compressor with a log manifold on a B34 only maintained VE to 5k but then fall flat on it´s face.

This makes sense (but do you mean similar sized compressor or similar sized turbine?). I'm afraid I don't like log manifolds with turbos, but I guess they have their place (and who am I to stop people having fun with them). In general, the larger the turbine the lower the backpressure (see the turbine maps on the Garett site as an example) and thus the less the effect upon an engines VE.[/quote]

One is 58mm compressor and the other 60mm compressor.
Turbine side I can´t say for sure.
But the tubular holset had better low end and high end over the log manifold setup. Holset probably 0.91 A/R against the other 0.68 A/R . I´ve seen the effects of good tubular manifolds directly and specifically split pulse as I have tuned identical engines with identical turbos
and one has got so much better breating ability then the other, only variable is the manifold. I´m hoping the one with a worse setup upgrades to a manifold and I can directly see the change.

Some of the swedes are recording positive pressure ratios at higher rpm´s . I.e backpressure in the exhaust manifold is less then the intake pressure. And the VE obviosuly goes through the roof. The limit of this gain will be when you start pushing fresh air into the exhaust . And I haven´t researched it enough but I´m pretty sure the gains will start off nice and strong and then tail off into a unlinear gain and chasing that is going to be useless.

[GeoffBob]

Or to further say that as backpressure goes up, VE goes down or visa versa.

Not so. As back pressure goes up the mass of air inducted into an engine will drop. Thus the turbo and exhaust have a very real effect upon engine VE..

But that´s what we are both saying

Sorry, I read your comment as meaning back pressure did not have an effect on engine VE (a misunderstanding).

It would be great if the inlet (boost) pressure went up in direct proportion to the increase in turbine back pressure, then at least a turbo engine would have the same VE as a NA engine (this would make calculations so much easier) but sadly it is not so, in particular at high rpm.

... But the tubular holset had better low end and high end over the log manifold setup. Holset probably 0.91 A/R against the other 0.68 A/R . I´ve seen the effects of good tubular manifolds directly and specifically split pulse as I have tuned identical engines with identical turbos and one has got so much better breating ability then the other, only variable is the manifold. I´m hoping the one with a worse setup upgrades to a manifold and I can directly see the change.

Excellent. It is fantastic that you have the chance to collect your own data such as this. You should get a job in a reserach facility where you can do this all day

On my Renault5 Turbo replica car I have an equal length tubular exhaust header feeding cylinders 1&4 and 2&3 into a twin entry Garett GT3267. I selected this setup for exactly the reasons that you have described.

Some of the swedes are recording positive pressure ratios at higher rpm´s . I.e backpressure in the exhaust manifold is less then the intake pressure. And the VE obviosuly goes through the roof. The limit of this gain will be when you start pushing fresh air into the exhaust . And I haven´t researched it enough but I´m pretty sure the gains will start off nice and strong and then tail off into a unlinear gain and chasing that is going to be useless.

This is one reason why being able to evaluate the VE of an engine (and having the tools to do it) is important. I would not expect the back pressure to be lower than the intake pressure so I will have to think a bit about how this is possible. I shall take your word for time being that it can be done, and I would expect this situation to require a turbine that is much larger than the compressor. How long it takes a spool up such a turbine could be a problem (but once again I must repeat that I have not yet thought this situation through).

[Gunni]

I can understand how it seems counter intiutive that the exhaust manifold can have lower pressure then the intake. But from what I have gathered it only happens when you have done your homework. And the setup is in sense faultless.
No leaks, perfect pulsation in the exhaust manifold, very large turbine compared to the compressor.
very large downpipe, perfect cam and cam timing for the setup as well.
Boost stays down, power is high.

It´s discussed in Maximum boost .

I don´t REALLY want my future job to be standing at a test bench messing with engines only.
Even people who dyno tune F1 engines get bored with it.

[GeoffBob]

Not counter intuitive, just impractical. I personally would not install a turbine with A/R greater than 0.85 on a daily driver, and I see that Corky Bell indicates (figure 3.13, page 34) that an A/R of greater than 1.15 is required for the pressure at the turbine to be less than the inlet pressure. I would personally not fit such a turbine to a road legal daily driver, but on a custom built drag car (where the turbo can be spooled up on the line) this would make perfect sense, and I can see would have a huge advantage.

I would expect that with an A/R of greater than 1.2 it would be possible to increase the period (and magnitude) of the inlet and exhaust valve overlap without the usual concerns of exhaust pushing back into the inlet. I can well believe that this idea has been well exploited in high power turbocharged drag cars, but its not exactly what I would want under the bonnet of my pickup truck I will certainly keep this idea in mind though if I ever decide to build a drag car, which I would one day love to do.