forced induction methanol engine 3

[twinturboash]

As the owner / driver of a sponsored drag car , running a twin turbo small block chev on methanol fuel , i want to share some data , and get some ideas please.

Initialy we spent 3-4 days on a engine dyno , to develop a fuel curve, and this is where some of the data has come from.

The sbc , 358 cubes , with a tunnel ram style intake manifold, and mechanical fuel injection ( constant flow) that has 2 nozzels per runner , enderle style nozzels, quiet a large plenum ontop of the tunnel ram, with twin 3 inch throttle bodies at the front.
twin 66mm turbonetics turbochargers feeding this engine.

Now,
from the data logging on the dyno, we found the temperature at the compressure housings outlet was around 148 deg C.
The temperature after the 2 fuel nozzels , at the intersection where the intake manifold bolts onto the cylinder head, was around 50 deg C.
The methanol fuel obvously does a really good job of cooling the intake air charge.

The boost pressure at the compressor outlet, and the intake plenum was the same , 29 psi recorded.
The boost pressure at the intersection of the tunnel ram and cylinder head was approx 9 psi lower......

Does the "ideal gas laws" , apply to dynamic air flow ??
If the intake air charge , is cooled does it directly affects the intake air charge pressure ??

I realize that the pulses in the intake runner change, but the equipment used , only records the max reading.
The intake runner of the manifold, is same size as the runner in the cylinder head.

Ideas for the reason for drop in boost pressure ???
regards
Ash

 

[JSteve2]

The ideal gas laws are going to still be valid at those pressures. With the quoted temperatures you should expect a little over 20% pressure drop (420K to 330K - ignoring that you've added moles of gas, which will add some error here). I'm not sure if you really mean boost pressure (i.e. 44 psi absolute, 29 psi boost) or absolute pressure of 29 psi. Assuming we're talking about a drop from 44 to 35 psi, that would match the expected temperature effect closely.

If we're really talking about 29 psi to 20 psi, then that's a little high (30%) and there is probably a measurement issue. However, I think when you said boost you meant boost so everything looks good.

 

[JSteve2]

I see there is still one issue you asked that wasn't addressed about the dynamic flow. Pressure is really a static concept, but dynamic pressure should be affected by temperature in a parallel manner to the static pressure, so yes the behaviors of the ideal gas law apply to dynamic systems at low pressures - except I'm pretty sure if you approach sonic velocity that would no longer be true regardless of the dynamic pressure.

 

[twinturboash]

hi jsteve2,
yes , i used the word boost as in whats seen on the boost gauge , for simple terms.
thanks for the input.


ash

 

[patprimmer]

Just to be clear, I have crewed with Ash and suggested he post here as I considered his operation as a professional operation and Ash has a good understanding of the science behind this, which I believe gives him the same rights of passage as both myself and Rod when we started here.

I am kind of hoping that as well as Jsteve, Rod, Greg, Isaac and Matty might contribute some thoughts.

In discussions one thought is where do you inject the methanol to optimise VE for the same energy input to the turbo but still get good ignition and burn rate to give optimum thermal efficiency.

The thought process got to how much of the fuel do you want o evaporate in the manifold, how much in the chamber while the valve is still open and how much on the compression stroke.

If you have very little fuel evaporate before the valve, how much extra fuel do you add to compensate for some fuel not evaporating and burning during the compression stroke and power stroke.

There is a point where if extra fuel is added to compensate for insufficient evaporation, the mixture is so wet that the ignition will not light it. This is partly set by the plug gap, volts, amps and duration of the spark and this is limited by the ignition system.

Ash

Are you measuring maximum or an averaged pressure. A restriction in the tube between the gauge and the point being measured would slow down response rate and average out the pressure or at least smooth out the pulses.

I expect he pulses in the runner may go slightly over plenum pressure and significantly below it during one full engine cycle.

The thing that I don't know is does the volume of fuel vapour fully offset the cooling effect due to the latent heat of vaporisation.

Latent heat of vaporisation is 1.154Mj/Kg

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[patprimmer]

Come on guys, surely someone has some thoughts or inspiration.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[evelrod]

Sorry, Pat. I have only a passing experience with turbocharging and a few alcohol fueled motorcycles. All were in the distant past. One thing I am a bit surprised about, the max boost pressure. From what I see on the little rice burners around here, that's pretty mild.

If I were to make a guess...some restriction(s) in the inlet and a very efficient manifold/cylinder head??? Perhaps it could be (I'm sure you would ck it) an instrumentation problem.

Also, if the intake charge is cooled perceptibly, it definitely will alter the pressure, all other things static. That much I found on a Turbo Coupe in the 80's. Certainly not to the degree expressed by the OP.

Rod

 

[Tmoose]

"The boost pressure at the compressor outlet, and the intake plenum was the same , 29 psi recorded.
The boost pressure at the intersection of the tunnel ram and cylinder head was approx 9 psi lower......"

Generally test ports stuck on the wall of a port or duct are reporting static pressure. Static pressure + velocity pressure = total pressure. If the compressor outlet and the head are connected by a relatively open runners /plenum ( no intercooler) Local cooling in a wide open vessel would not lower the pressure at the point of coolness. I'm wondering if possibly the higher velocity in the runner (temporarily?) exchanges velocity pressure for some static pressure. But, If the total pressure (pitot tube plumbed a special way) is 9 psi lower at the end of the runner, 9 psi is lost forever and I'm thinking the runner is "too small."

 

[patprimmer]

Come on guys

Surely someone has some thoughts on the net outcome of evaporative cooling increasing density vs air displacement by the gas generated by evaporation.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[JSteve2]

I'm not familiar with these engines - are they stoichiometric? If so, then about 1 mol of methanol is injected for each 7 moles of air.

Looking at each end, we should expect P1V1/n1T1 to equal P2V2/n2T2, where the 1s are the compressor outlet (CO) and the 2s are the intake manifold (IM). At the temperatures listed, I would expect virtually all of the methanol at the IM to be liquid phase.

If we estimate P1 as 43.6 psi, T1 as 421 K, and n1 as 1 mol, and V1 = V2, then P1/n1T1 = 0.1035. If we estimate P2 as 34.6 ps, T2 as 323K, and n2 as 1 mol (i.e. ignore any vapor generation) we get P2/n2T2 = 0.107, leaving an error of around 3.5% if we assume the CO values to be the "correct" values.

Our two big assumptions are methanol as liquid and V1=V2. If our methanol assumption is wrong, we have to have around 25% of the methanol remaining in the vapor phase (e.g. due to portions of the stream remaining at an elevated temperature) to explain the error completely. In my estimation that is too high.

By contrast, the volumes would only have to differ by 3.5% (could be simply non-ideal flows, runners being sized slightly differently, a bend, etc.) to fully explain the error. A total shift in the sensor data of about 1 psi would also explain the error (and we're comparing two sensors so the error is stacked). The temperatures would have to be off by about 10 degC to explain the error - again with two sensors being compared.

Therefore, my estimation is that this is pretty much explained by the data provided with an ideal gas law assumption. I doubt that vapor generation is causing any significant part of the noted error. Adding together normal sensor error with a slight volume variation (that could be explained by flow anomalies as well) is more than enough to account for the observed 3.5% difference (from ideality assuming equal moles) between the data points provided.

 

[patprimmer]

They are about 5.5:1 a:f where as stoich is about 6.5:1 I think. Anyway, they are quite rich for cooling and to ensure every molecule of oxygen is more exposed to a molecule of fuel vapour rather than a molecule of piston.

If you inject directly onto the inlet valve, you need extra enrichment to ensure oxygen and fuel vapour out weight oxygen and piston. At least that is what it seems.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[patprimmer]

By the way, on the weekend Ash ran 6.69 seconds @ 204.9 mph for the 1/4 at 2950' altitude, corrected air density. A PB in less than ideal conditions.

Turbochargers are still considered unconventional in altereds in drag racing. Car weighs about 1800#

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[twinturboash]

thanks guys, food for thought

the fuel system is flowed to be closer to 4.2- 4.2 :1 a/f
the race weight is 1900lbs

regards
ash

 

[JSteve2]

Good information. A/F doesn't really matter for this if in fact the fuel is still liquid. However, it does mean the percentage of methanol that would still be vapor is only about 14% to explain the 3.5% non-ideality. I would guess that the other factors (volume assumption and sensor uncertainty) still dominate.

 

[racecrafter]

In discussions one thought is where do you inject the methanol to optimise VE for the same energy input to the turbo but still get good ignition and burn rate to give optimum thermal efficiency.
Pat
Not to divert from topic but would you be willing to talk about what you learned about injector placement and how it affected atomization/charge cooling/and do you think the data would change greatly going from a constant flow system to EFI

 

[BrianPetersen]

Although I certainly don't have experience with it myself ... I've seen it reported that using water/methanol-injection before the turbo compressor improves the effectiveness of the compressor quite a bit. It absorbs the heat of compression so the density of the air inside the compressor remains higher, so it generates more boost pressure at the same compressor RPM. Since the compression is now closer to constant-temperature instead of adiabatic, it ought to require less work to do the compression, too.

The bad thing about this, is that large droplets are bad for the turbo compressor. It needs really good atomization. If you are running the engine on the fuel in question, it would probably be best not to inject *all* of it in this fashion - only just enough to bring the compressor outlet temperature down, and inject the rest into each intake port. The consequences of a backfire into the intake need to be considered. Injecting only a fraction of the fuel before the turbo compressor ought to reduce the severity of a backfire ...