Fundamental 2JZGTE flaw?

[JohnA]

...You cannot imply that because you reckon to be able to get 300hp at 1 bar from home tuning, GM should be able to get 15hp/psi from the 2JZ in stock trim....

I don't imply anything dude. GM made some errors of judgement when it came to those engines, that's partly why I started my site some time ago.

 

I'm wondering aloud why the measly 10bhp/bhp came to be.

It looks like a good, efficient design from any angle I've looked at it, and I can't see why they needed 11psi to add another 100bhp to the 2JZGE. Simple as that.

 

The way I see it, they don't appear to have 'skimped' on any single aspect of the design, so 140-150bhp should have been added to the 220bhp, not just 100.

 

If you want to use Vauxhall as an example, they skimped big time on intercooling. Fitting a proper FMIC I saw 238bhp on standard boost. That is 88bhp over the n/a engine's 150bhp. (but that is a 2lt engine!)

 

We may be going in circles here, but I still hope someone will come up with an original idea. Someone who's datalogged the stock Toyota ECU perhaps (I'm clutching at straws here )

[chilli]

hmmm well I can see why you're intrigued, not quite sure your going to find the answer you're looking for here though, but you never know...

 

perhaps you should get a datalogger and do some pioneering work to discover why, I'm sure we would all be interested to know

[JohnA]

I've got some ideas, which I could verify experimentally if I had a datalogger for the Toyota ECU.

 

If nobody on this board has ever had one, then I probably shouldn't have posted this thread here.

[Digsy]

Not sure you could even get such a thing off the shelf, John. Data logging functions are usually turned off in mass production ECUs.

 

On the other hand, simple data loggers that you plug into your laptop are pretty cheap to buy and easy to use, if you don't want to look at high-frequency events that is.

 

You seem pretty convinced there is a problem. What are your ideas?

[JohnA]

Surely someone here must have used the round plug under the driver's side....

 

How else can you read the sensor outputs under running conditions and see what's happening?

[Digsy]

How else can you read the sensor outputs under running conditions and see what's happening?

The Toyota manuals refer to the use of a breakout loom on the ECU, so I'm guessing that you can't read all the sensor outputs on the diagnostic port.

[Guest Usmann A]

Yep with a OBD scanner AFAIK you can see igntion timing and read outputs .. they I know of someone who did this when setting up thier E manage in the US.

 

It can be done, but the scanner is needed ,,

[JohnA]

Shame they don't make this any more. Nothing on eBay.com, SupraForums haven't got any leads either.

 

Surely someone here must have datalogged the stock ECU.

How else can you build a knowledge base? It's one thing talking about the ECU 'retarding' under knock or whatever, a different thing looking at the datalog and right after the knock event see advance shifting from 15 to 5 degrees. Or filling with race fuel and seeing ignition moving to 19 --- or staying put.

In my mind the former is speculation/wishful-thinking/talking-bollox, the latter is a sign of being knowledgeable. (I'm old fashioned perhaps hell I can't even send txt msgs, really I can't)

[Digsy]

The Toyota manuals refer to the use of a breakout loom on the ECU, so I'm guessing that you can't read all the sensor outputs on the diagnostic port.

Obviously you can then, and I stand corrected - but as far as I can see even if you could get one it isn't a data logger. Just a display unit.

 

Better than nothing though.

[JohnA]

Nothing indeed.

I contacted the only UK company I could find in the field and they do nothing for pre-2000 Toyotas.

Zilch

Niente.

[JohnA]

Not sure how the ignition timing behaves in real life, but the fuelling from mid-boost upwards is off it's rockers (remember when I was trying to lean it down a bit using pneumatic methods, with limited success lol....)

 

I understand why mr "T" would do that, I would have done the same if I were a mass-manufacturer, to account for fuel filters that are never changed, injectors full of crud, intercoolers with no fins, yadi yado yada.

 

...I would wager money on the car gaining 20-30bhp just from trimming off fuel to bring it closer to 12.5:1.

We're talking over 20% excess fuel here, in my engine's case this is a sad attempt to cool down a chamber that doesn't need cooling.

[Digsy]

I'm sure this was covered in another thread, but if this was a design flaw (or protection against failure from worn components) then everyone's would do the same? Is this the case?

 

As you say, there may be an allowance for a certain % over capacity in the intercooler but FWIW I think you can discount protection from running lean & hot due to a blocked fuel filter or dicky injectors. In my experience, P&E (performance and emissions) and durability work assumes proper maintenance. There are abuse tests, but if you tried to bullet proof the engine against everything the consumer could throw at it, you would drive yourself bonkers.

[JohnA]

I'm sure this was covered in another thread, but if this was a design flaw (or protection against failure from worn components) then everyone's would do the same? Is this the case?.

It's standard practice among major Jap manufacturers.

Do you think that out of the box, EVOs, scoobies etc run less rich? they don't.

.

As you say, there may be an allowance for a certain % over capacity in the intercooler but FWIW I think you can discount protection from running lean & hot due to a blocked fuel filter or dicky injectors. .

I don't know, maybe my turbos are worn and push less air than they should (resulting in abnormally rich readings, since the ECU is squirting open-loop at that stage)

It's hard to compare notes in reality, how often do you find people with stock(ish) cars and widebands fitted?

 

Some graphs I've seen from the US-spec indicate very rich operation at higher boost levels, maybe JSpecs are more 'normal', after all they have a lower-capacity fuel delivery system without making appreciably lower power.

 

In any case this sort of 'richness' alone wouldn't account for the bhp/psi discrepancy, which is the topic of this thread.

It's got to be something else.

[Digsy]

Just bringing this one back from the doldrums...

 

I went and had a chat with our head of R&D today - also our pressure charging guru - to try to get to the bottom of the "scaling bhp" issue, and whether upping the turbo hp by a factor of the boost pressure was valid, even given "optimised" conditons.

 

Not strictly the aim of this thread, but the answer I got may go some way to eventually helping to explain why the 2JZ seemingly produces less power per psi boost than John thinks it ought to.

 

The answer I got was that scaling by boost pressure is sound in basic theory but is pretty much "turbocharging 101". The reality moves away from this situation so quicky that it is of little use in real world turbo sizing - even on an rule-of-thumb level.

 

What he did show me was a little spreadsheet that he uses for empirical turbo sizing. Now, at the risk of sounding like I'm talking out of my neck, I'm not going to reproduce the spreadsheet here, for these reasons:

 

1) I don't pretend to be able to understand it all (yet).

2) My colleague hasn't given me permission (although I'm sure he wouldn't mind).

3) The really clever stuff in there is to do with the thermodynamics of the turbocharger and intercooler operation. However, if you set the conditions to be optimised as John specifies for any kind of empirical calculations to work, you can ignore this bit completely.

 

What I will do is explain how the not-so-clever bit works, and then if anyone wants they can reproduce the calcs for themselves.

 

If you want to scale hp from NA to turbo (or supercharged, for that matter) then what you have to use for your scale factor is the density ratio of the intake charge after the intercooler to atmospheric density.

 

Now, if you set the intercooler efficiency to 1 (which means that the intake charge temperature is the same as ambient temerature - no mean feat using traditional cooling methods) then yes, this ratio is the same as the ratio of manifold pressure to atmospheric pressure, as the ideal gas laws would lead you to believe.

 

However, since we are talking about empirical calcs here, my learned colleague immediatey adds another 10% onto the required pressure ratio to get from NA hp to turbo hp, so if your NA hp was 220 and you were shooting for 360hp from your turbo engine, your required pressure ratio would be 1.8 rather than 1.63, which means your required boost (at ambient intake temps) would be 0.8bar rather than 0.63 bar.

 

The reason for this extra bunce is the compensation that has to happen because of the different conditions inside the cylinder prior to ignition in a pressure charged engine: higher final compression temperatures and pressures, which equale retarded ignition and/or extra fuel to avoid detonation.

 

Another factor which is missed in the boost scaling, which actually works in the favour of the power per psi question, is that of friction:

 

An engine's net power at the flywheel is the sum of its indicated power and the (negative) friction power. This equals its brake power, which is the only really useful figure for the end-user.

 

Officianados will notice the same terminology as used for cylinder pressures - indicated, friction and brake, because in a roundabout way this is what this empirical methond is scaling by: brake mean effective pressure (BMEP).

 

The thing about scaling by boost pressure is that is also scales the frictional losses inside the engine, which is not a real condition. The frictional losses in the turbo engine will be similar (if not the same) as those in the NA engine. Therefore, if you want to scale the NA brake hp, you first have to make an estimate of the frictional losses, add it on to get the NA indicated hp, then scale that by the density ratio, then subtract the same friction power again to get the brake pressure charged hp.

 

The net result is a power figure that is slightly higher than you would expect by simply scaling by the density ratio alone.

 

So what figure to use as the frictional losses? I expected this to be quite low but to my amazement the figure plugged into the empirical calcs was - wait for it - 33% of the indicated power (or 50% of the brake power). That's right: half as much power again as you see at the flywheel goes to simply overcoming the internal friction in the engine - if the empirical figures are to be believed, and at this point I have no reason to disbelieve that friction losses don't lay in this kind of range (maybe between 33% and 50%).

 

So: In summary, the scaling factor for brake power is slightly lower than you would expect from just looking at boost pressure, and slightly higher than you would expect by just looking at density ratio.

 

As for moving away from the ideal situation, some more data needs to be filled in. Namely:

 

The turbocharger's isentropic efficiency - a measure of the real increase in density accross the turbo, taking into account the heat tha tthe turbo puts into the intake charge.

The intercooler's effectiveness - a measure of how cool the charge is once it has been through the intercooler. An effectiveness of 1 means that the intercooler returns the charge back to ambient temperature.

 

Both these parameters affect the final density ratio between atmospheric and plenum, and good ballpark figures are 0.7 for both. The real figures can be worked out fairly easily by measuring pre and post compressor temperatures and pressures, and pre and post intercooler temperatures, so if anyone has this data for the stock IC and the stock turbo setup, now is the time to post it up,

 

The other big factor is the "fudge" (for want of a better word) to account for igntion and fuelling. Once the real turbo and intercooler parameters are established, if this factor still needs to be outlandish to get the UK spec's 326 bhp out of the calcs, then we can probably say that John's assumption about running rich is probably correct.

 

The final peice of the puzzle would be a real world figure for friction power from an engine dyno, but I'm not holding my breath for that

[Guest Usmann A]

oooo, oooo, temps probes pre post turbo and cooler .....

 

 

4 fast reacting bosch racing sensors and a black box ... but will mean nowt to you guys ...

[Miguel]

oooo, oooo, temps probes pre post turbo and cooler .....

 

 

4 fast reacting bosch racing sensors and a black box ... but will mean nowt to you guys ...

 

 

You reckon

[Guest Usmann A]

edited -i misread your mig, plus im tooooooooo tired.

[Miguel]

well, not for Johns quest of stock based turbos and BHP/psi outcomes ..?

 

 

With you ...ill shut up ;-) Now get that car done

[Guest Usmann A]

haha, you can show me my mistakes on AIM ,lol

[Miguel]

[Guest Usmann A]

O, I thougt there was something wrong with my commment ....

 

anyway, just down to bad net interpreting thingie ...

 

Its coming along see the picces, LOL

[Miguel]

Link Link Link .......all excited now....ooowar hijacked

[JohnA]

Nice one Digsy.

 

As you might suspect, I am already aware of the density issues, i/c effectiveness and a few more on top. I wouldn't mention any of this of course on a public forum if there is no reason to 'up' the tech level of the thread (plus people get stroppy and abusive, don't know about you, it happens with me often lol...)

 

As it happens on my current experiments I have not allowed the density ratio to deteriorate, and the intercooling efficiency is pretty much 100%. Having not gone crazy with boost pressures (say 20psi++) I am fairly confident that the Toyota ignition and fuelling maps are pretty OK, especially with Tesco99RON fuel - it helps to ensure zero knock activity, we don't want 'safe maps' do we?

 

I wouldn't mind if you could email me that spreadsheet, I could compare it with others I've made in the past.

[chilli]

Nice post Digsy.

 

> I wouldn't mention any of this of course on a public forum if there is no reason to 'up' the tech level of the thread

 

I don't think anyone on here is going to get too scared by density and i/c effectiveness, it's not like we are all completely thick lol

 

Up the tech level as much as you can I say, it's quite possible that it is at the edge of your knowledge where the answer for the discrepancy you are looking for rests.

[Alex]

The digsmeister does it again...quite amazing how much frictional loss there is.

 

FWIW I don't think any IC core I've seen has ever claimed to be more than 92% (.92 out of 1) efficient...and most are high 70's to low 80's.