XSI Twin Turbo

[PugLee]

Just started my XSI 2.0L 8v Turbo build

Will be writing it all up on .net... (only because im too lazy to rewrite everything here lol)

http://www.theywhoshallnotbenamed/forum/...ost1966222

Enjoy

[Poodle]

How come you've gone with the 8v engine? Looks like an interesting project.

[Ruan]

No means to rain on your parade - whilst what you've proposed will "work" - you need to think about transient conditions and what happens then, you're thinking in overall pressure, you need to consider what happens when you put 5lbs of boost infront of a smaller turbocharger, then suddenly take it away - and put it down the inlet instead - what does that do to the manifold pressure, also what does that do to the smaller turbochargers compressor efficiency, since now the operating conditions are totally different...

There's really no need for the check valve you have in your setup - you aren't moving anything like enough air to require them. You'll quickly realise that as soon as you open it, the pressure will suddenly drop (read up about the compound effect), you might as well go full compound, since soon as the big turbo spools up, the pressure ratio will drop across the smaller turbo and therefore flow rate will increase - removing the need for the valve - remember the smaller turbo is now not producing much pressure, merely flowing the air that the larger turbo has compressed above ambient - you just need to control the exhaust gas flow across the small and large turbos well. Long as you set your wastegates up correctly (you'll probably want to do this with some electronics)

Refer to diagram below for proper compound setup:



Your limiting factor is not gaining more inlet pressure, since you can't use it realistically - rather spool capability - you might be wiser to use a modern BorgWarner EFR/Garrett GTX or similar with a appropriately sized turbine housing which would outperform a compound setup, cost similar and be more reliable under transient conditions. You can't shoot for insanely high inlet pressure without serious problems with knock, unless you're running on super dooper race fuel... An EFR has such a WIDE compressor map, you'll realistically be able to get a big turbo to spool fast, not surge *AND* provide a crapload of air at high RPMs.

Compound setups like you're suggesting are GENERALLY used on high power diesel setups, where 5-6bar boost is required - you can't realistically use more than about 25lbs without serious drops in compression and different designs of cylinder head which are better for combatting knock.

[PUGNUT_XS]

Good luck with the build, i converted my 306 to the 2.0 8v Turbo engine, still having a few niggles at the mo, but on the whole, pleased with it.  Have you allowed for uprating the gearbox?.  The BE3's limit stock is 220BHP.

[welshpug]

be limit is not 220 bhp, its torque limited, I know quite a few running significantly more than 220 bhp without any gearbox problems other than the few animal drivers, but they tend to break shafts.

I think there's another reason to stick to a good single turbo, which is the head, these engines just don't rev or flow as well as a 16v, so you may well not be able to make use of a twin charged setup anyway, just isnt the air flow to make it work.

[PUGNUT_XS]

be limit is not 220 bhp, its torque limited, I know quite a few running significantly more than 220 bhp without any gearbox problems other than the few animal drivers, but they tend to break shafts.

I think there's another reason to stick to a good single turbo, which is the head, these engines just don't rev or flow as well as a 16v, so you may well not be able to make use of a twin charged setup anyway, just isnt the air flow to make it work.

Agreed, you would need to do some work on the head to make it flow better, cat cams do a fast road cam for it too.  This engine is designed for low down grunt, which it does well, with the BE3, i was advised not to go over 220 BHP in stock trim.

I dont know how much fact there is in this but aparently the GTI-6 RS head will fit the J2TE block and run using the stock internals.  I can say that the head does fit as when i was building mine i tried it.  The RS block and the TE block are the same, just the TE has the turbo fittings on it, i compared them when i was building my car.

[welshpug]

id imagine they meant 220 lbft, not bhp.

indeed all xu10 blocks are much the same with just machining detail differences, presence of oil spray bars drillings etc.

[luke306rallye]

Mine's running a fair bit more than 220bhp with no problems

[Ruan]

Horsepower doesn't kill gearboxes.

[zx_volcane]

Horsepower doesn't kill gearboxes.

rappers do?

[Mattcheese31]

Horsepower doesn't kill gearboxes.

rappers do?

hehehehehe sorry, nothing contribute but that tickled me

[Ruan]

[PugLee]

Haven't done much recently, but engine should be out soon.

Currently speaking to a few turbo specialists about my setup, the check valve is to stop compressed air from the first turbo recirculating during stage 1

As for a 16v head, I was told the j2 block was shorter by around 2mm, so wouldnt work?

Haven't had any warnings with regards to gearbox though, apart from flywheel and clutch but those are a given

Ruan I will have a proper read later after when I'm on my laptop

[welshpug]

whoever told you that is talking bollocks, its the same block.

[PugLee]

Pug1off told me that... I'll have to look into a 16v head then

[fatlapit]

BE box will be fine
I advise moving to 16v single turbo as previously said a GTX unit or similar

But then again I`m biased to my car

good luck bud

---

also reading your thread a gt 28 as turbo 1 and a gt3782 ans a second is just silly, and your 3.5k budget is optermistic lol. you will struggle to make all that fit using the correct sized tubing for your exhaust. if your aiming for 500bhp just get 1 large gt28 bud.

[PugLee]

I would like to build something fairly unique, so may stick to the 8v head... if I can work out how to get it running

That price is for everything to get the engine running, so clutch, ECU, forged pistons etc, Ill be welding up the manifolds myself so saving some there

I have a T28, not GT28, its slightly smaller, it came up cheap so I grabbed it lol

This is what Ive based mine on...

[fatlapit]

U do realise just ecu sensors injectors will probably eat your budget up once youve been mapped.

[Ruan]

Notice that precise design is for Diesel motors which need very high boost pressures - turbocharged petrol motors don't... You simply don't need this - just run a single, GOOD turbo - I'm all for interesting stuff, but it's pointless, will eat money and the outcome will be poor since you simply cannot use the amount of boost pressure you're on about.

You also need to study the compressor maps if you are to make the setup you're on about work - the turbos you're intending on using are very poorly specced for that job.

[PugLee]

Notice that precise design is for Diesel motors which need very high boost pressures - turbocharged petrol motors don't... You simply don't need this - just run a single, GOOD turbo - I'm all for interesting stuff, but it's pointless, will eat money and the outcome will be poor since you simply cannot use the amount of boost pressure you're on about.

You also need to study the compressor maps if you are to make the setup you're on about work - the turbos you're intending on using are very poorly specced for that job.

Surely this setup can be run at lower pressures to suit?

1. I cant find a compressor map for the T28 and 2. Im still waiting to find out if my large turbo is in fact a GT40

And like I said, they came up cheap so I got them, Ill just sell one or both if I must

---

U do realise just ecu sensors injectors will probably eat your budget up once youve been mapped.

Trust me its all been accounted for

[welshpug]

Ive seen a rr graph from a Litchfield 4.2 ish stroked r35 gtr, on low boost (roughly .8 bar iirc) it made over 900bhp, they call 1.2 bar high boost and its making 1200 by then.

turbos on it aren't even that big!

iirc the stock IHI turbos are nigh on the same as an Impreza WRX, and theyll go to 600+, gives you an idea of the efficiency of the vr38 engine, and those turbos.

[Ruan]

Notice that precise design is for Diesel motors which need very high boost pressures - turbocharged petrol motors don't... You simply don't need this - just run a single, GOOD turbo - I'm all for interesting stuff, but it's pointless, will eat money and the outcome will be poor since you simply cannot use the amount of boost pressure you're on about.

You also need to study the compressor maps if you are to make the setup you're on about work - the turbos you're intending on using are very poorly specced for that job.

Surely this setup can be run at lower pressures to suit?

Well, you could, but that defeats the whole object of what that configuration is trying to achieve - that's designed to provide high boost pressure, which is precisely what you don't want - because you cannot combat the issues with detonation before you actually start being able to utilise the boost pressures provided by a system like that at low RPM.

Why have any of the complexity and inevitability that it'll all go wrong, jam a valve shut and either produce insane boost pressures or not make any, when your money, time and resources would be better spent getting a single _good_ turbo to work efficiently...

I see so many crazy, hairbrained ideas which are all well and good, but if it's too complex, it just won't get done.

---

Just actually begun to look at compressor maps to maybe attempt to prove to you how far out you are with those two...

With 3 seconds of googling I found a T28 compressor map which proves that at the maximum pressure it'll provide PR2.8 it'll provide around 25lb/min airflow - enough for around 250hp. At absolute maximum airflow on the choke line which corresponds to around PR2.0 - you're getting 33lb/min.

Start to look at the GT3782 compressor map - well, 25lb/min hasn't even hit the first efficiency island barely, move to 33lb/min - we're only just beginning to look decent here on the GT37 compressor map, but the first turbo will now not be able to provide any pressure, it's overspeeding and choking up... To move 25lb/min into the engine at 4000rpm, you'd need PR2.8 - where you're RIGHT up against the surge line, because now the primary turbo is running out of puff...

They're very badly matched - the T28 isn't a high pressure turbo and is too small anyway, the GT37 is too big anyway - you'd run it into surge because the small turbo can't move enough air. If you were looking for a setup to do 4bar of boost to provide 380hps worth of airflow, but maxing at around 4000rpm, a GT3271R and that GT3782 would be a nice combination. But you don't need high pressure to get high horsepower, just rev it to 7k and realise you only need 22psi boost at 7k to make 380hp - but you needed 50psi boost at 4k

TL;DR - get an EFR/GTX.

[PugLee]

Can you get the link for the T28 map?? Don't know why I couldn't find one!

Had a feeling they wouldn't match, oh well, stick one up on eBay lol

[fatlapit]

Any progress??

[PugLee]

Nothing physically happening to the car at the moment, but i have been working on mapping the compressor charts.

So here is some mathematics for your enjoyment!

This is for the T28

Its all in american units because 124 cubic inches is easier to work with thatn 2045cc

Cubic Inch Displacement (CID) = 124.8

Volumetric Flow Rate (VFR) = (124.8*7000)/3456 = 252.8 Cubic Feet/Min (CFM)

Ive assumed a Volumetric Efficiency of 85% so...

VE = 252.8*0.85 = 214.9 CFM

Mass Flow Rate (MFR) = (2.703*14.7*214.9)/(70+460) = 16.11 lbs/min (2.703 and 460 are constants, 70 is ambient air temp in fahrenheit)


Pressure Ratio (Pr) = (16+14.7)/(14.7-2) = 2.42 (The -2 is to account for pressure drop, -1 each from air filter and intercooler)


Temperature of air coming out of turbo (COMPt) = (((70+460)*2.42^0.283)-460-70)/0.75 = 200.8+70 = 270.8 Fahrenheit (f) (0.283 is a constant, the 0.75 is an assumed turbo efficiency of 75%)


Temperature of air coming out of intercooler (INTt) = (270-70)*0.75 = 150 f (Again, the 0.75 is an assumed intercooler efficiency of 75%)


Density Ratio (Dr) = ((70+460)/(150+460)) = 0.87*2.42 = 2.1

Now if I take my Dr and multiply it by my MFR = 2.1*16.11 = 34 lbs/min

Doing it all again @3500RPM gives me 17 lbs/min


So now i can map it on the compressor chart, looks a little something like this...

   

[fatlapit]

Wtf does that men

[Poodle]

It means the gt2871r is unsuitable for producing 1.2bar boost at 17lbs/min, as most suitably-sized petrol turbos will be. If you perform those calculations and produce a point left of the surge line (as per the yellow line in the image) you need a different turbo or a different target boost/airflow point. Tbf to the turbo 3500rpm is a very low number to be producing 1.2 bar on a petrol, especially if you want the turbo to be able to continue boosting to the redline rather than becoming a restriction higher up. Not sure if i've missed something, but that compressor map isn't for the T28 either, try this: http://www.mazdausa.com/MusaWeb/images/m...tent07.gif. Crap picture, but you can see the figures on the axes alright.

[Piggy]

Yeah the T28 and GT2871R are rather different units

[PugLee]

Yeah wrong map, couldn't find one for the T28, so I just found a turbo with same trim and A/R

3500 rpm was a bit of a shot in the dark tbh, just went with half of redline, although max redline is 8000 I dropped a 1000 to play it safe

[PugLee]

Ok I found the correct map for the T28! A lot of the Nissan S200 lads accept this as a suitable map, so its what im using.

Looks a lot better!!

   

32 lbs/min @7000rpm running 14psi

Will get my larger turbo mapped soon and put it up