Detroit 2008: BMW's twin-turbo 4.4L V8
BMW's twin-turbo 3.0L inline-six has been a rousing success, with nothing less than a Wards 10 Best Engines award sitting in its trophy case. The praise is understandable and well-deserved, as the engine uses a combination of forced induction and direct injection to serve up a torque curve that's far beefier and flatter than what should be expected from a engine of its size. So what next for the propeller brand? Simple -- take similar technology, add it to the company's 4.4L V8 and toss in a dash of unconventional packaging to make everything fit.
BMW put its latest jewel on display at the 2008 Detroit Auto Show, and we snapped some shots of it to give you an idea of why this engine is so special. We've included those photos in one of our galleries, and below the fold, we've thrown in a few observations.
Gallery: Detroit 2008: BMW 4.4L V8 twin-turbo
Undoubtedly, the most unique feature of the 4.4L TT is the packaging of the turbochargers in the valley between the cylinder banks.
This is made possible by the swapping of the intake and exhaust ports in the cylinder heads, with the exhaust now exiting towards the middle. This allows the turbochargers to be mounted in the valley with minimal plumbing, which means less bulk, lower weight and better efficiency.
Immediately after exiting the turbochargers, the exhaust gases enter the catalysts. It's important to keep the cats as close as possible, as the exhaust carries heat that is required for the converters' operation.
Packaging the cats in this location underneath the hood must present some interesting insulation challenges!
Instead of expensive variable-vane technology, each turbo makes use of fixed geometry and relatively simple wastegates.
Each compressor is fed via a black oval-shaped duct, as shown above.
Between the compressor outlet and the intake manifolds lie a pair of air-to-water intercoolers, one for each bank. Such an intercooler reduces packaging volume and tract length compared to a top- or front-mounted air-to-air heat exchanger.
Immediately downstream of the intercoolers lie a pair of electronically-controlled throttle bodies.
Below each centrally-located coil-on-plug (COP) assembly resides the direct injector, along with the feed lines and fuel rail. It's presumed that piezoelectric injectors are employed here, which are fast enough to allow several injection events per combustion cycle.
Each fuel rail gets its own fuel-pressure regulator. The operating system of this system is yet unknown, but we're guessing that it's similar to the 200 bar (3000 PSI) that is used in the engine's direct-injection predecessors.
Dished pistons are used to lower the compression ratio, with a compact ring package to reduce crevice volume and frictional losses.
The net effect is a peak of 400 HP, with 450 lb-ft of torque delivered between 1750 and 4500 RPM. In a sport sedan such as the 5 Series, this engine will likely erase fond memories of the musclecar era, and even in portly SUVs such as BMW's X5 and X6, the result should still be rather impressive.
Get a WordPress.com Blog
Reader Comments (Page 1 of 4)
2004m3driver 1:38PM (1/21/2008)
CSL for America PLEASE!!!!
Reply
why not the LS2/LS7? 1:41PM (1/21/2008)
Very tall. I cannot see how it'll fit in anything but an SUV.
Also, I do see how they've swapped the intake and exhaust, but I'm not sure how it helps too much, since unlike a supercharger, a turbo needs access to both anyway. The only advantage here seems to be that after returning from the intercoolers, the air doesn't have to fold back into the valley. And in return they get a problem where the hottest part of the engine is now right under the hood where it gets no airflow and might even bake the paint off the hood!
This does show one thing, direct air-air intercooling is dead. Air-water-air is so much smaller and easier to package once you work out how to keep it fed.
Reply
Dan 1:53PM (1/21/2008)
The packaging reduces the pressurized volume between the compressor and turbine, to reduce turbo lag. Compare this to the extensive turbo plumbing on a Subaru and you'll see the difference. Same goes for the intercoolers.
Carlos 2:02PM (1/21/2008)
They haven't swapped the hot or cold side? It runs off the manifold and the turbine is integrated into the intake manifold.
This setup is crazy lol. I can already see the swaps for this, 145iTT anyone?
why not the LS2/LS7? 2:35PM (1/21/2008)
No it doesn't reduce the pressurized volume between the compressor and turbine. Well, it does, but only by having short pipes. It still has to have one pipe go from the turbo to the cylinders and one come back from the other side of the cylinders to the other side of the turbo.
It isn't inherently shorter, it just flows the other direction. There is still a hose going from the "wrong" side of the heads (in this case the valley) around the block to the intake (in this case on the outside). Normally this would be the exhaust tube, it's just here it's an intake tube.
Additionally, I don't see why they have two turbos next to each other. Contrary to popular believe, twin parallel turbos don't decrease lag, because although the spools have less mass, they also have half as much hot exhaust gas pushing on them.
Carlos:
The turbo isn't integrated into anything, and where it is located is the exhaust manifold area. You're used to the intake being in the valley, but in this case they've reversed it.
naggs 2:48PM (1/21/2008)
the volume between the exhaust valve and the turbo is more important than the volume between the turbo and the intake valve for turbo lag
once the turbo is spooled up, it builds pressure in the intake manifold, it take a relatively short amount of time for that pressure to be felt at the intake valve.
compare that to the exhaust side where the turbo cant do anything until the hot gases hit the turbo.
if the 2 volumes were equal, then 90% of the lag would come from the exhaust side so it makes sense to trade smaller exhaust side for larger intake side.
besides, even if the turbo were right next to intake valve, you would still want to put an intercooler in there, mine as well put the turbo closer to the exhaust because you need some space for the intercooler on the intake.
naggs 2:58PM (1/21/2008)
"Additionally, I don't see why they have two turbos next to each other."
to go with 1 turbo, they would have to combine the flow of the 2 banks about 4 inches away from the exhaust valve. that would cause all kinds of problems i dont even want to have to think about. there are enough issues with cylinder cross talk when the intake is no where near that close, i can only imagine that they would get much much worse that close on the exhaust side of a turbo engine.
besides, more smaller turbos are more responsive (even if they have half the gas of one larger one) because the volume (length is not important) between the exhaust valve and the turbo is smaller, effectively moving the 2 smaller turbos closer to the cylinder than 1 larger turbo could possibly be.
why not the LS2/LS7? 5:38PM (1/21/2008)
naggs:
Every time the blowoff/bypass valve triggers (on a shift), the intake plenum falls in pressure and you have to start again. This is not the case with the exhaust manifold, it remains under pressure all the time.
Volume-wise, the size of the intake manifold is far more critical than the exhaust manifold. It is however nice to have a short exhaust manifold length before the turbo because the longer this tube is the cooler the charge gets before it hits the spool, which decreases the energy it imparts. The difference between 4 inches and 8 matters, but not as much as you make it out.
You could easily prevent cross talk with a dual entry turbo. The air from each bank would enter the hot side via separate nozzles, but would blow on the same vanes.
naggs 7:07PM (1/21/2008)
on a shift or any time the throttle is closed, there is a reduction in pressure on both the intake and exhaust side. the exhaust side, because the decrease in throttle decreases the cfm of air moving thru the engine and the intake side, because the blow off valve vents the pressure to prevent the turbo from being backed up by the closing throttle. once the throttle is opened again, the turbo is still spooled up (because most of the load was taken off) even though the pressure did drop on the exhaust side. the issue is complicated by tricks that the ecu can use to more quickly spool up the turbo. i know some engines can delay the spark until the exhaust valve opens to basically burn directly against the turbo.
the size of the intake manifold (between the throttle and the intake valve) is most definitely very important but that is different than the volume between the compressor and the intake valve which includes plumbing and intercooler. that volume is a factor in turbo response but much less so than either the exhaust or intake manifold
the intake manifold can be of any volume and is independent of the turbo plumbing.
Infra 7:16PM (1/21/2008)
Why not:
It's not the mass of the larger turbo that matters, its the radius. Radius is second-order (ie, it gets squared) when calculating the inertia of spinning objects.
Effectively, the ratio between flow volume and resistance would be disadvantaged compared to two smaller turbos that sum up to the same mass.
Infra 7:21PM (1/21/2008)
That said, assuming 400 HP is some guess and not an official BMW number, I would say this is not too impressive considering their 3.0L in the 335i really makes around 330 HP in dyno tests.
why not the LS2/LS7? 8:44PM (1/21/2008)
naggs:
The exhaust flow drops when the throttle plate closes. The pressure doesn't necessarily drop, although in practice I expect it does.
Both sides lose flow when the throttles close. But additionally, there is a "leak" in the intake plumbing, the BPV/blowoff valve. There is no such leak on the exhaust side. Thus the pressurization is more constant on the exhaust side.
Your argument about it taking time for hot gases to reach the turbo kind of pretends the system is starting from an off state. We're talking about a closed loop system, each part affects each other. Additionally, we're talking about a system with over unity gain. In fact, this is why it has a waste gate. To spin the turbo up, you need air from the cylinder to push the exhaust vanes, which then push the intake vanes which push more intake air, which then expands during burning and pushes on the exhaust vanes...
Having a large plenum on either side slows the response of the system. And in this way, I don't think BMW has solved any problem by switching this around. And I think we both agree that the new packaging seems to present some logistical difficulties.
Infra:
A "half size" turbo doesn't have a half-size (and thus 1/4 mass) spool. The pumping is proportional to the surface area of the vanes and turbine and thus a turbo that pumps half the air is more like 70% the size, and thus has half the mass, not 1/4.
Either of you, read "maximum boost" by Corky Bell for info on intake lag or the effects of parallel turbos. You'll kind of have to ignore the parts where he talks about carbureting and move to the later parts though.
Jeff 1:49PM (1/21/2008)
Wow, that is some bizarre packaging!
Knowing how heavy turbos, manifolds, catalytic converters, intercoolers and all the associated plumbing and braketry is, doing it this way must shift the centre of gravity of the engine way up! This should definitely stay SUV-only.
Reply
naggs 2:36PM (1/21/2008)
the weight is just the begging, its a really tall engine and THE HEAT, MY GOD THE HEAT!
the friggin cats would be right up against the back of the hood, everything in that area (hood, paint, windshield wipers motors and lines, wiper nozzles and lines, electronics, brakes) all of that stuff has to be able to handle EXTREME HEAT
there would almost be fewer problems if you mounted the engine upside down. there is certainly a very long list of very good reasons that NO ONE has EVER had a v8 exhaust in the valley and intake on the outside.
all of this trouble for better turbo response? hardly seems worth it.
mr.ed 5:10PM (1/21/2008)
...and the front/rear weight bias.
madgamer 4:15PM (1/21/2008)
The first thing I thought when looking at all of this was how incredibly difficult it would be to do anything with that aftermarket wise. The way those turbos and the cats are rammed in there (as well as how the engine has been redesigned) makes swapping the turbos or something very very hard.
Guenther 5:34PM (1/21/2008)
naggs- you know how the Germans love their heated wiper nozzles.
phoenexius 1:54PM (1/21/2008)
ha ha - random links to gallery of saab 9-3 conv - nice
Reply
spdracerut 2:03PM (1/21/2008)
I actually don't think it's a lot taller than a 'standard engine'. Take a look at the new M3 motor and you have a massive intake plenum that takes up pretty much the same space and height as this twin turbo setup. The flipside is that I bet the intake manifolds on this twin-turbo V8 don't take up as much space as turbo's hanging off exhaust manifolds would require on the bottom side. So maybe the motor can be placed lower in the chassis.
One thing that can be said is that this motor probably has some pretty insane throttle response for a turbo motor considering how short all the piping is between the exhaust mani > turbo > IC > intake mani.
Reply
naggs 2:12PM (1/21/2008)
this is a little taller than the v8 in the m3 and they needed a power buldge to clear that engine, another 2 or 3 inches would just look bizare