By now you know that our longterm 2010 Volvo XC60 has a lot of motor under the hood, even if you can't really see it when you pop the clamshell. The mill is torque-rich at low engine speeds and hustles the big wagon around with surprising authority. In short? It's got sauce.
For whence the sauce came, click the jump.
Let's begin at the beginnning. The XC60 T6 has a turbocharged 3.0-liter (actually 2953 cc) inline six that bears the inelegant engine code B6304T2. A single turbo huffs 8.7 psi into the mill, which develops 282 horsepower and 295 lb-ft of torque at its peak, which is actually not a peak at all.
And that, friends, is the beauty of boost -- the pointy heads in Goteburg can dial in with precision the amount of airflow the turbo provides at a given engine speed, and it is largely airflow through the engine that dictates torque. Provided there is adequate exhaust energy present to drive the turbine to drive the compressor, torque output can look, well, like the tabletop you see in the above dyno chart of the B6304T2. You'll never get this kind of area under the curve with a normally aspirated or even a supercharged engine.
The area of greatest focus by engineers of modern turbocharged passenger car engines is not maximum output or efficiency or emissions. It is transient response, or the amount of time after you press your Chuck Taylor down that the engine actually delivers the torque being commanded.
This is particularly difficult for turbo engines when the engine speed is low and you're pootling around at part-throttle. You give the throttle a little squeeze, expecting a little more torque, and you have to wait for that little extra fart of exhaust gas to drive the turbine to drive the compressor to pump a little extra fresh air, to which the fuel system adds a bit more fuel to make bigger farts. Producing this flatulence takes time, and the longer it takes the less happy the Converse-wearing dude is.
To help speed the process along, Volvo employs a twin-scroll turbo. That is, the housing that surrounds the turbine (exhaust side) wheel is split into two independent exhaust paths. These two paths, or volutes, dictate that the exhaust side of the engine be grouped into two bunches.
See here. Cylinders 1, 2, and 3 feed one of the volutes (the snail-looking parts of the housing), and Cyls 4, 5, and 6 exhale into the other volute. In this manner the exhaust manifold is essentially a pair of manifolds that each contains the exhaust of three cylinders.
This achieves two major objectives -- it delivers more pulse energy to the turbine wheel, and it keeps cylinders from contaminating each other with exhaust gas.
Here's how. Because this is a six-cylinder engine, a cylinder fires every 120 degrees of crank rotation. This means that when, say, Cyl 1 is puking a bunch of hot, high-pressure exhaust gas into the manifold, Cyl 4 is on its intake stroke, part of which involves having its exhaust valves briefly open at the same time as its intake valves are open. This is called valve overlap.
Picture a conventional inline-six exhaust manifold. The angry exhaust pulse from Cyl 1 travels down its runner and into the common collector that joins all six exhaust runners. Here, the pulse "sees" the nice low-pressure region of Cyl 4's overlap and really, really wants to be there. So much so that part of the exhaust pulse from Cyl 1 travels backwards up the Cyl 4's exhaust runner.
Now picture the divided twin-scroll manifold. Here, Cyl 4's overlap period is completely isolated from Cyl 1's pissed-off exhaust pulse, so instead of that pulse energy being drained off into Cyl 4, it is delivered to the turbine wheel instead where it can do some useful work. And so on for the remaining pairs of cylinders. Furthermore, because the collector regions of a twin-scroll manifold are each inherently lower-volume than the big single one of a conventional manifold, less energy is wasted in an unnecessary expansion.
The exhaust pulse recovery aspect of twin-scroll turbos is the primary reason for their existence. They simply build boost in a more linear and responsive fashion than do single-scroll turbos, improving time to boost and time to torque, every time you dip the throttle.
There's an additional benefit in dividing the manifold and turbine housing, though. Remember the exhaust pulse that traveled up the Cyl 4 runner in our conventional manifold? Not only did this bleed off pulse energy but also a bit of exhaust flow that instead is now taking up space in Cyl 4 that could otherwise be filled by fresh, oxygen-rich intake air. This phenomenon has many names, among them reversion, cylinder contamination, impaired volumetric efficiency, etc. Whatever you call it, it's bad, and it also heats up that cylinder somewhat, adversely affecting its knock resistance. Double bad.
With a twin-scroll arrangement, there is more freedom for the engine designers to employ more aggressive valve overlap too, since they're not forced to mix farts into their flowers.
The Volvo's turbo setup is not unique. Inline-six diesel engines have been doing this for decades but it has taken advances in metallurgy to allow twin scrolls to be used with the high exhaust gas temperatures of a gasoline engine. Particularly one running relatively lean air-fuel ratios at full load like this Volvo. When that divider wall between the scrolls gets really hot, it wants to taco and fracture radially, which isn't the best for durability. Modern high-nickel austenitic stainless steels have made this far less of an issue.
BMW's N54 engine achieves the same goal of enhanced pulse recovery as the Volvo but does it with two conventional turbos rather than one twin-scroll unit. However, the company's recent twin-scroll N55 suggests they're getting the same job done without the complexity (and cost) of the N54's two turbos and associated plumbing, hardware and sensors.
The now-defunct Cadillac SRX Turbo uses a similar arrangement as the Volvo, only with cylinders arranged in a vee rather than in a row. That one's sort of an odd bird as it's a hassle to plumb two banks of a V6 to a single turbo. It's simpler to use two turbos here (see Ford Ecoboost V6).
Anyway, a little insight into what makes the XC60 a torquey and responsive steed.
Jason Kavanagh, Engineering Editor
yellowbal says:
08:37 PM, 01/ 4/11
This article is awesome. Thank you Jason.
Could a turbo-like supercharger (like in the Miata) emulate this torque curve?
bmwm1 says:
08:42 PM, 01/ 4/11
Great article, thanks!
This T6 engine was just named to Wards Top 10.
The updated 2011 version is even better with 300 hp & 325 lb.ft of torque.
626gt says:
09:10 PM, 01/ 4/11
Interesting article - then I saw the mileage updates and the better mileage of the audi added the two together and thought "maybe an A4 vs. XC60 comparison" (engines, mileage, handling, utility, interior size/storage cost) maybe in the offering.? Regular wagon, tall wagon pros/cons. Through in a new TSX wagon for fun and share the impressions.
clarkma5 says:
09:17 PM, 01/ 4/11
Yes very well described in layman's terms; of course, as an engineer, I can't help converting this discussion into terms of fluid pressure, flow, and heat effects, but the fact that it's all there while using common language is impressive.
ampim says:
05:59 AM, 01/ 5/11
Great article Jason, although I do have one question. You mention that in a normal manifold setup, there's some exhaust gas that gets in to cylinder 4 and can increase the chance of knock, but isn't that similar to what the old EGR systems did? I though by having a little oxygen depleted exhaust gas in the cylinder, you would lower peak combustion temperatures. I guess that was more for NOx reduction, but wouldn't it also help prevent knock as well?
esoterica says:
07:30 AM, 01/ 5/11
GREAT article. Too bad Volvo briefly diverted their resources into their short-lived 4.4L V8 or they might be even more ahead of the pack on turbocharged engine technology (this engine actually debuted for MY2008)
DLu says:
08:21 AM, 01/ 5/11
I love how people are generally impressed by the power of this vehicle, which supposedly is a "practical" vehicle good for carrying people and cargo, yet plenty of people put down the Sienna for how useless it is to have a lot of power in a minivan. The XC60 T6 will never carry 7-8 people plus their luggage. Both minivans and SUV's/crossovers are frumpy and will be driven gently 99% of the time anyway. I smoke Cayenne Turbos and other such poseurs routinely in my fat minivan. So what.
I'm actually just ranting because I don't like how people associate positive attributes to certain body styles and pretend that there is a deeper or more practical reason behind it. Thanks for reading. :)
throwback says:
08:21 AM, 01/ 5/11
Thanks for a concise layman article. I expect to see more turbo engines focusing on low down torque in the future. At least for the USA with our slower speeds.
carguy622 says:
08:38 AM, 01/ 5/11
Great writeup.
subytrojan says:
09:07 AM, 01/ 5/11
Sweet blog entry, Jay! This is right up there with Dan's walkarounds!
I think I found out what turbo this uses thanks to Google. This engine apparently uses the BorgWarner K16.
http://www.btnturbo.com/parts/model/volvo/v70/bhp281/2007_b6304t2_4104.aspx
justinlink says:
09:15 AM, 01/ 5/11
@esoterica
Not sure what you mean by diverted; it was my understanding Yamaha did all the legwork for the 4.4 V8. I suspect the lengthy development time could be attributed to base engine (B6304) being developed during the Ford days. The base engine is still shared with Land Rover, and both the base and turbocharged B6304T2 engine are still built in the UK. But that's pure speculation.
Excellent article, Jason.
fundango says:
09:25 AM, 01/ 5/11
Nice post.
jkavanagh says:
09:50 AM, 01/ 5/11
@ampin, EGR acts as a diluent so it's great at part throttle -- it reduces NOx and reduces pumping losses incurred by the closed throttle for increased efficiency. Part throttle operation is also not knock limited so MBT timing is easily achievable.
At full load, max output is the primary objective and here EGR is no bueno since it displaces oxygen and erodes the detonation threshold. Even old school external EGR systems would disable EGR at full load for this reason.
- J Kavanagh
hybris says:
10:39 AM, 01/ 5/11
I like how this was written.
BigFordFan150 says:
01:30 PM, 01/ 5/11
Im 14 although I know this article is very well written, I still need some help figuring this out. Sorry
skunkworks says:
03:23 PM, 01/ 5/11
I Love ish like this, thanks Jay!
I wish you would have done this technical of a writeup on the S-AWC and AYC in the LT Evo X GSR. (Not too late)
sideswiper says:
04:34 PM, 01/ 5/11
This write up is an A+ piece of work and couldn't have been easy to do. These kinds of articles are my favorite to read, as many people like me understand the benefits of twinscroll turbos and SCs but not necessarily how they go about doing their job.
Have any of you considered writing a full how engines work, how turbos work, how leaf springs work, etc series of articles? They'd be destined to be better than the other crap I keep reading. This was written for the perfect intelligence level; not too dumbed down but not too mechanical either.
Kudos J.
sideswiper says:
04:38 PM, 01/ 5/11
Edit above, I said twin scroll sc, I meant screw.
church123 says:
06:56 PM, 01/ 5/11
Very concise Jason. That's not an easy topic to convey clearly to a wide spectrum of readers.
Next, how about a piece on how the torque ratings of these turbo motors really don't apply in the real world. I've never seen one of these sub 2000 rpm torque peak (or beginning of plateau) engines from VW, BMW, Volvo, etc. actually be able to reach peak torque anywhere close to those rpms unless you're lugging the engine in top gear - maybe.
AFAICT, they're all rated at static conditions on the dyno. For the layman, that means the engine is loaded up against the dyno at a fixed rpm until output stabilizes. Of course, that means that you get to keep putting energy into the turbo until you reach a stable boost condition. On the road that never seems to happen, and the engines can feel a touch soft, or even laggy under 2000-2500 rpm. But the car is still accelerating, and you reach some critical boost onset point where the turbo is able to fully spool an build target boost.
It doesn't mean these engines don't spool up quicker than their counterparts with 3000 rpm torque peaks, but actually achieving the rated torque at those rpms just isn't realistic.
jkavanagh says:
07:31 PM, 01/ 5/11
Funny you mention it, church123-- I originally had a section delving into exactly that (dyno vs real world torque curves). But this post was then getting waaay too long and meandering so I ditched it. For another day perhaps, since it's interesting stuff.
boost4 says:
02:49 AM, 01/ 6/11
This was a great read, Jason. Thanks. The point about diesels producing a lower EGT's reminds me of Variable Nozzle turbos and their somewhat delayed use in gas applications. Something regarding the slider mechanism that adjusts vane angle and acts as a wastegate of sorts would be susceptible to high temps and cause a boost regulation problem.
Now there are few gas applications that I can think of, 911 Turbo being one. The few people I knew with 89-90 only dodge 2.2 Turbos equipped with VNT turbos enjoyed long service life from the little T15 based units. As for diesels, Inside line had a VNT equipped engine for a time in your 05 Jetta TDI.
Is twin scroll the better, more cost effective solution for most applications?
subytrojan says:
09:22 AM, 01/ 7/11
Twin-scroll solutions are more expensive to implement.
This is why Subaru doesn't have them on any of their North American products; some of their Japanese engines have twin-scroll turbos. Using a twin-scroll turbo on a Subaru boxer engine would require the use of equal-length exhaust headers, which would eliminate much--if not all--of the distinctive Scooby boxer rumble.
This is also why the Mitsubishi Lancer Ralliart doesn't have a twin-scroll (if I remember correctly) and the Evolution does.
subytrojan says:
09:25 AM, 01/ 7/11
http://www.turbomagazine.com/features/0901_turbo_2009_mitsubishi_lancer_ralliart/index.html
"The Ralliart is more deboosted than detuned, relying on a TD04HL-15T-7 compared to the Evo X's TD05H-152G6-12T, which is one family bigger. The TD04 is a single scroll compared to the twin-scroll design of the Evo X system, and the TD04's footprint is considerably different, so a custom exhaust manifold will be required if you're wanting to upgrade to turbos outside of the TD04-flanged family. Ultimately, the proof is in the boost because the Ralliart runs a mild 13.3 psi, while the X uses 22.4 peak psi to generate its 291 hp."
jkavanagh says:
03:59 PM, 01/ 7/11
@boost4, only Porsche has a VNT on a gasoline application today. Making a VNT cope with a gasoline engine's high exhaust temperatures requires terrifically, horrendously expensive materials. That said, a VNT is theoretically sort of the "ultimate" in boost response and output.
@subytrojan, yes a twin scroll housing is more expensive than conventional one (but nothing like the cost of going to a VNT). Twin scrolls don't make a lot of sense on a flat four (Subaru) because the manifold is inherently extremely large volume due to the looooong runners. This tends to damp out pulse energy. So there's less to be gained by going to twin scroll. Not nothing, but less.
You can do twin scroll without equal length runners -- they are not a prerequisite.
-J Kavanagh
onramp says:
06:27 PM, 01/16/11
Pursuant to church123's astute observation, wouldn't this be an opportune car to throw some'a'dat "dynolicious" metrics into, to really explore where modern turbo lag is at, if any? How about some measured throttle response runs from idle at a stop and from a roll, in a tall gear?
How'abou'dit? Please, pretty please?!
myob says:
01:36 PM, 08/ 9/11
LOL DLu,
I think you have what is known as "fatherhood minivan bitterness syndrome". It sucks to have to pay $35K or more for a van, I know. Smoking Cayenne Turbos, eh? I would pay to see those two vehicles pull up at a stoplight and drag down the road to Beth, Bath, and Beyond.
The XC60 is hardly frumpy. Did you see its slalom speed? It handles better than many cars out there, for sure. The R design is even firmer and more fun to drive. It lacks only good steering feel to be on par with some cars that get the "sport sedan" title.
It won't ever carry 8 people, for sure. But it handles 2 adults and 3 dogs just fine. The power is not necessary, just makes it smoother and more fun at a tiny fuel economy price.
mark_brooklyn says:
12:48 PM, 12/20/11
Great article, And as an owner an Ice white 2010 Volvo XC60 T6- This is exactly what I'm talking about. We drive to Michigan from New York at least once a year and all I can tell you is it is truly a pleasure to open this baby up on the interstate- Im not talking about speeding either it just maintains nicely going up or down the mountains in PA. There is a little turbo lag if your going from 0 to full throttle, But after 25 mph the turbo lag is gone...you drop your foot and it responds..And let me tall you and inline6 its got a nice little growl, Not like that whiney high pitched sound my ford edge would make..anyway. Great article about a great engine in what is a great Car, My second year with it and its still runs great.
mark_brooklyn says:
12:48 PM, 12/20/11
Great article, And as an owner an Ice white 2010 Volvo XC60 T6- This is exactly what I'm talking about. We drive to Michigan from New York at least once a year and all I can tell you is it is truly a pleasure to open this baby up on the interstate- Im not talking about speeding either it just maintains nicely going up or down the mountains in PA. There is a little turbo lag if your going from 0 to full throttle, But after 25 mph the turbo lag is gone...you drop your foot and it responds..And let me tall you and inline6 its got a nice little growl, Not like that whiney high pitched sound my ford edge would make..anyway. Great article about a great engine in what is a great Car, My second year with it and its still runs great.