58 replies to this topic

[primer]

I was reading AMG-Mercedes PR material about their fancy, new M 157. This 5.5-litre V8 biturbo engine is supposed to be replacing the 6.2-litre NA AMG motor. Anyway, somewhere in the middle of that novella, there are these statements:

Quote

The flat torque curve ensures enormous pulling power in all speed ranges: 670 newton metres are already available at 1500 rpm, and the maximum torque of 800 newton metres is delivered just 500 rpm later, remaining constant to 4500 rpm. Even more effortless performance is ensured by the engine variant with the AMG Performance package, which has a peak output of 420 kW (571 hp). In this case the eight-cylinder delivers 875 newton metres of torque at just 2000 rpm, with a constant 900 newton metres available between 2500 and 3750 rpm.

Now this new engine is direct injection and twin turbocharged. Is it at all possible that the torque curve is flat (more or less) between the said RPMs? I was under the impression that there is still a very distinct peak torque at a certain RPM, but because there is so much of it in these monster motors, the manufacturers instead decide to talk about huge amounts available over a wider range of RPMs rather than the peak.

Incidentally, AMG- Mercedes seem to have poached the guy who wrote the press release for Mclaren MP4-whatever road car. You are awesome, we get it. Just show us the numbers and STFU.

[GrpB]

Yes, modern control systems with DI (not knock limited), very fine wastegate control and electronic throttles can artificially shape the torque curve to be perfectly flat for several thousand rpm by controlling for mass flow, usually up to what would be the mechanical torque peak. The trick is to keep the torque from falling off too hard shortly after and not 'making the number'.

[Greg Locock]

Very commonly done for drivability, and to preserve the transmission. You can also have different torque curves in different gears.

[kikiturbo2]

just to add variable valve timing which is used not only to "flatten" the torque characteristics, but along with some imaginative fuel management gives better turbo and with it throttle response..

it is interesting to see, on the other hand, how an artificiall "hump" in the torque graph (using alot of overboost at low RPM for example) will give you a subjective feeling of huge power increase while the car does similar numbers on the road..

[cheapracer]

and won't 500 lbs of torque at 1500 rpm be just wonderful at a roundabout on a wet day.

[gruntguru]

cheapracer, on Mar 2 2010, 10:51, said:

and won't 500 lbs of torque at 1500 rpm be just wonderful at a roundabout on a wet day.

As long as the car is 3000 kg.

[primer]

And has those 20" wheels and massive tires fore more gripz.

[gruntguru]

primer, on Mar 2 2010, 14:33, said:

And has those 20" wheels and massive tires fore more gripz.

Since Cheapy specified a wet day, I will stick with the narrower tyres thanks very much.

[jcbc3]

Just drive within the cars and, especially, your own limits, and you'll be just fine.

[gordmac]

I suspect the electronics will keep all but the very dedicated out of the ditches!

[imaginesix]

cheapracer, on Mar 1 2010, 19:51, said:

and won't 500 lbs of torque at 1500 rpm be just wonderful at a roundabout on a wet day.

Everybody seems to be reading this as sarcasm, but to me it sounds like real fun.

[dosco]

gordmac, on Mar 2 2010, 14:13, said:

I suspect the electronics will keep all but the very dedicated out of the ditches!

Hopefully the manufacturer includes a switch to disable teh driver aids. And by aids I don't mean the "high-five."

[primer]

Officially Official: BMW announces 2.0-liter four-cylinder turbo for U.S.

BMW said

Maximum output of 240 horsepower is achieved at 5,000 rpm, 1,500 rpm lower than in the normally-aspirated 3.0-liter inline six. The peak torque of 260 lb-ft, comes on stream at just 1,250 rpm. Not only is that 30% more torque than the aforementioned inline six, it also peaks 1,500 rpm earlier. The vigorous power comes on early and climbs steadily all the way to redline.

Peak torque at 1250 RPM?!

The idle would be what, around 1100 RPM? You could take off in second gear and then cruise in sixth/seventh gear all day with just 1300 RPM on the dial! This is like some commercial vehicle diesel engine.

Max power is also available at 5000 RPM. I'm very eager to get the bore/stroke numbers of this engine, it must be very long stroked to get such characteristics.

[WhiteBlue]

primer, on Jan 29 2011, 07:02, said:

I'm very eager to get the bore/stroke numbers of this engine, it must be very long stroked to get such characteristics.

http://en.wikipedia....enz_M278_engine

Quote

In contrast to the naturally-aspirated M273, however, the M278 features twin turbochargers, one per cylinder bank, producing up to 0.9 bar boost pressure... The basic M278 has a displacement of 4.7 liters (4663 cc) with a bore of 92.9 mm and stroke of 86 mm.

I would guess that MB would not use a waste gate but variable turbo vanes on this engine to improve responsiveness and limit the torque.

[primer]

WhiteBlue I meant the new BMW motor which also has a very flat and accessible torque plateau. AFAIK BMW have not revealed the dimensions as yet, but perhaps the Merc figures are a hint and the BMW is also mildly oversquare, and not undersquare.

[WhiteBlue]

primer, on Jan 29 2011, 13:08, said:

WhiteBlue I meant the new BMW motor which also has a very flat and accessible torque plateau. AFAIK BMW have not revealed the dimensions as yet, but perhaps the Merc figures are a hint and the BMW is also mildly oversquare, and not undersquare.

http://www.e90post.c...ad.php?t=477882

Edited by WhiteBlue, 30 January 2011 - 09:14.

[mariner]

These engines seem quite amazing in terms of fuel efficiency and fuel economy, espeically the BMW one. I suppose if you did a 1,400 bhp turbo 4 of 1.5 litres all those years ago this should not be too hard!!

However I would like to ask some questions.

1) I am no thermodynamist but I thought that turbo's had a pressure map which showed where they can't operate and where they are most efficent. I can see that twin scroll's help solve this but with all the waste gates, variable vanes (?) and variable valve timing how much or the ( real world) time do the turbos operate at the optimum zone in the boost map? Or putting it a different way have BMW pubished any BMEP data to match the torque/poer graph?

2) It would be interesting to know how much the all up engine weighs with so two turbo's , waste gates and VVT gear.

3) If these engines can give constant torque from 1100 to 4500 rpm etc. why do they need 6, 7 , 8 or even 9 speed auto gearboxes? Surely with so much torque the weight and cost of multi gear auto boxes should be unecessary?

Lastly and back to my admitted obsession with how much all of this has to do with getting good EU/US CO2/km test results would anyone hazard a guess where in the power curve the engine operates during the 4km urban /7 km non urban test cycle to get those CO2 results?

[primer]

Thanks, WB! I should have known better than browse autoblog and hope to get in-depth info.

BMW are guranteed another entry into Ward's 10 Best Engines. What a gem.

mariner, on Jan 30 2011, 09:52, said:

3) If these engines can give constant torque from 1100 to 4500 rpm etc. why do they need 6, 7 , 8 or even 9 speed auto gearboxes? Surely with so much torque the weight and cost of multi gear auto boxes should be unecessary?

The benefit is you can ride the torque in overdrive gears as much as possible, thus get better mpg numbers. If you can have peak torque at 1250 RPM (seems so odd to type that), no reason to keep the engine at higher RPMs while crusing at steady speeds is there? If you try to do this with a four/five speed gearbox perhaps the performance numbers will suffer a bit.

Edited by primer, 30 January 2011 - 10:11.

[WhiteBlue]

And in case no one noted the bore of 84 mm is bang on the GRE center spec. Expect to see derivatives of this engine in DTM, Rally and GT cars. I would not even exclude F1 mid term.

Ulrich Baretzky said

Source
We looked at the block, and decided it was 84 mm bore for example + or - 3 mm, so you could start with an 82 mm bore or you could go up to 87 mm bore.

Edited by WhiteBlue, 30 January 2011 - 10:32.

[Wuzak]

WhiteBlue, on Jan 30 2011, 14:28, said:

And in case no one noted the bore of 84 mm is bang on the GRE center spec. Expect to see derivatives of this engine in DTM, Rally and GT cars. I would not even exclude F1 mid term.

BMW have stated that they are uninetersted in F1 for the forseeable future. They're concentrating on touring cars.

[Wuzak]

WhiteBlue, on Jan 29 2011, 15:15, said:

I would guess that MB would not use a waste gate but variable turbo vanes on this engine to improve responsiveness and limit the torque.

Are not wastegates used to control the speed of the turbine, to prevent overspeeding and overboosting?

[gruntguru]

Wuzak, on Jan 30 2011, 21:12, said:

Are not wastegates used to control the speed of the turbine, to prevent overspeeding and overboosting?

VNT performs the Boost control function and more efficiently (lower BP) than a wastegate.

[Wuzak]

gruntguru, on Jan 30 2011, 15:37, said:

VNT performs the Boost control function and more efficiently (lower BP) than a wastegate.

So they don't need a wastegate at all?

[gruntguru]

mariner, on Jan 30 2011, 19:52, said:

1) I am no thermodynamist but I thought that turbo's had a pressure map which showed where they can't operate and where they are most efficent. I can see that twin scroll's help solve this but with all the waste gates, variable vanes (?) and variable valve timing how much or the ( real world) time do the turbos operate at the optimum zone in the boost map?

To produce a flat torque curve the turbo needs to supply approximately constant boost ie a horizontal line on the compressor map. The surge line defines the low flow (low rpm) limit and the upper limit will usually be determined by falling efficiency (and sometimes by turbo speed limitation in higher boost applications). The width of the efficiency island can be tailored to the application - usually at the expense of a lower peak efficiency. The BMW torque range is about 4:1 which is pretty extreme. It could be achieved using the turbo in the map below at a PR of 1.8, but the power extending to 7,000 rpm would pass the 60% efficiency line on this map. So the BMW engine would require a broader compressor flow range or staged multiple turbos.

[WhiteBlue]

gruntguru, on Jan 30 2011, 12:53, said:

So the BMW engine would require a broader compressor flow range or staged multiple turbos.

The BMW TwinPower system that was already used in the S63 V8 and the N55 L6 is a twin scroll system.

Quote

A Look At Twin Scroll Turbo System Design - Divide And Conquer?
May 2009 issue of Modified Mag
By David Pratte

Back in the day, most aftermarket and factory turbocharger systems featured simple log-style exhaust manifolds. But just like on normally aspirated engines, where exhaust manifold design has become recognized as a critical element to maximizing horsepower and torque output, there has been increasing attention paid to turbocharger and turbo manifold design. Divided or "twin-scroll" turbos and manifolds have emerged as the preferred design of many of the top tuners and even OEMs, showing up on high-performance models like the Mitsubishi EVO, Pontiac Solstice GXP and JDM Impreza STI. But what exactly are the differences between single-scroll (or constant pressure) turbo systems and twin-scroll (or two-pulse) turbo systems and how do these design differences impact overall engine performance?

Single-scroll systems have been in use for a long time, and for good reason. These systems are generally compact, inexpensive and extremely durable under the high heat they're exposed to. So from a simplicity of design, packaging and reliability standpoint, a single-scroll, constant-pressure turbo system is quite appealing-especially to the OEMs that must consider more than just power production. Although log-style or simple unequal-length turbo manifolds used by the OEMs can be tweaked for improved performance or replaced by a more sophisticated equal-length aftermarket manifold, this doesn't change the fact that there's a single exhaust gas inlet to the turbo's "hot side" turbine (which powers the "cold side" compressor, force feeding a denser and therefore more oxygen-rich air charge into the combustion chamber from the intake side). Because of this design limitation, single-scroll systems are not particularly efficient at low engine speeds or high loads. This decreased turbine efficiency contributes to turbo lag, something we've all probably experienced while driving a stock turbocharged vehicle.

One of the biggest limitations of most factory single-scroll turbo system is the restrictive nature of its log or compact unequal-length exhaust manifold. Keep in mind, the purpose of this manifold isn't just to channel exhaust gases to the turbocharger's turbine wheel; the manifold must be designed to allow exhaust gases to exit the combustion chamber of each cylinder quickly and efficiently. Also keep in mind that these exhaust gases do not flow in a smooth stream because the gas exits each cylinder based on the engine's firing sequence, resulting in distinct exhaust gas pulses. Next time you fire up your car, place your hand lightly over the exhaust tip (before it gets hot!) and you will feel these pulses. With a log-style or compact OE-style, unequal-length runner exhaust manifold like you'll find on SR20DET or USDM STI engines, the pulse from one cylinder can interfere with subsequent exhaust gas pulses as they enter the manifold from the other cylinders, inhibiting scavenging (where the high-pressure pulse draws the lower pressure gases behind it out of the combustion chamber with it) and increasing reversion (where exhaust gas flow is disturbed so much that its direction of travel reverses and pollutes the combustion chambers with hot exhaust gases). The trapped and wasted kinetic exhaust gas energy from poor scavenging and too much reversion also means higher combustion and exhaust gas temperatures, necessitating less aggressive ignition timing and reduced valve overlap as well as richer air/fuel mixtures (and higher NOx emissions).

Twin-scroll turbo system design addresses many of the shortcomings of single-scroll turbo systems by separating those cylinders whose exhaust gas pulses interfere with each other. Similar in concept to pairing cylinders on race headers for normally aspirated engines, twin-scroll design pairs cylinders to one side of the turbine inlet such that the kinetic energy from the exhaust gases is recovered more efficiently by the turbine. For example, if a four-cylinder engine's firing sequence is 1-3-4-2, cylinder 1 is ending its expansion stroke and opening its exhaust valves while cylinder 2 still has its exhaust valves open (while in its overlap period, where both the intake and exhaust valves are partially open at the same time). In a single-scroll or undivided manifold, the exhaust gas pressure pulse from cylinder 1 is therefore going to interfere with cylinder 2's ability to expel its exhaust gases, rather than delivering it undisturbed to the turbo's turbine the way a twin-scroll system allows.
The result of the superior scavenging effect from a twin-scroll design is better pressure distribution in the exhaust ports and more efficient delivery of exhaust gas energy to the turbocharger's turbine. This in turn allows greater valve overlap, resulting in an improved quality and quantity of the air charge entering each cylinder. In fact, with more valve overlap, the scavenging effect of the exhaust flow can literally draw more air in on the intake side while drawing out the last of the low-pressure exhaust gases, helping pack each cylinder with a denser and purer air charge. And as we all know, a denser and purer air charge means stronger combustion and more power, and more power is good!

But the benefits of twin-scroll design don't end there. With its greater volumetric efficiency and stronger scavenging effect, higher ignition delay can be used, which helps keep peak temperature in the cylinders down. Since cooler cylinder temperatures and lower exhaust gas temperatures allows for a leaner air/fuel ratio, twin-scroll turbo design has been shown to increase turbine efficiency by 7-8 percent and result in fuel efficiency improvements as high as 5 percent.

Combine these benefits with a well-engineered tubular equal-length manifold and the design strengths of a twin-scroll approach can pay even bigger dividends. "Equal length" simply refers to the length of the primary exhaust manifold tubes or runners that the cylinder head exhaust ports breath out into, which should ideally be of equal length before merging at a narrow angle at the collector so that the gases flow smoothly together into the turbine inlet. This helps maintain exhaust gas pulse energy, resulting in better boost response and overall higher turbo efficiency.
Designing a high-performance twin-scroll tubular manifold like those available from top tuners like Full-Race is no simple task. Fitting equal-length primaries into the tight confines of a turbocharged car's engine bay while maintaining proper radius bends and strong exhaust gas flow characteristics is a serious design challenge. Determining the best length and diameter of the primaries and angle of the merge collector also requires a lot of R&D, as does choosing the best wall thickness and material for the tubing itself. That's where Full-Race's team of highly educated mechanical engineers and years of constant refinement of their designs comes into play. According to Geoff at Full-Race, "Because of the increased turbine efficiency found in twin-scroll systems, twin-scroll manifolds can often use a smaller runner than a single-scroll design. However, due to the complex shape of the runners and the requirement for a second wastegate and dumptube (one for each side of the divided turbine) there's more mass and more parts which adds expense and complexity. Plus, twin-scroll turbos are physically larger than their single-scroll equivalents, so it's more difficult to make them fit our cramped engine bays." Overcoming these challenges means developing extremely robust manifolds that make smart use of the available space, something Full-Race does with the help of computer programs like SolidWorks and other proprietary processes.

All this hard work does translate to serious performance gains in the power-delivery department, particularly at spool-up and peak torque where sophisticated tubular twin-scroll manifolds properly matched to a twin-scroll turbo deliver superior airflow to single-scroll or OE twin-scroll designs. According to Geoff, "Our twin-scroll turbo kits have a higher average cylinder pressure and turbine efficiency, while single-scroll systems tend to have a higher peak cylinder pressure and exhaust backpressure. We have found the twin-scroll systems have higher backpressure at low rpm (which is good for turbo spool-up) and lower backpressure at high rpm (which is good for top-end performance). On the other hand, single-scroll systems have lower backpressure at low rpm (bad for spool-up) and higher backpressure at high rpm (which hurts top-end performance)." In order to realize the full benefit of a top-shelf twin-scroll system like one of Full-Race's, the manifold design and A/R ratio of the turbo must be spot-on, so it's best to get the help of a professional when choosing a turbo for this type of system.

It's certainly possible to generate huge power and great high-rpm performance with a single-scroll turbo system. There are plenty of examples of very high-horsepower, single-scroll turbocharged engines out there, but with single-scroll systems spool-up and response are much slower than with a twin-scroll design, yet twin-scroll systems still provide excellent top end performance. Although switching from single-scroll to twin-scroll can be expensive, for hard-core boost junkies who want much faster throttle response without giving up any top end, there is no better solution. With the added benefits of higher turbine efficiency, lower cylinder temps and EGTs which allow more aggressive timing and fuel mapping, and the freedom to run more overlap,twin-scroll turbo system design is really a perfect match for the high specific output engines featured in many of our favorite sport compact machines.

Send your feedback to dpratte@modified.com

[Dmitriy_Guller]

mariner, on Jan 30 2011, 04:52, said:

3) If these engines can give constant torque from 1100 to 4500 rpm etc. why do they need 6, 7 , 8 or even 9 speed auto gearboxes? Surely with so much torque the weight and cost of multi gear auto boxes should be unecessary?

From a performance perspective, flat power curve is what minimizes the need for gear changes, not flat torque curve.

[mariner]

primer " The benefit is you can ride the torque in overdrive gears as much as possible, thus get better mpg numbers. If you can have peak torque at 1250 RPM (seems so odd to type that), no reason to keep the engine at higher RPMs while crusing at steady speeds is there? "

I can see that point if there is a point of peak torque but the BMW engine develops max power @ 5,000 rpm and the torque ( per the graph above ) is flat from 1,250 to 5,000 rpm or 75% of full power rpm. So compared to a conventional power/torque curve there is no gearing optimisation to be gained from more gears. Trucks have multi gear boxes to keep revs at the peak efficiency point but they do not ( I beleive) have such a flat torque curve. For example the latest 16.4 litre Scania engine has torque flat from 1,000 to 1,350 rpm and max power at 1,900 rpm so its flat torque curve is only over 20% of max power rpm and not 75% as with the BMW engine.

So what is the advantage of lots of gears if you have constant torque over 75% of teh required rev range? I am not saying there is no benefit but I can't see it.

[primer]

mariner, IMO this would be the advantage of many close ratios (with a couple sacrificed to overdrive):

80-120 kph (4th/5th) in 6.0/7.1 seconds

If there are fewer ratios the RPM drop between gear changes would be higher, this will reduce performance^*. Surely revving from 2000 to 5000 RPM cannot be quicker than revving from 3000 to 5000 RPM, flat torque curve or not.


^*: Performance that 95% of the people will not need/use 95% of the time but is perhaps important to win the media contests and nurture the brand's 'power' image.

Edited by primer, 30 January 2011 - 22:23.

[Dmitriy_Guller]

mariner, on Jan 30 2011, 16:41, said:

So what is the advantage of lots of gears if you have constant torque over 75% of teh required rev range? I am not saying there is no benefit but I can't see it.

Acceleration would be an obvious advantage.

[gruntguru]

As Dmitriy has pointed out, it is the region of constant power that determines the gear spacing required for best acceleration. In this case with constant power from 5,000 to 6,600 rpm the maximum ratio spacing should be 6600/5000 = 1.32 to avoid falling out of the max power region. A six speed box for example with this spacing would have an overall ratio span of 1.32^5 = 4.0.

Whiteblue, the twin scroll system is a divided inlet turbine system which utilises blowdown energy to produce more turbine energy or lower exhaust back pressure. This technology has been around for a long time. Although it can be used to produce boost from lower engine speeds, it does not reduce the need for a broader compressor map (quite the opposite in fact) so the map I posted above would not cover the range required for the BMW.

[Wuzak]

gruntguru, on Jan 31 2011, 02:25, said:

Whiteblue, the twin scroll system is a divided inlet turbine system which utilises blowdown energy to produce more turbine energy or lower exhaust back pressure. This technology has been around for a long time. Although it can be used to produce boost from lower engine speeds, it does not reduce the need for a broader compressor map (quite the opposite in fact) so the map I posted above would not cover the range required for the BMW.

How do you get a broader compreesor map? Is it just detail design of the impeller, or do you need some sort of variable geometry to help with that?

[gruntguru]

Wuzak, on Jan 31 2011, 15:36, said:

How do you get a broader compreesor map? Is it just detail design of the impeller, or do you need some sort of variable geometry to help with that?

For turbo size compressors, it is impeller and housing design. As I mentioned in a previous post, there will be a tradeoff of max PR or peak efficiency.

[WhiteBlue]

gruntguru, on Jan 31 2011, 00:25, said:

Whiteblue, the twin scroll system is a divided inlet turbine system which utilises blowdown energy to produce more turbine energy or lower exhaust back pressure. This technology has been around for a long time. Although it can be used to produce boost from lower engine speeds, it does not reduce the need for a broader compressor map (quite the opposite in fact) so the map I posted above would not cover the range required for the BMW.

I have asked myself who supplies both twin scroll and VTG technology and found Borg Warner does.



I think that they may have combined the two technologies in one design. At least logic would tell me I would try to combine this. I would not be keen to invest much cost and design resources to get a highly efficient twin scroll system to regulate it with a waste gate.



This graph shows that BW supplies both technologies in the range from 150-220 kW. Is there a reason why they should not be combined?

[Wuzak]

gruntguru, on Jan 31 2011, 03:25, said:

Whiteblue, the twin scroll system is a divided inlet turbine system which utilises blowdown energy to produce more turbine energy or lower exhaust back pressure. This technology has been around for a long time. Although it can be used to produce boost from lower engine speeds, it does not reduce the need for a broader compressor map (quite the opposite in fact) so the map I posted above would not cover the range required for the BMW.

Grunt, from the blurb WB posted above I gather that the twin scroll system is of benefit to an L4, or a flat plane V8 (one turbo per bank) because of valve timing and firing order. I take it that there would be little advantage for an L3 or twin turbo V6?

[WhiteBlue]

Wuzak, on Feb 2 2011, 06:27, said:

Grunt, from the blurb WB posted above I gather that the twin scroll system is of benefit to an L4, or a flat plane V8 (one turbo per bank) because of valve timing and firing order. I take it that there would be little advantage for an L3 or twin turbo V6?

I agree that an L3 should be easily run with a single scroll turbo, but the V6 is a prime candidate for a twin scroll turbo. BMW does it with their N55 L6 and their S63 V8 engines. For V6 it should be even better because it should save the weight of a second turbo and allow the placing of the single twin scroll turbo in the V between the cylinder banks.

[gruntguru]

WhiteBlue, on Feb 2 2011, 16:22, said:

I agree that an L3 should be easily run with a single scroll turbo, but the V6 is a prime candidate for a twin scroll turbo. BMW does it with their N55 L6 and their S63 V8 engines. For V6 it should be even better because it should save the weight of a second turbo and allow the placing of the single twin scroll turbo in the V between the cylinder banks.

"Plumbing" is the issue on a V6. The exhaust tubes become too long and tortuous.

[gruntguru]

WhiteBlue, on Jan 31 2011, 19:39, said:

This graph shows that BW supplies both technologies in the range from 150-220 kW. Is there a reason why they should not be combined?

Yes there is. The twin scroll design needs to keep the two exhaust streams separate all the way to the turbine wheel. Interposing two sets of movable vanes while maintainig high efficiency presents a significantly greater technical challenge.

[WhiteBlue]

gruntguru, on Feb 2 2011, 10:04, said:

"Plumbing" is the issue on a V6. The exhaust tubes become too long and tortuous.

I would use the inside ports and make the scrolls clockwise and anti clockwise. It may be more difficult to cast but it would make for a much lighter piping.

gruntguru, on Feb 2 2011, 10:09, said:

Yes there is. The twin scroll design needs to keep the two exhaust streams separate all the way to the turbine wheel. Interposing two sets of movable vanes while maintainig high efficiency presents a significantly greater technical challenge.

Why would a twin scroll TC need two different sets of vanes? The sets of cylinders are always running with synchronous speed. So there is no need for different setting of the vanes. Only the piping and the scrolls need to be adjusted for the correct aerodynamics.

[gruntguru]

WhiteBlue, on Feb 3 2011, 02:08, said:

Why would a twin scroll TC need two different sets of vanes? The sets of cylinders are always running with synchronous speed. So there is no need for different setting of the vanes. Only the piping and the scrolls need to be adjusted for the correct aerodynamics.

You need a set of vanes in each passage. Yes they would not need independent control.

[WhiteBlue]

gruntguru, on Feb 3 2011, 00:10, said:

You need a set of vanes in each passage. Yes they would not need independent control.

So why not use the same vanes with sufficient width that both scrolls can enter the same set of vanes? That would be natural to do.

[gruntguru]

WhiteBlue, on Feb 3 2011, 18:20, said:

So why not use the same vanes with sufficient width that both scrolls can enter the same set of vanes? That would be natural to do.

You would be combining the streams before the turbine wheel. There would be some loss (don't know how much) of blowdown energy.

[WhiteBlue]

gruntguru, on Feb 2 2011, 10:04, said:

"Plumbing" is the issue on a V6. The exhaust tubes become too long and tortuous.

Quote

Source
New aluminum 2.8V6 Turbo producing 250 bhp/184 kW, exclusive to top-of-the-line Aero, or 230 bhp/169 kW in Arc and Vector variants (EU only). 24-valves, DOHC, variable valve timing, twin-scroll turbocharger feeding both cylinder banks. Hydroformed stainless steel exhaust manifolds with air injection for quick cold start warm-up.

It looks like GM is doing it already on their new Saab model.

[WhiteBlue]

gruntguru, on Feb 3 2011, 12:10, said:

You would be combining the streams before the turbine wheel. There would be some loss (don't know how much) of blowdown energy.

Using a waste gate would also mean loosing some energy. I don't see a particular geometric disadvantage for a twin scroll vs a single scroll by using one set of variable vanes. Merging the flow at the variable vane entry point should not be much different to merging them at static vanes.

[Scotracer]

WhiteBlue, on Feb 3 2011, 14:45, said:

Using a waste gate would also mean loosing some energy. I don't see a particular geometric disadvantage for a twin scroll vs a single scroll by using one set of variable vanes. Merging the flow at the variable vane entry point should not be much different to merging them at static vanes.

But it's not possible to get around a waste-gate, unless you want to build an intake capable of insane pressures and don't mind your engine blowing up each time the you go from closed throttle to WOT.

[WhiteBlue]

WhiteBlue, on Feb 3 2011, 15:45, said:

Using a waste gate would also mean loosing some energy. I don't see a particular geometric disadvantage for a twin scroll vs a single scroll by using one set of variable vanes. Merging the flow at the variable vane entry point should not be much different to merging them at static vanes.

Scotracer, on Feb 3 2011, 16:33, said:

But it's not possible to get around a waste-gate, unless you want to build an intake capable of insane pressures and don't mind your engine blowing up each time the you go from closed throttle to WOT.

Systems with VTG and without waste gate are standard in diesels AFAIK. No reason they should not work in gas engines with some careful thermal management. And as I discussed above they should also work with twin scroll. So far I haven't heard an adamant point against that idea.

[gruntguru]

WhiteBlue, on Feb 4 2011, 00:45, said:

Merging the flow at the variable vane entry point should not be much different to merging them at static vanes.

A typical twin scroll housing doesn't have vanes. The two flows are kept seperate all the way to the turbine wheel.

Edited by gruntguru, 03 February 2011 - 23:08.

[J. Edlund]

Wuzak, on Jan 30 2011, 12:46, said:

So they don't need a wastegate at all?

So far there is only one gasoline engine that uses variable nozzle area instead of a wastegate, which can be found in the Porsche 911 Turbo.

gruntguru, on Jan 30 2011, 12:53, said:

To produce a flat torque curve the turbo needs to supply approximately constant boost ie a horizontal line on the compressor map. The surge line defines the low flow (low rpm) limit and the upper limit will usually be determined by falling efficiency (and sometimes by turbo speed limitation in higher boost applications). The width of the efficiency island can be tailored to the application - usually at the expense of a lower peak efficiency. The BMW torque range is about 4:1 which is pretty extreme. It could be achieved using the turbo in the map below at a PR of 1.8, but the power extending to 7,000 rpm would pass the 60% efficiency line on this map. So the BMW engine would require a broader compressor flow range or staged multiple turbos.

You have forgotten what impact the changing volumetric efficiency of the engine will have on what pressure ratio is needed to produce a flat torque curve. Usually the turbocharger and engine is designed in such a way that the pressure ratio can be decreased with increased engine speed as the engine is capable of reaching higher volumetric efficiencies by conventional means.

With direct injection and variable valve timing the valve overlap can also be increased at low engine speeds. This allows some charge air to pass though the engine which can be helpful to increase compressor flow and prevent surge at low speeds. So if you just have enough turbine power, increasing the compressor flow isn't that difficult.

gruntguru, on Jan 31 2011, 00:25, said:

As Dmitriy has pointed out, it is the region of constant power that determines the gear spacing required for best acceleration. In this case with constant power from 5,000 to 6,600 rpm the maximum ratio spacing should be 6600/5000 = 1.32 to avoid falling out of the max power region. A six speed box for example with this spacing would have an overall ratio span of 1.32^5 = 4.0.

Whiteblue, the twin scroll system is a divided inlet turbine system which utilises blowdown energy to produce more turbine energy or lower exhaust back pressure. This technology has been around for a long time. Although it can be used to produce boost from lower engine speeds, it does not reduce the need for a broader compressor map (quite the opposite in fact) so the map I posted above would not cover the range required for the BMW.

Without twin scroll it have also been common to use for instance 4-2-1 exhausts on turbocharged four cylinders, the benefits are similar to using twin scroll, but obviously the "leak" between the two sides are much larger when using a 4-2-1 exhaust and a single scroll turbocharger.

Wuzak, on Jan 31 2011, 06:36, said:

How do you get a broader compreesor map? Is it just detail design of the impeller, or do you need some sort of variable geometry to help with that?

The angle of the impeller blades are one factor, increased back angle tends to give a wider range at the cost of some higher turbocharger speed for a given pressure ratio. Another possebility is to give the compressor housing map width enhancement grooves (also refered to as inducer recirculating bypass). Close to surge some flow will then flow out of the impeller and back to the inlet via the groove and near choke some air will flow into the impeller via the groove.

Wuzak, on Feb 2 2011, 06:27, said:

Grunt, from the blurb WB posted above I gather that the twin scroll system is of benefit to an L4, or a flat plane V8 (one turbo per bank) because of valve timing and firing order. I take it that there would be little advantage for an L3 or twin turbo V6?

Twin scroll is an advantage when the engine have overlapping exhaust pulses.

WhiteBlue, on Feb 3 2011, 15:38, said:

It looks like GM is doing it already on their new Saab model.

It is still a compromise. To take advantage of the blowdown energy, the volume of the exhaust manifolds need to be small. When you're fitting one twin scroll turbo to a V6, this will require longer exhaust pipes, which increase the volume between the engine and the turbine. Using two turbochargers instead will give a smaller exhaust volume and less turbocharger inertia.

That engine was by the way introduced 2006 or so.

WhiteBlue, on Feb 3 2011, 19:17, said:

Systems with VTG and without waste gate are standard in diesels AFAIK. No reason they should not work in gas engines with some careful thermal management. And as I discussed above they should also work with twin scroll. So far I haven't heard an adamant point against that idea.

Diesels operate with much lower exhaust temperatures and increased engine load have a much smaller impact on the exhaust mass flow than with a gasoline engine which makes it a more suitable application for variable geometry turbines. Infact, diesels have in the past often used fixed geometry turbochargers without wastegates. A diesel doesn't have the same need to control the boost as a gasoline engine since the fuel flow doesn't follow air flow. Adding a wastegate, and later variable geometry, to the diesel engine turbocharger have however been helpful to improve low speed boost as the turbine and turbine housing can be made smaller.

[gruntguru]

J. Edlund, on Feb 6 2011, 08:34, said:

You have forgotten what impact the changing volumetric efficiency of the engine will have on what pressure ratio is needed to produce a flat torque curve. Usually the turbocharger and engine is designed in such a way that the pressure ratio can be decreased with increased engine speed as the engine is capable of reaching higher volumetric efficiencies by conventional means.

With direct injection and variable valve timing the valve overlap can also be increased at low engine speeds. This allows some charge air to pass though the engine which can be helpful to increase compressor flow and prevent surge at low speeds. So if you just have enough turbine power, increasing the compressor flow isn't that difficult.

If the VE is optimised for high rpm as suggested in your first paragraph, it might be more "difficult" than you suggest to scavenge during overlap at low rpm, as this requires a positive pressure differential acoss the engine. Even more difficult with VTG and single scroll meaning no capture of blowdown energy.

[Powersteer]

combining the two can be made effective but each of the twin scroll would have its own set of vanes. the twin scroll relies heavily on the wall that splits the exhaust until the turbine wheel so that the exhaust gas speed does not get interrupted when one scroll is in use so this set has its own vanes within its own scroll.

[WhiteBlue]

To summarise the discussion: We are much more likely to see twin scroll TCs with waste gate in gasoline engines. VTG would require twin vanes and have thermal problems. Perhaps future systems will develop in that direction.