two small turbos better than one big turbo(?)

[Artsi]

I've been pondering whether two small turbochargers would offer any benefit over one large turbocharger. I've ploughded throught the net, and found some material suggesting that two small turbo's are a better way to go than one large turbo.

Please have a quick peek at linked forum thread. I do welcome any commentary over this subject on this eng-tips forum. Thank's.

http://www.dieseltruckresource.com/dev/showthread.php?t=104290

 

[Fabrico]

One advantage is that, although in total "twins" take more space, each side can usually be crammed into a smaller area.

A disadvantage is that it cost about twice as much.

Operationally; they've been around a long time and seem to work very well on many different engine arrangements.

 

[gijim]

I don't know if this is true or not, but I have heard that originally, twin turbos were "good" because they could spin up faster than one big turbo.

However, apparently today's big turbos spin up fast enough already, especially with ball bearing center sections.

 

[hvychvy81]

I believe toyota and nissan run one smaller turbo and one larger on their twins setups. This allows for the most power through almost all rpm ranges.

 

[Artsi]

I know of Toyota's system. They shut down other turbo at lower revs, so it'll receive all of the available exhaust gas flow. The remaining turbo will jump in as engine revs clime up.

BMW's latest 335i does not have this above mentioned arrangement. It just simply has two small turbo's working in perfect harmony, parallel to one another.

As mentioned in above linked forum discussion, when turbo size is reduced by 50%, the inertia is reduced 75%.
In other words, half the tubo size, have only quarter of inertia forces.

If this math holds, there is indeed great sense to BMW's logic in bolting on two small turbochargers instead one really big turbocharger.

Can anyone verify the inertia math assumption?

 

[gijim]

This is true, but you have to look at the turbine and compressor diameters. For 2 turbochargers, one that flows X and one that flows 2X, the diameters of the wheels are not necessarily proportional. The one that flows twice as much may only be 50% larger in diameter.

Additionally, the drag from the oil film bearings is likely equal in most turbochargers, since they all use similarly sized bearings.

So... once you get all that extra stuff added up, you might see that you don't really gain much if anything. The problem is you can't really model some of it, so without an actual installation you'll probably never know.

 

[Warpspeed]

It really depends on a whole lot of issues, and you need to make a value judgement.

First thing is that half the exhaust energy fed through each of two exhaust turbines with inertia proportional to the square, and flow proportional to the cube is not going to (in theory) produce less lag. But it may actually do so, if the alternative requires a massively large total exhaust manifold volume to feed a single central turbo. For boxer or Vee engines, twins may be better for response, for an inline engine, perhaps not.

The next thing to think about is where you plan to opearate on the compressor flow map. All compressor flow maps tilt to the right, available flow increases as pressure ratio increase.

With only one turbo, if run at a relatively high pressure ratio, a single mid sized turbo may well do the job.

But if you requirew vast volumes of air at low boost pressure, multiple smaller turbos are the most practical way to achieve it.

For example a four cylinder engine producing 400Hp will need to run high boost pressure, and a single turbo would be best suited.

A big block V8 in a boat producing 400Hp would only require low boost pressure. Twins would be a much better choice. Now you may think that a 400 Hp big block really doesnt require a turbo. But if you wanted a genuine continuously rated 400Hp at modest Rpm, it is really a far better solution for many reasons than a much higher Rpm normally aspirated engine.

It is just a case of weighing up the alternatives for a particulr application. If it doesnt make much difference, a single turbo will always be cheaper.

 

[Artsi]

Simulation of turbocharged SI-engines
- with focus on the turbine

Doctoral thesis - Fredrik Westin

KTH School of Industrial Engineering
and Management TRITA – MMK 2005:05
Royal Institute of Technology ISSN 1400-1179
SE-100 44 Stockholm ISRN/KTH/MMK/R-05/05-SE



1.3.2 Sequential systems
A more extreme case of variable geometry is sequential turbocharging. Instead
of varying the geometry of one charger the number of chargers is changed.
The reason for using sequential turbocharging is to widen the flow range for
the boost pressure required. Since still only one compressor stage is used it
does not necessarily increase the boost pressure.
Volvo tested sequential systems presented in an SAE paper in 1991 [1.26] as a
part of a methanol engine program. The engine was a six-cylinder unit and the
boost pressure target was as usual, slightly below 2 Bar abs. They state that a
parallel turbo system has around 30% lower inertia than a single turbo with the
same top end flow characteristics, and thus should have a benefit in terms of
response.

 

[innovator]

Interesting Artsi that you quote Fredrik Westin as I have spoken to him (I judged his car at Formula student and a year later he was in my judging team)about twin turbos. He said that 2 small turbos were the way to go for general all round use. He was talking about having one turbo on 2 ports and the other on the other 2 ports. They would spin up and get on boost a lot quicker yet still give reasonable top end power.

I did not really get into any more technical details at the time.

John

 

[Artsi]

I thank you John. That made my day

 

[Warpspeed]

Thirty percent less inertia maybe, but will thirty percent less inertia spin up faster with fifty percent less exhaust flow ?

 

[Artsi]

I understand it's 30% less inertia opposed for all available exhaust flow.
Otherwise quicker spool theme would not materialise.

Do we agree(?)

 

[Warpspeed]

Yes, but you cannot add the inertia of two small turbos. For spool time, each operates quite independantly and each will only have half the total exhaust energy to work from.

That is the problem. A turbo half the size will have one quarter the internal flow area and seventy percent of the inertia.

As you scale downwards, flow falls much faster than inertia.

Twin turbos will not be that much smaller physically than a single. Try doing some actual sizing to see that this is true. The inertia of each turbo will certainly be less, but not that much less than a single. The killer is that only half the energy is available to run each single.

 

[mgray10]

As related to V type, canted cylinder engines, two turbochargers usually offer a distinct advantage over one.

From above, it was briefly touched upon that two turbochargers are usually smaller per unit, but indeed larger when summated and compared to a single. This will, in many cases, permit close proximity installation with regard to the exhaust runner or port. On V engines, you will usually find the turbocharger manifolds or tubular headers feeding directly into the turbine housing. This distance is usually 24 inches or less of linear manifold port length. Comparatively, single unit installations are usually predisposed to having much longer expanses of tubing before reaching the housing inlet.

For example, the single installation requires that both manifolds merge together into one pipe. In the compactness of modern engine compartments, this is usually done behind or under the engine. In rare instances, the manifolds or headers are aimed forward and have a short distance (maybe another 24 inches) to traverse where they meet symmetrically at the housing flange. When this is not possible, a large pipe is used in accordance with maximum flow requirements.

Just as diameter of a tube has a seemingly skewed relationship with maximum flow capabilities (when looked at by the uninformed) so does minimum flow. By using that large pipe, you lose a lot of velocity whilst not at peak flow. With two tubes sized for maximum flow/2 and half the exhaust flow, the difference between minimum flow and maximum is not as great. This can be modeled by using the cross section of the two tubes and a common length.

As was also mentioned, the broad flow and performance required of gasoline engines (compare RPM to that of a diesel) requires the same in a compressor map. Because of the limited low-flow capabilities of these compressors, using two effectively moves the flow requirement higher in the compressor map. As you can see, the compressor map is usually much more efficient at higher pressures and flow rates than low pressures and flow rates. If you size the twins too large, you forfeit this advantage. Think of it this way, if you ran 4 respectively smaller turbos, they would provide better efficiency down low and roughly equal efficiency mid and high in the map versus two, and two would do the same versus one. So, four is that much better than one. With simple regard to the compressors, this progression goes on and on. When you start factoring in real world constraints and duty efficiency, you see why people use one, and not four.

Hope this helps a bit.

 

[samplesecond]

A good rule of thumb is to follow the path of least resistance-ie, simplicity...a clean and simple solution has never failed me.

With regards to the twin vs. single arguement, I think you really need to evaluate the project as a whole-outlining objectives etc.

As mentioned above, twin turbochargers will have a total inertial advantage whilst yielding similiar top end performance-hence quicker spool..but thats only part of the story.

Something that MUST be emphasized is system reliability. Twice as many components, twice as many potential issues.
Unless space constraints require the use of two smaller turbochargers (v-engines)..use a properly sized single, with an appropriately designed ex/int manifold, and be done with it.

 

[SPOrotary]

Technically a single turbo is more efficient. several things are in it's advantage: less surface area for heat to escape before it enters the turbine, better flow/inertia ratio, less total system weight, higher compression efficiency possible, higher air pressure/flow rates for the backpressure it creates, the entire system can be made shorter and lose less pressure over the whole system.

But there is one big disadvantage: the single big turbo only has all these advantages over a very small operating range.

Where as a multi-turbo system misses a percent or two in overall system efficiency it more than makes up for that by achieving close to peak efficiency over a wider operational range. This is the key to why multi-turbo systems can achieve better drivablity in real world applications.

Current project car: '98 BMW 316i
Main targets: 950kg, 250rwhp, min. 1G lateral grip. Special: Twincharger system, hydraulic ebrake, New suspension.

 

[Majik]

In regards to wondering why BMW chose parallel twins on their new engine. I got a look at one of those cars this morning. There's no room for a single. That's it. I honestly think you might have to pull the engine to change a turbo. Reminds me of a z32 300zxTT engine bay where everything was assembled by umpaloompas and they're all out of stock of those at the dealership. So as fabrico mentioned. Packaging is the reason they go sometimes and this is definitely one of those times.

 

[patprimmer]

Also, a "twin turbo" badge has marketing appeal over just "turbo" badge

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[wgknestrick]

All you have to do to answer this question is look at what people do with specific engines and follow these rules

1. Keep it simple
2. 1 turbo is always better than 2 unless it adds unnecessary piping

Look at the NOPI series to see what wins out there as it is amazing what they get HP wise out of those engines and most seem to use 1 big turbo. You add about 50lbs extra for each turbo + piping you install and you never really get that extra HP or response out of it.

In general:
V-engines - twin turbo
anything else (including boxer engines) - single

Drive-ability comes from proper turbo sizing more than anything

 

[SMOKEY44211]

My vote is for single. Reynolds numbers are better. Over all power output potential higher. 1/2 cost and complexity (is that a word?). Packaging in some cases may tip the scale in favor of dual setup and inertia considerations for spool up time. At the risk of sounding like a broken record " all engineering is a compromise". I would be interested to know what configuration you decide to persue.---------Phil