Turbocharger/supercharger Compressor Type

[DannytheRadomski]

So I know nearly all turbochargers use centrifugal compressors, and while a some superchargers are centrifugal, most people know of superchargers as positive displacement. I was reading into axial compressors (like a in a jet engine) and according to what I was reading they are the most efficient compressors. As I read more the consensus seemed to be that centrifugal pumps are better at producing high pressure, but axial compressors have a higher airflow. I was thinking about the possibility of using these in tandem, an axial compressor followed by a centrifugal one. Any thoughts on compressors (especially axial ones) are welcome and appreciated.

I did find axial superchargers online, but there were no dyno comparisons to centrifugal superchargers. None of the info was very good, so I was hoping you guys might know a little bit.

[jd74914]

Just some quick thoughts since I'm from the land of airplane engines...

If you want to learn about this stuff you'll have to read a turbomachinery book. Just note that the subject is not trivial; typically it's a senior-plus class for undergrads or a full grad student class. I think the best books on the subject are by Mattingly and Baskharone.

Selection of turbomachine type is probably most easily done by Cordier Diagram (look on google, there is too much to write here) and deals with the mass flow, pressure rise, speed, and other normalized sizing constants. Coupling of axial stages to centrifugal stages is pretty rare; the only machine I'm aware of that does this is the PW150A turboprop motor. In terms of airflow, an automotive supercharger is at the bottom of the barrel (remember, even the largest automotive engines are orders of magnitude smaller than aircraft engines, and have orders of magnitude higher pressure rise requirement that many industrial combustion blowers), so while I haven't actually calculated position on a Cordier Diagram, I'd be very surprised is a centrifugal compressor was not the correct choice. Note that they can be staged like axial rotors, they just see a little more pressure drop in the guide vane and ducting after each stage.

Also, note that when looking at efficiencies, axial compressors typically enjoy max efficiencies only a few percent higher than centrifugal. Inefficiency in a blower pales in comparison to other SI engine inefficiencies (friction, etc.). The difference is so negligible I don't think you would ever see it in any dyno plot.

[Racer-X]

My take on this is you end up having to weigh throttle response with efficiency and peak power. A positive displacement supercharger isn't very efficient but offers the best throttle response (and I love them for that), versus a large turbo that is going to take forever to spool up and is relativity efficient once it builds boost. You end up scarfing throttle response and power (in a transient sense) as you shoot for more efficiency and more peak power. Think of a compound engine, very efficient for a piston engine but is the least rev happy thing I can imagine.

I could see an axial compressor, or some sort of two stage set up, on a generator or ship engine that runs at one efficient speed its whole life. I don't think all that inertia is going to be good in an auto-x car's powertrain.

[RenM]

people rule out superchargers as a viable option way to soon. Especially given the circumstances in FSAE.

- Modern Superchargers do have a very good efficiency that is no worse then radial compressors in the FSAE mass flow range.
- They don't have any surge problems. You can use the complete map and create a pressure rise from almost zero mass flow on.
- immediate throttle response is very important on the very small straights

Available turbochargers like the gt12 do have a bad system efficiency combined with a small engine, because they create large pumping losses particularly during acceleration.

If you do the calculations you will be surprised about the efficiency comparison of a supercharger vs. a turbocharger in FSAE.

Adding to jd regarding the original question of the axial compressors:
Unless you have variable guide vanes the axial compressor does achieve the high efficiency only in a very small mass flow range. Volutes of a radial compressor are much better suited to the varying mass flows seen in internal combustion engines.
I too don't see any potential for an axial compressor.
There is however potential for an axial turbine. Because the turbine typically makes for 2/3 of the turbochargers inertia, using an axial turbine can greatly reduce your inertia and allow a quicker throttle response. The high pressure ratio of a centrifugal turbine is not needed in FSAE application. However designing such a turbine is probably beyond possibilities for fsae teams and not worth the time invested.

[Smeller]

What do you think about something like this? It should be 12v instead and you could get by with a slightly larger battery and what you can get extra from the alternator.

http://www.phantomsuperchargers.com/fts-tq18024v.html

http://www.ft86club.com/forums/showthread.php?t=39719

[Drew Price]

That takes over 4 horsepower to drive and is only capable of a pressure ratio of 1.3.... sooooo.... (At 24VDC and 3,000 watt consumption it take 125 amps to drive that thing. Even a very larger battery wouldn't last long with that draw. At 12V it would take 250A.)

Also that looks way too small to be a 3,000 watt capable electric motor.

I don't think I'd give it much thought.

[Drew Price]

Having now glanced thru that FRS/BRZ thread it's an interesting way to do it, could probably be made viable. A few PSI on such a small FSAE engine without the drawback of the oil control problems with an exhaust driven turbocharger is at least decent justification.

[RenM]

Well there are very capable and small electric motors from hobby rc applications. The main problem i see here is the map of the compressor used. It only fits to way bigger engines.

[Z]

Danny,

RenM's post (half way up page) answers most of your questions. I assume the modern superchargers RenM is talking about are the "screw type" variations on the traditional Roots blower. These are compact, efficient, give great low-down, instant torque, are relatively easy to fit (cf. turbo), and I am surprised they are not more common in FSAE. Their main disadvantage is that they draw power from the engine (more than, say, a turbo, though turboes can impede the gas flow through the engine), but this is somewhat offset by lower frictional losses because the engine doesn't have to spin as fast for same power.

The electric-radial(centrifugal)-blower above most definitely DOES DRAW POWER from the engine, at least over the course of an Enduro event. As noted, the 3 kW motor will quickly flatten its batteries (unless they are E-car sized!), so an upsized alternator is going to have to work hard to keep the batteries topped up. My guess for round-trip efficiency of IC-engine->alternator->batteries->electric-motor->blower is well below 50% (maybe <30%?). In other words, not very good for Fuel Economy.

An alternative to the electric-motor driven radial-blower is simply to drive a radial-blower mechanically from the engine. This is quite common and requires a multi-stage drive, because the blower has to rev to 100k+ rpm. Usually there is a toothed rubber belt in there somewhere to smooth out any crank impulses. This approach is more efficient than the electric-drive (+alternator+++), but gives an increasing boost curve with engine speed, rather than the constant boost of Roots or always-on electric-blower. Some people prefer the extra boost only at higher revs.

Finally, (as pointed out by RenM) an axial-fan can have great efficiency, but usually only over a small operating range. Also, for any significant boost in IC engine terms, the axial-fan will have to be multi-stage (a series of impellor-stator-impellor-stator+++), which makes it more difficult/expensive to manufacture.

The earliest jet engines (= gas-turbines) used single-stage radial-blowers to compress the air, and sometimes also radial-turbines to extract the energy from the heated air (so they looking a lot like a car turbo). I think one of the main reasons they later went to all-axial designs was to reduce the frontal area of the engine, for lower aero drag (plus, of course, the slightly better efficiency of the axial design). However, I believe some modern, small gas-turbines are going back to the all-radial design, because overall it is more compact and simpler, and therefore also cheaper!

Sometimes "optimum efficiency" isn't everything. :)

Z

[murpia]

Danny,
However, I believe some modern, small gas-turbines are going back to the all-radial design, because overall it is more compact and simpler, and therefore also cheaper!
Z
This is especially true of turbo-shafts for helicopters, which don't need to be concerned with frontal area, or having their air intakes at the front.

http://www.mtu.de/en/products_services/military_business/programs/mtr390/index.html

Regards, Ian

[Z]

Ian,

Thanks for the link. The PDF file shows it as a two stage radial (centrifugal) compressor, followed by two (???) axial stages for the turbines (I guess one to drive the compressors and the other to drive the shaft, but can't see too clearly).

Interesting is the quoted specific fuel consumption, at about 280-300 g/kWh. Aside from the usual Engineering industry's abomination of units ("kilowatt-hour" for energy!!!, with "hours" being the standard units from bus-timetables!!!) this fuel efficiency is quite poor. Translated into more sensible units of "percentage of fuel-energy converted into useful shaft-energy" it comes out at about 25-30% (depending on energy content of the fuel). This is mediocre by petrol engine standards, and piss-poor compared with diesels.

Of course, 1,000+ kW from <180 kg does sound good. But when you add the mass of fuel required for longer journeys, then fuel efficiency quickly becomes more important (ie. consider the extra tonnage of fuel to lug around). Essentially, the gas-turbine is a simple, cheap way of getting a lot of power from a reasonably small and reliable bit of machinery (because low parts count), albeit at rather poor fuel efficiency.

As a warning to you students, jet-engine spin-doctors often quote the thermal efficiencies of their aeroplane engines as being better than diesels (ie. typically >45%). However, the jet-engine calcs are done at ambient (= exhaust) temperatures of something like -50 C (it's cold up there above the clouds!). Run a diesel up there and it will be even more efficient....
~~~~~o0o~~~~~

Because it is bucketing down rain outside ..... here is one sexy-arsed bit of machinery for you all to consider ... :)

http://www.madviolinist.com/myblog/wp-content/uploads/2010/11/Auburn-851-SC-Speedster_3.jpg

This ~1934 Auburn 851 Speedster has an all-alloy, flat-head (= side-valve), straight-eight engine, with a radial-flow (centrifugal) supercharger driven mechanically from the engine. You can see the SC just under the carburettor. (SC has vertical shaft, so horizontal shaft driven by crank nose runs next to engine, then 90 degree bevel-gearbox, and shaft up to SC?)

http://www.wallpaperup.com/uploads/wallpapers/2013/09/11/146429/47a6eb7bab322f6058b9df5f3f1b2171.jpg

A centrifugal-blower like this is a good match for a side-valve engine. At low revs the lazy, large capacity engine gives easy, tractable, and quiet cruising (side-valves are very quiet). At higher revs the small valves start to restrict airflow, but that is just when the blower kicks in. Hence a steady increase in torque and power...

Z

[d.haddon]

Z,

In addition to the reasons you quoted above, radial compressors have seen little use in aircraft engines due to very high weights and gyroscopic forces. As engines were upscaled for more power, these rotors became much more heavy, (partly to cope with the very large pressure ratio, orders of magnitude higher than seen by a single stage of an axial compressor) and the very high rpm of jet engines led to much greater gyroscopic forces forces being experienced.

The smaller radial compressors don't have this problem, and when scaled down, the final stages of axial compressors have VERY thin blades, so moving to radial compressors (and turbines) can increase simplicity and reliability. When considering the size of an automotive application, it seems like there is much less that can go wrong with a single-stage radial blower versus a multi-stage axial compressor.

Regards, Damon.

[VFR750R]

Don't forget about inertia. Multi-stage axial compressors work fine at steady state, but have much higher inertia than a single stage radial compressor working at higher pressure ratios.

I also seem to remember that axial compressors are only more efficient than radials beyond a certain size. At really low airflows which necessitate a very small inlet size and wheel size, an axial loses a ton of efficiency out at the tips of the blades (sealing) and the fact that there is such a surface speed gradient from the base to tips of the blades.

Small compressors of both types lose efficiency due to a number of effects such as higher operating speeds, and higher surface area to flow rate, and higher leak percentages.

[MCoach]

I currently have nothing to contribute to this thread (and hope to not derail it!), but Z, I would like to congratulate you on your 1000th post on this forum. That's a lot of words! I ..err... have no sort of prize, but maybe a trip to the local ice cream shop could be good enough.

[Big Bird]

Wow, Z, congrats
(celebrating the occasion with maybe my shortest post ever....)