> I guess I'll have to jump in on this one. I did quite a bit of
> research
> and experimentation on turbos. I have real experience with running
> them
> up to 18000 ft. My plane has a total of 150 hrs of turbo time and 50
> hrs N/A time. I really should update my website with a discussion of
> turbo selection so I don't have to keep re-typing this...
>
> I agree that the TO4B series is probably the best choice for size,
> cost,
> and availability. Paul is right that the stock turbo will produce
> sufficient power, but at that power level it is over speeding and
will
> not last long. That cannot be fixed by putting on a larger waste gate
> alone. Turbo speed(in a given turbo) is simply a function of mass
flow
> ( i.e. desired horse power) and pressure ratio. That is regulated by
> the throttle (i.e. how much h.p. you use), not the waste gate. A
> bigger
> waste gate will only reduce the back pressure created for a given
> position on the compressor map. But given an altitude and power
> output, changing the waste gate will not affect turbo speed, surge,
> stall, over speed, or thermal efficiency.
>
> The waste gate only determines what portion of the exhaust energy is
> flowed through the turbine. So the idea is to feed the turbine only
> the
> energy that is needed to create the desired boost. Anything more just
> increases the back pressure and reduces the efficiency of the system
> (different than compressor thermal efficiency).
>
> In my plane, only the stock turbine housing will fit. It is amazingly
> compact. I originally ran the stock series 5 turbo. It was good
for 70
> flight hours, of which only about 15 were high altitude/ moderate
> boost
> hours. It died. My replacement (current) turbo is the TO4B 60-1
> compressor attached to the stock turbine housing that has been
reamed
> to
> accept a TO4 p-trim turbine. After 80 flight hours, including at
least
> 40 hrs of high duty, it shows no sings of being abused. This turbo is
> capable of making at least 40psi of boost (MAP well above 80in.
at sea
> level).
>
> I need a turbo that size for proper efficiency and compressor
speed at
> altitude, not high boost. Since I will never use that much boost I
> opened up the waste-gate. The waste gate has been ported and the
> flapper
> removed. In this configuration, full throttle at S/L will give me a
> MAP
> just under 50", which is perfect.... I am only using the exhaust
> needed
> to obtain the desired boost.
>
> This translated to just over 30" MAP at 16k feet altitude and WOT.
> Again, perfect. Minimum loss of efficiency from back pressure,
minimum
> loss from throttle body pumping. Aircraft Speed approaching Vne.
> Compressor speed well into the green according to the compressor map.
>
> It sounds like the end of the story, but I still plan some changes to
> the turbo when I get around to it. I plan to put the flapper back in,
> and aggressively clip the turbine blades. The waste-gate has another
> important function in my set up. Besides bypassing unneeded exhaust,
> it
> helps to regulate and smooth out the natural variations in boost.
> These
> variations only really become evident when running WOT, particularly
> when running lean of peak. With the flapper removed (waste gate
> passage
> fully open all the time) I have found that I can either aggressively
> lean the mixture, or operate at WOT, but not both at the same
time. If
> I try to do so, the natural variations in engine RPM have an
amplified
> effect on boost and engine RPM. i.e. a slight decrease in RPM
leads to
> decreased exhaust energy which leads to decreased boost which further
> decreased RPM. It all happens rapidly and seems like someone is
> playing
> trombone with the throttle. At less than WOT, the throttle valve
tends
> to dampen out these variations. The flapper valve on the wast gate
> (connected to the stock automatic actuator) did the same thing
when it
> was connected. I am not sure why running rich also fixes the problem,
> but it does.
>
> So here is the IMHO Dave Leonard simple guide to selecting a
turbo for
> the aircraft 13B: A compressor in the size neighborhood of the TO4B
> 60-1 (since it is oversized and underused the exact size does not
> really
> matter as long as it is in the neighborhood - check the maps,
many are
> in this size range). Then, the turbine should be as BIG as you can
> fit... (A/R in excess of 1.0). Have some sort of automatic regulation
> on the waste gate, then make the wast gate big enough and/or the
> turbine
> clipping aggressive enough so that full WOT will give you what you
> decide is the maximum MAP you want to be able to achieve (for the
time
> being I recommend less than 50" MAP at sea level).
>
> The amount you need to clip you turbine blades will depend on the
> turbine and compressor that you finally choose. A larger turbine and
> waste gate is probably preferable to a more aggressive clip.
> Alternately, a huge aircraft style wast gate with a normalization is
> probably appropriate if you prefer.
>
> BTW, keep in mind that if you go to an automotive guy to get the
turbo
> built they will try to talk you out of it. Very few of them
understand
> what you need, they seem to be too in tune with what the car guys
> want/need.
>
>
>
> And finally, regarding mass vs. volume for turbine sizing... My first
> issue with the above discussion it that you discounted mass flow as
> the
> preferred way to size the turbo. I initially (years ago) agreed with
> you
> that volume flow seemed more approiate, and it will work (and some
> turbo
> maps are in volume). But the density of the air does matter. The
> reason is that we need some sort of standard reference pressure
to use
> when deciding how much volume we need in order to achieve the desired
> h.p. from the engine. Since h.p. is mass dependent, I have found that
> is far easier to use mass flow on the turbine maps. This then will
> correlate directly to the proper higher pressure ratio for use on the
> compressor map.
>
> Another way to look at it: since the pressures are different in
> different parts of the intake system, then volume flow is going to be
> different depending on whether you are looking at compressor inlet,
> outlet, throttle body etc... but the mass flow is the same throughout
> the system. If you play with it long enough, you will see that
mass is
> the better measure.
>
> David Leonard
>
> Turbo Rotary RV-6 N4VY
> My websites at:
>
http://members.aol.com/_ht_a/rotaryroster/index.html
>
>
http://members.aol.com/_ht_a/vp4skydoc/index.html
>
http://leonardiniraq.blogspot.com
>
>
> Thanks for your input on this Dave. So far you have got more
> out of the largely stock 2nd gen turbo than anybody else.
>
> I agree with you on mass flow. So many air molecules
> combined with so many fuel molecules equal HP.
>
> You state:
> "A bigger waste gate will only reduce the back pressure created
> for a given position on the compressor map."
>
> This is not clear. I am not sure I agree. If no other changes
> are made a larger opening in the waste gate WILL change the position
> on the compressor map because less turbo HP is going into the
> compressor. The limit case is no exhaust gas going to the
> turbine at all. The position on the compressor map will be at the
> very bottom in that case.
>
> I think not everybody wants 50 inches of Hg boost at sea
> level or flat rating to 16,000 feet.
> Suppose a larger (or different type) waste gate alone gave 29
> inches of HG at sea level (no boost) and the waste gate was
> gradually closed to get more inches of Hg boost at altitude.
> In other words flat rating but maybe not to 16,000 feet as you have
> done. Better yet a red line RPM was established for the turbine
> and the waste gate controlled to keep the turbine within
> the red line RPM. Would that not be cheaper than increasing
> the size of the compressor to slow the turbine?
>
> I think that would be a good starting point for most people.
>
> The HP of a given turbine is a function of the blade speed to
> exhaust gas velocity ratio. Dynamic pressure if you will.
> Dynamic pressure is both a function of velocity and air
> density. As the blade goes faster than about 50% of gas speed
> HP starts to drop as turbine torque is a linear inverse function
> of turbine RPM. Chart 2. Therefore I agree that using a larger
> compressor (A to B) reduces turbine RPM and increases the HP put
> into the compressor IF the RPM is beyond the 50% gas velocity point.
> Chart 4. Note the similarities with a combustion engine
> and a propeller. The HP curve can be thought of as
> potential HP. Only when the HP curve intersects the load
> curve have you extracted any appreciable HP from the turbine.
> Another good analogy is dyno load verses prop load.
>
> One might get the impression this is a very complex subject.
> One would be right :)
>
> Paul Lamar
>
>
> Thanks for jumping in Dave.
>
> I have nothing against the maps that use mass flow. If you want a
> turbo because you want to increase your horsepower you have to think
> in terms of mass. The Garrett website gives good info on this. It's
> a simple matter to figure out how much mass flow you need based on
> your air-fuel ratio and BSFC. You can then look at the compressor
> maps and choose a compressor that will give you the mass flow you
> need.
>
> But what do you do if you don't want to increase your horsepower,
> but want to carry your existing horsepower to a higher altitude?
> Then you have to think in terms of volume, because you're going to
> need more volume in the thinner air -- not more mass because you're
> not after more power. (Same mass, more volume) And those compressor
> maps are no help at all for us.
>
> Still, it's no problem if you do the arithmetic. Then you can use
> the mass flow maps just fine. The mass and volume relation is fixed
> in the ideal gas law. That's why I posted the formulas. I have
> reposted them here in a simple 4-step form. I will also put them
> into a spreadsheet and that will make the calculations easier.
>
> Thanks for describing your setup. With 50 inches of manifold
> pressure it makes considerably more power than a normally aspirated
> engine at sea level -- a ground-boosted aircraft engine, in other
> words. Nothing wrong with that -- it's always nice to have extra
> power.
>
> What I'm interested in though is a turbo normalized setup that
> is dead simple from a flying point of view. A good example of this
> is the Mooney 231 or Turbo Arrow, neither of which has any kind of
> waste gate adjustment whatsoever -- either automatic or
> pilot-controlled. Same thing on the Seneca turbocharged twin.
>
> There is just a small pipe upstream of the turbo that bypasses some
> of the flow. That's it. It's basically a fixed waste gate, but it can
> be adjusted on the ground. (See attached picture). The turbine and
> compressor are sized to achieve the desired boost and turbine speed
> as you climb. This can be done just as easily on the rotary if we
> choose the right size turbine and compressor.
>
> Another simple setup is the turbo Skylane and Saratoga which do have
> a waste gate that is linked to the throttle. You fly the plane just
> like a normally aspirated bird. You push the throttle in for takeoff
> and pull it out a bit for cruise.But you do have to monitor manifold
> pressure more closely in climb and descent. A lot of STC'd
> turbo normalizing systems use this setup too.
>
> Notice that the waste gate is closed at takeoff. Why doesn't the
> turbo over speed or the manifold pressure over boost? Because the
> turbine is big enough that the sea level exhaust energy can't spin
> it up. The exhaust energy is a function of exhaust speed across the
> turbine, which is a function of pressure drop across the turbine,
> which, in turn, is a function of the pressure delta between the
> turbine outlet (which is ambient) and manifold pressure. This
> pressure delta increases as you climb, provided you can maintain sea
> level manifold pressure. So the exhaust speeds up, and the turbo
> spools up the compressor, adding just the right amount of boost you
> need to keep manifold pressure up -- provided you have sized the
> compressor and turbine just right.
>
> We can use this natural balance, if you will, to our advantage in
> designing a turbo system that needs no pilot control whatsoever. I
> like the simplicity of this approach because I find that simpler
> systems make for more enjoyable flying. That's why you see the move
> toward single-lever control -- something turbine pilots will tell
> you is a very good thing.
>
>
> Regards,
>
> Gordon.
Gordon,
Good information!
I agree on the bypass system and advocated that sort of system a little
while ago, however my suggestion was a bypass to the turbo with a
gate to
restrict either the exhaust or the bypass. My reasoning is that you
can't
overspeed the turbo if you restrict the incoming exhaust to the turbo.
I also believe that by remotely locating the turbo you will reduce the
speed and heat of the exhaust.
I still believe this is the best way to set-up an appropriately
sized turbo.
It allows for variation and fine tuning as well. Once set up it
doesn't need
to be varied in flight.
It appears from what you have stated, others think this is the right
way of
doing things as well.
George ( down under)
George,
I should point out that the fixed wastegate idea will not work very
well on ground-boosted engines. If you want a turbocharger to make a
lot of extra power at sea level, you can forget about the simple
approach.
The reason again is the exhaust pressure. A turbonormalized engine
like I am proposing is typically run at about 35 inches of manifold
pressure for takeoff power. This is just a touch higher than ambient
and just makes the engine a little more efficient. It overcomes a
little pressure loss that may be present at the compressor inlet due
to air filter or bends in the tubing, etc. Also a little bit of
pressure is lost between the compressor and the manifold, so this
little bit of boost, about 2.5 psi, is just enough to make the
engine run at good efficiency. It's not really adding much power.
So look at the pressure delta between manifold and turbine outlet --
it is very small, so there is not much pressure drop at all across
the turbine, so not much exhaust energy. Even though you are at
maximum rpms and takeoff power, your big turbine is not spinning
very fast at all and your compressor is only making a small bit of
boost.
Now, when you have a ground-boosted engine that is running 50 inches
of manifold pressure at taekoff, the pressure delta across the
turbine is significant. That means a lot of gas velocity and a lot
of energy imparted to the wheel. And if you also have a small
turbine, it will spin up to a very fast speed.
You actually need to open the wastegate at takeoff power to prevent
overspeeding and overboosting. So there goes the simple hands-off
system out the window.
Sure, you could divert some of the flow ahead of the turbocharger.
This might be an idea worth considering for those who insist on
using the stock turbo and get the wastegate big enough. But what
happens at altitude when that little turbine runs out of steam? Then
you can't close that little pipe. So your compressor slows down and
there goes your boost -- just when you need it most in the thin air.
The example I worked through with the formulas used 35 inches of MAP
at takeoff and was able to carry that full 35 inches to 20,000 feet
-- full takeoff power available. With nice and cool intake charge
without an expensive and space-robbing intercooler.
The formulas give us a pretty good idea of the compressor that will
do the job, but now we need to apply the same kind of logical
approach to the turbine size and configuration. This will be the
hard part because, as Paul pointed out, the turbo manufacturers
don't give much info on turbines. That's because for car turbos, you
pick the compressor size first and then just take the turbine that
goes along with it.
For a turbonormalized application though, I would bet that we need a
much bigger turbine than those compressor-turbine combinations they
package for cars. We will need to get some good data on turbines.
And to really do this right we need to instrument the rotary's
exhaust and get some good pressure graphs.
Regards,
Gordon.
The Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site.
http://www.rotaryeng.net
Copyright 1998-2006 All world wide rights reserved.