# Centrifugal Impellor in place of Axial Ducted Fan

### Help Support HomeBuiltAirplanes.com:

#### Culleningus

##### Well-Known Member
The diffusers used on model turbine compressors drive the air neatly round 90 deg (in side view).
This provides a different efflux geometry, compared to the regular spiral arrangement used on turbochargers for example, which exhausts out of the side.
If the pressure rise is better with a Centrifugal (single stage) in the context of a compressor can such an arrangement (with the outer diffuser ring) be used in place of a Ducted Fan in free air for propulsion?
What are the comparative efficiencies if so?
Secondly what about heat? I wondered if the exhast gases from an IC motor were added to the efflux, and re-ignited (in a stainless steel chamber) is there a capacity for 'boost' therein?

Last edited:

#### Dan Thomas

##### Well-Known Member
The diffusers used on model turbine compressors drive the air neatly round 90 deg (in side view).
This provides a different efflux geometry, compared to the regular spiral arrangement used on turbochargers for example, which exhausts out of the side.
If the pressure rise is better with a Centrifugal (single stage) in the context of a compressor can such an arrangement (with the outer diffuser ring) be used in place of a Ducted Fan in free air for propulsion?
What are the comparative efficiencies if so?
Secondly what about heat? I wondered if the exhast gases from an IC motor were added to the efflux, and re-ignited (in a stainless steel chamber) is there a capacity for 'boost' therein?
Centrifugal compressors move a relatively small amount of air and raise it to a relatively high pressure. They do it like the axial compressor: They accelerate the air, imparting energy to it, and then it is slowed in the diffuser, which raises its pressure. Two cetrifugals in series can get 350 psi.

That sort of thing isn't of much use in replacing a fan. The fan or propeller accelerate the air without slowing it to raise pressure.

But I've often wondered if a squirrel-cage fan might do the trick. It would accelerate huge volumes radially, and a housing similar to a radial engine cowling would be needed to direct the flow aft. I think the internal drag of all those vanes and the cowl would eat up too much of the energy, though. In comparing boats of identical horsepower, one driven by a propeller and the other by a jet pump, the jet's efficiency can be as much as 40% lower than the prop's. Its attraction is safety around swimmers, and the ability to operate in shallow waters.

A jet engine burns about 25% of the air pumped through it. That's why an afterburner can work; there's still plenty of oxygen available for further heating and expansion (and therefore velocity) of exhaust gases. The system must be designed so that only velocity is increased, or back pressure would stall the engine. The highest pressure in any turbine engine is found in the diffuser just behind the compressor section and ahead of the combustion cans. After that, the pressure falls and velocity increases enormously.

Dan

#### Lucrum

##### Well-Known Member
Log Member
Homemade "jet" engines have been made from turbo chargers. I haven't seen anything that I'd want to strap myself and an airframe to though.

There was some experimentation along the lines of what you're talking about done in the early years of aviation up through WWII. All of which was pretty much abandoned once the jet engine came into widespread use.

You can look up Motorjet - Wikipedia, the free encyclopedia. Unless I've misunderstood this is similar to what you're asking about.

Scott F. Hall's Homage To the Motorjet (a.k.a. Afterburning Ducted Fan)

Some NACA History ch8

#### Davefl42

##### Well-Known Member
Also look up the Napier Nomad. It was a diesel two stoke with turbo compounding of the exhaust flow.

#### Culleningus

##### Well-Known Member
Dan
The comparative inefficiency of centrifugals is I think is in part resultant of diffuser design (having to move the flow round through more acute angles)
As for the 'amount' of air they each move I would argue the centrifugal does better. In single stage format it gives a bigger pressure rise. So Im not sure anyones taken the troubke to compare radial 'diffusers' with axial 'stators' for cold air propulsion.
But like you rightly say the losses associated with the radial diffuser fixed blades are 'likely' greater than those with the axial stator fixed blades, it seems where no cross section area taper intake/exit is employed . The same is 'definately' true for spiral volute housings like on a turbo (charger), which is poor.
That being so it is harder to stall a centrifugal by backpressure making it more suitable for combustion as a 'single stage' fan. It is also less sensitive to upstream changes in pressure & better eqpd therefore 4 vacuum ops.
In this respect my enquiry was a little contradictory, in that the requirements for cold thrust is that backpressure (and compression) its been demonstrated has a negative effect. Therefore intake/exit taper for ducted fans is of the lowest magnitude (typically a few degrees max).

The best of the worst crop of designs kindly offered above (but none the less interesting) in respect of 'motor jets' were those that were really just ducted axial fans, with a 'tap' for combusting a 'percentage' of the efflux rather than vice versa. But the net gain from such was unjustifiable. I never figured how those Kamikaze motorjets worked, because they are the exception to this rule. I can only think they did something else as an 'exception' to the above. Or else they wouldnt have got up to speed, with motorjet combustion alone as the videos only too clearly testify.

Im not sure Whittle would have confessed to building a 'motor-jet' since I would argue in this case the IC motor was only used for starting (contrary to what one of the links above was making out!)

Dave

Last edited:

#### Culleningus

##### Well-Known Member
Further to the above:

I wondered if anyone had actually obtained any meaningful figures/drawn graphs from tests pertaining to comparisons of centrifugal and axial air-fans, for cold air propulsion.

1)25cc leaf blower (@ same 'impellor as fan' diameter measure peak thrust + rpm)
2)25cc ducted fan (@ same 'fan as impellor' diameter measure peak thrust + rpm)
3)25cc with centrifugal 'bell shaped' turbo-charger impellor @ same diameter and model jet turbine type diffuser (measure peak thrust + rpm)

As well as 'rpm'
the axial stator;centrifugal diffuser(if any); fan/impellor design; volute housing design etc all have a marked effect (variable). But in general terms 'how much' more STATIC thrust would the ducted axial fan generate than the other pair?

Stage 2:
How much DYNAMIC thrust (but I guess that'd be harder to calibrate.

This would provide the feesability for which any potential concept exploration of no.3)might then be justified in terms of time.

Dave

Last edited:

#### wsimpso1

##### Super Moderator
Staff member
Log Member
Centrifugal compressors were used in early jets - The Whittle Engine (UK) had a centrifugal compressor, as did the early Soviet jet engines (admittedly Whittle Engines) used in the MiG 15. In turboshaft engines for helos, APU's, etc, centrifugal compressors and turbines are commonly used, where they facilitate axial compactness at the expense of increased diameter.

Billski

#### Toobuilder

##### Well-Known Member
Log Member
If the pressure rise is better with a Centrifugal (single stage) in the context of a compressor can such an arrangement (with the outer diffuser ring) be used in place of a Ducted Fan in free air for propulsion?
What are the comparative efficiencies if so?
Secondly what about heat? I wondered if the exhast gases from an IC motor were added to the efflux, and re-ignited (in a stainless steel chamber) is there a capacity for 'boost' therein?

If I understand your post correctly, you are thinking of driving a centrifugal rotor with an internal combustion engine as a means of generating cold thrust. The centrifugal rotor would be in place of a ducted fan.

As discussed, the centrifugal rotor as used in turbine engines is a fairly inefficient "air mover". It works in turbine engines because it does more than simply feed large volumes of air to the combustor, it also has to accelerate the air for compression in a single stage. This acceleration and compression consumes a great deal of energy that is regained in the combustion/expansion process in a jet, but totally lost in a "cold" thrust application. In other words, you are "paying" for volume AND compression, but only get to use the volume... Not a good return on investment. A ducted fan is more efficient at moving "pure" volume.

#### Culleningus

##### Well-Known Member
I think the rate of movement is higher in any case with the centrifugal. At a given intake diameter, lets say.
But the problems as Dan cited above are more bound up with drving the air through acute angles.
The fact that a single stage centrifugal works better in compression/evacuation for a 'single stage' unit is as as Ive said previously is more to do with its ability to tolerate more pressure variation at both ends, without stall.

#### Culleningus

##### Well-Known Member
It was in the context of improved diffuser design, which made me raise the issue of whether such losses could be reduced to a suitable level for meaningful cold-air proplusion. But it strikes me that surface/skin friction alone will take care of that..

#### Toobuilder

##### Well-Known Member
Log Member
It was in the context of improved diffuser design, which made me raise the issue of whether such losses could be reduced to a suitable level for meaningful cold-air proplusion. But it strikes me that surface/skin friction alone will take care of that..
So in your comparison (DF vs CF), the input power is not an element for consideration?

#### Culleningus

##### Well-Known Member
It was in this context that the thrust of my query was to determine if the diffuser/impellor combo pictured below (except I pictured a stator in my minds eye in there too), was likely to provide a more meaningful result against an axial ducted fan of the same intake/outflow cross-sectional areas?

At, I might add, the same shaft HP

#### Culleningus

##### Well-Known Member
Heres another, sorry the last was so BIG
YES intake/outflow cross sections same as DF, power applied same for both.

[/URL][/IMG]

#### Starman

##### Well-Known Member
The stator in the second image is better suited for driving the air rearwards but it could be optimized much more with 90 degree bends to make the ir actually go straight back. In any case, a centrifugal compressor is the worst possible way of providing thrust because:

#### Dan Thomas

##### Well-Known Member
A centrifugal compressor has to turn at very high speeds to compress its air. A squirrel-cage fan has a larger diameter and can be made with a wider blade set to move much more air at lower RPMs and pressures, which is what you need for thrust purposes. The centrifugal compressor in a turbine engine can produce pressure rises of as much as 15:1 per stage (axial is typically 1.25:1 per stage, so lots of stages are needed), but that pressure rise is mostly useless for our purposes. The compressor increases pressure much more than velocity, see, and once the air passes through the diffuser its speed has dropped to well below 100 feet per second.
It's barely moving. The flame front of kerosene isn't all that fast.

The diffuser is a divergent duct. That is, the walls and vanes have channels of increasing cross-section so that the air is slowed, and, according to Bernoulli, its pressure is increased as the velocity falls. That's why the highest pressure in any jet engine is found in the diffuser; as it enters the combustion cans and heat is added, its velocity rises and pressure falls. If the pressure didn't fall there, it would back up into the compressor and stop the airflow and put the fire out.

Any diffuser offers drag. Every blade in the rotor and stators of an axial compressor is draggy, and the diffuser has axial flow through a divergent duct. In fact, every blade on the stator forms a divergent duct with its neighbor to form a divergent duct, to slow the air and increase its pressure.

It's hard to beat the propeller as an efficient thrust producer for the low-speed aircraft we fly. It's a rotating wing of high aspect ratio and presents minimal drag for the thrust it produces. Like the wing, it can generate more thrust (lift) than it produces in induced drag, which is also why the propeller is used on big ships and small boats, instead of the old paddlewheel whose drag was higher than the thrust produced.

Dan

#### Culleningus

##### Well-Known Member
Many thanks Dan. Youre certain the pressure rise is in the diffuser, not in the combustor? due to flare & good old Benoulli. If so it clarifies alot. Earlier I think you cited skin friction drag which is defo a serious problem.
But Im not sure these (propellor) rules are applicable in this context. Like Dan says that'd be like common factoring the effect of blade area for both 'paddles' and 'screws' both of which work in different ways.
I was actually rather hoping someone could offer quantification of the relative thrust efficiencies between a ducted fan with stator and the carefully shaped centrifugal compressor stage (above) with diffuser as well as stator which Starman rightly points out would help reduce efflux 'spin' (not sure about the 90deg tho.)
Tall order me thinks.
Good to know its prob. wasted energy trying. ...unless ..the inefficiency you cite is justifiable in some way sense or form

eg.NOISE?
or SAFETY

Dave

Last edited:

#### pie_row

##### Well-Known Member
Nothing taken out of context. .
You changed what you wrote and what you quoted of me after the fact. And so this statement of yours is a lie. And in doing this you abused you moderator privileges.

Last edited:

#### Dan Thomas

##### Well-Known Member
Many thanks Dan. Youre certain the pressure rise is in the diffuser, not in the combustor? due to flare & good old Benoulli. If so it clarifies alot.
Remember this picture?

Here's another:

In both pictures you will see that the highest pressure is in the diffuser, before the combustors. It has to be that way; a higher pressure in the combustors would prevent any further entry of incoming air, wouldn't it?

This stuff sometimes isn't intuitive. Turbine engines, like a lot of other machinery, make use of energy exchanges between pressure, velocity, and temperature. When we slow airflow, we increase its pressure. When we increase its pressure we raise its temperature. In the combustors, the pressure falls, but that doesn't mean power is diminishing. The addition of teriffic heat there causes expansion, an increase in volume, which is not used to create higher pressure as inside our piston engines; it's made to create high gas velocities that are then used to drive the turbines. See how pressure and temperature both drop enormously as the gases pass through the turbine section, representing another energy exchange.

Here's the converging-diverging duct theory, just to confuse you further:

Dan

#### Toobuilder

##### Well-Known Member
Log Member
...Any diffuser offers drag. Every blade in the rotor and stators of an axial compressor is draggy, and the diffuser has axial flow through a divergent duct. In fact, every blade on the stator forms a divergent duct with its neighbor to form a divergent duct, to slow the air and increase its pressure.

Dan
Spot on in almost every respect. One minor correction is that the compressor section forms a convergent, rather than a divergent duct. The rise in pressure in the compressor section is due to the fixed stator vanes recapturing the mechanical acceleration caused by to the rotating compressor blades. The compressor blade accelerates a section of air, the fixed vanes capture and compress, then offer the compressed air to the next stage, where the process is repeated. You will notice in the cut away that the physical volume of the duct is also decreasing (converging) with every stage as it moves aft to help the acceleration process. It is the compressor discharge pressure (CDP) that is delivered to the diffuser section (a classic divergent duct) for deceleration and ultimate pressure peak. You are correct that this is the maximum pressure in the engine; the combustor only adds energy in the form of heat (fuel), and the rapid expansion of gasses provides the thrust, but the resulting pressure rise still remains below the peak found at the combustor inlet.