Centrifugal fan?

[quorneng]

The current EDFs are to all intents and purposes the equivalent of a turbofan but without the high velocity part.
This means that if the original was a pure jet either it has to have a small capacity fan or the ducts have to be sized as if the plane had been powered by a turbofan.

So why not a centrifugal fan that, at model sizes at least, produces a higher pressure ratio than an axial fan thus generating its thrust from a smaller volume of air albeit requiring more power to do so.

This 5" (130mm) fan from a wet and dry vacuum only weighs 1.6oz (46g) but is capable of absorbing at least 1200W with an inlet of just 1.9" (47mm) diam.


An axial ducted fan with that inlet would be hard put to absorb much more than 200W.

Just a thought.

[kyleservicetech]

Originally Posted by quorneng

The current EDFs are to all intents and purposes the equivalent of a turbofan but without the high velocity part.
This means that if the original was a pure jet either it has to have a small capacity fan or the ducts have to be sized as if the plane had been powered by a turbofan.

So why not a centrifugal fan that, at model sizes at least, produces a higher pressure ratio than an axial fan thus generating its thrust from a smaller volume of air albeit requiring more power to do so.

This 5" (130mm) fan from a wet and dry vacuum only weighs 1.6oz (46g) but is capable of absorbing at least 1200W with an inlet of just 1.9" (47mm) diam.An axial ducted fan with that inlet would be hard put to absorb much more than 200W.

Just a thought.

I think that's been discussed before in either wattflyer or other websites.

Methinks the applications for a centrifugal or axial fan are quite a bit different, and the airflow issues with a centrifugal fan would limit its usefulness in our models.

Other wattflyer reader opinions???

[constantCrash]

5" is a 127mm fan. that's without any form of cowl to re-direct the air out the back. you'd have to have a BIG plane to fit that fan in. That is unless you mounted it horizontally wich would probably cause some interesting adverse yaw affects.

Also what are the losses from having the air turn 90 degrees twice?

[quorneng]

The fan outlet annulus diam is not much bigger than the fan itself, no more than 5.5" diam and you could get that in a 60" span Vampire.

As in this case the fan is intended to maximise pressure rather than flow the duct losses may not be that serious.

I am not saying it will work but it is not quite as far fetched as it sounds.

A centrifugal turbo jet actually turns the air through 90 degrees four times.

[Bill G]

Interesting stuff Quorneng. Always good to be experimenting with ideas and thinking outside of the box.
The FW190TL was to be based on a centrifugal turbine:
http://tanks45.tripod.com/Jets45/His...w190/Fw190.htm

I built one, but of course I cheated and used an EDF.

[tampaflyer]

simply put.. if it was better the ducted fans would have not become those like the viofan,shubler, weomtecs...

but you can always do it differently.. it might fly and it would be differntly.

[quorneng]

A centrifugal fan will never be 'better' than a single stage axial unit for simply producing thrust.
However if you want to stick to scale inlet and exhaust ducts on something like a Vampire then a centrifugal fan may well be able to create more thrust than an axial unit.

[Glacier Girl]

well if it doesn't work for propulsion, you could always use it to make a sorta Jericho Trumpet for a large scale Stuka. All it was was an air operated centrifugal fan used to make the wail.

Hmm, mix the motor/fan to the planes throttle with and on/off switch, and bingo you have a dive bomber with sound.

Funny, I was actually pondering this with a unit out of a little hand held vacuum.

[quorneng]

I think calling it a centrifugal fan is misleading.
What I have in mind is to create more thrust that an axial EDF for a given inlet size by raising the pressure to accelerate the exhaust.
In a centrifugal compressor it is called an impeller.
The problem with this approach is that raising the pressure to gain thrust is a squared function where as raising thrust by increasing mass flow is linear. So trying to gain thrust over an EDF is quickly going to result in a severe battery weight penalty.

I have a nice big 3 view of what I think is a Mk1 Vampire.
The question is how big would it have to be for the scale inlets to match the 47mm diam inlet of the 5" impeller?
It comes out at 48" span and the fuselage is close to 6" diam. So far so good.

For comparison it would be useful to know how much thrust would a 47mm EDF could be expected to produce.
The problem is there isn't one but a good 40mm on 4S can give 10oz (300g), a 55mm double this, so lets say 15oz (425g) for a 47mm.
The task then is to establish if the 5" impeller can produce more than 15oz thrust and how much power would it require to do it.

The final issue is could I build a 48" span Vampire and get it to fly satisfactorily on as little as 15oz thrust?
Of all these questions the last one worries me least.

[Flite-Metal]

Finally, we've arrived at where this discussion should have started...

At this point in history you need not address your issue as an experiment or even a DYI. That has served
to hinder more projects than help as typically few of the DYI's utilize sustainable resources in their effort
to address their latest challenge, then they move on to another personal challenge.

Thrust (pounds) to weight (pounds) ratio. The tried and tru(e)th ratio of 1.67# of airframe can be flown for
every 1# of exhaust outlet measured thrust, regardless of the type of fan. Nothing else matters, whether it
be a hampster or squirrel powered "air mover". 1.67 = 1 is the go/no go factor.

[quorneng]

Flite-Metel
I would not disagree in principle but the danger of a single factor is it does not cover every situation.
This 36" ducted fan flew very nicely with a ratio of 1:1.82.

And it could maintain height with a ratio in the order of 1:3.

I rather expect a big light Vampire might well be similar.

[Flite-Metal]

The ratio I provided is a relative example to show the "plane's flying surfaces" aka sq.in.,
air foil, aspec ratio and parasitic drag of a "scale" jet will function properly at a thrust to
weight ratio of 1:1.67. As the flight and control surfaces grow, the ratio will change. This
is a ratio observed during too many hours of head down, butt up trying to get 2 flights a
day....back 30 years ago... ;^)

[Bill G]

Originally Posted by quorneng

Flite-Metel
I would not disagree in principle but the danger of a single factor is it does not cover every situation.
This 36" ducted fan flew very nicely with a ratio of 1:1.82.
Attachment 157122
And it could maintain height with a ratio in the order of 1:3.

I rather expect a big light Vampire might well be similar.

Some EDFs will fly well with a lower thrust to weight ratio than that. My little brushed EDF50 airliner is likely around 1:4. You and I are experimenters, that are willing to develop a spec for whatever situation we build for. The rules of thumb are for those who want to put something together with a more guaranteed result, without extensive testing and prototyping. We don't mind going back to the drawing board, while some wand conservative specs to avoid that. The only issue I have with rules of thumb, is if they are stated as written in stone.

I'm sure you are correct about a big, light Vampire also. While my 30mm version in not big, it has a thrust to weight ratio of roughly 1:4, although there is a bit of a wing cheat also. The undercamber wing definitely helps it fly, as a low powered floater. It's on my You Tube WTFLYR channel, and flew even better when the battery was not old as dirt, at the time of the video.
BTW, if this thread stays up long enough, some nutcase like myself is going to want to try this fan idea.

[Octavius]

A Jet is a little different. A centrifugal jet produces thrust at the combustion chamber(s). I tried building a vertical type UFO that had to duct 100% of the air in 90 deg turns twice and it ended up producing almost no thrust so I have some doubt in this project, but you should still give it a try if you feel passionate about it.

[quorneng]

Just as experiment I taped over one of the exhausts of a wet and dry vacuum to give a single side outlet. Of course very inefficient ducting but it actually generated sufficient thrust (about 3oz) to just move the vacuum on its castors across the floor and as I was holding the inlet hose there was no suction effect either.
All I have to do is improve that thrust by a factor of 5!

[JetPlaneFlyer]

Quorneng,

The vac may well use 1200W but it used this power to increase pressure, not to generate large airflow. Pressure wont propel a plane, for that you need airflow and your centrifugal fan is never going to deliver enough of that to fly. Not only would you have to increase thrust by a factor of 5 but you would have to decrease weight by a factor of maybe 20.. bottom line, it aint gonna happen.

I've seen the same question asked over and over though usually the subject is a leaf blower, same problem through, too little thrust and much too much weight.

sorry

Steve

[quorneng]

"Pressure won't propel a plane"
I think you will find thrust is generated by a combination of mass and velocity and velocity is created by pressure.
In reality all EDFs must alter the pressure otherwise the air would not flow down the duct.
However increasing mass flow is a linear function, twice the mass, twice the thrust and requires twice the energy but raising pressure is a squared function, twice the pressure, twice the thrust but it needs four times the energy to do it.
A good EDF 'system' is about twice as powerful as it needs be to fly so in theory a centrifugal impeller can get away with being just half as efficient at producing thrust and still work, particularly if coupled with a good aerodynamic airframe.

We shall see how close I can get.

[JetPlaneFlyer]

Quorneng,

Thust (aka force) = mass x acceleration (Newton's second law of motion)

Note that pressure doesn't appear in the formula. You may have a point that there is some small pressure change in order to generate the acceleration of the air but the fact remains that it's the acceleration that's needed to generate thrust. The fan must be able to pass high volume of air if it's ever going to create a large acceleration on a large mass in order to create large thrust.... that's what Newton told us. Your argument seems to be 'Ahh, but i can use the excess pressure to create more airflow', but that doesn't really work. If you allow a centrif fan to blow into an open duct (like an EDF) then the pressure drops to atmospheric and the air flow increases to a point but it very soon hits a 'wall' because the design of the centrif fan doesn't allow high airflow (convoluted and narrow flow path). The fan basically becomes a very inefficient air mover rather than an efficient pressure increaser.

Basic rules:
Centrif: good for low flow high differential pressure
Axial: good for high flow low differential pressure

For propulsion you need high flow, period

By all means go build one, and i do hope it meets with some success, with your amazingly light building then you have more chance than most of making it work.. But if it does fly it would be in-spite of the centrif fan, not because of it. If you want to use small scale like ducts then best EDF option is one of the small EDF units that spin at very high RPM. I have a 65mm unit that can easily take 800W and more, even 50mm units will (Irecall) take 500W or thereabouts with specialist high kv EDF motors.

Steve

[quorneng]

Steve
You use a nozzle to convert pressure into velocity.
This creates a change in momentum and thus thrust, just as every propulsion jet engine ever built does.

[JetPlaneFlyer]

Quorneng,

Yes, a nozzle creates high acceleration but if fed by a centrif then the mass flow will be relatively low. Obviously if you can create high enough acceleration by shooting the air at extremely high velocity through a small nozzle then you could produce reasonable thrust with relatively little mass flow. But there is a big problem with this approach; it's highly inefficient.

The problem is kinetic energy is proportional to velocity squared so this tells us gram for gram of thrust it takes twice as much energy to produce thrust by accelerating air faster compared to moving a larger mass. This is why large props are more efficient at making thrust than small ones, and explains why helicopters have huge 'props'.

Turbojet engines are a different animal altogether and really cant be sensibly compared to EDF. The centrif fan used on some turbines is there to compress air for combustion. The actual thrust comes from the huge volume flow of hot gas that's generated in the combustion chamber which is many times more in volume that the air flow through the compressor. Unfortunately with an EDF flow in = flow out because no additional volume is created in combustion.

If anyone can do it you can, but i still think that your results would be better if you used a high RPM axial fan on the type commonly used in EDF's. If the airframe is build large and light it's perfectly possible to use scale inlet and outlets on EDF jets.

[quorneng]

The lash up test stand.
No shround as at this stage I simply wanted to test the motor performance and impeller balance as it does no longer has a couple of kg of electric motor attached to it.

A remarkably cheap 2700kV heli motor. It is only 230W so it is probably too small. The impeller is mounted on a modified collet type prop driver but at the moment retains the big steel washers used in the original vacuum application.

I have in mind to use a full diameter 'annular' nozzle. The air at the impeller tip is travelling fast so it seems sensible to just turn it through 90 degrees rather than slow it down in a scroll venturi to gain pressure only to speed it up again at a conventional exhaust nozzle.
Of course in a true turbojet you have to do exactly this to create a working heat cycle.

[quorneng]

JetPlaneFlyer
As a first step towards a centrifugal EDF I am indeed going to build a large light conventional EDF with scale size inlets, probably a Vampire.
By sizing the inlets ducts to match the 47mm dia of the centrifugal unit it will also become the test bed for it.
A cheap 55mm out runner EDF.

Interestingly the actual fan area is almost exactly the same as a 47mm dia duct.
The fan placed over a Vampire fuselage section scaled to give an inlet area of a 47mm dia duct. The span would be 48" (1220mm).

The first task is to build a complete test duct to see how much thrust is lost with the bifurcated inlet.

[quorneng]

The first run of the centrifugal unit 'test' unit.

YouTube Video

I only had a cheap 30A ESC spare so only about 1/3 power to avoid 'smoking' it. I have a 40A on order.
At the moment it is running in the opposite direction to the vacuum giving swept forward impeller blades. In theory this should give a higher exit velocity at the expense of pressure.
A long way to go but at least it runs smoothly.

[dkrhardy]

I am NOT even conversant when it comes to EDF's, but 2 things come to mind right off the bat.

1- weight
2-AMA says no metal in EDF's , so what about this one?

Interesting concept and all new to me.




Don

[quorneng]

dkrhardy
There are metal bladed EDFs.
The concept I am exploring here is just how much thrust can a centrifugal unit produce.
For any given power it will never be as efficient as a normal axial EDF in producing thrust but it is a relatively simple way of getting a lot of energy into a relatively small airflow, more than a single stage EDF ever could.
The actual centrifugal unit need not be particularly heavy but the battery power required to drive it might well prove to be so.