Welcome, Guest. Please login or register.
October 21, 2014, 12:30:49 AM
Home Help Search Calendar Login Register
+  FlyQuiet
|-+  Free Flight
| |-+  Free Flight Discussion (Moderator: mike)
| | |-+  Ducted fan for sport free flight
« previous next »
Pages: [1] Print
Author Topic: Ducted fan for sport free flight  (Read 4740 times)
boffin
Hero Member
*****
Offline Offline

Posts: 369



View Profile
« on: March 19, 2007, 07:38:50 PM »

Hi guys,

I have been impressed with the modellers flying free flight Rapier powered jet models and wondered if I could put together a ducted fan electric free flight model to avoid the repeat  cost of the expendable motors.

Accordingly I have acquired a 50mm DF driven by a 12 mm GWS IPS brushed motor . I expect 1.5 amps current draw and 70-75gm thrust from a unit without any downstream devices. This is in the same ball park as the quoted thrust for a Rapier L2. I have cadged a plan for a Crossley Bell XS1 which fits the motor well.

Has anybody done this before and what advice would you offer?
How would you treat the duct downstream of the fan? I feel that the fan is a 2nd Law Newtonian device where thrust is a function of d/dt(mv) , so maximising the velocity would maximise the thrust. My options for the downstream duct therefore are........

1. Parallel (surely this is not the way?)
2. Convergent duct accelerating the air stream........what convergent angle and what area ratio?
3. Laval nozzle accelerating the flow to the throat and then expansion?

Anyone with any thoughts/experience/practical results?

Regards, Ian
Logged
PDR
Sr. Member
****
Offline Offline

Posts: 126


Quidquid latine dictum sit, altum sonatur.


View Profile
« Reply #1 on: March 19, 2007, 11:32:15 PM »

Having done a bit of playing with the GWS40 fan with the Fiego 12mm (5800rpm/v) brushless motor, my comments would be:

Thrust is essentially d(mv)/dt as you say, but closing the duct down to boost the efflux velocity simply restricts the mass flow, so the thrust remains about the same. But there is some benefit because it allows you to adjust the efflux velocity to match your desired flying speed. I suspectthe maximum thrust probably occurs with a slightly convergent duct where the convergence is just enough to keep the flow laminar

I don't think you'll get anything like 75g of thrust at 1.5A. The brushless version in the 40mm fan draws up to around 7A from a 3s lipo and develops between 90 and 160g of thrust (depending on intake and duct geometry).

HTH,

PDR
Logged

Personally, I have nothing against work, particularly when performed quietly and unobtrusively, by someone else. I just don't happen to think it's an appropriate subject for an 'ethic'.
boffin
Hero Member
*****
Offline Offline

Posts: 369



View Profile
« Reply #2 on: March 20, 2007, 07:43:15 PM »

PDR,
Thanks for your input. I will not bother too much about the duct downstream. It will be a design of convenience rather than of first principles.......slightly convergent fits the design OK and it can be made easily.

The 75g/1.5 amps comes from a Knight and Pridham supplier quote on a 44 mm brushed KP02 driven DF which on 4.8v goes to 54gm (allegedly) . I corrected for current draw / diameter / voltage for the bigger fan to give 75gm (but very much an estimate).

Another way.......Taking the smallest motor / prop combination that I have data for (Voodoo from Atomic) all these motors run circa 8gm thrust/watt, so rule of thumb, quick calc, ball park expected thrust on a 50 mm DF on 2s1p / 1.5 amps would be close on 90gm......reduce this to 70-75gm for lousy assumptions on efficiency/diameter.

So, have to disagree and if I have to go to 7 amps, the IPS motor and the smoke fairy will make acquaintance. If your data is kosher, this stuffs the project, not enough thrust to make it frightening.

To settle the matter, I will dust off my thrust rig and see what I can achieve......but later.

I too have a 12mm Feigao outrunner but Kv=4100 rev/min/volt. Have just screwed it into a Microstick AUW 110gms. Very fast and powerful motor and super smooth......excellent value for money.
 
Best regards, Ian
Logged
PDR
Sr. Member
****
Offline Offline

Posts: 126


Quidquid latine dictum sit, altum sonatur.


View Profile
« Reply #3 on: March 21, 2007, 11:25:42 AM »

Thanks for your input. I will not bother too much about the duct downstream. It will be a design of convenience rather than of first principles.......slightly convergent fits the design OK and it can be made easily.

The 75g/1.5 amps comes from a Knight and Pridham supplier quote on a 44 mm brushed KP02 driven DF which on 4.8v goes to 54gm (allegedly) . I corrected for current draw / diameter / voltage for the bigger fan to give 75gm (but very much an estimate).

Another way.......Taking the smallest motor / prop combination that I have data for (Voodoo from Atomic) all these motors run circa 8gm thrust/watt, so rule of thumb, quick calc, ball park expected thrust on a 50 mm DF on 2s1p / 1.5 amps would be close on 90gm......reduce this to 70-75gm for lousy assumptions on efficiency/diameter.

So, have to disagree and if I have to go to 7 amps, the IPS motor and the smoke fairy will make acquaintance. If your data is kosher, this stuffs the project, not enough thrust to make it frightening.

To settle the matter, I will dust off my thrust rig and see what I can achieve......but later.

I too have a 12mm Feigao outrunner but Kv=4100 rev/min/volt. Have just screwed it into a Microstick AUW 110gms. Very fast and powerful motor and super smooth......excellent value for money.

There's another variable - the pitch of the fan. I'm rather disdainful of static thrust measurements, having gone to a lot of effort to build an accurate thrust measuring rig when I first got into this electric malarky a few years ago. My rig measures thrust, torque, current, voltage, rpm and a few other tnings (the grandmother's age transducer proved difficult to calibrate but the SWMBO "little job" detector worked flawlessly). Having built the thing I spent a week of evenings collecting data over a range of power systems from 60w (shock flyer) to 700w (Axi 4120/18), and then spent several evenings crunching the numbers in Excel. The overall conclusion what that it tought me nothing useful whatsoever, other than the verify the electric motor output power equation and thus show that the torque/rpm transducers were superfluous.

The reason is simple - the static thrust from any power system can be almost any value you choose - if you you want more you just reduce the prop pitch. This variable is utterly dominant, and completely masks any useful thrust information that you might otherwise get. I've got a datapoint showing nearly two pounds of thrust from a geared 70w system (which would give the shockie a thrust/weight ratio of six to one!) but as the efflux speed is less than 3mph it's essentially useless for any practical purpose other than a helicopter. The rig is sitting on a shelf now, because the exercise told me that I just need to measure the input power and use the motor parameters to derive the output power. This is done in the model, and just uses a current clamp and a DVM - a much simpler solution!

I suspect that your reference fan has a low pitch, which would limit the achievable airspeed of the model - possibly not unacceptably so, but that depends on the model. My reference system (GWS 40 fan, Fiego 5800, Thunderpower 3s480) is used in a Robotbirds LA159 foamy fun-jet. It flys quite well (Mike could give an opinion as he was there when we flew it) but I was hoping for more. This installation has a 250mm(ish) exhaust duct, and I suspect there's a lot more performance to come by playing with the taper-ratio of the duct to match the efflux velocities. To some extent you can use the duct taper to "correct" an inappropriate fan pitch, but the process will be lossy for anything but small corrections.

All IMHO of course - YMMV.

HTH,

PDR
Logged

Personally, I have nothing against work, particularly when performed quietly and unobtrusively, by someone else. I just don't happen to think it's an appropriate subject for an 'ethic'.
mike
Administrator
Hero Member
****
Offline Offline

Posts: 2103



View Profile
« Reply #4 on: March 21, 2007, 11:31:26 PM »

This is good stuff guys.

I'm trying to get on to an instant moderator's training course to put Boffin's thread back on track without spoiling the thrust augmentor stuff in any way.

your neophite moderator,

Mike Fantham
Logged
mike
Administrator
Hero Member
****
Offline Offline

Posts: 2103



View Profile
« Reply #5 on: March 22, 2007, 01:50:57 PM »

I'm trying to get on to an instant moderator's training course to put Boffin's thread back on track without spoiling the thrust augmentor stuff in any way.

Thanks to guidance from Richard, have managed to split this topic and moved Jef's Augmenter cone thread into its own new topic.

regards
Mike Fantham
Logged
mike
Administrator
Hero Member
****
Offline Offline

Posts: 2103



View Profile
« Reply #6 on: March 22, 2007, 05:30:25 PM »

This installation has a 250mm(ish) exhaust duct, .....

I'm confused.

250mm is a large diameter.

250 sq mm is 18 mm diameter.

I guess it's the latter but it sounds small.

regards

MF


Logged
PDR
Sr. Member
****
Offline Offline

Posts: 126


Quidquid latine dictum sit, altum sonatur.


View Profile
« Reply #7 on: March 22, 2007, 05:57:14 PM »

This installation has a 250mm(ish) exhaust duct, .....

I'm confused.

250mm is a large diameter.

250 sq mm is 18 mm diameter.

I guess it's the latter but it sounds small.

You're mixing up the LA159 EDF model with those ramjet trials. The duct in the LA159 is around 250mm LONG; the diameter tapers from 40mm to around 35mm. The ramjet duct is 250mm diameter to get the shredded newspaper to swirl properly to promote reliable ignition in the pre-combustion chamber.

HTH,

PDR
Logged

Personally, I have nothing against work, particularly when performed quietly and unobtrusively, by someone else. I just don't happen to think it's an appropriate subject for an 'ethic'.
Steve_B
Jr. Member
**
Offline Offline

Posts: 14


View Profile Email
« Reply #8 on: October 25, 2007, 10:31:56 PM »

Boffin,
IMHO the Crossley Bell XS-1 (great little design that it is) would not be a good choice for EDF conversion...
For a start it's a model of a rocket powered plane (not a jet) therefore design has no inlet to supply air to the fan. Because it's a rocket the rear nozzle openning is also very small, way too small to be used for an EDF outlet. At the original scale I actually doubt that the fuselage at it's widest point would be big enough to accept an EDF 50. Also the  EDF claimed thust at 70g is about 6 -7 times more than a typical rapier L2 which would have 10 - 12g of thrust. The original XS-1 model is only 13" span and for Rapier power would have an all up weight in the region of 25 - 30g. The EDF unit and battery by themselves would be about 40g, therefore the wing loading would be at least twice that of the Rapier power model (which is no 'floater' in the first place)...

You could get around some of these issues by doubling the size of the model but then you would need to redesign the structure... also the lack of air inlet and the small size of the rear nozzle would still be problems.


If you want a model that would take the EDF 50 (or better still the EDF 40) what about the Aerographics Hawker Hunter?. It's over twice the size of the Bell and is intended for the K&P EDF44, but it's underpowered with that motor so the GWS with LiPo battery may be just what it needs.

Steve
Logged
boffin
Hero Member
*****
Offline Offline

Posts: 369



View Profile
« Reply #9 on: October 26, 2007, 10:51:27 AM »

Steve,

Thank you for your reply. I came to the conclusion that I was treading on thin ice on this endeavour, so I built a Fullarton sheet Rezenbe as a test bed for the EDF50 to clear the air. The spec was 250mah 3.7v LiPo controlled by a FET timer. The AUW came out at a massive 85gm but the model flew exceptionally well free flight with a decent climb out and typical canard pitch stability. I have had much fun with this corruption that the 'proper' DF has taken a back seat. The learning points have been......

1. On a single LiPo cell a GWS EDF50 does not produce 70gm of static thrust....30-40 gms would be a more relaible estimate.
2. This is enough to give an 85 gm Rezenebe a lively performance
3. The 13mm brushed motor draws 1.5 amps at 3.7volts and is more than adequate for free flight. A Feigao brushless would not be a good idea as the model would be too fast and very difficult to trim.

I have EDF50 and EDF40 units to hand and will build something (but not a Bell) Sabre/MIG/Hunter are candidates. Project is in limbo waiting for a mandrel for me to construct an inlet/diffuser system. The 250mah LiPo will be used but I intend  a Zombie or FlytProf controller to limit the current and give me some chance of trimming it out.

Regards Ian





* rezenebe 001.jpg (84.33 KB, 1293x654 - viewed 259 times.)

* rezenebe 003.jpg (45.23 KB, 747x612 - viewed 223 times.)
« Last Edit: October 26, 2007, 01:52:53 PM by boffin » Logged
Steve_B
Jr. Member
**
Offline Offline

Posts: 14


View Profile Email
« Reply #10 on: October 26, 2007, 12:09:37 PM »

The Rezenbe looks like it should be a lively performer.
I think if you go scale then you are likely to need to go to two lipo cells. The EDF 40 and EDF50 are designed to deliver their rated thrust when operated at 7.2V (no surprise you are down on thrust then)
For a scale model you will need to go with a pretty big airframe just so you can get reasonable sized inlet and outlet duct areas. This larger scale airframe is sure to come in much heavier than your Rezenbe. For example the Aerographics Hunter typically comes in around 150g AUW. Your estimated 30 -40g of thrust wont be enough for any model approaching this weight especially the relatively inefficient aerodynamic properties of a scale jet.

Good luck with whatever scale project you do take on.

cheers
Steve
Logged
Pages: [1] Print 
« previous next »
Jump to:  


Login with username, password and session length

Powered by MySQL Powered by PHP Powered by SMF 1.1.16 | SMF © 2006-2008, Simple Machines Valid XHTML 1.0! Valid CSS!
Page created in 0.116 seconds with 19 queries.