4-Max 4S 50mm EDF V2 – What am I doing wrong?

[Futura57]

August 2023 I purchased a FMS V2 50mm EDF 4S 4500kv 11 blade (here) for my new prototype Strange Quark model. The performance numbers quoted are; 32.9A, 486W & 1,086g static thrust. This was my first ever EDF purchase, so on receipt I did a quick check it would run ok using a 4S Lipo, unbranded 40A ESC and servo tester for the throttle. The fan ran smooth so I put it back in the box until recently, when I installed it in my model. The model’s regular intake area is approximately double the fan area and there are additional cheat holes close to the fan, about the same area of the fan. The outlet is 200mm long and decreases to about 85 / 90% of the fan area at the exit. Now that the model is ready for its maiden, I made a few measurements.

 

The model’s AUW is 870g, so I was expecting a thrust to ratio far in excess of 1. I’m using an Overlander 4S 2200mAh Lipo rated 35C (here) and a SkyWalker 50A ESC (here). The throttle-stick / ESC I calibrated in the usual manner.

 

On full throttle my Watt meter measures a maximum 34.84A and 512W. The Voltage drops quite quickly during the first 10 seconds or so from 14.9V down to 14.2V

 

With the current and power measurements 5 to 6% higher than the specification I was expecting good thrust figures. I placed the model nose down on some kitchen scales and zeroed the scales. On full throttle the scales were reading around the 660g mark. I can only surmise that my model’s ducting is woefully inadequate. Removing the hatch with the cheat holes opens up 2 or 3 times the fan area right next to the fan and makes no discernible difference to the measured thrust. So, is my outlet tube the culprit?

 

I’ve never been particularly impressed with my Overlander 4S 2200mAh Lipo performance. On a warm day I barely get more than 4.5 minutes on my EFlite Ultimate 3D, which I do not prop-hang or fly aggressive 3D. I have no way to measure cell internal resistance at this time.

 

I’m beginning to wonder if I have the EDF I ordered and paid for. The box and instructions are not marked with V2 anywhere, but I don’t know if they should.

 

So, what is the likely cause of my underwhelming thrust measurement? I do believe my model will fly adequately, but perhaps I was naïve to expect the specified EDF static thrust. Any suggestions would be appreciated.

 

[Graham Bowers]

I think my first action would be to contact the vendor. He has a very good reputation and is probably best placed to dig down to the required detail.

[Chris Walby]

Voltage sag is very apparent with edf's as they are quite high current demand. Simple solution is a higher C rating lipo

 

I like GNB 2200mAh 4S 100C LiPo's as they give plenty of current without the voltage sag, but other high C rating lipos are available.

 

 

You could parallel two of your 4S2200 together and redo the static test to prove the point around voltage sag!

 

My Fokke Stick uses a 4S3300 at about 120A, but it gets to a point where as you increase the current so the voltage rolls off so I don't actually get any more power as P=VI

   

PS it was a wing left over from a crashed model and was build to do one thing.....go very very fast!

 

PPS - I have a twin EDF that if held at WOT will do 3 min 5 sec and kill the lipo <5% left...don't ask how I know 😞

 

 

Edited January 23 by Chris Walby

[Futura57]

What puzzles me is my current and power reading is a little higher than the EDF specification, yet the thrust is way down at 660g vs 1,085g. Surely my ducting must be suspect to deliver such poor thrust. 🤔

[Lipo Man]

36 minutes ago, Futura57 said:

What puzzles me is my current and power reading is a little higher than the EDF specification, yet the thrust is way down at 660g vs 1,085g. Surely my ducting must be suspect to deliver such poor thrust. 🤔

Actually those figures aren’t at all surprising. The quoted thrust figures for EDF units are usually measured without any intake or exhaust “plumbing” and are never achieved in a model installation. I’d be pretty happy with 660g from a 50mm in a model. Sounds like your intake is about as good as you can get - I’d try to shorten the thrust tube if you can. Also check the percentage of fan swept area it reduces to - too small loses a lot of static thrust. Good luck!

 

Post script - just found video of a thrust test of my FMS 50mm fan in a Tony Nijhuis mini Hunter model. Mine was only a 3s fan and I get just over 500g static thrust. Your figure doesn’t look terrible to me. 🤔

Edited January 23 by Lipo Man

[EvilC57]

35C doesn’t seem a high enough rating for a lipo being expected to power an EDF to me. OK for something with a moderate performance and a propeller maybe, but for an EDF I’d expect to use a higher rated battery - more like 60C or more.

[Futura57]

I have tried a GNB 4S 1550 60C Lipo but I'm getting very similar readings to my Overlander 2200s, including the voltage drop. I may order some GNB 2200/2300s 100C to maintain a similar CG.

[Simon Chaddock]

Fuura57

I have to agree with LiPo man.

The published figures will be for the EDF in "free air" with no ducting and the bell mouth fitted. This static thrust figure will not be equalled with an inlet and exhaust duct in place. The longer and more complex shape the ducts are the bigger will be the "duct losses".

What most EDF can make use of is the the "flying" thrust which can be a bit more than the static thrust due to the planes forward motion creating some dynamic inlet pressure. This is different to most props where the thrust falls away as the airspeed rises. 

This explains the use of catapult launches for some models and the long take off run for EDFs with undercarriage.

Ducts basically have two loss making features. Duct wall skin friction that slows down the boundary layer next to the skin and pressure drops caused by changes in the airflow direction as this is likely to cause additional turbulence slowing down the air velocity.

The primary duct requirement is that the walls of both the inlet and exhaust are as smooth as possible to keep the boundary layer as thin as possible. The inlet duct should have a constant area from the inlet to a point where the duct can be gradually reduced to exactly match the EDF body.

 

Your inlet duct certainly sounds to have sufficient area. A fact rather confirmed by the fact when you enlarged the cheat hole it made no significant difference. At the inlet area you state you really don't really need cheat holes anyway.

In principle with an EDF unless your ducting has obvious limitations that can be corrected you just have to live with what you get.☺️

I hope this helps.

Edited January 23 by Simon Chaddock

[Futura57]

Thanks guys. It seems my expectations are a little high 🤔. It is my first EDF experience after all. With the EDF on full chat the model certainly feels like it will take a hand launch easily, so she's ready to go when the weather plays ball.

[FlyinFlynn]

...Also remember that static thrust is measured in Chinese grams.

[EvilC57]

And that the static thrust figures the manufacturers quote will likely be somewhat optimistic, like car manufacturers MPG figures.

[PeterF]

I carried out some proper static thrust tests on Vasa 55mm fans with Tenshock motors (high quality gear) and at 490W I could only get 820g thrust bare fan from a 55mm fan. I find the claimed >1000g thrust at 480W to be too high, getting 20% more thrust from a fan with an FSA that is 20% smaller is just not realistic. Reducing fan size means that to get the same thrust you need a higher velocity because you have less flow area. Moving a smaller mass flow to a higher velocity is always more energy intensive which is inherent in fans, thrust is proportional to velocity^2 and power is proportional to velocity^3, ye can't break the laws of physics.

[George Worley - 4-Max]

Not sure why you didn't talk to us as your first port of call.

If we sold them as V2 then they are V2. there is no difference in the markings to distinguish between a V1 and a V2.

You say that the outlet is chocked down 85-90%, is the compaired to the fan diameter or to the FSA?

Any reduction of the exhaust tube will greatly reduce the thrust but it will increase the air speed.

If you want more thrust increase the diameter of the exhaust tube.

Is the inlet ring still fitted?

We recommend a 60C good quality battery, like one of ours

I would also try a 4S 2600mAh if you have the room.

Have you set the ESC to "high" timing?

Is the battery fully charged when conducting these tests?

Battery temperature at this time of the year will affect it's performance.

Is the battery warm, around 20 degrees when conducting the tests?

The figures we quote are in free air with the inlet ring and at nominal voltage.

As soon as you fit it in a model those figures will drop and it will vary from model to model.

 

Hope this helps

[Martin McIntosh]

I was about to suggest a 4-Max battery but was beaten to it, they really are a better option to the ones you are using.

I have a Hawk, 90mm 12s fan, as yet unflown and despite some mods the static thrust is very disappointing indeed.

[Futura57]

On 24/01/2024 at 10:20, George Worley - 4-Max said:

Not sure why you didn't talk to us as your first port of call.

If we sold them as V2 then they are V2. there is no difference in the markings to distinguish between a V1 and a V2.

You say that the outlet is chocked down 85-90%, is the compaired to the fan diameter or to the FSA?

Any reduction of the exhaust tube will greatly reduce the thrust but it will increase the air speed.

If you want more thrust increase the diameter of the exhaust tube.

Is the inlet ring still fitted?

We recommend a 60C good quality battery, like one of ours

I would also try a 4S 2600mAh if you have the room.

Have you set the ESC to "high" timing?

Is the battery fully charged when conducting these tests?

Battery temperature at this time of the year will affect it's performance.

Is the battery warm, around 20 degrees when conducting the tests?

The figures we quote are in free air with the inlet ring and at nominal voltage.

As soon as you fit it in a model those figures will drop and it will vary from model to model.

 

Hope this helps

 

Thank you for your suggestions @George Worley - 4-Max

I've just had another look at my EDF setup...

 

It turns out my exit pipe is hardly chocked at all. I forgot I had made a minor change during construction compared to my original plan from nearly three months ago. After reading advice in this EDF forum my original design aim was for 85 - 90%, considered optimal. My actual measured pipe exit area is 2100 sq mm. If I calculate area using fan casing body diameter of 52.5mm (which is the inside diameter of the front of my 200mm long exit tube) gives area = pi x 26.25 x 26.25 = 2165 sq mm. So ratio is 2100 / 2165 = 97%. If I use the 50mm diameter as basis then exit is bigger at 107%. I'm not sure which is FSA, but tube exit / inlet is close to unity either way.

 

The inlet ring is still fitted.

ESC timing is set to High. Actually it made no discernible difference to the current/power/thrust when I changed it to medium. Maybe this means my ESC is garbage?

These tests I carried out with the fan hatch removed, giving additional 'cheat hole' area over three times the fan area in close proximity to the fan.

The battery is fully charged.

Everything is being done at room temperature 20/21 degrees Celsius. And at sea level 🤣

 

I will test fly the model to assess the performance and duration. I will then decide if it's worth investing in high C lipos and perhaps a better quality ESC.

 

 

[Lipo Man]

1 hour ago, Futura57 said:

 

Thank you for your suggestions @George Worley - 4-Max

I've just had another look at my EDF setup...

 

It turns out my exit pipe is hardly chocked at all. I forgot I had made a minor change during construction compared to my original plan from nearly three months ago. After reading advice in this EDF forum my original design aim was for 85 - 90%, considered optimal. My actual measured pipe exit area is 2100 sq mm. If I calculate area using fan casing body diameter of 52.5mm (which is the inside diameter of the front of my 200mm long exit tube) gives area = pi x 26.25 x 26.25 = 2165 sq mm. So ratio is 2100 / 2165 = 97%. If I use the 50mm diameter as basis then exit is bigger at 107%. I'm not sure which is FSA, but tube exit / inlet is close to unity either way.

 

The inlet ring is still fitted.

ESC timing is set to High. Actually it made no discernible difference to the current/power/thrust when I changed it to medium. Maybe this means my ESC is garbage?

These tests I carried out with the fan hatch removed, giving additional 'cheat hole' area over three times the fan area in close proximity to the fan.

The battery is fully charged.

Everything is being done at room temperature 20/21 degrees Celsius. And at sea level 🤣

 

I will test fly the model to assess the performance and duration. I will then decide if it's worth investing in high C lipos and perhaps a better quality ESC.

 

 

Your tube definitely isn’t choked - that’s actually a lot more than the fan swept area. The outer diameter of the fan is 50mm, so that’s the size to use in your calculation. However, for swept area you have to subtract the area blanked off by the motor - which is probably about 26mm diameter. Once you take that away and take your 90% calculation you’re actually aiming for something more like 1,290 square millimetres.  I don’t know what the effect of having a tube that widens rather than contracts would be - I’d imagine you could lose some max speed. 
Good luck with your test flight!

[Futura57]

5 hours ago, Lipo Man said:

Your tube definitely isn’t choked - that’s actually a lot more than the fan swept area. The outer diameter of the fan is 50mm, so that’s the size to use in your calculation. However, for swept area you have to subtract the area blanked off by the motor - which is probably about 26mm diameter. Once you take that away and take your 90% calculation you’re actually aiming for something more like 1,290 square millimetres.  I don’t know what the effect of having a tube that widens rather than contracts would be - I’d imagine you could lose some max speed. 
Good luck with your test flight!

Understood @Lipo Man. Thanks.  I really have cocked-up a bit on my first EDF attempt. But this could be good news because in version 2 I can significantly reduce the cross-sectional area of the back end of the exit tube. This will leave sufficient room at the rear fuselage to mount the elevon servos, relocating them from the front fuselage and removing the  need for snake links. Then I can use a slightly heavier higher capacity lipo to rebalance at the CG.  I will first fly as is. Then I may experiment by adding a short convergent nozzle at the exit. In a later version I may also incorporate a vectored thrust nozzle. It's all fun at the fun factory.

Edited January 25 by Futura57

[Lipo Man]

1 hour ago, Futura57 said:

Understood @Lipo Man. Thanks.  I really have cocked-up a bit on my first EDF attempt. But this could be good news because in version 2 I can significantly reduce the cross-sectional area of the back end of the exit tube. This will leave sufficient room at the rear fuselage to mount the elevon servos, relocating them from the front fuselage and removing the  need for snake links. Then I can use a slightly heavier higher capacity lipo to rebalance at the CG.  I will first fly as is. Then I may experiment by adding a short convergent nozzle at the exit. In a later version I may also incorporate a vectored thrust nozzle. It's all fun at the fun factory.

It may well fly great as it is. Only one way to find out! Good luck!

[PeterF]

We are getting some answers here. As stated above fsa is fan swept area, i.e. the area that the fan blades sweep, the annulus between the ID of the shroud and the OD of the motor housing and to get a 100% fsa exhaust the diameter of the exhaust duct MUST be smaller than the fan shroud, i.e. the exhaust duct is always a cone reducing to the outlet to compensate for the motor. Keeping the exhaust duct at the same diameter as the fan shroud is bad news because it reduces the thrust because the air slows down. I noted above that I did some detailed thrust checks on some 55mm fans and I also looked at the effect of different sized exhaust ducts and a parallel exhaust duct with 127% fsa reduced the thrust from 825g to 739g but reduced the power as there was some energy recovery as the air velocity decreased. See the table below.

 

Duct type	FSA, %	rpm	Thrust, g	Power, W	g/W	Vel m/s
No duct	100%	55260	825	492	1.68	60.5
Round	127%	55380	739	461	1.60	50.8
Oval	100%	54120	833	503	1.66	60.8
Round	100%	54300	823	504	1.63	60.5
Round	95%	53700	823	499	1.65	62.0
Round	90%	53580	820	509	1.61	63.6
Round	85%	53400	836	517	1.62	66.1
Round	80%	53040	829	523	1.59	67.9

[Simon Chaddock]

Futura57

The way I look at it is you can afford to have slower moving air in the inlet duct. Even at 1.2 the FSA the air will still be travelling faster than most planes fly so the inlet is actually "sucking" a little bit can actually create negative drag. Obviously if the inlet is too big the air entering will be slowed down and will create positive drag.

However the efficiency rule for any low pressure duct is what ever velocity it has - keep it!  Any change in area will result in a change of velocity that will require energy to restore, so to limit inlet losses its area what ever it is should remain constant all the way to the EDF.  

Duct efficiency is particularly noticeable on the exhaust duct where the air velocity is higher. 

The air has maximum velocity as it leaves the fan and at that point the duct is the fan swept area. Then what usually happens is it is dumped into a duct the size of the fan case which is about 1.2 times the FSA along with a relatively square end to the motor that creates turbulence. All this looses some energy from the air stream and thus thrust.

Ideally the motor should have a long tapering cone on it, although not really practical on many EDFs and the duct diameter should be reduced at the same time to maintain the FSA and thus maintain the fan airstream velocity. Finally how quickly the exhaust duct reduces to 85% FSA is a balance between increasing the air velocity unnecessarily early in the duct and the efficiency of the final 85% nozzle.

 

From all this in any EDF "system" it is not the ultimate power itself that is the final decider but how well the "system" converts that power into thrust.

There is no absolute "right" only a whole series of compromises unique to a particular airframe and duct layout, hence the fascination.      

 

[Simon Chaddock]

I know I am going a bit over the top but this picture shows what I mean about the EDF exhaust of a typical 50 mm EDF.

You can see that the motor has a larger diameter at 26mm  than the fan hub at 22mm. Si in theory the exhaust duct should be a bit larger than the fan case to maintain the FSA. Then at the end of the motor the duct suddenly becomes less tan the FSA. In this case to make matters worse there must be considerable turbulence from the bearing support lugs spinning at 30,00+ rpm. This must play havoc trying to achieve any sort of smooth airflow even if the duct is reduced to the FSA in a reason able manner.

This is my attempt at a printed duct to go from the fan case to an FSA size duct.

The duct fits over the EDF case and initially maintains the case diameter. It expands slightly over the motor area and then smoothly reduces to the true FSA diameter.

Of course it would be better if the motor had a cone fitted of the rear end to reduce the turbulence and to maintain the FSA over the length of the duct reduction. This particular fan & motor are so well balanced I rather "chickened out" of attempting one.

Does it make any difference? Very hard to tell as the benefit will be small but it does make me feel better for at least trying.   

 

[Futura57]

On 26/01/2024 at 10:37, PeterF said:

We are getting some answers here. As stated above fsa is fan swept area, i.e. the area that the fan blades sweep, the annulus between the ID of the shroud and the OD of the motor housing and to get a 100% fsa exhaust the diameter of the exhaust duct MUST be smaller than the fan shroud, i.e. the exhaust duct is always a cone reducing to the outlet to compensate for the motor. Keeping the exhaust duct at the same diameter as the fan shroud is bad news because it reduces the thrust because the air slows down. I noted above that I did some detailed thrust checks on some 55mm fans and I also looked at the effect of different sized exhaust ducts and a parallel exhaust duct with 127% fsa reduced the thrust from 825g to 739g but reduced the power as there was some energy recovery as the air velocity decreased. See the table below.

 

Duct type	FSA, %	rpm	Thrust, g	Power, W	g/W	Vel m/s
No duct	100%	55260	825	492	1.68	60.5
Round	127%	55380	739	461	1.60	50.8
Oval	100%	54120	833	503	1.66	60.8
Round	100%	54300	823	504	1.63	60.5
Round	95%	53700	823	499	1.65	62.0
Round	90%	53580	820	509	1.61	63.6
Round	85%	53400	836	517	1.62	66.1
Round	80%	53040	829	523	1.59	67.9

 

Thank you @PeterF At the end of the day, depending upon the model, I could be prepared to accept, say, 739g versus 836g IF it saves a whole lot of complexity and fiddling around. BUT, clearly any exit duct converging to between 100% and 85% FSA gives near optimal performance to within a few grammes and Watts. So I don't see it as super-critical. I do, however, need to get my model in this range and I should see the greatest gains.

 

 

 

1 hour ago, Simon Chaddock said:

I know I am going a bit over the top but this picture shows what I mean about the EDF exhaust of a typical 50 mm EDF.

You can see that the motor has a larger diameter at 26mm  than the fan hub at 22mm. Si in theory the exhaust duct should be a bit larger than the fan case to maintain the FSA. Then at the end of the motor the duct suddenly becomes less tan the FSA. In this case to make matters worse there must be considerable turbulence from the bearing support lugs spinning at 30,00+ rpm. This must play havoc trying to achieve any sort of smooth airflow even if the duct is reduced to the FSA in a reason able manner.

This is my attempt at a printed duct to go from the fan case to an FSA size duct.

The duct fits over the EDF case and initially maintains the case diameter. It expands slightly over the motor area and then smoothly reduces to the true FSA diameter.

Of course it would be better if the motor had a cone fitted of the rear end to reduce the turbulence and to maintain the FSA over the length of the duct reduction. This particular fan & motor are so well balanced I rather "chickened out" of attempting one.

Does it make any difference? Very hard to tell as the benefit will be small but it does make me feel better for at least trying.   

 

 

Understood completely @Simon Chaddock and thank you. Since my fabrication techniques are pre-historic single curvature tubes (i.e. not 3D printing) then close enough is good enough for me. More time to browse the BMFA News 😜