Experimental EDF Jets and Other Ideas

[telnar1236]

Now that I have a day behind me, I have a couple conclusions from the v2 flight. The big one is that I feel ready to start getting more into designing the 80mm version. Performance was good and it was stable, especially when at speed. Slow flight will be improved a lot by flaps, but it was fine, even if it looked terrible. Based on the control authority the stabilators gave me, I think I need to include ailerons in the big version - especially at low speeds they weren't really as effective as I would have liked and I had to set the roll deflection the same as the pitch deflection.
I also did some static testing to determine the thrust of the EDF when installed in this setup - I measured 620g which fell off to around 580g over time which suggests that this wasn't a very good installation of the 50mm EDF unit and goes a fair ways to explain why the speed wasn't as fast as I was hoping. The CFD predicted drag at a 91 mph speed was 490g which isn't all that much of a loss of thrust so I'm feeling better about how it performed.
I also printed a shorter version of the duct for the 80mm EDF to make sure the inlet wasn't overly restrictive (I was worried that was why the 50mm EDF was underperforming) and thankfully it looks like the issue is the lack of the EDF bell mouth in this setup and not the inlet size. Uninstalled, the 80mm EDF gives me 3.2 kg of thrust with a bell mouth, but only 1.6 kg of thrust without. With the duct, it was boosted back up to 2.2 kg. This isn't as high as I would have liked compared to the thrust with the bell mouth, but it is still plenty to do what I want it to. I also have some ideas about how I might be able to improve this value without increasing inlet area by changing the internal duct geometry to get more thrust out of it.

 

[L Edge]

I appreciate the compliments but I'm really a pretty average jet pilot. They're certainly not beginner planes and I've been flying for a while, but they're also not too bad to fly. The takeoff looks extremely sketchy and looked the same every flight, but I barely touched the sticks - it rocks back and forth and looks awful, but then it recovers itself and flies off normally.

Orientation is every bit as much of a problem as you're saying though - it's pretty hard to see against the clouds. Incidentally this is why I had such poor luck with the pylon racer design - it's somewhat smaller and even on 4s about 20% faster and I couldn't keep track of it - as a plane the pylon racer was actually nicer to fly than the jet.

Point I am trying to get across is if you want to be a jet jockey, you need to start flying EDF's now. Look at your set up to land, for example, what it takes to bring it in unharmed.

You just don't drop a foamy and pick up an EDF and do what you do with it. Worst case setup is tool along full open, EDF dies at 100mph, can you turn it around, float it in and land without damage. How do you slow down and flare(gear or no gear)(ride the ground effects) to prevent crashes. You should be able to do that with a foamy. Do you feel the stall coming on?

Did you watch the TV program showing training Navy pilots for carrier duty. If you are unaware and didn't learn by practice, no way are you going to hit the deck(night time worse) and be a jet jockey.

 

[telnar1236]

I've continued to mess around with thrust testing my 80mm EDF with slight tweaks to the inlets. I've tested 6 different setups now, as follows from worst to best:

1. EDF with no inlet: 1600 g thrust
2. EDF with sharp 100% FSA inlet and internal bell mouth: 1950 g
   
   
3. EDF with sharp 100% FSA inlet and straight duct to fan: 2200 g
   
   
4. EDF with sharp 110% FSA inlet and straight duct to fan: 2350 g
   
   
5. EDF with rounded 100% FSA inlet: 3000 g
   
   
6. EDF with stock bell mouth: 3200 g

The best was the bell mouth inlet which barely lost any thrust at all so now I need to figure out how to incorporate that into my airframe. To start with I've rounded the inlet mouth with a 4mm radius since this is the largest radius I could achieve without needing to change the overall shape of the fuselage.

This is a fascinating result, and not at all what I expected. It suggests that the vast majority of the losses in the ducting occur due to the turbulence immediately after the inlet and that sharp inlet lips kill somewhere around 1/3 of the thrust. Part of why I find this fascinating is that sharp inlet lips appear all over the place in both store-bought planes and homemade designs including most of my own designs. I assumed the losses were maybe around 5-10%, but this seemed to have a much greater impact than not giving enough inlet area even. I currently have the final designed inlet geometry printing, to make sure I'm not assuming the bell mouth I tested translates to a rounded inlet lip incorrectly.

 

[Houndpup Rc]

I've continued to mess around with thrust testing my 80mm EDF with slight tweaks to the inlets. I've tested 6 different setups now, as follows from worst to best:

1. EDF with no inlet: 1600 g thrust
2. EDF with sharp 100% FSA inlet and internal bell mouth: 1950 g
   View attachment 254158
3. EDF with sharp 100% FSA inlet and straight duct to fan: 2200 g
   View attachment 254157
4. EDF with sharp 110% FSA inlet and straight duct to fan: 2350 g
   View attachment 254159
5. EDF with rounded 100% FSA inlet: 3000 g
   View attachment 254160
6. EDF with stock bell mouth: 3200 g

The best was the bell mouth inlet which barely lost any thrust at all so now I need to figure out how to incorporate that into my airframe. To start with I've rounded the inlet mouth with a 4mm radius since this is the largest radius I could achieve without needing to change the overall shape of the fuselage.
View attachment 254153
View attachment 254154
View attachment 254155
This is a fascinating result, and not at all what I expected. It suggests that the vast majority of the losses in the ducting occur due to the turbulence immediately after the inlet and that sharp inlet lips kill somewhere around 1/3 of the thrust. Part of why I find this fascinating is that sharp inlet lips appear all over the place in both store-bought planes and homemade designs including most of my own designs. I assumed the losses were maybe around 5-10%, but this seemed to have a much greater impact than not giving enough inlet area even. I currently have the final designed inlet geometry printing, to make sure I'm not assuming the bell mouth I tested translates to a rounded inlet lip incorrectly.

Wow, very interesting!

 

[telnar1236]

I tested the actual inlet for the plane, and got 2950 g, so I'm feeling pretty happy with that. I'm not sure if the slightly lower thrust when compared to the 10mm bell mouth is due to the smaller 4mm radius, different profile, or battery being slightly lower since I'd tested a couple of the other profiles in between, but it's pretty much the same thrust as the 10mm bell mouth and I don't need to redesign the entire fuselage, so I'll take it.

I'm going to see if I can cut just the inlet off the 50mm version to replace it with the new design to see how much of an improvement it gives at higher speeds since I'm not sure if this is just a low-speed thing, or something that will benefit the plane throughout its speed range.

 

[L Edge]

I tend to look at the dynamic losses that are greater after the fan. Statically, you are sucking the flow in, dynamic, your getting ram air and from experience, it comes from after the fan (higher internal pressure and faster flow) and its the bends and transition shape (turbulence) and frictional loses of the ducting.

Tried circular, oval, rectangular, and this one below.

Did a number of flights with different shapes doing transitions as smooth as I could. This one was a dog, Rough estimate to measure speed between 50 ft using toilet paper on grass. I like your suggestion on EDF at exit.(think X-29, just like horizon's SR-71 would work.)

On my SR-71, after it was flying, added plastic nose cones for the nose. To see how much loss it had, notice acceleration was slow when I threw it, when up to speed with ram air , it was definitely slower with nose cone(choked -EDF about 6" downstream) and 12'" exhaust ducting.

That was my approach to EDF's. I never reduce exit area over 5% and as always, designed to keep it light.
I did help my F-117 improve thrust at the exhaust since it was really quite rectangular and losses were great.

 

[quorneng]

"EDF at the back"
The reason I went for that configuration was most EDF planes exhaust ducts simply extend the EDF body and only reduced it further down stream. This meant there was a sudden increase in duct area (about 20%) immediately after the motor. Even if the motor had an 'after body' to streamline the flow it would still mean the exhaust airflow slowed by 20% unless the duct exactly maintained the FSA. In most cases the duct was only reduced by the "nozzle" with an area of say 85% of the FSA. Such significant changes in duct area are far from ideal as far a airflow losses are concerned.
I felt as EDF manufacturers tended to quote maximum thrust figures of an EDF in 'free air' with no exhaust duct or inlet apart from the bell mouth. I felt that if the EDF had to be ducted for appearance it was better to have it exhausting directly to air as done by the manufacturer's testing and then try to make the inlet duct as generous as possible to try to keep the inlet duct airspeed similar to the plane's flying speed i.e. the fan would see no inlet suction.

 

[telnar1236]

Thankfully nozzle design has a couple fewer moving parts than inlet design. Even though you have the sudden change in area from the end of the motor, it's still significantly better to have the fan in a duct. You get turbulence and thrust loss from the end of the motor can regardless, so it's better to have it in a duct where this will result in higher pressure behind the motor and less separation is my thinking.
This almost 20 year old manual for the 55mm hyperflow EDF is still the best documentation I've ever seen from a manufacturer on EDF performance. It shows that a straight duct (134% FSA, equivalent to no duct) has less thrust than a 99% FSA nozzle which is pretty near equivalent to a 90% FSA nozzle. I wish more manufacturers would release this kind of information because it would take a lot of the guesswork out of the design process.

In an effort to try and reduce the effect of the sharp end of the motor, I've incorporated a shorter reduction in area at the end of the motor into some of my designs, but I'm not sure how well it works.

I want to test if this design or a more continuous reduction in area results in more thrust at some point, but both seem very efficient. I used the design in the picture in a test nozzle for the 80mm EDF that I'm using in this project and it achieved the same thrust as with no nozzle at all to within my ability to measure it, or possibly slightly higher thrust even ( it's hard to tell what is the result of the geometry and what is the result of other factors). And if you're going to go fast, a reduced nozzle area is absolutely essential. A great example is this youtube video - the same plane without a nozzle achieved only 134 mph, but with an 85% area nozzle, it achieved 158 mph - a substantial increase with the nozzle being the only change.

 

[L Edge]

Thankfully nozzle design has a couple fewer moving parts than inlet design. Even though you have the sudden change in area from the end of the motor, it's still significantly better to have the fan in a duct. You get turbulence and thrust loss from the end of the motor can regardless, so it's better to have it in a duct where this will result in higher pressure behind the motor and less separation is my thinking.
This almost 20 year old manual for the 55mm hyperflow EDF is still the best documentation I've ever seen from a manufacturer on EDF performance. It shows that a straight duct (134% FSA, equivalent to no duct) has less thrust than a 99% FSA nozzle which is pretty near equivalent to a 90% FSA nozzle. I wish more manufacturers would release this kind of information because it would take a lot of the guesswork out of the design process.
View attachment 254176
In an effort to try and reduce the effect of the sharp end of the motor, I've incorporated a shorter reduction in area at the end of the motor into some of my designs, but I'm not sure how well it works.
View attachment 254177
I want to test if this design or a more continuous reduction in area results in more thrust at some point, but both seem very efficient. I used the design in the picture in a test nozzle for the 80mm EDF that I'm using in this project and it achieved the same thrust as with no nozzle at all to within my ability to measure it, or possibly slightly higher thrust even ( it's hard to tell what is the result of the geometry and what is the result of other factors). And if you're going to go fast, a reduced nozzle area is absolutely essential. A great example is this youtube video - the same plane without a nozzle achieved only 134 mph, but with an 85% area nozzle, it achieved 158 mph - a substantial increase with the nozzle being the only change.

Been watching this guy for a long time. You show the 158 run, did you take a look his previous run video with the 134 and notice the difference? I would not agree about the exhaust nozzle info. What do you think?

 

[telnar1236]

Been watching this guy for a long time. You show the 158 run, did you take a look his previous run video with the 134 and notice the difference? I would not agree about the exhaust nozzle info. What do you think?

I can't vouch for the accuracy of either but both seem plausible. The EDF used in that plane is certainly capable of sending a plane to 150+ mph as there are a ton of videos of the Freewing Zeus going that fast and I don't see anything obviously wrong with the 3D printed design, so I think I believe he hit 158.
From a theory perspective, I think the thrust tube does a lot to reduce turbulence, but I haven't actually gotten thrust numbers with vs. without. If you have numbers like that, it would be great to see them.

Testing and analyzing different configurations is what this project is about, so I guess I should probably look at a range of different options on my test stand and see how they perform, and then maybe stick them on a plane and see what happens.

 

[L Edge]

I can't vouch for the accuracy of either but both seem plausible. The EDF used in that plane is certainly capable of sending a plane to 150+ mph as there are a ton of videos of the Freewing Zeus going that fast and I don't see anything obviously wrong with the 3D printed design, so I think I believe he hit 158.
From a theory perspective, I think the thrust tube does a lot to reduce turbulence, but I haven't actually gotten thrust numbers with vs. without. If you have numbers like that, it would be great to see them.

Testing and analyzing different configurations is what this project is about, so I guess I should probably look at a range of different options on my test stand and see how they perform, and then maybe stick them on a plane and see what happens.

What I was hoping for was to see what the actual value of with/without exhaust ducting speed runs. With the gusty winds/calm day data, how can you compare? Top speed, I am not interested in, yes, prove, choking it really pays in lots of thrust with good reliable data. When EDF's first came out years ago, at the club field I would measure the exhaust diameter(especially TVN) and get their name EDF brand and size, and and compare data found on the internet where most did not exceed 5% area. Tried some at 5% less.

That's why I just explore exhaust ducting shapes and effects.. Hey, when I see a reduction in speed just visually, at least I now know what I have to improve if possible. Some you can solve, others not. The F-117 was one I got it to fly, not the fastest. And no gyro is needed.

It is more fun to get a jet flying that whether it does 96 or 98mph?

 

[Mr Man]

What I was hoping for was to see what the actual value of with/without exhaust ducting speed runs. With the gusty winds/calm day data, how can you compare? Top speed, I am not interested in, yes, prove, choking it really pays in lots of thrust with good reliable data. When EDF's first came out years ago, at the club field I would measure the exhaust diameter(especially TVN) and get their name EDF brand and size, and and compare data found on the internet where most did not exceed 5% area. Tried some at 5% less.

That's why I just explore exhaust ducting shapes and effects.. Hey, when I see a reduction in speed just visually, at least I now know what I have to improve if possible. Some you can solve, others not. The F-117 was one I got it to fly, not the fastest. And no gyro is needed.

It is more fun to get a jet flying that whether it does 96 or 98mph?

True, but if it was 96vs140mph… 🤣

 

[telnar1236]

What I was hoping for was to see what the actual value of with/without exhaust ducting speed runs. With the gusty winds/calm day data, how can you compare? Top speed, I am not interested in, yes, prove, choking it really pays in lots of thrust with good reliable data. When EDF's first came out years ago, at the club field I would measure the exhaust diameter(especially TVN) and get their name EDF brand and size, and and compare data found on the internet where most did not exceed 5% area. Tried some at 5% less.

That's why I just explore exhaust ducting shapes and effects.. Hey, when I see a reduction in speed just visually, at least I now know what I have to improve if possible. Some you can solve, others not. The F-117 was one I got it to fly, not the fastest. And no gyro is needed.

It is more fun to get a jet flying that whether it does 96 or 98mph?

Sounds like I need to test

 

[L Edge]

True, but if it was 96vs140mph… 🤣

But I do have a do have a DarkStar and others that hits between 80 and 100mph.

 

[Piotrsko]

True, but if it was 96vs140mph… 🤣

68 %. in scale, not even significant and measurement errors are higher than that. Even a doppler speed gun is going to be off most of the time for our purposes. Remember math class and right triangles?

 

[telnar1236]

But I do have a do have a DarkStar and others that hits between 80 and 100mph.

Not too shabby

 

[telnar1236]

What I was hoping for was to see what the actual value of with/without exhaust ducting speed runs. With the gusty winds/calm day data, how can you compare? Top speed, I am not interested in, yes, prove, choking it really pays in lots of thrust with good reliable data. When EDF's first came out years ago, at the club field I would measure the exhaust diameter(especially TVN) and get their name EDF brand and size, and and compare data found on the internet where most did not exceed 5% area. Tried some at 5% less.

That's why I just explore exhaust ducting shapes and effects.. Hey, when I see a reduction in speed just visually, at least I now know what I have to improve if possible. Some you can solve, others not. The F-117 was one I got it to fly, not the fastest. And no gyro is needed.

It is more fun to get a jet flying that whether it does 96 or 98mph?

I did my testing and it seems like it very much depends on the EDF which is interesting.

For the 50mm Powerfun EDF L Edge is completely right and the best static thrust is achieved with no thrust tube of any type. My design for the area ruled thrust tube is tied for second worst, so with this fan, it certainly doesn't work. By far the worst is just a straight thrust tube with no taper resulting in a 117% FSA design. That loses almost 16% of the total thrust and also would be very poor for speed. If a thrust tube is necessary, a 100% FSA thrust tube is the best for static thrust and if going purely for speed an 80% FSA thrust tube is probably the best since it only has about 2.3% less thrust than the 100% FSA design but a much higher efflux speed due to the smaller nozzle area.

However, for the 80mm FMS fan, a thrust tube very slightly boosts thrust

I have some suspicions as to why this is the case. Specifically, the FMS fan uses an inrunner which means the motor is a slightly smaller diameter than the hub of the fan which means the overall EDF has a pretty much straight line along the side of the internal parts of the EDF. However, the Powerfun EDF uses an outrunner motor with a diameter larger than the hub of the fan which results in the duct area being reduced when the motor starts. This is supported by the fact that the threaded adapter for the various thrust tubes reduced the thrust by the same amount as having the threaded adapter plus the 100% FSA thrust tube despite being extremely short so that the motor was actually still sticking out the back. Most of the losses seem to occur in the red boxes, so thrust tube length doesn't seem to be all that detrimental, at least if the length is reasonable.

 

[telnar1236]

Next step is to build some kind of experimental plane and do some speed tests

 

[L Edge]

Next step is to build some kind of experimental plane and do some speed tests

I wish you could try this one.
Build a 3 D cone directly behind the motor(small gap for cooling?) based on a 1 pylon support to hold it firm (part of your exhaust duct) that you screw on that is in line with one of the three EDF fan supports. I believe by controlling the drastic expansion area you should be able to improve the exhaust thrust.
This is based on all the exhaust shapes that the drastic shape changes certainly reduces the thrust as I have tested. You might want to explore shapes of the cone if it feasible.

 

[telnar1236]

I wish you could try this one.
Build a 3 D cone directly behind the motor(small gap for cooling?) based on a 1 pylon support to hold it firm (part of your exhaust duct) that you screw on that is in line with one of the three EDF fan supports. I believe by controlling the drastic expansion area you should be able to improve the exhaust thrust.
This is based on all the exhaust shapes that the drastic shape changes certainly reduces the thrust as I have tested. You might want to explore shapes of the cone if it feasible.

It's something I've thought of looking at, but I'm a bit worried about damaging my EDF units. My understanding is that the cooling air flows through the EDF from the back to the front so whatever design I come up with would still have to supply the back of the EDF with high pressure air