Experimental EDF Jets and Other Ideas

[telnar1236]

Ok, this time it didn't self-destruct as soon as I tried a hard turn - that said, I still never reached the top speed despite a successful flight. About 3/4 throttle on 4s is the limit of my flying skill which is 110 mph in more or less level flight and low wind - anything more than that and the plane gets away too fast.

Like with other fast prop planes, it also wants to roll at full throttle until it gains speed which made orientation even more of a challenge. I'm guessing it doesn't have too much more in it on 4s, maybe another 5-10 mph at full throttle. Based on the increase in power at 6s, top speed should be somewhere upwards of 140 mph, but I am not a good enough pilot to do that safely. The 4s top speed matched my prediction from CFD pretty well, so I think my 150 mph level flight estimate is probably about right on 6s.

Unfortunately, on the second flight, the ESC decided to light on fire - it was a 60 A ESC in a plane that draws at most 50 A, but it had been crashed probably 4 or 5 times which was why I was using it in this plane - because it might crash again - and something in it must have reached the point of failure. The plane hit the ground going over 100 mph and obliterated itself. These are the parts that were big enough to pick up - probably about half the plane was in 1/4" confetti after the crash. If you look closely at the ESC, you can see where it caught fire.

I don't think I'm going to rebuild this, at least in its current form. It's too fast/small for me to fly safely and it has served its purpose fully to demonstrate the performance of the wing and tail combination.

I'll upload the STLs in this thread in case a better pilot is interested in building it, with the caution that you really need to be a good pilot to handle something this small and fast. It gets away from you quick.

 

[telnar1236]

So, on to the lessons learned from this design.

1. The overall configuration of my custom laminar flow airfoil on the wing and a NACA 0008 airfoil on the tail which is in line with the wing works well and provides a wide speed range. Takeoff was very easy relative to other attempted fast planes I've built, and the top speed was very high for something this small.
2. I need to design a larger airframe to help maintain orientation at these speeds. I regularly fly 100+ mph jets, but they are 3x the pylon racer's size which makes them much more manageable. Therefore, I'm going to go with one of my 80mm EDF designs instead of a twin 50mm EDF, since I doubt I will be able to keep track of something the size of a twin 50mm jet at 150 mph.
3. My fast jet should have a separate BEC for safety. It's something I already do in my larger jets, like my F-106, but the ESC failure today confirms its importance.
4. My fast jet will need dedicated airbrakes in addition to its flaps in case I need to belly land. In general, I'm used to heavy planes wanting to sink quite rapidly, but with the significant reduction in drag due to the design, the pylon racer was like a very fast heavy glider, and I actually had to do a power off circle of the field to let it slow down enough to land. I've never had another plane handle similarly power off.
5. My fast jet should be more close-coupled than the pylon racer. It was fine to fly, but I'd rather have something that responds a bit slower in pitch. Fortunately both of my 80mm concepts are already like this since I suspected this might end up being something I would prefer.
6. The launch dolly works very well and I should use it for other similar prototype designs where I don't want landing gear to reduce drag and don't want retracts either due to size or complexity. I performed three takeoffs from it, all of which went flawlessly. This isn't necessarily useful for the remainder of this project, but it will be useful for future projects.

 

[Mr Man]

Ok, this time it didn't self-destruct as soon as I tried a hard turn - that said, I still never reached the top speed despite a successful flight. About 3/4 throttle on 4s is the limit of my flying skill which is 110 mph in more or less level flight and low wind - anything more than that and the plane gets away too fast.

Like with other fast prop planes, it also wants to roll at full throttle until it gains speed which made orientation even more of a challenge. I'm guessing it doesn't have too much more in it on 4s, maybe another 5-10 mph at full throttle. Based on the increase in power at 6s, top speed should be somewhere upwards of 140 mph, but I am not a good enough pilot to do that safely. The 4s top speed matched my prediction from CFD pretty well, so I think my 150 mph level flight estimate is probably about right on 6s.

Unfortunately, on the second flight, the ESC decided to light on fire - it was a 60 A ESC in a plane that draws at most 50 A, but it had been crashed probably 4 or 5 times which was why I was using it in this plane - because it might crash again - and something in it must have reached the point of failure. The plane hit the ground going over 100 mph and obliterated itself. These are the parts that were big enough to pick up - probably about half the plane was in 1/4" confetti after the crash. If you look closely at the ESC, you can see where it caught fire.
View attachment 253789
I don't think I'm going to rebuild this, at least in its current form. It's too fast/small for me to fly safely and it has served its purpose fully to demonstrate the performance of the wing and tail combination.

I'll upload the STLs in this thread in case a better pilot is interested in building it, with the caution that you really need to be a good pilot to handle something this small and fast. It gets away from you quick.

Wow, sounds like it had potential!

 

[Mr Man]

Too bad it crashed though…

 

[telnar1236]

Wow, sounds like it had potential!

Maybe a bit too much potential - I think I might redesign it as a more conventional pylon racer with fixed gear and a modeled in canopy on 3s since the motor is still fine which should still give it a speed of around 80 or 90 mph at some point, but that's a project for some time in the future.

Like I said in my original post, the goal of this thread isn't necessarily to end up with one plane as it is to come up with the design techniques to get this kind of performance and in that respect the prop design was very much a success, even if it didn't last more than a couple flights

 

[Houndpup Rc]

I might have to try this.....

 

[Piotrsko]

One of the problems of self design: sometimes the solutions are counter intuitive. I would think twitchy continues to be a problem, getting worse at higher speeds. All my stuff gets harder to change direction at speed so the pattern gets really huge sometimes, but a thumb twitch doesn't immediately scatter airplane parts across the field or fold a wing joint

Props have a hard limit, cant go faster than about .75 pitch times rpm.

 

[Houndpup Rc]

One of the problems of self design: sometimes the solutions are counter intuitive. I would think twitchy continues to be a problem, getting worse at higher speeds. All my stuff gets harder to change direction at speed so the pattern gets really huge sometimes, but a thumb twitch doesn't immediately scatter airplane parts across the field or fold a wing joint

That's why you have dual rates and expo

 

[telnar1236]

I might have to try this.....

Hope you have more luck with it than I did

Everything should be PLA except the motor mount with a 0.4 mm nozzle. The wings should have 1 wall and 4% infill, the fuselage 2 walls and 5% infill, the tail surfaces 1 wall and 5% infill, the control surface links (elevator, aileron torque tubes and aileron control horns) 4 walls and 80% infill, and the spinner 3 walls and at least 1mm thick top and bottom skins but no infill.

The motor mount should be a more heat-resistant plastic like ABS and should be printed with 3 walls and at least 50% infill.

The aileron and elevator hinges are meant to use a 1.5mm metal pin. The dolly is meant to use any wheels with 3mm hole.

If printing the spinner in PLA, it must be white or natural PLA as anything else may soften in the sun and fail. I used carbon fiber nylon which let me get away with making it black and made it a lot stronger.

 

[Houndpup Rc]

Hope you have more luck with it than I did

Everything should be PLA except the motor mount with a 0.4 mm nozzle. The wings should have 1 wall and 4% infill, the fuselage 2 walls and 5% infill, the tail surfaces 1 wall and 5% infill, the control surface links (elevator, aileron torque tubes and aileron control horns) 4 walls and 80% infill, and the spinner 3 walls and at least 1mm thick top and bottom skins but no infill.

The motor mount should be a more heat-resistant plastic like ABS and should be printed with 3 walls and at least 50% infill.

The aileron and elevator hinges are meant to use a 1.5mm metal pin. The dolly is meant to use any wheels with 3mm hole.

If printing the spinner in PLA, it must be white or natural PLA as anything else may soften in the sun and fail. I used carbon fiber nylon which let me get away with making it black and made it a lot stronger.

Sweet and thanks! Would PTEG work for the motor mount? (I have just heard it smells less than ABS)

 

[telnar1236]

Sweet and thanks! Would PTEG work for the motor mount? (I have just heard it smells less than ABS)

I've never used it, but possibly - it also depends on how hot your motor gets. There isn't a ton of airflow to the motor - just through the two holes in the spinner where the prop forces the air through, so it can get a bit toasty in there on 6s. It would probably be ok since 85 C for petg isn't too much lower than ABS at 105 C but abs is probably better suited.

 

[Houndpup Rc]

I've never used it, but possibly - it also depends on how hot your motor gets. There isn't a ton of airflow to the motor - just through the two holes in the spinner where the prop forces the air through, so it can get a bit toasty in there on 6s. It would probably be ok since 85 C for petg isn't too much lower than ABS at 105 C but abs is probably better suited.

Okay, thanks!

 

[telnar1236]

One of the problems of self design: sometimes the solutions are counter intuitive. I would think twitchy continues to be a problem, getting worse at higher speeds. All my stuff gets harder to change direction at speed so the pattern gets really huge sometimes, but a thumb twitch doesn't immediately scatter airplane parts across the field or fold a wing joint

Props have a hard limit, cant go faster than about .75 pitch times rpm.

That's why you have dual rates and expo

Expo is definitely the only thing making this flyable. I had about 50% on the ailerons and 10% on the elevator.

The challenge wasn't so much that is was twitchy as just that it was fast - I struggled to keep orientation on it with its size and speed. It's actually somewhat stable, especially at around 1/2-3/4 throttle, but just too small and fast for me.

 

[Houndpup Rc]

Expo is definitely the only thing making this flyable. I had about 50% on the ailerons and 10% on the elevator.

The challenge wasn't so much that is was twitchy as just that it was fast - I struggled to keep orientation on it with its size and speed. It's actually somewhat stable, especially at around 1/2-3/4 throttle, but just too small and fast for me.

Yeah, my radjet had 60% on elevator and 70% on the roll!

 

[telnar1236]

I think I've made my decision on the design for the jet - it will the 80mm super saber lookalike. The increased internal volume is worth the increased drag and therefore lower top speed. The other design is more suited to a very fast plane with a lower flight time so I'm going to keep it in my back pocket for future speed competitions or if I decide I really want to go crazy fast.

There are a couple things contributing to this decision. First, the pylon racer design broke through 100 mph relatively easily on 4s and on a 6x6 prop which validates the design of the wing and the streamlining methods I used.
Second, I did end up modeling the Freewing Mirage 2000 in CFD to get a comparison and I was quite surprised by what I found. It has a reputation as one of the faster foam 80mm EDFs out there, but even being as nice to the design as I could (not modeling any of the draggy scale details and not trying to model the cheater inlet) it came out as having 40% more drag than the design and selected, and 74% more drag than the alternative design. Both designs also have much more efficient ducting and do not require cheater inlets and both would have internal control horns and no draggy details further decreasing drag. Therefore, both should be somewhat faster and the Mirage 2000 would meet my speed target, so both of my designs also would. Landing speeds should be similar although the mirage 2000 can high alpha which means the inflight stall speed is lower.

I think my next step is to build a 50mm test version of the jet using the EDF unit from my attempted speed plane, or maybe even a 3s EDF I have sitting around to keep the speed manageable. I'm tempted to just go straight into building the 80mm version, but I think testing a smaller version is probably the better thing to do. This will be focused on stability and stall performance instead of speed. The similarities between my design and the F-100 suggest it might be less than happy at higher angles of attack and I'd rather not lose a larger jet if it turns out that way.

 

[telnar1236]

The design for the 50mm prototype is intentionally very simple, meant to take minimal time in CAD and to be able to be printed and assembled very quickly. It is scaled down to 73% of the scale of the 80mm jet which reduces the area of the inlet to 53% which will provide the right inlet area for a 50mm fan. Instead of using the full set of control surfaces, it will fly solely on elevon mixing on the horizontal stabilizers. The 80mm version is currently planned to use both elevon mixing and ailerons, but if elevons work well on their own, it would be nice to be able to have full span flaps instead. The airframe will mostly be LW-PLA with infill instead of an internal structure to make it easier to model.

It will use a 50mm fan on 3s - with a design this slippery, keeping it slow may end up more of a challenge than making it fast and even the 3s design should be quite quick. While primarily intended to test stability and handling, it will also let me validate my predictions for top speed a bit better than the pylon racer since it should be slow enough for me to actually follow and it uses an EDF which should have a more representative thrust curve. My current prediction is a bit over 100 mph which seems somewhat optimistic for a 3s EDF of this thing's size, so it will be interesting to see how it matches up.

It's maybe about half modeled right now, but with how simple it is meant to be to design and build, it should be done sometime this week or maybe this weekend. One of my goals in this project is to use my 3D printer and simple designs like this and the pylon racer to rapidly iterate on my design so that I don't get too lost in the theory and end up with something that doesn't fly well. And since I'm not buying new electronics for the prototypes and instead basing them around parts I already have, it's very cheap to do too.

Unfortunately, there is a TFR in my area this weekend so the earliest I'll get to fly will probably be sometime next week.

 

[quorneng]

That is a nice design.
My slight concern is the area of the inlet duct.
It is a nice "free flowing" design but it looks like its cross section area will be a bit less than the FSA causing an increase in inlet pressure drop. At high speed there will be some dynamic pressure recovery but at slow speed the EDF will be a bit "choked" and its thrust reduced.
Remember a turbojet is not fan. Its air inlet duct requirements are less than that required for the exhaust duct which is the exact opposite of an EDF.

Note also the F100 Super Sabre layout was optimised for super sonic flight. Any resulting limitation in low speed thrust simply meant it needed the long USAF runways to take off. Its aerodynamic configuration also resulted in it having by far the worst attrition rate for a jet fighter at the time. You may find your plane has some nasty habits!

 

[telnar1236]

That is a nice design.
My slight concern is the area of the inlet duct.
It is a nice "free flowing" design but it looks like its cross section area will be a bit less than the FSA causing an increase in inlet pressure drop. At high speed there will be some dynamic pressure recovery but at slow speed the EDF will be a bit "choked" and its thrust reduced.
Remember a turbojet is not fan. Its air inlet duct requirements are less than that required for the exhaust duct which is the exact opposite of an EDF.

Note also the F100 Super Sabre layout was optimised for super sonic flight. Any resulting limitation in low speed thrust simply meant it needed the long USAF runways to take off. Its aerodynamic configuration also resulted in it having by far the worst attrition rate for a jet fighter at the time. You may find your plane has some nasty habits!

That's the same thing I think looking at the inlet - it looks way too small to me too - but it's actually greater than the entire fan frontal area including the motor on the 50mm version. The fan frontal area is 2116 mm^2

While the inlet area is 2289 mm^2 or 108% of the fan frontal area, and therefore even more than that relative to the FSA.

I think what gets me, and it seems like others as well is that there isn't a real sense of scale for this plane. It's 38" long, so decently big for a 50mm EDF. The duct is also somewhat flatter than on most planes which contributes as well. On the 80mm version, the duct looks even worse, but it's very carefully designed to maintain exactly 101% FSA since I don't want the losses at higher speeds from trying to cram too much air through the fan either.

Trust me, I'm very well aware that EDFs are not turbojets - I've had my fair share of designs where I've needed to compensate for insufficient inlet area with a cheater inlet and I was very careful to avoid that in this design.

 

[telnar1236]

That is a nice design.
My slight concern is the area of the inlet duct.
It is a nice "free flowing" design but it looks like its cross section area will be a bit less than the FSA causing an increase in inlet pressure drop. At high speed there will be some dynamic pressure recovery but at slow speed the EDF will be a bit "choked" and its thrust reduced.
Remember a turbojet is not fan. Its air inlet duct requirements are less than that required for the exhaust duct which is the exact opposite of an EDF.

Note also the F100 Super Sabre layout was optimised for super sonic flight. Any resulting limitation in low speed thrust simply meant it needed the long USAF runways to take off. Its aerodynamic configuration also resulted in it having by far the worst attrition rate for a jet fighter at the time. You may find your plane has some nasty habits!

Those possible nasty habits are what the 50mm version is there to try and iron out - my biggest concern is the swept wing and possibly deep stall/sabre dance tendencies, but the washout at the wingtip and design of the horizontal stabilizer should help reduce that. The hinge angle on the horizontal stabilizer is a bit out of line with the airflow which means that as the stabilizer deflects more, either up or down, the plane should become more directionally stable. I'm used to flying my 70mm F-104, so I can accept some bad habits, but hopefully there shouldn't be too many.

The important thing to remember about this design is that it's not an F-100, no matter how much it resembles one - the wing planform is somewhat different and the tail is quite different. On the fuselage, the inlet design is also different with a 10-degree slope to help it work better at higher angles of attack, and the overall fuselage is a somewhat different shape too. This design evolved out of some of my previous concepts and came to resemble a super sabre almost by accident, though I'll admit I leaned into it at the end when I swept the horizontal and vertical stabilizers. Incidentally, the reason for the 45-degree wing sweep is for printability - at the low subsonic speeds we fly at, it provides very few drag advantages, but the wing root leading edges on my pylon racer didn't print very well and sweeping the wing lets me fix that.

 

[telnar1236]

I thought it might be interesting for people to see how the jet design has progressed over time since I've only posted a couple snapshots of time from the process. Each version is c a number for concept a number.
First was the c1 which was meant to use twin 50mm EDFs and mostly existed to get something on paper and would certainly not have achieved the performance I want. The wing airfoil was a NACA 0012 airfoil.

Then was the c2 which is basically a twin 50mm EDF version of my trainer jet and would also not get close to what I want it to do. In this version, I changed the airfoil to a NACA 2412 for more lift.

Then came the c3-1 which I've already posted a later version of, but this original version still had the NACA 2412 airfoil.

Then the next version was the c3-2 which improved the streamlining and switched to the laminar flow airfoil all following versions of the design use. The wingspan also increased 10%.

Then the C3-3 with the tail moved to the bottom of the vertical stabilizer and both the vertical and horizontal stabilizers switched to the thinner, lower drag NACA 0008 airfoil. The NACA 0008 airfoil is used in the tail surfaces of all following designs. This is the first version that I think has a good chance of being able to meet the various goals I set out in my original post. However, it struggles with internal volume, especially in the wing, so I would have had to have spent a ton of time figuring out better ways to fit everything.

So at the same time as the C3-3 I started on the C4 which switches to a single 80mm fan and has a more streamlined fuselage, but keeps a similar wing design. However, it also struggles with internal volume, this time for the batteries.

So I looked at a design with cheek mounted inlets instead in the form of the C5-1.

This helped, but it was draggier, so started to look at optimizing the wing and ended up with the C5-2. The lower aspect ratio wing is substantially lower drag which helped recover some of the performance I had lost out on with the new fuselage. However, the fuselage was still a bit too draggy and there wasn't that much more space inside. The C5-2 wing however is essentially an unswept version of the wing on the final design.

So I combined the c5 wing with the c4 fuselage to get the c6-1.

And then to try and make it nicer to print I swept the wing 45 degrees to get the c6-2. The performance of the wing should be about the same as the unswept version according to CFD, but it will have less severe overhangs on the section of the wing that is printed as part of the fuselage.

Then again in an effort to try and fix the internal volume problems, I looked at putting the inlet in the nose. The C7-1 had the inlet split around the battery compartment on either side, but the fuselage had to be pretty fat to accommodate that which made it very draggy. This version didn't even make it far enough into the evaluation to have a wing in the first place.

And then finally the decision ended up as one between the c6-2 and the c7-2. The c7-2 changes out the split inlet duct for one going under the battery bay and does a better job blending the canopy into the fuselage. I think the really draggy thing on the c7-1 was the split duct, so doing away with that helped a lot.

I ended up picking the c7-2 since it has much more internal volume than the c6-2 and that volume is more easily accessible and therefore useable. The cost is that it is still draggier than either of the c6 concepts. However, to get the same kind of useable internal volume in one of c6 designs as in the c7-2, the fuselage would need to be scaled up enough that it would have more drag too.