3d printed EDF/BDF experiences and tests

[IcedStorm777]

Ok so, I am going to preface this by saying I know there is a specific discussion for the legitimate BDF's but I really want to try making my own, I will make many mistakes but am going to see if i can run some experiments. So here are the goals:

1. Make a 3d printed EDF that is satisfactory and can produce X amount of thrust with Y amount of run time (I will fill these numbers in once I have done the calculations for goal three.

2. Run experiments with said 3d printed EDF including things like compression, nozzles, and exhaust shape (De laval, aerospike, etc.)

3. Put these EDF's in RC versions of some of these: F9F Swept Wing Cougar, F-11 Tiger, F-5 Freedom Fighters, and MB 339.

4. Have a fun time, don't die trying, and post findings.

So a few questions right off the bat. Would it be beneficial to have a sort of compressor stage? Here is what i mean by this. With 3d printing i can easily make complex geometry shapes. The beginning of the fan tube (these will all be done in tubes so i can test different diameters and geometries. So the beginning of the nozzle would be a fairly average size and there would be a standard EDF Fan blade. Now on the same motor shaft that is driving the EDF Fan blade there would be, say, four compression blades. Immediately following the standard EDF Fan blades there would be four compression blades and as the compression blades go further back, the tube would compress. The nozzle would than open back up similar to a De Laval nozzle. My theory is as such: The fan will suck the air past it but as the air goes through the compression blades and as the tube compresses the air will be forced to move much faster to escape with the same pressure which would greatly increase thrust to energy power. The reason that the nozzle will not be super constricted to lower the spool time. Any thoughts are appreciated, I will post pictures and progress over the weekend if i find time, i am very busy this week so we will see how this works out. Thanks

 

[leaded50]

De Laval nozzle is based on compressed heated gas, who would expand .... air as in a edf wouldnt expand, its the push of the rotorblades who make the "thrustpower", is will "loose" air velocity when out of the tube.
Reason for thrusttubes is a bit smaller in end (5-7%), is to keep consistent, or a bit faster velocity of the air pushed. Where motor is, its a space where air isnt pushed. It has been shown though, this increase in thrusttube isnt as affected by eg a 12 blade vs a 5 blade. (the 5 blade would need it more for best effect)
To restricted in-air, or too small thrusttube opening would decrease thrust of the EDF.
Compression of the air through with more fans inside could function, but it will be needed a lot of testing of diferent fans/ size/ to find correct values to raise effectiveness. And will need a special powersystem A turboprop eg, have a combustion, gasses, heat to raise the power.. that we will not have by a fan drawing/blowing air.

 

[The Hangar]

Ok so, I am going to preface this by saying I know there is a specific discussion for the legitimate BDF's but I really want to try making my own, I will make many mistakes but am going to see if i can run some experiments. So here are the goals:

1. Make a 3d printed EDF that is satisfactory and can produce X amount of thrust with Y amount of run time (I will fill these numbers in once I have done the calculations for goal three.

2. Run experiments with said 3d printed EDF including things like compression, nozzles, and exhaust shape (De laval, aerospike, etc.)

3. Put these EDF's in RC versions of some of these: F9F Swept Wing Cougar, F-11 Tiger, F-5 Freedom Fighters, and MB 339.

4. Have a fun time, don't die trying, and post findings.

So a few questions right off the bat. Would it be beneficial to have a sort of compressor stage? Here is what i mean by this. With 3d printing i can easily make complex geometry shapes. The beginning of the fan tube (these will all be done in tubes so i can test different diameters and geometries. So the beginning of the nozzle would be a fairly average size and there would be a standard EDF Fan blade. Now on the same motor shaft that is driving the EDF Fan blade there would be, say, four compression blades. Immediately following the standard EDF Fan blades there would be four compression blades and as the compression blades go further back, the tube would compress. The nozzle would than open back up similar to a De Laval nozzle. My theory is as such: The fan will suck the air past it but as the air goes through the compression blades and as the tube compresses the air will be forced to move much faster to escape with the same pressure which would greatly increase thrust to energy power. The reason that the nozzle will not be super constricted to lower the spool time. Any thoughts are appreciated, I will post pictures and progress over the weekend if i find time, i am very busy this week so we will see how this works out. Thanks

I'll be following. @chris398mx has done a bit of work with BDF's and might be able to give some advice.

 

[quorneng]

IcedStorm777
Nothing wrong with experimenting but remember EDFs have been around for quite a while so the chances are slim that any major improvement has been missed.
With an EDF or any cold fan for than matter it is all about the mass flow. Pressure, and the more the better, is only significant if you move into a combustion cycle.
With an EDF there are two conflicting goals.
The maximum thrust for a particular diameter, usually at the expense of efficiency.
The highest thrust for a given input power, usually expressed as g/W.
Given that most planes are modelled on full size aircraft that use 'combustion' jets engines it tends to favour the maximum thrust type EDF with relatively short flight durations.
It is noticeable that Airliners with large turbo fans, where most of the thrust is created from a cold fan, can work well with EDFs in scale ducts.

I will follow your experiments with interest.

 

[IcedStorm777]

IcedStorm777
Nothing wrong with experimenting but remember EDFs have been around for quite a while so the chances are slim that any major improvement has been missed.
With an EDF or any cold fan for than matter it is all about the mass flow. Pressure, and the more the better, is only significant if you move into a combustion cycle.
With an EDF there are two conflicting goals.
The maximum thrust for a particular diameter, usually at the expense of efficiency.
The highest thrust for a given input power, usually expressed as g/W.
Given that most planes are modelled on full size aircraft that use 'combustion' jets engines it tends to favour the maximum thrust type EDF with relatively short flight durations.
It is noticeable that Airliners with large turbo fans, where most of the thrust is created from a cold fan, can work well with EDFs in scale ducts.

I will follow your experiments with interest.

I appreciate all the feedback. I totally recognize that I (a simple minded 16 year old won't be making any major improvements but figure it could be a fun project. I am also planning on using a friends resin printer for the blades so they will be MUCH smoother. I am generally doing this project more for fun but also want to test a few things as stated. Another thing I am interested in is what different airflow and compression configurations could do to the sound. Thanks, hoping to do some work on it over the weekend.

 

[telnar1236]

Looking over some of the replies I figured I could weigh in on some slightly more technical things. Overall, I don't like giving super-technical responses about RC planes, but this is a pretty technical subject matter that requires a decent grasp of some fluid dynamics.

Nozzles: The first thing to understand is Bernoulli's principle and conservation of energy. In subsonic flows, when a duct gets narrower, flow velocity increases and pressure decreases. This is called a nozzle. Similarly, when a duct gets wider, flow velocity decreases and pressure increases. This is called a diffuser. However, at supersonic speeds, flow behaves differently, and an expanding cross-section behaves as a nozzle and a contracting cross-section behaves as a diffuser. For an EDF, you want a subsonic nozzle that achieves a good velocity without getting too narrow and significantly increasing losses. De Laval Nozzles and aerospike nozzles are both supersonic nozzles (technically a combination of subsonic nozzle and supersonic nozzle) and would not function with a normal EDF. It is worth noting that, while supersonic nozzles typically function with hot gasses and while they do exchange heat for kinetic energy, their design is based on Mach number. A de Laval nozzle would work just fine with a room temperature gas if that gas was at a high enough pressure for the flow to choke across the throat, but achieving that pressure without a turbine or rocket is difficult. For an EDF unit, there is less design work possible with a subsonic nozzle, but that does not mean there is none. Look up motorjets if you are interested in some of what has been done with full-scale aircraft. With no real data, just spit-balling, I would guess that you could boost thrust by maybe 5%-10% which is not nothing.

Compressor stages: Additional stages in series will increase thrust, but at the cost of efficiency. You are better off spinning a single stage faster, increasing the diameter, or running multiple EDFs in parallel. In a real jet engine, they are necessary because the engine generates energy internally by burning fuel. Compressing the air first significantly increases efficiency and power output. To extract energy, the exhaust gases power a turbine that drives the compressor, with the remaining energy being used for thrust. In an EDF unit, where the power is supplied externally, increasing the internal pressure won't be as useful, and it is more efficient to generate the thrust in a single stage. Adding multiple stages adds significant complexity too. The only real exception to this would be if you really need a high efflux speed, but again, at the the speeds we are flying, that is not a real consideration. Also consider that to increase pressure, you need to spin each stage faster than the previous stage or increase the blade pitch. Thus simply stacking sets of identical blades will simply lose efficiency with little benefit.

Materials: This is probably the most important. EDF blades spin very fast and experience significant loads. Even professionally made EDFs can fail violently. Be aware of the limitations of 3D printing and stand far back from any EDF units you are testing and out of the plane of the blades. If you really want high performance, try making 3D printed molds for a unit with molded fiber glass blades. It's pretty easy to do, and this is a DIY version of how some of the best professional EDF units are made.

I should note, that with enough power, your initial concept of a multi-stage electric compressor followed by a de Laval nozzle would work. It's just very far beyond what any available power systems can come close to achieving and there are more efficient solutions. If you really want to go crazy this is a link to an article about a true electric jet being developed in China. Plasma Jets May One Day Propel Aircraft - IEEE Spectrum

I look forward to seeing what you design. The F11F and F-5 are some of my personal favorites and I would love to see models of both fly.

 

[IcedStorm777]

Looking over some of the replies I figured I could weigh in on some slightly more technical things. Overall, I don't like giving super-technical responses about RC planes, but this is a pretty technical subject matter that requires a decent grasp of some fluid dynamics.

Nozzles: The first thing to understand is Bernoulli's principle and conservation of energy. In subsonic flows, when a duct gets narrower, flow velocity increases and pressure decreases. This is called a nozzle. Similarly, when a duct gets wider, flow velocity decreases and pressure increases. This is called a diffuser. However, at supersonic speeds, flow behaves differently, and an expanding cross-section behaves as a nozzle and a contracting cross-section behaves as a diffuser. For an EDF, you want a subsonic nozzle that achieves a good velocity without getting too narrow and significantly increasing losses. De Laval Nozzles and aerospike nozzles are both supersonic nozzles (technically a combination of subsonic nozzle and supersonic nozzle) and would not function with a normal EDF. It is worth noting that, while supersonic nozzles typically function with hot gasses and while they do exchange heat for kinetic energy, their design is based on Mach number. A de Laval nozzle would work just fine with a room temperature gas if that gas was at a high enough pressure for the flow to choke across the throat, but achieving that pressure without a turbine or rocket is difficult. For an EDF unit, there is less design work possible with a subsonic nozzle, but that does not mean there is none. Look up prop jets if you are interested in some of what has been done with full-scale aircraft. With no real data, just spit-balling, I would guess that you could boost thrust by maybe 5%-10% which is not nothing.

Compressor stages: Additional stages in series will increase thrust, but at the cost of efficiency. You are better off spinning a single stage faster, increasing the diameter, or running multiple EDFs in parallel. In a real jet engine, they are necessary because the engine generates energy internally by burning fuel. Compressing the air first significantly increases efficiency and power output. To extract energy, the exhaust gases power a turbine that drives the compressor, with the remaining energy being used for thrust. In an EDF unit, where the power is supplied externally, increasing the internal pressure won't be as useful, and it is more efficient to generate the thrust in a single stage. Adding multiple stages adds significant complexity too. The only real exception to this would be if you really need a high efflux speed, but again, at the the speeds we are flying, that is not a real consideration. Also consider that to increase pressure, you need to spin each stage faster than the previous stage or increase the blade pitch. Thus simply stacking sets of identical blades will simply lose efficiency with little benefit.

Materials: This is probably the most important. EDF blades spin very fast and experience significant loads. Even professionally made EDFs can fail violently. Be aware of the limitations of 3D printing and stand far back from any EDF units you are testing and out of the plane of the blades. If you really want high performance, try making 3D printed molds for a unit with molded fiber glass blades. It's pretty easy to do, and this is a DIY version of how some of the best professional EDF units are made.

I should note, that with enough power, your initial concept of a multi-stage electric compressor followed by a de Laval nozzle would work. It's just very far beyond what any available power systems can come close to achieving and there are more efficient solutions. If you really want to go crazy this is a link to an article about a true electric jet being developed in China. Plasma Jets May One Day Propel Aircraft - IEEE Spectrum

I look forward to seeing what you design. The F11F and F-5 are some of my personal favorites and I would love to see models of both fly.

Wow, that was awesome!! I got up at 5:00 in the morning and that comment literally made my day. I love this whole field and that is a lot of good info i need to adjust for and use. I am currently very focused on wrestling and another project so until season is over i probably won't do much work on this. Thanks!

 

[quorneng]

You can certainly increase the static thrust using a nozzle.
I did some simple experiments using various sized 3D printed aerodynamic bodies to reduce the nozzle area from100% to 80% FSA all in the same 76 mm duct.

The 100% FSA body. In this case the body creates an annulus nozzle.

This was at very low power for an EDF so I was not really expecting any measurable difference, however, low and behold, the highest thrust was indeed achieved at the recommended 85% FSA. It was only a 5% increase but it was still for 'free' so it was used.

 

[LitterBug]

Octo-UFO from pre-BDF dates. I've thought about scaling it up. 1.9" props made it not worth the effort to do much that might add extra weight.