How to Make a 3D Printing Coreless DC motor (Brushless)

Method
I came up with the idea of making this project after I finished my graduation project and dissertation. As graduation approached, I didn’t have enough time to make a large scale project, so I used readily available materials to make the motor.
Rationale
A 3D printer can print in many materials, but what can it not print?
Magnets, copper, metal bearings, iron cores and carbon brushes
What are the essential components of the motor?
A motor is made up of a magnet, copper windings, bearings, and a shaft. The iron core is not necessary, just like coreless motor has no iron core.
Let’s make a coreless DC motor without a brush.

You might find that there are many motors that do not have magnets
There are four possibilities for motors that do not have magnets:
1.Use energizing winding to replace magnets: energizing winding needs power supply from electric brush, not even to say that its relatively large size poses obstacles for the 3D printing machine to print out its shell. 
2.The rotor of the asynchronous motor is composed of a bunch of coils or  aluminum frames. This will enlarge the total volume of the motor and there is no available driver suited to the asynchronous motor. 
3.The rotor of the switched reluctance motor is an iron. Although the composition is simple but how can I make an iron?
4.For motors made in other principles: ultrasonic motor (piezoelectric effect), electric power motor (electric field) and etc. I cannot make them without driver and materials.

Mechanical design:
Integral assembly drawing (with no drawing coils)
 
This is the integral explosive view. The brown color is for the top and bottom caps, the white one is for the bearings, the black one is the shaft, rotor is highlighted in blue and the green one is the stator shell. There are 36 grooves on the stator shell for installing coils.

There surfaces on the rotor are magnetized. Hence, the two biggest surfaces are the magnetic poles: One is N pole and the other is S pole. There is no need to know which one is N pole and S pole during installation. It should be fine as long as you know the adjacent poles are the opposite ones and you make the installation in the NSNS order. Originally, every magnetic pole was made of three magnets but later on it was changed into two magnets for greater efficiency. 

Assembly:
Material list: 3D printed rotor, 3D printed shell, 3D printed top and bottom caps, two bearings, one optic axis, eight NdFeB magnets (used in pairs as there is no suitable thickness), as well as enameled copper wires (not pictured, any thickness will do)

For better torque output, the optic axis and rotor should be fixed tightly. Hence, you need to force the optic shaft in to the rotor with hammer. After this, the two sides of the rotor will be somehow twisted and cannot be fit properly in the bearing, you’d better polish it a little bit. 

The bearing seats of the top and bottom caps are designed as slightly large so that the rotor can be put easily into them.
Try first to see if the rotor can be put into the stator.
To prevent parts flying out from the rotor because of centrifugal force, you need to wrap up the rotor with adhesive tape. However, do not use too much adhesive tape or you will prevent the motor from spinning freely.
Wind the coils consistently using a certain rule: My 36 groove coils are winded in this order: A1+?A2+?A3+?B1+?B2+?B3+?C1+?C2+?C3+?A1-?A2-?A3-?B1-?B2-?B3-?C1-?C2-?C3-?A4+?A5+?A6+?B4+?B5+?B6+?C4+?C5+?C6+?A4-?A5-?A6-?B4-?B5-?B6-?C4-?C5-?C6-?

Wind coil A1+ from the prominent teeth on the top of the stator shell, through the stator’s inner shell, and wind the teeth at the stator shell’s bottom. Then wind along the circular arc grooves at the bottom of the stator shell to A1- teeth. Wind the A1-teeth through the inner shell of the stator to the teeth from the stator shell’s upper end. Wind along the circular arc groove on the top back to the A1+teeth. Repeat this by winding the other four coils (The space for winding in my design is only allowed for four coils...you can wind more rounds if you still have enough space but to make sure all grooves have equivalent coils). Until now, the winding in the two grooves are complete. Repeat this process in the other grooves before the winding is finished. 

The rest is wiring. Connect A1- to A2+?B1- to B2+?C1- to C2+?A2- to A3+?B2- to B3+?C2- to C3+?A3- to A4+?B3- to B4+?C3- to C4+?A4- to A5+?B4- to B5+?C4- to C5+?A5- to A6+?B5- to B6+?C5- to C6+, it’s just like each phase coil is connected with six coils. Then connect A6-?B6-?C6- together?which turns into a three phase coils like the star coupling method. Then you can use A1+?B1+, C1+ as the three-phase input terminal of the motor. 

Q1: Why wind like this?
There are altogether 36 grooves in the stator, three-phase coils, with each phase having 12 grooves. As there are four magnetic poles of the rotor, if the rotor turns 90 degree angle, the external magnetic field from the coil in the same position with the stator will change from N pole into S pole, or change from S pole to N pole. Therefore, A1+ and A1- at two ends of the A1 coil?the flow direction of A1+ and A1- coils are always the opposite, as current flows from the top to the bottom on one side , it will flow from bottom to the top from the other side). In order to get Ampere force from the same direction under the left-hand rule, you need to have 90 degree interval in space, which means that there should be 9 grooves in between the two sides. This will guarantee that every time when A1+faces one magnetic pole (such as N pole), A1- just perfectly faces the other magnetic pole (S pole).

The other coils should follow the same rule, which means that there should be a nine-groove interval. 

If the rotor I designed only has two poles, the interval between coils should be 18 grooves. 

Q2: Why wire in this way?
You may use the a UAV brushless ESC (electronic speed control) during the test. Normal ESC controls phases by changing the counter electromotive force. If the counter electromotive force is not high enough, and if the controller cannot identify the phase-change time, you may not be able to change phase. Therefore I have used a series of coils to improve the counter electromotive force. 

The reason to use the star-coupling method was because such practice is usually applied onto low power motor. One thing for sure is the star-coupling method (A1+ connects B6-?B1+ connects C6-?C1+ connects A6-?can also rotate. The star-coupling method or the Y-connection method will change the voltage current of the motor winding and further change the motor’s performance. 
 After installing the bottom cap, insert the rotor to see if it can touch the copper in the inner shell of the stator. If there is any contact, you should try as much as possible to flatten the cooper wire or you may need to change the rotor design: Make the rotor smaller and increase the air gap between the stator and the rotor.  
You’d better install the top cap well.
3D printing coreless DC motor without brush is finished.
 Let’s have a test by taking a brushless ESC You’d better choose the ESC which allows for relatively higher voltage to rotate itself. 

Test the equipment: DC regulated power supply (You can also use battery but it can not regulate voltage), an electronic speed controller (no brush), remote control and receiver (for the convenience of regulating the PWM of the input ESC without brush. You can also use varieties such as steering gear tester and Arduino), a laser reflection tacheometer (You need to attach a piece of paper slip on the motor’s rotation shaft. When the motor rotates, the paper slip will constantly reflect the laser from the tachometer so that you can test the rotating speed)

Result: 
If you purchase all the essential materials, you need one day for the 3D printing, and one more day for the winding, assembling, and debugging. 

Under idling condition, the maximum speed can reach as fast as 2800rpm. Loud noise can be caused by loose winding. This is due to the fact that the copper will vibrate fiercely once the rotor starts rotating. As the friction force of the bearing in use is rather big, you can think of using bearing with smaller friction force to increase the rotation speed. 

As the air gap between the stator and the rotor is big and there is no iron core, the leakage inductance will be relatively big. Besides, as winding is not dense enough, the power and efficiency will not be very high as well. However, as an interesting project, I think it has realized all its purposes. 

The existing problem: Concentricity of the object that printed out in 1.3D is not very high. Therefore, the air gap between the stator and the rotor should be big enough, which may increase leakage inductance and enlarge vibration. 2. Wind with hand becomes loose easily, which may cause frictions between rotor and stator. 

Finally: Magnet with stronger power may not be the best choice. If the magnetic power is too strong, the back-EMF will be too high if the motor is in rotation speed, which may prevent the winding from injecting enough current so that the motor cannot rotate normally. My initial idea was to use three magnets for each pole but the motor was failed to rotate well. I think two magnets are enough. 

I was intended to make more calculations but I could not find any data about the magnetic field strength of a magnet. I also wanted to measure the output mechanical power but I didn’t because there was no dynamo-meter. 

To be honest, as long as you understand the working principle of the motor, you can make the motor in operation more easily. 
I believe in the future, the breakthroughs in 3D printing technology and materials will enable us to make personalized motor that we just need more easily, such as 3D printing high performance magnets, high-precision metal 3D printing machine to print out complicated silicon steel sheet, winding, yoke and etc. Admittedly, there is still a long way to go for the 3D printing motor to put into practical use.