• Air-core Brushless motors:

But before going to the voice coil actuators, let me at least give a try to aircore axial brushless motors with the flat coil type.

Like this.

Plus a good reason:

“Different materials have different saturation levels. For example, high permeability iron alloys used in transformers reach magnetic saturation at 1.6–2.2 teslas (T), whereas ferrites saturate at 0.2–0.5 T. Some amorphous alloys saturate at 1.2–1.3 T. Mu-metal saturates at around 0.8 T.”

“Iron is desirable to make magnetic cores, as it can withstand high levels of magnetic field without saturating (up to 2.16 teslas at ambient temperature.) Annealed iron is used because, unlike "hard" iron, it has low coercivity and so does not remain magnetized when the field is removed, which is often important in applications where the magnetic field is required to be repeatedly switched.”

Source is wikipedia.

Iron is 1.7 times denser than ferrite, so you would face the same problem with weight.

Magnetic permeability of iron powder and resin per weight:

A rough guideline based on available data is:

1. 10% iron powder by weight: Permeability ~1 (almost no magnetic response).
2. 30% iron powder by weight: Permeability ~2–5.
3. 50% iron powder by weight: Permeability ~10–20.
4. 70% iron powder by weight: Permeability ~30–50.
5. 90% iron powder by weight: Permeability ~50–100+.

My brother in christ, I had to make this stupid thing by hand.

Every type of modifier simply doesn’t work for some reason, even this thing just… Keeps changing diameter and dimensions even though I didn’t do anything for that to happen.

Anyway, right now I’m so done with Blender that I simply don’t wanna count the amount of turns and blah blah blah.

In the image there are 18 vertical layers and each layer has 2324mm³ (according to blender), all of them together weigh around 0.112 kilograms. 0.112 kilograms x 3 slots x 2 stators = 0.6777 kilograms. So you need to add a little more layers, so the final thickness of each slot is around 2cm. You can reduce it a little bit if you wind up each layer in the hexagon style (either vertically or horizontally).

I would suggest you add cooling channels because I don’t think the flat coil wound motor would have enough surface area, but to do that I would need to 3D model this thing again with the cooling channels in mind.

And boi, I won’t do that.

If I decide to stick with the aircore, I will just add some sticks/cylinders in the mold. Another idea would be to add paraffin filaments through the winding, so when the resin cures I can just melt it away and leave really tight packed cooling channels.

I just woke up and I feel way calmer, and I thought of a way that could help with the aircore motor.

You see, in a linear electromagnetic motor, or a voice coil on itself, you could make them work even without the magnetic cores, because the electromagnetic field starts from the wires themselves.

The way I thought of dealing with the lack of magnetic core to focus the electromagnetic fields was to increase the surface area of the coils, allowing for every wire of both coils to interact with each other. Reducing the distances as much as possible. Since the density of the magnetic flux is way lower, maybe the aircore would benefit from a back iron to reflect back the electromagnetic field. I think a practical way of doing that would be to make several thin stators or make a wobbly U shaped stator with as much surface area as possible.

This time I won’t be making the coils by hand, I will just compare the density of a section of wires to a solid piece of mesh with the same dimensions.

The 60 wires (20% increase over 50 wires for the 300kw motor) weigh around 1.2528 grams and the solid cube in the right weighs 1.62 grams, so the wires are 1.29 times lighter.

… Couldn’t I have done that since the beginning and saved up time…?

First, the easier version:

Parallel stators with a single 1mm wire, so every stator has 1mm of thickness.

This time I just copy pasted the arc vertices and scaled down, I didn’t make a spiral or anything like that. Also, may god have mercy on whoever tries to wind a 1mm thick flat coil in the shape of an arc.

Wait… I think there is something wrong with the motors…

Now that I went to make the flat coil motor… These 3 arc coil have 1mm of thickness each and 0.111 grams of weight (even after the 1.3 reduction).

0.111 grams x 3 slots x 2 stators = 0.666 kilograms.

That is the exact weight of the 100 kilowatt motor before the increase in 20% of weight. 😐

The literal motor that I’ve made that is supposed to have 20 parallel wires and weigh 2 kilograms in total weighs as much as a full motor that has these 3 stators with 1mm thickness and a single wire with also 1mm thickness.

If you take the weight and calculate:

• 1 slot = 0.100 kilograms
• 1 slot x  3 slots in total x 50 stators (for every parallel wire) x 2 for the moving stator = 30 kilograms.

So all the work I’ve done here so far was useless?

… Or is Blender just messing up with me?

I think I get it.

I didn’t count the amount of turns in the material, initially I said it would be 12 turns with 50 wires for every 99 slots, but I reduced it by ⅓. So it would be 4 turns per 50 wires, right?

0.111 kilograms per slot / 20 wires (100 kilowatt motor with 20% extra wires) = 0.00222 kilograms.

And look at that, it actually is 0.002 kilograms and it is around 3 turns of wire. I guess I got concerned for nothing.

If I have that much space, then wouldn’t it mean that I could fit multiple coils in a single stator?

Well, I won’t be making whatever many spirals on the damn thing to find out, but I can take the total weight of all wires in the model and divide it by the weight of the intended target.

So, its total weight is 0.020817 kilograms, divided by 0.0022 kg = 9.37702702703

So every slot allows for around 9 wires each, which means I would need what? 3 stators for 23 wires in total.

So, 3 stator disks for 267 amps x 2 moving stators = 14 stators in total

Every stator has 1mm of thickness, with 1mm of space between them = 20.5mm of thickness. 2 centimeters.

For the 300 kilowatt motor, 3 times as much stators = 42 in total. 62mm of thickness

By the way, since the cables have to go out axially, it would be necessary to literally hammer them until they flatten to avoid obstruction in the air gap.

… I feel like these motors really need a protective shell. I thought of leaving to 3D model it in the structure section, but I don’t know…

• Well, NOW I get to actually 3D model the wobbly motor:

Am I even doing this right? Like I said in the energy generation section, a generator with the same power output (370 kw) weighs 175 kilograms. And that thing was meant to work at 60,000 RPM. The lower the rpm and the bigger the torque, the higher the weight. So I’m actually right? At all?

Wouldn't it be better to stick with hydraulics?

I have no f*cking idea what I’m doing, so in case all of this doesn’t work, I should at least attempt at 3D modeling the hydraulic actuators, valves etc…

But before that, let’s remember the thingap air-core brushless motors:

This 100 kilowatt is said to have 385 mm of diameter, 223 mm of width and only 8mm of thickness. If you input the numbers on volume calculators, this motor would weigh around 20 kilograms.

Assuming it is made out of solid copper and assuming it is only the weight of the copper wires, not the bearings, metal armature etc. Using the 1.3 reduction in weight, it would be 15kg.

If it was made out of aluminum, it would weigh around 4.5 kilograms.

(they also make “U-shape” rotors with double magnets)

Although I feel a little bit more confident on the weight of the brushless motors I’ve designed, but a little more skeptical at the same time.

After all, supposedly, my 300 kilowatt motor would weigh 5 kilograms, while their 100 kilowatt motor would weigh the same.

Assuming that the 22 actuators in the upper body (5 actuators in the shoulder + 1 in the elbow + 3 actuators in the wrist x 2 arms + 4 actuators in the torso = 22) weighs 5 kilograms each, it would weigh in total 110 kilograms in total. Assuming that the 14 actuators in the lower body (3 in the hip + 1 in the knee + 3 in the ankles x 2 legs = 14) weigh 15 kilograms each, it would weigh 210 kilograms in total.

I will need to actually build the motors to check if they will actually output 300 kilowatts with only 3 kilograms.

• Anyway, actually, finally, definitely 3D modeling the wobbly motor:

In blender you can use the “Array” modifier, select the option “object offset” and then scale the object, this will allow a sequence of materials that will increasingly get bigger than the previous one.

These two plates have more or less the same volume as the aluminum coils in both the 300 kilowatt motors (left) and 100 kilowatt motors (right).

I just reduced them to a single plane in this shape, so the Blender can show me the surface area. Then it is just a matter of finding a shape that can achieve the same surface area.

Now I will increase their surface area.

I used a “Path” curve on blender and then applied the “Screw” modifier without turning it into a spiral.

This is just for the 100 kilowatt motor, by the way. I think that I should make a wobbly shape in both directions.

Well, the 100 kilowatt was pretty tame… But the 300 kilowatt one was…

This thing looks like a turd.

Well, I will try a proper U shape or even a donut shape.

I just used the basic shape “Torus” on blender and added a “subdivision surface” modifier

First try, I didn't even need to modify it that much.

Although, the 300 kilowatts has 25cm of width and the 100 kilowatts has just 6.1cm.

… Now that I think about it, I need to change it. It doesn't have any space for a rotor and bearings…

Well, you would need to divide the stator/rotor in two and screw it in the middle. Or else I don’t see how it could be done…

But now that we are in the subject, the “U”/Torus shaped motor can be quite a pain to align (if not impossible).

So I just added a few radial options besides the axial ones, but they ended up too large (15cm of thickness for 100 kilowatt alone), so I still prefer the axial flux 1mm thick stators.

I didn't feel like I did a good enough job, so I tried again to model the wobbly motor, but it ended up being a trapezoid or something like that.

Since this one has straight lines, I do think it will be easier to align, and I think I will keep with this design.

Due to some shenanigans it ended up with 298 mm of diameter and 87.1 mm of thickness for the 300 kilowatt motor and 298 mm with 29.7mm of thickness for the 100 kilowatt one.

I changed the 100 kilowatt motor one, it is just the 300kw with less vertices, this way you can make either using the same mold.

You know what? I think I will use this design instead.

I will add the back iron and cooling channels for it.

My brother in christ, I took all day to make this stupid cooling channel in a spiral. I used the “curve = toroid” on Blender, then added the “shrinkwrap” modifier and then I deleted some parts of the toroid (like half of its inner laterals”.

They are supposed to have 5mm, but like always, blender keeps modifying their sizes for whatever reason.

Also, for some reason I feel stupid, idiotic, like a clown, like an imbecile, a laughing stock.

How could this possibly ever work?

How could I possibly ever finish it?

I took all day long to make this goddang toroidal motor, but in the end I can only cringe at the hell on earth that it will be to wind this thing by hand.

I will just keep going with the conventional radial thingap motor, this axial motor be damned.

• Thingap Aircore actuators:

In the matter of compactness, I really dislike this motor design, but in the ease of building it, this is the best way I could think of.

You can make concentric motors like I showed before, but again, it will be a pain to make them.

Unfortunately, I wrote the cooling part before this part. But in essence, using nitrogen gas (or just plain ambient air), you could reduce the temperature of the motors to -70 or even higher if you use materials that can survive cryogenic temperatures.

I said that using liquid coolant at -70ºC would be dangerous due to frostbite, but working with air is way safer, lighter and easier. You could go down to cryogenic temperatures, but not every material can survive these temperatures. On top of that, if you get too cold, air/nitrogen will become liquid and it will be dangerous to work with them.

Thus I can reduce the weight/size of motors by 40%, if you don’t want to work with a -70ºC air cooler, you can just increase the size of everything by 40>#/b###.

The thingap aluminum motor with 100 kilowatt of power will have 90mm of thickness and 300mm of diameter, the 300 kilowatt one will have the same diameter, but have 270mm of thickness.

Of course, this is the thickness of the coils themselves, the supporting structure may be thicker.

The outer ring/cylinder has an inner diameter of 295 mm and the inner ring/cylinder has an outer diameter of 293.

I added an inner cylinder with 1 mm of thickness to show the space where copper wire will be the outer thick cylinder (where the cooling channels are) will be the back iron.

I will also copy the structure of wheel rim’s for the structural support of the brushless motors.

Dunno if this will be relevant to say or not, but the motors are two cylinders, one smaller than the other. But if I had both with the same thickness of material, the outer cylinder would have almost 3 to 5 times more material, and thus, 3-5 times more strength. So I’ve made each cylinder disks/rotors/stators with really different thickness in an attempt to compensate for that difference.

If it will have a significant impact on performance I don’t know.

…But that is not necessarily applicable to the arms, the arms will work like a longer lever.

I’ve made this design for the structure to attach to the motors, but after stopping to think about it, how I’m going to make a composite with such shape?

Every time it is like this, I make a design then I have a moment of clarity and I discard the design…

I think this is better, the profile of the rim is in there and you can modify it, but I just added the option anyway.

Now that I think about it again, what is the point of making the rim if I don’t even know what the bearings will look like? What if its dimensions are completely different from this idea?

… I will just make the cooling channels and make the rims at the bearing section…

There

… Or not. Is this even a good enough pattern for cooling channels?

I will make the structural holes and cooling holes like I’ve done in the first model of the motor.

Left a space of 3mm between the two rotor/stators, but it is because it is to add the aluminum coils.

However, I still don’t feel very confident with simply gluing the coils to the rotor/stators. I thought of using some kind of plastic fibers, like embroidery of some kind. But that would be very labor intensive…

Unfortunately, the outer diameter now is 330mm.

Now a new problem that I forgot to take into consideration: weight of the armature.

If you take the density of epoxy resin + 20% per weight of iron powder (2.5 g/cm³) and the volume of the armature of the motors, it will weigh around 8 kilograms for the 100 kilowatt motor alone.

Meaning I will have to take weight out again.

By the way, I’m not using blender's built-in volume calculator, because it doesn’t really work.

… Now I recalculated the surface area for the axial stator and I need just half of it, but even with half of the thickness, it is still 4 kilograms in total, which would be 12kg for the 300 kilowatt version.

Okay, I’m recalculating everything from scratch.

Assuming that the 100 kilowatt has a weight of around 0.6 kilograms, it would have a volume of 222 cm³, a rectangle with 17mm (17 wires with 1mm) of width, 1mm of height and 13,059mm of length would have the same volume and a surface area of a rectangle with 2220cm² of area and 1mm of thickness. A cylinder with the same surface area would have around 17mm of height and 300mm of diameter.

Fixing the values, now I have… 10 kilograms for both stator-rotors…

Now I reduced the thickness of both stator-rotors, the inner one has 55mm of thickness and the outer one has 10mm and now they weigh 1.5kg each, totalling 3kg. 3kg x 18 actuators on the upper body = 54 3kg x 3 times more output = 9 kg for structure + 1.4 kg for the coils = 10.4kg x 14 actuators for lower body = 145.6 145.6 kg + 54kg = 199.6 kg in total.

I mean, I do intend on using natural fibers for the composite and they normally have a similar density to pure resin (1.3 to 1.5 g/cm³), and since the resin will be 20% iron powder, it would be lighter.

The only way to reduce it even further would be to leave only the aluminum coils, which won’t be viable.

That or adding some strong foam, the only type that comes to my mind would be metal foams and/or aerogels. Which I don’t think are viable to make.

Maybe the aluminum foam could work, because they are normally made of hollow steel spheres in an aluminum matrix. But you wouldn’t get much benefit from the magnetic permeability… I think? You could envelop the spheres with sodium silicate and iron powder….

… Or just use thermal insulation (like ceramic fibers) as spheres with iron powder as the filling material in the aluminum metal (sodium silicate has a density of 2.3 g/cm³, so it won’t be able to have too much iron powder to increase its density to 2.5 g/cm³).

There are also metal sponges/wools that are actually pretty cheap to buy, you could also use them.

You could impregnate these sponges/foams (and even crumpled aluminum/iron/nickel foil instead of doing all this work) with resins, PU foams etc.

I listed earlier a video from tech ingredients and how a simple layer of fiberglass composite could double to triple the strength of a foam, but since I’m going to work with metal foam… Still, it is up to you.

By the way, I used the option “make manifold” and “distorted” in the “clean up” section of the 3d-print add-on that calculates the volume of meshes. And it finally calculated the volume correctly.

I wasted too much time on this thing, and it still didn’t reduce the volume enough.

… I should’ve done a similar approach to the change in volume the holes have so I can find a sweet spot.

This one was just 3 times lighter.

Also, I’m trying to make it in a way that it will be easier to make a 3D print and/or a mold.

If I actually made a 3D lattice, how would you take it out of a mold? You would need to sacrifice the entire thing for every attempt on making the motor.

Also, I’m choosing circles because they are the best structure against torsion and good at compression too.

Structural Shapes Ranked and Reviewed - Which one Wins?

I couldn’t find the video in which I saw such information, but I think this one is relevant the same way.

Now because of all of this, the motor now has 40cm of diameter. super.

My brother in christ, this thing still couldn’t reach 7.5 times less volume than the original, only 4.8.

Well, after trying a crapton of different designs, I finally settled on this one.

Essentially, I divided the structural part of the motor in 3 disks 55mm apart (or you could just use one).

Obviously, you would need to add the epoxy + 20% iron powder between the spaces for the back iron.

Then you do 3 more for the 300 kilowatt motor and so on.

The outer armature/back iron weighs 1.7394 kg in total, the inner armature/back iron weighs 1.2 kg in total, + 0.6kg of the coil = 3.6 kg in total. For the 300 kw motor = 10.8kg. 14 actuators in total for the lower part = 151.4kg.

I think I would need to take out 1 of the armatures then:

1.1596 kg for outer structure + 0.8 kg inner + 0.666 kg coil = 2.6kg x 3 times the output = 7.8 kg x 14 actuators on the lower body = 110.2kg in total.

And this is like, around 40 kilowatts per kilogram. Twice the density of Evolito’s motors.

You know what? I said “Axial flux motor be damned”, but now I’m curious on how it will perform.

First, simply inputting the surface area of the 100 kW motor's rotor-stator (2220cm²) into a circle area calculator will give you around a flat disk with 53cm of diameter, but you could divide that into different stators.

Well, if they had 306mm of outer diameter and 150mm of inner diameter, I would need literally 2 rotor-stators for one side alone, 4 in total.

The only way I could make this thing as light as the radial motor was using tubes with 1mm of thickness and 10mm of outer diameter and 70mm of length, they aren’t meant to go all the way through the diameter of the coils, but to support them. I also choose to do this way because it will work kinda like a centrifugal fan, but I don’t think it will move too much air taking into consideration that it won’t make full rotations.

(I should’ve done that to the radial motor, no?)

(did it, but it didn’t change much)

Again, you don’t have to do exactly how I 3D modeled it, you could use foams, sponges etc.

The 100 kilowatt axial flux motor has 306mm of diameter and 40.3mm of thickness, the 300 kilowatt one has 104mm of thickness.

Besides the tubes I added an extra layer with 1mm thickness between the tubes and the coils, this would be the epoxy resin + 20% iron powder, so you could make the tubes out of something else or just keep the epoxy composite.

You could also use stringers instead or with the tubes. 

Source: https://www.researchgate.net/publication/318502058_Skin-stringer_component_with_self-healing_functionalities 

I will make a 3D model with stringers on it, just because I think they would work better.

Also, I think I’ve made a mistake due to my lack of organization and confusion.

Essentially, the resulting weight of aluminum coils with 99 slots and 66 poles should have been 6, but I reduced to ⅓ of it because of the other motors that resulted in 2. Then I added +20% to compensate for inefficiencies (90% to 80%) 2.4 kg, then I reduced it by 40% due to the cooling at -70ºC resulting in 1.4kg for the 300 kilowatt motor and 0.48kg for the 100 kilowatt.

Observation:

It should be 2 kilograms, not 1.4 kilograms because the 2kg already assumes it has the conductivity of copper. To compensate you would need to make it heavier by 15% of the 1.4kg of weight, or 0.2kg heavier. For the air core axial flux brushless motors, you just need to add 1 disk with 2 coils for the 100kw and extra 3 disks with 6 coils for the 300kw. Which in turn would make the 100kw air core axial flux motor have 60mm of thickness and the 300kw would have 150mm of thickness.

So, 0.48kg of aluminum coil + 0.3125kg for the stringers + 0.41715kg for the back iron disks = 1.2857 kilograms for the 100 kilowatt motor. 22 actuators in the upper body x 1.2857 kg = 28.2 kg

I could multiply it by 3 for the 300 kilowatt motor but it is not exactly 3 times more mass due to the way it is organized. It weighed around 3.8kg.

3.8kg x 14 actuators in the lower half = 53.2kg.

This is a power-to-weight-ratio of around 78,000 watts (104 horsepower) per kilogram.

In both cases I’m assuming the stringers are made out of 20% iron powder 80% epoxy resin (density is 2.5 g/cm³).

Here they are, the 100kw one has 44.3mm of thickness, the 300kw has 105mm and both have 306 mm of outer diameter and 150mm of inner diameter.

In the end I will stick with the Axial Flux motors.

Also, the 3D models are named as:

Xangô-VC = Xangô Voice Coil 

Xangõ-3-AF = Xangô 3 Axial Flux

Xangõ-3-TG = Xangô 3 ThinGap

Observation:

Finally, I do think you should use plastic staplers (or a fiber embroidery) to absolutely secure the coils to the rotor-stator’s structure.

https://www.thingiverse.com/thing:4540908 (you could 3D print or DIY the staples to the appropriate size and weight).

Although you can use extra strong epoxy and/or metals, this is a flat aircore coil for a 100-300 kilowatt motor. The electromagnetic field alone could bend the rotor/stators and/or rip the coils out of their structure.

I’m not an engineer and I didn’t even attempt to simulate the structural strength of any part at all.

Be aware.

Every time I think “I will settle with this design”, a couple of days later I start wondering if it was really the best design that I could’ve come up with.

Wouldn’t a design similar to the first one I’ve made better for construction? I mean, it would be even easier than this one. But I don’t know if the iron powder + epoxy composite would be a good enough core material…

Observation 2:

It just dawned upon me that you could simply use laser cutting (or mechanically cutting) a 1mm thick aluminum plate instead of using wires wound by hand.

I will try to make a 3D model of such a coil.

I’ve made the 3D model, but I don’t think it is a very good result, you really can’t tell by just looking at it, but a lot of the sections are slightly bigger than 5mm.

I had the “brilliant” idea of simply making a coil that is the space between the coils and then cutting it out of a solid arc mesh.

Just now it occurred to me that I shouldn’t care about the number of parallel wires the same way I shouldn’t care about the number of turns.

(literally every page of this project)

As long as it weighs 2 kg to 0.6 kg of aluminum coil, and as long as it fits inside the rotor-stator, it doesn’t need to have X or Y parallel wires.

Then, you could have whatever parallel wires and whatever the number of 1mm thick wires that fits in the stator-rotor it will be amperage that you will be able to apply.

So it can end up being 800 amps or whatever.