[mliben]

Silver is crazy expensive compared to copper, and the tradeoff isn't worth 6%. Also, 3d-printing pure copper is relatively new, and I'm not sure there is as much of a business case for 3d-printing silver (from the perspective of the companies making these metal AM machines), since the demand is lower because of the cost. Lastly, there are actually some loss mechanisms in the motor where the lower resistivity of silver would actually hurt you (proximity effect from flux crossing through the stator conductors, producing eddy currents).

We want to start simple with cooling, hence the water/glycol. There certainly could be some opportunity to use something different (maybe with certain fuel cells and liquid hydrogen already onboard?). Regardless, the thermal resistance from hotspot to coolant is dominated by conduction resistances inside the motor, and is less a function of the convection resistance from the housing to the coolant.

We will be sending out a newsletter occasionally, there should be a link at the bottom of our website.

Thanks for the questions!

[baybal2]

Silver is heavier than copper.

That's why people were so upset when carbon nanotube yarns happened to be poor conductors.

People were thinking of super light motor windings.

[mliben]

Aluminum has the best conductivity/mass out of all of the common conductors. Motors are actually more volume constrained than mass constrained for the windings, which is why copper is typically used there instead.

[LeifCarrotson]

Copper is also 60% better in the thermal conductivity properties as well, which is another critical property.

Power density in an electric motor is really based on how fast you can remove heat from the motor. I'm involved in sizing industrial servomotors, but even there you have 1s/10s/60s power ratings.

I wonder if H3X can post higher power levels for takeoff, assuming it starts cold and the flight plan calls for throttling back after a certain altitude is reached. And even in the event of an immediate 150% power return to runway after a 150% takeoff, the motor might only have slightly degraded the winding insulation; it can almost certainly exceed its ratings once for long enough to get back to the ground.

[mliben]

Yes, thermal conductivity is of the utmost importance at the continuous current densities we are designing for.

That being said, typically the effective thermal conductivity of the winding (perpendicular to the axis of current flow) is limited by the insulation (strand and/or turn) and the encapsulation/varnish. As a result, changing the thermal conductivity of the conductors themselves will have much less impact on the total thermal resistance (from winding hotspot to coolant) than changing the insulation and encapsulant thermal conductivities.

At these very high power densities, the thermal RC time constants inside the motor are very short (small motor = small thermal capacity, low thermal resistance by design). Therefore, even for a "short" 10 minute takeoff, most of the motor will have already hit thermal steady state. As such, the motor needs to be able to run at takeoff power continuously. There has been a lot of fun discussion elsewhere in this thread about how to tackle that aspect of the problem (given that takeoff power is typically 3x cruise power).

I will say that we are working on developing a high thermal conductivity (> 1 W/m-K) and high temperature (> 300 C) insulation system.

[airbreather]

In medium to large generators the winding pitch makes a big difference, but this is also optimised for fault current (as in the winding pitch selected) depnding on the installation being compact or spread out.

[faeyanpiraat]

It’s just a weird idea, but cant you use hollow copper winding, and run the coolant liquid trough that?

[mliben]

Yes, this is one variation of what is referred to as "in-slot cooling". Keep in mind that putting coolant inside the slot removes precious conductor cross-sectional area. There is a tradeoff to be analyzed, and in some situations it can make sense. However, in-slot cooling does lock you in to having a liquid cooling system, whereas the design we currently use (coolant channels integrated into shared housing) could be modified to be air-cooled if it makes sense to do so.

[h0l0gr4ph1c]

LOL, just had exactly the same thought.