Given that the deceleration phase recycles the energy into the system, and that the load/unload time is projected to be 5 minutes, I'd not think of it a big problem. The design can cope with frequent stops relatively well. A stop or two in each large state, should the system be used to go cross-country, would be OK, although it doesn't really make economical sense for it to go cross country anyway. Hyperloop only makes sense for short hops where you don't need to go potty.

An acquaintance has a small demo unit made as an (expensive) novelty item sitting on his desk. It's an aluminum pendulum and a 3 phase linear motor (just because, as he says). Runs off a couple AA rechargeable batteries. The pendulum is a disk and can be converted to a balanced disk by removing a weight. Once converted, you flip a switch and instead of going back-and-frth, it can spin up to 20kRPM in about a minute. The configuration is entirely immaterial, it's really a very basic thing, electrodynamically speaking. After you press the brake button, it similarly stops in about 40 seconds, while recharging the batteries.

There would be no attendant - what for, you can't walk inside of one anyway. Those are sit-only capsules. You probably can't exit your seat at all.

When it comes to emergency response, the default scenario is to reach the destination if mechanically possible. The whole system is designed to complete the journey of all capsules enroute with no external power and no sunlight - there's a lot of power-smoothing batteries at the accelerator sites. They have enough power not only for propulsion, but also to run all of the other systems, possibly for hours. If there was a local blackout, the stations would likely stay up as if nothing happened, the capsule traffic would merely be halted if there was no sunlight.

If further traver is not mechanically possible (many reasons here), the solution is to (mechanically) brake and crawl to the nearest station or emergency access point. There will be small electric motors and wheels to push you along at a modest pace (say 60mph?). The tube repressurization is a passive thing, so not a big concern - if a capsule signals that the onboard life support is down and the backups are down as well, a bunch of valves open and that's the end of it. Probably the repressurization could also be used as a stand-in for mechanical braking; the air-induced drag would surely stop the capsule rather quickly. A failure of the compressor would do the same thing although probably too slowly, there's storage for gas bearing air such that whatever braking mode is used, the gas bearings wouldn't run dry, so to speak. I'm sure the system would be engineered to behave. That's what engineering is for.

The whole "trapped passengers" issue is I'd think a bit overblown. The major active systems in a capsule are mostly not unlike those on a locomotive and on an airplane: a compressor, an electric motor, control and power electronics, a battery bank. Propulsion is external - the capsule merely has an aluminum stator sticking out from it. Due to low drag, the capsule is coasting without propulsion for 98% of the length of the route. For it to keep coasting, the compressor needs to keep on spinning, and you must have no leaks in the water coolant loops. They're not sure yet to what extent the active tilt control would be used.

The idea is nifty, and it's sorely needed. I think that if nobody else in the western world would pick it up, we'll end up seeing it somewhere in Asia. I'd like to be among the first passengers once it's open to the public :)

I'd say 2g turns are OK.

You also need to buy the easements to allow passage of your track above all of the land it passes above. That space also belongs to the land owner!

Just to give you an idea: a 2kW 40kRPM liquid cooled brushless motor is just a tad bigger than a D-size 1.5V battery. If you wanted it air cooled, it'd not only run hotter due to higher thermal resistances to the fins, but you'd need 4x+ the volume due to the size of the fins, and you'd be wasting another 100W to blow the air through those fins.

Oh, those are definitely possible to rebuild if you've got the right tools and can still get the parts. They are quite amazing to work on - rebuilding even a modern internal combustion engine is a downright boring thing after you've done a modern front wheel drive transmission.

In automation, you'd often use sealed brushless motors that are conduction or convection cooled through the case. Those have artificially low power densities since there's inadequate cooling of the stator. I can have a 2.5kW sealed servo brushless motor, or a 10kW-40kW (RPM-dependent) liquid-coooled motor in the same volume and roughly same mass. In cars, where weight and volume do matter, you need forced cooling for the motor. If your motor is under the car, and especially at the wheels, air cooling is out of the picture. You need liquid cooling.

I don't know where you got the idea that I imply that electric brushless motors somehow need maintenance. The only things that go bad on them are the bearings, and it doesn't take much to design those to outlast both the car and the owner.

As for the unsprung mass mattering only to race cars: ha ha ha.

Interesting. I think with LyX open I can keep up with anyone handwriting almost any math, perhaps with a few hours worth of practice if I'm rusty at the moment.

I mostly agree. A meeting half-way, for me, would be for students to print out the handout notes, formatted such that there are breaks between paragraphs for you to write your own notes in.

There's a lot of wisdom in that sentence. Kudos!

There's an easy solution for that. Just write your notes down on Surface Pro. As much as I adore Apple hardware, Surface Pro is about the only PC that I'm willing to put up with. I'm kinda sad in fact that it's not an Apple product. It's fucking fantastic for note taking and drawing. There's nothing else like it. I mean it. Surface Pro with Autodesks's Sketchbook Pro is pretty much all I'm using for my engineering sketches and diagrams - it has replaced my paper notebook and scratch paper. After a couple of weeks it's quite amazing how cool my notes look, even though I'm not a very artsy person at all.

They've been optimized to hell and back. Adding 10% more mass to them would make their life practically infinite. Alas, those 10% were optimized out, at quite a significant cost, too, since you need some fairly heavyweight modelling tools to pull that off. Heavyweight as $100k per seat, that is. I'd imagine you might pay more for this "infinite" life in gas than in costs of whatever repairs are needed, so it still may be a net win. Well, the companies doing those modeling tools definitely always win :)

AFAIK, modern automotive transmission and engine design practice models all those "little" things that are routinely overlooked when you design stationary machines with factors of safety above 4. An automotive transmission these days is designed like a helicopter transmission would be, except that the safety factors and nominal overhaul periods are higher, but they are still fairly well modeled. If you do it by an undergrad textbook aided by design codes, or even by an "experienced" engineer, you may get a design where the real factor of safety and real life ranges between 1/4 to 4x of what you intended. You do an automotive or an aerospace design, and you're between 1/1.3 to 1.5/1, and the tools you use to pull that off cost more than the education of yourself and a bunch of your buddies, combined. I don't know if it's a good thing...

Install menumeters and make sure you don't have a CPU hog.

95% of what?