3 ms·

I would suspect that the conditions behind the VLEO satellite are close enough to a vacuum that vacuum chamber results are actually relevant.

by phire 2y ago

throwup238 2y ago
Getting to ultra high vacuum like in higher orbits requires a multistage vacuum pump. To get the right conditions for VLEO they only need to skip the final stage turbomolecular vacuum pump or run it slower to decrease its effectiveness - it's basically just a bunch of fan blades that use transfer of momentum to bump molecules out of the vacuum chamber and keep them out. I think the harder part is simulating the relative velocities of the molecules of air. They can try to use the same mechanism as a turbomolecular pump to accelerate the remaining molecules at the test setup, but AFAIK the maximum velocities in a turbomolecular pump are on the order of 1km/s, not 8km/s like in VLEO. Quick back of the napkin math says it'd need tip velocities of 4km/s which translates to over 380,000 RPM with a 4 in blade radius. I'm not sure there any materials capable of withstanding those forces. At 0.5m diameter it's a more reasonable ~76,000 RPM but that's basically the size of a commercial jet engine that probably caps out at 10-20,000 RPM.
- MadnessASAP 2y ago
  I know more about turbines then I do vacuum pumps, and do agree that about 1000 m/s tip velocity is pretty much the material limit. I suppose something like the solid disc in a tesla turbine might exceed that however I don't know if that's useful or if you'll get even close to 8km/s out of that. Going beyond spinning discs there's the USAFs "Hypervelocity Wind Tunnel 9", that can hit mach 14 or 4800 m/s, which, still not good enough.
  - throwup238 2y ago
    Turbomolecular pumps are basically turbines except they’ve got more blades in a different configuration, they’re driven by a motor instead of extracting energy, they have no method for convective cooling, and can’t use any lubricants or cooling fluid because they would vaporize and contaminate the vacuum. I’ve got more experience with turbojet engines than turbines so I’m not sure how that changes the calculus on the last two points. Turbomolecular pumps use active electromagnetic bearings and require very precise manufacturing to match blade radius with the chassis and balance the blades. The material science for the blades is mostly the same. I’m sure they’d could get plenty of useful data at mach 14 but sadly I doubt the USAF wind tunnel machinery would work in a vacuum environment without melting immediately, and it’s probably way bigger than even the biggest vacuum chambers NASA has. I’ve never heard of the tesla turbine before so I don’t know how the angular momentum translates to the molecules.
  - thereisnospork 2y ago
    Presumably you'd have an array of 'slow' ion thrusters upstream of your test article in your vacuum chamber to generate your flow. Matching temperature and velocity profile exactly is probably fairly tricky, but producing nominally 8km/s of very dilute gas shouldn't be particularly difficult.
  - master_crab 2y ago
    I suppose something like the solid disc in a Tesla turbine might exceed that It’s been a few years since I studied these types of issues, but I assume the relevant mechanical deformation from these speeds (creep) is a problem no matter the geometry.

I would suspect that the conditions behind the VLEO satellite are close enough to a vacuum that vacuum chamber results are actually relevant.

by phire 2y ago

I would suspect that the conditions behind the VLEO satellite are close enough to a vacuum that vacuum chamber results are actually relevant.

throwup238 2y ago

Getting to ultra high vacuum like in higher orbits requires a multistage vacuum pump. To get the right conditions for VLEO they only need to skip the final stage turbomolecular vacuum pump or run it slower to decrease its effectiveness - it's basically just a bunch of fan blades that use transfer of momentum to bump molecules out of the vacuum chamber and keep them out. I think the harder part is simulating the relative velocities of the molecules of air. They can try to use the same mechanism as a turbomolecular pump to accelerate the remaining molecules at the test setup, but AFAIK the maximum velocities in a turbomolecular pump are on the order of 1km/s, not 8km/s like in VLEO. Quick back of the napkin math says it'd need tip velocities of 4km/s which translates to over 380,000 RPM with a 4 in blade radius. I'm not sure there any materials capable of withstanding those forces. At 0.5m diameter it's a more reasonable ~76,000 RPM but that's basically the size of a commercial jet engine that probably caps out at 10-20,000 RPM.

MadnessASAP 2y ago
I know more about turbines then I do vacuum pumps, and do agree that about 1000 m/s tip velocity is pretty much the material limit. I suppose something like the solid disc in a tesla turbine might exceed that however I don't know if that's useful or if you'll get even close to 8km/s out of that. Going beyond spinning discs there's the USAFs "Hypervelocity Wind Tunnel 9", that can hit mach 14 or 4800 m/s, which, still not good enough.
- throwup238 2y ago
  Turbomolecular pumps are basically turbines except they’ve got more blades in a different configuration, they’re driven by a motor instead of extracting energy, they have no method for convective cooling, and can’t use any lubricants or cooling fluid because they would vaporize and contaminate the vacuum. I’ve got more experience with turbojet engines than turbines so I’m not sure how that changes the calculus on the last two points. Turbomolecular pumps use active electromagnetic bearings and require very precise manufacturing to match blade radius with the chassis and balance the blades. The material science for the blades is mostly the same. I’m sure they’d could get plenty of useful data at mach 14 but sadly I doubt the USAF wind tunnel machinery would work in a vacuum environment without melting immediately, and it’s probably way bigger than even the biggest vacuum chambers NASA has. I’ve never heard of the tesla turbine before so I don’t know how the angular momentum translates to the molecules.
- thereisnospork 2y ago
  Presumably you'd have an array of 'slow' ion thrusters upstream of your test article in your vacuum chamber to generate your flow. Matching temperature and velocity profile exactly is probably fairly tricky, but producing nominally 8km/s of very dilute gas shouldn't be particularly difficult.
- master_crab 2y ago
  I suppose something like the solid disc in a Tesla turbine might exceed that It’s been a few years since I studied these types of issues, but I assume the relevant mechanical deformation from these speeds (creep) is a problem no matter the geometry.