Could a bladeless "Tesla Turbine" type air compressor be used as a high vacuum pump? I have a wacky idea for a new vacuum pump, I'm trying to figure out why it shouldn't work.  I'm interested in learning about Tesla Turbines, specifically when used in reverse as compressors or vacuum pumps.  There is a LOT of hype on these things, but I haven't found much real data on how well they work as vacuum pumps, and what equations govern their design. The best reference I have is this 2008 study.
This source seems quite good, but only studies the flow at atmospheric pressure and above, and gives equations for designing Tesla type disc blowers.  It is unclear to me whether these equations hold true all the way down into the Knudsen and molecular flow regime. 
Clearly, if the inlet was sealed off it would pull some (marginal) vacuum, as any centrifugal pump would.  But what would be the limit of this vacuum?  
The design of such a vacuum pump feels, subjectively, very similar to Gaede and Holweck molecular drag pumps.  Given small enough disc spacing, would a Tesla pump continue to pump effectively in the molecular flow regime?  Given large enough diameter blades, could it back itself to atmospheric pressure?  Why has this not been tried before? (i.e., what dumb concept am I missing in my understanding of these pumps?)
 A: Tesla Patented a multi stage pump specifically as a High flow high vacua pump, primarily intended to be use with steam condensers to evacuate any gases that get into closed loop systems from packing seals or the like and to achieve very high degrees of vacuum in the condenser while being able to drastically reduce the size of the condenser, or increase the performance of existing condensers.check it out. Its different from the original patent. (sorry i know this post is originally from a while ago so i hope it still helps)
as you get closer and closer to deep vacuum the pumps have less and less work to perform. While technically yes the efficiency drops at this point, the actual energy required to maintain the vacuum reduces as there is less and less mass flow rate through each stage of the pump.
NIKOLA TESLA BRITISH PATENT 179,043 - IMPROVED PROCESS OF AND APPARATUS FOR PRODUCTION OF HIGH VACUA
https://teslauniverse.com/nikola-tesla/patents/british-patent-179043-improved-process-and-apparatus-production-high-vacua
I'm currently in the process of reproducing the multistage high vacua pump/compressor specifically to be able to have a compressor that can supply enough CFMs for me to actually Test the tesla turbines that im making now. Check out the videos of my turbine build so far on my youtube channel www.YouTube.com/CharlieSolis I will be using .3mm to .5mm thick 150mm diameter Carbon Fiber discs with .25mm disc spacing. these discs should be capable of safely operating at 80,000-100,000RPM, which is very optimal as the compressor capabilities are directly proportional to the square of the compressor peripheral speed. If i remember correctly thats something close to 800m/s. to be honest im not even sure whats supposed to happen in the compressor volute, as in whether or not the supersonic speed fluid will be caused to continue increasing in speed because of the increasing area of the volute since its already going supersonic. Does the volute become like the diverging section of a De laval nozzle or are there different actions at play?
That being said i plan to utilize a more compact design that was patented by donald durant in 1975.
https://patents.google.com/patent/US4025225A/en
but i will be doing dual intake from both sides left and right and exhausting through a final stage in the center to cancel out the thrust forces via the pumping action.
A: The use of a Tesla high vacuum pump is indeed very compelling. I did some research during university for its use as a fluid pump, quite similar to the 2008 one you cited.
I guess the design approach might be that of the many particles. At ambient temperature, the average speed of air molecules is about 500 m/s, which is comparable to the tip speed of a high rpm rotating disk. The molecules impacting the surface will, in some measure, exchange momentum. This allows for a substantial modification of the average direction of the speed from approximately none prevalent to tangential, thus driving the residual gas to the periphery, where they can be driven away with a backing pump. The limit of the resulting vacuum is to be determined with experiments and simulation, however in principle it might be similar to that obtained by turbomolecular pumps, which also modify the average direction of the gas molecules.
I don't think it can back itself up, due to geometric reasons, but a multistage approach might be interesting. The potential here is for a device a bit less expensive than a turbomolecular pump with similar performance.
A: Tesla turbine works because there is friction between discs and air, but as you approach vacuum there is less and less friction and I guess the efficiency would drop badly before you reach say 95% vacuum.
But I'd like to hear about the turbine used as compressor too.
