Hacker News Reader

Comment by TheDudeMan | Hacker News Reader

So what's the actual reason no one uses this type of turbine?

The video frustratingly says, "People think it's inefficient. It's not." End of story.

I think the limiting factor on this type of turbine is the material you use to make the disks. As the disks get bigger the rotational stresses increase. To be practical the stress induced would need to be below the ultimate tensile strength of the material. Since these turbines need to spin at extremely high rpms and need to be fairly large to be efficient, there's no practical material that they can be made of.

Given that the point he makes is that they've got high torque at low RPMs and that's what makes them interesting, I'm not sure that's relevant here. Frustratingly it doesn't go into any more details than that.

> and need to be fairly large to be efficient

This statement is not true. Actually the inverse is true, the smaller you can make the disks, the better the efficiency.

See https://onlinelibrary.wiley.com/doi/full/10.1002/ese3.1417

> As disc outer diameter decreases from 100 to 50 and 20 mm, the isentropic efficiency of the miniaturized turbines increases significantly from 21.46% to 30.78% and 32.21%, respectively.

Not Op, but I suspect the use of "efficient" here is used more in terms of "usefulness of output" rather than pure turbine efficiency.

For example, you might have a VERY efficient turbine, but if they're constrained to be small (and therefore low-output), you then need to have a squillion of them, which is not an efficient use of money or effort expended, in terms of output power total.

Now there's an interesting optimisation problem...

We've all played enough Factorio/Satisfactory to know what the end result is going to look like.

It seems like while the turbine works great, the rpm output is too high, with too little torque, and you loose most of the gains when gearing it down, which is also difficult to do

That’s what I’ve read, haven’t built one

Lesics (another Youtube channel) argues that current turbines are already 90% efficient. For Tesla Turbines to reach the same efficency it needs to spin in really high rpm (maybe 50 000), which makes it impossible to built disks for it. Those disks would have to be 2-3 meters in diameters for industrial applications and couldn’t withstand the forces.

This wouldn't surprise me.

Modern turbines are very, very optimized, because they are used in systems like power plants. Here, every fraction of a percent of efficiency is massive. Hence, modern turbines are optimized as much as many, many smart people can do. This in turn means, even if Tesla just designed something pretty efficient many years back (which would be really impressive), it just loses in modern, industrial applications.

Like, people have been looking at better ways to generate electricity from nuclear or fusion reactors. But current steam and steam turbine systems have been optimized so hard and far, it's extremely challenging to find anything better than just heating water and jamming it through a turbine. Because current steam turbine systems are just that good.

It reminds me of the recent thread about hypothetical usages of the recently discovered and not yet confirmed room temperature super conductor. So many people are just arguing that a potential material that was discovered 2 or 3 months ago would be worse than silicon based electronics we've been optimizing for 40 - 50 years, or steel-based long range transfer we've been optimizing for longer.

Also, people forget that Tesla was more of a “throw together until it works” inventor, not a scientist. He didn’t believe in goddamn electrons, even though that was already the accepted theory of the time! While many of his inventions are fascinating and genius, I think there is an unhealthy mythicism and hype around him.

As for current turbines, I remember they are grown from a single crystal, as any kind of joint would be a breaking point in the material. They also have to abide by very strict size restrictions, at that RPM a single millimeter wobble can have catastrophic damage.

How does lesics arrive at that figure? In my turbine design course we were taught that the absolute maximum is around 57% else the outgoing fluid won’t have enough kinetic energy to get out of the way of the incoming

I don't know. Here is the Video: https://www.youtube.com/watch?v=AfCyzIbpLN4 And here with timestamp: https://youtu.be/AfCyzIbpLN4?t=448

Tesla invented this because he wasn't happy that cavitation was reducing the life of bladed turbines.

A plastics injection molding mechanical designer in China told me rotational blades are the hands down most irritating thing to try to injection mold because conventional tolerance factors that are very well suited to the production of most other products will often result in imbalance, noise, vibration, wear and failure in rapidly rotating bodies.

I've seen a five axis mill that is used to cut turbine blades (in Poland, through an industry contact of mine), it was an absolute work of art and I totally believe the injection mold designer. Everything was compensated for, including tool wear and heat injection. the whole thing was cut in one run without any tool changes or re-registering. One set of tools for one blade.

That is the kind of process engineering and fabricating that is fascinating to me.

It is. The company that built the system makes very nice stuff.

This is the Chinese parent company with subsidiary in Germany:

https://www.inovance.eu/industries/cnc-machine-tools

And this is the notice about them acquiring Powerautomation, the German company that designed and manufactured those controllers:

https://www.inovance.eu/news/details/inovance-technology-eur...

PA was a fantastic company but like many such specialist producers they didn't have a clear succession path laid out so they eventually got sold.

I’m similarly interested. Does anyone know of an introductory book on modern production technology? It somehow seems to be an impenetrable topic to me without actually doing a major in some kind of engineering.

This would be a great start but it is a pretty large book:

https://books.industrialpress.com/9780831136369/the-cnc-hand...

Also, the materials science and engineering handbook. It's floating around on the web in PDF form if you just want to skim it but if you really need it you should probably buy a copy (warning: huge book):

https://www.amazon.com/Materials-Science-Engineering-Handboo...

I used to work at a plant like this, we were acquired by Siemens. The machining process itself is less critical than the balancing and even the welding of these components, which is highly engineered and controlled.

I can imagine. Manufacturing tolerances will have some minor effect on balance and efficiency, failure of a weld or large imbalance or resonance would be fatal. Even so, the machining process itself was quite impressive to someone who is used to 'regular' CNC machining without working to such close tolerances. The movement of the cutter was very interesting too, hypnotizing if you will :) I forgot what the material was, these blades were for aircraft turbines.

What kind of turbine blades did you work on? Gas turbines for stationary applications or for aircraft or something else? Do you remember what material they were made out of?

The problem there is almost certainly people using "efficient" when they mean "fast" or "productive." Easiest example for that, in the past, was people mistakenly thinking gas stoves were more efficient than electric. Not even getting into induction, which can also be fast, the standard electric stove tops were far more efficient than the gas counterparts.

So, my guess on why people don't use this is that they can get more output from other designs. Even if they are "less efficient."

This reminds me of how annoying “efficient” appliances are. To comply with increasingly strict energy and water regulations, modern washing machines have small capacities. But that just means I have to do two loads. It’s not actually efficient at all. Certainly not time efficient. Each load probably has some energy and water overhead as well, so more is worse. Also, since agitators require energy, they are becoming less common, so the clothes are less clean, too, and that probably leads to extra loads being done.

The way we talk about and regulate efficiency is all messed up.

I think we do use this turbine but with thicker liquids and in reverse as a pump -- maybe it was in oil or some of the chemical industries?

It has terribly low stall torque. Basically it spins really fast and really efficiently, until you actually try to put any kind of load on it and make it do anything. The load slows down the disks, which disrupts the boundary layer and causes a negative feedback loop which makes it less and less efficient.

So load it with an electrical generator and use power electronics to present a constant load. Problem solved.

Many old high efficiency engine designs were discarded because they had poor torque characteristics at the bottom end of their speed range. This is because they were coupled directly to the load through gears. That does not have to be a requirement any more.

A constant load is only useful under limited circumstances, you would need to be either constantly starting and stopping your generator or wasting energy with a dummy load for this to work.

Could you do it? Sure. Does it completely negate the advantages of the turbine? Absolutely.

depends who you ask. If you watch these characters they come and go. They often argue with each other. Charlie is the latest and certainly the least ridiculous.

The best argument might be that current manufacturing techniques have caught withe requirements.

I plan to build one.

Please keep us appraised of your progress!

It's a casual desire and a component of a larger ambition. I hope Charlie's proves useful.

I do plan to start documenting and publishing but I'm still a few months out.

Thank you for the encouragement. It always helps.

You cannot have very high torque on this thing.

Correct.

The size of the disks compared to a traditional turbine would easily consume a four to six story building for, say, steam energy generation. They would also need to spin at 50k+ rpm to reach their claimed 90% efficiency. Thats a blade tip speed of mach 13. Not easy.

Reply on news.ycombinator.com