The article also references another article by the same author about cavitation and propellors, which is studied with water tunnels.[1]

The water tunnel was invented by Sir Charles Parsons. Parsons had invented the compound steam turbine, the long shaft of finned wheels seen inside all modern turbo devices, in 1884. The first turbine product was turbo-generators for power stations. It took a while for demand to build up, since this was only five years after Edison's light bulb demo.

So Parsons decided to try ship propulsion. A prototype boat was built with one turbine and one propeller. Too much power was available, and the propeller was cavitating, spinning in its own self-created vacuum and going nowhere. A demo for the British Admiralty was an embarrassing flop.

This was unexpected. Before Parsons' turbine, nobody had enough engine power to force a propeller into cavitation. Parsons had to start studying propeller design.

So Parsons built the first water tunnel, in Newcastle.[2] There was a closed loop of water, maybe two meters around, and a window to look in at the propeller. Parsons could see the cavitation bubbles.

Having built a debugging tool, Parsons was able to try out propeller designs, and came up with some workable high-speed propellers. So it was time to rebuild the test boat.

That was Turbinia, version 2. One boiler, three turbines, three long propeller shafts, and three propellers on each shaft, spaced over a meter apart. Way overpowered. First boat with serious vroom. Turbinia was far faster than anything else on the water.

Turbinia's public demo went down in history. The British Navy had prepared a huge parade of warships for Queen Victoria's Diamond Jubilee fleet review. The might of the British Empire was shown off to the rest of the world. Everybody who was anybody, including the Queen, was there. Parsons brought Turbinia up behind the fleet and went to full power, turbines screaming. Turbinia zoomed through the columns of ships and disappeared into the distance. Some Navy patrol boats gave chase, but couldn't possibly catch Turbinia.

The British Admiralty wasn't pleased with this. But they couldn't do much. Sir Charles Parsons, son of an Irish lord, was a peer of the realm. He could only be tried by the House of Lords, which was more concerned about British naval superiority than the Admiralty being embarrassed. Also, there's a story that when the Admiralty sent some people over to Parsons' offices to chew him out, the Prussian naval attache was just leaving. Within a few years, Parsons turbines powered two new destroyers, and after that, the entire next generation of naval vessels.

[1] https://jordanwtaylor2.substack.com/p/killer-bubbles

[2] https://research.ncl.ac.uk/marinepropulsion/resources/fundam...

[3] https://en.wikipedia.org/wiki/Turbinia

If you happen to be in Newcastle, you can see the Turbinia at the Discovery Museum:

https://discoverymuseum.org.uk/

It looks impressively fast even to modern eyes.

Old tech, we simulate such tunnels now. Wind and water. For cars, airplanes, ships, surfwaves, river fixes.

CFD isn't perfect, reality still has a few suprises, and serious R&D teams use CFD to hone ideas while still using wind tunnels, real water, etc on models or real life at 1:1 scale to final test and verify.

Your claim aside, we still see articles and papers such as:

(2019) https://global.ctbuh.org/resources/papers/download/4200-will...

(2025) https://www.dlubal.com/en/support-and-learning/support/knowl...

asking if wind tunnels and "real" scale models can be fully set aside.

As has been the case for decades now, the answer is still "almost" and "soon".

CFD isn't perfect but better and cheaper. About once a year we try the real tunnel, but for the rest CFD is better. Esp. now with the new graphic cards

I worked at NASA Ames for seven years on a system which supported wind tunnel tests. For a while my team worked in a cold and noisy computer room under the 80x120 wind tunnel test section. We supported around seven of their large wind tunnels.

We had a PDP-11 system for data acquisition and real time display of data. That sent data to a VAX for data storage and near real time calculations and then to an Evans and Sutherland system for visualization. I worked on the RMS data storage and CLI.

We supported tests on airplanes, helicopters, submarines, space capsules, and even a tractor trailer. Since then I've seen a few tractor trailers on the road with flaps at the rear which help smooth the airflow.

One time our customer found a bug while a test was in progress. A co-worker and I went into the control room next to the test section. We could hear the air roaring and feel the building shaking while we worked on finding the issue. No pressure like that!

My dad worked on the Unitary Plan wind tunnel. I've been inside it and in the control room one time when it was running. Indeed is loud.

I think it pulls up to 100 megawatts when running.

If you want to know more about the European transonic wind tunnel , there’s a great podcast from OMEGA Tau

https://omegataupodcast.net/76-the-european-transonic-windtu...

Such a fun topic. And then there are some things tunnels can't do, or can only do in the most limited form, like analyse post-stall behavior. You can't very well have your model tumbling free in your flow, it might crash and make a mess. (Well, there have been attempts, some of them successful, but also quite limited in their parameters.)

So you have to resort to a free flight model, scaled down to simulate the parameters required. There's a (fascinating, for those so inclined) treatise on it here[0].

[0] https://ntrs.nasa.gov/api/citations/20110012492/downloads/20...

Having spend some time there, I really like the R J Mitchell (designer of the Spitfire) wind tunnel - used to be at Farnborough until they took a section of it to Southampton Uni. Real sense of history using it!

Most of our use is validating sensors for use outside wind tunnels, that is showing our on-bike drag sensors match wind tunnel measurements (it's for serious cyclists and triathletes). Watching live drag (more accurately, CdA) change with body position is fascinating - and tells me my back can't handle the most aero position for more than a few tens of seconds.

> turbulent air is flow (...) propagating fractally almost to the quantum scale.

Wow, do we have measurements that show this?

One interesting thing I learned when I visited Kitty Hawk is that the Wright brothers created their own wind tunnel to test different wing profiles. It’s really impressive how systematically they worked. True engineers.

Great article. Nice review... Helps to see the mathematical equations and numbers, like Bernoulli and Reynolds's, compared to graphs and flow tunnel simulations.

There is a third type of tunnel not really discussed here: hypersonic wind tunnels. They are closed but not looped. At one end are giant tanks of compressed and chilled air, almost liquid. At the other, even bigger expansion tanks filled with vacuum. Tests last only seconds and are more like explosions than wind.

https://asiatimes.com/2023/06/chinas-jf-22-hypersonic-wind-t...

In the western world, hypersonic testing has been traditionally done live. For groups like NASA, it was always a tossup between biulding a monster tunnel, or just strapping the test articles to the sharp end of a rocket and letting the upper atmosphere be the tunnel.

I worked next door to one in UQ. It was cobbled together from structurally strong steel parts, said to include former gun barrels, and major mining equipment prop shafts. I still have a pressure plate diaphragm, explosively burst like a bubble when the air is forced through it to make the hypersonic shock wave. The chamber is big enough for a wee 1/72nd scale model thing: typically a ramjet engine.

I used to have a teletype 33 from the same lab. At the end, it was only being used to submit orders to university Central stores for cleaning products.