Comment by geocrasher | Hacker News Reader

Comment by geocrasher | Hacker News Reader

Indeed, Mach 3.5 is the declared top speed in this story, but higher speeds are implied.

     The Mach eases to 3.5 as we crest 80,000 feet... Screaming past Tripoli, our phenomenal speed continues to rise... I realize that I still have my left hand full-forward and we're continuing to rocket along in maximum afterburner... The TDI now shows us Mach numbers, not only new to our experience but flat out scary.

He mentions unusually cool temperatures allowing higher speeds than usual, and the aircraft is performing better than ever. So who knows. 3.6? Did a perfect set of circumstances, mixed with a rather urgent need to depart, take them closer to 3.7?

It's truly incredible that the A12/SR-71 aircraft was designed without the benefit of modern computing. In Ben Rich's book "Skunk Works" I recall there a bit about a comment that somebody once made regarding Kelly Johnson, something like "That Swede can see air". Truly a compliment to Johnson's almost uncanny ability to design aircraft for a specification. The U2 was another example of such ability.

> It's truly incredible that the A12/SR-71 aircraft was designed without the benefit of modern computing.

One of the huge things that computers give us is the ability to simulate where the design limits actually are - you can "relatively" easily design a device that needs to do "at least X" but once you're done, you don't really know how far past X you can go - computer simulations can help discover where and what exactly fails.

(Some limits like boat "hull speed" are easier to find than others).

The computer allows you to cut down your design so that it will do "at least X and no more" so you can save a few cents and then the plane falls apart when the pilot asks it to do Mach 3.6 but that's OK because the manual said it will top out at 3.5

"Any idiot can build a bridge that stands, but it takes an (computer assisted?) engineer to build a bridge that barely stands"

It's actually an interesting problem from a "total effort" viewpoint.

There's zero reason to design a kitchen to last 100 years if the average one will be remodeled in 7 and even the oldest won't go past 20.

But other things can be designed to last significantly longer, and be used way past their "original expected lifespan" - many of the buildings in Europe, for example, have lasted centuries.

True.

That said, w.r.t old buildings, selection bias may play a significant role, perhaps bigger than intentions

And maintenance costs, too. You may have a bridge that's 100 years old but you're paying 1/10 a new bridge in maintenance every year - not noticeable.

There are also other factors. Imagine you gave a 500yo bridge that is low maintenance and has stood the test of time and all those nice things, but you can't carry the weight of modern vehicles, or if it's not wide enough for modern vehicles etc etc

One of the many tradeoffs in design is how much do you want to spend in preparing for anticipated future demands.

Striking example in my home town of Zagreb. Communist city planners at the time designed new neighbourhood with plenty of parking space because one day productivity advancements will give every family a car to own and they need to have a place to park it!

A few decades later you couldnt find a parking space anywhere because every family had two or three cars!

Also incredible most of the titanium came from the USSR but through indirect routes and suppliers with the Soviets being unaware.

Which is really odd, as Titanium is not that rare.

Are we sure that story about the Titanium coming from the USSR is accurate?

It's a fairly well known and documented story, do you have anything similar in evidence rather than just conjecture?

I think it's mostly anecdotal (I have heard it too, but I can't think of where)

Not a metallurgist, but if I had to guess it was hard to find in large enough quantities at the time. I also understand that it's a lot harder to work with than say aluminum or steel

Speaking of which, King Louis served his guests with aluminum silverware, since at the time it was more valuable than gold due to the difficulty of extracting it at the time, despite its abundance.

"Skunk Works" by Ben Rich is a first hand account by the (eventual) lead of that Lockheed division, and he describes this in his book.

https://www.amazon.com/Skunk-Works-Personal-Memoir-Lockheed/...

Why deplete your own supplies when you can use your opponents?

Titanium is very common in the Earth's crust. But turning titanium minerals into pure, ductile titanium metal is more energy intensive and complex than making most other metals:

https://chem.libretexts.org/Bookshelves/Inorganic_Chemistry/...

At the time, the USSR had already invested significant resources into making titanium for submarines. Buying titanium under false pretenses was easier than increasing the domestic industrial capacity for making titanium, particularly when the plane was not going to be produced in large numbers.

Related:

https://en.wikipedia.org/wiki/Losos-class_submarine

The Alfa submarine would seem to be more relevant.

I don't disagree but the Piranha is damn cool. Crew of 9!

Also the Papa-class (only one built), which preceded the Alfa and remains the fastest submarine ever built. And the Sierra-class submarines which followed the Alfa class.

https://mentourpilot.com/sr-71-when-cia-used-soviet-titanium...

There was a government arranged meeting betweeen Boeing and Soviet engineers to discuss working with titanium for the SST: https://www.smithsonianmag.com/air-space-magazine/the-titani...

In the alpine climbing world Soviet climbers used titanium gear that was considered very good and light. Some of it was from scraps and spare material, but they had access to titanium.

Second the recommendation Ben Rich's "Skunk Works" as a good read. A fascinating tale of hardware development in an earlier age.

Temperature is the limiting factor for speed, I remember SR71 pilots on a panel saying the aircraft flew faster in winter than in the summer for that reason.

I'm fuzzy on details but I remember reading it had some pretty crazy thermal expansion. I think the pilot had to wait quite a while (hours?) before opening the cockpit to give the plane time to cool enough.

It's also cool that the whole thing was machined 'loose' so that it would expand as it heated up. Apparently the plane would leak fuel and oil when it was on the ground ...

Yep, JP-7 [1]. Special low vapor pressure to mitigate accidental ignition hazard.

The Sled leaked like a grad student's '69 VW bus in the hangar. Leak-free above > 1.5 Mach.

(edit to add link)

[1] https://en.wikipedia.org/wiki/JP-7

What does winter mean at 80 000 ft and Mach 3?

The jet stream is insanely fast at that altitude mixing the air, and at Mach 3 the surface weather changes in fractions of a minute.

Winter means that your nearest pole is tilted away from the sun. Your half of the planet is cooler, especially towards the poles.

For instance the James Webb telescope gets clearer pictures in summer because in winter the astronauts breath can fog up the lens.

Such a great book that is becoming harder to find a decent copy of these days.

There's a great sketch that Kelly Johnson did of the A12. This was in September of 1958(!) and he had a rough top speed of Mach 3.0, with a possibility of Mach 3.2. And it existed on nothing but a sheet of paper and in his mind. No computers with advanced software and super computers with petaflops of power.

https://www.thesr71blackbird.com/History/CIA/a-12-blackbird-...

First time I've seen that sketch, but it looks like Popular Science lifted it directly for an article from the 80's that I remember distinctly about Project Aurora.

Adrian Newley is aerodynamist and the current chief designer of Red Bull (current champions and top of the season). He has won more F1 champions than any other designer.

He draws his cars on paper.

Granted, I'm sure someone digitizes his designs and runs them through a CFD suite, if for no other reason that wind tunnel time rule limited.

Still.

It’s a nice story to read about rock star engineers but F1 race car aerodynamics is not done on paper.

Sure concepts to try can be sketched out on paper. Things like wing arrangements, rough intuitive ideas about where the clean and dirty air is. But everything in F1 is an iterative loop between a lap time simulation, vehicle dynamics, and aerodynamics.

Vehicle performance is largely a trade off between Cornering dynamics (typically demanding more downforce) and straight light speed (typically demanding less drag) which is why different tracks have different setups.

Basically balanced between cornering sections and straight line sections to figure out the optimal balance of lift/drag and the dynamic behaviour of the rest of the car in response to those aero set up effects.

Obviously they use computers. My point is that Newley, (claims at least) to do his design work on paper.

Which, is not that surprising. If you're good enough to have a draughtman digitize your designs for you, you're free to draw what you want on a five by four foot surface and not fiddle with a GUI and a 30?, 40? inch screen.

Overnight they print you the latest drawing on the plotter and you're back to you doing you. Id live to have a secretary print me my work and then digitize it in the evening.

His name is Newey BTW.

What you're describing is not how engineering is done.

There's no overnight printing of the drawings.

The 3D CAD models are designed and available in real time, they are used in various CFD and FEA models. The only time drawings are printed out are dimensioned production drawings sent to manufacture the parts.

https://fractory.com/engineering-drawing-basics/

There are no digital draftsmen, it's sort of semantics but design engineers are working directly with parametric 3D models, they can produce 2D drawings from the models.

As a senior engineer Newey is likely reviewing Wind tunnel data, CAD designs of concepts, CFD/FEA Data, laptime sim results, and collaborating "over the shoulder" of less senior engineers responsible for specific aerodynamic packages, who then feed that info in collaboration with design engineers and surface modelers to update the CAD models.

Im sure the workflow is complicated, just like you said. Im sure lots of Console licenses and Ansys licenses are being used.

And yet here's Newey drawing on a piece of paper. And he hardly looks like a young chap in the 80s.

Who knows, maybe this photo was for the cameras. Maybe he lies that he draws his cars to throw off the competition. But, for what it's worth he claims in interviews that he draws his cars in pencil.

https://www.racetechmag.com/wp-content/uploads/2017/01/Scree...

Reply on news.ycombinator.com