Hacker News new | ask | show | jobs
by icegreentea 3574 days ago
Should probably begin by saying that there's no fundemental reason why China could not eventually develop their own modern turbofans without outside assistance. But obviously now is better for China.

Designing and building modern turbofan engines is incredibly dependent on your mastery of materials science and fabrication. These are areas of science and engineering that are notoriously "art like" or even "black magic like". The accrued knowledge that sits in Rolls-Royce or GE or Pratt-Witney don't really exist in textbooks, and they aren't taught in classes either. This knowledge and know how was generated over 60 years of trial and error, and are jealously guarded, not just by the manufacturers, but also by their governments.

Things to consider: modern turbine blades are already pushing the absolute edge of lightweight-temperature resistant materials. And then modern designs will run these blades at temperatures that are nominally above the material failure point, because we've come up with ways to cool the structures. Oh, and the temperature resistant alloys in use? Secret. Empirically determined, and likely unique.

3 comments

This being China...what are the chances they can buy existing designs and reverse engineer them and improve then so they don't start from scratch. They've done it before...
Well, they've almost certainly been trying that. And as the OP said, it hasn't been paying off in a big way. If they've been doing it (I'm guessing they have), then they've probably benefitted, just not enough for the big win.

Once you get into materials and fabrication, having a finished example to examine is rarely going to lead to a big win. If the secret sauce is in the fabrication (HOW do you make a single crystal blade? HOW do you inflate corrugated titanium? HOW do get this type of grain structure with this nickel-iron-manganese mix? HOW do you get a laser weld in this impossible again?), then a single example isn't -THAT- useful.

Reversing engineering a modern turbofan(remember, the basic principles of turbofans are well known) is more akin to reverse engineering Intel's fab process by peaking at a finished i7 (with AFM and electron microscopes) than reverse engineering say... a sensor fusion module once you get the assembly code.

Which is why the technology transfer associated with the engine deal is a big deal. It should (I mean, its possible that the terms are all fucked up) give the Chinese an oppourtunity to learn all of these hows and whys (albeit for a slightly dated design, but still a relatively modern one).

Yep, sending the prints over is just getting started.

Similar to razor blades and little balls that go into writing pens. Few companies can make them well. China has to import little balls from Japan (at least for high quality pens).

Sending over work instructions will help (somewhat). China will need veterans of the process to work with local Chinese vendors, explaining how to do each step.

The Ukrainian staff will resist handing over their work babies. Also, the subcontractors (heat treaters, metal workers) will not hand over their secret sauce.

China will figure it out eventually. This will help. Still many years away from mastering it though. They will have to source from Ukraine for quite a while.

In that case, China needs to build some good analog of KGB.
They have it, and Boeing and Airbus and US and European counterintelligence people are watching them.
How much of that trial and error can be skipped with modern computer simulations?
In the field of materials and fabrication? Not that much. Fundamentally, the problem is that for many of the problems that we consider interesting in materials or fabrication, meaningful analogs to say.. the Navier Stokes equations don't even exist.

In the field of fluid and thermodynamics? Probably quite a bit. This is why I tried to emphasize the materials and fabrication bit.

Things like metallurgy & fluid dynamics are notoriously difficult to simulate; this combines them in exciting ways.
Pretty much this. Incredibly hard to design from scratch; cheaper to buy an existing design and start from there.