Researchers want to build a dam with a distributed 3D printer

[Daniel Sims]

Forward-looking: People have been building structures with 3D printing techniques for years, but one group of researchers wants to bring AI-driven construction to a new scale. If successful, it could construct a dam more quickly and cheaply than with human labor.

Researchers from Beijing's Tsinghua University recently unveiled plans to build an almost 600ft-tall dam using 3D printing methods. It could be the biggest 3D-printed, AI-built structure ever made, but that may also depend on the definition of "3D printing."

Builders have already saved time and money by constructing homes using large 3D printers to make concrete layers. Last year, a fabrication company used robotic wire-and-arc additive manufacturing to 3D print a metal bridge in the Netherlands. The Tsinghua University proposal, however, doesn't involve any such techniques.

The researchers want to complete the currently-in-construction 590ft Yangqu hydropower plant on a Tibetan portion of the Yellow River using a legion of AI-driven robots, excavators, trucks, bulldozers, pavers, and other vehicles. Combined with an automated scheduling system, the researchers refer to the system as a single massive 3D printer.

Machine translation from Chinese

No humans would be directly involved in building the dam. An AI would slice a 3D model of the project into layers, then assign each layer to the unmanned machines sequentially. The AI would automatically plan material collection, driving routes, and placement. It can also analyze vibrations to determine build quality. Humans would only mine the fill rocks.

Aside from advantages in speed and cost, the machines could better withstand hazards like low oxygen levels, work around the clock, and researchers think they are less prone to human mistakes. The most apparent drawback is the potential loss of jobs that would normally come from such a massive project.

The dam is slated for completion in 2024 and will deliver almost 5 billion kilowatt hours of electricity to Henan province.

The Chinese government very well could sign off on the proposal given the success of another large 3D printing construction project in Shanghai last year. The Wisdom Bay bridge is an almost 30-foot-long Bluetooth-operated bridge that can unfold in under a minute.

Image Credit: Weibo

Permalink to story.

https://www.techspot.com/news/95802-researchers-want-build-dam-distributed-3d-printer.html

 

[yRaz]

Okay, so this is one of the few topics I actually know a s*** about. First off, 3D printing concrete was never a thing, it's called slip-forming and was around WAY before 3d printing. Basically all highways, Bridge decks and concrete skyscraper support columns use slip forming. People just used 3D printing as a marketing term to make money. Is slipforming a house useful? Yes, but it's also a lot slower than pouring a footer and often weaker. It needs to be wet and the wetter it is the weaker the concrete is. Slip forming also has the drawback of having difficulty working around rebar. In some structures you can use glass fiber reinforced concrete, but there are limits to how large you can make something without using steel reinforcement. Pre and post tensioned concrete is also basically necessary for large structures. You cannot do that with glass fiber. Glass fiber also has the drawback of being abrasive and wearing out the molds of the slip-forms.

but now lets get into the mechanics of slip form machinery. They are all hydraulic based and they DO break down. I'm not saying it's impossible, but I find it unlikely that AI will be able to when(not if) a hydraulic motor fails.

But there is also the issue of "cold joints" in this case. This is where dry concrete meets wet concrete. It is a weak spot and you need special care to keep the structural integrity at these points.

EDIT: The cold joints in concrete are similar to the weak points in 3D printed plastic

 

[Mr Majestyk]

Dams be damned. Another Chinese eco-wrecking venture.

 

[Uncle Al]

While it sounds like a good idea, only the test of time will tell. We have seen too many things that promised the moon, but failed in the long run and as technology zooms ahead, how many of these little "miracles" will be left behind to the point that replacement parts for repairs will cease to exist, making the great tech project a seemingly failure.