3 ms·
Technically impressive, but unfortunately feels basically like a PR stunt. The title suggests that the melted material will be removed soon but is not realisti
by pvaldes 2y ago
Technically impressive, but unfortunately feels basically like a PR stunt.
The title suggests that the melted material will be removed soon but is not realistic. At this moment should say "will try to remove 3 grams from the >800 tons of melted fuel. At this pace, the place will be safe by natural means much before the robot will end the work.
Apart of this, I wonder... is this stuff still usable? There is something that could be done with this grams of melted dust that would worth the risk of spreading it around?
...
Second: why to design a robot for that, or move electronics close to this area so they can fry in seconds?.
The idea that must be some sort of robot (because some sort of idealized human rescuer figure is what the public expects) is not very smart. What they need is maybe (IMAO) a sort of "anteater-tongue machine" that just will try each crevice until finding the correct path. A sort of strictly mechanical machine moved from a safe distance with a real human brain taking the decisions much faster.
If you need to move a tube there are a lot of solid mechanical joints in the market tested since decades for every possible need. All needed is an axis able to rotate, a motor placed very far away, and a removable set of arms that would pick and vacuum the stuff and would double as "fuel rods" or something later when will be unavoidably contaminated. Just cut the tube in appropriate lengths and put the contaminated rod to sleep in the pool with the other stuff.
If electronics are an obstacle, just remove it entirely and use a stochastical movement instead a directed one.
Radioactive dust in the path could be grabbed with some type of greasy or sticky material before to be collected by a moving ring. No need to design a complicated claw here. There are several metallic types of grease available that could jail the dust and maybe even, dunno, provide benefits to block the dust radiation somehow into a matrix of metallic particles.
If you need to see the path, use echolocation or maybe run a camera forward for one second to retreat it fast to safety. The "tongue" could be used as a rail for the camera.
Alternatively, they could made the path safer first deploying a lead jacket made with tubes or plates where the robot can move. If the room is radioactive, line it in lead lames. That would also remove coins and crevices and would made the route much safer for people and machines and easier to navigate until the final part.
- aloisklink 2y agoI was at Fukushima Daiichi on Monday, and they explained that they only want to take a few grams of the melted material to test the composition of the melted fuel/debris, e.g. what elements are in it. Even just planning to remove all of the melted fuel is a long way away. I can't remember if they were talking about Unit 1 or Unit 2, but from what I understood is that due the collapsed rubble above/around the reactor, they only have a very narrow opening, which means they've struggled to use larger robots. And I believe the robot operating centre is a bit of a distance away from the reactors too (probably so that the robot operators don't need to wear protective equipment). > What they need is a sort of "anteater-tongue machine" that just will try each crevice until finding the correct path. I think the problem is that once you get through all the debris, there's a big cavern. Hence why they're using some sort of crane robot on a rope (it reminds me a bit of a [claw machine game][1]). [1]: https://en.wikipedia.org/wiki/Claw_machine https://en.wikipedia.org/wiki/Claw_machine
- pvaldes 2y ago> I think the problem is that once you get through all the debris, there's a big cavern. I see. Different situation then.