js
Flashlight Enthusiast
I don't know, actually. I thought it was an analogy, but perhaps there are solders meant to go superconducting in LHe, and which will solder Niobium. Dunno. Maybe IMSabol knows?
Large Hadron Collider goes online tonite!
- Thread starter Chuck289
- Start date
Help Support Candle Power Flashlight Forum
I don't know, actually. I thought it was an analogy, but perhaps there are solders meant to go superconducting in LHe, and which will solder Niobium. Dunno. Maybe IMSabol knows?
js
Flashlight Enthusiast
OK. I talked to one of the physicists who is on the LHC machine advisory committee about the magnet incident, and specifically asked him how they joined the bus bar to the magnet leads. Turns out, it is soldered! It's just normal solder of a copper bus bar onto a copper lead, but embedded in the copper are lots and lots of niobium filaments. So the idea is that when the leads and bus bars are cooled to liquid Helium temperature, the niobium filaments have zero resistance, and thus this is where the current likes to go, as it is a current divider that is zero resistance vs. low resistance, so where possible, all current goes to the niobium. And the small bits where the current must cross solder and copper boundaries have such a low resistance that it doesn't present much of a voltage drop even at very high currents.
The reason it is done this way is that, apparently, there is no way to join niobium to niobium for a super conducting joint. If it isn't cast or machined as one continuous piece, then, no dice. So they embed niobium filaments (wires) into copper and join it in the usual ways.
As for magnet damage, it is too early to tell for sure, but it is looking as if one of the magnet coils was indeed damaged. And, also, I was right about the bus bar exploding / vaporizing! That must have been something. Crazy the amount of stored energy involved here.
The reason it is done this way is that, apparently, there is no way to join niobium to niobium for a super conducting joint. If it isn't cast or machined as one continuous piece, then, no dice. So they embed niobium filaments (wires) into copper and join it in the usual ways.
As for magnet damage, it is too early to tell for sure, but it is looking as if one of the magnet coils was indeed damaged. And, also, I was right about the bus bar exploding / vaporizing! That must have been something. Crazy the amount of stored energy involved here.
TedTheLed
Flashlight Enthusiast
very interesting. still I wonder what was wrong with the solder joint that caused it to heat up? Solder not applied correctly? niodium bearing copper wires not placed properly before solder?? just curious.
(btw) if the current encountered resistance at 0 volts, wouldn't that be a voltage rise, not a drop?? or on rereading did you mean just that the solder areas have very low resistance at normal temps? )
(btw) if the current encountered resistance at 0 volts, wouldn't that be a voltage rise, not a drop?? or on rereading did you mean just that the solder areas have very low resistance at normal temps? )
js
Flashlight Enthusiast
errr, I'm not sure what you're getting at, Ted. You mean the current divider thing? I just meant that the current will go through the niobium where it can precisely because it has zero resistance, and where it must go through the solder or copper, it only has a short way to go, and the copper is very cold, so the total resistance of that part of the bus bar is very low. Not zero, but very small.
Flashanator
Flashlight Enthusiast
NA8
Flashlight Enthusiast
- Joined
- Jun 4, 2007
- Messages
- 1,565
If the bus bar vaporized it sounds like the solder joint connection wasn't a big concern. Or did I miss something ?
Edit: Nevermind. Here's the part I missed (per js):
"the problem at the LHC was at a bus bar connection between super conducting magnets, in a whole series of cryogenically connected SC magnets. There was quench protection on the magnets themselves, but not on the bus bar. This is where the failure happened, and the bus bar went normal conducting, got hot, developed high resistance, and boiled off the cryogenic cooling for that section of magnets. This caused a quench and an interlock trip, of course, but about half of the stored energy in the magnets (which is enormous) went into the resistive load presented by the bus bar. And freaking EXPLODED IT. VAPORIZED IT."
Last edited:
js
Flashlight Enthusiast
NA8,
Yeah. I should clarify: the solder joint developed slightly higher resistance. This in turn caused more heat to form in the joint. This in turn caused the entire bar, or a significant portion of it, to go normal conducting. It was never meant to handle this kind of current while being normal conducting! This raised the resistance of the whole bar up to a high enough value that it presented a resistive load for all the stored energy in the magnets, and POOF!
Yeah. I should clarify: the solder joint developed slightly higher resistance. This in turn caused more heat to form in the joint. This in turn caused the entire bar, or a significant portion of it, to go normal conducting. It was never meant to handle this kind of current while being normal conducting! This raised the resistance of the whole bar up to a high enough value that it presented a resistive load for all the stored energy in the magnets, and POOF!
TedTheLed
Flashlight Enthusiast
ok, so, to reiterate, exactly what did the goof who soldered the sucker do wong?
..and do you think they'll put sensors on all the busbars now?
..and do you think they'll put sensors on all the busbars now?
Last edited:
Btw, most damage wasnt done by the bus bar "explosion":
At some point an electric arc was formed (not exactly sure where, and why, but my guess it was just induction from those collapsing magnet fields in combination with the severed shunt resistors). This burned right through the whole damn assembly (including the vacuum of the dewars, and the beamline).
They did have interlocks to prevent major damage, but they never accounted for the possibility of a vacuum leak that consists of liquid helium flowing through a fist-size hole. So they either didnt close quickly enough or were physically destroyed.
As a result, the (former) vacuum got severe overpressure (iirc something in the order of several atmospheres, which caused the most damage to the dipoles (yanked them out of their foundation). I think is was 20+ dipoles and a couple of quads that were affected that way.
Thanksfully, there are enough spares to build a small synchrotron light source by itself...
At some point an electric arc was formed (not exactly sure where, and why, but my guess it was just induction from those collapsing magnet fields in combination with the severed shunt resistors). This burned right through the whole damn assembly (including the vacuum of the dewars, and the beamline).
They did have interlocks to prevent major damage, but they never accounted for the possibility of a vacuum leak that consists of liquid helium flowing through a fist-size hole. So they either didnt close quickly enough or were physically destroyed.
As a result, the (former) vacuum got severe overpressure (iirc something in the order of several atmospheres, which caused the most damage to the dipoles (yanked them out of their foundation). I think is was 20+ dipoles and a couple of quads that were affected that way.
Thanksfully, there are enough spares to build a small synchrotron light source by itself...
js
Flashlight Enthusiast
IMSabbel,
Don Hartil, who is on the advisory committee, told me that they did NOT have monitoring on the bus bars to detect just such a failure as this. And he should know.
Perhaps we are misunderstanding each other?
Don Hartil, who is on the advisory committee, told me that they did NOT have monitoring on the bus bars to detect just such a failure as this. And he should know.
Perhaps we are misunderstanding each other?
