Pisa april 2008 - We had at disposal a 7.23 Tesla superconducting magnet, part of a Varian VXR300 NMR spectrometer, irreversibly damaged by the impact with a nitrogen cylinder, brought too close to it. We decided to section the magnet, following the example presented in www.jeolusa.com (in this site, search for “Magnet Destruction”) and taking advantage of the schematic drawing of a similar Varian magnet kindly supported by engineer Vanni Piccinotti.
In order to have a look into the liquid nitrogen and liquid helium vessels, it was necessary to cut six windows (80cm x 30cm the widest one, 22cm x 20cm the smallest one) in as many aluminium layers, 2 – 5 mm thick. This task was performed by technician Paolo Nardini by means of an angle grinder: about 30 wheels were consumed! It required about 15 hours to complete the cutting and to smooth the metal edges (Fig. 1).

The direct inspection of the magnet inner structure stimulates observations and questions, related both to basic and applied chemical and physical subjects.
1) The external aluminium shell encloses the high vacuum chamber (p ~10-5 Pa), which can communicate only with the laboratory through the valve behind the magnet. This shell is sealed to both ends of the bore tube (that is the axial cylindrical channel which hosts the NMR probe). Of course, all this is at room temperature.
2) The liquid helium vessel, containing the superconducting coil, is located in the lower half of the magnet and can communicate only with the laboratory through two channels, one necessary for establishing the electric connections during magnet energizing and the other one for helium service (Fig. 2). When the magnet is operating, this aluminium vessel is at the liquid helium temperature (~ 4 K) .
3) The liquid nitrogen vessel is located in the upper half of the structure and can communicate only with the laboratory through a channel for nitrogen service. Its aluminium wall is connected to an aluminium shell which (non-hermetically!) encloses the underneath helium vessel (Fig. 3). This whole structure is at the liquid nitrogen temperature (~ 77 K) when the magnet is working.
4) Two more aluminium screens (non-hermetically sealed) are present (Fig. 4):
- the first is interposed between the external shell (~ 293 K) and the one at 77 K, in order to block the radiation from the hot layer inwards; its equilibrium temperature will be intermediate between the two mentioned above;
- the second screen is interposed between the one at 77 K and the helium vessel at 4 K, in order to block the radiation from the hot layer inwards; its equilibrium temperature is evaluated ~ 20 K.

Polished aluminium has very low emissivity, which further decreases with lowering temperature, and mainly reflects radiation. Therefore, the described structure efficiently limits the energy flow from the room to the liquid helium vessel and the helium boil off rate. (This “multi-shieldings” structure extends also around the bore tube).
Of coarse, the thermal contact between layers must be avoided as much as possible. This is ensured by the high vacuum condition between the external wall and the liquid vessels and by the magnet structure itself: each shell and vessel is steadily positioned inside the previous one by triplets of supporting cables fixed on their upper or lower parts (Fig. 5). The cables are made of a synthetic, thermally insulating material, with excellent mechanical properties even at quite low temperatures. It is easy to verify that the demise of this magnet was due to the strong impact, which deformed the external shell and put it into contact with the first internal screen (Fig. 6): this has reached the room temperature, at once and permanently, and, consequently, it has been impossible to cool down the magnet again.

Inside the liquid helium vessel, we can clearly see:
- The superconducting magnet coil, presumably composed of a Nb/Ti alloy, clearly coated by Cu (Fig. 7);
- Various shimming superconducting coils, necessary to optimize the magnet basic homogeneity. This is performed at the final stage of the energizing procedure (Fig. 8);
- The connectors necessary to supply power for energizing the magnet and the shimming coils;
- The load resistors necessary to protect the magnet from disastrous overheating, if a quenching occurs when the liquid helium level is to low.

This old, moderately low field magnet does not show any device for shielding the magnetic stray field, either passively (for instance by means of a steel shield) or actively (for instance by a peripheral coil powdered by reverse current).

Donata Catalano – 15 april 2008