[This blog covers my experience of repairing a key magnet coil without which machine would have been dead and all due to a photo I happened to notice somewhere without searching for it!]
I had taken over as Project Leader of ADITYA Tokamak(A nuclear fusion reactor) at Institute for Plasma Research(IPR), Gandhinagar in early 1996. See the related blogs at :
http://ranjan.in/my-research-journey-to-brain/
http://ranjan.in/5-days-miracle-in-nuclear-fusion-reactoraditya-tokamak/
Tokamak consists of many magnetic coils of large size to create the right configuration of magnetic field to trap very very high temperature plasma away from any material body. One of the very critical magnet coils in ADITYA was TR-1. This was in the form of a large solenoid with weight of about 1.5 tons and length of more than 100 cm. This is placed in the center of machine (in fact the first magnet coil to be placed) at the time of assembly. Slowly the rest of the machine is assembled around it. It is not visible after assembly of the machine as it gets embedded inside the inner hole of the vacuum vessel.
TR-1 was designed to hold 20 KV of insulation but after assembly process somehow it was found to have insulation failure around 7.5 to 8.0 KV. Many days it would fail even earlier. It depended very critically on the environmental conditions of temperature and humidity. As ADITYA hall was very big, air conditioning system used to take time to cool down the hall as well as bring down humidity levels for it to become operational. It was daily practice to put on A/C machine and wait for couple of hours before right levels of temperature and humidity levels were achieved. At this stage an insulation test was carried out to find out if we can go ahead with machine operations. Sometimes especially during rainy seasons, it would almost be impossible to operate the machine due to this weakness. This also hampered the full range of operation of machine as we could not use power supply to its full designed values. We had learnt to live within the limits as it was not felt feasible to try to take it out and repair it.
I believe it was some time in late 1997 or early 1998, in less than two years of my taking over as Project leader, that we found that TR-1 insulation was failing at much lower value of 2.5 to 3.0 KV. We waited for several days trying to hope that somehow insulation would improve but nothing of that kind happened. Finally we had to realize that this was not going to improve by itself and we need to do something to get it back on track. I was not in IPR during the time of assembly of machine during 1986-89 period and so we had meeting with senior scientists who were involved with design and assembly of machine.
Standard process of repairing TR-1 was to remove it and get it repaired or replaced. For this one had to remove it out of the machine. For this one had to disassemble the machine fully to access it. One of the senior scientist remarked that this whole process of disassembly and assembly would take at least 2 years. Another remarked that once you disassemble it, it would not be simply be possible to reassemble it in the first place since it was so difficult to assemble it the first time itself. I was very disappointed and sad that we found ourselves in such a situation of hopelessness and were looking at a dead machine!
In the next couple of days, I happened to walk in to the office of one of the senior scientist’s office casually for some conversation. I noticed in one corner of the room, a photograph from the time of assembly of the machine pinned on a soft board and not visible prominently. This “photo” is shown on the left side here. An idea came in my mind after seeing this photo about how I could remove TR-1 without disassembling the whole machine.
As seen in photo, TR-1 was in cylindrical shape with two metal cap at each end. These metal caps were connected to each other by a number of steel rods. On top cap, there were hooks to move this assembly by crane. Vacuum vessel was tightly surrounding the cylinder with very little gap. So it was not possible to take out the whole assembly. After seeing this photo, an idea came in my mind. Can we remove the top cap by removing the tie rods connecting the two caps? In that case TR-1 cylinder becomes free to take it out. However there was nothing to hold it with since hooks were part of top cap!
So now the problem was to remove the 1.5 ton cylinder without anything to hold it from. Inner part was hollow like in a solenoid. So I thought of expanding something inside it so that it presses against the inner walls and then using frictional force to lift it out, at least partly, so that there was a gap between bottom cap and cylinder to insert some kind of hook at the bottom to take it out fully. Whatever expands inside should have good friction coefficient so that it does not slide off while we try to grip it from inside when trying to lift it. Any slip can either make it fall after a while damazing it or it may not come out at all.
So I thought of using vehicle brake shoes! Since they have good friction to stop the vehicle in speed. I requested one Technical Asst to go to the market and see if he can find a vehicle brake shoe, whose outer diameter matched closely with the inner diameter of TR-1 to get maximum surface contact. He went to the market area of Ahmedabad dealing with vehicle spare parts. One vendor said he can supply one and it would cost Rs 800. He purchased the pair of brake shoes. After purchasing the vendor mentioned that this is brake shoe of Tata Pickup 407 and it costs only Rs 400. He charged double due to knowledge and if he had told him the vehicle model, he could have purchased the same from other shops ! I was fine with that respecting knowledge!
Now we needed to put something to expand and make the brake shoe expand inside TR-1 cylinder. We used a Jackscrew and welded the brake shoe to the two sides of it as shown in the picture. Now we could hang it inside the TR-1 and use rod to rotate the inner expansion assembly and let it press it against the inner wall of TR-1. However we did not know how much to expand. Too much expansion would apply too much force and possibly crack the TR-1 copper tubes. Too less pressure and it would come loose. We did not have scientific way of judging this!
Anyway we decided to try our luck! Many of us went to the top of machine including myself. I always like to be in the front line! The tie rods were removed one by one. I think there were six of them. After that the top cap was removed.
While we were expanding the jackscrew, I heard a crackling sound! My heart skipped a beat! I asked them to stop fearing damage to the magnet coil. We had made arrangements so that when we tried to lift the TR-1 out, some one from bottom would insert spacers between bottom cap and cylinder. This was to ensure that if the cylinder slips back, it would not fall down. So we used the crane to lift it a bit – it worked to pull it and after 3-4 mm lift, we inserted spacers at the bottom as per plan. We moved up further to about 10 mm and again added additional spacers. Slowly we created enough gap that, we could insert hooks to the bottom of cylinder and take it out and carry it down to the floor.
Finally we could repair the TR-1 magnet coil and tested it to full insulation of 20 KV by removing certain defects (probably it was some sharp corner!). This not only brought ADITYA Tokamak back from near death but it also allowed us to operate it to full operating parameters. As we started going up in parameters, we faced other strange problems and solved them as well. I would share those in subsequent blogs.
Common sense and presence of mind helps in scientific missions also where textbook knowledge fails us sometimes!