Helping handles

The ambitious international project to build and operate the world’s largest nuclear fusion reactor, ITER, is in its earliest stages.

Intended to find out whether fusion is, as hoped, a practical source of energy for electricity generation, it is still at the site-clearing stage; the first buildings have barely started going up at Cadarache, southern France. However, development of the essential sub-systems for the reactor are well advanced.

In Finland, a team at VTT, the national technical research centre, is developing a system for one of the trickiest jobs in the plant’s maintenance — replacing the parts of the doughnut-shaped, or toroidal, reactor chamber that receive the greatest bombardment of sub-atomic particles from the plasma that undergoes fusion.

These parts are at the base of the toroid, in a circular trench-shaped structure called the diverter. It is here that the neutrons and helium nuclei generated by the fusion of deuterium and lithium end up, and where their excess energy is exhausted. Although there is still discussion over whether the sections of the diverter will be made of beryllium, like the rest of the plasma chamber, or of tungsten, it is certain that the constant bombardment with energetic particles will degrade the material.

The diverter will probably have to be replaced several times during ITER’s lifetime, perhaps three times in 20 years, and that will be true of any power stations that result from the project.

There is more involved than just pulling a component out and bolting another one in place. The plasma chamber will become radioactive during the facility’s operation, so the task must be carried out remotely.

The components that make up the diverter are big: the structure is made up of 54 separate cassettes, each mounted individually on rails inside the reactor vessel, and each cassette is 3.5m long, 2.1m high, and will weigh about 10 tonnes. Even more awkward, the cassettes will have to be removed, and the new ones moved into place, via maintenance tunnels underneath the toroid, but these will also house the systems — pipes for coolant, electrical cabling, and magnetising equipment for cooling the chamber. The space available is cramped, with only a millimetre of clearance available on either side of the cassette.

The research team led by Mikko Siuko has just taken delivery of the prototype of the maintenance robot. VTT’s laboratories at Tampere house full-size mock-ups of various parts of the ITER toroid, and these will be used to test the system, while also developing the control system and, eventually, training the future operators of ITER.

‘The aim is to develop the mechanics, the control system, the operator interfaces and tools so that, when the real test comes in 2018, we will have everything here ready to operate, and also experienced operators,’ said Siuko.

The maintenance robot, known as a CMM or cassette multifunctional mover, is a combination of several robots working together. The CMM is the carrier unit. Made mostly from stainless steel, it houses an electric motor-driven pinion system that moves the machine along a rack on rails within the maintenance tunnel.

On the top edge of the carrier is a robot arm, which can hold tools for locking and unlocking cassettes from their position within the plasma chamber. The front of the carrier can hold a variety of units called end-effectors, which are the parts that lock on to the cassette to move it in and out of position and carry it through the tunnels. These are made from cast aluminium. The entire system is about 4.5m long, with the effector plate, the part that locks on to the cassette, measuring 60cm x 60cm.

The entire system combines to make a robot arm that can move the effector plate through a wide range of motion. The movement of the robot’s parts is handled by hydraulics, using demineralised water as the working fluid. ‘Demineralised water isn’t activated under radiation, so we won’t be making more radioactive waste,’ said Siuko. ‘Also, it doesn’t ruin the environment inside the chamber if there’s a small water leak. If we were to get oil drops inside the reactor, we would have deep problems; but a little purified water won’t harm the reactor itself.’

Like most of today’s engineering systems, the CMM started off as a 3D simulation. ‘We had a 3D model of the diverter, which gave us all the geometry and the limitations of the space in which it has to work, and the components it will handle,’ said Siuko. ‘We then put some block models together to develop some ideas, to look at the mechanics and kinematics that we needed to confront. But the main thing was that, as early as possible, we made a numerical simulation model of the device, so that we could see all the flexibilities of the materials and the loads. that showed us where the critical point of the design was.’

It might seem computer simulation has advanced so far that it can be used as the single design tool for the process, but Siuko insisted this is not the case. ‘Virtual prototyping shows just the behaviour considered and modelled,’ he said. ‘Therefore, the real hardware is needed to expose the clearances, show the flexibility of structures, and to point out the weak points of the proposed design.’

There were considerable constraints in designing the CMM, said Siuko. Temperature, although critical to the design of the toroid, was not a problem; the reactor generates internal temperatures in excess of 150 millionºC, but by the time the maintenance operations are carried out, the plasma-facing surfaces have cooled to 50°-100°C. Radiation, however, is still a problem. ‘All plastic materials have to be protected from the radiation, and any semiconductors have to be closed off as well,’ said Siuko.

The project has two parts: design of the equipment and design of its control system. The latter task faces constraints that may be even more arduous than the mechanical problems. The space in which the CMM will work is so confined there is no room for cameras, so the operators will work blind.

‘This is the great advantage of creating the numerical simulation as early as possible,’ said Siuko. ‘It means that, even while the device itself is being manufactured, we can already be designing the control system.

‘The most difficult thing in operating the device is that the effector plate, the part that locks on to the cassette, goes into the chamber sideways and grabs this 10-tonne weight, then lifts it by one corner. There is no balancing weight, so it’s an uneven situation,’ said Siuko.

‘From a control point of view, when you start lifting and lowering the cassette there is a lot of flexibility, both in the robot and in the cassette itself. You have to know the position of the cassette very precisely, and you can’t see it. We were able to start working on this problem very early, even though we won’t start testing the system until next year.’

Siuko stresses that the device is a prototype, which might be revised considerably before the final robot is built.

‘At the moment, we don’t know how many robots ITER will need,’ he said. ‘We’re involved in the ITER design process, and we can see what kind of changes are made to the overall design via the PLM system. How many we need depends on how fast the device can remove and replace diverter cassettes. We’ll have about two months to replace the entire diverter, and we might need up to three robots to do it in that time.’

However, Siuko also believes VTT’s work could have applications in other industries. ‘What we’ve developed on the control side, and the user interfaces, will be useful in a variety of situations,’ he said.

‘Even in conventional power stations, there are complex structures that need to be maintained where there is limited visibility and no access for humans.’