by Corrado Gargiulo. Published: 05 December 2012

The fourth working group of the ITS upgrade brings together a diverse mix of physicists, engineers, technicians.

The main mission of our group is to design and construct the lightest possible mechanical support that will maintain the ITS silicon sensors in a very accurate position while at the same time it removes the heat dissipated by the front-end electronics. The ITS silicon sensors will be arranged in seven cylindrical concentric layers around the beam pipe. The segmentation of these layers in the azimuthal direction leads to the lowest level of modularity in the ITS: the stave.

The new ITS will consist of 7 layers of silicon sen¬sors supported by a ultra-light carbon fibre structure

The stave is the smallest operable part of the detector and for the three inner layers will be constituted of 9 pixel silicon sensors (30 mm x 15 mm, 50 µm thick) aligned with micrometric precision and served by an electrical bus that will be glued on a support structure with 30 cm length and 1.5 cm width.

For these inner staves the minimum amount of structural material to place in front of the sensor is defined by a material budget fixed as low as 0.3% X0 per layer. Meeting this requirement was one of the main challenges that our group had to confront.

The new ITS Inner barrel: sketch of the building blocks constituting a generic stave (right) and front view of their positioning around the beam pipe to form the Inner Barrel (left).

Another demanding request to our group was to locate innermost staves and sensors of the ITS at a minimum distance of 2 mm from the beam pipe wall. Placing the silicon detector so close to the fragile beryllium beam pipe was a delicate task for the mechanical design and integration.

This task was made even more challenging by the requirement of having a layout that would allow rapid access to the ITS modules during the yearly LHC shutdown. This meant that the new ITS should be translated approximately 3 meters along the beam pipe in order to be easily accessible which also means that the new ITS should be moved out of the inner bore of the TPC in which the current ITS is placed.

Following the above requirements we tried to identify and characterize different materials, processes and cooling technologies that would be suitable for the construction of the new ITS.

First, we tested 50 micron thick silicon dummies to get a better sense of how to handle such a thin sensor. We then tried to tune the design of the stave structures on these new findings and looked for the lightest and stiff materials like carbon fiber reinforced plastic which are currently used in aerospace.

Handling a dummy silicon 50 µm thick

We worked to minimize the structural material in front of the sensors in order to fulfill the tight requirement on the material budget. For that purpose we started with designing a thin wall structure and we ended in drawing a line to trace the direction of one carbon fiber wound around a mandrel in few turns. The result was a stiff structure as light as 0.6 grams over a length of 30 cm and 1.5 cm in width.

We then started to couple the structure with a cooling system to remove the heat dissipated by the silicon sensors. We glued a cooling pipe, embedded in the carbon fibers, running along the entire stave length and back which would enable us to get in and out at the same side of the detector where access is provided. For that purpose we had to use tubes with a very small diameter (almost 1.5 mm), a minimum tube wall thickness (35 µm) while at the same time we had to maximize the contact area with the carbon fibers which curry away the heat (dissipated at a rate of 0.3-0.5 W/cm2) from the silicon sensors to the pipe. In order to reduce the material in front of the sensors we tested smaller cooling tubes and considered various micro-technologies like micro-channels (200 µm x 200÷800 µm) dug into different substrates, polyimide and silicon. The stave structure complete with the cooling system resulted to weigh between 1,5 and 2 grams.

Stave prototype with embedded cooling system for the ITS Inner Barrel

In order to validate the different mechanical and cooling solutions we decided to glue on the stave the thin silicon sensors (dummies), aligned at high accuracy, and we added heaters which were simulating the dissipated power. The extensive thermal tests carried out on these samples allowed to optimize the cooling design and come up with validated structures.

Of course this was just one step as the ITS inner layers consists of 48 staves. That means that once we had a solid verified design for one stave we had to think how they could be grouped together in order to form a cylindrical layer. We built the three inner layers and, through light composite end-wheels, we assembled them to constitute the ITS Inner Barrel. The Inner Barrel prototype gave us the opportunity to go through the whole production and integration process and resulted in an extremely useful structure to handle, rotate, mount and dismount, to be used with our colleagues of the other groups to discuss and decide on services and integration.

ITS staves assembled in three layers to form the Inner Barrel prototype

Following the work done for the inner three layers of the upgraded ITS we are now looking in the design of the four external layers. The different requirements in term of the sensors that will be used (pixel or strips), the material budget (0.8 % X0 per layer ) and the required length (stave length up to 1,5 m) lead us to explore new design routes albeit different from those discussed for the inner silicon detectors of the ITS.

Finally, although we developed the ITS design starting from the stave, the lower level of modularity, we had to consider the whole ALICE detection system and how the ITS would be integrated into that. We developed clear design principles about the installation of the new ITS in ALICE and how to gain a fast access to it. These principles were taken into account in designing the different levels of the ITS going down to the level of single staves. We foresaw a detector barrel divided in two halves to get around the beam pipe. A rails system guides the two halves of the detector in and out-of ALICE for about 3 meters, closing them against the beam pipe, where a gap of only 2mm is left.

Over the next few months our group will have an intense design, prototyping and testing activity. Specifically we will be optimizing the design and production of the inner layers, we will develop a design for the external layers and we will validate and detail the ITS integration process into ALICE. Of course all these wouldn’t have been possible without the hard work of the members of our group and I would like to deeply thank all of them: Pieter, Sergey, Yapp, Yannick, Thomas, Claudio, Andrea, Alessandro, Paolo, Manuel, Enrico, Jose, Michele, Cosimo, Irene, Arturo and Romualdo. It has been a real pleasure working with them and we feel excited watching the progress made thus far.