The Norwegian ATLAS project Faglig rapport August 31st 2003

1.0 Introduction

ATLAS is conceived as a general purpose instrument which should be able to exploit the full physics potential of the Large Hadron Collider (the LHC). The proton-proton collisions at the LHC will represent the luminosity and energy frontier in particle physics, and it is expected that studying these collisions will result in major advances in particle physics. It is the goal of this project to contribute substantially to the construction of ATLAS, and thus contribute to a timely installation and commissioning of the detector whichis expected to see its first collisions early 2007. This report concentrates on activities and progress made since last year. More extensive descriptions of the project are found in previous reports.

1.1The Norwegian ATLAS MoU, deliverables and background information.

The preparation for construction of the silicon detector system for ATLAS started in 1992 and was a natural development of the expertise and know-how built up for the DELPHI-experiment. During the first phase, prototyping and radiation hardness studies were the key issues, and a number of Cand. Scient. and PhD students have been working on these topics over the years. For the production phase, the work of the various groups and national responsibilities for detector component delivery are defined in the ATLAS Memorandum of Understanding (MoU). The MoU was signed in June 1998 by Norway on the basis of an agreement about the eight year project described in this status report. The Norwegian ATLAS project has three main components:

  1. Development, purchase and testing of silicon detectors for the ATLAS SCT. This is the most costly component. The Norwegian groups have purchased and tested 1950 silicon strip detectors, sufficient for the innermost of the four barrel cylinders (145 cm long and radius equal 30 cm) of the SCT. The detectors are developed for operation in a high radiation environment at bias voltages of up to 350V. The silicon detectors are discussed further in chapter 2.
  2. Construction of about 450 modules for the SCT, corresponding to all modules on the innermost barrel cylinder. This is a collaboration between the particle physics groups in Bergen,Oslo and Uppsala, who has formed ‘the Scandinavian cluster’ which is responsible for the delivery of the SCT modules. This is the most demanding task in the coming years in terms of competence, workload, manpower and infrastructure. The workload is shared as follows:

Bergen:Detector inspection and testing and delivery, Tests of complete modules

Oslo: Mounting of the four detectors on baseboards. Leakage current tests of the assembly, tests of complete modules.

Uppsala: mounting of hybrids, bonding of the modules, metrology of the modules, tests of compete modules.

See chapter 3 for details of the Norwegian commitments to module construction.

  1. The ATLAS Common Fund contribution. Most of this contribution is the delivery of Liquid Argon and Nitrogen tanks by SB-verksted (Industrial manufacturer of mechanical constructions) in the city of Drammen. The tanks were designed in collaboration with the NorwegianUniversity of Science and Technology (NTNU), Trondheim. The tanks are part of the cryogenics system for the ATLAS Liquid Argon Calorimeters. In terms of cost this is second largest Norwegian contribution. The Common Fund activities are discussed in chapter 4.

The ATLAS project has involved a substantial investment in facilities at the local institutes. The infrastructure upgrades include 2.5 Mkr in basic equipment and rooms, and 15-20 man-years of work funded by the Universities. Special construction equipment for around 1.5 Mkr has been purchased, covered by the Norwegian Research Council, over the last 5 years.

In addition, Norwegian groups are involved in the development and construction of the evaporative cooling system for the ATLAS ID. The commitment, as agreed in the MoU, amounts to covering 37% of the cost. Unfortunately, the development and production cost of the cooling system has been substantially underestimated. Furthermore, a PhD student is involved in the development of the DCS (Detector Control System) for the ID.

Finally, in collaboration with the “Norwegian Physics Analysis Project”, ATLAS relevant software development and physics studies are performed (see chapter 5). A Ph. D. student of the University of Bergen has been granted a stipend by this project to work with physics studies relevant for ATLAS, and in Oslo a PhD student has a university grant to perform such studies.In addition NFR has recently given a separate PhD. grant to a student who will participate in SCT module testing in addition to doing physics studies for ATLAS. Several Cand. Scient. students are also working on studies of the physics in ATLAS. Some of this activity is taking place in close collaboration with theorists affiliated to the ‘physics analysis project’, in particular with professor Per Osland from Bergen.

1.2 Financial overview.

The Norwegian ATLAS budget is intended cover costs specified in the Memorandum of Understanding (MoU), travels and operation, a minimal Dr.scient grant program and expenses for a Post. Doc. and for some manpower on the technical side. The silicon detectors are funded through a separate budget line (“Tungt utstyr”).

One problem is that the detector prototyping costs have been underestimated by 600-1000 kNOK. The most important overrun was caused by one extra prototyping run in 1998-99 for silicon detectors to test the radiation hardness of oxygenated silicon. This was advocated as very promising by the RD48 (ROSE) collaboration, and had to be taken very seriously. It was invested substantially in detector prototypes from SINTEF to do the necessary studies. Furthermore two preseries of detectors were ordered from SINTEF (80 and 50 detectors in each). Although the detectors were found to be of a high quality in general, the rate of out of spec. detectors was relatively high, and serial production detectors were ordered from Hamamatsu.

The two most cost intensive items are now paid from our CERN Core account. These are:

  1. Payments for all detectors delivered by SIntef and Hamatsu. SINTEF has received CHF 151767 for preseries detectors and Hamamatsu has received CHF 1326056 for the delivery of 1950 series detectors, resulting in a total cost of CHF 1477 823.
  2. The CORE account has been charged for a total of CHF 1438 157 for the four cryogenic tanks. This includes overruns due to the bankruptcy and restructuring of the company delivering the tanks in 2002.This in-kind contribution has been acknowledged by ATLAS as a contribution to the common fund of 1.15 MCHF. In addition, payments have been made to SINTEF/NTNU in Trondheim of NOK 794043 for their technical help within cryotechnology (valve design etc.) as part of the project follow up from our side. This follow up has been essential for the success of the project.

The total cost of the whole project is still not settled. It is clear that infrastructure and running costs for detector testing and module building has been much more costly than originally assumed. In addition number of financial uncertainties remain. These uncertainties are detailed below:

  1. Costs to construct the cooling system have increased very significantly.With a Norwegian commitment for 37% of the total cost, the present estimate is an overrun of about 2 MNOK).
  2. Detector testing: In our original planning, this should be performed by a technician employed by the University of Bergen. Unfortunately this person has deceased. For most of the work the project has had to finance a temporary replacement.
  3. In order to take part in SCT detector module testing inBergen, we have employed a research assistant to the end of 2003. We will prolong this to at least the end of 2004. It is absolutely necessary to do this work in order to deliver detector modules in a timely fashion to ATLAS.The person employed will enroll as a PhD student.
  4. Expenses in connection to the integration/installation of the Inner Detector have to be covered (C&I) . This may amount to a total of 2 MNOK.
  5. ATLAS has now started to collect funding for Maintenance and Operation (M&O). The cost for this for the period 2002-2007 is estimated at 508 kCHF with about 200 kCHF to be spent in the period 2002-2005.
  6. ATLAS overrun on Common Fund projects of the order of 10%. The Norwegian part to cover this would be a cash contribution of 312 kCHF (1.5 MNOK).

We do not expect to be able to recover significant amounts from the manpower budget any longer, as we now are in a period where manpower is very marginal compared to what is required to fulfill our obligations. Expenses have been close to the budgeting up to now, but a deficit has been accumulating. We have paid our CORE contributions in a timely fashion. This year we have to reduce our CORE payment compared to our original plan. Compared with our original MoU commitments in CHF this would not have been a problem., as we have been doing payments into CORE at favorable exchange rates. However , in view of the added ATLAS costs to completion and the collection of funding for M&O , as introduced and discussed in the RRBs 2001-2003,the overall budget seem to give a deficit. ATLAS has already granted us flexibility to do some of these payments in 2006, and hopefully new NFR projects for the period 2006 to 2011 can cover some of the added costs.

Planning / 1998 / 1999 / 2000 / 2001 / 2002 / 2003 / 2004 / 2005 / Total / Comment
Travel and operation / 1100 / 1100 / 1100 / 1100 / 1100 / 1100 / 1100 / 1100 / 8800
Project manpower / 570 / 850 / 850 / 850 / 850 / 850 / 850 / 530 / 6200
Ph.D grants / 170 / 510 / 680 / 680 / 510 / 510 / 510 / 170 / 3740 / 11 MY
Detectors / 5000 / 3800 / 8800 / MoU
ASICs, cooling, etc / 1350 / 2200 / 650 / 200 / 4400 / MoU
Common Fund / 3000 / 3000 / 800 / 1500 / 1500 / 9800 / MoU
Total / 41740

The table above shows the original funding of the project. The budget for grants and manpower has since been adjusted to cover inflation. In addition ATLAS-Norway has received a special contribution of 572000 CHF from CERN as payback of an old Norwegian loan to CERN.

2.0 Production and Testing of Silicon Detectors for the SCT

All the roughly 10000 detectors for the SCT barrel are delivered by Hamamatsu, including our order of 1950 detectors.In a joint effort with the university all infrastructure necessary to perform all relevant probing of the detectors has been acquired. This includes a cleanroom (class 10000), a computer operated probe station, microscopes capacitance metres, amperemetres and a switching system. Substantial expertise in the use of this equipment has been acquired by members of the group. All detectors have now been tested as described below.

2.2Detector testing during production.

Although equipped for extensive detector tests, acceptance tests of the serially produced detectors have been limited to a thorough visual inspection and leakage current measurements at different voltages. The quality of the detectors has been excellent and only a handful have been returned to the manufacturer for replacement. As a precaution, the leakage current of detectors have been tested up to 500V, and it is found that about 99 % of the detectors show no sign of breakdown up to this bias voltage. Out of the 1950 detectors received, only 16 had to be rejected after the inspection.

The figure above shows leakage currents at a bias voltage of 350V for allHamamatsudetectors measured. (Figure by L.G. Johansen)

- 1 -

3.0 SCT Modules Production

- 1 -

Preparations for assembly of detectors on baseboards at University of Oslohave taken place over the last 6 years. The equipment was bought in 1997 after several years of work to prepare a cleanroom for the module production. The ATLAS Barrel Modules are fully specified and consist of 4 silicon detectors, a TPG (thermal pyrolytic graphite with very high thermal conductivity) baseboard and a wrap-around double hybrid structure with 12 integrated readout circuits. The Scandinavian cluster receives fully populated, bonded and tested hybrids from Japan. This has significantly lessened the total workload originally envisaged for the cluster.

3.1 Recent developments

Here is a short summary of the developments over the last 12 months:

  • A total of 75 detector baseboard assemblies have been produced to the required mechanical precision of 5 microns.
  • In Uppsala 38 of these have been equipped with hybrids, bonded and tested.
  • At present, module production has halted as some changes have had to be made on the jigging for hybrid mounting in Uppsala. This is holding up ‘site qualification’, an internal SCT procedure established to ensure that the good quality of the modules throughout the production in the clusters.
  • VME systems in Oslo and Bergen for electrical readout are fully tested. Some work remains in order to perform complete module characterization in Oslo. In Bergen the systems has just been upgraded to be able to handle up to 5 detector modules in long term tests. To be able to test more than one module at the time, the power supplysystem needs to be expanded.

3.2 Module serial production and testing.

The flow of material is as follows :

  • The silicon detectors are tested in UiB and sent to UiO.
  • The TPG baseboardsare produced at CERN and delivered directly to UiO.
  • Four detectors and a baseboard are assembled to a detector-baseboard assembly at UiO, and then transported to Uppsala.
  • The hybrids, fully populated, bonded and tested are delivered to Uppsala from KEK (Japan).
  • Bonding the detectors in pairs, mounting and bonding of the hybrid takes place in Uppsala. The number of electrical bonds to be made are about 3100 for each module.
  • Finally, the modules are distributed for electrical testing at the three institutes in the cluster.
  • The modules will then be delivered to Oxford for mounting on the SCT barrel structure.

A procedure for aligning and mounting the modules has been developed during the past year to mount the detectors. Due to the precision required a number of realignment steps has to be foreseen, the mounting and gluing time per module-baseboard assembly typically requires five hours for two persons. After gluing the detectors on the baseboard, a total curing time of 33 hours is required, (30 hours at room temperature and 3 hours at 30 C. The current production plan is to complete 6 modules per week. This is feasible, but requires a very dedicated effot from the project members. Four jig-sets are now in operation. Tests made so far indicate that the construction task itself is not going to be the bottleneck but rather the electrical testing of hybrids and complete modules, and rework of faulty modules if needed. This is the reason for installing module test equipment in all three laboratories of the production cluster.

Atlas-Norway purchased already in 1997 the basic equipment for module construction. The infrastructure costs have been higher than foreseen in the original planning but has partly been absorbed by keeping one Post.Doc/Engineer position open when possible.

The Manpower for module production consists of:

  • Ole Dorholt, is working as technical coordinator for the Norwegian ATLAS activities, and leads the module construction and testing, and Torkjell Huse is employed as engineer at the project. Both these engineers have several years of experience in the SCT at this point. Dorholt has experience all the way back to work for DELPHI.
  • A third engineer, Kjell Martin Danielsen, is responsible, part time, for the database setup.
  • A dedicated technician, Anders Fylling, is hired for the module construction phase at UiO.
  • The mechanical workshop of UiO, lead by Finn Hostad has a long experience at CERN. The workshop was vital in the build-up of the clean-room and general infrastructure at the Institute as well as for the tooling for this project (precision jigs etc.).
  • In Bergen the module testing team consist of doctoral student Ola Øye (employed as a research assistant), with engineer Bjørn Pommersche and technician Karen M. Hovland. Logistics and coordination is taken care of by engineer Arne Solberg.
  • The overall quality control of the modules and reporting to the SCT community is the responsibility of Bjarne Stugu in cooperation with Richard Brenner from Uppsala.

This team, together with the physicists of the project should be sufficient to get the serial production and testing done.

The sequence of electrical tests is now established and necessary equipment and software

is in place. The picture above shows a prototype module under test in Bergen.

4.0 ATLAS Common Fund.

The ATLAS Common Fund contribution corresponds to 43% of the total Norwegian funding, as specified in the MoU. Funding shall be allocated to items as magnets, mechanical constructions, infrastructure cooling and ventilation, etc. The contributions can be provided in cash or as items delivered to ATLAS with predetermined specifications and price. A contract for the delivery of cryogenic tanks is placed in Norway as described below.

The remaining of the common fund commitment will be paid directly to the ATLAS common fund account.

4.1Norwegian delivery of cryogenic tanks

This project is now completed with success. The problems this project were facing last year have thus all been solved. We refer to last year’s report for an extensive description of these problems. The Norwegian company SB-verksted has delivered four cryogenic tanks for the ATLAS liquid argon calorimeter proximity cryogenics. They have been tested at CERN and found to satisfy all requirements, in particular the heat and gas leakage requirements which were found to be a problem of the first tank delivered. Due to the restructuring of the companies involved, the total price has been higher than anticipated. Some of these additional costs might be taken over centrally by ATLAS (82 kCHF), in case we do not find Norwegian funding by 2006.

The cryogenics system has to maintain uniform temperature conditions of the calorimeters which are immersed in liquid argon baths, by making use of saturated liquid nitrogen as cooling source. In addition to the temperature control the system should handle the purge, the cool down, the filling, the emptying and the warm up of the cryostats housing the detectors. SB-verksted has been in close contact with SINTEF-NTNU/Trondheim to ensure proper expertise for design and QA of the tanks, and the Norwegian ATLAS project has been financing this activity. Now SB-verksted has acquired the necessary competence to produce large liquefied gas tanks, and hopefully this will result in new contracts in the LNG market to the company.