3.2.2 Plans for next 6 months... 19

Transcription

1 REPORT TO THE STFC OVERSIGHT COMMITTEE ON THE 24 TH APRIL 2009 ON ATLAS UK RESEARCH AND DEVELOPMENT TOWARDS THE REPLACEMENT TRACKER FOR A HIGH LUMINOSITY LHC UPGRADE University of Birmingham, University of Cambridge; University of Glasgow; Lancaster University; University of Liverpool; Queen Mary, University of London; Royal Holloway, University of London; University College London; University of Manchester; University of Oxford; Rutherford Appleton Laboratory; University of Sheffield Table of Contents 1. Executive summary International Context Overview of UK Contributions International Priorities for the Next 12 Months Calls upon contingency Notes on Reporting Actions from Previous OsC Meeting Action Action Action Action Action Action Action Work package summaries WP Commentary on project to date Plans for next 6 months Summary of major deliverables WP1 financial statements Resource Usage WP1 milestone plan Gantt chart WP2: Radiation backgrounds measurements and simulations Commentary on project to date Plans for next 6 months Summary of major deliverables WP2 Financial statement Resource Usage WP2 milestone plan Gantt chart WP Commentary on project to date Plans for next 6 months Summary of major deliverables and their financial allocation WP3 Financial Statement Resource Usage

2 3.3.6 WP3 milestone plan Gantt chart WP4 Microelectronics and Interconnect Research and Development Commentary on project to date Plans for next 6 months WP4 financial statements Summary of major deliverables and their financial allocation WP4 milestone plan Gantt chart WP Scope of WP Commentary on project Oct 2008 Apr Plans for next 6 months WP5 financial statements Summary of major deliverables and their financial allocation Resource Usage WP5 milestone plan Gantt chart WP Commentary on project to date Plans for next 6 months WP6 financial statements Summary of major deliverables and their financial allocation Resource Usage WP6 milestone plan Gantt chart WP Commentary on project to date Plans for next 6 months Summary of major deliverables and their financial allocation WP7 financial statements Resource Usage WP7 milestone plan Gantt chart WP8: Data Acquisition and Read out Summary on project to date Plans for next 6 months WP8 financial statements Summary of major deliverables and their financial Resource Usage WP8 milestone plan Gantt chart Commentary on finances UK Global Deliverables Risk Register

3 1. EXECUTIVE SUMMARY 1.1 International Context The planning for the upgrade programme is determined by a number of factors, not least the timing of the LHC ramp up and the likely dates when physics priorities and real running experience (including accurate dose predictions) are likely to be available. The current thinking in ATLAS is to aim for a Letter of Intent (LoI) of about 100 pages for 2010, linked with an interim Memorandum of Understanding (MoU) with funding agencies, with planning for both now started in earnest. This is to be followed by the Technical Proposal (TP) for the Upgrade programme in 2011/12 linked to the SLHC discussion/approval now expected at about that time. In parallel, Technical Design Reports (TDRs) for sub systems will be developed with some, such as that for the tracker upgrade, needed on about the same timescale to allow adequate time for procurement and construction. Clearly final MoUs need to be negotiated and responsibilities agreed before large item procurement (such as sensors or ASICs) can proceed, and this process may define the schedule. From a machine point of view, a date of 2017/18 for the long shutdown is now regarded as the most realistic. The management structure for the tracker upgrade is also in transition, with the recognition that individual component R&D is coming to successful completion and module prototyping and demonstrators for the different aspects of the tracker stave taking priority. The UK has been key in driving this transition, producing the first demonstrator modules and much of the detailed mechanical and electrical design needed for first stave production. The central role of the UK has very recently been further recognised by inclusion of the PI in the overall ATLAS Upgrade Steering Group with specific responsibilities for the micro strip tracker. 1.2 Overview of UK Contributions Since the last meeting in October, two further international meeting of all the ATLAS Upgrade activities have taken place at NIKHEF 1 and CERN 2, as well as a major meeting on micro strip and pixel mechanics at Berkeley 3. The UK typically contributes ⅙ th of the presentations at ATLAS Upgrade workshops and for these last two, at NIKHEF and CERN, also contributed the tracker upgrade overview talks. Simulation work has come to the forefront, with concern that the proposed layouts may not have sufficiently robust pattern recognition, possibly leading to a greater need for pixel (as opposed to short strip) layers at higher radii and reinforcing the interest within the UK for this technology as part of the proposed next phase programme. An outline of planning from 2010 to 2013 has been presented as Statement of Interest to PPAN s meeting of 27/28 November 2008 and with formal feedback encouraging us to proceed to full proposal from STFC on 19 th January Radiation simulation remains a critical issue with many systems aspects such as cooling and mechanical design, dependent on accurate predictions. These clearly require input from the radiation fields measured when first collisions take place (and additional steps to extrapolate from 10 TeV data) as well as more detailed modelling of the likely accelerator configuration, shielding and detector layout. UK groups have also continued to dominate the studies of irradiated detectors, including studies to doses well

4 beyond that expected for the strip tracker region at the SLHC. Crucially, first results for miniature Hamamatsu detectors irradiated with glue on their surface show no additional degradation and studies with mixed neutron and proton irradiation to the fluences expected at the inner layer of the micro strip tracker also reveal no unpleasant surprises. Undoubtedly one of the major highlights of the ATLAS Upgrade meeting at CERN was the detailed evaluation of the new ABCN μm CMOS read out chip and the very fast implementation of a 20 chip hybrid and first demonstrator module with two hybrids bonded to a full size Hamamatsu SLHC prototype sensor. This programme, from the first read out of ABCN25, to the hybrid design, fabrication, assembly and evaluation, to the first module assembly and test, to the entire read out chain from which all these results (and the only results on ABCN25 to date) were obtained was entirely based in the UK. The different UK work packages worked seamlessly together to produce and demonstrate the very first ATLAS SLHC prototype module and we are now supplying hybrids and DAQ across the collaboration so that colleagues can begin their own module prototyping and testing programmes. As described in our last submission, our experience for the current ATLAS tracker, was that the 6 chip per side ABCD hybrids took 5 iterations to achieve the required noise performance and stability. That the ABCN25 first prototype hybrids easily meet noise specifications, provides great confidence that hybrids designed to match the details of the electrical and mechanical interfaces of the first full length staves should also work as required. The success of the glued samples also builds confidence that the programme to develop novel interconnect technologies and effectively process the hybrid circuit directly onto the sensor, should also be radiation hard. The first wafers post processed with layers that correspond to those required for such an approach have been returned and are under detailed evaluation. New (temperature dependent) effects on the radiation hardness of the optical fibres have been identified by the UK and we are focussing on understanding these. This unexpected and highly important result necessitates a more extensive radiation programme than previously planned, leading to our first call on contingency. This comes about as the allocated budget for WP5 of 29k is insufficient for the additional irradiation campaign needed to properly explore this effect. This additional programme is clearly vital, since the operating temperature in ATLAS will be significantly below room temperature and the problems look to become more pronounced as the temperature is reduced. Powering remains a critical issue internationally, and with the ABCN25 and SPi ASICs a major programme of multi module tests is anticipated to meet one of the key deliverables of the UK programme. While DC DC conversion alternatives to serial powering continue to develop elsewhere, none are yet close to the current and voltage step down capacities needed for this application and all are a long way from being suitable for evaluation from a circuit noise perspective. For this, and first phase electrical stave studies, DAQ systems able to handle many hybrids reading out together need to be developed, along with supporting the initial DAQ developments which have so successfully allowed the UK to dominate the first ABCN25 studies. Finally, the UK has fully embraced the stave concept, having played a central role in providing evidence to the review which led to its adoption as the baseline design. The UK has continued to provide key FEA and prototyping studies to understand many of the concerns with the approach in terms of differential expansion and the thermal integrity of the proposed structure. This work has led to many improvements which were the subject of the Berkeley workshop, which led to the UK activities becoming much more integrated with 4

5 the US programme. Our work on the electrical tapes, which interface neatly with the hybrid design activity, has led to us taking international responsibility for this area and we continue to lead the cooling programme, which is common between the pixels and the strips. Indeed, with the micro strip tracker adopting mechanics much closer to that proposed for the pixels, the two programmes have greatly increased their commonality and our engineering programme provides crucial inputs used in both. The reworking of the management structures within the ATLAS Tracker Upgrade programme is also attempting to recognise the much greater level of integration now possibly between the pixel and micro strip subsystems. 1.3 International Priorities for the Next 12 Months The completion of layout studies and development of a robust tracker design from both pattern recognition and radiation tolerance perspectives is crucial to the completion of any Technical Proposal. Furthermore, the module programme needs to advance in two directions: 1) towards proving the individual modules in terms of radiation hardness, thermal cycling and robustness against noise pick up and 2) demonstration of multi module systems, both linked together with serial powering and operated in close proximity on a stave like section with appropriate hybrid and bus tape configurations. In all cases, the concerns are increased noise, mechanical failure, connector or electrical failure, loss of thermal performance or loss of sensor performance. Both these activities will be central to the UK programme over the coming period, with these multi module demonstrator systems planned for assembly and investigation at UK institutes. A major programme of irradiation of large area objects, which requires scanning modules or stave sections for weeks in proton beams at the CERN PS or Karlsruhe cyclotron, is now envisaged to which the UK is also contributing its expertise from carrying out a similar programme for the current SCT modules. The above areas are expected to provide the most significant international programmes on the upgrade tracker in 2009/10. In parallel, the international programme envisages several thermal/mechanical stave prototypes to test issues of mechanical stability, thermal and mechanical integrity after thermal cycling, distortions and vibrational analysis. The delivery of dummy sensors to the collaboration by Hamamatsu should allow rapid progress in development of staves with increasing mechanical complexity and realism, building towards objects with full cooling and realistic heater elements that can be put through a much more rigorous test regime than would be dared with the first electrical stave. Both the above programmes are expected to see a major contribution from the UK, with the module and prototyping work making UK institutes a natural place to carry out much of the necessary construction and investigations. The nature of the programme and the distribution of components mean that multiple prototype systems for different studies are to be anticipated internationally, and some of these may well pursue options which are not favoured by us, with a view to having sensible comparisons at the time of any final review. However, the parallel programme of producing the first electrical stave will be managed differently. This was always expected to prove difficult to achieve within the initial UK R&D programme, but the better than expected results with ASICs, hybrids and the first module have led to colleagues in the international community wanting to press ahead quickly with the first stave electrical demonstrator. We are arguing for this project to be based at CERN, allowing maximum continental European collaborative effort to be brought to support US 5

6 and UK personnel. It is likely that many of the modules, the tapes and possibly the carbon fibre facing and sandwich honeycomb for the stave could be provided, or at least prepared, by the UK. We would almost certainly also play a significant role in the DAQ, with US colleagues, for this and all the other multi module systems. 1.4 Calls upon contingency As discussed above and in the WP5 section, there is a need for more capital within WP5 to pay for the additional costs of the irradiation campaign needed to fully explore the temperature dependence of the fibre behaviour after SLHC doses. The collaboration requests that the committee consider our proposal for 20.5k of additional capital funds from the project contingency for WP5 to pay for the additional irradiations of fibres at differing temperatures corresponding to those anticipated in the experiment. It is not expected that this will incur any additional manpower costs. No other activity looks to require such a call on contingency with progress towards key deliverables matching the revised schedules presented at the last meeting. 1.5 Notes on Reporting 1. In the financial tables we have not split out the new money from that in the Rolling Grants, because the only way to make these numbers meaningful would be to do the same exercise for the 18 months already reported. After discussions with STFC we concluded that the effort to generate this information was not worth the gain it would provide. 2. A table of deliverables has been introduced for each work package. This has 3 columns, one for the deliverable, one the value of the deliverable, and one for the project phase. The project phase describes if this deliverable is likely to fall within the original grant period or overlap with the next phase which will be the subject of a proposal this Summer, following the positive reception by PPAN of the SoI. In the case of work packages 1 and 2 no value has been declared for these as they do not really have defined hardware deliverables. 3. As requested at the last meeting we have included travel into the cost breakdowns. The collaboration believes the numbers presented here are correct, however the breakdown between work packages is almost certainly not accurate. This is due to approximately 20% of the personnel working on more than one work package, and their travelling thus invariably being across several work packages. The impact of this is worsened by these personnel being the heaviest travellers (PI and project engineer for instance). 4. When studying the critical path information in the milestone tables only items upon the critical path have been identified, items not upon the critical path have been left blank, this does not indicate that they have been missed. 6

7 2. ACTIONS FROM PREVIOUS OSC MEETING Below is the list of actions given to the collaboration by the committee following the previous meeting. Following each, the collaboration provides its response or a reference to the response should it be detailed later in the document. 2.1 Action 1 All milestones should be consistent and relevant, should specifically relate to the allocated funding, and should include an overall milestone list, milestones for the previous period and milestones for the coming period. Within each work package there should be consistency between milestones shown in the sub tables. A narration on any changes to the milestones should also be included. The Collaboration should define the critical path and indicate the impact on the critical path where appropriate. The Committee also wanted Table 6.2, Overall Milestone List, in WP6 to be reformatted and resubmitted by 15 th January Action: Collaboration The collaboration refers the committee to the relevant work package sections, where we believe we have improved the quality of the milestones, as requested. 2.2 Action 2 The Collaboration should re issue the risk register as soon as possible focusing on risks to the project, not to the schedule. The Collaboration should also consider the impact of failure of one element of the UK contribution to the entire UK contribution. The revised risk register should also be submitted to the Office by 15 th January Action: Collaboration Completed as requested on 15 th January The revised version is referenced from this document in section Action 3 In light of the changes to WP5 (Opto electronics), this work package should be revised in its entirety by the next meeting, including costings and a breakdown of effort. The Collaboration should also provide an organogram in the next report showing the management of WP7. Action: Collaboration WP5 revision presented below. WP7 organogram is included in WP7 reporting (section 3.7) 2.4 Action 4 The Collaboration should give their best estimate of the effort breakdown of the international programme by responsibility, if not by finances, with some indication of the competitive element. This could be represented as a responsibility table showing percentage UK contribution and should be submitted to the Office by 15 th January

8 Action: Collaboration Completed as requested upon 15 th January Action 5 The Committee informed the Collaboration that they were broadly sympathetic to the Collaboration s desire to have one Oversight Committee for all ATLAS upgrades but would need to discuss this in more detail as and when appropriate based on future proposals. The Committee advised the Collaboration that they could create some Working Allowance out of their existing award but that there would be no access to the Contingency unless a need arose and a case could be created for this. Potentially the Collaboration could bid to use the Contingency to finish the project if there was a cost over run in order to make a clear distinction between the R&D and next phase. The Office discussed the discrepancies in the finance tables with the Collaboration. The Collaboration agreed to separate out rolling and standard grant effort per work package in future finance tables. Action: Collaboration After the last meeting the collaboration considered how best to respond to this action. This was taken forward and discussed with STFC and it was concluded that for this action to be useful it would need to retrospectively break the data to date down into RG and new money. This was concluded to be an unnecessary burden to place upon the collaboration. Within this document you will find the costs described as before, however, to give some frame of reference to the amount of new spend within each work package, a breakdown is given for each on the proportion of rolling grant and new money making up the workpackage. The collaboration apologises to the committee for not realising the difficulties in executing this action at the last meeting. 2.6 Action 6 The deliverables per work package should be identified for this project, matched to financial allocation and should allow clear distinction between the R&D phase of this project and the next phase. Action: Collaboration This information can be found within each section below in tabular form. However, the distinction will always be a rather artificial one for most activities since this is an R&D not a construction programme. Even in the proposal, items were identified that have long lead times and need to be purchased during this period to be available without long delay for the next phase. Most activities have therefore been defined as terminating on 31 st March 2010, except where posts started late and run on or where the activities feed into the electrical stave programme completion and testing at CERN. However, the posts in this programme 8

9 were assumed in the SoI to be core to the next, pre production phase, of the tracker upgrade project starting 1 st April 2010 and funding for these is bid for in the participating institute Rolling Grants. Clearly therefore, the deliberations of the PPGP will have impact on the scope of the future programme, complicating the timescales for a sensible proposal to be prepared. 2.7 Action 7 The Committee thanked the Collaboration for their comprehensive report. The Collaboration agreed that as they would be reporting on a six month period for the second meeting that their next report would be shorter than the first one had been. The Committee also discussed the format of and attendance at future meetings with the Collaboration. At future meetings, the Collaboration should be represented by the PI, Chair of the ATLAS UK Collaboration Board and Project Engineer in addition to two Work package Coordinators of their choice unless the agenda of the meeting dictated otherwise and/or a specific need was to be addressed. The next meeting should include a general presentation based on the report lasting no more than 30 minutes, giving key updates from the preceding six months and projections for the next six months. Milestones and finance tables should be included. The Collaboration should also give a presentation on design interfaces to include power, number of channels, serial powering, adequate use of copper, data handling, cooling and mechanical issues, outlining how the interface worked, who was involved and the potential impact. Action: Collaboration The collaboration presents this second report as an update which is significantly shorter than the last, 132 page, presentation. 9

10 3. WORK PACKAGE SUMMARIES 3.1 WP Commentary on project to date In the last six months, the software for simulation and performance studies of upgrade layouts has progressed from essentially a development prototype, to become a tool used by a number of groups for different upgrade related studies. A pixel system simulation has been implemented along the same lines as the Geant4 based strip simulation, which has been changed to reflect the fixed barrel lengths of the current strawman layout, as shown in Figure 1. The standard ATLAS tracking package has been adapted to work with the upgrade simulation, and performance studies for tracks in the barrel are in progress previous crashes were due to a mixture of bugs and memory limitations. At the present time, the pixel end caps can be simulated but not analyzed due to a current limitation in the standard software infrastructure. Figure 1. Left: fixed length barrel layout strawman from November 2008 Tracker Upgrade Workshop. Right: end view visualization of the barrels as implemented in the Geant4 simulation. Common upgrade related software and configuration files are now maintained in a shared area to facilitate participation by the growing international community. Layouts have been implemented which are variations of the current strawman; the current ATLAS silicon detector layout has also been implemented with corresponding simplifications to the material in order to facilitate comparisons between upgrade and current simulations, and, eventually, to data when it becomes available. The amount of material in the simplified simulation has been checked against detailed accounts for the strip barrel, and comparisons in other parts of the simulation are underway. Simulated data files, incorporating single and 400 pileup events per bunch crossing, have been generated, run through reconstruction, and shared for performance studies. Resources and methods for grid production have been prepared. In addition, fully simulated minimum bias events have been used as a starting point for Level 1 trigger studies, which use a histogram method in the strip layers for finding high pt tracks (compared to the standard tracking package, which uses combinatorial and Kalman filter methods, and extend to much lower ptʹs). 10

11 The first reconstruction results, using Geant4 and the older, Geant3 based ATLSIM package, were presented at the ATLAS Tracker Upgrade Workshop in November 2008, and highlighted the stark challenges faced by ATLAS in the SLHC environment. With standard track quality criteria, the efficiency for high pt muons does not appear to decrease appreciably with pileup, but the fake rate rises dramatically. The latter effect can be inferred from the top curve of Figure 2 (right), which counts the number of reconstructed tracks as a function of the number of pileup events for the standard threshold of 7 hits: since the number of primary particles (and consequent secondaries) rises linearly, any rise faster than linear rise can be attributed to fakes. At the Tracker Upgrade Workshop, it became a concern that pattern recognition may be impossible in the SLHC environment with any conventional layout. (Figure 2 also shows the rise in CPU consumption of the standard ATLAS track reconstruction algorithm as pileup is increased; the same behaviour is seen in both full and fast simulation, and presents its own practical difficulties.) time (mins) cpu time/event vs pileup mean pileup Figure 2 Left: CPU time for ID track reconstruction vs pileup, with a transverse momentum threshold of 5 GeV/c. Right: number of reconstructed tracks vs pileup, for different requirements on the number of hits included in the fit. These results shifted the focus of the upgrade simulation programme, from completing the simulation in all its aspects and proceeding as rapidly as possible to physics simulations, to understanding pattern recognition itself in greater detail. Subsequent studies in the barrel have suggested that tighter requirements, in particular on the number and type of hits on the track, may be able to reduce the fake rate to a manageable level with a small loss of efficiency in finding primary tracks. The lower curves of Figure 2 illustrate the effect of increasing the requirement on the number of hits. Related studies have been carried out using the current ATLAS simulation with pileup increased to comparable occupancies, showing similar results. Most of these preliminary studies, however, do not incorporate realistic hit inefficiencies yet, and moreover consider neither track environments such as jets, nor important secondary particles such as conversion electrons. The preliminary studies were discussed extensively during a workshop held in Oxford on January 2009, gathering experts across the simulation, reconstruction, and tracking performance communities. Along with stimulating increased collaboration among the widely dispersed, expanding community of experts, the workshop resulted in common 11

12 definitions to be used for tracking efficiencies and fake rates, and an evolving discussion document which prioritizes the issues to be addressed and the plan for doing so. Progress on these studies was reported at the ATLAS Upgrade Week of February In addition, the upgrade simulation effort has been integrated more closely into the ATLAS experimentʹs software effort, with members joining the new Simulation Working Group and Layout Task Force. Goals and concerns from previous period, WP 1 Goals Status Produce sizable Monte Carlo samples of simulated pileup, and some signal, for general use. Investigate the feasibility of physics analysis with an ATLAS upgrade. Concerns Resource competition with commissioning and early data analysis. Performance and robustness of reconstruction software. Some pre mixed samples have been produced already and shared widely in the upgrade and trigger simulation communities for performance studies. Concerns raised about pattern recognition shifted these studies to increase focus on tracking efficiency and fake rate. Status Deferred until LHC resumes in late Track reconstruction software is now working, though with a pt threshold of 1 GeV/c. Concerns about pattern recognition in the ID being addressed. Full detector simulations not yet run Plans for next 6 months Upgrade simulation activity will be directed, as noted above, to understanding pattern recognition behaviour in a more general sense than evaluating one particular strawman layout. The initial focus for these studies will be the barrel, where tracking is relatively well understood in the current ATLAS detector. As the pixel endcaps are implemented (which requires a low level software infrastructure fix, the best solution for which is currently under discussion), these studies will be extended to larger η. Material profiles will be refined for the barrel and endcaps, in particular the transition region between them, in light of current material budgets. The fast simulation package, Atlfast IIF, which has also been adapted for upgrade use and shares infrastructure with the full simulation, will be compared with the full simulation in preparation for studies demanding larger samples. A number of technical improvements are needed for tracking performance software, as well as consideration of tracks in light and b quark jets, and important secondary particles. The main questions identified at the Oxford workshop concern whether there is need for additional high granularity layers (and where they might be, if so), the role of pixels and strips in pattern recognition and track parameter resolution, and the role of different types of hits or holes in recognizing a good track, especially in light of realistic hit efficiencies. Addressing these questions should enable the upgrade community to develop an 12

13 understanding of pattern recognition behaviour which will guide further evolution of the layout in the face of realities of engineering and, when available, actual LHC data. Full simulations of the complete upgraded detector are still required to evaluate physics performance and provide feedback for LHC machine planning. It appears likely that some level of optimization or region selection will be required in order to run these simulations on a normal grid connected computer. Certain aspects of the current ATLAS simulation, such as the cavern background, also remain under developed. The new Simulation Group will help involve the relevant experts in order to address these shortcomings and increase the usability of the software for both current ATLAS and upgrade studies. The work of upgrade simulation will be an important input for the Layout Task Force, which was started during the February 2009 ATLAS Upgrade Week and will begin work in the next month. This task force is expected to bring together experts from the simulation as well as detector and engineering communities to produce an overall upgrade layout, for the whole detector, suitable for a Letter of Intent in It is anticipated that the work of WP1 will adapt accordingly. Goals and concerns for next six months, WP1 Goals Complete pixel endcap implementation for simulation and tracking, and validate overall ID software. Check material profile in simulation against current ID. Study general pattern recognition behaviour over a range of layout options. Physics simulations with full detector. Concerns CPU time and memory consumption may exceed resources available on grid nodes, slowing progress on sample generation and reconstruction. Fix to basic software infrastructure to enable endcap use may trigger a cascade of related fixes Summary of major deliverables Deliverable and description Completed software for simulating, digitizing, and reconstructing events in a range of potential upgrade layouts, including pixels and strips, and barrel and endcaps. A range of layout specifications for evaluation using the above software. The specification data should be in a form suitable for archiving and sharing among research groups. Simulated data samples with at least several of the layouts specified above. The samples should be shareable among research groups. Report on performance evaluations relevant to designing the upgraded Inner Detector. Project phase R&D R&D R&D R&D 13

14 3.1.4 WP1 financial statements Approved (excluding contingency) Transfers Actual spend to Feb 09 Projected spend this year Actual spend (2+3) Projected spend (2+4+5) Actual (6-1- 1a) (1) (1a) 2007/ /2009 (3) (4) 2009/ /11 (6) (7) University Staff Effort Costs* Liverpool Sheffield Oxford Lancaster University Sub-Total STFC Lab Costs RAL PPD RAL Technology STFC Lab Sub-Total Equipment Other Directly Allocated costs (eg consumables) Travel Total (Excluding VAT and WA) Working allowance VAT WP1: Finance Summary (all figures in k) Actual spend in previous years (2) Current year 2008/09 Latest estimate of future requirement (5) Total Variance Projected (7-1-1a) Total (including VAT & WA) Contingency (Held by STFC) 1 Excluding Working Allowance and VAT * The University staff effort recorded in this table should be the 80% amount STFC pays, including academic time Use of columns: (1) = The amount approved by STFC (1a) = This column should be used to show any virements between headings, for example when Working Allowance is used, the amount should appear as a debit in the WA row and then credited to the relevant row (2) = The actual spend in previous financial years, by year (3) = The actual spend in the current financial year up to the most recent quarter (4) = The total projected spend for the current financial year, including any expenditure so far (ie actual spend this year plus predictions of remaining spend this year) (5) = Projected spend for the remaining years (6) = The actual spend so far (7) = Projected spend over the whole duration of the project (ie actual spend so far plus predictions of remaining spend to project completion) The variance columns show the difference between the actual and projected amounts and the approved amount. 14

15 3.1.5 Resource Usage Table of staff %age FTE WP1 WP1 2007/ / / /11 NAME INSTITUTE Allport P P Liverpool Vossebeld, J Liverpool 10 Tovey Sheffield 10 A.N.Other Nicolas Sheffield 70 A.N.Other Nicolas Sheffield 70 Abdesselam (SLHC) Oxford Tseng Oxford Roger Jones Lancaster Peter Love Lancaster Stephen Haywood RAL 5 5 Total Grand Total WP1 milestone plan Table 1.1: Milestones achieved in the last six months Milestone Work Milestone Target Date Status No. Package M1.3 WP1 Simplified geometry schema design Feb 07 Update: strips done previously; pixel barrel now done, with pixel endcap soon M1.6 WP1 Strawman layout and initial mods implemented in simplified geometry Apr 07 As in M1.3, now including pixel barrel, and pixel endcap soon Table 1.2: Overall Milestone List for WP1 Milestone No. Work Package Milestone As at Sep 08 As at Mar 09 (Changes in Bold) Delay due to UK Other Collaborators Affects Critical Path? See Note Reason for change M1.1 WP1 Software infrastructure Nov 06 Nov 06 modifications M1.2 WP1 Initial document of Feb 07 Feb 07 design constraints M1.3 WP1 Simplified geometry Feb 07 Feb 07 schema design M1.4 WP1 Initial generator level Feb 07 N/A Monte Carlo samples M1.5 WP1 Hard coded geometry Apr 07 Apr 07 15

16 in main software release removed M1.6 WP1 Strawman layout and initial mods implemented in simplified geometry M1.7 WP1 Pattern recognition software for tracking adapted for new layouts M1.8 WP1 First version of layout evaluation software used on initial layouts M1.9 WP1 Further studies using cosmic ray or combined test beam data M1.10 WP1 Quarterly reports to ATLAS upgrade management M1.11 WP1 Material studies with 900 GeV data M1.12 WP1 Initial studies of minbias/jet data with 14 TeV M1.13 WP1 Quarterly report on relevant physics topics to ATLAS upgrade management Apr 07 Apr 07 Jun 07 Apr 08 Y Y Jun 07 Jun 07 Y N Sep 07 N/A Sep 07 Jun 08 Mar 10 N Y LHC delayed Jan 09 Oct 10 N Y LHC delayed Apr Gantt chart 16

17 3.2 WP2: Radiation backgrounds measurements and simulations Commentary on project to date The WP2 programme is separated into two main areas, measurements at the LHC, and simulations for the SLHC. The main priority 6 months ago was to be ready for radiation background data taking with first collisions. The WP2 post doc participated in the preparations to read out various radiation detectors and monitoring systems. This involved operation shift training. Unfortunately the early stoppage of the LHC meant no collision data being taken, so the focus switched back to simulations. The first task was the completion and writing up of the poly moderator study [1]. This work was presented at the Amsterdam Tracker Upgrade Workshop in November Discussions at the workshop led to the request for further simulations to be performed to include additional engineering considerations. One of the conclusions of the poly moderator study was that the preferred design from an installation perspective would result in a radiation fluence increase of ~30%, compared to the original proposal. The decision on whether this is acceptable or not has yet to be taken. More recently, effort has gone into updating the FLUKA geometry and material description. It had been hoped to find a way of converting automatically the very detailed ATLAS detector description (described in GeoModel) into the format used by FLUKA, but this has proved non trivial. Furthermore, it seems the full ATLAS simulation does not have a complete description of the various shielding elements, which are important for radiation background. These issues have been recognised by ATLAS management, and a new group has been set up to push this forward. This is partly motivated by the fact that there are large uncertainties in the cavern background fluence predictions, and until these are understood better, the scale of the muon system upgrade cannot be evaluated. An additional request from the Amsterdam workshop, not planned in the scope of the original WP2 proposal, was an assessment of the radiation backgrounds for the proposed insertable b layer (IBL) in phase I of the upgrade. These simulations required zooming in on in regions around the interaction point, and looking at fluences and doses with a finer level of detail than done before. This work was reported on in a recent ATLAS Upgrade Week [2], and some results are shown below. 17

18 It can be seen that for fluences in the pixel region, there is little z dependence, and that the impact of a Gaussian bunch profile compared with a single collision point is negligible. We also show that for small radii, the impact of the rest of ATLAS is small, and the fluences are, as expected, dominated by particles directly from the interaction region. Finally, there have been strong requests for our participation in beam loss related simulations. The WP2 post doc has experience in this area from the Tevatron collider at Fermilab, so the request for our involvement by the ATLAS community is natural. However, this work is not covered in the original WP2 proposal, so our level of participation is limited. The WP2 post doc has been working with a N. Mokhov, a recognised world expert in experiment and machine background simulations [3]. [1] Neutron Moderator Studies for the ATLAS Inner Tracker Upgrade, EDMS Id [2] Updated IBL radiation estimates, [3] Progress report on Mars implementation of ATLAS cavern, Goals and concerns from previous period, WP 2 Goals Status Produce preliminary set of radiation Completed background estimates New JM design and revised inner Completed being written up tracker fluences Machine interface studies Completed Concerns Status Delays in post doc appointment Project delayed ~6 months 18

19 3.2.2 Plans for next 6 months A major task over the next 6 months will be working with the ATLAS simulation group in order to produce an updated and accurate FLUKA and G4 geometry and material description. These geometries will be simplified versions of the detailed GeoModel description, which would then be used as a basis for comparisons with measured data. It is also being proposed to benchmark the G4 predictions with the FLUKA results, which has more confidence in the low energy regime and for nuclear physics. It should be noted that for low energy neutron and photon simulations, the fluences are sensitive to material composition, so it is important to model this as accurately as possible. Recently there has been a lot of effort to measure the material mass and content in ATLAS, which has been implemented in GeoModel. Because LHC start up is now planned for 10TeV centre of mass energy, and not 14TeV, additional simulations will have to be performed to take this into account. Continue collaboration on machine background studies. This work impacts most of the detector subsystems, including the inner detector, so understanding the consequences is important. Because of delays to LHC data, we will also attempt to initiate earlier than planned the simulation deliverable 2.3. Goals and concerns for next six months, WP2 Goals Update FLUKA geometry and material description to be in line with ATLAS Athena GeoModel details. Repeat FLUKA simulations for 10TeV. Collaborate and complete machine background studies. Ensure readiness of data taking machinery for October 2009 Concerns Not clear whether new geometry description should be done automatically or by hand (former obviously preferred)? Summary of major deliverables Deliverable and description Preparations for data taking, implementing detailed geometry and material description Beam line activation design Radiation monitoring data taking and analysis Minimum bias measurements Publish analyses of data Project phase Current Future Future Future Future 19

20 3.2.4 WP2 Financial statement Approved (excluding contingency) Transfers Actual spend to Feb 09 Projected spend this year Actual spend (2+3) Projected spend (2+4+5) Actual (6-1- 1a) (1) (1a) 2007/ /2009 (3) (4) 2009/ /11 (6) (7) University Staff Effort Costs* Sheffield University Sub-Total STFC Lab Costs RAL PPD RAL Technology STFC Lab Sub-Total Equipment Other Directly Allocated costs (eg consumables) Travel Total (Excluding VAT and WA) Working allowance VAT WP2: Finance Summary (all figures in k) Actual spend in previous years (2) Current year 2008/09 Latest estimate of future requirement (5) Total Variance Projected (7-1-1a) Total (including VAT & WA) Contingency (Held by STFC) 1 Excluding Working Allowance and VAT * The University staff effort recorded in this table should be the 80% amount STFC pays, including academic time Use of columns: (1) = The amount approved by STFC (1a) = This column should be used to show any virements between headings, for example when Working Allowance is used, the amount should appear as a debit in the WA row and then credited to the relevant row (2) = The actual spend in previous financial years, by year (3) = The actual spend in the current financial year up to the most recent quarter (4) = The total projected spend for the current financial year, including any expenditure so far (ie actual spend this year plus predictions of remaining spend this year) (5) = Projected spend for the remaining years (6) = The actual spend so far (7) = Projected spend over the whole duration of the project (ie actual spend so far plus predictions of remaining spend to project completion) The variance columns show the difference between the actual and projected amounts and the approved amount. 20

21 3.2.5 Resource Usage Table of staff %age FTE WP2 WP2 NAME INSTITUTE Paganis Sheffield Dawson Sheffield A.N.Other Majewski Sheffield 17 A.N.Other Nicolas Sheffield Total Grand Total / / / / WP2 milestone plan Table 2.1: Milestone achieved in last six months Milestone Work Milestone Target Date Status No. Package M2.5 WP2 Complete Athena G4 inner detector radiation background predictions. Dec 08 Complete Table 2.2 Overall Milestone List for WP2 Milestone No. Work Package Milestone As at Sep 08 As at Mar 09 (Changes in Bold) Delay due to UK Other Collaborators Affects Critical Path? See Note Reason for change M2.1 WP2 Preliminary set of inner tracker fluence predictions M2.2 WP2 New JM design and revised fluence predictions M2.3 WP2 Low activation beamline design M2.4 WP2 Ensure SLHC solutions do not compromise inner tracker program M2.5 WP2 Complete Athena G4 radiation background predictions M2.6 WP2 Publish analysis of LHC data M2.7 WP2 Benchmark codes using LHC data Jan 07 Jan 07 Sep 08 Sep 08 Nov 10 Dec 09 Jun 08 Jun 08 Dec 08 Dec 08 Mar 10 Nov 10 LHC Delayed Jun 10 Dec 10 LHC Delayed 21

22 3.2.7 Gantt chart Unchanged since #1 (3 rd Quarter 2008) 22

23 3.3 WP Commentary on project to date Miniature detector characterisation programme The irradiation campaign of miniature detectors with different strip isolation technologies (p spray and p stop) has been continued with reactor neutrons, 26 MeV and 23 GeV proton irradiations (at the research reactor of Ljubljana, the cyclotron of the University of Karlsruhe and the CERN PS facilities, respectively). The measurement of the charge collection efficiency at various equivalent doses and with different particles has confirmed that the results after irradiation of the Hamamatsu (HPK) sensors are in line with the comparable amount of data collected with detectors produced by Micron Semiconductor within the framework of the RD50 experiment. These studies have been pioneered for several years by UK groups and are now been repeated and confirmed by other institutes from Europe, US and Japan. For the first time, one UK group has pushed the charge collection measurements to neq cm 2, well into the region of the SLHC pixel B layer (possibly located at 3.7 cm from the beam line).figure 3 shows the collected charge as a function of the bias voltage (CC(V)) characteristics after this unprecedented dose. Figure 3. Collected charge as a function of the bias voltage for 140μm and 300μm thick detectors after 1.5 and neq cm 2. These are the recognised reference for assessing the radiation hardness properties of segmented planar detectors. One UK institute has also performed the first measurements (and the only available to date) of charge collection efficiency with miniature strip sensors after mixed irradiations. The mixed irradiation (up to 1x10 15 neq cm 2 ) has been performed with neutrons first and then protons in equal doses, to mimic the result of the radiation environment for strip sensors in the SLHC, where the damage will be produced by a mixture of fast neutrons and charged hadrons. Figure 4 shows the CC(V) characteristics of micro strip sensors irradiated to an integrated dose of neq cm 2 (ATLAS Upgrade qualification dose for micro strips), with reactor neutron only and with an equal dose of reactor neutron and 26MeV protons. It can be seen that the same CC(V) is measured with the baseline sensor electrode configuration n implant strips in both cases. It should also be noted that this performance is adequate to guarantee full tracking efficiency at 500V. A slightly better performance is obtained with a difference substrate (n type MCz): but this material requires a much more expensive processing (double sided lithography, not 23

24 available, for example, with HPK) for producing the sensors. While this may be of interest for the pixel layers, where the scale of the silicon order is more commensurate with the capabilities of the more niche European suppliers, its use is not required for the CC(V) performance needed for the micro strips where the p type sensors are perfectly adequate and offer significantly reduced cost. Figure 4 CC(V) characteristics p FZ and n MCz micro strip sensors irradiated to an integrated dose of neq cm 2 with reactor neutron only and with an equal dose of reactor neutron and 26MeV protons (mixed). The ratio of the two fluxes depends on the radius, but an almost equal dose of neutron and charged particles is expected at the innermost (and therefore most exposed) layer of strip sensors. This measurement is the best reference to the specific needs of micro strip detectors for the ATLAS upgrade. Also, a significant programme to characterise important operational parameters, like the inter strip resistance, inter strip capacitance and punch through voltage protection (PTV) against high ionisation events, has been carried out and completed with all the different types of geometries included in the HPK mask set. These measurements are important to optimise processing parameters like p stop and p spray doses, geometry of the PTV structures in p stop devices etc. The measurement will be repeated to check these processing parameters with the new improved mask set. They will inform the decisions on the optimal geometrical parameters for the final production sensors. The UK institutes will play a major role in this process. Characterisation of sensors for the WP4 interconnection programme Several 4 wafers with 1x1 cm 2 micro strip detectors from Micron have been provided for deposition of metal routing on BCB isolation for R&D on advanced interconnects method for WP4. These wafers have been returned from ACREO after the BCB deposition and metal patterning and after the initial verification of connectivity, they have been characterised with standard micro strip detector measurements. The results of IV, CV and inter strip capacitance measurements on un diced and diced non irradiated detectors have shown encouraging results. This will allow the UK institutes to proceed to the next crucial part of the programme, namely the irradiation with x rays and protons. The results of this programme will explore if this interconnectivity method can be suitable for hadron collider detector applications. 24

25 Full size detector measurements Five large area (98mm x 98mm, 320mu thickness) HPK sensors (with 1280 n strips in 4 segments, high R p type substrate, p stop isolation) have been delivered to the UK (3 as part of the pre series order and 2 sent from US for further investigation). IV, CV, inter strip resistance (ir) and capacitance (ic) and full strip tests have been performed on all of them. The CV, ir and ic measurements all exhibited the expected performance. It was known from HPK measurements that the IV characteristics of these devices were not satisfactory for ATLAS upgrade requirements, with a breakdown voltage below 600V in many cases. This result led to a modification of the mask set for the production of the remaining pre series detectors, entailing a delay relative to the original programme, of more than 1 year. First results from the company with the new designs and processing look good now to ~1000V. The full strip tests on large area sensors delivered to the UK confirmed the very high strip yield of the devices, with a total result close to 100% (the worst sensor has 99.54% of functioning strips, with 3 devices reaching 100%). One large area sensor has been used for the build of the first ATLAS module: two hybrids with 20 ASIC s chip each have been attached to the detector (one bridge and one directly glued on the sensor). The results of this activity are reported in the WP4 section. From these results for the initial 3 UK sensors (which suffered lower than acceptable breakdown) but show good strip yield and unchanged behaviour after hybrid gluing) and the HPK results on the corner runs and March 09 pre series, we conclude that the sensors being delivered in August 09 should be suitable for the first electrical stave programme and multi module tests. However, the stave requires 24 sensors + spares and so the first electrical stave at CERN will be built in collaboration with the rest of ATLAS to ensure adequate components. Original HPK full size sensors delivery schedule Actual and anticipated HPK full size sensors delivery schedule Pre series Split 1 Split 3 Pre series Corner runs Date Dec 2007 Spring 2008 # of sensors Action Spring 2008 April (low breakdown) Preseries August 2009 Preproduction Mask set design Preproduction Mask set design Mask set redesign February 2009 July Preseries March 2009 Date End 2008 End 2009 Sensor Order Spring 2009 February 2010 Date # of sensors 24 Table 1 Explanation of current Hamamatsu (HPK) delivery plans to the UK. Note the global order has been for 132 devices in total. 25

26 The low breakdown voltage of the pre series sensors obliged a revision of the mask design and processing to recover optimal performance. This redesign was tested by HPK with a processing run called corner run with differing isolation implants explored. Three sensors from this run were released without fees to the UK outside the present commercial order with HPK. All showed excellent breakdown characteristics when measured by HPK. In Table 1 the evolution of the definitions of the different sensor deliveries are outlined. Note that it is now assumed that what HPK call the pre series (as shown in the table) will be used for first module and stave prototyping, whereas the pre production order which will need to see minor mask alterations to be stave compatible (one set for axial and one for stereo sensors), is assumed as a final deliverable of this phase of the project to allow preproduction phase module and stave prototyping and qualification to proceed without 6 months design and production delay. Studies of gluing of miniature detectors An important issue for module construction is to establish if the hybrid flex circuit hosting the front end readout electronics can be directly glued onto the silicon detector surface or if it needs a bridge support structure. The test of the effect of the glue on the detector surface in changing the electrical properties of the devices is a first important step in this direction. Two types of glue have been used for the study: Araldite epoxy 2011 (used in the construction of ATLAS SCT) and Fuller Epolite electronics grade epoxy 5313 (used in CDF Run II). The glue was applied within the active area, over the surface of the sensors, without spreading over the guard rings. The electrical properties (IV, CV, ic and ir) were measured before application and after application and curing. A serious degradation of the electrical properties has been found (in term of reduction of the breakdown voltage and of the ir) after Araldite epoxy gluing, while stable properties have been measured when Fuller Epolite was used. A further step was taken, using the Epolite glue to attach a layer of kapton and a layer of kapton with aluminium on two microstrip detectors to mimic the attachment of a real hybrid. Similar satisfactory results have been obtained after electrical characterisation of these devices. Also charge collection measurements were in line with those for unglued sensors. Two Epolite glued miniature sensors were irradiated at the CERN PS to neq cm 2. After irradiation they have been characterised by electrical and charge collection measurements. No changes in the breakdown performance are observed after irradiation with the glued detectors (up to 1000V). The reverse currents (measured at 25 o C) are consistent with the received fluence. No measurable effects from the epoxy on the surface have been seen from any electrical measurements. The charge collection measurements also show no measurable effect of glue relative to similar irradiations, both in the collected charge as a function of the bias voltage or in the cluster size and shape. Figure 5 shows the spectrum of the charge deposited by a minimum ionising particle on an epolite glued detector, with a piece of alumina kapton circuit biased to 1000V. No effect is seen by comparison to an un glued sample. In conclusion, the epolite glue to surface appears to introduce no further degradation effects after proton irradiations, giving strong support to pursue the direct gluing of the electronics circuits on detectors for the WP4 programme. 26

27 Figure 5 Energy spectrum (from minimum ionising particles) measured with an epolite glued detector, with a piece of alumina kapton circuit (picture on the left) biased to 1000V. No effect is seen in comparison to an un glued sample. Irradiation of full-size sensors, modules and mechanics There is a clear need to irradiate full size sensors under realistic conditions (applied bias and with an electric field across the AC oxide) to qualify the devices. Also detectors with powered hybrids will have to be qualified during and after irradiation. The dimension of the full size samples (10x10cm 2 detectors + mounting frames + services) do not allow irradiations in the available neutron facilities traditionally used. Also, high energy protons are needed to efficiently irradiate through the hybrid. CERN PS is the suitable facility for this type of irradiation. In fact, it has been efficiently used for the irradiation program of the present SCT, where a 2d scanning stage was used to cover the 6x6cm 2 area of the devices to the required final fluence of 2x10 14 neq cm 2. On the other hand, it will not be possible to use the same irradiation method as in the past, with the sensor plane orthogonal to the 1x1 cm 2 beam spot. The larger size of our present samples and the higher qualification fluence of 10x10 14 neq cm 2 prevents the use the previous set up due to the insufficient intensity of the CERN PS beam (it would require up to 400 days of irradiation to reach the target fluence scanning over the 100 cm 2 surface). A new irradiation set up is being designed. The new ATLAS SCT Upgrade irradiation box will be installed in the CERN PS IRRAD 5 facility located in the east hall which consists of a scanning table (X Y parallel bedways) placed on SKF telescopic foot (Z with 400mm of adjustment), a rotary motor and a motor controlling the inclination for axis alignments. The samples will require significant cooling (to ~ 10 o C) for irradiation under bias (both detectors and hybrids). Preliminary work for providing the mechanical supports and to explore the sample cooling dynamics is taking place in the UK (several restrictions are imposed on the possible mechanical mounting by the need of demount for measurements after irradiations, inclined mount for optimal exposure to the beam, limitation of the possible materials used for support and cooling due to the high radiation environment, etc). This activity is performed in close collaboration with the CERN person in charge for the irradiation zone. Goals and concerns from previous period, WP 3 Goals Status Complete electrical characterisation and QA tests Completed on the 5 received of full size detectors. sensors 27

28 Prepare for irradiation of some full size detectors. Perform systematic irradiation of miniature sensors to confirm the radiation hardness of the substrate. Concerns Availability of analogue read out systems for signal and noise studies. HPK delivery schedule. Ongoing activity Ongoing activity Status A new system is now available (Alibava) 18 months delay Plans for next 6 months The delivery of 6 p stop full size HPK ATLAS07 detectors that was expected in December 2008 has been further delayed to end of March The remaining part of this order consists of 18 main sensors (4 p stop, 14 p spray) expected in August The delay was due to a correction of the mask set that has been used to process the first batch of sensors (including the 3 p spray + p stop sensors delivered to the UK in February 2008). This first batch of pre series detectors showed electrical performances (namely a breakdown voltage as low as V) at the limit or below the ATLAS quality specifications. A mask correction was implemented with the advice and agreement of ATLAS experts from the international community, including members of this work package. Some (four) of the detectors produced by HPK for qualifying the new mask set have been released in February 2009 to the UK institutes outside the present commercial order, with no extra cost for WP3. The batch to be delivered in March 2009 (and part of the UK order to HPK) will consist of six large area detectors with common p stop strip isolation. The priority will be to carry out the full qualification programme with IV, CV, VRT, IR, IC and strip quality measurements. The qualification programme will be repeated in different UK institutes to ensure readiness for future large volume delivery. An irradiation programme of some of the large area devices will take place when the cool irradiation box for the CERN PS will be ready. The plan (to be agreed between the ATLAS UK institutes and CERN) is to install the irradiation facility by Summer Also, irradiation of miniature detectors from the same wafers of the full size sensors will continue to confirm the properties of the substrates are consistent with the earlier batches. In particular, 76 miniature detectors and 25 monitor diodes have been delivered in February They will be irradiated with neutrons and protons to different neq doses (0.15, 20, 50 and 100x 14 cm 2 ). International planning is expected to begin around the pre production order of sensors with the same technology as proven for the pre series, but with mask topology changes made to give stave specific strip orientations and layout. 28

29 Goals and concerns for next six months, WP 3 Goals Characterisation of 6 p stop large area detectors delivered in March 2009 (pre series order) Characterisation of 4 p stop and p spray large area detectors delivered in February 2009 (HPK runs, delivered free, outside of UK order) Make some of these devices available for module programme, after testing Irradiation and test of the various geometry miniature detectors made with the corrected mask set Preparation and installation in the CERN PS Irrad 5 facility of the ATLAS Full Size detector cool box. Reception (August 2009) of the last 18 pre series full size sensors from HPK Concerns Further delays from HPK Possible problems for accessing the CERN Irrad 5 facility with delays of the full size irradiation programme Summary of major deliverables and their financial allocation Deliverable and description Financial allocation from WP3 (kgbp) Project phase Provision of sensors to WP4 interconnect programme Investigation and choice of substrate type in view of adequate radiation hardness 15 RD50 wafers post processed 48 Current Choice of optimal interstrip isolation geometry 48 (this activity overlaps with the previous one) Current Set up for full size device and module irradiation 12 (+ 48 as above) Current Order of 24 pre production sensors 49 Future Production of detector specification document for market survey and tendering Summary of overall activity (no direct cost) Future WP3 Financial Statement The predicted spend of capital in 2010/11 is 49k. This may be inconsistent with the manpower allocation of 0 in this period but is explained by the long lead time upon the pre series sensors. 29

30 Approved (excluding contingency) Transfers Actual spend to Feb 09 Projected spend this year Actual spend (2+3) (1) (1a) 2007/ /2009 (3) (4) 2009/ /11 (6) (7) University Staff Effort Costs* Liverpool Cambridge QMUL Glasgow Lancaster University Sub-Total STFC Lab Costs RAL PPD RAL Technology STFC Lab Sub-Total Equipment Other Directly Allocated costs (eg consumables) Travel Total (Excluding VAT and WA) Working allowance VAT WP3: Finance Summary (all figures in k) Actual spend in previous years (2) Current year 2008/09 Latest estimate of future requirement (5) Total Projected spend (2+4+5) Actual (6-1- 1a) Variance Projected (7-1-1a) Total (including VAT & WA) Contingency (Held by STFC) 1 Excluding Working Allowance and VAT * The University staff effort recorded in this table should be the 80% amount STFC pays, including academic time Use of columns: (1) = The amount approved by STFC (1a) = This column should be used to show any virements between headings, for example when Working Allowance is used, the amount should appear as a debit in the WA row and then credited to the relevant row (2) = The actual spend in previous financial years, by year (3) = The actual spend in the current financial year up to the most recent quarter (4) = The total projected spend for the current financial year, including any expenditure so far (ie actual spend this year plus predictions of remaining spend this year) (5) = Projected spend for the remaining years (6) = The actual spend so far (7) = Projected spend over the whole duration of the project (ie actual spend so far plus predictions of remaining spend to project completion) The variance columns show the difference between the actual and projected amounts and the approved amount. 30

31 3.3.5 Resource Usage Table of staff %age FTE WP3 WP3 NAME INSTITUTE Allport P P Liverpool Affolder A A Liverpool Casse G L Liverpool Dervan P Liverpool Tikkanen T - left July 08 Liverpool 10 3 Tsurin (replaced Turner) Liverpool Whitley M Liverpool Wormald M P Liverpool Carter Cambridge Hommels Cambridge Robinson Cambridge Sigurdsson Cambridge Beck QMUL C.Buttar Glasgow R.Bates Glasgow L.Eklund Glasgow A.Cheplakov Glasgow F McEwan Glasgow F. Doherty Glasgow Harald Fox Lancaster Alex Chilingarov Lancaster John Statter Lancaster Total Grand Total / / / / WP3 milestone plan Table 3.1: Milestones achieved in the last six months Milestone Work Milestone Target Date Status No. Package M3.11 WP3 6 Sensors being shipped March 09 Partial Delivery The situation is in line with the previously stated milestone plan. Monitoring the smaller tasks within the work package we feel that we are on target to meet the next and subsequent milestones based on the pre series sensors playing the role of pre production ones. Table 3.2: Overall Milestone List for WP3 Milestone No. Work Package Milestone As at Sep 08 As at Mar 09 (Changes in Bold) Delay due to UK Other Collaborators Affects Critical Path? See Note Reason for change M3.1 WP3 01/07 order RD50 sensors Not funded Not funded 31

32 M3.2 WP3 Complete mask design for full size SCT wafers (for production with HPK) Completed July 2007 Delay in early studies for finalising the design M3.4 WP3 Delivery of HPK first 3 large area and several miniature sensors. Completed February HPK Mask mistake from HPK. Redesign for large area electrical sensors. These will be used for initial electrical stave programme M3.5 WP3 Delivery of mechanical sensors Completed July 2008 HPK Priority given to the production of electrical sensors. Delay on these has affected the delivery of the mechanical detectors. M3.6 WP3 Completion of irradiation program with RD50 sensors Completed June 2008 Not funded. Usage of detectors from other sources (RD50). M3.7 WP3 Decision on optimal substrate type Completed July 2008 Performed using Japan funded ATLAS06 miniature sensors. Distribution of these devices and technical time for irradiations led to completion in Summer 08. M3.8 WP3 Complete mask design for pre series detectors and place order. June 2009 Completed July 2008 Redefinition of sensors planned for first electrical stave as those from the preseries run as defined by HPK. M3.9 WP3 Complete mask design for WP4 interconnect studies Completed Feb 2008 (using copy of existing masks) M3.10 WP3 Delivery of sensors for WP4 interconnect studies Completed April 2008 M3.11 WP3 Delivery of preseries sensors December 2009 March 2009 (6 pieces), August 2009 (18 pieces) Yes The completion of the order of the 27 preproduction sensors (that in the original plan would have informed the second 32

33 M3.12 WP3 Delivery of preproduction sensors. December 2009 part of the order of a further 24 sensors) was originally due at the end of 2007). These first pre series sensors are now to be used for the module and stave programmes. Summer 2010 N HPK Yes Delay in HPK This second part of the HPK order (further 24 detectors for UK) might be revisited in the light of the status of the overall programme M3.13 WP3 Test of pre series sensors completed and delivered for stave programme 03/09 Mid to end 2009 M3.14 WP3 First irradiation tests with full size sensors (pre series instead of the delayed preproduction) July 2010 End of summer 2009, but with preseries sensors instead of preproduction (delayed by at least 12 months) N HPK Initially expected with full size preproduction sensors, now anticipated with full size pre series sensors. M3.15 WP3 Designs complete for pre production sensors End 2009 N HPK Yes The final design for the ATLAS upgrade sensors will need to be informed by the results of the present pre series programme, Gantt chart 33

34 3.4 WP4 Microelectronics and Interconnect Research and Development Commentary on project to date This section reports on the status of the ASICs, the hybrids, results from a first module and the development post processing ABCN25 status In order to gain time, the first ABCN25 wafer was diced and ASICs were distributed to the collaboration for testing without first wafer probing. The yield seems to be very high with all ASICs mounted to a test PCB or hybrids so far working. A systematic measurement of the performance is well underway and so far the ASIC has met the design specifications. The UK has produced all of the early performance results. Some results are shown in the following hybrid section. A US/UK project has been initiated to design a buffer control chip (BCC) which will interface ABCN25 ASICs on a hybrid to signals on a stave. The design is almost complete and the ASIC will be fabricated through MOSIS and is expected in July 09. This device will allow the implementation of the serial powering schemes discussed below. The design and purchase of the electronics and DAQ for a wafer probing system has started but the implementation was given lower priority because of the high yield observed to date. The plan now is to complete the initial detailed evaluation of the ABCN25 and use the results to define an optimised wafer probe test Hybrid Status The UK designed and fabricated a conservative first single finger hybrid with Stevenage circuits in the UK. The hybrid is 24mm wide and is shown below fully equipped with 20 ABCN25 ASICs (2560 channels). The hybrid production yield was also high (89%) and the electrical performance fully met specification. An individual channel noise of ~400enc was measured at 40MHz (the same as for a single ASIC). First tests at 80MHz show the same performance. The internal regulators which will be important for optimising the power distribution have also been shown to work well. The next step will be to produce hybrids suitable for the stave. These will be derived from the existing hybrids with a 3 layer plus shield topology, slightly reduced dimensions and with the connectors replaced by wirebonding. An important development area is related to having the ability to populate hybrids and bond them directly to sensors. 34

35 Figure 6. First hybrid equipped with 20 ASICs Figure 7 Test results from first hybrid showing uniform gain and noise Module tests The UK hybrid has been made up into a single sensor module shown in Figure 8 below. This will be the basic electrical building block on a stave and the correct performance of such a unit is a major electrical milestone. The UK assembled the first ATLAS module with a bridged hybrid at one end and a directly glued hybrid at the other. 35

36 Figure 8 First assembled module with a directly glued hybrid and a bridged hybrid The first test results from this module are extremely encouraging. The input noise with a 2.5cm strip connected is ~600enc for both the bridged and directly glued hybrids as can be seen in Figure 9 below. (This used a hybrid designed for a bridged module not for direct gluing and so a ground shield layer had to be implemented by hand in this case.) Figure 9 Noise for the ASICS on the directly glued hybrid. The missing channels have understood bonding problems linked to the placement of the hybrid on the sensor. We note here that with the adoption of the stave that the thermo mechanical study should involve the complete stave; it can not be divided into a module plus support. Consequently the work is now treated in WP Post-processed Development Status After some delays the first batch of wafers have been processed by ACREO and measured by us. This R&D programme is now planned to go in 3 stages. The goals of the first stage were to study the impact of post processing on sensors. The processing consisted of a single BCB layer etched to open contact windows and a top patterned metal layer. 36

37 DC measurements have shown an acceptable production yield and no adverse effects on the sensor. The depletion IV curves (Figure 11) are well behaved. The capacitance seen by the strips increases as expected (Figure 12). Figure 10 Layout of post processed wafer (The yellow are the ground planes) Figure 11 Measurement of the bias current for post processed sensors Figure 12 Interstrip capacitance as measured for post processed wafers. Note this is not the capacitance seen by the ASIC amplifiers (ref Chilingarov) 37

38 Goals and concerns from previous period, WP4 Goals Status Finish design and manufacturing of 0.25 μm ABCN25 test PCBs. Evaluate ASIC performance. Design, manufacture, and prototype testing of 0.25 μm ABCN25 wafer probing system (electronics and software). Continue with specifications of final ASIC chip set (0.13 μm ABCN25, MCC, power ASIC, DCS ASIC, ) for use in experiment. Build and evaluate mechanical hybrid/module pre prototypes to confirm thermal FEA studies. Evaluate use of differing hybrid substrate materials as they become available. Manufacture, populate and test singlefinger 0.25 μm ABCN25 electrical prototype hybrid for use in first module prototypes and powering studies. Build and electrically test first prototype module with two hybrid fingers on a sensor. Beginning the design of the supermodule, 0.25 μm ABCN25 hybrids Manufacture and testing (including 40 MHz readout and irradiation studies) of first sensor post processing prototype. Beginning of 2 nd post processing prototype design or first hybrid onsensor prototype depending on results of 1 st post processing testing. Concerns The functionality and availability of 0.25 μm ABCN25. Any delay in functional ABCN25 ASICs impacts on the majority of program. Available of materials for prototyping, especially uni directional carbon carbon. Manpower, especially if commissioning of current ATLAS has unexpected problems. Production complete. Evaluation ongoing but ASIC appears to fully meet specification In progress. Assigned lower priority because of the very high yield achieved In progress with strong UK input. No longer relevant (moved to WP7) Completed with Stevenage Electronics, Ltd, UK. First module built. Preliminary results available In progress Production and probe testing completed. Readout and irradiation planned. In progress. Revised plan is to factorise the project and build a thin film standalone hybrid first Status OK. Excellent yield and performance See WP7 Currently OK. Biggest pressure is in the area of DAQ development Plans for next 6 months 38

39 The goal for the next 6 months is to complete the evaluation of the ASIC, Hybrid and singlesensor module and then design and produce the components for a stave. As soon as the ASIC tests are complete, the rest of the production wafers need to be purchased. The wafer probing system will need to be up and running to select the good die. We have allowed for two iterations of the hybrid in our planning finalizing the design and manufacturability. There is increasing complexity as the hybrid will have to interface both to the sensors and to the stave bus. The BCC will need evaluation and integration onto the stave. As the first electrical stave will consist of modules built throughout the collaboration there will be additional delay in standardizing components dimensions and interfaces for the hybrid. The post processing programme will continue with the irradiation of the diced postprocessed sensors to measure interface charges. In parallel with this work a thin film hybrid will be designed and built. Its performance will be compared to the current Kapton hybrid Goals and concerns for next six months, WP4 Goals Complete the evaluation of ABCN25 ASICs and place production order. Evaluate BCC Commission ASIC test station Design and prototype Stave hybrid Assemble single sided module with stave hybrid Design and prototype thin film hybrid Irradiate post processed sensors and evaluate performance. Prepare thin film hybrid design for post processing Concerns WP4 financial statements Notes about figures below: 1) Outstanding 18K ASIC commit not yet in FRS 39

40 Approved (excluding contingency) Transfers Actual spend to Feb 09 Projected spend this year Actual spend (2+3) Projected spend (2+4+5) Actual (6-1- 1a) (1) (1a) 2007/ /2009 (3) (4) 2009/ /11 (6) (7) University Staff Effort Costs* Liverpool Cambridge QMUL Glasgow University Sub-Total STFC Lab Costs RAL PPD RAL Technology STFC Lab Sub-Total Equipment Other Directly Allocated costs (eg consumables) Travel Total (Excluding VAT and WA) Working allowance VAT WP4: Finance Summary (all figures in k) Actual spend in previous years (2) Current year 2008/09 Latest estimate of future requirement (5) Total Variance Projected (7-1-1a) Total (including VAT & WA) 1, , Contingency (Held by STFC) 1 Excluding Working Allowance and VAT * The University staff effort recorded in this table should be the 80% amount STFC pays, including academic time Use of columns: (1) = The amount approved by STFC (1a) = This column should be used to show any virements between headings, for example when Working Allowance is used, the amount should appear as a debit in the WA row and then credited to the relevant row (2) = The actual spend in previous financial years, by year (3) = The actual spend in the current financial year up to the most recent quarter (4) = The total projected spend for the current financial year, including any expenditure so far (ie actual spend this year plus predictions of remaining spend this year) (5) = Projected spend for the remaining years (6) = The actual spend so far (7) = Projected spend over the whole duration of the project (ie actual spend so far plus predictions of remaining spend to project completion) The variance columns show the difference between the actual and projected amounts and the approved amount. 40

41 3.4.4 Summary of major deliverables and their financial allocation Deliverable and description Financial allocation from WP4 (kgbp) Project phase Development, production & evaluation of ABCN25 incl. wafer probe test system 70k R&D Single finger hybrids & modules 25k R&D Stave hybrids development & production 65k R&D Post processed sensor 25k R&D Thin film hybrid 25k R&D Integrated hybrid and sensor 60k R&D/pre series.resource Usage Table of staff %age FTE WP4 WP4 NAME INSTITUTE Allport P P Liverpool 5 5 Affolder A A Liverpool Greenall A Liverpool Smith N A Liverpool Muskett D A Liverpool 10 Whitley M Liverpool Wormald M P Liverpool Hommels Cambridge Parker Cambridge Goodrick Cambridge 5 10 Shaw Cambridge Carter QMUL Beck QMUL Gannaway QMUL Morris QMUL L.Eklund Glasgow V.O'Shea Glasgow John Melone Glasgow Lewis Batchelor/Ass. Tech. RAL 40 Jeff Bizzell/Analogue Eng. RAL 20 Richard Holt RAL John Matheson RAL Peter Philipps RAL Giulio Villani RAL Marc Weber RAL Mike Tyndel RAL RAL TD RAL TD Total Grand Total / / / /11 41

42 3.4.5 WP4 milestone plan Table 4.1: Milestones achieved in the last six months Milestone Work Milestone Target Date Status No. Package M4.1 WP4 ABCN25 FDR Feb 08 Complete M4.2 WP μm ABCN25 ASIC Jan 09 Complete evaluated M4.7 WP4 Prototype electrical hybrid Feb 09 Complete tested M4.11 WP4 Results on 2 layer postprocessing Mar 09 Pre irradiation complete Table 4.2: Overall Milestone List for WP4 Milestone No. Work Package Milestone As at Sep 08 As at Mar 09 (Changes in Bold) Delay due to UK Other Collaborators Affects Critical Path? See Note Reason for change M4.1 WP4 ABCN25 FDR Feb 08 Feb 08 N Y M4.2 WP μm ABCN25 ASIC Jan 09 Feb 09 evaluated M4.3 WP μm ABCN25 available Jun 09 Sep 09 Y Y Reprioritised; not on critical path M4.4 WP4 MCC FDR Jan 10 Jan 10 BCC earlier M4.5 WP4 Module concept fixed Jun 09 Jun 09 M4.6 WP4 Thermo mechanical Feb 09 Xxx Moved to WP7 hybrid tested M4.7 WP4 Prototype electrical Mar 09 Mar 09 hybrid tested M4.8 WP4 Prototype module Apr 09 Apr 09 electrically tested M4.9 WP4 Noise studies with Sep 09 Sept 09 multiple modules complete M4.10 WP4 Hybrids available for Nov 09 Nov 09 super module M4.11 WP4 Results on 2 layer postprocessing Feb 09 Feb 09 M4.12 WP4 Results on 4 layer postprocessing Nov 09 Feb 10 Thin film first M4.13 WP4 Results on 6 wafer Jun 10 June 10 post processing M4.15 WP4 SPI flip chip complete July 09 July 09 M4.16 WP4 Post processed flip chip results Sep 10 Sep 10 42

43 3.4.6 Gantt chart The Gantt chart has been updated to show progress for each task. Note Three of the tasks are greyed out : 1) ABCN13 project delayed (noted at time of approval) 2) Module Thermo mechanical studies moved to WP7 3) Flip chip program delayed (noted at time of approval) 43

44 3.5 WP Scope of WP5 In the light of the international developments in the field of R&D on optoelectronics readout for the SLHC, the scope of this work package has been substantially revised with respect to the original proposal submitted in In 2007 the Oxford Group was invited to join the international Versatile Link Project (VL) [ i ]. This was submitted to both the CMS and ATLAS upgrade steering groups at the end of 2007, and was positively reviewed by each collaboration. The programme of WP5 is now fully embedded within this international effort, much better streamlined and provides the UK with a leading role in a very important cross experiment international R&D programme for the SLHC. The VL project will concentrate on fast bi directional digital optical data transmission at rates up to ~5 Gbit/s with an emphasis on SLHC level radiation resistance, low power and low mass components. The project concentrates on the technical development of the link and finishes at the pre production readiness stage. No production level commitments are envisaged at this stage. The versatile link system is foreseen to operate in point to point or point to multipoint bi directional 4 topologies and will be made available in multimode (MM) and/or single mode (SM) versions operating at 850nm or 1310nm wavelength respectively. It will have serial data interfaces and will be protocol agnostic (within the limits set by the DC balance requirement on the data stream). Examples of applications could be a DAQ link implemented in a point to point configuration while a TTC link could be built around a point to multipoint topology. The VL project will develop and qualify active electro optic components such as packaged laser transmitters and pin diode receivers, as well as validate passive components such as fibres, connectors and couplers. The ASIC design work associated with the transmit and receive functionality of the versatile link (laser driver and transimpedance amplifier) will be carried out in the framework of the GBT common project [ii]. The high level definition of the system will be under the responsibility of the experiments interested in using the versatile link. The official start date of the project was The versatile link project is broken down into 6 concurrent work packages, 3 at the system level and 3 at the components level. Each work package is under the technical and financial responsibility of one institute. The institutes report to their individual funding agencies concerning their work packages. The project as a whole reports to the experiments via their SLHC steering groups. Oxford took the responsibility for the Passive Components work package which is lead by Issever. Academia Sinica is also on this work package as a participating institute. 4 Bidirectionality is understood here in its most simple sense, as transfer of data in two directions over two fibers. 44

45 Passive Optical Components The passive optical components which will be required for the Versatile Link are fibres, connectors and any fibre couplers 5 that may be needed to implement the various topologies under investigations. The fibres will also require protective jacketing around individual fibres (for inside the active volume of the detector) and multi fibre cables for the longer run from the detector to the counting rooms. Since systems based on multimode transmission at 850 nm and single mode at 1310 nm are being developed, suitable variants of the passive components need to be identified for both 850 nm and 1310 nm. For the fibres, this implies that Graded Index (GRIN) fibres for use at 850 nm and SM fibres for use at 1310 nm will be considered. The main tests that will be performed will be to verify the radiation tolerance for the expected SLHC doses (up to 500 kgy(si) [ iii ]). Other studies will look at mechanical properties of fibres, in order to understand reliability issues, particularly after irradiation. The bandwidth of fibres after irradiation will be also evaluated to ensure that they are suitable for a 5 Gbits/sec operation. The components will be checked under realistic environmental conditions. The programme for the evaluation of radiation tolerance of the passive optical components has started and the status and plans will be described in the following sub sections. To minimise risks during production, two sources for each component type will be identified. Fibres The bandwidth for the fibres should be sufficient for data transfer at a rate of ~5 Gbits/s over lengths of up to 100 m. A commercial SM fibre 6 has already been tested with a gamma source to the expected doses and has shown very encouraging performance [ iv ]. Another possibility for SM fibre is to use the same fibre currently used for the LHC machine, which has already been demonstrated to have excellent radiation tolerance [ v ]. One commercial 7 and one prototype GRIN fibre 8 have also been tested to SLHC doses with good results. This prototype fibre is expected to become commercially available during As the fibres show important annealing with time and light (photo bleaching), the tests have all been active, in which realistic light levels have been used and the performance of the fibres measured during the irradiation. Extensive studies of the dose rate effects for the GRIN fibre were performed showing that the damage for a given total dose, increased with dose rate as anticipated. This fibre has the required bandwidth and will therefore be used as a reference for future tests. In addition bandwidth evaluations of the irradiated fibres will be performed. Attempts to find other sources of radiation tolerant GRIN fibre are being pursued Thus far all the existing tests have been performed at room temperature or higher. Further studies are needed before these fibres can be qualified for the low 5 Also referred as splitters. 6 Corning SMF 28 7 Corning Infinicor SX+ 8 Draka, Super Rad hard fibre. 45

46 temperature operation ( 20 C) that would be required for them to be operated inside SLHC silicon trackers. Mechanical performance tests of unirradiated fibres have started and will be repeated on irradiated samples. The aim is to determine if the mechanical reliability is degraded by irradiation. These involve tensile strength measurements using a tensile test machine, equipped with a 1kN load cell at CERN. This is a destructive test, as the breaking point is used to determine the strength of the fibre at a given deformation rate. Some example data for the SMF 28 fibre are shown in Figure 13. Different pull speeds can be used to perform a reliability analysis. Figure 14. Stress versus fibre extension (transfer function) for 3 samples of 50 cm lengths of unirradiated SMF 28 fibre at a pull speed of 10 mm/ minute. The ends of the curves indicate the breaking points. The fibre was preloaded with 500 MPa before the dynamic test started. Fibre Couplers and Connectors The first irradiation tests with fused taper couplers have given very good results for both GRIN and SM fibres as the damage is consistent with that expected from the lengths of fibre used in the devices. Further tests will be performed to understand if there are any significant differences between passive and active radiation tests. However the first radiation tests with a SM Planar Lightwave Curicuit (PLC) coupler showed significant sensitivity to radiation. Therefore this type of device could only be considered in regions of lower total dose. However other versions of PLC coupler will be surveyed to determine if more radiation tolerant devices are available. The LC connector has been identified as the most suitable small form factor connector for single fibres for both SM and MM systems. The first radiation tests have given very encouraging results, although more work is still needed to understand all the systematics in the measurements. The fibre plans may involve the 46

47 use of multi way fibre ribbons, therefore the radiation tolerance of multi way connectors like the MT 12 9 will be evaluated. Fibre Cabling The Versatile Link will use individual fibres, therefore protective jacketing will be required. The longer runs of fibres from the detector to the counting rooms will require the use of multi fibre cables or equivalent protection systems. Their radiation tolerance will need to be evaluated. Since the concern will be about changes in mechanical strength, it will be necessary to test with both ionizing and non ionizing sources. Therefore a comparison of mechanical strengths will be performed before and after irradiation Commentary on project Oct 2008 Apr 2009 We have written the paper which summarizes the results of our irradiation run in The paper will be submitted to JINST by the time of the OsC meeting. The most important outcome of this irradiation run is that the ambient temperature of fibres during irradiations looks to have a significant impact on the performance degradation of the fibres which is discussed later in more detail. We made progress in the understanding of how to perform the mechanical tests on the non irradiated fibres and are now able to measure good quality transfer functions and are also able to prevent breakage at the disks where the fibres are mounted. The environmental chamber is set up in Oxford and we are in the process of programming LABVIEW drivers to control the chamber remotely. In the last 6 months we have made firm contacts with Draka and have established that they will be starting commercial production of the prototype radiation hard fibre that we tested. From the contacts with other fibre manufacturers we have a possibility of a candidate fibre from Polymicro and we are currently performing mass spectrometry analysis to determine if there is any P in the core. We havenʹt made progress on the coupler market survey, although we have verified that a fused taper couplers made with rad hard fibre, has the expected radiation tolerance. We are in the process of preparing the next active gamma irradiation test at SCK. We concluded that after the analysis of the results of the irradiation run in 2008 that the next test needs to be performed at 20C because we found out that the performance degradation of fibres looks to be heavily dependent on the temperature. The lower the temperature is, the higher the observed attenuation seems to be. We are designing, in collaboration with SCK, a cold vessel which will allow us to cool the fibres during the irradiation down to 20C which is the temperature the inner detector volume will be operating at. This apparently large dependency of the attenuation on the temperature was not expected, and means that the irradiation programme will be much more extensive than we had foreseen. Each possible fibre candidate needs to be irradiated at cold temperatures and any new candidate would need to be irradiated in the high dose and in the low dose (cold) source at SCK increasing the costs. If the upcoming cold test this spring reveals that the shortlisted fibres are not suitable for SLHC, WP5 will need to rapidly find new candidates and tests them again. 9 MT: Mechanically Transferrable splice. 47

48 We did not progress on the passive irradiation tests of couplers. The Versatile Link management had decided to put the mechanical tests for now at a lower priority and to focus the efforts on market surveys and irradiation tests. We got a STFC CASE studentship with Ericsson UK awarded and we are trying to fill this studentship. We have a lot of difficulties to get eligible students to apply, although the position was very widely advertised. Discussion of the Temperature Effect on the Fibre Performance Figure 15 shows the relative attenuation vs. time for the SMF 28 single mode fibre. All the high dose rate (22.5 kgy/h) exposures show a feature of very rapid attenuation (a in Figure 15) occurring almost instantaneously with immersion in the high dose 60 Co source followed by a somewhat less rapid decrease in attenuation (c in Figure 15). This effect is not present in any of our lower dose rate tests done at SCK CEN (1.01 kgy/h) or BNL ( kgy/h). In addition, as shown in Figure 15, the effect re appears after the initial exposure when the vessel containing the fibres was taken out of the radiation environment (first stop of radiation label in Figure 15) and then briefly returned to the radiation environment (second start of radiation label in Figure 15). This feature appears in all the other high dose fibre test as well. We conclude the process causing this attenuation spike is, to lowest order, a reversible process. Figure 15: Relative attenuation of the SMF 28 single mode fibre vs. time to the end of the full run at SCKCEN in The SLHC exposure started at approximately the 2.5 hour mark and ended after 30 hours of exposure. Note that the vertical axis is in db. Dose rate was 22.5 kgy/h and total accumulated equivalent dose was 650 kgy (Si). We are currently uncertain as to the exact cause of the fast attenuation and then rapid recovery but we have strong suspicions that it is related to the vessel temperature in 48

49 the high dose rate facility coupled with the known natural annealing behaviour of optical fibres. The fibres in the high dose rate facility were held in place by solid aluminium disks of a diameter of 220 mm and an average thickness of 10 mm. Three such disks were used. Gamma ray photons at the energy of 60 Co decay will Compton scatter off the electrons in any material. In each such scattering energy is deposited in the material. If the flux of gamma rays is high enough that energy will eventually manifest as a measurable source of heat. The Compton Scattering heating effect was certainly present in the high dose rate test shown in Figure 16. The explanation given is that the initial observed quick rise in attenuation is due to the presence of pre existing defects within the fibre that naturally arise in any manufacturing process. In this case self trapped holes (STH) in the glass matrix are identified as the likely cause. STH put trap sites in the band gap between the conduction and valence band and a large influx of radiation will quickly fill those trap sites with electrons which will act as scattering centres for optical transmission. The effect is most pronounced in the visible region of the spectrum but has a continuum which extends well into the infrared. The fact that the relative peak size is significantly higher in our 850 nm fibres than in our 1310 nm tests supports this point. Optical fibres also have a strong annealing behaviour to ambient temperature [ vi ],[ vii ]. The temperature dependence in the case of the fibres presented in [vi],[vii] was quite strong (see Figure 18) especially for the Ge doped one. The multimode fibres we tested are all Ge doped, and we therefore have reason to suspect they too will exhibit strong temperature dependence. As the temperature of the fibre increases bonds that are associated with trap sites break releasing the trapped electrons and improving the transmission characteristics of the optical fibre. We believe that we are seeing first, the high attenuation caused by the immediate filling of trap sites from STH (a in Figure 15). These sites fill very quickly in a high dose environment to a maximum that is reached within minutes of the exposure. In this particular case though, the temperature rise in the vessel was not far behind and as this temperature increased the STH bonds were broken and the attenuation fell as a result (c in Figure 15). Radiation does, obviously, damage an optical fibre so this recovery period is limited both by the fact that the damage process continues and by the fact that the temperature profile did stabilize after two hours in this particular test (at around the 4.5 hour mark in Figure 17 lower plot). We believe the fact that a new minimum attenuation is reached at about this same time indicates that the temperature annealing process is complete and the continuing damage from the gamma ray source takes over. There is much that is still not well understood. Some parts of the path the fibre takes out of the upgraded LHC detectors are likely to be in cold environments ( 20 o C), consequently for a future study we think it is very important to take further samples from our same Infinicor SX+ reel and repeat this test under both constant room temperature and temperatures approaching 20 o C to determine the sensitivity of this fibre to temperature variations in a radiation environment. 49

50 Figure 16: Plot of the temperature within the high dose rate vessel as a function of time The first rise in temperature is when the vessel was first lowered into the radiation area. The decrease occurs after the vessel was removed for several hours to study photo bleaching effects. The final temperature increase was due to a second exposure before the end of data taking. Figure 17: In this plot the attenuation of all the multimode fibres under test in the high dose rate area are shown vs. time but the time axis has been greatly expanded. Beneath this plot is the plot of the ambient temperature in the vessel containing these fibres plotted vs. time on the same scale. 50

51 Figure 18: Temperature dependence of loss increase at 1.3 μm due to gamma irradiation for 0.4 h at 10 kgy/h [vi]. 51

52 Goals and concerns from previous period, WP5 Goals Analyze irradiation data and document results in form of a report (to be published in the future). Perform mechanical test on irradiated and non irradiated fibres. Setup environmental test for passive components. Continue market survey of couplers and fibres. Prepare next active gamma radiation test of fibres Prepare passive gamma radiation tests of couplers Define with the Versatile Link project management the details of the future tests (irradiation, mechanical and environmental). Concerns Mechanical tests of fibres are performed at a facility in CERN. We have to share the equipment to perform the dynamic pull tests with others at CERN and have not control over the access to the machine. This makes the progress on these tests slow since we need to collect a lot of statistics. Manpower is always an issue. Status Paper is written and to be submitted to JINST. Progress made, fibres are not breaking at the disks and good quality transfer functions measured. Environmental chamber is set up. We are programming the LABVIEW drivers for the chamber which is essential for the environmental test setup with passive components. Good progress on fibre market survey. No progress on the coupler market survey. In progress. Components for irradiation are in our hands. We are building a cold vessel to perform irradiations at 20 C. No progress. But we irradiated peltier coolers at IPAS as part of the cooled irradiation test preparations. At the face to face Versatile Link Project Management meeting in September 08 it was decided that WP5 should put the mechanical tests for now at a lower priority and should focus the efforts on market surveys and irradiation tests. Status The situation at CERN is the same. We made progress on the tests but it was a slow progress. We put a bid for a material testing system into the RG application. We also agreed with Ericsson that the CASE student will be able to perform mechanical tests at the premises of Ericsson, but this will not happen before summer 2010 and is dependant upon filling the position this year. We applied for a STFC CASE studentship with Ericsson and got it awarded (start 52

53 Oct 2009), we are encountering many difficulties filling this position though. Until now we had foreseen one passive irradiation test in the next year. But depending on the results of the data analysis of our SCK run in Aug 2008 and the discussion with Taiwan and the VL project the number of passive irradiations could increase. This is just to flag a possible change in the plan of WP5 depending on internal and external reasons. Based on the analysis of the irradiation test results we concluded that a cooled irradiation test of fibres has a higher priority than further room temperature passive tests Plans for next 6 months We will perform the cold irradiation of fibres at 20 C at the SCK gamma source. Based on these tests we will know if the currently identified fibre candidates are still suitable for SLHC operation. These results will define details of the next foreseen irradiation tests in September Within the next 6 months, the results of the 2008 irradiation test in JINST should be published. We aim to find a suitable student for our CASE studentship with Ericsson such that the student can start his/her PhD in October Now that we are able to produce good quality transfer functions (see Figure 14) we will aim to collect more statistics and determine the mechanical reliability of unirradiated fibres. Within the next 6 months we intend to start to test the environmental resistance of fibres, connectors and couplers in an environmental chamber which has a temperature and humidity control. We will focus on the market survey for more couplers and once we have purchased possible candidates we intend to irradiate them passively at a gamma source in Taiwan. Suitable couplers which perform well after the passive irradiations will be irradiated actively at SCK. We will get in touch with industry and identify possible cabling and packaging options for fibres. We will define the details of the Phase 2 of the versatile link project and will seek approval for it from the VL project steering board. Once this is done the plans for the October 2009 to October 2010 of the WP5 project will be known and will be reported at the next OSC meeting. 53

54 Goals and concerns for next six months, WP5 Goals Perform cold irradiation test of fibres and monitor performance of the fibre during the irradiation (so called active test). Based on the tests results of the cold irradiation we will determine the details of the next active irradiation at SCK. Publish irradiation results of 2008 in JINST Find student for CASE studentship with Ericsson Collect statistics on the mechanical tests of non irradiated fibres Discuss with SC Lee new students and get new students Irradiate fibres passively at a gamma source in Taiwan for mechanical tests Tests passive components in the environmental chamber Perform market survey for couplers. Perform passive irradiation of couplers. Understand power instabilities in the test system O(0.5dB) Get in touch Brookhaven for mass spectroscopy of polymicron fibre Define phase 2 of versatile link project and get approval for it from VL management. Concerns We are concerned that we will not find an eligible and qualified candidate for the CASE studentship for this year. Speed of progress in the mechanical tests of fibres The strong temperature dependency of radiation induced damage in fibres means a more intense testing programme than we had foreseen and the irradiation costs will be higher than expected. The increased scope of the irradiations will cost more as detailed; we need to finance this additional work WP5 financial statements The expenditures for WP5 for other allocated costs (e.g. consumables, beam time) in FRS (FK21500) start in October Until April 2009: Actual spend: 17,352 Committed spend: 0 Total: 17,352 Projected total spend until March 2009: 18,000 Projected total spend until March 2010: 49,500 Approved: 29,000 According to our current estimates, we will overspend to achieve the programme including temperature dependent studies. As discussed above the implication of the temperature dependence attenuation is extending the scope of the irradiation programme and we foresee a higher number of irradiations than we had planned originally. 54

55 Until the end of the project we estimate to have 4 cold irradiations at the low dose source and 3 irradiations at the high dose source in SCK. The beam cost of each of these irradiations is 4000, plus 500 for consumables for each test, adding up to a total of WP5 is expected to have left for March 2009 to March 2010 meaning we would overrun the budget by In the Standard format finance table below it is apparent that the capital and consumable costs are predicted to overspend. The total WP appears to still be in credit due to the manpower underspend ( 28k). As was described in the previous OsC report the manpower underspend is due to Issever receiving an academic post at Oxford and thus being charged to the project at a very different (lower) rate. This money is part of the RG contribution so this saving is not accessible to the project. 55

56 University Staff Effort Costs* Approved (excluding contingency) Transfers WP5: Finance Summary (all figures in k) Actual spend in previous years (2) Current year 2008/09 Actual spend to Sept 08 Projected spend this year Latest estimate of future requirement (5) Actual spend (2+3) (1) (1a) 2007/ /2009 (3) (4) 2009/ /11 (6) (7) Projected spend (2+4+5) Actual (6-1- 1a) Oxford Total Variance Projected (7-1-1a) University Sub-Total STFC Lab Costs RAL PPD RAL Technology STFC Lab Sub-Total Equipment Other Directly Allocated costs (eg consumables) Travel Total (Excluding VAT and WA) Working allowance VAT Total (including VAT & WA) Contingency (Held by STFC) 1 Excluding Working Allowance and VAT * The University staff effort recorded in this table should be the 80% amount STFC pays, including academic time Use of columns: (1) = The amount approved by STFC (1a) = This column should be used to show any virements between headings, for example when Working Allowance is used, the amount should appear as a debit in the WA row and then credited to the relevant row (2) = The actual spend in previous financial years, by year (3) = The actual spend in the current financial year up to the most recent quarter (4) = The total projected spend for the current financial year, including any expenditure so far (ie actual spend this year plus predictions of remaining spend this year) (5) = Projected spend for the remaining years (6) = The actual spend so far (7) = Projected spend over the whole duration of the project (ie actual spend so far plus predictions of remaining spend to project completion) The variance columns show the difference between the actual and projected amounts and the approved amount. 56

57 3.5.5 Summary of major deliverables and their financial allocation Deliverable and description Financial allocation from WP5 (kgbp) Project phase Have a portfolio of passive components based on functionality, environmental resistance and mechanical reliability and radiation resistance Identify and document cabling and packaging options; Define and get approval of phase two of the Passive Component work package programme. 29 (49.5) R&D R&D R&D Resource Usage Table of staff %age FTE WP5 WP5 NAME INSTITUTE E Tech RG Oxford Huffman Oxford Issever Oxford Jones Oxford Weidberg Oxford Issever (RG) Oxford Halliday (RG) Oxford Lau, W Oxford M Tech RG Oxford Lau, P Oxford Fopma Oxford Total Grand Total / / / / WP5 milestone plan Table 1: Milestones achieved in the last six months Milestone Work Milestone Target Date Status No. Package M1.3 WP5 Feedback to companies Oct 2008 Done 57

58 Table 3: NEW Milestone List for WP5 Milestone No. Work Package Milestone As at Sep 08 As at Mar 09 (Changes in Bold) Delay due to UK Other Collaborators Affects Critical Path? See Note Reason for change M1.1 WP5 Completion of first Aug Aug 07 N N gamma irradiation test of fibres at SCK. M1.2 WP5 Completion of gamma Aug Aug 08 N N irradiation tests of fibres, splitters and connectors at SCK M1.3 WP5 Feedback to companies Oct 2008 Dec 2008 N N Met with Ericsson and Draka at CERN in December 08 M1.4 WP5 Submit Paper to JINST N/A Apr 2009 N N New milestone M1.5 WP5 Cold ( 20 C) irradiation N/A June 2009 N N New milestone of fibres at SCK M1.6 WP5 Completion of passive irradiation tests in Taiwan June 2009 June 2009 N N M1.7 WP5 Completion of active irradiation tests at SCK M1.8 WP5 Feedback to companies Oct 2009 Oct 2009 N N M1.9 WP5 Determine mechanical performance degradation of irradiated fibres M1.10 WP5 Determine mechanical performance degradation of irradiated fibres M1.11 WP5 Determine environmental resistance and reliability of passive components M1.12 WP5 Define portfolio of passive components M1.13 WP5 Identify preliminary cabling and packaging options M1.14 WP5 Identify cabling and packaging options M1.15 WP5 Definition of Phase Two of WP2.3 of VL Project M1.16 WP5 Approval of Phase Two of WP2.3 of VL Project July 2009 Sep 2009 N N This test was postponed because we needed to perform first a cold irradiation in order to define the details of this test. Jan 2009 Nov 2009 N N Slow progress due to problems with fibre mounting Sep 2009 Sep 2010 N N Slow progress due to problems with fibre mounting Oct 2009 Sep 2010 N N Missing Labview Drivers Oct 2009 Sep 2010 N N Mechanical and Environmental tests progressing slower than expected. The temperature dependency of the attenuation of irradiated fibres is extending the radiation programme. Apr 2009 Apr 2009 N N Sep 2009 Sep 2009 N N Mar 2009 Jun 2009 N N VL Project Management shifted this date. Sep 2009 Sep 2009 N N 58

59 3.5.8 Gantt chart Mechanical tests of non irradiated fibres take longer than expected. The setup of the environmental chamber takes also longer than expected. The radiation programme of WP5 is extended due to the discussed temperature effects we seem to be seeing in our 2008 tests. 59

60 3.6 WP Commentary on project to date The highlights of the past 6 months have been the delivery and testing of the ABCN25 and SPi chips. This is the culmination of almost 2 years of design work. The chips are functional and will eventually allow us to estimate the ultimate electrical performance of serial powering, build miniature SP circuits and make a major leap in our understanding of SP systems in general. We have also made significant progress on protection schemes for serial powering. We developed, simulated and prototyped one possible scheme. We coordinated a series of meetings with other groups working on the topic and are close in achieving convergence on the specifications of the system. Our other activities involve the testing of the constant current source prototype; comparison of alternative SP implementations in 130 nm ABCnext (the variants differ in the use of LDOs or integrated DC DC step up or down charge pumps); design of the BCC chip with LVDS drivers for AC coupling, involvement in testing of AC LVDS coupling. We did not proceed with the design of SP plug in boards as planned since the module and hybrid interfaces are not yet defined. We did not participate as planned in the testing of the LBNL stave since SCT commissioning and ABCN25 DAQ development prevented prolonged stays in the USA. Organizational and coordination work is still a significant burden, but is very beneficial for the overall progress of the field. We continue to coordinate much of the international power distribution R&D and drive serial powering R&D. ABCN25 and SPi chips The chips contain several power management and serial powering blocks. The combination of ABCN25 and SPi allows us to test three alternative serial powering architectures (see Figure 19). The ABCN25 contains a low drop out (LDO) regulator to derive 2.2 V analogue voltage from the nominally 2.5 V digital voltage. It also contains an internal shunt regulator and shunt transistors for serial powering in the W scheme and an independent set of shunt transistors for operation in the M scheme when combined with the SPi chip. SPi is a complex chip which contains shunt regulators and transistors for stand alone power regulation and many other blocks. In order to characterize these chips, we specified, designed and manufactured a number of PCBs and set up several DAQ systems. The SP blocks have been tested in various configurations mostly with single chips, some on a 20 chip hybrid (for a photograph of the 20 chip hybrid see WP4). The SP blocks are fully functional and can be used for future prototyping. The noise of the ABCN25 chip is similar with and without serial powering circuitry for the W and M schemes, as it should be. SPi is fully functional, its noise is low and it is able to maintain shunt currents of several amps. A photograph of SPi mounted on different PCBs is shown in Figure

61 Figure 19 Three serial powering schemes: (Top) W scheme with internal shunt regulator and shunt transistors operating in parallel. (Middle) M scheme with internal parallel shunt transistors driven by a single external shunt regulator. (Bottom) External shunt regulation using SPi. Figure 20 SPi bump bonded to a daughter PCB (left). The daughter PCB on its mother board (right). Protection The most relevant single point failure mechanism of SP systems seems to be the interruption of the current flow through the modules due to a broken connection, e.g. broken wire bonds. Solidly engineered connections and/or a redundant serial powering trace would be important to minimize the failure probability. Another very effective protection against this failure is shown in Error! Reference source not found.. A dedicated by pass circuit (switch) can be enabled to provide an alternate current path if the connection to/through the module 61

62 becomes faulty. Several protection schemes are currently being specified and prototyped with discrete components. Figure 21 Protection schematic: A single DC power line drives the protection circuitry which consists of an on board oscillator, rectification and voltage amplification stage, etc. Goals and concerns from previous period, WP6 Goals Submission of SPi chip (FNAL responsibility) Completion of DAQ systems for SPi and ABC Next read out and control Exploration of SPi bumping process Characterization of SPi and of ABC Next power blocks Specification, layout and population of SP plug in board Completion of 30 module stave construction and testing (at LBNL) Contribution to SP system specification: protection scheme; slow control (DCS); grounding and shielding concept Concerns Status SPi was received Completed Found one, ongoing Completed Delayed Delayed; planned for mid year Very good progress Status 62

63 Delay in SPi submission SPi or ABC Next might not be functional RAL manpower is marginal due to SCT commissioning, international power co ordination tasks and STFC restructuring Dependence on non UK collaborators SPi was received Chips are functional Remains a concern Remains a concern Plans for next 6 months ATLAS considers making a baseline decision on the power distribution scheme by the end of It is desirable to reduce the number of options in serial powering before this review. 1) The performance and other features (e.g. real estate, mass, number of dedicated chips) of the three serial powering schemes should be compared in detail with each other. For the measurements we will either design or modify a suitable hybrid or use the aforementioned plug in board. This study will provide crucial input for the selection of the favoured SP scheme within ATLAS. For similar studies with modules see 3). 2) We will continue our studies on protection schemes. Our current design is for powering ABCDs. We will extend this to ABCN25s. We will converge with Bonn and BNL on the specifications of the final system and prototype and test it with discrete components (either at RAL or BNL). 3) We will need to gain experience with SP and DC DC on a silicon module. Modules are just becoming available in small numbers. Tests on modules include noise measurements (in particular for DC DC buck converters), estimate of thermal performance, effect of pin holes and more. We would also like to power several modules in series, but this will strongly depend on the availability of an electrical test bed. In addition we will continue the development of the constant current source; specify and test powering circuitry for the 130 nm ABCnext frontend prototype (which might well be submitted in 2010 only); continue our work on grounding and shielding, DCS and AC LVDS coupling. 63

64 Goals and concerns for next six months, WP6 Goals Detailed comparison of the three serial powering schemes Converge on the specifications for a protection scheme Contribute to testing of the 30 module stave (at LBNL) Provide input to the 2. iteration of the constant current source Characterize a module with SP and DC DC Get serial powering and integrated DC DC blocks into a 130 nm ABCnext prototype Concerns RAL manpower is marginal due to SCT commissioning, international power coordination tasks and STFC restructuring. The detailed comparison between schemes partially relies on a student. The availability of a functioning module might be delayed. The time scale for design and submission of the 130 nm ABCnext prototype is open. Dependence on non UK collaborators 64

65 3.6.3 WP6 financial statements Approved (excluding contingency) Transfers Actual spend to end Feb 08 Projected spend this year Actual spend (2+3) Projected spend (2+4+5) Actual (6-1- 1a) (1) (1a) 2007/ /2009 (3) (4) 2009/ /11 (6) (7) University Staff Effort Costs* Liverpool University Sub-Total STFC Lab Costs RAL PPD RAL Technology STFC Lab Sub-Total Equipment Travel Other Directly Allocated costs (eg consumables) Total (Excluding VAT and WA) Working allowance VAT WP6: Finance Summary (all figures in k) Actual spend in previous years (2) Current year 2008/09 Latest estimate of future requirement (5) Total Variance Projected (7-1-1a) Total (including VAT & WA) Contingency (Held by STFC) 1 Excluding Working Allowance and VAT * The University staff effort recorded in this table should be the 80% amount STFC pays, including academic time Use of columns: (1) = The amount approved by STFC (1a) = This column should be used to show any virements between headings, for example when Working Allowance is used, the amount should appear as a debit in the WA row and then credited to the relevant row (2) = The actual spend in previous financial years, by year (3) = The actual spend in the current financial year up to the most recent quarter (4) = The total projected spend for the current financial year, including any expenditure so far (ie actual spend this year plus predictions of remaining spend this year) (5) = Projected spend for the remaining years (6) = The actual spend so far (7) = Projected spend over the whole duration of the project (ie actual spend so far plus predictions of remaining spend to project completion) The variance columns show the difference between the actual and projected amounts and the approved amount. 65

66 3.6.4 Summary of major deliverables and their financial allocation Deliverable and description Financial allocation from WP6 (kgbp) Project phase Conceptual power system design Specification, design or characterization of custom building blocks for SP including: ABCnext SP blocks, SPi, constant current source Final validation of SP concept using an electrical stave Finish implementation of SP for 0.13μm ABCnext design 65 k 100 k 65 k 40 k Current (R&D) Current Future Future The costs for WP6 are effort dominated Resource Usage NAME INSTITUTE Affolder A A Liverpool Jeff Bizzell/Analogue Eng. RAL 10 Richard Holt RAL John Matheson RAL Giulio Villani RAL Marc Weber RAL RAL TD RAL TD Total Grand Total WP6 milestone plan Table 6.1: Milestones achieved in the last six months Milestone Work Milestone Target Date Status No. Package M6.3 WP6 Characterization of 1. SPi 10 Feb 09 Complete chip prototype M6.4 WP6 Characterization of 0.25 μm ABCN25 SP blocks 4 Mar 09 Complete 2007/ / / /11 Table 6.2: Overall Milestone List for WP6 Milestone No. Work Package Milestone M6.1 WP6 Conceptual power system design M6.2 WP6 Finish implementation of SP for 0.25μm As at Sep 08 As at Mar 09 (Changes in Bold) Delay due to UK Jan 08 Complete N N 6 Apr 08 Complete N N Other Collaborators Affects Critical Path? See Note Reason for change 66

67 ABCN25 design M6.3 WP6 Characterization of 1. SPi chip prototype M6.4 WP6 Characterization of 0.25 μm ABCN25 SP blocks M6.5 WP6 Specs and testing of constant current source M6.6 WP6 Specs and design contribution to protection electronics (discrete components) M6.7 WP6 Final validation of SP concept M6.8 WP6 Supermodule powering scheme validated M6.9 WP6 Finish implementation of SP for 0.13μm ABCnext design 10 Feb 09 Complete 4 Mar 09 Complete 31 Jul 09 4 Aug Jan Feb 10 N Y N Electrical test bed delayed 23 Aug 10 1 Nov Gantt chart The gantt chart and M 6.7 has been slightly updated to reflect the delayed production of an electrically functional supermodule. It is not unlikely to recover the delay and thus we chose to leave the following section of the gantt chart unchanged. 67

68 3.7 WP Commentary on project to date Overview The critical issues facing the engineering have been further clarified during the six months being reported. In particular the major design issues have been identified which are associated with what has come to be called the hard bonded stave design. A hard bonded stave structure has no sliding components. Previous designs were described as glued, however the compound used, CGL, is not elastic, but flows under sufficient stress. The UK has pursued a hard bonded option, and major progress has been made in both identifying the problems with such a structure, and finding solutions to the problems. Important progress has also been made in the areas of material properties, cooling system parameterisation; engineering of a stave design; FEA of structures, preparation and use of measurement systems. During the period the test beam work required has been crystallised into a single programme and resides under the umbrella of WP7. The OsC requested that the organisational structure of WP7 be reported, it is shown below. The balance of the report is broken down according to this structure, except for the first section, which is a brief report on the WP7 goals from the last report. OsC Project Leader, Project Engineer Tapes (Weidberg) WP7 Manager, Project Engineer Tracker WP Managers Meetings WPx Test Beam (Casse) Material Properties (Canfer) Weekly WP7 Meetings 4/year Physical WP7 Meetings Test Box Meetings Stave Design (Wilmut) Thermal Measurements (Bates) Design Meetings Materials meetings Etc.. Prototype manufacture (Jones) Cooling Systems (Viehhauser) Pipe Studies (French) The web site: is the main mechanism for tracking the ongoing work of the group. 68

69 Basic Design The report relates to the work directed at the design shown below. The support structure consists of two skins of zero CTE carbon fibre separated by a honeycomb carbon. Cooing is achieved with an embedded cooling pipe which makes a U turn at one end of the stave. The pipe is glued into a pocofoam strip, which acts as a heat spreader, joining pipe to skins. Cooling of detectors, which are glued to the skins is via conduction to the embedded pipes. The structure is hard bonded, meaning that there are no sliding joints. To relieve stresses in pipe and foam, options are under consideration, which include wiggling the pipe, the stave core may also be made thicker. The read out interface, SMC, is located at one end of the stave on the section which stands proud of the main stave structure. The stave is attached to the surface of a support cylinder via one edge with a locating and locking mechanism. Figure: Cut away drawing showing basic stave construction. Stave design process: In the last OsC submission it was reported that the design process that had converged on a mechanically simple stave design. Over the last 6 months the concept has been iterated and improved. The thermal stress distribution within the stave has been investigated, and attempts made to design a robust low stress structure with rigid adhesives. There has also been extensive modeling of the thermal performance, considering various options for the on stave thermal systems. Key to this analysis has been numerous mechanical material properties measurements which has involved a significant portion of the collaboration. These results were used to compare FEA with mechanical results on small prototypes (the construction of which has also been developed). In addition to this detailed analysis the on cylinder edgemounting mechanism has been refined and servicing and integration scenarios examined. The key problem discovered is that in a hard bonded design, the stress levels in the materials after cool down is closer to the yield values than is desirable. Significant 69

70 effort has been expended attempting to understand ways to ameliorate this problem, with some success. Convergence on a viable design remains one of the major goals for the coming period. End plate Central boss Outer tube End stopper Tapered tap Tapered cup Latest version of the locking mechanism, evolved from first tests. Thermal contraction measurement: Past experience of using FEA to predict thermal contraction suggests that in some case it will not be a reliable predictor of behaviour. In particular, stave contraction, where composite materials are the main structural element, will be unlikely to be accurately predicted. Techniques for directly measuring thermal contractions have been and are being developed. These include the use of ESPI and an extensometer measurement from tensile test applications, inside an environmental chamber. The latter has already been employed with the result that the thermal contractions of the stave are greater than FEA predicts by a factor of two and efforts are underway to understand this in detail. Materials Properties Materials properties are required as inputs for finite element modelling packages. Due to the low stresses in service, and novel nature of some of the materials, relevant properties are not available. Mechanical and thermal measurements are being conducted by Glasgow, Liverpool, QMUL and at RAL. The materials being measured are carbon fibre composites, carbon foams, carbon carbon and bonding materials. Thermal Measurements During the last 6 months the in plane thermal measurements of materials has been applied to poco foam and TPG over a wider range of temperatures down to 20C. The value of thermal conductivity of poco foam measured at 20C was 60 Wm -1 K -1. This is in the process of being measured as a function of temperature. The thermal isolation of the apparatus at Glasgow has been improved with the use of a second thermal shield and better thermal anchoring of wires that enter the apparatus. This has allowed the apparatus to be operated down to 50C, which covers the operational 70

71 range of the experiment. The thermal conductivity of the materials of interest will be re measured over this larger temperature range in the coming month. CGL (an AIT product) which is the compliant thermal grease replacement suggested by LBL for the fixation of the poco foam inserts into the stave has been purchased. The thermal conductivity of this (and other adhesives) has been measured with the Liverpool line source conductivity probe method. The value of thermal conductivity obtained was / 0.08 Wm -1 K -1.This is lower than the manufactures quoted value of 4.1 Wm -1 K -1. However the CGL was not cured to 80C for the Liverpool measurements. Several adhesive samples, including CGL, have recently been taken to Karlsruhe, Germany, to be irradiated with protons to observe the effect of irradiation on the thermal performance of the gel. The thermal conductivity of CGL is in the process of being measured in the now commissioned thermal interface tower of Glasgow and the existing apparatus at QMUL. This will allow the thermal conductivity to be measured as a function of pressure and temperature for Cu CGL and carbon carbon CGL interfaces. Mechanical Measurements Since the stresses on a detector structure are low, and stiffness is a design driver, mechanical measurement efforts have focussed on methods to measure modulus at low stress. Little is available in the literature, indeed a key reference dates from Initial measurements showed that the available mechanical testing machine did not have sufficient axial alignment to measure modulus at low stress, below 10MPa. A jig has been designed and fabricated to provide this alignment and it has been demonstrated that it is capable of measuring modulus of aluminium at these low stresses. Efforts are now moving to carbon fibre composites and novel materials such as pocofoam. Camera y = 97152x Stress MPa Strain The Figure shows a measurement of the Young s modulus of carbon fibre composite at low stress using the new non bending jig. 71

72 Tapes One of the critical issues for the electrical performance of the tapes is the use of multi drop TTC in order to minimise the number of lines from the SMC to the hybrids. A prototype electrical stave was designed to study this. This stave has now been constructed and bonded down to a carbon fibre base plate. The stave contains 24 dummy hybrids in order to test the electrical transmission performance. The driver uses an M LVDS chip which is designed for this type of application. The first tests looked at the eye diagrams at the receivers and gave very encouraging results. The next steps will be to use a data loop back and an FPGA board to measure the Bit Error Rate (BER). The BER will be measured using both LVDS and M LVDS receivers. Further tests will look at the effects of feedback to allow for non balanced codes. A first look at the layout for a fully functional stave tape has been made. After these studies with the first prototype are completed, the design of a fully functional prototype tape will be made. Cooling & Pipes Work on R&D on pipes and fittings has continued. There is now a facility at RAL, which allows pipes to be tested over a large range of pressures (vacuum to 150bara) and temperatures ( 80 C to >50 C). Vacuum and overpressure leak tests can be performed and pipe diameter variations under pressure measured. Qualification of prototype pipes has started using this setup. Currently the focus is on thin wall stainless steel tubes, but Titanium pipes will be added to the study, due to increasing interest in this material in the collaboration. R&D on pipe bending has continued. The limits on bend radii for different pipes while maintaining the OD of the pipe is being examined. The facility to optically measure the OD along the bend has been added during the reporting period. Pipe joining using Orbital Tungsten Inert Gas (TIG) welding is under study, for use in joining thin wall pipes. The requirements are non standard, but high quality welds have proven possible with high repeatability. Quality assessment procedures, based on industry standards, to evaluate these welds have been developed. This technology is proposed for the final integration of the sub detectors into the ID so it will be used next to sensitive readout electronics. It is therefore necessary to demonstrate that it will not affect the electronics and a programme to test this will be starting soon. Preliminary trials from the Micro Focus CT using a Swagelok welder show excellent results: the structure of the material is little affected within the heat affected zone and the concentricity and uniformity of the tube wall are still within the manufacturing tolerance. The specification of requirements for the pipe work and fittings for the future cooling system is being driven from the UK (ATU SYS ES 0003), as well as the requirements for the final cooling system itself (ATU SYS ES 0004). Both sets of requirements are well advanced and discussed within the collaboration. The system requirements are expected to go through an Upgrade Steering Group approval soon. Work on measuring two phase fluid dynamics properties such as pressure drops and heat transfer coefficients between the evaporating coolant and the pipe wall is continuing. Completion of the measurements of these properties for C3F8 has been 72

73 held up by problems with the plant, where the low flow operation caused cavitation in the pump. This, and other, problems have been recently overcome and this programme should conclude soon, after which CO2 will become the exclusive focus of the groups. Prototype Fabrication Several small scale prototype structures to examine specific properties of staves have been fabricated. These have been aimed at understanding stave contraction, the quality and performance of the glue bond between pipe and poco foam and the Destructively tested pipe in pocofoam thermal performance of a CO2 cooling system used with a stave like structure. Based on the results of initial prototyping, improved techniques and tooling are being developed and a new round of prototype manufacture is underway. The aim is to provide measurements which will inform thermomechanical FEA allowing the stave design to be firmly grounded in engineering reality. In addition, experience will be gained on the manufacture of stave like sandwich structures which will be invaluable as the development of tooling to manufacture full length staves proceeds. Irradiation at the CERN PS Several aspects of the overall tracker upgrade programme require a dedicated irradiation programme. During the reported period, this programme has been rationalised and included in WP7 as a package of work. This is because the programme will include testing radiation hardness of materials and the evaluation of stave prototype systems. The key work during the period being reported has been to assess the degree to which reuse of old equipment is possible and development of a plan for the future work. For the upgraded tracker prototypes, the cooling plant requires upgrade. A DAQ system needs development to allow evaluation during irradiation. Project Engineer: In his role of project engineer, Ian Wilmut is now an active member of the upgrade project office with the role of maintaining the whole upgrade schedule. This gives a unique perspective on the direction of the project and assists UK groups in understanding the various factors driving the project. Given Ian s mechanical background his placement within the PO keeps him very close to the CERN engineering effort of the upgrade, in which he is actively involved. As part of this work he has supervised development of the service module concept, in which the services for cylinders are created as separate objects, with a view to massively simplifying the installation process: 73

74 View of a section of the proposed Service Module Goals and concerns from previous period, WP7 Goals Status Verify operation of locking mech Completed Consolidate materials properties Ongoing, significant progress made Complete FEA on stave design Significant progress, ongoing Build First stave prototype Completed Complete SMC, tape, lock mech design Completed HTC in CO2, compare with simulation Major progress, ongoing Concerns Status Plans for next 6 months During the next six months the WP7 group expects to make major progress with several of its ongoing efforts. In particular it is expected that the work on measuring HTC for C3F8 as a function of several variables will be concluded and the CO2 programme become advanced; low stress measurements of material properties will be understood and an acceptable level of comprehension of pocofoam, skins and cores for staves achieved. It is also expected that measurements of adhesives, both before and after irradiations will be completed, including developing an understanding of their application to foam structures. The design of a hard bonded stave, complete with SMC location, mounting scheme and tape design will be completed. The design will be studied for both thermal and mechanical performance using FEA. Prototype staves with a range of functions will be built over the next six months. These will include short sections built with accurate materials and longer staves which may be fabricated using representative materials. The former will be required 74

75 to understand thermal and material properties and the latter will allow both mechanical measurements and fabrication methods to be developed. The CERN PS irradiation facility will be built, in particular the cooling system for modules in the beam will be designed and fabricated. Goals and concerns for next six months, WP7 Goals Fabricate short length prototype staves with accurate materials Fabricate long length prototype staves with representative materials Complete HTC measurements with C3F8 and C02 Measure thermal performance of proposed structures; incl. CGL, other adhesives Complete a stave design and FEA, including mounting mechanism Complete test box system for tests in beam Make preliminary measurements of stave prototype mechanical, thermal properties Produce realistic tapes for use in prototype stave programme Concerns Viability of low mass hard bonded stave design Summary of major deliverables and their financial allocation Deliverable and description Financial allocation from WP7 (kgbp) Project phase Fully developed bus cable 35 Current Complete design of thermal path including all testing, evaporative facility and baseline component qualification Stave structure design and development, including prototyping Pre production stave structures (3 off) for mechanical, thermal and electrical testing 80 current & future 60 Future 20.5 current WP7 financial statements It is notable that RAL PPD costs are significantly down on our previous estimates, (which were lower than our original allocation). This is due to recruitment constraints within PPD. It was previously expected that within the course of the project PPD would be able to address this. Now, two thirds of the way through the grant, it is apparent that RAL will be unable to recruit soon enough to spend this money before April It is hoped that recruitment will be possible in September 2009, if this is the case it would be possible to make use of this effort only if this effort is rolled forward into 2010/11. The data presented below does not do this, as it would give a false sense of the certainty. 75

76 University Staff Effort Costs* Approved (excluding contingency) Transfers WP7: Finance Summary (all figures in k) Actual spend in previous years (2) Current year 2008/09 Actual spend to Feb 09 Projected spend this year Latest estimate of future requirement (5) Actual spend (2+3) (1) (1a) 2007/ /2009 (3) (4) 2009/ /11 (6) (7) Liverpool Sheffield Cambridge QMUL Oxford Glasgow Lancaster Total Variance Projected Actual (6-1- Projected (7- spend (2+4+5) 1a) 1-1a) University Sub-Total 1 1, , STFC Lab Costs RAL PPD RAL Technology STFC Lab Sub-Total 1, Equipment Other Directly Allocated costs (eg consumables) Travel Total (Excluding VAT and WA) Working allowance VAT Total (including VAT & WA) 2, , ,164 2,697-1, Contingency (Held by STFC) 1 Excluding Working Allowance and VAT * The University staff effort recorded in this table should be the 80% amount STFC pays, including academic time Use of columns: (1) = The amount approved by STFC (1a) = This column should be used to show any virements between headings, for example when Working Allowance is used, the amount should appear as a debit in the WA row and then credited to the relevant row (2) = The actual spend in previous financial years, by year (3) = The actual spend in the current financial year up to the most recent quarter (4) = The total projected spend for the current financial year, including any expenditure so far (ie actual spend this year plus predictions of remaining spend this year) (5) = Projected spend for the remaining years (6) = The actual spend so far (7) = Projected spend over the whole duration of the project (ie actual spend so far plus predictions of remaining spend to project completion) The variance columns show the difference between the actual and projected amounts and the approved amount. 76

77 3.7.5 Resource Usage Table of staff %age FTE WP7 WP7 NAME INSTITUTE Allport P P Liverpool 5 10 Jackson J N Liverpool 5 10 Vossebeld, J Liverpool Affolder A A Liverpool Jones T J Liverpool Sutcliffe P Liverpool Carroll J L Liverpool Muskett D A Liverpool Paganis Sheffield French Sheffield Carter Cambridge Carter QMUL Martin QMUL Beck QMUL Gannaway QMUL Morris QMUL Mistry QMUL C.Buttar Glasgow V.O'Shea Glasgow C.Buttar Glasgow R.Bates Glasgow A.Cheplakov Glasgow F.McEwan Glasgow J.Melone Glasgow Lau, W Oxford Nickerson Oxford Drumm Oxford Viehhauser Oxford Wastie Oxford Yang Oxford Weidberg Oxford M Tech RG Oxford Jones Oxford Handford (DO) Oxford Tacon, J Oxford Hawes, B (RG) Oxford Dawson (+Halliday) Oxford P Lau (+Ottewell) Oxford Fopma Oxford E Tech Oxford Roger Jones Lancaster Harald Fox Lancaster Ian Mercer Lancaster Tad Komorowski Lancaster 6 40 Richard Apsimon RAL 10 Martin Gibson RAL / / / /11 77

78 Jeff Bizzell/Analogue Eng. RAL 40 Richard Holt RAL John Matheson RAL Marc Weber RAL Steve McMahon RAL 10 0 Mike Tyndel RAL Stephen Haywood RAL 5 5 ATLAS new physicist RAL 0 10 Ian Wilmut RAL TD RAL TD RAL TD Total Grand Total WP7 milestone plan Table 7.1: Milestones achieved in the last six months Milestone Work Milestone Target Date Status No. Package M7.7 WP7 Custom connector decision 18/01/08 Completed* M7.8 WP7 Connector On hold* M7.9 WP7 Connector On hold* M7.18 WP7 Cooling pipe pressure tests 04/05/09 Complete *The selection of a stave design has largely made redundant the concept of a custom connector, so this programme is on indefinite hold until such time as it may become relevant again. It is replaced by a selection and evaluation of commercial custom connectors, a process which requires approximately the same resource. 78

79 Table 7.2: Overall Milestone List Milestone No. Work package Milestone As at Sep 08 As at Mar 09 (Changes in Bold) Delay due to UK? Other Collaborators? Affects Critical Path? See Note Reason for change M7.1 WP7 Open evap feasibility 29/06/07 Complete N N M7.2 M7.3 Select Overall Concept 21/12/07 Complete N N M7.4 Select Bascit Macro 23/12/08 Complete N N support M7.5 WP7 Supermodule Geometry 21/12/07 Complete N N M7.6 International supermodule 8/6/08 Complete N N M7.7 WP7 Custom connector decision M7.8 WP7 Connector demonstration 18/01/08 complete Y Y 16/01/09 Indefinite delay M7.9 WP7 Connector design 26/06/09 Indefinite Y Y delay M7.1 WP7 Power Tape abl vs etch 30/05/08 Complete N N 0 M7.1 Select Cu/AL 21/5/09 Complete N N 1 M7.1 2 Establish bandwidths 14/01/09 14/01/09 N N M7.1 3 M7.1 4 M7.1 5 Demonstrate Realistic Tapes 01/03/09 01/03/09 N N WP7 Tape connection 29/08/08 Complete N N method WP7 First tape pre series 18/12/09 18/12/09 Y N M7.1 6 M7.1 7 M7.1 8 M7.1 9 WP7 Poco foam, grease, TPG measurements 29/5/09 29/5/09 N N HTC, pressure drop 04/05/09 04/05/09 N N measurements Cooling pipe pressure 04/05/09 complete tests Welding test review 17/08/09 17/08/09 M7.2 0 M7.2 1 M7.2 2 M7.2 3 WP7 Prototype supermodule 19/12/08 15/06/09** N Y WP7 WP7 WP7 Second proto supermodule 26/06/09 11/12/09 Final module thermal 03/07/09 12/06/09 design Complete Supermodule 04/01/10 29/07/10 evaluated **Early prototypes exist, so milestone M7.20 could be listed as complete, however the plan is to build something larger as the first prototype and so this date is delayed. 79

80 3.7.7 Gantt chart 80

81 3.8 WP8: Data Acquisition and Read-out Summary on project to date A DAQ for support of front-end chip evaluation and module R&D During the last 6 months, the work done in the context of this work package had been focused on providing a DAQ for reading out the ABCN25 front end chip upon its arrival. Prior to the release of the ABCN25, the VHDL code describing the design was ported to an FPGA development board, allowing for a hardware emulation of the chip to test and debug the DAQ system. This approach has proved very useful, both to get the appropriate hardware such as cables and patch panels in place, as well as the convergence towards a working version of the MuSTARD firmware and SCTDAQ software prior to the arrival of the chip. Thanks to the joint effort, the MuSTARD/SCTDAQ system was operational as soon as a preliminary, untested batch of ABCN25 chips were made available in October Upon their release, ABCN25 chips were fitted to single chip test boards as well as prototype hybrids (see Figure 22). Both were successfully read out using the MuSTARD and updated SCTDAQ software, and first results were gathered in time to be presented at the ATLAS Upgrade Workshop in November Figure 22: Prototype hybrid (see WP4 report) attached to SCTDAQ adapter board Since then, the MuSTARD/SCTDAQ system has successfully been used to establish the characteristics of the ABCN25. Currently, the performance of multiple hybrids carrying up to 20 chips each, coupled to a large sensor is being investigated. Also, the system is used in conjunction with a National Instruments commercial acquisition system to test the remainder of the ABCN25 chips prior to dicing. One of the new features of the ABCN25 chips is the capability of 80MHz data transfers. Although MuSTARD is limited to 40MHz readout speed, FPGA development boards are used to intermediately buffer incoming 80MHz data streams, retransmitting them at 40MHz. This additional hardware facilitates access to the full feature set of the ABCN25 design in legacy mode. 81

82 A demonstrator DAQ for supermodule readout Much progress has also been made since the last report on the realisation of the necessary hardware for a DAQ to read out multiple modules, leading towards DAQ suitable for read out of a supermodule/stave. Our US collaborators have recently designed a High Speed Input/Output (HSIO) board for readout of Linac Coherent Light Source (LCLS) detectors. We are collaborating with them to specify the modified requirements for a prototype board for multiple module readout. A first prototype of the HSIO board arrived in the UK in January, with an adapter board following in February. Both are displayed in Figure 23. The HSIO board seems to offer the capabilities necessary to achieve the goals set for this work package, see Paragraph Figure 23: High Speed Input/Output (HSIO) board (top), with adapter board (bottom). 82

83 Goals and concerns from previous period, WP 8 Goals Status MuSTARD/SCTDAQ Upgrade for Successfully completed on time. full ABCN25/MCC based prototype modules support Definition of Supermodule DAQ specifications, components and market survey on candidate commercial components Concerns Continued buildup in availability of DAQ experts Plans for next 6 months Superseded by High Speed I/O board developments (see text) Status Concern remains, mainly due to delays in the startup of LHC experiments. With the MuSTARD/SCTDAQ system now operational, only minor debugging remains before the upgraded SCTDAQ software is made available to the wider international collaboration. As the bulk of the ABCN25 chips are not yet tested and diced, it is preferred to thoroughly debug the software prior to release. Another key ingredient on the development roadmap of a detector module is the Module Controller Chip (MCC). However, the MCC development has suffered further delays due to as yet incomplete specifications by the collaboration. To avoid halting the progress in development of modules and staves because of this, our US colleagues initiated the design of the Buffer Control Chip (BCC), an ad hoc solution providing the very basic MCC functionality. The BCC is designed by a US/UK working group, and is due for submission at the end of March to become available in June this year. Both the MCC and BCC are specified to transmit data at 160MHz rate, using a specific data format. As these characteristics exceed the MuSTARD capabilities by far, new DAQ readout hardware would be necessary. The prime candidate for the successor to MuSTARD would be the High Speed Input/Output (HSIO) board, depicted in Figure 23. As the name suggests, the HSIO board features many high speed interfaces, both on industry standard connectors (XFP cages) and on a large user connector. It contains a large, powerful FPGA (Xilinx Virtex4 FX60), capable of driving links up to speeds of 10Gbits/s. USB and conventional Ethernet interfaces provide less sophisticated, relatively easy to use, communication channels. The HSIO board is intended to connect to an adapter board, which provides the interface to the actual hardware that needs reading out. This adapter board (a UK/US collaboration) contains interfaces to the existing single chip test boards and hybrid driver boards, as well as dedicated LVDS signal drivers to send and receive data to/from a stave. By design, re issuing an adapter board is relatively straightforward. If required by the application, specific adapter boards can now be fabricated without major investments in manpower and/or money. 83

84 The HSIO, equipped with the appropriate adapter board, can serve all communication channels necessary to operate a stave, including sampling data of many channels at high speed. Moreover, new versions of the HSIO, adapted to meet future requirements, are feasible and would take relatively little time. The above characteristics make the HSIO a prime candidate for the central component of the DAQ system required to reach the goals as set for this workpackage. Consequently, UK firmware and software effort will now be concentrated on the HSIO to provide the functionality of SLOG and MuSTARD, and interface with the SCTDAQ software framework. The stave controller will likely be based on components developed and selected by the Gigabit transceiver (GBT) and versatile link projects. As the outlook for these projects is to make use of similar commercial interfaces as present on the HSIO prototype, it appears well prepared for providing the interface to future versions of the stave. Looking further ahead, a performance assessment using the HSIO provides a reference benchmark feeding into the DAQ design studies concluding WP8. All the above developments hinge on the availability of WP8 software and firmware effort. This is largely provided by experts who are also instrumental in the start up phases of the ATLAS SCT. Thus the ongoing delay in LHC collisions is a continuing cause for concern for WP8. Much progress has been made recently, but the pace of future development is vulnerable to the availability of these experts. Goals and concerns for next six months, WP8 Goals Development of firmware for HSIO to read out ABCN25 chips and BCC outputs. Modifications of SCTDAQ software for sending and receiving data from BCCequipped modules using the HSIO board Concerns Continued build up in availability of DAQ experts, in particular beyond May

85 3.8.3 WP8 financial statements Approved (excluding contingency) Transfers Actual spend to Feb 09 Projected spend this year Actual spend (2+3) Projected spend (2+4+5) Actual (6-1- 1a) (1) (1a) 2007/ /2009 (3) (4) 2009/ /11 (6) (7) University Staff Effort Costs* Liverpool Cambridge UCL University Sub-Total STFC Lab Costs RAL PPD RAL Technology STFC Lab Sub-Total Equipment Other Directly Allocated costs (eg consumables) Travel Total (Excluding VAT and WA) Working allowance VAT WP8: Finance Summary (all figures in k) Actual spend in previous years (2) Current year 2008/09 Latest estimate of future requirement (5) Total Variance Projected (7-1-1a) Total (including VAT & WA) Contingency (Held by STFC) 1 Excluding Working Allowance and VAT * The University staff effort recorded in this table should be the 80% amount STFC pays, including academic time Use of columns: (1) = The amount approved by STFC (1a) = This column should be used to show any virements between headings, for example when Working Allowance is used, the amount should appear as a debit in the WA row and then credited to the relevant row (2) = The actual spend in previous financial years, by year (3) = The actual spend in the current financial year up to the most recent quarter (4) = The total projected spend for the current financial year, including any expenditure so far (ie actual spend this year plus predictions of remaining spend this year) (5) = Projected spend for the remaining years (6) = The actual spend so far (7) = Projected spend over the whole duration of the project (ie actual spend so far plus predictions of remaining spend to project completion) The variance columns show the difference between the actual and projected amounts and the approved amount. 85

86 DAQ for R&D support Due to the delays mentioned in the first report submitted to the oversight committee, WP8 spending is not yet on par with the initial projections. Also, as existing equipment is being re used for providing a DAQ in support of module R&D, the cost incurred so far has been minor. As mentioned in the previous section, the specification definition of the MCC is delayed, and consequently its prototyping is also pushed forward. The cost of the MCC substitute, the BCC, will be covered by our US colleagues, although a minor contribution could be foreseen once the chips are ready to be released in July As a result, the 5k projected spending related to the MCC prototyping and read out, previously allocated to the period is transferred to DAQ development R&D The prototype HSIO, and its adapter board were purchased last year, and it is expected that the required power supplies, cabling etc. will be bought before the financial year ends. The total expenditure for this prototype system is about 4k. Although the HSIO is a very cost effective solution for our DAQ needs, multiple systems will have to be bought for installation at the various UK institutes subscribed to this work package. A complete HSIO based system, including a DAQ PC, is expected to cost about 5k. With 4 institutes participating this would add up to 20k. Additional cost would be incurred when the adapter boards are revised to meet application specific requirements. When reading out the final stave prototype, it may be desirable to run multiple HSIO boards in parallel, requiring high performance switching equipment to route data from the HSIO boards towards dedicated DAQ PCs. As the stave prototype should become available in , the projected spending is adjusted accordingly. The HSIO has provisions for multi Gbps interfaces on board. This would allow for the incorporation of GBT prototypes, providing an interface to a possible future Super Module Controller. These developments are expected to get underway in 2010, so budget is allocated accordingly. 86

87 3.8.4 Summary of major deliverables and their financial Deliverable and description Financial allocation from WP8 (kgbp) Project phase SCTDAQ for readout of ABCN25 test structures HSIO based DAQ for readout of ABCN25+BCC based modules Multiple HSIO based demonstrator DAQ systems for readout of many modules Modular, scalable DAQ system for stave readout 7 Current 12 current & future 16 Future 29 Future Resource Usage Table of staff %age FTE WP8 WP8 NAME INSTITUTE Greenall A Liverpool Carter Cambridge Hommels Cambridge Lester Cambridge Goodrick Cambridge Hill Cambridge Robinson Cambridge Shaw Cambridge Barr Oxford Butterworth UCL Wing UCL Attree UCL Warren UCL Postranecky UCL Bruce Gallop RAL Peter Philipps RAL Total Grand Total / / / /11 87

88 3.8.6 WP8 milestone plan Table 8.1: Milestones achieved in the last six months Milestone Work Milestone Target Date Status No. Package M8.1 WP8 Complete software modifications for ABCN25 28/09/2008 Successfully completed Table 8.2: Overall Milestone list for WP8 Milestone No. Work Package Milestone M8.2 WP8 DAQ ready for full ABCN25/MCC testing M8.3 WP8 Assessment of HSIO performance as a DAQ benchmark M8.4 WP8 Readout of prototype stave M8.5 WP8 Complete design study of scalable DAQ for readout of multiple staves As at Sep 08 As at Mar 09 (Changes in Bold) Delay due to UK Other Collaborators Affects Critical Path? See Note Reason for change 30/04/ /07/2009 N Y N Delay in MCC development 30/10/ /10/2009 N N 30/03/ /03/2010 N N Yes 30/08/ /08/2010 N N Gantt chart As the HSIO is the prime candidate hardware platform for the development of the prototype DAQ, the milestone reading previously Technology and component choice for DAQ design study has now been converted into M