Saga of the IBL Staves Franck Cadoux on

Saga of the IBL Staves Franck Cadoux on behalf of the ATLAS IBL engineering team (and Eric Vigeolas in particular) 1

Saga of the IBL Staves q q q Topics to be discussed IBL project & its challenges Stave design and its history… Stave manufacturing & Testing From a bare object up to integration What we’ve learnt…(still learning!) 2

Saga of the IBL Staves IBL project & its challenges Beam pipe Current PIXEL detector IBL location (envelope in blue) PIXEL after de integration @ CERN 3

Saga of the IBL Staves ü ü ü IBL project & its challenges IBL project is really short in time (driven by LS 1) Beam pipe is now part of the IBL Envelopes are tight ( ID=56 mm < IBL < OD=85 mm…see next!) Tricky operations in the pit (insertion thru 7 m long pipe called IST) 14 Staves (detectors) cooled @ -40°C with an evaporative CO 2 system IST (ID 85 mm) IPT (ID 58 mm) 4

Saga of the IBL Staves IBL project & its challenges IBL overview ü 14 Staves (or Local support) arranged cylindrically around the IPT (one of the two key structures, or Global supports) ü The IST (or IBL Support Tube) is a 6. 6 m long cylinder tied to the PIXEL ü The IPT will position the Staves and support their services IST cutout (to look at the IBL package) IPT with services 5

Saga of the IBL Staves IBL project & its challenges IST and IPT manufacturing challenges ü ü ü IST and IPT manufacturing (both carried out in Seatlle) Material is a prepreg of K 13 C (same as for Staves!) Driven by weight reduction (extremely rigid CFRP… fiber around 900 Gpa) IST is 0. 45 mm thick, 6600 mm long, 5 segments jointed in a second step IPT is 0. 45 mm thick, except in its central area (Staves)… 0. 325 mm thick!! IST prototype segment IPT proto assembly with rings 6

Saga of the IBL Staves Stave design and its history Main design requirements (Stave) ü ü ü As usual the outstanding features are: Lightweight(X 0), Stiffness, Stability Support 2 types of modules to be cooled down to -40°C all way long (650 mm) Stave fixation to “Global structure” driven by the IBL insertion into the pit! Radiation hardness for every material (350 MRad at maximum) Overall Stave planarity should be within +/- 150 µm (see later…) Cable bus (flex) bonded to Stave due to tight envelopes e Stav IPT 7

Saga of the IBL Staves Stave design and its history IBL Stave evolution over the last 4 years (coordinated by CPPM Marseille) ü Based on the PIXEL experience, several concepts have been studied… ü … on CFRP type for omega shape (M 60 vs K 13 C), carbon foam (Poco vs K 9), cooling pipe (CFRP vs Titanium) and the stave shape! ü Everything led by the material optimization, reliability… and manufacturing 8

Saga of the IBL Staves Stave design and its history IBL Stave Design and details… A bare stave (w/o module) is 748 mm long 3. 5 mm thick, and weighs about 26 g Made of 12 parts glued together K 9 foam choice after LBNL investigations: Low density (0. 22 g/cm 3)… but good Thermal conductivity (40 W/m. K) ü Stave end parts made of peek CF (inspired by PIXEL staves…) ü CFRP Face plate inserted between foam and Module Prevent TH grease penetration, increase the stiffness, and allow for HV protection (parylene coating) ü ü 9

Saga of the IBL Staves Stave design and its history Some details on design… Developed by Wuppertal, gluing tests on 2 samples (between Ti pipe and foam)… Clearly shows glue penetration issues This led to some new ideas (face plate in K 13 C/RS 3)… thanks to developments by SLAC + Seattle (avoid the grease underneath module to be soaked up by carbon foam) Short length proto by SLAC (tested @ Uni. Ge) Layup of K 13 C facings (Omega and face plate) - [0°/90°/0°] - Overall thickness: 195µm (quite thin) K 9 CNC machined Tomography of K 9 plates from Allcomp 10

Saga of the IBL Staves Stave design and its history Some details on design… Thermal Grease is scraped over the module surface by means of mask and centering tools (see next) Tested without the face plate (grease tends to be soaked up or diffused into the foam pores) Grease being applied onto the Stave face plate + scraping over the mask… All those testes performed on prototypes led to the validation of the STAVE before mass production Result of thermal grease “footprint” when mask is taken out (in this way, the grease amount is well known) 11

Saga of the IBL Staves Stave manufacturing & Testing Manufacturing was done in Germany (led by Wuppertal with IVW)… ü An initial target of 32 bare Staves + 10 protos to be manufactured (a bit less to be loaded at the end of the history) ü Based on the PIXEL Stave experience (molds, counter molds, CNC machining) ü Graphite material used for molds … f foam ü …due to high temperature curing (@120°C)… o g n i n hi Remac ü Co curing was abandoned (material saving OK, but delimitation still an issue… further developments needed! ü Such a work (12 pieces to be assembled) is long process and difficult for mass prod. Tomography of first proto PIXEL shape legacy Now it’s a V-shape due to Flex bonding CFRP wrapping @ 12 bars pressure CFRP Omega at 195µm (good homogeneity) 12

Saga of the IBL Staves § Stave manufacturing & Testing Main steps of the Stave manufacturing…. K 13 C/RS 3 prepreg unrolling (for Omega + face plate shaping) K 9 foam machining (6 blocks per stave) End blocks machining (peek CF) Stave is complete 12 steps are needed to complete 1 bare Stave (post machining required some fine tuning…) 13

Saga of the IBL Staves § Stave manufacturing & Testing First lessons from the Stave “pre production” step (improvements on fabrication process) Strip on cooling pipes found on 2 staves vacuum bag removal during stave gluing process was suspected First staves prototypes showed cracks on Omega Machining process was changed successfully … no crack observed afterwards … 14

Saga of the IBL Staves § Stave manufacturing & Testing Thermal qualification (on “Stavelet” + Stave full length proto) ü Several dummy staves were tested to select the best design (thermal grease, glue…) ü 2 final staves equipped with Si heaters + CO 2 cooling Si heaters + thermal sensor ü TFOM (Thermal Figure Of Merit) < 15°C. cm 2/W (over full length) (30°C. cm 2/W is the maximum) ü 350 thermal cycles [-30°C; +40°C], 50 cycles [-40°C; +50°C], 10 pressure cycles [1; 66 bars] and pressure shock up to 150 bars! No TFOM change has been seen! 15

Saga of the IBL Staves § Stave manufacturing & Testing Some metrology results… ü An extensive campaign of metrology was done to record any change in planarity before and after thermal cycling (with and w/o Flex glued to the Stave) ü After Stave Module loading, thermal cycling is performed systematically to select things… so far, the deformations are close to the limits, but still OK! Metrology at Uni. Ge Staves with modules 16

Saga of the IBL Staves § Stave manufacturing & Testing Some detailed FEA were done to “help” designing the Staves ü The policy was to get several lab working on Stave FEA to X-Check results (CPPM, LAPP, CERN, INFN Milano, Uni. Ge) ü Both mechanical and Thermo mechanical FEA have been carried out (ANSYS, Abaqus) ü Tests on prototypes to get to the “MODEL calibration”… Flex Extensive FEA’s to validate the choice of the Face plate option (1 ply, 3 plies, a few cut out at module location…). For ∆T= -60°C, the ∆Zmax =15 µm… instead of 65 µm w/o face plate! No measurement of TH deflection has been done on proto 17

Saga of the IBL Staves § Stave manufacturing & Testing Comparison between Test and FEA (mechanics, simply loaded stave) ü Stave is held by its handling frame (used from early step to final integration) ü A concentrated loading is done at the ¼ of stave length (see below) Stave with Flex weights Stave onto the handling frame: “final” boundary conditions … Tends to lift up slightly loading er t Bett 15 han ence iffer %d 18

Saga of the IBL Staves § From a bare object to Integration The first phase is about “Module Loading” onto the stave… ü Aluminum jigs have been designed and successfully tested on the first protos ü Every step of the process is safely recorded in a Data Base (useful in case of pb) ü Precision is given by the jig thru the Handling frame Loading jig (main tool) Stave on its handling frame 19

Saga of the IBL Staves § From a bare object to Integration The main steps of the loading process… Now ready for final QA (electrical tests, thermal cycling, metrology…) Repair on module is still doable… but risky! 20

Saga of the IBL Staves § From a bare object to Integration onto the IPT @ CERN (building SR 1) ü Pretty long object (the cooling line brazing is done before integration)… 7 m long! ü A dedicated Integration Stand is under fabrication (Uni. Ge is responsible)… sort of multi purpose object to …Integrate, store, insert thru the IST the IBL package Brazing technique under testing @ CERN + Brazing Stand The Integration stand Multipurpose container (MPC) + Loading stand (central area) 21

Saga of the IBL Staves § From a bare object to Integration onto the IPT @ CERN (building SR 1) ü Stave are assembled first, then the services are connected …and tested! ü The challenge is to insert a stave with very close neighbors (< 0. 9 mm clearance… so envelope check will be an issue! Loading stand IBL _ MPC (6. 6 m long) Envelope checks Stave Handling frame Loading stand IPT 22

Saga of the IBL Staves § What we’ve learnt… Over the 4 years of manufacturing… ü Stave manufacturing is a lengthy operation difficult to “duplicate” among institutes (tooling, jigs, know how…)… so to be thought about for the next Upgrades! ü Material procurement was always tricky (K 13 C, foam, pipe…)…even if for the layup we took advantage of the IST-IPT business! ü So many pieces to glue together… but at the end, the result is pretty satisfactory (planarity, rigidity… and close to FEA predictions!). ü Main drawback is about the brittleness of the Omega + Face plate (due to very high modulus fibers)… special care applied to the post machining § Towards the next years… ü Still some challenges to be faced on Stave Integration (envelope issues) ü About K 13 C layup…is it really needed ? ? Seems to be “too much”… ü Think about even thinner cooling pipe, and smaller diameters? ? To save further material… ü Try to optimize the amount of glue… especially at cooling line joint ü … but the feeling that we are close to some current limits / mechanics…(material, glues, ? ) ü Optimize the links (number of fixations) to the Global structures (like IST) is challenging 23