LARP Internal Review of the MQXFA Mechanical Structure Design and Functional Requirements 7/16/2015 Dan Cheng, Helene Felice Development of the Shell-Based.

Slides:



Advertisements
Similar presentations
LQ Reaction & Impregnation Tooling Internal Review F. Nobrega, N. Andreev May 7, 2007.
Advertisements

EuCARD-HFM ESAC review of the high field dipole design, 20/01/2011, Maria Durante, 1/40 EuCARD-HFM ESAC Review of the high field dipole design Fabrication.
Dan Cheng H. Felice, R. Hafalia QXF Structures Video Meeting July 10, 2014 Coil Handling for Magnet Assembly Status Update.
S. Caspi, LBNL HQS Progress Report High Field Nb 3 Sn Quadrupole Magnet Shlomo Caspi LBNL Collaboration Meeting – CM11 FNAL October 27-28, 2008.
Helene Felice Joint LARP CM20 / Hilumi Meeting April 8 th to 10 th 2013 Napa, CA, USA HQ02 Assembly summary and Next steps.
LARP QXF 150mm Structure Mike Anerella / John Cozzolino / Jesse Schmalzle November 15, nd Joint HiLumi LHC-LARP Annual Meeting.
LHQ Coil Fabrication Experience and Plan Miao Yu 04/08/2013.
Coil Design and Fabrication Miao Yu Outline Introduction Coil Design –Coil Pole –Coil End Parts –Coil Insulation Coil Fabrication –Winding.
MQXFS Assembly Procedure J.C Perez 4 th Joint HiLumi LHC-LARP Annual Meeting November 17-21, 2014 KEK, Tsukuba N. Bourcey, D. Cheng, B. Favrat, H. Felice,
SQXF Shell & Yoke Assembly Procedure
H. Felice - P. Ferracin – D. Cheng 09/19/2013 Update on structure CAD model.
Helene Felice HQ02 Assembly status 02/28/2012. HQ02 layout COIL 16COIL 15 COIL 17COIL 20 VIEW LOOKING AT LEAD END A C BD Sector A Sector B Sector C Sector.
H. Felice - P. Ferracin – D. Cheng 09/11/2013 Update on structure CAD model.
DOE Review of LARP – February 17-18, 2014 Helene Felice P. Ferracin, M. Juchno, D. Cheng, M. Anerella, R. Hafalia, Thomas Sahner, Bruno Favrat 02/17/2014.
2 nd Joint HiLumi LHC – LARP Annual Meeting INFN Frascati – November 14 th to 16 th 2012 Helene Felice Paolo Ferracin LQ Mechanical Behavior Overview and.
Fred Nobrega. 9 Dec 2014 Model Design & Fabrication – FNAL Fred Nobrega 2 Outline Background Mechanical Design Model Magnet Features Coil Technology and.
Development of the EuCARD Nb 3 Sn Dipole Magnet FRESCA2 P. Ferracin, M. Devaux, M. Durante, P. Fazilleau, P. Fessia, P. Manil, A. Milanese, J. E. Munoz.
SC magnet developments at CEA/Saclay Maria Durante Hélène Felice CEA Saclay DSM/DAPNIA/SACM/LEAS.
MQXF length P. Ferracin Video-meeting Structure WG September 18th, 2014.
LARP Prototype Assembly Steps Towards Q1 and Q3 Production Dan Cheng MQXF Design Review December 10-12, 2014 CERN H. Felice, R. Hafalia, D. Horler, T.
LARP Prototype Assembly Steps Towards Q1 and Q3 Production Dan Cheng MQXF Design Review December 10-12, 2014 CERN.
Fred Nobrega, Nikolai Andreev 21 September, 2015.
Helene Felice, Dan Cheng Support structure modifications Update on discussion at CERN.
S. Caspi, LBNL HQ Progress and Schedule Shlomo Caspi LBNL LARP Collaboration Meeting – CM13 Port Jefferson November 4-6, 2009.
Helene Felice 11/16/2011 LARP Collaboration Meeting 17 1 st HiLumi Collaboration Meeting CERN Status on HQ Coil Design and Fabrication.
H. Felice, D. Cheng, H, Pan, M. Reynolds, A. Pekedis, T. Lipton MQXFS1 update 07/27/2015.
MQXF support structure An extension of LARP experience Helene Felice MQXF Design Review December 10 th to 12 th, 2014 CERN.
Subscale quadrupole (SQ) series Paolo Ferracin LARP DoE Review FNAL June 12-14, 2006.
MQXFS Assembly Procedure J.C Perez N. Bourcey, B. Favrat, P. Ferracin, Ph. Grosclaude, M. Juchno, L. Lambert, P. Moyret, J. Parrilla Leal, N. Peray, T.
Hybrid Structure with Cooling John Cozzolino LARP Collaboration Meeting Port Jefferson, NY November 4-6, 2009.
Magnet design, final parameters Paolo Ferracin and Attilio Milanese EuCARD ESAC review for the FRESCA2 dipole CERN March, 2012.
Long Quadrupole – G. Ambrosio 1 LARP Collaboration Meeting – FNAL, Oct , 2008 BNL - FNAL - LBNL - SLAC Long Quadrupole Giorgio Ambrosio LARP Collaboration.
Helene Felice HQ Test Results Review Thursday December 16 th Overview of HQ coils and Magnet.
LMQXFA Cold Mass Assembly Antonios Vouris Fermilab February 3, 2016.
LARP Collaboration Meeting Racetrack Coil Fabrication
MQXF Workshop on Structure, Alignment, and Electrical QA Feb 2-4, 2016 Dan Cheng J. De Ponte, H. Felice, S. Kincaid, T. Lipton, S. Myers, A. Pekadis, M.
Dan Cheng, Helene Felice 4/8/2015 MQXFS-D Build-up LBNL “Lessons Learned”
1 MQXF Workshop on Structure, Alignment, and Electrical QA Feb 2-4, 2016 Options for modifications to parts and procedures Options for modifications to.
MQXF Workshop on Structure, Alignment, and Electrical QA Feb 2-4, 2016 Dan Cheng J. DiMarco, C. Hernikl, D. Humphries, S. Myers, X. Wang MQXFA Magnet Fiducialization.
Overview short model design and structures P. Ferracin on behalf of the MQXF collaboration MQXF Workshop on Structure, Alignment, and Electrical QA CERN.
MQXFB design, assembly plans & tooling at CERN J.C Perez On behalf of MQXF collaboration team MQXF Workshop on Structure, Alignment and Electrical QA.
DOE Review of LARP – February 17-18, 2014 SQXF Coil Design and Fabrication Winding and Curing Miao Yu February 17,
MQXFB Prototype structure and plans at CERN J.C Perez On behalf on MQXF collaboration team Joint HiLumi-LARP Meeting & 24 th LARP Collaboration Meeting.
DOE Review of LARP – February 17-18, 2014 Helene Felice P. Ferracin, M. Juchno, D. Cheng, M. Anerella, R. Hafalia, Thomas Sahner, Bruno Favrat 02/17/2014.
LARP DOE Review, 7/9/2012Long Quadrupole – G. Ambrosio 1 Long Quadrupole Program Giorgio Ambrosio DOE Review of the LARP Program SLAC July 9-10, 2011 LQ.
QXF Coil working group 2/11/16 Dan Cheng J. DiMarco, C. Hernikl, D. Humphries, S. Myers, X. Wang MQXFA Coil Fiducialization Proposal.
MQXFSD3 laminated structure J.C Perez N. Bourcey, B. Favrat, P. Ferracin, P. Moyret.
Dan Cheng 8/27/2015 MQXFS1 Observations and Feedback from the Assembly Processes, Part II.
LHQ Coil Design and Features Miao Yu Fermilab May 09, 2012.
2 nd LARP / HiLumi Collaboration Mtg, May 9, 2012LHQ Goals and Status – G. Ambrosio 11 LHQ Goals and Status Giorgio Ambrosio Fermilab 2 nd LARP / HiLumi.
3 rd ESAC Review, 27 th February to 1 st March 2013, CEA Saclay Fresca2 Dipole Structure Assembly J.C Perez on behalf of Fresca2 collaboration team.
QXF magnet integration Paolo Ferracin Joint LARP/CM20 HiLumi meeting Napa Valley, CA, USA 8-10 April, 2013.
Dan Cheng QXF Structures Working Group November 7, 2014 Implications for the split-shell MQXFS concept.
1 Long QXF Mirror Fabrication (MQXFPM1) Rodger Bossert HiLumi-LARP Collaboration Meeting May 18-20, 2016 SLAC R. Bossert - HiLumi Collaboration Meeting.
1 MQXFS Mirror Fabrication R. Bossert, G. Chlachidze, S. Stoynev HiLumi-LARP Collaboration Meeting May 11-13, 2015 FNAL.
LHQ Coil Fabrication Experience and Plan Miao Yu 04/08/2013.
Page 1 TQC Mechanical Model Igor Novitski / Rodger Bossert May 24, 2006 Results of MM#4 and MM#5.
QXF Coil Fabrication & Tooling Reaction / Impregnation
Coil Design and Fabrication
HQ Mirror Assembly R. Bossert
TQS Structure Design and Modeling
Mechanical tolerance analysis for the MQXF prototype structures
SQXF Mirror QXFSM1 Strain Gauges
FRESCA2 Update on the dipole design and new calculations
Helene Felice 02/15/2013 HiLumi Meeting HQ Program
Glyn Kirby Magnet AssemblyTechniques.
MQXF coil cross-section status
Design of Nb3Sn IR quadrupoles with apertures larger than 120 mm
Design of Nb3Sn IR quadrupoles with apertures larger than 120 mm
Helene Felice 02/15/2013 HiLumi Meeting HQ Program
Presentation transcript:

LARP Internal Review of the MQXFA Mechanical Structure Design and Functional Requirements 7/16/2015 Dan Cheng, Helene Felice Development of the Shell-Based MQXFA Structure, Part II

Overview LQ Design Features and Experience HQ Design Features and Experience Experience with MQXFS Short Models Summary Outline 07/16/20152D. Cheng

TQ LQ HQ Scale up of Nb 3 Sn magnet technology has been developed over the past 10+ years in the LARP collaboration The MQXFA magnet is a culmination of lessons learned from recent magnets, LQ, HQ, and MQXFS 07/16/2015D. Cheng3 Shell-Based Magnet Structure Development Fabrication Parts Design Assembly Feedback

LQ EXPERIENCE 07/16/2015D. Cheng4

Design Features: – Segmented shell for stress distribution – Used alignment pins to align shell to pad – Load pads were assembled from 2” thick profiles – Implemented Master Key bladder packages for assembly Full-length magnet used half-length masters, bladders, and keys – Used stainless steel axial rods instead of aluminum Processes Employed: – Shell procured as a single, 3.7 m long shell cylinder before cutting into four equal length segments – Pre-assembled shorter shell-yoke assemblies vertically – Short shell-yoke subassemblies were joined together into longer assemblies (horizontally) – Coilpack was slid into place on surrogate master key (same as what was done in LR) 07/16/2015D. Cheng5 Features of the LQ Structure

Forged and honed a full length (3.7 m) long shell – The LR shell was also initially fabricated in this manner – Ensured that each shell segment’s diameter would match at transitions Cylinder was then cut into four equal length pieces Each segment had pin slot features machined for shell to pad alignment (in-house, after receipt of segments) – To ensure alignment of shell-yoke subassemblies Limitations of this method – Severely limited potential fabrication vendors by long length requirement – Long lead times required for procurement – Required additional in-house fabrication work 07/16/2015D. Cheng6 Shell Procurement & Fabrication

Full-length assembly process was as follows – Each of the four ¼-length shell-yoke segments were preloaded separately (vertically) – Two segment pairs were then joined together, creating two half-length subassemblies These were “floated” together on air pallets Tie rods and gap keys were removed and replaced with half-length versions – Both of these half-length subassemblies were joined together Yoke tie rods were removed and replaced with full-length versions Large, flat granite surface plates were required for this assembly process 07/16/2015D. Cheng7 Assembling the 3.7 m Shell-Yoke Structure

Single, planar smooth surfaces for load keys and bladder slots were also beneficial – Load pad was an assembly of 50 mm thick laminations—not uniform, as in TQ pads – Small variations could affect the ability to insert long, narrow bladders and/or key & shim packages Trapezoidal shape of the master keys allowed insertion of coil pack without having to be fully aligned in orthogonal axes – Tapers on each side allows for slight mismatches at initial insertion – Side to side assembly tolerances were 0.005”-0.009” in contact with pads/yokes and masters – Profile machining is a little more complicated with angular features – First parts arrived bowed These parts were individually bent back into shape 07/16/2015D. Cheng8 LQ Master Keys

Incorporated two lifting fixtures for coil pack assembly operations for maneuvering lower and upper coil pairs – Relied upon “compound” lifting and rotating operations – Coils were lifted and paired with the use of lever arms maneuvered by hand – This method required several technicians to perform this operation (not resource efficient), and not technician dependent Tooling for coil rollover operations were literally “rolling” operations – Functional, but not efficient use of space and resources 07/16/2015D. Cheng9 Long Coil Handling

Coil packs were built with both Fuji paper and SG measurement data – Pressure response, combined with the SGs, showed mismatches in radial contact We needed better collar radial shim package control 07/16/2015D. Cheng10 Radial Contact Examination

LQ load pad profile was made via 2” thick laminations assembled together into 3.7 m long assemblies – TQ load pads were fabricated as monolithic “blocks” Two SST pieces on each end sandwiched the iron piece in the middle For lifting, tapped holes were drilled in these pad assemblies for directly attaching lifting swivels for handling – This is not necessarily a scalable method of handling these long pieces; relies upon the rigidity of the assembly to self-support 07/16/2015D. Cheng11 Load Pad Assemblies, Laminated

Coilpack was built up on the insertion table “raft” with surrogate master key for assembly After coilpack assembly was completed, insertion table/raft was moved into position and aligned prior to coilpack insertion Coilpack was inserted into magnet structure by sliding on surrogate master key – Friction effects and metal on metal sliding are not ideal 07/16/2015D. Cheng12 Coilpack Assembly and Insertion

Wires were not properly routed and protected during axial prep and preload on LQS03 – Pinched VT instrumentation wires when the magnet was axially preloaded – Problem was noticed only on hipot checkout of magnet, after preload operations were complete – Diagnosis and repair required several days’ examination and effort, including unloading/re-loading and in-situ wire repairs 07/16/2015D. Cheng13 LQ Instrumentation Wiring LQ, Return End LQ, Lead End Pinched VT wires in LQS03 From:

What we learned: – Master keys were integral to the design and assembly of the long structure assembly process – Joining shell-yoke subassemblies is a viable option to making long magnets – Half-length master key assemblies were effective in the preload operations – Collar shimming and coil radial contact was important for the preload of the magnet coils What we wanted to do better: – Long shell machining started with one piece, but limited fabrication options, and added time and cost – Long coil handling tooling was functional, but not efficient – Load pad handling was also functional, but not necessarily scalable – Coil pack insertion method is not necessarily scalable – Instrumentation wire routing at magnet level needed improvement 07/16/2015D. Cheng14 Summary of LQ Experience

HQ EXPERIENCE 07/16/2015D. Cheng15

HQ Design Features: – Bolted aluminum collars introduced between pads and coils – Keys incorporated between coils and collars for alignment and improved field quality – Incorporated master key package concept from LQ Process Features: – Vertical shell-yoke assembly 07/16/2015D. Cheng16 Design Features of the HQ Structure

First keys were machined out of Aluminum – Length was same as pole groove – Shimmed to fill the gap width using G10 layers Initial collar design had sharp corner at gap – Coil arcing (due to coil oversize/over-compression issues) to this sharp edge was observed during early tests Changes implemented over HQ program: – Aluminum keys were changed to full-length G10 keys, shimmed to size only at pole groove – Sharp collar edges were chamfered – Introduced ground plane insulation Kapton “wings” to electrically isolate collar edges 07/16/2015D. Cheng17 Collars & Coil Alignment Key Performance

HQ01-style coils – Oversized coil conditions – CMM standards were not the same across the Labs HQ02- and HQ03-style coils – Coil sizes were closer to nominal dimensions – Coil CMM model was standardized Refined methods to calculate coil and collar radial shims and packaging – Better CMM data helped – Better control and adjustment of the radial shims between collars and coils 07/16/2015D. Cheng18 Coil Shim Size Calculations Measurement data is used to determine best fits, required shims, etc. for assembly Components are shimmed, assembly operations are performed

LQ & HQ had some history of “losing” strain gauges upon cooldown – Adhesive cure temperature could not be adequately controlled to manufacturer’s specifications However, the shell gauges still performed reliably… – Tape securing thermal compensators was also causing these gauges to be lost New procedures developed for HQ03a gauges: – SGs using new bonding procedures performed well; epoxy was able to cure at RT and remained intact when cold – Temperature compensators were held in place w/o affecting the bonded SG – Switch from SK- to WK- gauges (with solder pads) attached required less time to wire up 07/16/2015D. Cheng19 Strain Gauge Performance SGs using new procedures SGs using old-style procedures

What we learned – We refined the process of calculating what shims were required to ensure the contact between collars, coils, and alignment keys – Collars required ground plane protection in addition to better control of radial shimming – Full length (G10) alignment keys only needed to be shimmed to size at pole island region – Strain gauge installation process was refined, which improved reliability 07/16/2015D. Cheng20 Summary of HQ Experience

MQXFS EXPERIENCE 07/16/2015D. Cheng21

Design Features of the MQXFS Structure – Shells features pins to align yokes in each quadrants – Coils incorporate a bonded GPI layer – Introduced alignment key feature between pads and collars – Retained master key packages – One axial endplate is threaded for the axial loading rods Process & Tooling Features of the MQXFS – Single vertical shell-yoke subassembly with stacked shell segments – Dedicated rollover tooling for coil manipulation and GPI – Coilpack insertion rails for coilpack/structure integration 07/16/2015D. Cheng22 Features of the MQXFS Structure Note: Structure was designed in collaboration with CERN, and the components were procured by CERN and shipped to LBNL.

Incorporated vertical assembly process – We were able to preload several shell segments at once – Requires slightly more than 2x the shell length of crane hook height – Working pedestals required for access to top of assembly Single pins were installed in one shell/yoke quadrant only – Pin slots weren’t machined in correct location after shell was cut Seismic support of stacked shells can be improved Bladder support cross tooling (also supplied by CERN) required high pressures to preload yoke gaps 07/16/2015D. Cheng23 Shell Assemblies Seismic Straps In MQXFS Note: Nominal-length shell segment was cut in half

MQXFSD1 Yoke-Shell Subassembly 07/16/2015D. Cheng24 High bladder pressures (6, ,000 psi/40 – 48 MPa) were required for the initial assembly – Bladders reached this pressure without incident – However, we encountered some bladder failures upon cycling after reaching these pressures CERN’s redesigned bladder support 2 bladder slots per quadrant (N. Bourcey)

MQXFSD revealed rollover tooling needed modifications (cutouts) to accommodate GPI application of coils – Changes were incorporated for MQXFS1 coils processing – Increased efficiency of manipulating and processing coils, with only two technicians required GPI was also purchased in custom width rolls, requiring trim to length only Tooling for pre-creasing the corners will need optimization GPI was originally bonded to OD and Midplane – Bumps and wrinkles were observed, due to localized heat applied during layer application – Midplane portion was peeled back to allow it to “float” 07/16/2015D. Cheng25 Coil GPI Layer Application

Even though this outer-pick coil lifting tool was patterned from LQ experience, we found some issues keeping the paired coils together – We tested this tooling with the MQXFS dummy coils – We added alignment key segments to hold coil in place, and were able to move the MQXFS1 coils without issue 07/16/2015D. Cheng26 Lifting tooling mod

MQXFSD1 Strain Gauges 07/16/2015D. Cheng27 CERN gauges (half-bridge/AC) were also installed on dummy coils and shells – In addition to LARP-style gauges, these will provide an apples-to-apples comparison of measurements Benefits, if standardized – Half-bridge configuration and installation will further improve throughput SG installation – Simplified SG hardware, software, and cabling See Helene’s Part I talk

Collars were designed with too large of a radius (~1.25 mm nominal gap) – Required many (5-6) layers of 0.010” thick G10 to shim correctly – First MQXFS coils appear to be <0.003”R oversized, worst case Proposed collar design will now be 0.5 mm/0.020” over nominal coil radius – Impregnated OR is mm – Therefore, new proposed collar radius is mm, down from 115 mm Radial shims – G10/G11 Sheets do not come in long enough lengths for single length application – Thinnest G10/G11 sheet available is mm/0.005”; thickness still not consistent – Plan to replace with custom width rolls of Kapton of various thicknesses ≤0.005” 07/16/2015D. Cheng28 Collars and Shimming

Fuji paper application – TQ/LQ => MQXF represents ~40% increase in pressure area, using virtually the same bolting torque (100 in-lb) – Purchasing “Super Low” LLW Fuji paper (71~355 PSI / 0.5~2.5MPa) to obtain better readings Coil pack assembly spud support – Requires positive engagement with the coils during assembly 07/16/2015D. Cheng29 Collar Pack Experiences

MQXFSD1 Master Key Shims 07/16/2015D. Cheng30 Alignment and load keys required many shims in nominal dimensions – Each load key required 0.120” (~3 mm) thick worth of shims to start – Alignment keys required ~2 mm per side of shims (still with clearance) => We plan to increase nominal thickness of these keys to reduce shims needed Load keys – Mild steel on mild steel master key surfaces seemed to gall (even with Moly dry lubricant) – These keys will be changes to bronze material

What we learned – We were able to preload several shells in one shell-yoke preload assembly step – Yoke/Shell alignment may only require 1 quadrant pin for clocking, which will not over-constrain system – Custom width GPI coil layers facilitated ease of preparation – Strain gauge measurement systems are showing comparable results What we’d like to do better – Seismic support of the shell-yoke subassembly needs improvement – Bladder support cross requires double slots for lower pressure operation => more reliability – GPI creasing technique/tooling to be improved – Collar radius will be adjusted to reduce amount of radial shims required Collar shims will be custom width sheets of Kapton rolls, also for ease of prep – Pressure paper sensitivity was not high enough – Coil pack installation support needs to positively engage coils during assembly – Alignment and load keys will be adjusted in thickness to reduce the amount of shim stock required, and also be made out of bronze instead of steel 07/16/2015D. Cheng31 Summary of MQXFS/D Experience

Key takeaways from our lessons learned – LQ We learned manufacturing and assembly techniques for long magnets and their related structures Tooling related to the length scale up from short models was functional, but more efficient designs are required – HQ We refined the process of calculating radial shim packaging due to large variances of coil sizes that were observed The addition of the alignment key feature also required better manipulation of coil CMM data and overall shim package control – MQXFS We are learning ways to improve our efficiency and throughput of building these magnet structures Our tooling for manipulating the coils during preparation and assembly improved our efficiency with processing with fewer resources than before Some greater efficiencies can still be gained by adjusting some part dimensions and incorporating the use of different materials 07/16/2015D. Cheng32 Summary of the Lessons Learned in Shell-Based Structure Development

APPENDIX 07/16/2015D. Cheng33