Slides - Agenda

Stato del progetto
MAGIX
Giovanni Volpini
a nome della collaborazione
Frascati, riunione CSN V, 27 luglio 2015
MAGIX & INFN participation
to HL-LHC
CERN-INFN Collaboration Agreement
MAGIX
Design, construction and test of
WP1
CORRAL
the
five
prototypes
of
the
corrector magnets for the HL
CERN endorses MAGIX WP1 & WP2
deliverables and milestones, contributing with
527 k€, through the collaboration agreement
KE2291/TE/HL-LHC
interaction regions of HiLUMI
WP2
PADS
2D & 3D engineering design of
the D2 magnets
Development of HTS coil for
WP3
SCOW-2G
application to detectors and
1
accelerators
Low-loss SC development for
WP4
SAFFO
application to AC magnets
MAGIX is a INFN-funded research project,
(GrV, «Call») whose goal is to develop
superconducting technologies for
application to future accelerator magnets.
It includes four WP’s, two of which are
relevant to HL-LHC
2014-2017, 1 M€ + personnel funds
2
INFN already involved in FP7HiLumi (UE-HILUMI, GrV)
WP2 beam dynamics, LNF
WP3 magnets, MI-LASA
WP6 cold powering, MI-LASA
Economic Plan
93 m.u. (WP1) + 18 m.u. (WP2) (INFN personnel only, no University employees)
Costs
CERN contribution
Personnel
521,900 €
50%
260,950 €
Materials for magnet
construction
113,000 €
75%
84,750 €
Costruction & test
363,000 €
50%
181,500 €
Travels, lab operation,
tooling, R&D activities
305,000 €
0%
TOTAL
0€
527,200 €
781,000 €
WP1 (679k€) + WP2 (102 k€)
Payment linked to deliverables
Giovanni Volpini
CERN, 30 June 2015
Schedule
HiLumi-MAGIX schedule
2014
v. February 2014
2
3
4
5
6
7
2015
8
9 10 11 12 1
2
3
4
5
6
7
2016
8
Project Management
WP0
M 0.1
Feb 2014
Kick-off meeting with specification transfer
M 0.2
Dec 2014
1st year activity monitoring
M 0.3
Dec 2015
2nd year activity monitoring
M 0.4
Dec 2016
3rd year activity monitoring
M 0.5
Jun 2017
4th year activity monitoring
M 1.1
M 1.2
D 1.3
D 1.4
Jul 2014
Dec 2014
Mar 2014
Mar 2015
Oct 2016
Dec 2015
Mar 2016
Jul 2016
Oct 2016
Apr 2015
July 2016
Feb 2017
Mar 2017
June 2017
M 2.1
D 2.1
June 2015
M 2.2
D 2.2
WP1
CORRAL
D 1.1a
D 1.1b
D 1.2
M 1.3
M 1.4a
M 1.4b
M 1.5
M 1.6a
M 1.6b
M 1.6c
WP2
*
*
**
***
***
****
****
****
Sextupol engineering design.
Sextupol construction.
Preliminary 2D design of the five magnet types
Preliminary 3D design of the five magnet types
Executive design of the five magnet types
MgB2 quadrupole design.
Octupole and decapole construction
Quadrupole and dodecapole construction
MgB2 quadrupole construction
Test of the sextupole
Test of the octupole and decapole
Test of the dodecapole and quadrupole
Corrector magnet test report
Corrector magnets final check, packing and transport to CERN
PADS
2D magnetic design to minimize the cross talk between the two dipoles.
Dec 2015
2D mechanical design.
M 2.3
Feb 2016
3D magnetic design including the coil ends.
M 2.4
Apr 2016
Quench preliminary analysis.
M 2.5
Jun 2016
3D mechanical design with the axial pre-load study.
Dec 2016
Final Engineering design.
M 2.6
Notes
*
**
***
****
1
D 2.3
These two deliverables are grouped in one in the MAGIX project
Note the change of scope wrt to the MAGIX project
These two milestones are grouped in one in the MAGIX project
These two milestones are grouped in one in the MAGIX project
Explanation
Activity
Milestone
Deliverable
Giovanni Volpini, CERN 30 June 2015
9 10 11 12 1
2
3
4
5
6
7
2017
8
9 10 11 12 1
2
3
4
5
6
7
8
9 1
Milestones
&Deliverables
50%
June 2015 The INFN-CERN collaboration steering committee (F Bodry, L Rossi, E Nappi A
Zoccoli) has endorsed the work done so far, authorized the transfer to INFN of 260 kE, in
return of the deliverables completed so far.
Giovanni Volpini, CERN
30 June 2015
WP1
WP1
CORRAL
CORRAL
CORrettori per le Regioni di interazione ad Alta luminosità di Lhc.
Obiettivo: progettazione, costruzione e collaudo dei prototipi di cinque magneti
correttori multipolari per le regioni di interazione ad alta luminosità di LHC, sulla
base delle richieste del progetto HL-LHC dal CERN.
Verrà verificata sia in fase di progetto che attraverso la costruzione di un
prototipo, la fattibilità di soluzioni che utilizzino il MgB2 anziché il più
convenzionale Nb-Ti.
Giovanni Volpini, CSN V, Catania 23 luglio 2014
6
Corrector magnet
inventory
Iron yoke
Iron yoke
SC Coils
SC Coils
Mechanical
support
Mechanical
support
SC Coils
150
Mechanical
support
OD460
From 6-pole to 12-pole
magnets exist in both normal
and skew form (the latter is
shown)
150
OD320
The superferric design was chosen for ease of construction, compact shape,
modularity, following the good performance of earlier corrector prototype magnets
developed by CIEMAT (Spain).
Giovanni Volpini, CERN
30 June 2015
Sextupole
layout
Yoke
Coil
D320
Wedge
CuBe
TieRods
194
5.8 mm thick iron
laminations, machined
by EDM
Yoke
Bridge
Flux-return plates
Coil winding &
impregnation tooling
Insulation scheme:
-wire w/ S2 glass 0.14 mm thick (on diameter)
-ground insulation:
G11, 2 mm thick plates on both sides of the coil, including the wire exits
G11 thin, flexible layer on the inner wall of the coil;
S2 tape on the outer wall
Resin inlet/outlet
Base plate
mandrel
Resin inlet/outlet
Top plate
Giovanni Volpini
CERN,
26 February 2015
Closing cap
(defines the
impregnation
chamber)
Winding station
Controlled wire tension 10 N
(51 MPa)
Telecentric camera system
Single Coil Sample Holder
Goals:
1) To test a coil in “realistic” conditions
to identify major faults in the
design/assembly;
2) To commission the “small” magnet
test station, to be used to test sextupole,
octupole and decapole
y
z
x
Test results
First test at 4.2 K
Current increased by steps at 0.3 A/s. Quench induced with heaters at 90, 160, 200 and 220 A. Ramp up to 260
A (no quench induced at this current value by choice). No spontaneous quench occurred.
Test at subcooled LHe
Significant heat load in the bath prevents from reaching a temperature lower than 2.5 K. Main reason is the
thermal shield, whose temperature decreases very slowly. Current ramp up to quench.
Four training quenches occurred at
295 A (2.56±0.04 K) or 80% of the s.s. at this T
318 A (2.60±0.04 K) or 87%
“
329 A (2.72±0.05 K) or 91%
“
325 A (2.85±0.06 K) or 91%
“
The magnet
operates at 40%
on the load line
Training at 4.2 K
Current ramp up to quench at 0.3 A/s
First quench at 280 A, then repeated increasing the ramp rate up to 5.7 A/s (limited by power supply in this
configuration). In total 14 quenches at 280 A , or 95% of the s.s. limit.
E.M. Forces
A magnetic plate creates along the normal of the coil plane an e.m. force pattern more resembling to that
experienced by a coil during its operation inside the magnet.
Fx (normal to the coil plane, half coil)
Fy (normal to long axis, half coil)
Fz (normal to long axis, half coil)
2.9 kN @ Iop,
1.5 kN @ Iop,
0.6 kN @ Iop,
here
reached at about
“
“
300 A
250 A
180 A
Green light to the magnet construction !
Sextupole assembly
Iron laminations
Cu traces for coil-to-coil junctions
Duratron plate
s.s. rings
Giovanni Volpini
CERN, 30 June 2015
Coil
MgB2 development
We are focusing the desing around an
innovative solution. We call it Round Coil
Superferric Magnet (RCSM)
Simple, circular coil shape, cost effective.
Expecially suited to strain-sensitive
materials, like MgB2
We consider a sextupole configuration;
different multipoles may be realized
replacing the iron
Preliminary design (milestone M1.3)
almost completed.
14
Old & Older
15
A rule (symmetry) changer
No matter how a sextupole magnet is done,
it is invariant by a 120 degree rotation.
A 60 degree exchanges the “north” and
“south” poles; if we reverse the current
direction as well, the field is globally
unchanged!
A RCSM is invariant by 120 degree rotation.
A rotation by 60°, amounts to a “mirroring”
w.r.t. a plane normal at z-axis, at z=0. No
change in current.
No overall mirror symmetry.
This difference of the symmetries has profound consequences on the harmonics properties: a
“traditional” layout has no even (“forbidden”) harmonics, and no net solenoidal field; a RCSM
has also even harmonics, that vanish when integrated from -∞ to +∞, and a net solenoidal
field. More complex configurations may suppress the latter, at the price of net even harmonics.
G. Volpini, J. Rysti
CERN 15 July 2015
Finanziamento 2016
Giovanni Volpini
CSN V Frascati, 27 luglio 2015
17
Persone coinvolte
+ 1 tecnologo TD a partire da settembre 2015
Giovanni Volpini
CSN V Frascati, 27 luglio 2015
18
WP1 Revised schedule
to completion
2015
New dates
M
M
M
M
M
M
M
1.3
1.4a
1.4b
1.5
1.6a
1.6b
1.6c
D 1.2a Oct-15 Nov-15
D 1.2b
Apr-16
Dec-15
Mar-16 May-16
Jul-16 Dec-16
Oct-16
Apr-15 Oct-15
Jul-16 Sep-16
Feb-17 May-17
D 1.3 Mar-17 Jun-17
D 1.4 Jun-17
1
Executive design of b3 b4 b5
Executive design of a2 and b6
MgB2 magnet design
Octupole and Decapole Construction
Dodecapole and Quadrupole Construction
MgB2 magnet construction
Test of the sextupole
Test of the octupole and decapole
Test of the quadrupole and dodecapole
Corrector Magnet Test Report
Corrector magnets Final Check & transport
2
3
4
5
6
7
2016
8
9 10 11 12
1
2
3
4
5
6
7
2017
8
9 10 11 12
1
2
3
4
5
6
7
8
9 10 11 12
3 month potential
delay on the last
deliverable
Milestone
Deliverable
CERN-INFN Collaboration Agreement End
Modified
From the minutes of the CERN-INFN steering committee:
“There is a potential 6 month delay (that so far has not impacted on Deliverables)
due to longer design and R&D. It will appear as delay of next deliverable. However,
G. Volpini is confident that part of the delay is recoverable, and the potential delay
on the last deliverable could be of the order of three months. The CERN coordinator
expresses an extremely positive judgement on the work done so far and supports
the effort of INFN to recover the schedule without impacting of the quality of the
work. However, the correctors are not on the critical path of HL LHC IR magnets
schedule.”
19
Giovanni Volpini
CSN V Frascati, 27 luglio 2015
WP2 PADS
Progettazione dipoli di separazione (D2) per
l’upgrade di luminosità di LHC.
Trasparenze a cura di P. Fabbricatore
Giovanni Volpini, CSN V, Catania 23 luglio 2014
20
The central critical point of D2 magnets
Close to IR (ATLAS and CMS) two D2 magnets (one per side) have the same field orientation
Multipole (units)
70
60
b2 105
50
b2 95
40
30
20
10
0
0
1
2
3
4
5
3
4
5
Field (T)
20
0
-20
Multipole (units)
Preliminary studies (BNL):
High harmonics due to
cross talk!!!!
-40
-60
-80
-100
b3 105
-120
-140
b3 95
-160
0
1
2
Field (T)
A solution was found!
Since the magnet suffers from magnetic cross talk between the
two coils, we need to actively compensate this cross talk in a wide
magnetic field range in the two apertures. This was done involving
a strategy based on three pillars:
a) No iron is placed in between the coils (so limiting saturation
effects);
b) Each coil is asymmetric in a way to cancel the magnetic cross
talk each other.
c) The yoke is suitably profiled for keeping constant the harmonic
components .
Based on these concepts a 2D magnetic optimization was carried
out leading to an acceptable field quality with a limited variation of
the multipoles as the magnet field is raised from the injection
value to the maximum value (4.5T).
Configuration INFN_3_3_6
2D Lay-out
Rh
Dy
Dx
R
Main Characteristics
Characteristics
Aperture
Units
mm
Number of apertures
Value
105
2
Distance between apertures (cold/warm)
mm
188.00/ 188.45
Cold mass outer diameter (min/max)
mm
570/630
Magnetic length
m
7.78
Bore field
T
4.5
Peak field
T
5.20
Current
kA
12.050
Temperature
K
1.9
Loadline margin
(%)
35
Overall current density
A/mm2
443
Differential inductance per meter
mH/m
3.509
Stored energy
MJ
2.18
Differential inductance
mH
27.3
Superconductor
Nb-Ti
Field quality
b2 and b3 optimised at a B field slightly lower than 4.5 T (4.3 T)
3
2
1
b2
0
b3
b4
-1
b5
b6
-2
b7
b8
b9
b10
-3
0
1
2
3
4
5
B (T)
Coil end design is 90% done
End with connections
b3 (black) and a3 (red) components in the connection end
Integrated harmonics still to be optimised (ex. b3 int. -5 unit)
Mechanics: Plan B (single collar)
Before the two-collar option, a solution
with a single collar was completely studied.
Not satisfactory because the complexity of
the assembly and the differential thermal
contraction. However this solution is still
there as plan B in case of major problems
with double collar option.
Mechanical effects on b2 and be field harmonics
(computed for a previous version)
Presently we have engineering drawings of the 2D cross section
Conclusions
• The activities related to D2 development are progressing
faster than foreseen
• The 2 D magnetic and mechanical designs were completed
• The 3D magnetic design is close to be completed
• Still to do: 3 D mechanical design, 3D engineering and
quench analyses
Finanziamento 2016
Giovanni Volpini
CSN V Frascati, 27 luglio 2015
32
Persone coinvolte
Giovanni Volpini
CSN V Frascati, 27 luglio 2015
33
WP3
SCOW-2G
SuperCOnducting Windings - 2G tapes
Obiettivo: Progettazione, realizzazione e
caratterizzazione di avvolgimenti dimostrativi a doppio
pancake realizzati in YBCO coated conductors, per
verificare l’applicabilità di questa tecnologia alla
realizzazione di magneti in alto campo, quali ad esempio
quelli richiesti nei canali di raffreddamento muonici.
Responsabile: Umberto Gambardella (SA)
Giovanni Volpini, CSN V, Catania 23 luglio 2014
34
Attivita’ in corso
1. La completa caratterizzazione, in termini di corrente di trasporto in
funzione del campo applicato, di nastri di diversa fabbricazione
(SuperPower, SuNAM e SupeOx).
2. Analisi delle problematiche di impregnazione e di terminazioni dei
primi avvolgimenti.
3. Il potenziamento delle stazioni di test a temperatura variabile con
campo magnetico di background e cryogen free.
Giovanni Volpini, CERN
14 January 2014
35
Apparato sperimentale
Portacampioni
Magnete cryogen-free
64.96
14
T=65 K B=0 T
64.94
Temperature (K)
12
Voltage (mV)
Alle temperature di 77K, 70K e 65K, per ogni
valore di campo magnetico è stata registrata la
caratteristica I-V del nastro, ricavando da questa
la corrente critica.
Le misure sono state effettuate sia in flusso di
gas che in bagno di azoto liquido pompato. La
temperatura è misurata da un termometro
Cernox.
A destra è raffigurato un esempio di
caratteristica I-V misurata, insieme
all’andamento in temperatura La deriva termica
è contenuta a 0.12 K.
10
8
6
4
2
0
200
64.92
64.9
64.88
64.86
64.84
64.82
250
300
350
Bias current(A)
400
450
64.8
200
250
300
350
Bias current(A)
400
36
450
Caratterizzazione dei nastri
YBCO 2G
10
1
0
1
2
3
4
m H (T)
0
5
6
7
Ic @10mV/cm (A)
100
SuNAM
H + tape surface
SuperPower M4-217
H + tape surface
Ic @10mV/cm (A)
65K
70K
75K
77K
c
I @ 10mV/cm (A)
SuperOX H + tape surface
100
10
1
65K
70K
75K
77K
0
1
2
3 4
B (T)
5
6
7
8
65K
70K
75K
77K
100
10
1
0
1
2
3
4
m H (T)
5
6
7
0
I confronti delle prestazioni tra questi nastri individuano come
miglior nastro a zero campo quelli SuNAM, mentre per le prestazioni
in campo (ad es. a 2T) i nastri SuperPower risultano decisamente
migliori. Comunque il rapporto qualità/prezzo (o più precisamente
$/kA m) risulta a favore di SuNAM, ed è con questo tipo nastro che
intendiamo proseguire le attività.
Giovanni Volpini, CERN
14 January 2014
37
Test avvolgimento
e impregnazione
Per la realizzazione di avvolgimenti e’ necessario che le spire
siano ben ferme, cosa che normalmente si ottiene
impregnando i magneti con resine epossidiche.
Nel caso dei nastri HTS 2G, gli stress termici tangenziali
introdotti durante il raffredamento dalle contrazioni
dell’impregnante sulla struttura esterna del nastro (cap layer
in rame), facilmente degradano le prestazioni in corrente del
nastro.
Anche noi abbiamo riscontrato questo problema su diversi avvolgimenti di prova.
Abbiamo quindi realizzato test coil da 5+5 spire impregnate con differenti soluzioni per la
resina, ciclati più volte in bagno di LN2.
Solo quando non si e’ manifestata degradazione alcuna in almeno 4 cicli termici siamo
passati ad incrementare il numero di spire. L’ultima bobina di test è composta da 25+25
spire avvolte su un diametro di 40 mm e, con il tipo di resina messo a punto nelle
precedenti prove, abbiamo riscontrato una sostanziale tenuta della corrente critica in
autocampo a 77 K Ic=90 A, rispetto a 4 cicli termici.
Giovanni Volpini, CERN
14 January 2014
38
Attrezzatura per il collaudo
dei doppi pancake
Per il collaudo dei doppi pancake è stato attivato il VTI a flusso (Variable Temperature Cryostat) in
He gas inserito nel magnete cryofree da 12 T. Questo rappresenta un sistema economico e
flessibile per testare piccoli avvolgimenti realizzati su diametro utile di 40-50 mm. Con
avvolgimenti da 50+50 spire (massimo stimato come dimensione avvolta su 40 mm compatibile
con il nostro VTI) si potrebbero raggiungere campi >2T a 40 K (con bias da 200 A).
Una delle attrezzature in corso di realizzazione nel 2015
riguarda la supporteria per il test dei pancake nel VTI a flusso
di He comprensiva di un discendente che fornisce
contemporaneamente il sostegno meccanico ed i discendenti
di corrente. Attualmente i componenti sono in fase di
lavorazione
Dopo la pausa estiva partira’ la realizzazione di un
criostato in bagno di LN2 per il magnete cryofree,
per alloggiare pancake con maggior numero di spire
(a T= 65 K). Per questa attività esistono già disegni
preliminari del contenitore in acciaio con diametro
utile 80 mm. Restano da definire la supporteria ed i
passanti di corrente per sostenere lo stack di
pancake.
Giovanni Volpini, CERN
14 January 2014
39
Avvolgimento
cryogen-free
E’ prevista la realizzazione di un avvolgimento compatibile con la modalità cryogen free. Per
svolgere utilmente queste prove dovevamo incrementare le prestazioni dell’attrezzatura
mostrata sopra (sx), innalzando la corrente dei passanti fino ad almeno 300 A.
Ad oggi sono stati realizzati una coppia di adduttori di corrente HTS da 400 A con nastri
ReBCO da 12 mm mostrati a dx. Nel 2016, quando sarà meglio definito il layout finale dello
stack di pancake, ed i relativi former per la modalità cryogen free, si passerà alla
progettazione dei componenti interni della camera
Giovanni Volpini, CERN
14 January 2014
40
Nonostante un ritardo nell’avvio delle attività si ritiene che rispetto alla programmazione
2015 abbiamo raggiunto i seguenti risultati:
1.
realizzata una stazione di test nel magnete cryogen free 12T con un inserto VTI a
flusso forzato di He per misure su nastri HTS;
2.
realizzato e testato uno dei componenti cruciali per la stazione di test cryogen
free, cioè gli adduttori HTS da 400 A;
3.
progettato il controcriostato in bagno LN2 per il magnete cryofree da 12T, adatto
a contenere avvolgimenti di dimensioni fino a 80 mm;
4.
progettato supporto per test pancake fino a 50 mm da inserire nel VTI in campo
magnetico;
5.
acquisito conoscenze sulle proprietà dei nastri di ultima generazione di
fabbricanti diversi;
6.
acquisito conoscenze per impregnare avvolgimenti senza degradazione delle
prestazioni a seguito di cicli termici;
7.
realizzate bobine “significative” (25+25 spire, con campo sul conduttore a ≈0.3 T
in LN2) con primi approcci a terminazioni rigide.
Giovanni Volpini, CERN
14 January 2014
41
Attività 2016
Nel 2016 il gruppo si concentrerà nella realizzazione dello stack a 10 pancake, da inserire
nel VTI per raggiungere un campo da almeno 2T, e compatibilmente con le forze
meccaniche, in presenza di un campo di background.
Inoltre si avvierà la progettazione dello stack a 10 pancake in modalità cryogen free
(raffreddato solo per conduzione) e la meccanica per l’inserimento nella camera cryogen
free.
Giovanni Volpini, CERN
14 January 2014
42
Finanziamento 2016
Giovanni Volpini
CSN V Frascati, 27 luglio 2015
43
Persone coinvolte
Giovanni Volpini
CSN V Frascati, 27 luglio 2015
44
The End
Infinite Stack
Now, let’s consider the case with
an infinite number of RCSMs
stacked on top of each other with
alternating orientation (RCSM∞).
Closest to a 2D case for these
magnets.
For the radial field, same
symmetries apply as for RCSM1.
Longitudinal field Bz cancelled by
successive magnets.
G. Volpini, J. Rysti
CERN 15 July 2015
46
Two coils
• Two magnets with mirror
orientation, and reversed current
(RCSM2).
z
It possesses reflection symmetry w.r.t. a plane normal to z-axis,
but –surprisingly- it has no other symmetry (apart from 120° for
sextupole).
Therefore
+∞
𝐵 𝑑𝑧
−∞ 𝑧
=0
so not net z-component
But it turns out that its harmonic content is very high, lacking
those symmetries which “cancel” specific harmonics.
G. Volpini, J. Rysti
CERN 15 July 2015
2D Mechanics
LHC vs. HL-LHC corrector magnet
comparison chart
[A]
[T.m]
[H]
14 150
24.57
182
1.00
0.807 1.247
4.7 150
1.28
132
0.06
0.111 0.118
50
7.8 150
1.28
132
0.06
0.111 0.118
2
S
3
N
MCSX
3
S
MCSSX
70
4
N
MCOX MCSOX
70
16 100
4.4 150
1.41
120
0.04
0.087 0.152
4
S
70
22 100
3.2 150
1.41
120
0.04
0.087 0.152
5
N
150
1.39
139
0.03
0.095 0.107
5
S
150
1.39
139
0.03
0.095 0.107
6
N
29.2 150
4.35
167
0.086
0.430 0.229
187/II
[kJ]
6
S
150
0.92
163
0.017
0.089 0.052
MQSX
MCSTX
MCOSX
MCTX
MCSTX
70 2,116 550
[m]
Differential
Inductance
@ Iop
Magnetic
Length
Integrated
field at
r=50 mm
Operating
Current
Stored
energy
Aperture
[A] [mH] [mm]
Rev 9 July 2014
[J]
Inductance
Order
Type
mm
Operating
Current
HL-LHC
Aperture
Stored
energy
LHC
70
70
39 100
6
94
80
Giovanni Volpini, CERN
30 June 2015
Test Station
460
500 A current leads
SC bus-bars
Al-clad NbTi SC
cable from Mu2e TS
300
λ-plate for subcooled
LHe operation
Single coil test stand
(later 6-pole …)
coil under test