5

10

15

20

25

30

35

40

45

50

55

60

0 20 40 60 80

PH

Del

ay (

ms)

Power (W/cm2)

HQ02a2, OL, 14.6kA

HQ02a2, OL, 12 kA

HQ02a1, OL, 12 kA

HQ02a1, OL, 14.6 kA

Simulation 12 kA (OL)

Simulation 12 kA, tau = 90 ms

Simulation 14.6 kA (OL)

HQ02a2, IL, 14.6 kA

0

10

20

30

40

50

60

70

0 10 20 30 40 50 60 70 80

Hea

ter

dela

y (m

s)

Longitudinal coverage (mm)

50 W/cm2

150 W/cm2

B = 5 T

Lines with 25 W/cm2 steps

Design optimization of the protection heaters for the LARP high-field Nb3Sn quadrupoles

M. Marchevsky, D. W. Cheng, H. Felice, G. Sabbi, Lawrence Berkeley National Laboratory, T. Salmi, Tampere Institute of Technology, G. Chlachidze and G. Ambrosio, Fermi National Accelerator Laboratory, V. Marinozzi, University of Milan, E. Todesco, European Organization for Nuclear Research (CERN)

7 – 9 T5 - 7 T

short US CERNMagnet length (m) 1 4 7Heater width (mm) 20 20 20

Heater thickness (mm) 0.025 0.025 0.025Station length (mm) 25 25 25

Station distance per turn (mm) 105 105 105Station resistance (Ω) 0.025 0.025 0.025

SS resistivity (Ωm) 5.0E-07 5.0E-07 5.0E-07Cu resistivity (Ωm) 5.0E-09 5.0E-09 5.0E-09Cu resistance (Ω) 0.0056 0.0056 0.0056

Cu width (mm) 5 5 5Cu thickness (mm) 0.010 0.010 0.010

No. of stations per turn (dimless) 7 30 53Total resistance (Ω) 0.43 1.84 3.24

Voltage (V) 64 275 487Current (A) 150 150 150

Power (W/cm2) 112.5 112.5 112.5

Peak power density 50-150 W/cm2

HFU voltage up to 450 VHFU current: up to 220 A HFU capacitance: 4.8-19.2 mFDistance between heating stations: up to 120 mm(Could be related to the transposition pitch of 109 mm)Trace parameters:Kapton Insulation thickness: 50 µmStainless Steel thickness: 25 µmCopper thickness: 10 µmGlue thickness: up to 25 µm Coil surface coverage by trace: < 50 % IL Distance from heater to coil or voltage taps: 4 mm or more

• Large spacing L between the heating stations -> higher surface power density -> shorter t1, but longer t2 of the quench propagation between the heated areas

• Small spacing L between the heating stations > smaller heater power -> longer t1, but shorter t2 of the quench propagation between the heated areas

Power density > 50 W/cm2 is desirable

Reducing insulation thickness is critical for achieving a shorter heater delay

Superconducting accelerator magnet normally require active quench protection: upon detecting the quench, the goal is to create the largest normal zone in the shortest possible time. One well-known way to use a protection heaters that spreads quench propagation within the coil winding

As heaters are placed on the coil surface, there is an insulation barrier between heater and the supercondutor, and hence, an associated heat diffusion time.

When coils are short, a continuous heating strip of higher-resistivity material (usually stainless steel) can be used. Example shown: protection heater of the “HQ” LARP high-field quadrupole

For a long magnet, a continuous heater strip cannot be used due to large resistance, and therefore a concept of “heating stations” is used to induce quench zones with some periodicity along the coil. Example shown: heater of the LARP “LQ: long quadrupole.

Heater delay time (interval between firing the heater and the voltage development across the conductor) measured experimentally for the LQ magnet as function of the peak heater power

Uniform strip (t1 , t2=0)

Strip with heating stations (t1 , t2)

Numerical simulations of heater delay

A model has been developed calculate the heater delay; details can be found in:

Making heater periodicity commensurate with the twist pitch of the cable may be expected to further improves heater performance.If p = 2nw and l = (2n+/-1) w , then the supercurrent in all strands

This approach can potentially improve heater efficiency, as all cable strands will get resistive and start dissipating heat at once.

of the cable segment of length L=nl can be “interrupted” simultaneously by the normal zones created with n heating stations.

Experimental heater patterns for the existing LARP magnets

1. Heater patter for the HQ03 – new LARP quadrupole. Heating station periodicity is 2x the cable twist pitch.

Heater for the QXF - new LARP magnet for LHC luminosity upgrade

2. Heater pattern for the outer layer of the LHQ – a 3m LARP coil. The top portion design is based on the optimization algorithm [1-3] to provide a simultaneous quenching of high and low-field zones.The bottom portion is the old-style “LQ” pattern. The test is to be completed in September 2014.

2LPo2K-03

Introduction SQXF Coil 1 - a modification of the periodic heating station concept

Reduced width of the copper-plated bridges (more space available for holes)

Increased width of the heating station

Cooper-plated terminals of the heating stations to improve current flow uniformity

Optimization is done by minimizing the sum of (PH delay + quench propagation time between the HS).

Other designs with selective copper plating are currently in development for the QXF coils:

The following list of heater parameters have been by the LARP-CERN collaboration for QXF magnet development:

Inner layer heater layout:

Four heater strips per outer layer and two per inner layer are planned for redundancy

Power (W/cm2)

LHS (mm)

Period (mm)

PH delay (ms)

Tot. Delay (ms)

50 60 106.7 20.7 25.475 40 88.2 19.1 23.9

100 30 106.70 18.5 23.22125 30 87.2 17.1 22.8150 30 96 16 22.6

Power (W/cm2)

LHS (mm)

Period (mm)

PH delay (ms)

Tot. Delay (ms)

50 80 142.2 28.4 34.675 55 121.3 25.9 32.5

100 40 103.00 25.1 31.4125 40 116.2 23.2 30.6

150 40 128.4 21.6 30.4

0

5

10

15

20

25

30

35

0 10 20 30 40 50 60 70 80

Hea

ter

dela

y (m

s)

Longitudinal coverage (mm)

50 W/cm2

150 W/cm2

B = 7 T

Lines with 25 W/cm2 steps

Outer layer heater layout:

Alternative inner layer heater layout:

*by T. Salmi

*by E. Todesco

5 segments (303.5 mm length) will provide simultaneous quenching of all strands.

½ twist pitch base (60.7 mm period) :OL: 18 segments (Rheater = 1.6 ; W P/A (straight) = 161 W/cm2 at 150 V)IL: 16 segments (Rheater = 1.73 ; W P/A (straight) = 137 W/cm2 at 150 V)

Selective copper plating is the way to scale up the heater length to long QXF. (Still an option for the upcoming QXF coils)

T. Salmi, et al., IEEE, Trans. Appl. Supercond., 24 (3), 2014T. Salmi, et al., IEEE, Trans. Appl. Supercond., 24 (4), 2014T. Salmi, et al., Proc. WAMSDO, 2013

More simulations and experimental data:https://indico.cern.ch/event/311824/contribution/9/material/slides/1.pdf

0

10

20

30

40

50

60

70

20 40 60 80 100

PH d

elay

[ms]

Magnet current / SSL [%]

HQ02a (1.9 K)

HQ02a (2.2 K)

HQ02 (4.5 K)

HQ02b (1.9 K)

Sim., HQ02 (1.9 K)

Sim., HQ02 (4.5 K)

HQ01e (1.9 K)

HQ01e (4.4 K)

Sim., HQ01e (1.9 K)

Sim., HQ01e (4.5 K)

HQ01e, C8 (4.5 K)

HQ01, 25 µm polyimide

HQ02, 75 µm polyimide

Experimental results versus simulations for the HQ magnet

0

5

10

15

20

25

30

35

40

0 10 20 30 40 50

Hea

ter

dela

y (m

s)

HS length (mm)

50 W/cm2

150 W/cm2

B = 9 T

Lines with 10 W/cm2 steps

Simulation result (adjusted for the pitch length of 109 mm): τRC = 36 ms, R = 5.6 Ω, I = 80 A HS length = 18.32 mm period = 91 mm P(0) = 130 W/cm2

PH delay = 13 ms propagation between HS = 4

ms