fessia paolo, david smekens and the whole te-msc-mdt section

26
Fessia Paolo, David Smekens and the whole TE-MSC-MDT section Insulation strategies and materials for incoming SC magnets

Upload: summer

Post on 23-Feb-2016

76 views

Category:

Documents


0 download

DESCRIPTION

Fessia Paolo, David Smekens and the whole TE-MSC-MDT section. Insulation strategies and materials for incoming SC magnets. Outline . Working conditions Ground Insulation Voltages T echniques Cable insulation Nb -Ti Nb 3 Sn Quench heaters. Conditions of max voltage. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Fessia Paolo, David Smekens and the whole TE-MSC-MDT section

Insulation strategies and materials for incoming SC magnets

Page 2: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Outline

• Working conditions• Ground Insulation

• Voltages• Techniques

• Cable insulation• Nb-Ti• Nb3Sn

• Quench heaters

Page 3: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Conditions of max voltageThe maximum voltage is reached during the quench of the magnet

that up to that point was in an He bath (hp. 1.9K). We can assume that we need to be able to insulate the magnet in

He gas at 75 K 1 bar (conservative)

0

100

200

300

400

500

600

700

800

900

1000

1100

0 10 20 30 40 50 60 70 80 90 100 110 120 130

Volta

ge to

gro

und

[V]

T [K]-100

0

100

200

300

400

500

600

700

800

0 10 20 30 40 50 60

Volta

ge to

gro

und

(V)

Hot spot Temp (K)

MQXC Dump Voltage v hot spot.

Courtesy G. Kirby Courtesy E. Todesco

MQXC Nb-Ti 120 mm bore quad voltage increase MQXF Nb3Sn 150 mm bore quad voltage increase

Page 4: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Environment dielectric in function of temperature and pressure

100

1000

10000

0.001 0.01 0.1 1 10 100

Brek

adw

own

volta

ge [V

]

Electrode distance [mm]

Air Breakdown voltage in function of electrode distance 275K and 1 bar

275 K AIR 1 bar

100

1000

10000

0.001 0.01 0.1 1 10 100

Brek

adw

own

volta

ge [V

]

Electrode distance [mm]

Helium and Air Breakdown voltage in function of electrode distance in few selected pressure and temperature conditions

275 K P 1 bar 275 K AIR 1 bar

100

1000

10000

0.001 0.01 0.1 1 10 100

Brek

adw

own

volta

ge [V

]

Electrode distance [mm]

Helium and Air Breakdown voltage in function of electrode distance in few selected pressure and temperature conditions

275 K P 1 bar

275 K AIR 1 bar

75 K P 1 bar

3.5

Page 5: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

GROUND INSULATION

Page 6: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Vq• Starting point: quench voltage

Vmaxr

• Maximum reference voltage: assumed to be Vq plus 450 V+150 V (difference voltage between circuits/and or extraction voltage)

Vtql• Voltage for test qualification in liquid helium =1.2 Vmaxr

Vtqgas• Voltage for test qualification in He gas = 0.5 X Vtql

Vtfdel• Voltage in air equivalent to Vtqgas: Vtfdel = 3 X Vtqgas

ΔVcryo

, ΔVcold

• Difference in voltage between 1st and last test during cold test and cryostating =0.1 X Vmaxr

Vtaf• voltage for final assembly test= Vtfdel+ ΔVcryo+ ΔVcold

Vtai• Voltage for initial assembly test= Vtaf+ 0.2 X Vmaxr

1000 V

1600 V

1920 V

960 V

2880 V

160 V

3200 V

3520 V

Vds 3X1600

V+500 V= 5300 VVds air=

3.5X Vds= 18550 V

Page 7: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Examples the ground insulation of the MQXC

Page 8: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Few Magnet examples I LHC MB

Material:Polyimide foils 0.125 mm thick

Number of foils: 4

Minimum number of foils continuous: 2Minimum overlap 7 mm

Creep path >> 7 mm

MQ

Material:Polyimide foils 0.125 mm thick

Number of foils: 4

Minimum number of foils continuous: 2Minimum overlap 10 mm

Creep path >> 10 mm

MQXC

Material:Polyimide foils 0.125 mm thick

Number of foils: 4

Minimum number of foils continuous: 2Minimum overlap 10 mm

Creep path >> 10 mm

4-5 layer Polyimide 0.125 mm each

minimum overlap 10 mmCreep path >> 12 mm

Page 9: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

INTER-TURN

Page 10: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

The Nb-Ti insulation schemes (polyimide based)

Insulationtype 1st layer 2nd layer 3rd layer

MB

11 mm wide,no gap,

50 µm thick

11 mm wide, no gap,50 µm thick

50% overlap 1st layer

9 mm wide, 2 mm gap,69 µm thick,

cross wrapped

MQ

11 mm wide,no gap,

50 µm thick

11 mm wide, no gap,37.5 µm thick

50% overlap 1st layer

9 mm wide, 2 mm gap,55 µm thick,

cross wrapped

EI3

9 mm wide,no gap,

50 µm thick

3 mm wide, 1.5 mm gap,50 µm thick

cross wrapped

9 mm wide, 1 mm gap,69 µm thick,

50% overlap 1st layer

EI4

9 mm wide,no gap,

50 µm thick

3 mm wide, 1.5 mm gap,75 µm thick

cross wrapped

9 mm wide, 1 mm gap,69 µm thick,

50% overlap 1st layer

Page 11: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

New enhanced scheme

Page 12: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Specifications the Nb3Sn inter-turn insulation system

Voltage withstand level in cryogenic conditions

100-200 V

Mechanically withstand large compression stresses and some bending and shear

Stand High Radiation dose

(typical 40-50 MGy with shielding 100 MGy without shielding)

Do no deteriorate in cryogenic conditions

Be of repeatable, constant and

controllable thickness

Parts to be applied before Nb3Sn reaction shall withstand reaction

(typically 650 C for 100 hours)

Do no damage cable during the application

Turn to turn insulation system

Page 13: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Cold Maximum gradient

Azimuthal

Shear

MQXF preliminary coil stresses evaluation

13

Page 14: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Nb3Sn Turn to turn insulation Possible Approaches

Standard approach Insulator based approach Reinforced approach• “Thick” fiber cable

protection• Resin impregnation to

provide the dielectric

• Add a thin thermal resistant insulator (MICA) that provides the dielectric

• Impregnation is “only” structural

• Add a thin thermal resistant thin insulator (MICA) the provides the dielectric to “thick” fiber cable protection

• Insure that the fiber+ impregnation is redundant in terms of insulation

• Better to get fibers that do not present carbon deposition after reaction

• Local resins bubble could lead to local shorts if the fiber is not thick enough or damaged

• Impregnation provides second level of defense in case of 1st insulator breaks, but it is not necessarily dimensioned in terms of distance to keep all the inter-turn in case of fiber squeezing

• Relaxed specification on the fiber

• Better to get fibers that do not present carbon deposition after reaction

• Impregnation provides second level of defense in case of 1st insulator breaks, but it is not necessarily dimensioned in terms of distance to keep all the inter-turn in case of fiber squeezing

• Absence of carbon will reassure also about the bonding impregnation to fiber

• Main issue real estate, but would be possible to limit this application to the insulation between turn and metal posts (winding post and end spacers?)

Page 15: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

100

1000

10000

0.001 0.01 0.1 1 10 100

brea

kdow

n vo

ltage

[V]

Electrode distance [mm]

Gas Helium electrical breakdown voltage in function of electrode distance at 75 K and 1 bar

0.2 mm350 V

0.1 mm250 V

0.3 mm450 V

0.4 mm550 V

If we assume that we have inter-turn voltage in the order of max 100-200 V this could lead to design for about 300 V-600 V cryogenic conditions.

In case of needing good insulation also in presence of cracks, this lead to cable distances of about 0.2-0.4 mm.

Page 16: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Standard and reinforced approach Material considerationsPossible cable insulation processes evaluated at CERN1. Tape

1. Problems in finding in Europe adequate tape quality for geometry and presence of polyester guiding wires on the edges. No S2 tape found on the Europe market

2. Problems to achieve good overlapping2. Sock type sleeve

1. Very good results2. Problem to find European source3. Limited to cable of about 100 m

3. Direct cable braiding1. See following slides

Page 17: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Direct braiding on cableWe have insulated

- 10 mm wide cable- 15 mm wide cable- 21 mm wide cable

Very promising under all point of view.Only issue, and experienced by us: educate the company to increase

Q.C., awareness to avoid and identify possible cable damages that could

occur during the braiding

Page 18: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Insulator based approach

Braiding of 11 TEX plus MICA

Winding of copper coil 11 TEX+ MICA

Page 19: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Yarn for braidingUse standard yarn with organic sizing

for the braiding

Easy, fast and cheap process

Carbon residuals after reaction

Use standard yarn with organic sizing

Thermally or chemically de- size

Apply the palmitic acid or other sizing on

the yarn

Braid

No carbon residuals

Long, complex and costly operation

Probable deterioration of fibers during handling

without sizing and thermal desizing

Use special yarn with temperature stable

sizing

Braid

No carbon residuals

Easy and fast process

Average costs

AGY933 S2 glass

Page 20: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

20

Fiber Glass: tensile test during process

876

550

210

10

1131

1540

1870

1600

759

440 390480

425365 344 300

195 200 181

1837

100

100200300400500600700800900

10001100120013001400150016001700180019002000

Ulti

mat

e St

reng

th [M

Pa]

Page 21: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

21

Glass fiber after 100 hours at 650CAGY S 2 493 AGY S 2 933 AGY S 2 636 E glass

S2 933 temperature behavior

Page 22: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Metal component belonging to the coilsThese parts (end spacer, winding posts) have to be strongly insulated to the coil and the insulation applied has to withstand reaction. Present practice is to

overlap an insulation coating (Plasma Spray Al203) and add an extra protective cloth of fiber glass that will be later impregnated wit the coil

R6

R5R4

R3R2R1

Central Post Spacers set R (GΩ) @ 1000 V ( R1 to R6)

R (GΩ) @ 2500 V (R1 to R6)

Central post 1 1 73 52

Central post 1 2 73 52

Central post 2 1 73 52

Central post 2 2 73 52

Central Post 1 after

Thermal cycle @ 650 0C

1 73 52

Central Post 1 after

Thermal cycle @ 77 K

1 73 52

Page 23: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

QUENCH HEATERS

Page 24: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Quench Heaters

Requirement for Q.H. insulation to

coil

In cryogenic condition withstandV=1.2 (Vq+900 V)

Provide very efficient thermal

transfer to the coil(typical thickness 0.125 mm

or less)

Be robust and allow shaping to

follow the coil geometry

Vq Vq Vq Vq 250 V 350 V 500 V 1000 V

Vmic-c 1380 1500 1680 2000

Vmic-w 455 495 554 660

VQH-coil 1366 1485 1663 1980

VtfQHdel 1366 1485 1663 1980

ΔVcryo, ΔVcold 90 90 90 90

VtQHaf 1546 1665 1843 2160

VtQHai 1726 1845 2023 2340

Page 25: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Other issues to remind• A Nb3Sn impregnated coil has a limited thermal conduction and

limited electrical insulation: this leads to problems install effective and electrical safe quench heaters to protect magnets.

• Quench heaters at the Outer Radius might not be sufficient, as the heat does not immediately penetrate the coil. Inside the bore and inside the inter-layer of the coil, QHs are very technologically challenging.

• Instrumentation (voltage taps) are also difficult to install as the have to go through the insulation (either VTs installed before reaction, and survive the reaction and potting ; or installed after impregnation by digging into the epoxy to reach the bare conductor). All instrumentation circuit should be tested with the magnet and wire have sufficient voltage insulation rating

Page 26: Fessia Paolo, David Smekens and the  whole  TE-MSC-MDT section

Questions ?

LOGO