CuZr-Mo bimetals for CLIC accelerating structures for CLIC - - PowerPoint PPT Presentation

▶
cuzr mo bimetals for clic accelerating structures for
SMART_READER_LITE
LIVE PREVIEW

CuZr-Mo bimetals for CLIC accelerating structures for CLIC - - PowerPoint PPT Presentation

May 23, 2023 155 likes •512 views

CuZr-Mo bimetals for CLIC accelerating structures for CLIC accelerating structures Introduction Introduction HIP diffusion bonding Explosion bonding p g Brazing-thermal treating Others Forum on Materal and Surface Technologies - 1 CERN

slide-1
SLIDE 1

CuZr-Mo bimetals for CLIC accelerating structures for CLIC accelerating structures

Introduction Introduction HIP diffusion bonding Explosion bonding p g Brazing-thermal treating Others

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

1

slide-2
SLIDE 2

Bimetals for CLIC. Introduction.

  • CLIC (Compact Linear Collider) two beam scheme:
  • Acceleration of main e+/e- beam to 1.5+1.5 TeV: use high accelerating

field to limit the machine length

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

2

  • Demonstration of technical feasibility. Structure technology development

(amongst other issues)

slide-3
SLIDE 3

Bimetals for CLIC. Introduction.

Hybrid Dumped accelerating Structures (HDS) concept.

▲ Copper prototype of ¼ of HDS structure containing 10 cell cavities. ▲Detail of HDS cells

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

3

g ▲Detail of HDS cells geometry.

slide-4
SLIDE 4

The material problem. High H field regions.

Periphery: regions with

pulses of magnetic field

◄Surface magnetic

inducing surface currents

∆T= 56 K, 2.3x1010 cycles

l ti i g field distribution in HDS cell.

pulsating compressive

stress 0 to 155 MPa

fatigue surface damage

▼Surface of Cu and CuZr specimens after equivalent tests of

use of CuZr, or improved

h i l t th hi h thermal induced fatigue (laser simulation) mechanical strength high conductivity alloy.

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

4

CuZr Cu

slide-5
SLIDE 5

The material problem. High E field regions.

Iris: regions with surface

electric field >300 MV/m

◄Surface electric

(originally)

high field and breakdown

events field distribution in HDS cell. events

geometry modification

▼Accelerating structures in Mo and Cu after RF

use of Mo, or alternative

refractory metal. and Cu after RF tests at SLAC.

But found that for low

breakdown probability Mo is less favourable : the baseline is now all the structure made of copper- b d t i l !

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

5

based material !

slide-6
SLIDE 6

Aims

Bi t lli CLIC l ti t t b d d f h d

  • Bimetallic CLIC accelerating structures may be needed for enhanced

performance

  • Regions of high electric field:

pure Mo (or alternative refractory & low P pure Mo (or alternative refractory & low Pvap

alloy)

  • Regions of high pulsed currents:

C Z ll UNS C15000 i th b t

CuZr alloy UNS C15000 in the best

achievable temper state (or alternative copper alloy with best match of conductivity and fatigue resistance: ODS Cu, CuCrZr, …)

Find a way of producing bimetallic raw material with:

  • suitable geometry
  • sufficient bond strength and soundness
  • limited effect on bulk material properties

Explore and develop bonding methods Quantify their performance

set up acceptance criteria

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

6

set up acceptance criteria

slide-7
SLIDE 7

Bonding techniques

  • Promising bonding techniques explored so far:
  • HIP diffusion bonding (METSO Oy /FI)
  • Explosion bonding (Research Institute of Impulse Processes / BY)
  • Brazing (Politecnico di Torino /IT, running)
  • Also attempts of coextrusion and vacuum casting
  • Also attempts of coextrusion and vacuum casting
  • Reminds on the materials:
  • Pure Molybdenum
  • Unsuitable DBTT if recrystallised (recrystallisation may occur at

temperatures above 900 °C depending on time exposure and previous cold k) work)

  • CuZr, UNS C15000
  • Solution annealing at 900 °C to 980 °C (5 min to 30 min), followed by rapid

cooling

  • Artificial ageing 450 °C to 550 °C (1 h to 4 h)
  • The best (mechanical strength) temper states are obtained by addition of

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

7

cold work before or/and after ageing.

slide-8
SLIDE 8

Precipitation hardening (applied to CuZr)

The aim is to have

400

The aim is to have

  • fine precipitates for optimum mechanical strength
  • and minimum Zr dissolved in the matrix for

maximum conductivity Cu-Zr equilibrium diagram

220 200 205 316 358

250 300 350 400 Cu-OFE OS050 C15000

The way is with a series of thermal treatments: 1. Solution heat treatment: go and stay in the solubility region (900 to 980 ºC) to dissolve Cu5Zr 2. Quench: end by fast cooling to metastably keep

Solid (Cu) + C Z li id Cu-Zr liquid

69 45 100 41 200 54 64 90 50 91 15 92

50 100 150 200 TB00 Solution annealed, no cold worked TF00 Aged, no cold worked TH02 Cold worked 45% and aged

2. Quench: end by fast cooling to metastably keep that dissolved microstructure at RT 3. Age or precipitate: stay a limited time at intermediate temperature (450 to 550 ºC) to favour fine precipitation

Cu-Zr liquid Solid (Cu)

972 ºC

15

50 Rp0.5 (MPa) Rm (MPa) A (%) (%IACS) Yield Strength Ultimate Strength Ductility

  • Elec. Cond.

fine precipitation

  • Do not over-age: by doing to hot or two long such

that precipitates become coarse

  • Avoid a too slow cooling
  • r

ure ºC

Solid (Cu) matrix +

  • r

Avoid a too slow cooling

  • To boost hardening cold-work can be added before

and/or after ageing

  • r

Temperatu

Cu5Zr

  • r

0 15 %Zr

Weight %Zr

20 ºC Temperature Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

8 0.15 %Zr C15000

g

T Time

slide-9
SLIDE 9

HIP diffusion bonding

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

9

slide-10
SLIDE 10

HIP diffusion. Bonding advantages and concerns

HIP-diffusion bonding (Metso OY /FI)

Cylindrical configuration Mo insert CuZr matrix Cylindrical configuration, Mo insert, CuZr matrix Attempt to have the CuZr in a solution treated state right after the

HIP cycle

  • HIP temperature set up to match solution treatment

temperature of CuZr (900 ºC)

  • Cooling after HIP as fast as possible (HIP quenching)

Cooling after HIP as fast as possible (HIP quenching)

Test pieces produced for characterization One piece produced for machining a first bimetallic HDS prototype

Concerns:

Soundness of the bond Avoid recrystallization of the Mo insert

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

10

Achievable strength on the CuZr matrix

slide-11
SLIDE 11

HIP diffusion. Test pieces produced

1st piece:

  • Ø50 mm x 100 mm, insert Ø 5 mm

,

  • from extruded CuZr bar commercially available

2nd piece:

  • more careful preparation of surfaces

faster cooling capable size for HDS structures

  • Ø 87 mm x 300 mm, insert Ø 8.6 mm
  • from forged CuZr bar for functional diameter

3nd piece: 3

piece:

  • Ø 87 mm x 300 mm, insert Ø 8.6 mm
  • insert coated with pure opper

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

11

slide-12
SLIDE 12

HIP diffusion. Tests performed.

Bond characterization

  • Bond characterization
  • Bond strength
  • Shear tests
  • Pull tests
  • Pull tests
  • Fractography
  • Microstructure
  • Metallography OM-SEM

0.5µm

50000

Al (111) [Holder]

Mo

  • Metallography OM-SEM
  • EDS diffusion profile, phases
  • XRD
  • Mo base material (recrystallisation?)

Mo CuZr

)

40000

  • Mo base material (recrystallisation?)
  • Metallography / hardness
  • CuZr base material characterization

As received

Lin (Counts)

20000 30000

02) [Holder]

25 30

ion [w%]

Cu Mo Zr

  • As received
  • Grain size
  • Hardness
  • Tensile tests

10000

Al (20 Cu (111) Mo (110) Mo2Zr (311)

10 15 20

malised concentrat

Zr

CuZr matrix Mo insert

  • Tensile tests
  • Electrical conductivity
  • After further treatment
  • Idem

S f f M d f AS Ni filt Fil AS ll d

2-Theta - Scale

36 40

5 2 2 4 6 8

Distance from the interface [µm] Norm

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

12

  • Idem
  • Machining test
slide-13
SLIDE 13

HIP diffusion. As produced. Bond strength

  • 1st piece showed several detachments
  • 2nd piece

2 piece

  • Shear:
  • 30 % shear in CuZr, 70 % interface
  • Pull:

☺ Plastic deformation of CuZr before breakdown ☺ Final breakdown partially through Mo 1 over 4 presented poor adhesion

1 st 2 nd

average

1

average

Shear strength ( MPa)

2 0 6 1 6 2 1 7 4 1 6 3 1 7 4 1 5 3

1 6 5 ± 9

Pull strength ( MPa)

detached 1 8 3 1 8 7 1 5 4 detach.

1 7 5 ± 1 8

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

13

( )

slide-14
SLIDE 14

HIP diffusion. As produced. Bond microstructure

1 st piece

  • Showed local lack of adhesion
  • Regions with porosity at the interface
  • Regions with porosity at the interface
  • Composition profile:
  • diffusion of Cu in Mo along 14 µm, in

agreement with calculations agreement with calculations

  • Intermetallic layer Zr-rich, confirmed by

XRD

Mo CuZr Mo CuZr 100

Al (111) [H ld ]

70 80 90

ration [w%]

Cu Mo Zr

40000 50000

Al (111) [Holder]

30 40 50 60

ised concent CuZr matrix Mo insert

Lin (Counts)

20000 30000

[Holder]

10 20 30

Normal

10000

Al (202) Cu (111) Mo (110) Mo2Zr (311)

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

14

5 5 10 15 20

Distance from the interface [µm]

S f f M d f AS Ni filt Fil AS ll d

2-Theta - Scale

36 40

slide-15
SLIDE 15

HIP diffusion. As HIPed. Bond

C Z CuZr Mo 2 nd piece

  • Diffusion Cu in Mo over

D C A

CuZr Mo

B A B

Diffusion Cu in Mo over 16 μm

☺

Continuous interface, low porosity CuZr Mo

  • Intermetallic layer (A)
  • Three characteristic areas in

the CuZr side:

  • B : precipitates with Zr and Mo
  • C : area depleted in Zr
  • D : precipitates with Zr

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

15

p p

slide-16
SLIDE 16

HIP diffusion. As HIPed. Mo insert

  • Grain size : 70 μm
  • Hardness : 230 HV

(if recrystallized: 140 160 HV) (if recrystallized: 140–160 HV) M h t t lli d d it

Mo has not recrystallized despite

  • f the long stays at 900 ºC

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

16

slide-17
SLIDE 17

HIP diffusion. As produced. CuZr matrix properties

  • Effect of long HIP cycles on the

CuZr matrix

  • softening

CuZr before HI P CuZr after HI P 1 st 2 nd 1 st 2 nd G i i 7 2 1 1 1

  • softening
  • but very limited grain growth

Grain size ( µm ) 6 0 7 2 ( het.) 7 0 1 1 1 ( het.) Hardness ( HV5 ) 1 2 2 8 7 4 8 5 4

  • Mechanical and electrical

properties

CuZr As HIP 2nd CuZr As HIP 1st 1st (by Metso) 199212 217 200 250 58 88 78 82 64 55 84 50 100 150 58 36 40 55 50 Rp0.2 (MPa) Rm (MPa) A (%) %IACS

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

17

slide-18
SLIDE 18

HIP diffusion. Post HIP treatments

HIP “ h d” d

350 400

  • HIP “quenched” + aged

316 358

350 400

212 199 215

200 250 300

1 st HIP quenched 2nd HIP quenched 1 st HIP quenched + aged 550 ºC 3h 212 199 215 200 205 316 220

200 250 300

1 st HIP quenched 2nd HIP quenched 1 st HIP quenched + aged 550 ºC 3h *TB00Solution annealed nocold worked 86 40 82 64 36 88 70 55 94

50 100 150 200

86 40 82 64 36 88 70 55 94 41 54 64 90 50 91 1 92 69 45 100

50 100 150 200

*TB00 Solution annealed, no cold worked *TF00 Aged, no cold worked **TH02 Cold worked 45% and aged *Cu-OFE OS050 * ASM Handbook ** O k

50 Rp0.5 (MPa) Rm (MPa) A (%) (%IACS) Yield Strength Ultimate Strength Ductility

  • Elec. Cond.

15

50 Rp0.5 (MPa) Rm (MPa) A (%) (%IACS) Yield Strength Ultimate Strength Ductility

  • Elec. Cond.

** Outokumpu

HIP “quenching” is not fast enough to

properly solution anneal the alloy

HIP + sol. treat. & quenched

  • Trials to do HIP + solution annealing +

water quench + ageing

Systematic detachment of the Mo insert is

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

18

  • bserved in water quenched specimens

As HIP

slide-19
SLIDE 19

HIP diffusion. Machining tests

  • Successful milling of an HDS6

structure from 2nd piece as HIPed

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

19

slide-20
SLIDE 20

Summary of results and future. HIP diffusion

Sum m ary of results

Bond soundness

Good strength and absence of interface voids are attainable

after HIP cycle

Mo insert Mo insert

Does not recrystallise during long HIP cycles

Strength of the CuZr matrix Strength of the CuZr matrix

No proper ageing can be obtained No cold-w ork (easily) applicable

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

20

slide-21
SLIDE 21

Summary of results and future. HIP diffusion

Future studies and developm ents

  • Machine the first bimetallic HDS prototype using a HIP DB rod
  • Machine the first bimetallic HDS prototype using a HIP-DB rod
  • Study alternatives to improve homogeneity of the bond:
  • trial with pure Cu interlayer sputtered on the Mo rod is underway It has to
  • trial with pure Cu interlayer sputtered on the Mo rod is underway. It has to

be characterized.

  • Optimize strength and electrical conductivity of CuZr matrix after HIP

by a post-HIP ageing

  • Study possible post-treatment to add cold-work to the matrix: Cold

( ) l d h b ll b Isostatic Pressure (CIP), explosion around the bimetallic bar

  • Study performance applied to other alloys ( Cu-OFE, GlidCop, …) and

(refractory) metals (refractory) metals

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

21

slide-22
SLIDE 22

Explosion bonding. Bonding process principle

“Solid state welding process that is used process that is used for the metallurgical joining of dissimilar

  • metals. The process

uses the forces of controlled detonations to accelerate one metal plate into metal plate into another creating an atomic bond… is considered a cold- welding process which allows metals to be joined without losing their pre- losing their pre- bonded properties.”

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

22

slide-23
SLIDE 23
  • Explosion. Bonding advantages and concerns

Explosion bonding (Research Institute of Impulse Processes /BY)

Flat configuration using CuZr back plate already in an optimum Flat configuration, using CuZr back plate already in an optimum

cold-worked and aged state

Bonding process does not affect the starting temper of Mo and CuZr Test pieces produced for characterization Pieces for machining of bimetallic HDS prototype available

C

Concerns

Soundness of the bond Avoid damaging of the starting materials Avoid damaging of the starting materials

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

23

slide-24
SLIDE 24
  • Explosion. Test pieces produced
  • Series 1: Piece with a groove
  • 0.8 mm Mo and pure Cu proposed by the

I tit t Institute

  • Series 2: Three flat pieces for characterization
  • Mo thicknesses: 1 mm, 2 mm and 3 mm

from Plansee grade FG013,

  • CuZr plates machined from Luvata rods Ø50 mm,

40%CW and aged

  • Series 3 and 4: Two new pieces available for

production of structure prototypes

  • Same explosion conditions as 2 mm of Series 2
  • Mo thickness 2 mm
  • CuZr plate as previous pieces
  • CuZr plate from Hitachi hot extruded rod of Ø100

mm.

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

24

slide-25
SLIDE 25
  • Explosion. Series 1
  • Crest wave morphology of the interface

Mo

  • In the flat region good interface integrity

Below the groove produced by rolling Mo is not

Cu

  • Below the groove produced by rolling, Mo is not

debonding but delaminating

Cu M Cu Mo Mo

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

25

slide-26
SLIDE 26
  • Explosion. Series 2
  • US inspections enables to detect

bonding defects (found in the f piece of 1 mm Mo)

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

26

slide-27
SLIDE 27
  • Explosion. Series 2
  • Microstructural observation
  • Different wave amplitude

Mo

100 µm Piece 1 mm Mo

p

  • No voids or cracks at the interface
  • No cracks on the Mo

I l li t 30 f Cu

  • Irregular line at approx 30 µm away of

the interface visible only on etched sample

Piece 2 mm Mo Piece 2 mm Mo Piece 3 mm Mo

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

27

slide-28
SLIDE 28
  • Explosion. Series 2
  • Hardness increase in the narrow distorted regions adjacent to the

interface.

S l 2 M Sample 2 mm Mo

450,0 500,0 350,0 400,0

)

CuZr Perpendicular CuZr Paralel 250,0 300,0

dness (HV0.2)

CuZr Paralel CuZr Average Mo Perpendicular Mo Paralel Mo Average Mo, close to intertfce 100,0 150,0 200,0

Hard

Mo, close to intertfce CuZr, close to intertfce CuZr, close to intertfce 0,0 50,0

  • 1
  • 0 5

0 5 1 1 5 2

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

28

1 0,5 0,5 1 1,5 2

Position (mm)

slide-29
SLIDE 29
  • Explosion. Series 2
  • Bond strength
  • Tests adapted from standards

1 cm

p

  • Shear test: ASTM B 898 modified
  • Pull test: Ram Tensile Test MIL-J-24445 A

modified

1 cm

modified

  • Higher strength values than Series2
  • Comparable to as HIP diffusion bonded
  • Fracture always on the Mo close to the

interface showing a wavy aspect 1 st 2 nd

1 mm 2 mm 3 mm Shear strength up ( MPa)

1 1 0 ±19 2 6 1 3 0 9 2 2 5

Shear strength dow n ( MPa)

2 4 9 2 2 7 2 0 0

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

29 Pull strength ( MPa)

  • 7 7

1 2 1 ±7 2

slide-30
SLIDE 30
  • Explosion. Machining test
  • Successful milling test of a piece with three teeth

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

30

slide-31
SLIDE 31

Summary of results and future. Explosion bonding

Sum m ary of results

Feasibility demonstrated Feasibility demonstrated First prototypes for fabrication are available CuZr maintains its temper state CuZr maintains its temper state

Future studies and developm ents Future studies and developm ents

Machine the first bimetallic HDS prototype using the available

plates plates

Adapt the geometry from flat to (locally) cylindrical

St d h th t h i l t i d ti d l

Study how the technique scales to series production and large

pieces (flatness homogeneity, …) St d pe fo mance applied to alte nati e allo s (GlidCop )

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

31

Study performance applied to alternative alloys (GlidCop, …)

slide-32
SLIDE 32

Brazing - Thermal treating. Principle

  • Simultaneous brazing and thermal

treatments (Politecnico di Torino / IT)

  • Flat configuration, using CuZr back plate.

Flat configuration, using CuZr back plate. Possibly other.

  • Attempt to have the CuZr in a solution

treated state after brazing and fast cooling + ageing in the same furnace

  • Brazing temperature set up to coincide

with solution treatment temperature of CuZr: 975 ºC, Cu-Ni-Ge braze,

  • Thermal cycle under vacuum except for

the “fast” cooling from 975 ºC to 500 ºC by injection of N2

  • Continuation with ageing cycle at 500 °C

3h 3h

  • Test pieces to be produced for

characterization

  • Concerns:
  • Soundness of the bond
  • Avoid recrystallization of the Mo insert

A hi d t th f th C Z t i

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

32

  • Achieved strength of the CuZr matrix
slide-33
SLIDE 33
  • Brazing. First results
  • Series of shear test specimens for deciding number of braze folds
  • Shear strength not much dependent of number of braze folds

Shear strength (MPa) 178,18 146,78 124,55 154,49 135,74 107,89 158,07 126,75 102,51 120 140 160 180 200 a 20 40 60 80 100 MPa 3-1 3-2 3-3 2-1 2-1 2-3 1-1 1-2 1-3 3 fold 2 fold 1 fold Specimen

  • Some lack of wetting of the braze
  • Severe loose of hardness with respect to

the original 40 %cw and aged temper: 120 HV vs 45 HV

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

33

120 HV10 vs 45 HV10

slide-34
SLIDE 34

Summary of results and future. Brazing

Sum m ary of results

Bond strength seems comparable to previous techniques Bond strength seems comparable to previous techniques Bond soundness is doubtful CuZr annealed (such as in HIP quenching) CuZr annealed (such as in HIP quenching)

Future studies and developm ents Future studies and developm ents

Production of a capable piece for machining prototypes

Ad t th t f fl t t l ll li d i l

Adapt the geometry from flat to locally cylindrical

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

34

slide-35
SLIDE 35

Other techniques

HI P diffusion bonding, CuZr/ Mo Coextrusion, Cu/ Ni/ Mo 1 0 m m 1 0 m m

▲Coextrusion with intermediate layer (Lutch) intermediate layer (Lutch). ◄Vacuum casting over a lid i t (St k)

Forum on Materal and Surface Technologies - CERN 21/11/2008

  • G. Arnau Izquierdo

35

solid insert (Starck).