Metallographic Studies of Electron Beam Welded Copper Lids - Posiva

Working Report 2014-23

Metallographic Studies of Electron Beam Welded Copper Lids: EBSD Studies of the Cross-Section of XK049 at 323deg Taru Karhula

June 2014 POSIVA OY Olkiluoto FI-27160 EURAJOKI, FINLAND Phone (02) 8372 31 (nat.), (+358-2-) 8372 31 (int.) Fax (02) 8372 3809 (nat.), (+358-2-) 8372 3809 (int.)

Working Report 2014-23

Metallographic Studies of Electron Beam Welded Copper Lids: EBSD Studies of the Cross-Section of XK049 at 323deg Taru Karhula Tampere University of Technology Department of Materials Science

June 2014

Working Reports contain information on work in progress or pending completion.

ABSTRACT This work is part of Posiva’s spent nuclear fuel disposal canister sealing development. Posiva has welded a full-scale lid to a canister 450 mm or 890 mm tall at Patria Aviation Facilities. At Tampere University of Technology, Department of Materials Science (TUT DMS), metallographic and electron microscopy studies of electron beam welded copper lid welding have been carried out. The methods used in the metallographic studies are presented together with the results. In this report a part of the welding test program is analyzed. The results of the electron backscatter diffraction (EBSD) measurements of the cross-section of the test weld XK049 at 323deg are presented here together with the methods used. The aim of this study was to estimate the residual stresses present in the EB-weld using EBSD technique. In previous study based on EBSD reference curves with tensile test samples it was found out that the recrystallized, substructured and deformed fractions of grains, the occurrence of 1.5º and 2.5º misorientations, the average of mean intra-grain misorientation AMIS, and the hardness could be related to the applied strain. The hardness results for the samples used in this report are presented in another report (to be published 2014). Based on the recrystallized, substructured and deformed fractions of grains, the estimated strain level in the sample was 0.014-0.019. Thus, the estimated maximum residual stresses in the weld were in the range 50-56 MPa. Keywords: EBW, electron beam welding, copper, lid, canister weld, EBSD, electron backscatter diffraction, residual stress.

Elektronisuihkuhitsatun kuparikannen metallurgiset tutkimukset: EBSD tutkimukset XK049 hitsin poikkileikkauksesta kohdalta 323° TIIVISTELMÄ Tämä työ on osa Posivan käytetyn ydinpolttoaineen kapselin sulkemiskehitystyötä. Posiva Oy on hitsannut täyden mittakaavan kansia 450 mm pitkiin kanisteriputkiin Patria Aviation Oy:n tiloissa. Tampereen teknillisen yliopiston Materiaaliopin laitoksella (TTY MOL) on tutkittu näitä elektronisuihkuhitsattuja kansihitsejä metallografisin ja elektronimikrosopian menetelmin. Tämän raportin yhteydessä on analysoitu vain osa hitsauskoeohjelman hitseistä. Tässä raportissa esitellään hitsin XK049 poikkileikkauksen tulokset sekä tutkimuksissa käytetyt menetelmät. Tämän tutkimuksen perimmäinen tarkoitus oli arvioida jäännösjännityksen suuruutta kyseisessä elektronisuihkuhitsatussa kanisterissa käyttäen EBSD-tekniikkaa. Aikaisemmassa tutkimuksessa vetokoenäytteiden EBSD-datan perusteella todettiin, että rekristallisoituneiden, alirakenteisten ja deformoituneiden rakeiden osuudet, 1.5° ja 2.5° misorientaatioiden esiintymät, keskiarvo rakeiden sisäisestä keskimääräisestä misorientaatiosta sekä kovuus näyttivät olevan riippuvaisia näytteen venymästä. Saman näytteen kovuusmittaustulokset on esitelty toisessa raportissa (julkaistaan 2014). Rekristallisoituneiden, alirakenteisten ja deformoituneiden rakeinen osuuksien mukaan arvioituna näytteessä oleva plastinen venymä oli 0.014-0.019. Tästä arvioituna, jäännösjännityksen maksimitaso oli 50-56 MPa. Avainsanat: EBW, elektronisuihkuhitsaus, kupari, kansi, kapselin hitsaus, kapselin hitsaus, EBSD, jäännösjännitys.

PREFACE This report is one part of the residual stress evaluation of the EB-welded nuclear waste canister made of copper. Target of the residual stress evaluation is to assess level of the residuals stresses of the weld and also effect of the welding parameters and stress relief on residual stresses. Evaluation can be divided on following parts: -

Evaluation deformations and residual stresses on EB-welds using numerical modeling (finite element method, FEM) Residual stress measurements of the EB-welds using different measurement methods Destructive testing of the weld to assess residual stresses

Results of the EBSD studies of the lid weld XK049 are reported in this report. Target was to evaluate work hardening and deformation of the weld caused by residual stresses and estimate residual stresses using calibration curves defined in Posiva working report WR2013-14. The macroscopic evaluation and hardness measurement of the lid weld XK049 to evaluate plastic deformation is reported in separate report along with this report.

1

TABLE OF CONTENTS ABSTRACT TIIVISTELMÄ PREFACE LIST OF SYMBOLS AND ABBREVIATIONS ................................................................. 3 1.

INTRODUCTION .................................................................................................... 5

2.

WELDING PARAMETERS ..................................................................................... 7

3.

RESEARCH METHODS ......................................................................................... 9 3.1.

Identification of sample ................................................................................... 9

3.2.

Sample preparation ....................................................................................... 10

3.3.

EBSD studies ................................................................................................ 11

4.

MAIN RESULTS ................................................................................................... 13

5.

ANALYSIS AND DISCUSSION ............................................................................ 17

6.

5.1.

The remaining plastic strain in the weld ........................................................ 17

5.2.

Estimated residual stress in the weld ............................................................ 17

5.3.

General notices for strain and stress determinations .................................... 18

CONCLUSIONS.................................................................................................... 19

REFERENCES ............................................................................................................. 21 APPENDICES............................................................................................................... 23

2

3 LIST OF SYMBOLS AND ABBREVIATIONS A

breadth of cross-sectional sample

[mm]

AA

working distance

[mm]

AF

focal length

[mm]

ACX

oscillation input in x-direction for EB-machine

ACY

oscillation input in y-direction for EB-machine

AMIS

average of the mean intra-grain misorientation

av(G)

average of grades

av(s)

average of penetrations

[mm]

α

angle of groove

[º]

β

angle of slot

[º]

B

measure from the weld centerline to lid side of cross-sectional sample

[º]

[mm]

BC

band contrast

BSE

backscattered electron

C

distance from inner surface of tube to inner surface of lid

[mm]

D

distance from inner surface of tube to outer surface of tube

[mm]

Delta

the highest minus the lowest average for each factor

E

distance from weld centerline to inner surface of tube

EBSD

electron backscatter diffraction

EBSP

electron backscatter diffraction pattern

εeng,pl

plastic engineering strain

[mm/mm]

εeng,tot

total engineering strain

[mm/mm]

εtr,pl

plastic true strain

[mm/mm]

F

distance from outer surface of lid to inner surface of lid

[mm]

[mm]

FESEM Field Emission SEM h

height of an imperfection

[mm]

h1

height of a spike, h1 = S – s

[mm]

4 hmax

maximum height of welding imperfections in certain welding distance

[mm]

HAZ

heat affected zone

IB

beam current

[mA]

Ifoc

focus current i.e. lens current

[mA]

IPF

Inverse Pole Figure

L

width of slot

[mm]

L osc

longitudinal oscillation measure

[mm]

MAD

mean angular deviation

s

penetration depth without spiking

[mm]

sintact

clean penetration i.e. minimum intact fusion zone thickness

[mm]

smin

penetration minimum in certain welding distance

[mm]

smax

penetration maximum in certain welding distance

[mm]

Δs

variation in penetration depth in certain welding distance

[mm]

S

penetration depth to the tip of a spike

[mm]

Smax

maximum penetration depth (to the tip of a spike) across the welding distance

[mm]

SE Fit

standard error of fits

SEM

scanning electron microscope



standard deviation



stress

[MPa]

T osc

transversal oscillation measure

[mm]

UB

accelerating voltage

[kV]

v

welding speed

[mm/s]

x

axis in the direction of welding

y

axis in the direction of the side of welding

z

axis in the direction of the weld penetration

w

width of weld or width of an imperfection

[mm]

5

1. INTRODUCTION This report contains the methods used at Tampere University of Technology, Department of Materials Science (TUT DMS), in electron microscopy studies of electron beam welded copper samples welded by Posiva at Patria Aviation Facilities. The aim of the work was to evaluate the residual stresses in the electron beam welded canister sample using electron backscatter diffraction (EBSD). A reference curve with tensile tests of known strains was earlier constructed [1].

6

7

2. WELDING PARAMETERS The EB welding of the full-scale lid to a 890 mm tall canister was carried out by Posiva at Patria Aviation. In the test XK049 the welding was a circumferential joint of a spent nuclear fuel canister containing slope up, full penetration weld, overlap and slope down. Test XK049 was welded before the repair work and calibration of the EB welding equipment. Lid weld test XK049 was welded for an extensive residual stress testing program and metallographic and EBSD studies were conducted after those tests. The section examined at TUT DMS was machined from the area of the full penetration weld, in the weld diameter area of θw1 = 323°. The accelerating voltage UB was 150 kV. The welding of lid weld tests was carried out in a medium vacuum (9.7×10−3 mbar) using a vertical electron beam. The welding parameters, based on the results of earlier full penetration tests, were: 

Welding beam current IB 320 mA



Welding speed v 2.5 mm/s



Lens current IL 2390 mA



Oscillation pattern E32 using the relative set values of ACX 4.2 and ACY 3.8

The welded material was phosphorus micro-alloyed oxygen-free copper Cu−OFP with P= 30−70 ppm, O < 5 ppm, H < 0.6 ppm and S < 8 ppm.

8

9

3. RESEARCH METHODS 3.1.

Identification of sample

The lid welding experiments were marked with running numbers starting from one, and the number of an experiment was indicated with a three-number code, e.g. XK049. For the metallographic studies, the welding samples were cut into smaller pieces which were marked with identification numbers. A sample with parallel surfaces was needed for the EBSD studies. The first (front) surface of the cross-section was cut at θw = 323°. The second (back) surface was cut 10 mm after that, which means that this surface was not cut exactly in radial direction. The sample was marked with the angular location of the first surface: XK049-323. The sides of the cross-sectional samples were numbered as “1” and “2”. Side 1 is the front side and side 2 is the back side of the sample when the samples are ordered by welding direction. In the naming of the images, the co-ordinate axis system shown in Figure 1 was used. The zero points of the axes are located at the starting point of the welding. The ‘x’ direction indicates the welding direction. The ‘y’ direction indicates perpendicular welding, i.e. towards the outer side of the tube. The ‘z’ direction points to the depth of welding.

Figure 1. The co-ordinate axis system used.

10 3.2.

Sample preparation

The transverse cross-section samples were cut by wire electric discharge machining WEDM from the ring containing the weld. The cross-section was ground using P160 and P320 grinding papers to remove the excess oxide layer and to obtain sufficient smoothness of the surface. The sample was etched with technical grade nitric acid (HNO3). The sample was first “rinsed” with nitric acid by immersing it for 5−10 s, washed with water, and then actually etched by immersing for 30 s in fresh nitric acid. After etching the sample was washed with water and ethanol and dried with a drier. After the macroscopic studies the cross-section sample was cut at the depths z=17 mm, z=35 and z=52 mm, as shown in Figure 2. These ca. 17 mm high samples were then prepared using several grinding and polishing steps. EBSD analysis requires specific conditions for the preparation of the surface. To obtain a good quality EBSD pattern, residual deformation or stress due to mechanical polishing must be removed [5]. As a consequence, the sample preparation presents many stages. First the samples were mechanically wet ground with silicon carbide papers from P320 to P1200. Polishing was carried out with 9 μm and 3 μm diamond suspensions. Final polishing was performed with colloidal silica particles. Between the stages the samples were carefully washed with water and cleaning agent. Finally the samples were cleaned in an ultrasonic bath for 10 min in distilled water with cleaning agent and for 10+10 min in ethanol. The samples were stored in desiccator. Just before EBSD studies the samples were cleaned in a plasma cleaner.

Weld z=5 mm

Tube z=5 mm

Lid z=25 mm Weld z=25 mm

Figure 2. The locations of the EBSD measurement samples.

11 3.3.

EBSD studies

Electron backscatter diffraction (EBSD) measurements were performed using Zeiss ULTRAplus UHR FESEM microscope with the HKL Premium-F Channel EBSD ultrafast Nordlys F400 detector provided by Oxford Instruments, see Figure 3. EBSD was first discovered nearly a hundred years ago, but a true utilization has been made in the last two decades mainly because of the development of fully automated EBSD systems [3]. EBSD technique provides crystallographic information (orientation maps, orientation relationships between two phases, grain boundary disorientations, etc.). Briefly, the principle of this technique is to exploit information given by backscattered electrons from the specimen. Diffracted electrons are collected on a phosphorus screen where they form Kikuchi pseudo-bands. From the Kikuchi patterns, the corresponding orientations of the crystal lattice are calculated automatically using the HKL Channel 5 software. [4]

Figure 3. Zeiss ULTRAplus UHR FESEM microscope with the HKL Premium-F Channel EBSD ultrafast Nordlys F400 detector.

12 EBSD was scanned in the center of the weld in 2 depths: the center of the EBSD maps was in the depths z=5 mm and z=25 mm (measured from the post-welding machined surface). The base material in the tube was measured from the sample A in the depth z=5 mm and in the lid from the sample B in the depth z=25 mm (location of the samples A and B is shown in Figure 2). The scanned area of a map was 4.4 mm wide and 3 mm high with the step of 5 µm, which give 533886 measurement points for each map. During EBSP analysis the detection was made using band centers with 6 to 7 bands. The EBSP imaging was set to 28 ms with 4x4 binning and gain 4. In the HKL software the minimum angle to define a grain boundary was 1°. The EBSD data were noise reduced using a ‘‘wildspike’’ correction and a five-neighbour zero solution extrapolation. For the deformation maps the HKL Tango software measures the internal average misorientation angle within each grain. If the average angle in a grain exceeds the userdefined minimum angle to define a subgrain, (θc), the grain is classified as being "deformed". Some grains consist of subgrains whose internal misorientation is under θc but the misorientation from subgrain to subgrain is above θc. In that case the grain is classed as "substructured". All the remaining grains are classified as “recrystallized”. [6] In the deformation maps the minimum misorientation (°) to separate subgrains was set to θc=1° and to separate grains to 7.5°. The grains with minimum deformation (recrystallized) are marked by blue, the grains containing low angle boundaries (substructured) by yellow and the deformed grains by red.

13

4. MAIN RESULTS The various EBSD maps of the weld areas in the depths z=5 mm and z=25 mm are shown in Appendix A; and the maps of the lid and tube base material areas are shown in Appendix B. The maps presented in the Appendices are: 

EBSD Band contrast maps with twin ( 60° Ʃ3) boundaries (red), small angle (2°-10°) boundaries (blue) and high angle (>10°) boundaries (green)



Inverse Pole Figure (IPF) Maps with direction X0 with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black)



All Euler Maps with twin ( 60° Ʃ3) boundaries (yellow), small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black)



Recrystallized fraction maps showing deformed (red), substructured (yellow) and recrystallized (blue) fractions and with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black)



Local Misorientation Maps with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black)



The maps showing the average of the mean intra-grain misorientation AMIS

The results of the EBSD data are summarized in Table 1 and in Table 2. The recrystallized, substructured and deformed fractions in the studied weld and base material areas are also shown in Figure 4. In Table 1 the EBSD indexing is given before noise reduction (NR), since this shows the actual level of the indexing; NR will reduce the zero solution points. In all the cases the indexing was very good. The mean angular deviation (MAD) is given in degrees specifying the averaged angular misfit between detected and simulated Kikuchi bands [5]. In all cases the MAD was low indicating a high degree of fit between theoretical and measured EBSPs. The amount of twin boundaries is very low (10°) boundaries (green).

APPENDIX A

28

EBSD maps of weld areas

Figure 9. z=5 mm. IPF Maps with direction X0 with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

Figure 10. z=25 mm. IPF Maps with direction X0 with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).


29

APPENDIX A

Figure 11. z=5 mm. All Euler Maps with twin ( 60° Ʃ3) boundaries (yellow), small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

Figure 12. z=25 mm. All Euler Maps with twin ( 60° Ʃ3) boundaries (yellow), small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

APPENDIX A

30


Figure 13. z=5 mm. Deformed (red), substructured (yellow) and recrystallized (blue) fractions, small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

Figure 14. z=25 mm. Deformed (red), substructured (yellow) and recrystallized (blue) fractions, small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).


31

APPENDIX A

Figure 15. z=5 mm. Local Misorientation Maps with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

Figure 16. z=25 mm. Local Misorientation Maps with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

APPENDIX A

32


Figure 17. z=5 mm. Average of the mean intra-grain misorientation AMIS maps. The color scale is 0°-15°.

Figure 18. z=25 mm. Average of the mean intra-grain misorientation AMIS maps. The color scale is 0°-15°.

EBSD maps of lid and tube

33

APPENDIX B

APPENDIX B EBSD maps for the lid and tube base materials: 

EBSD Band contrast maps with twin ( 60° Ʃ3) boundaries (red), small angle (2°-10°) boundaries (blue) and high angle (>10°) boundaries (green)



Inverse Pole Figure (IPF) Maps with direction X0 with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black)



All Euler Maps with twin ( 60° Ʃ3) boundaries (yellow), small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black)



Recrystallized fraction maps showing deformed (red), substructured (yellow) and recrystallized (blue) fractionsand with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black)



Local Misorientation Maps with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black)



Average of the mean intra-grain misorientation AMIS maps with the scale 0° to 15° and scaled based on the maximum of the area.

For the report: Metallographic studies of electron beam welded copper lids: EBSD studies of the cross-section of XK049 at 323deg

APPENDIX B

34


Figure 19. Lid z=25 mm. Band contrast maps with twin ( 60° Ʃ3) boundaries (red), small angle (2°-10°) boundaries (blue) and high angle (>10°) boundaries (green)

Figure 20. Tube z=5 mm. Band contrast maps with twin ( 60° Ʃ3) boundaries (red), small angle (2°-10°) boundaries (blue) and high angle (>10°) boundaries (green)


35

APPENDIX B

Figure 21. Lid z=25 mm. IPF Maps with direction X0 with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

Figure 22. Tube z=5 mm. IPF Maps with direction X0 with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

APPENDIX B

36


Figure 23. Lid z=25 mm. All Euler Maps with twin ( 60° Ʃ3) boundaries (yellow), small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

Figure 24. Tube z=5 mm. All Euler Maps with twin ( 60° Ʃ3) boundaries (yellow), small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black)


37

APPENDIX B

Figure 25. Lid z=25 mm. Deformed (red), substructured (yellow) and recrystallized (blue) fractions, small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

Figure 26. Tube z=5 mm. Deformed (red), substructured (yellow) and recrystallized (blue) fractions, small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

APPENDIX B

38


Figure 27. Lid z=25 mm. Local Misorientation Maps with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).

Figure 28. Tube z=5 mm. Local Misorientation Maps with small angle (2°-10°) boundaries (gray) and high angle (>10°) boundaries (black).


39

APPENDIX B

Figure 29. Lid z=25 mm. Average of the mean intra-grain misorientation AMIS maps. The color scale is 0°-15° (white = over 15°).

Figure 30. Tube z=5 mm. Average of the mean intra-grain misorientation AMIS maps. The color scale is 0°-15° (white = over 15°).

APPENDIX B

40


Figure 31. Lid z=25 mm. Average of the mean intra-grain misorientation AMIS maps, scaled based on the maximum of the area.

Figure 32. Tube z=5 mm. Average of the mean intra-grain misorientation AMIS maps, scaled based on the maximum of the area.