19 th International Symposium on Zirconium in the Nuclear Industry - - PowerPoint PPT Presentation

Characterization of Hydrides and the α-Zr Matrix in Zirconium Alloys Effects of Stresses, Microstructure and Neutron Irradiation on Hydride Texture, Terminal Solid Solubility and Dislocation Structure

• P. Vizcaino, A. V. Flores, M. A. Vicente, J. R. Santisteban, G. Domizzi, A. Tolley, A. Condó,
• J. Almer

19th International Symposium on Zirconium in the Nuclear Industry

Monday May 20 to Thursday May 23 2019 The Midland; Manchester United Kingdom

OUTLINE

• Design of the experiments: APS, LNLS, TEM, etc.
• Materials under study: characterization
• Data reduction and analysis

Results

• Hydride texture/matrix texture

✔Effect of an applied stress ✔Link between microscopic and mesoscopic scales (Perovic´s model)

• Effects of stress on the solubility limit
• Hydride dislocation densities by synchrotron X-ray diffraction
• Hydride dislocation densities by TEM and HRTEM observations

General Conclusions Experimental

Materials under study: APS experiments

Zr-2.5Nb CANDU type microstructure, axial surface Hydride blister grown in the laboratory: X-yay beam scan scheme

Axial Radial Hoop

Sample for X-ray diffraction

3 mm 4 mm

Blister

Zr-2.5Nb CANDU type microstructure axial surface

Materials under study: APS experiment

Four dog-bone samples hydrided: 44, 55, 67 and 130 H-wtppm Hydride morphology before the in-situ experiment

Hoop

L2: 130 wt-H ppm E1: 55 wt- H ppm

Neutron irradiated (~1022 neutrons/cm2) and unirradiated Zircaloy-4 Fully recrystallized microstructure

Materials under study: other experiments

iii) Unirradiated fully recrystallized and hydrided thin foil was also prepared for TEM and HRTEM studies, ~ 3000 H-wtppm) i) Planar sheet: 5 x 4 mm for XRD in the Bragg-Brentano geometry at LNLS, Brazil ii) Thin foil: φ~3 mm for TEM and HRTEM

Hoop Axial

Deuterided irradiated Zircaloy-4, ~ 400 Heq-ppm)

10μm

SEM image

• f a

Zircaloy-4 highly hydrided sample TEM, Zircaloy-4 hydrided. Red arrows indicate hydrides.

APS 1-1D x-ray beam station. Experimental setup for the cylindrical specimens containing the hydride blister Experimental setup

• Beam energy: 80 Kev, monochromatic beam
• Beam size: 300 x 300 μm
• CCD image obtained in 0,1 sec (2048 x 2048 pixels
• f 200 x 200 μm2)
• Cylindrical specimen of hydrided (blister) Zr-2.5Nb

with principal axes in the three directions, radial, hoop and axial The X-ray beam goes through the cylindrical sample and an area detector collects the diffracted rings. The cylindrical sample is rotated along the radial direction (ψ) in angle y from -90 to 90º in 5º steps

 

Pole figures

• Running around a diffraction ring (angle χ)

corresponds to scan along two parallel lines that virtually cross the center of the pole figure (radial direction).

• Rotation of the specimen around the radial

direction (ψ) reflects as a rotation of the lines around the center of the pole figure.

Sample holder Radial Axial-Hoop plane Sample

LNLS (Brazilian Synchrotron) XRD#1 beam station

Experimental setup for the Bragg-Brentano experiments

• Geometry: θ-2θ, 0.05o step, 20º≤2θ≤130º
• Photon beam of 8.08 keV
• λ = 1.542484Å (Δλ =0.000001Å)

(b)

Hydride peaks Close to a blister (α-Zr and δ-hydride are ~ 50%): ✔δ-(200) peak is considerably wider than the α-(10-11), but ,

TEM Bariloche Atomic Center, Arg.

200 nm

(a) (c) (c)

1 0 1 0 1 1 2 0 1 1 0 2 0 2

(b)

2 2 2 2

TEM images and SAD patterns from hydrides and Zircaloy-4 matrix. (a) Hydride with orientation relationship [0002]α // [111]δ and (1 1-2 0)α// (2- 2 0)δ (b) SAD pattern of a hydride with Z= 111. (c) SAD of the matrix with Z= 0002.

Unirradiated Zircaloy-4: two

• rientation

relationships for the hydrides with the α-Zr matrix: [0002]α // [111]δ and (1 1-2 0)α// (-2 2 0)δ ✔instrumental resolution is identical due to the close proximity between the α and δ peaks

APS experiments: data reduction

Four main phases are identified, the α, β and ω phases of Zr and the δ−hydride, with varying intensities for different azimuthal angles, due to the crystallographic texture of all phases. Images were transformed into 72 “traditional” 2θ diffractograms by slicing the rings azimuthally with an angular section of 5o The position, integrated area and FWHM of all peaks observed in each diffractrogram were defined by least-squares fits, in order to define texture, elastic strains and dislocation densities of the δ-hydride, as described below. Specimen L2 (130 wt H-ppm), unloaded condition

within the blister, ~90%

• f δ-hydride phase

blister interface ~50% of δ-hydride phase

• utside the blister (in the

parent material), with a concentration of ~200 wt ppm H At first glance, the three pole figures display a similar pattern regardless of the position and the very different hydride concentrations and morphologies A closer look shows: ✔a nearly perfect orthorhombic symmetry, ✔those measured where hydrides appear as a minority phase display certain degree of asymmetry (~15o to de radial direction)

Texture of the α-Zr matrix plays the major role in the structure of the hydride ODF RESULTS Hydride texture/matrix texture: crystallographic texture of delta hydride

Hydride ODF: the orientation relationship [0002]α//[111]δ, (11-2 0)α//(20-2)δ has been applied to all the crystal

• rientations

composing the α-Zr ODF, weighted by their intensities We have determined many intensity pole figures for α-Zr phase

Identical precipitation probability in all α−Zr grains regardless of its orientation (perhaps explain the

• bserved asymmetries)

The agreement between experimental and calculated pole figures is extremely good, confirming that the hydride-parent crystal relation [0002]α//[111]δ and (1 1-2 0)α//(2 0- 2)δ is by far the most commonly observed They were used to produce the

• rientation distribution function (ODF) of

the α-Zr phase

RESULTS Hydride texture/matrix texture: crystallographic texture of delta hydride

Basic texture components of a Zr-2,5Nb pressure tube

Asymmetries can be explained in terms of a texture simplified model assuming four texture components:

Colored points indicate the positions of the components in the ODF of the microstructure

RESULTS Crystallographic texture of delta hydride

Regions of the α-(0002) and δ-(111) pole figures where the ideal orientations of the synthetic ODFs manifest on a recalculated figure pole of a blister

Blister: 90% hydride Far from the blister, low H content Interesting to note the rotation of ~15o from the hoop direction a change in the texture component precipitation probability Explanation in terms of the Perovic’s model [*]: For circumferential hydrides far from the blister the elastic strain field that appears in the matrix after hydride precipitation makes the arrangement of htilted hydrides (b) energetically more favorable over arrangements hhoop (a) or hRadial (c) Perovic´s representation of circumferential and radial hydrides in a Zr-2.5Nb pressure tube precipitated in α-Zr grains of either mhoop, mtilted or mradial orientations

[*]. V. Perovic • G.C. Weatherly • C.J. Simpson. Hydride precipitation in α/β zirconium alloys. Acta Metall. Vol. 31. No. 9, (1983), pp 1381-1391.

This confirms that, although the texture of the α-Zr does plays a role in precipitation of circumferential

• r radial hydrides, its connection with the actual

morphology

• f

hydride clusters is not straightforward: it involves a complex interplay with the morphology of α-grains, the type of grain boundaries and other factors such as external and internal stresses and the detailed spatial distribution

• f defects

RESULTS Crystallographic texture of delta hydride

Radial Hoop

implies

• Beam energy: idem
• Beam size: idem
• CCD image idem
• Dog-bone specimens of Zr-2.5Nb with principal

axes in the three directions, radial, hoop and axial

• Samples were placed in a MTS tensile test

machine

APS 1-1D x-ray beam station. Experimental setup for dog-bone experiment

The cycles (three) from room temperature up to 400oC were performed without stress (A), at 130 Mpa (B) and at 225 Mpa (c) keeping them 5´ at 400oC.

Unstressed (L2), circumferential hydrides Cooled stressed at 225 Mpa (L2), circumferential and radial hydrides

• NO SAMPLE ROTATION POSSIBLE

44 H- wtppm 55 H- wtppm 67 H- wtppm 130 H- wtppm

Hoop

Four dog-bone samples hydride samples

RESULTS Effect of external stress on the texture of the δ-hydride

During the experiments the dog-bone specimens remained at a fixed orientation: the white dotted circle contains the observed orientations Before After Reorientation implies: ✔Favoring hydride precipitation in hHoop or htilted instead of other orientations ✔Changes in the intensity of the d-hydride spots in the pole figures but not in the position ✔Turning energetically favorable some Perovic´s cluster arrangements of radial hydrides

Journal of Nuclear Materials 304 (2002) 96–106

Hydrogen solubility limit determinations in Zirconium alloys

DSC technique, very used for TSS determinations ✔It measures volumetric changes in the temperature of the sample (thermal flux apparatus) during the process ✔It is a dynamic technique It does not distinguish microstructural anisotropies like: ✔Preferential orientation of the grains ✔State of stress of the whole material, and much less a particular family of grains Thermal cycle in the dog-bone APS experiment

Heating-cooling rate: 20oC/min

✔(a) φ=0° correspond to hydrides precipitating in α-Zr crystallites of the mHoop orientations ✔(b) φ=70° correspond to hydrides precipitating in α-Zr crystallites

• f

both mHoop+mTilted

• rientations which represent the bulk of the

hydride texture. azimuthal angles ✔Integrating around the complete Debye ring (c), for the same CCD image the presence of hydrides becomes barely visible and it would easily lead to wrong conclusions. Details of the diffractograms recorded during cooling for the L2 specimen at the temperature of 270°C, with and without applied load.

RESULTS Effects of stress on the solubility limit

TSSP determinations: area of the δ-(111) peak and the position of α-(0002) peak were followed during cooling: ✔We observe an increase of ~20°C in TSSP by application of the 225MPa load and the hysteresis is reduced in 30oC ✔It is clear from (a) that the concentration of hydrides precipitated in mHoop grains increases from ~50 wt ppm H in the unloaded condition to ~120 wt ppm H under load ✔From the δ-(111) peak area, we can define the volume of precipitated hydrides, ✔From the α-(0002) peak position we determined the concentration of H in solution

RESULTS Effects of stress on the solubility limit

Before precipitation: H in solution in mHoop grains is ~120 wt ppm H in both the loaded or unloaded

• conditions. This means:

✔H redistribution occurs during precipitation due to the diffusion of H from grains with lower precipitation temperatures into grains

• f higher precipitation temperatures

✔H has migrated from mHoop grains into neighboring grains

• f

any

• f

the α-Zr

• rientations

RESULTS TSS as a function of the applied stress for two texture components: mHoop and mTilted Important result: the average rate o f change of TSSP is 0.08 oC/Mpa, two orders

• f magnitude higher than the value estimated by Shi, of 0.0007 oC/Mpa

Material: Zircaloy-4 fully recrystallized and with quasi-radial texture, a highly hydrided sample (3557 ppm) and a non hydrided sample

Hydride dislocation densities by TEM

hydrides dislocations hydrides Second phase particles

✔The dislocation density in the non hydrided material (pure α-Zr matrix) lies between 5.0 x 1012 m-2 and 2.0 x 1013 m-2 ✔Dislocation densities from 1.7 and 2.3 x 1014 m-2 in the α-Zr matrix surrounding hydrides

Findings:

✓ The density calculated for the hydrides lies between 1.6 x 1014 and 1015 m-2

RESULTS

Studies on hydrides were based on the analysis of the position and intensity of the observed hydride diffraction peaks, it became clear that some δ-hydride peaks were considerable wider than α-Zr peaks This experimental data were obtained for the blister, in particular δ(111) and δ(222) peaks are much narrower than δ(200) and its higher order δ(400), δ(220) and δ(311) peaks, with ΔK similar to those observed for αZr.

This leads to the inquiry of the structural nature of the hydride

The width difference between δ and α phases is marked for some hydride planes and is only subtle for others

Hydride dislocation densities by synchrotron X-ray diffraction

✔Physical broadening results from crystallite size, defect type and density. The line FWHM (or peak breadths) are obtained for both, the real sample and the “instrumental” diffractogram The experimental broadening: convolution between the instrumental broadening of the diffractometer and the physical broadening introduced by the sample ✔The instrumental broadening results from the wavelength and angular spread of the incident beam, the finite sizes of beam, sample and detectors Two facts are clear: ✔The dislocation density of the hydride is 2

• rder of magnitude higher than that
• bserved in α phase, with values as higher

as 3 x 1016 m-2 ✔There is a systematic increase of the dislocation density of both phases as the H content increases.

Hydride dislocation densities by synchrotron X-ray diffraction RESULTS

For low H content the dislocation density in the hydride starts at a value around 2 x 1015- 5 x 1015 m-2, indicating that even for low H content a high dislocation density is observed We understand the observed dependence of the dislocation density with the H content as a footprint on how the precipitation elastic strain converts to plastic deformation and redistribute between it δ and α matrices with different hardness

Hydride dislocation densities by synchrotron X-ray diffraction RESULTS

CONCLUSIONS

• 1. For both circumferential and radial hydrides precipitation occurs in α-grains from a variety of
• rientations, but with a clear preference for α-Zr crystals with their c-axes at an angle of 20o from the

hoop direction (mTilted grains).

• 2. A smaller volume fraction precipitates on crystallites having their c-axis along the tube hoop direction

(mHoop grains).

• 3. Application of an external stress changes precipitation probabilities depending on the angle between

the applied load and the c-axes of the α-Zr grains. Hydride precipitation increases on α-grains with their c-axis stretched by the external stress and diminishes on grains where the c-axis is compressed. This manifests as a pronounced change in the orientation of hydride clusters (from circumferential to radial) but only in a moderate change in the crystallographic texture of the hydride.

• 4. Application of a tensile load along the hoop direction during thermal cycling changes precipitation

and dissolution temperatures, reducing the hysteresis in both grain families.

• 5. Precipitation temperatures increase most for mHoop grains at a rate of (0.08±0.02)oC/MPa, in

disagreement with previous theoretical estimations presented in the literature. At a load of 225 MPa the precipitation temperature in mHoop grains is larger or equal than in mTilted grains.

• 6. Application of a load while H is in solution does not result in significant redistribution between

different grain orientations.

• 7. TEM observations in agreement with X-ray results show that dislocations densities in the δ-hydrides

are large (1.6 1014 to 1016 m-2), much larger than the matrix values (~1013 m-2) and increase as the strength of the α-Zr matrix increase (either due to higher dislocation densities or to a higher density of hydrides) The conclusions of the present research program can be summarized in the following points:

Thanks for your attention!

Diffraction vector Hoop // to the Radial

a

Hoop Axial

a a q q Hoop Diffraction vector Basal planes sin 𝜄 =

𝜇 2𝑒 = 0.15 Å 2 𝑦 2.57Å = 0,0292~ 𝜄 = 1.67𝑝 a is a small angle too

The error associated to the machining of the dog bone specimens should be

• f this order