P. Hofmann and M. Markiewicz

7/28/2019 P. Hofmann and M. Markiewicz

1/53

KfK 4729Juni 1994

Chemical Interactions betweenas-received and pre-oxidizedZircaloy-4 and Inconel-718 atHigh TemperaturesP. Hofmann, M. Markiewicz

Institut fr MaterialforschungProjekt Nukleare Sicherheitsforschung

Kernforschungszentrum Karlsruhe


2/53


3/53

Kernforschungszentrum Karlsruhe. Institut fr MaterialforschungProjekt Nukleare Sicherheitsforschung

KfK4729

Chemical interactions between as-received andpre-oxidized Zircaloy-4 and

Inconel-718 at high temperatures

P. Hofmann, M. Markiewicz*

*Comisi6n Nacional de Energia At6micaAv. dei Libertador8250, 1429Buenos AiresArgentina

Kernforschungszentrum Karlsruhe GmbH, Karlsruhe


4/53

Als Manuskript gedrucktFr diesen Bericht behalten wir uns alle Rechte vorKernforschungszentrum Karlsruhe GmbHPostfach 3640, 76021 Karlsruhe

ISSN 0303-4003


5/53

AbstractIsothermal reaction experiments were performed in the temperature range of1000 - 1300 oe in order to determine the chemical interactions between Zircaloy-4fuel rod c1adding and Inconel-718 spacer grids of Pressurized Water Reactors(PWR) under severe accident conditions. It was not possible to apply even highertemperatures since fast and complete Iiquefaction of the components occurred asa result of eutectic interactions during heatup. The liquid reaction productsformed enhance and accelerate the degradation of the material couples and thefuel elements, respectively. Only small amounts of Inconel are necessary to Iiquefy large amounts of Zircaloy.Thin oxide layers on the Zircaloy surface delay the beginning of the chemical interactions with Inconel but cannot prevent them. In this work the reaction kinetics have been determined for the system: as-received and pre-oxidized Zircaloy4/lnconel 718. The interactions can be described by parabolic rate laws; theArrhenius equations for the various interactions are given.Chemische Wechselwirkungen von Zircaloy-4 im Anlieferungs- und voroxidierten Zustand mit Inconel718 bei hohen TemperaturenZusammenfassungZur Bestimmung der chemischen Wechselwirkungen zwischen Zircaloy-4Brennstabhllrohren und Inconel 718-Abstandshaltern von Druckwasserreaktoren unter schweren Unfallbedingungen wurden isotherme Reaktionsexperimente im Temperaturbereich 1000 - 13000e durchgefhrt. Hhere Temperaturenkonnten nicht angewandt werden, da eine schnelle und vollstndige Verflssigung der Materialien infolge eutektischer Wechselwirkungen schon in der Aufheizphase stattfand. Die entstehenden flssigen Phasen beschleunigen die weitere Zerstrung der Materialproben bzw. der Brennelemente. Nur geringe Mengenan Inconel sind notwendig, um groe Mengen an Zircaloy zu verflssigen.Dnne Oxidschichten auf der Zircaloy- Oberflche verzgern den Beginn der che-mischen Wechselwirkungen mit Inconel, sie knnen jedoch nicht verhindert werden. Im Rahmen dieser Arbeit wurde die Reaktionskinetik fr die Systeme nichtoxidiertes Zircaloy/lnconel und voroxidiertes Zircaloy/lnconel bestimmt. Die verschiedenen chemischenWechselwirkungen knnen durch parabolische Zeitgesetze beschrieben werden; die Arrheniusgleichungen werden angegeben.


6/53

Contents

1. Introduction

2. Experimental Details

3. Results3.1 Chemical behavior3.2 Reaction kinetics3.3lnfluence of an oxide layer on Zircaloy/lnconel interaction

4. Discussion

5. Summary and Conclusions

1

1

2256

8

9

6. Acknowledgement

7. References

.................................................. 1010

List of Tables

List of Figures

.........................................................

.........................................................11

12


7/53

- 1 -

1.lntroductionIn severe reactor accidents the temperature within the core exceeds 1200C,which is the maximum tolerable temperature for design-basis accidents. At thesehigh temperatures the various materials of the fuel-rod bundles interact chemically with each other and with the steam of the environment. With respect tocore damage initiation and progression, eutectic interactions are of special interest since these interactions result in the formation of liquid phases at low temperatures. Experiments with fuel-rod bundles, which were heated up to 2000C,c1early showed, by optical observations and temperature measurements, the significant impact of Inconel spacer gridson the early melt formation and damagepropagation within the fuel assembly at rather low temperatures [1, 2]. Inconelspacer grids fail due to Iiquefaction below their melting point (around 1450C), asa result of eutectic reactions, of nickel and other Inconel alloy components withzirconium. The resulting liquid reaction products relocate and enhance and accelerate the degradation of the fuel rod bundle.Given this body of evidences, a very important aspect is to study the reaction kinetics and the onset of liquid phase formation between Zircaloy c1adding (Zry)and Inconel718 caused by the chemical interactions.

The Zry c1adding was used in the as-received condition and in the pre-oxidized(10,20,45, pm oxide layer thicknesses) condition. An oxide layer on the surface ofZry can initially prevent the interaction with Inconel, but the oxide is rapidly dissolved in the Zry at higher temperatures, during an incubation period, ta. The result of the separate-effects tests are the basis for the description of the materialinteractions by modelswhich can be used in computer code systems.2. Experimental DetailsThe isothermal annealing experiments were performed with Zircaloy-4 (Zry) capsules into which short cylindrical rods of Inconel718 had been pressed and c10sedgas-tight by a conical Zry plug. Figure 1shows the components of a compatibilityspecimen before being loaded.The annealing experiments were performed in a tube furnace under flowing


8/53

- 2 -

argon. The specimens were heated up to about 900C and then placed into thepre-heated furnace whieh had the desired annealing temperature. The annealingtime started when the temperature of the specimen was 20 K below the annealing temperature. The specimens were cooled down to room temperature outsidethe furnace.The temperatures investigated ranged from 1000 to 1300C and the maximum annealing time was 300 minutes. The highest temperature whieh could be examinedwas determined by the condition of the reaction couple,with limits imposed bythe failure of the specimen due to the onset of liquid phase formation followedby a fast and complete Iiquefaction of the Zry crucible at slightly higher temperatures.After annealing, the specimens were mechanieally cut and thenmetallographieally prepared for examinations of the reaction zones with an optieal mieroscope. Some of the specimens were etched to be able to recognize better the various phases. The thicknesses of the reaction zones were measured atfour locations of the interface. The maximum thiekness of the reaction zones wasused to evaluate the kinetie data. In addition, some Scanning Electron Mieroscopy(SEM)/Energy Dispersive X-Ray (EDX) or SEM/Wave Length Dispersive X-Ray(WDX) examinations were performed to obtain information on the chemical compositions of the reaction products and diffusion zones.The Zry crucibles from the pre-oxidized Zircaloy/lnconel interaction experiments,were oxidized in an Ar/02 mixture in a tube furnace. The selected temperaturewas 1000C and the times needed to produce the oxide layer thieknesses of 20and 45 11m were 6 and 25 minutes, respectively. Metallographie examinationswere done to check the actual Zr021ayer thieknesses.3. ResultsThe chemieal interaction between Zircaloy and Inconel is a diffusion controlledprocess and can therefore be described quantitatively by Arrhenius equations.Besides the identifieation of the reacting species, phases, and the temperature ofonset of liquid phase formation, one of the main results was the determination ofthe reaction kineties, whieh may be used in a code system to prediet the materialbehavior at high temperatures.3.1 Chemical behaviorThe chemieal interaction between Inconel 718 (chemieal composition of about 53


9/53

- 3 -

wt. % Ni, 19 % Cr, 18 % Fe, 6 % Nb+Ti, 3 % Mo) and Zircaloy-4 shows that a l iquid reaction zone develops at temperatures as low as 1000C. The melting pointsof Inconel718 and Zircaloy-4 are 1335C and 1760C, respeetively [3, 4].Although many elements are involved in the Zry/lnconel chemical interaction, thebinary phase diagrams offer a reasonable explanation of low temperature liquidphase formation. The (Ni - Zr), (Fe - Zr) and (Cr - Zr) phase diagrams indicate thateutectic interactions occur in these binary systems (Figures 2,3) [5, 6]. The (Ni - Zr)system has four eutectic temperatures, with the lowest eutectie point at about960C on the Zr-rieh side. The diffusion of approximately 2 wt. % Ni into -Zr results in the formation of a liquid phase; even at this low temperature (Fig. 2) theeutectie temperature on the Ni side is about 1170C. The lowest eutectie temperature exists in the (Fe - Zr) system at about 930C also on the Zr-rieh side, while theeutectic temperatures in the binary (Cr-Zr) system are 1332 and 1592C, Le. muchhigher.The cross-sections of the Zryllnconel reaction specimens after annealing at 1000,1100, 1150 and 1200C for various reaction times are shown in Figure 4. Thecross-sections of both as-received and pre-oxidized Zry/lnconel reaction specimens (initial Zr02 layer thiekness: 10 11m) are presented for comparison. Theeutectie interactions, whieh occur at 1000C and higher, result in the formation ofliquid reaction products and voids form as some of the molten material relocates.Although the specimens were annealed inan upright position, the reaction zonesof the as-received Zry are not symmetrieal around the circumference. Apparently,very small forces are sufficient to induce asymmetrie reaction zones in the Zrywall. First interactions have been noticed at 1000C afterannealing times longerthan 5 minutes. In all cases, the reaction in the Zry was much stronger than inInconel. This may be explained by phase diagram considerations (Fig. 2, 3). Onlysmall amounts (> 5wt. %) of Ni, Fe or Cr are necessary to Iiquefy large quantitiesof Zr. Also, since the eutectic temperatures in the (Ni - Zr), (Fe - Zr), and (Cr - Zr)systems are lowest on the Zr-rieh side, Zr can be Iiquefied at much lower temperatures than Ni, Fe, or Cr. Liquefaction starts as soon as the Iiquidus line of the twophase field has been reached.The cross sections of Zry-4/lnconel 718 specimens and typieal microstruetures produced bythe chemieal interactions are shown in Figures 5 through 10.The solidified reaction zones were examined with SEM/EDX. Besides the determination of the element distribution, quantitative measurements were


10/53

-4-

performed. A typical result of a specimen annealed at 1150C for 1 minute isshown in Figure 11. As can be recognized, the molten reaction zone consists oftwo layers formed due to the eutectic reaction, the morphology of the reactionzone was observed in the whole range of temperature and time examined. Theelement distribution shows Cr enrichment and Ni depletion in the thinner reaction layer I adjacent to the Inconel. In the reaction layer 11, Zr, Ni and Fe areuniformely distributed, Cr is connected with precipitates in this zone.The integral chemical analysis of the solidified melt of reaction layer I shows thefollowing chemical composition: about 56 wt. % Zr, 23 % Cr, 9 % Fe, 5% Ni, 5%MO,2 % Nb. The integral analysis for the reaction layer 11 reveals: 82 wt. % Zr, 12% Ni, 3 % Fe, 2% Cr, 1% Sn, which is very similar to the chemical composition ofthe eutectic point on the Zr-rich side in the (Ni - Zr) system (Fig. 2). From the binary (Ni - Zr) and (Fe - Zr) phase diagrams and the chemical composition of the reaction zone it is c1ear that the diffusion of Ni and Fe into Zry is initially responsiblefor the low-temperature liquefaction of Zircaloy. The further liquefaction of Zrymay then occur faster by dissolution of the Zr in the liquid reaction product.The metallic (Zr, Ni, Fe, Cr) melts in the two reaction layers decompose into various phases on cooldown. In Table 1 the results are entered of the quantitativeanalysis of the phases in the reaction zone.A relation between the microstructure observed in the reaction zone and the ternary (Zr - Ni - Fe) and (Zr - Cr - Fe) phase diagrams at 1000C [7] is shown in Figure11. The integral chemical composition of the reaction zone (dark areas in thephase diagrams) Iie in the l iquid region for the reaction layer 11 and in the Iiquid/solid region for the reaction layer I. The chemical analysis is in agreementwith the microstructure observed in the reaction layers.During the isothermal interaction the liquid phase formed due to the (Ni - Zr) and(Fe - Zr) eutectic interactions penetrates along the grain boundaries of the Zr (Cr,Feh phase, which develops in the reaction layer I (Cr enriched zone) due to thediffusion of Zr into Inconel. This liquidisolid interaction dissolves to a certain extent the Zr (Cr, Feh phase; the amount and size of this phase decrease with increasing distance from the Inconei/liquid interface. In the reaction layer 11,whichis l iquid at the reaction temperature, the Zr (Cr, Feh phase is present only in avery limited amount (Fig. 12). During cooldown the liquid alloy of reaction layersland 11 starts to solidify and decomposes into the -Zr and Zr2 (Ni, Fe)


11/53

- 5-

phases. The Cr2 (Zr, Fe) phase was observed only in the reaction layer I at temperatures above 1150C (Fig. 11).3.2 Reaction kineticsThe behavior of the solid/solid contact interface of Zircaloy/lnconel in the temperature range 1000 - 1200C shows the pronounced damage of Zircaloy, wh ich isindicated by the extended thickness of the liquid reaction zone resulting from thechemical interactions. To be able to determine the reaction kinetics between Zry4 and Inconel718, the thicknesses ofthe reaction zones were measured as a function of the reaction time. The isothermal growth of the reaction zone (dissolutionof Zry and Inconel) is a linear function of the square root of time (parabolic ratelaw) for all temperatures examined, indicating a diffusion-controlled process. Themaximum thicknesses of the reaction zones in the Zry crucibles and Inconel rodsin the as-received conditions are Iisted in Table2. The squares of the reaction zonethicknesses in Zry and Inconel are plotted as a function of the annealing times inFigure 13.Ni and Fe diffuse into solid Zry and initiate Zry liquefaction as soon as the solidlliquid phase field has been reached. To show the differences in the amount ofdissolved Zry and Inconel, the thicknesses of the reaction zones in Zry and Inconelare plotted in Figure 14 as a function of temperature for an annealing time of 5minutes. Small amounts of Inconel dissolve large amounts of Zry. Above about1250C complete liquefaction of the specimens occurs.The reaction zone growth rates for the Zry/lnconel interactions are Iisted in Table

~ and plotted as a function of the reciprocal temperature in Figure 15. The experimental data have been fitted by the following equation.

Ix2/t = A 'exp (- B/RT) Iwherex (cm): reaction layer thickness, t (s): annealing time, A: pre-exponential factor, B(J/mol): activation energy, T (K): annealing temperature, R(J/mol'K): 8.314 gasconstantThe growth rate equations determined for the temperature range of 1000 to1200C are:


12/53

-6 -

for Zircaloy-4x2/t (cm 2/s) = 4.435 104exp (-252093/RT)

for Inconel718x2/ t (cm 2/s) = 2.884 '105'exp (-294962/RT)

These equations can be extrapolated to lower temperatures, but not to highertemperatures since fast and complete liquefaction of the materials takes place atabout 1250C.3.3lnfluence of an oxide layeron Zircaloy/lnconel interactionThe chemical interaction between pre-oxidized Zircaloy-4 and Inconel718 (initialZr02 layer thicknesses: 20 and 45 11m) indicate that the oxide layer delays the beginning of the interaction and reduces the reaction rate, but cannot prevent interaction. The cross section of Zry/lnconel reaction specimens annealed at 1000,1100, 1150 and 1200C for various reaction times are shown in Figure 4. The crosssection of both as-received and pre-oxidized specimens (initial Zr02 layer thickness 10 11m) are presented fo r comparison. Inconel does not interact withstoichiometric Zr02, but the oxide layer wil l be dissolved by Zry, while formingoxygen-stabilized a-Zr(O) during a time-dependent incubation period, ta. The delay in interaction (liquid phase formation) is determined by the time which is necessary for the oxygen to diffuse from the Zr021ayer into Zry forming a "metallic"oxygen-stabilized a-Zry(O) phase which can then interactwith Inconel.The experimental results with the pre-oxidized Zry specimens have shown thatthe Iiquefaction of the specimens has shifted to higher temperatures. The chemical reaction rate was slower when compared with the specimens tested in the as-received condition. Figures 16 and 17 show the cross sections of pre-oxidizedZry/lnconel specimens and a detail of the reaction zone produced by the chemicalinteraction at 1000 and 1200C. The extent of the reaction in the as-received Zrycondition is presented for comparison. It is c1ear, that the protective character ofthe Zr02 layer on the Zry surface, which acts initially as a barrier for theZry/lnconel interactions, depends on the temperature, initial oxide layer


13/53

-7-

thickness and t ime. Figures 18 and 19 show the cross sections and typicalmicrostructures produced by the chemical interaction between pre-oxidized Zryand Inconel as a function of time. Liquid phase formation starts before the oxidelayer has completely disappeared. Figures 17 and 19 show that a thin porous oxide layer remains in the l iquid reaction zone; the same effect was seen in thechemical interaction between Zry-4 and 1.4919ss (AISI 316) [8]. The oxygen fromthe Zr02 layer diffuses into Zry, stabilizing the a-Zry(O) phase, and forms a substoichiometric Zr02-x oxide layer. In this way, Zr f rom the oxide layer can interactwith Ni and Fe, producing a l iquid phase on the Inconel side which penetratesthrough the cracks of the brittle oxide layer reaching the a-Zry (0) phase.The thicknesses of the reaction zones in Zry and Inconel measured for the preoxidized Zry/lnconel reaction specimens are Iisted in Tables 4 and 5 for an initialZr02 thickness of 20 and 45 11m, respectively.The square of the reaction thicknesses versus the annealing times are plotted in Figures 20 and 21. The differences inthe thicknesses of the l iquid reaction zones between Zry and Inconel are smallerat 1200C and above than at lower temperatures; the same has been found in theexperiments performed with specimens in the as-received condition. The reasonis that at the temperature of about 1170C eutectic interaction develops betweenNi and Zr on the Ni-rich side (phase diagram in Fig. 2); only small amounts of Zr 3wt. %) are necessary to produce a liquid phase at this temperature by diffusion processes.The calculated isothermal growth rates are Iisted in Tables 6 and 7 for the initialZr02 thicknesses of 20 and 45 11m, respectively, and plotted versus the reciprocaltemperature in Figures 22 and 23. The growth rate equations and 'to' values havebeen determined fram the x2versus t correlation of the experimental data. In Figures 24 and 25 a comparison is made of the growth rates of Zry and Inconel, respectively, as a function of the reciprocal temperature fo r the different conditions of the Zry/lnconel reaction couple. The results indicate that a Zr02 layer delays the chemical interaction between Zry and Inconel. However, the protectiveeffect of the oxide layer could be inhibited at the contact point of both materialsi f the interaction took place in an oxidizing atmosphere. To c1arify this point, afew experiments were done in steam atmosphere using the configuration schematically represented in Figure 26. The interaction was demonstrated by severalmolten zones on the external surface of the Zry tube in the regions of contact. Across section of a sampie heated at 1150C for 3 min. is


14/53

-8-

shown in Figure 27. Liquid phase formation due to the chemical interactions isnot prevented under conditions of unlimited steam oxidation.4. DiscussionThe main purpose of this work has been to determine the kinetics of reaction between Zircaloy and Inconel with a view of understanding and describing analytically the chemical interactions between the Zircaloy-4 fuel rod c1adding tubes andInconel 718 spacer grids in integral experiments Iike the CORA tests [9]. Althoughsome interaction experiments have been described in the literature [10, 11], noArrhenius equation had been previously been developed from the data. In allcases, the maximum thickness of the reaction zone was used in the evaluation ofthe kinetic data to obtain conservative results.In the Zry/lnconel system, a considerable amount of liquid phase is formedaround 1000C; the amount of liquefied material increases with increasing temperature and time. The nickel-zirconium and iron-zirconium phase diagrams(Figs. 2, 3) show that due to eutectic interactions early melt generation has to beexpected which initiates melt progression within the fuel assembly at low temperatures.The results of the present single-effect tests are in agreement with results of integral tests (CORA experiments) where fuel rod bundles were heated to temperatures of about 2000C [9]. In all cases, damage to the bundle was initiated byZircaloyllnconel and Zircaloy/stainless steel interactions. Localized Iiquefaction ofthese components started at around 1200C. The low-temperature melt formation initiates bundle damage and causes the chemical attack of the fuel rods dissolving the Zry c1adding tubes and part of the U02 fue!. By this process, moltenfuel formation and relocation as weil as early fission product release can takeplace even weil below the melting point of Zry. The melts relocate towards coolerregions of the co re where they may cause coolant channel blockages upon solidification.Thin oxide layers on the Zry surface can delay the chemical interactions withInconel, but cannot prevent them. Zr02 interacts with metallic Zr while formingoxygen-stabilized a-Zr(O) resulting in a sub-stoichiometric Zr02. a-Zr (0) is capable of interacting with the Inconel alloy components. The oxygen dissolved in theZry lattice exerts an influence on the diffusion and/or dissolution processes; thereaction rates are slower. In all other cases, the reaction of Zry with oxygen orsteam is more favorable thermodynamically than the reactions with


15/53

-9-

Inconel, but as the integral bundle meltdown experiments [9] show, a steam environment cannot prevent the chemical interaction between Zry and Inconel.The chemical analysis of the reaction zone in the single-effect tests showed agreat similarity in the chemical composition of some phases (developed duringthe chemical interactions) compared with the analysis of integral CORA experiments [9] and chemical analysis on selected sampies from the TMI-2 reactor core[7], fo r example the phases Zr2 (Ni, Fe) and -Zr.The chemical interactions which take place during the integral fuel bundle testscould be explained only by the single-effects test results. The complexity of a se-vere reactor accident scenario with the resulting multitude of materials interactions make a complete interpretation of the many possible chemical interactionsextremely difficult and in some cases even impossible.

5. Summary and Conclusions- Inconel 718 in contact with Zircaloy gives rise to chemical interactions which

can be described bya parabolic rate law.

- As a result of eutectic interactions at temperatures as low as 1000C, the liquidphase causes a fast and complete Iiquefaction of both components of the reaction couple at about 1250C.

- Onlya small amount of Inconel is necessary to Iiquefy large amounts of Zry.- Thin Zr02 layers on the Zircaloy surface delay liquid phase formation and reduce the rates of reaction, but cannot prevent the chemical interaction.In experiments of short durations, the Zr02 layers shift the Iiquefaction temperature to higher values; the required incubation period depends on the initial oxide layer thickness and on the temperature.Liquid phase formation due to the chemical interactions at low temperaturesis not prevented under unlimited steam oxidation conditions.


16/53

- 10-

6. AcknowledgementThe authors wish to thank Dr. T. Haste (AEA Technology, Winfrith) for his thorough and critical review of this paper.This work was partially funded by the Commission of the European Communitiesunder the Reinforced Concerted Action Project "Core Degradation", ContractNumber FI3S-CT92-0001.

7. References[1] S. Hagen, P. Hofmann, G. Schanz, L. Sepold; Interaction in Zircaloy/U02Fuel Rod Bundles with Inconel Spacer at Temperatures above 1200C; Results of the Experiments CORA-2 and CORA-3, KfK Report 4378,1990.[2] P. Hofmann et al.; Low-Temperature Liquefaction of LWR Core Components, Severe Accident Research Program ReviewMeeting, Brookhaven National Laboratory, Upton, NewVork, April 30 - May 4,1990.[3] Metals handbook, Vol. 1, 10th edition, 1990, ASM International, MaterialsPark, OH, USA.[4] G. Miller; Zirconium; Academic Press Inc., Publishers 111, N. V., 1986[5] T. B. Massaiski; Binary alloy phase diagrams, Vois. 1,2, American Society forMetals, Ohio, 1986.[6] D. Arias, J. P. Abriata; The Fe-Zr system, Bulletin of Alloy Phase Diagrams,Vol. 9, No. 5, 597 - 604,1988.[7] H. Kleykamp, R. Pejsa; Chemical and X-Ray Diffraction Analysis on SelectedSampies f rom the TMI-2 Reactor Core, KfK 4872; May 1991.[8] P. Hofmann, M. Markiewicz; Chemical interaction between as-receivedand pre-oxidized Zircaloy-4 and stainless steel at high temperatures;KfK 5106,1994.[9] S. Hagen, P. Hofmann, G. Schanz, L. Sepold; Results of the severe CORAfuel damage experiments CORA-2 and CORA-3; KfK 4376, 1989.[10] B. Daniel, A. Nichols, J. Simpson; The Iiquefaction of alloys identified withstructural materials in a PWR core; AEEW-M 2250, Dec. 1985.[11] F. Nagase, T. Otomo, H. Uetsuka; Interaction between Zircaloy tube andInconel spacer grid at high temperature; JAERI-M 90 -165,1990.


17/53

- 11 -

List of TablesTable 1: Quantitative analysis of phases in the Zircaloy-4/lnconel 718 reaction

zone (Fig. 12).Table 2: Measured total thicknesses of reaction zones in Zircaloy and Inconel for

the as-received Zircaloy-4/lnconel 718 diffusion couple as a function oftemperature and time (Fig. 13).

Table 3: Reaction zone growth rates in Zircaloy and Inconel for the as-receivedZircaloy-4/lnconeI718 diffusion couple (Fig. 15).

Table 4: Measured total thicknesses of reaction zones in Zircaloy and Inconel fo rthe pre-oxidized Zircaloy-4/lnconeI718 diffusion couple; initial Zr02 oxide layer thickness: 20 11m (Fig. 20).

Table 5: Measured total thicknesses of reaction zones in Zircaloy and Inconel forthe pre-oxidized Zircaloy-4/lnconeI718 diffusion couple; initial Zr02 oxide layer thickness: 45 11m (Fig. 21).

Table 6: Reaction zone growth rates in Zircaloy and Inconel for the pre-oxidizedZircaloy-4/lnconel 718 diffusion couple; initial Zr02 oxide layer thickness: 20 11m (Fig. 22).

Table 7: Reaction zone growth rates in Zircaloy and Inconel fo r the pre-oxidizedZircaloy-4/lnconel 718 diffusion couple; initial Zr02 oxide layer thickness: 45 11m (Fig. 23).


18/53

- 12 -

List of Figures

Fig.9:

Fig.4:

Fig.5:

Fig.8:

Setup of the reaction couples.Binary phase diagram of the Ni-Zr system [5].Binary phase diagramsof the Fe-Zr and Cr-Zr systems [5, 6].Chemical interactions between Zircaloy-4 and Inconel 718 at dif ferenttemperatures; influence of a thin Zr02 layer on the reaction behavior(right column of pictures).Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1000C/5 min.Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1000C/15 min.Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1100C/2 min.Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1100C/10 min.Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1150C 1min.

Fig.10: Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1150C 12min.

Fig.7:

Fig.6:

Fig.1:Fig.2:Fig.3:

Fig. 11: Chemical composition ofthe Zircaloy-4/lnconel718 reaction zone after1min at 1150C.Fig. 12: Isothermal section of the Fe-Zr-Ni and Fe-Zr-Cr system at 1000C [7]. Relation between the observed microstructures and the phase diagrams (Table 1).Fig.13: Maximum reaction zone thicknesses in Zircaloy and Inconel for theZircaloy-4/lnconel718 system between 1000 and 1200C (Table 2).Fig.14: Comparison of the reaction zone thicknesses in Zircaloy and Inconel forthe Zircaloy-4/lnconel718 system versus temperature;annealing time: 5 min.Fig.15: Reaction zone growth rates for the Zircaloy-4/1nconel 718 reaction system (Table 3).Fig.16: Chemical interactions between Zircaloy-4 and Inconel 718; 1000C/30min. Influence of a 20 11m thick Zr02layer.Fig.17: Chemical interactions between Zircaloy-4 and Inconel718; 1200C/5 min.Influence of Zr021ayers of different thicknesses.Fig.18: Chemical interactions between pre-oxidized Zircaloy-4 and Inconel 718as a function of time.


19/53

- 13 -

Fig.19: Chemical interactions between pre-oxidized Zircaloy-4 and Inconel 718as a function of time.Fig.20: Maximum reaction zone thicknesses in Zircaloy and Inconel for the preoxidized Zircaloy-4/lnconel 718 systems; initial Zr02 layer thickness: 20p.m (Table 4).Fig. 21: Maximum reaction zone thicknesses in Zircaloy and Inconel for the preoxidized Zircaloy-4/lnconel 718 system; initial Zr02 layer thickness: 45p.m (Table 5).Fig.22: Reaction zone growth rates for the pre-oxidized Zircaloy-4l1nconel 718system; initial Zr021ayer thickness: 20 p.m (Table 6).Fig.23:Fig.24:

Reaction zone growth rates for the pre-oxidized Zircaloy-4/lnconel 718system; initial Zr021ayer thickness: 45 p.m (Table 7).Reaction zone growth rates in Zircaloy for the as-received and preoxidized Zircaloy-4/lnconel-718 system; influence of Zr02 layer thicknesson reaction rate.

Fig.25: Reaction zone growth rates in Inconel for the as-received and preoxidized Zircaloy-4/lnconel-718 system; influence of Zr02 layer thicknesson reaction rate.Fig.26: Schematic of the Zircaloy-4 tubellnconel 718 grid system spacer used inexperiments under oxidizing conditions (steam).Fig.27: Chemical interactions between Zircaloy-4 tube and Inconel 718 after 3min at 1150C in steam atmosphere.


20/53

-14-

Table 1: Quantitative analysis of phases in the Zircaloy-4/Inconel718 reaction zone (Fig. 12).Reaction layer Element Cone. in at. % Phase(see Fig. 12)

Zr 95I, 11 Sn 2 -Zry

Ni, Fe, Cr 3Zr 68

1,11 Ni 22 Zr2 (Ni, FehFe 7Cr 3Cr 48

I, 11 Zr 34Fe 11 Zr (Cr, FehMo 4Ni 3Cr 65

1 Zr 12Fe 11 Cr2 (Zr, Fe)Mo 7

Ni, Nb 5


21/53

-15-

Table 2: Measured total thicknesses of reaction zones in Zircaloy and Inconelfor the as-received Zircaloy-4/1nconel718 diffusion couple as afunction oftemperature and time (Fig. 13).Temperature Time Reaction zone thickness, 11mSpecimen

oe min Zircaloy-4 Inconel718

29 1000 30 675 230

30 " 60 790 255

28 " 120 1290 55031 1100 10 720 200

32 " 15 975 275

36 1200 3 995 507

35 " 5 1315 700


22/53

- 1 6 -

Table 3: Reaction zone growth rates in Zircaloy and Inconelf o r t h e as-received Zircaloy-4/lnconeI718 diffusioncouple(Fig. 15 ).Reaction zone growth rate, x2/ t

Temperature cm2/soeZircaloy-4 Inconel718

1000 2.1710-6 3.2510-71100 9.40.10-6 7.7410-7

1200 5.6610-5 1.5510-5

Rate equations:For Zircaloy-4

x2/ t (cm2/s) = 4.435104exp (-252093/RT)For Inconel718Ix2/t (cm2/s) = 2.884lOSexp (-294962/RT)

R = 8.314 J/mol'K


23/53

- 17 -

Table 4: Measured total thicknesses of reaction zones in Zircaloy and Inconel forthe pre-oxidized Zircaloy-4/lnconeI718 diffusion couple; initial Zr02oxide layer thickness: 20 llm (Fig. 20).

Time Reaction zone thickness, llmSpecimen Temperaturemin Zircaloy-4 Inconel

009 1100 120 240 140037 " 180 1890 514011 1200 5 400 255010 " 10 840 975012 " 15 1440041 1300 1 740 856049 " 1.5 1008 1130


24/53

-18-

Table 5: Measured total thicknesses of reaction zones in Zircaloy and Inconel fo rthe pre-oxidized Zircaloy-4/lnconeI718 diffusion couple; initial Zr02oxide layer thickness: 45 11m (Fig. 21).

Time Reaction zone thickness, 11mSpecimen Temperature min Zircaloy-4 Inconel043 1100 180 290 110044 " 300 1095 306023 1200 5 190 105022 " 10 595 485024 " 15 850 930047 1300 1 440 290048 " 2 1040 1255


25/53

- 19-

Table 6: Reaction zone growth rates in Zircaloy and Inconelforthe pre-oxidized Zircaloy-4/lnconeI718diffusion couple; initial Zr021ayer thickness: 20 11m(Fig.22).Reaction zone growth rate, x2/t

Temperature cm 2/soeZircaloy-4 Inconel718

1100 9.76-10-6 6.79-10-71200 3.18-10-5 2.95-10-51300 1.51-10-4 1.80-10-4


x2/(t-to) (cm2/s) = 20952exp (-246427/RT)For Inconel718

x2/(t - to)(cm2/s) = 1.7271013exp (-507395/RT)R = 8.314 J/mol'K

to values, soe Zry-4 Inconel

1100 7175 69121200 314 3391300 24 20


26/53

- 20-

Table 7: Reaction zone growth rates in Zircaloy and Inconelforthe pre-oxidized Zircaloy-4/lnconeI718diffusion couple; initial Zr021ayer thickness: 45 11m(Fig.23).Reaction zone growth rate, x2/t

Temperature cm 2/soeZircaloy-4 Inconel718

1100 1.5510-6 1.1210-71200 1.1410-5 1.4210-51300 1.48.10-4 2.49.10-4


x2/(t-to) (cm2/s) = 5.44109'exp (-41 0306/RT)For Inconel718

x2/( t - to) (cm2/s) = 5.41019exp (-699018/RT)R = 8.314 J/mol'K

to values, soe Zry-4 Inconel

1100 12900 97411200 276 3401300 68 40


27/53

-21-

L . . -__ conical plugrod of Inconel 718

Zircaloy - 4 crucible

Fig. 1: Setup of the reaction couples.


28/53

Atomic Percent Zirconium0 10 20 30 40 50 60 70 80 90 100

,1""1" .. ,1'1' " .. ,I .... , I! ;, , I , !, . , , 1 I , l; , I 1 l ~ O C18001 At1600-1 L / tU t ~ o c /.uo C Il1400K ~~

t..Q)0.- 1 2 0 0 ~ \ ________ ~ _ o c : J \ I l l 0 0 ~ C \ ~ o J " / Ir) I\ )E- I\ )

1000 -I 1\Ni ) I , I I I IW lrnDC V I \ / ILJ80C3.4-~ r n e 6~ '- ~ ~N ".:z: :z i3 Z t2 ~ O C- Z -z :z: z cr.Zr)800 0 10 20 30 40 50 60 70 80 90 100

Ni Weight Percent Zirconium Zr

Fig.2: Binary phase diagram ofthe Ni-Zr system [5].


29/53

-23-

100

100Fe

91l.lIZ 91Z'C

(o:Fe)

90

90

920'CL. _. _._. _. _ . _ . 7 . ~ ~ : C _. _. _._ .._._. _._.: lIa,nalle Trllndormlllion1~ Z r F e 3,, 470'C

"\ \'. 1462'Cr - ~ - - - - - - - - - -,,I

1673'55.1.I1,

Alomic Percenl Iron50 60 70I I t I 'I30,

L

20I

o:Zr)600

400

200

uo

20001 Ip~ 8 5 5 ' C

1600- "" ', '1600 \ '\I ,, ,, ,, ,1400 \ """ ," ,*(pZr)\ /, , ,1200 \ \ I

::l \ \ ~ 1 6 . 2 /-J , \,1000 1 " ' 974'CL-. _-&.' 926C \ - - - ~ - - - - - - - - - _ ..._ - - - - - - - - - - -


30/53

-24-

10000 (/15min- - - - -Zry -4--

Inconel 718 - -

1100 0(/2minreaction zone

(liquid ...at temperature)

1150 0 (/2min

Z r y - 4 : 1 O ~ 1 1 l 1 h i c k ZrO 2layeron 1.0.

Fig.4: Chemical interactions between Zircaloy-4 and Inconel 718 at differenttemperatures; influence of a thin Zr02 layer on the reaction behavior(right column of pictures).


31/53

-25-

31

2reaetion zone ( liquid ~at temperature )

Zircaloy - 4 e ,. '

Zry

reaetionzone

Inconel

11000O(/5min IFig.5: Chemical interactions between Zircaloy-4 and Inconel718;

annealing conditions: 1000C/5 min.


32/53

2

1

-26-

Zry - 4reaction zone

(liquid at temperature)Inconel 718

2mrn'>-----1

Inconel

11000 (/15min IFig.6: Chemical interactions between Zircaloy-4 and Inconel718;

annealing conditions: 1000C/15 min.


33/53

-27-

i ': e,"rjZircaloy - 4 .

12

Zircaloy-4

3

, 1 0 0 ~ m ,

once molten reaetion zone

InconelI

Zryinterface

Fig.7: Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1100(/2 min.


34/53

once malten reaction zone

1

localized melting of theZircaloy crudble

-28-

32

Inconelinterface

reactionzone

Zry

1100 0(/10min

Inconel718." .

-j

. '

-Iinterface

Fig.8: Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1100(/10 min.


35/53

2

-29-

penetration of the Inconel rodinto Zifca!oy crucib!e

Zry-4

Ineone! 718

[TI IIneonel

reaetionzone

ZryI

1150 0(/lmin

interface- Zry

Fig.9: Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1150C 1 min.


36/53

-30-

23

12mm

t - - - - - - - I L ~ o o ~ m : jonce mol ten r n "\t"t'.nn zone

readionzone

Ineonel

Zry

. readionzone

Ineonel

1150 0(/2min

interface

Fig.10: Chemical interactions between Zircaloy-4 and Inconel718;annealing conditions: 1150C / 2 min.


37/53

-31-

Inconel 718Zr

Ni

r--Zr:49, Cr:26, Fe:12, Mo:6, Ni:4, Nb:3wt.%] [ ~ Z r : 7 8 , Ni:9, Cr:5, Fe:4, (Mo+Nbl:4 wt.%][- - Cr: 55, Zr: 18, Fe: 11, Mo: 10, (Ni+Nbl: 5wt.%

Cr

Fer--Zr:95, Sn:3, (Ni+(r+Fel:2wt.%

] ( ~ Z r : 8 2 , Ni:13, (Fe+(rl:S wt.%

/1150 0U1min Ifig. 11: Chemical composition of the Zircaloy-4/1nconel718 reaction zone after

1 min at 1150C.


38/53

-32-

reaction layer I /

reaetion zone - - - -

1000C/15min

\

T-Fe

solid attemperature

....'()

Z,Ftl Z,Ft l FtZr120 40 60-- al.%Zr - p-Zr

1000 0 (

l-Fe n-Zr

lAI: integral analysis of reaction layer I (BI: integral analysis of reaction layer 11

Fig. 12: Isothermal section of the Fe-Zr-Ni and Fe-Zr-Cr system at 1000C [7]. Relation between the observed microstructures and the phase diagrams (Table 1).


39/53

VJVJI

no Zr02 layer,as-received__ reaction zone in Inconel_ reaction zone in Zircaloy1000 0 (

I::..

1200 0 (112000(I 1100 0 (

" "1100 0 (,,", ,",,""," I::..00"-'; _----------10000(

0" _-1::..-- I::..1 5 I

,,.,,,IIIHili'0If/I.

2

QJCoNCo...roQJCl::

N Xl/ )l/ )QJC~ u.c...

N EE

Annealing time t , hfig.13: Maximum react ion zone thicknesses in Zircaloy and Inconel fo r the

Zircaloy-4/lnconeI718 system between 1000 and 1200C (lable 2).


40/53

-34-

1500

e: :L

QJc:oNc:o:z:'- ', . "QJc::::

1000

500

1000

Zircaloy-4/Inconel 718annealing time: Smin

reaction zone in Zry\

1100

Liquefactionof thespeclmensabove 12500 (

\reaction zone in Inconel1200

Temperature, oeFig.14: Comparison of the reaction zone thicknesses in Zircaloy and Inconel for

the Zircaloy-4/lnconeI718 system versus temperature;annealing time: 5 min.


41/53

-35-

Temperature,O(1200 1100 1000no Z r 0 2 ~ y e r , a s - r e c e i v e d_ reaetion rate in Zircaloy_ _ reaetion rate in Inconel

V ).............E Ix2/t=44354 exp (-252092/RT) I- '..0- 10-4........ /....>< AQJ. . 0 -n: Ic.....r::::...0- 0~ 10-5 "...c:n "Jc::

/"Nc::0 " A:+= x2/t=288455 exp (-294962/RT) "- ' ":I 10-6 "Jc: : """~ = 8 . 3 1 4 J/ molKI "" 0 ,10-7 7 8

Reciprocal temperature , 1fT , 104/K

Fig. 15: Reaction zone growth rates for the Zircaloy-4/lnconel 718 reaction sys-tem (Table 3).


42/53

as-received Zry

pre:'oxidized Zryinitial Zr02 layer: 2 0 ~ m

-36-

voiddue tomeltrelocation

loconel/lry eutecti.cinteractions

Inconel

Inconel/pre-oxidized Zry

1000C/30min

lX-ZriOI

Fig.16: Chemical interactions between Zircaloy-4 and Inconel 718; 1000C/30min. Influence of a 20 11m thick Zr02layer.


43/53

-37-

as-received Zryno Zr02 layer

Inconel/Zircaloy eutectic interadions

1mmt----I

pre-oxidized Zryinitial Zr0 2 layer: 2 0 ~ m

pre-oxidized Zryinitial Zr0 2 layer: 4 5 ~ m 2mmt - - - l

In(l}nel

\--reactionzone

1200C/Smin

5 0 ~ mI------i

Fig.17: Chemical interactions between Zircaloy-4and Inconel718; 1200C/5 min.Influence of Zr021ayers of different thicknesses.


44/53

-38-

12000 ( / 5min

2mm1----1

Uquefiedmaterial

12000 ( / 10min

liquefiedmaterial

12000 ( / 15min

Initial Zr02 layer thickness: 4 5 ~ m

Fig.18: Chemical interactions between pre-oxidized Zircaloy-4 and Inconel 718as a function of time.


45/53

-39-

gap

Inconel --+---..

12000[ / 5min

reaetionzone In -i--........Inconel

12000( / 10min

.o4----i-- Zircaloy

initial interface

initial interface

reaetionzone InInconel

20llffiI-------l

12000( / lSmin

Initial Zr02 layer thickness: 4 5 ~ mFig.19: Chemical interactions between pre-oxidized Zircaloy-4 and Inconel 718

as a funct ion of time.


46/53

1300 0 (N JUOOO ( initial Zr02 thickness : 20E /11000 ( __ reaction zone in InconelE 2 :1 _ reaction zone in ZircaloyN x AV) IV) IJc: Iu.c. I~QJ Ic: 1 I;0 : Iv I~ .. 0u Iru tJ I "",,,1100 0 (J0::: ,,"'"."'", I ,,"OIY i *,,"-1 2 3 4 5Annealing time t , h

Fig.20: Maximum reaction zone thicknesses in Zircaloy and Inconel fo r the preoxidized Zircaloy-4/1nconel 718 systems; initial Zr02 layer thickness: 2011m (Table 4).


47/53

.....I

,/11000 (

initial Zr02 thickness : 45 11m_ _ reaction zone in Inconel_ reaction zone in Zircaloy

K_-- - - - - - _&--1100 0 (345Annealing time t , h

1300 0 (N E ,12000 (E 2 ',12000 (N x ,VJ ,VJQJ Ic::

~ ,-~ '1QJc::0 1Nc::0~ I;rtJQJ0::: v,tu-- 1 2

Fig. 21: Maximum reaction zone thicknesses in Zircaloy and Inconel fo r the preoxidized Zircaloy-4/1nconel 718 system; initial Zr02 layer thickness: 45p.m (Table 5).


48/53

-42-

Temperature,oe1300 1200 1.100

V I""'0oIE1...1-c::........I....... 10-4-""'0oI>.\ " "

x 2 / l t - t ~ = 5 . 4 10 19 exp 1-699018/RTI \ I I\ .R-:8.314 J/ molK.7 0\10- . . L . . - - - - - - - - r - - _ ~ _ _ _ I7 8

Q J,e-.J::-E 10-5c:::nQ Jc:::o......c:::o:.;::'- '~ 10-6

Reciprocal temperature , 1/T , 104/KFig.23: Reaction zone growth rates fo r the pre-oxidized Zircaloy-4/1nconel 718

system; initia l Zr021ayer thickness: 45 11m (Table 7).


50/53

-44-

Temperature,O(1300 1200 1100 1000

cu-"Oc.... .c--~ 10-5c:::ncucoNc:o:z:"-'t"O~ 10-6

re action rate in Zirca[oy- - no Zr0 2 layer,as-received (t 0=0)_. _ initial Zr02 thickness : 20 pm_ _ initial Zr02 thickness : 45 pm

10.7. . . 1 . - - - - - - - - - , - - - - - - - - . . . . . ;7Reciprocal temperature I 1/T , 104/K

Fig.24: Reaction zone growth rates in Zircaloy for the as-received and preoxidized Zircaloy-4/lnconel-718 system; influence of Zr021ayer thicknesson reaction rate.

8


51/53

-45-

Temperature,O(1300 1200 1100 1000

reaction rate in Inconel-- no Zr02 layer,as-received (t0=0)_ ._ initial Zr02 thickness : 20 ~ m_ initial Zr02 thickness : 45 ~ m

G.I,'-. c- 10-5cnG.Ic:oNc:o:z:~",~ 10-6

10-7 . . . . 1 . - - - - - - - - , - - - - - = - - - - - - - - - - 17

Reciprocal temperature , 1fT , 104/K 8

Fig.25: Reaction zone growth rates in Inconel for the as-received and preoxidized Zircaloy-4/1nconel-718 system; influence of Zr021ayer thicknesson reaction rate.


52/53

-46-

' ~ - welding

Inconel 718

end cup

Inconel grid/Zry -4 tube\ ....--y ..-,

\--- \---...

gas suppl

Fig.26: Schematic of the Zircaloy-4 tubellnconel 718 grid system spacer used inexperiments under oxidizing conditions (steam).


53/53

-47-

1.0.

0.0.

Zry - Inconel --tP-Icontad region

1--------

once malten readion zone

1

readionzone

ZryI

liquid attemperature

interface

11150 0(/3min Isteam environment

P. Hofmann and M. Markiewicz

Documents