ML20204D501
| ML20204D501 | |
| Person / Time | |
|---|---|
| Issue date: | 12/04/1978 |
| From: | Picklesimer M NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| References | |
| NUDOCS 7812200310 | |
| Download: ML20204D501 (110) | |
Text
{{#Wiki_filter:_. jN 4,
# UNITED STATES y- 4 NUCLEAR REGULATORY COMMISSION
, WASHINGTON, D. C. 20555
***s CEC 4 W3 MEMORAF90M FOR: File FROM: M. L. Picklesimer Fuel Behavior Research Branch
SUBJECT:
COMMENTS ON THE SYMPOSIUM ON EMBRITTLEMENT OF ZIRCALOY BY OXIDATION, SIXTH WATER REACTOR SAFETY RESEARCH INFORMATION MEETING, NATIONAL BUREAU OF STANDARDS, GAITHERSBURG, MARYLAND, NOVEMBER 6, 1978 < In as much as the research program on embrittlement of Zircaloy by > oxygen at the Argonne National Laboratory is nearing completion, the FBRB felt it desirable to hold a small symposium'on the topic, with papers invited from knowledgeable workers in the field. The best oppor-tunity appeared to be as part of the agenda (enclosure 1) of the Sixth WRSR Information Meeting, NBS, November 6-9, 1978. Four presentations L were made-at the symposium, one each from the USNRC, AECL, UKAEA, and ;
) JAERI programs on embrittlement criteria. Copies of the handouts are l , attached, as is an updated referenceable report by P. D. Parsons, SNPDL/
{ UKAEA on his review of embrittlement criteria. The presentations are reviewed individually below,.and then discussed collectively. When i sufficient published reports are available, a Research Information 4 l Letteri(RIL) on Embrittlement Criteria for Oxidized Zircaloy Fuel Element i Cladding will be prepared and issued. ZIRCALOY CLADOING EMBRITTLEMENT IN A LOSS-OF-COOLANT ACCIDENT ' i P. D. Parsons, UKAEA P. D. Parsons, Springfields Nuclear Power Development Laboratory, UKAEA, discussed the work he has done on evaluating the USNRC ECCS embrittlement criteria as stated in 10 CFR 50, Appendix K for applicability to the UK Steam Generating Heavy Water Reactor (SGHWR), which uses a thinner wall Zircaloy-2 cladding than is used in LWRs in the USA. The handout (enclo-
. sure 2).is an abstract of the formal report also appended (enclosure 3). )
Mr. Parsons conducted a review and evaluation of the earlier work upon l which the present USNRC ECCS criteria are based, subsequent modifications
, suggested by Pawel, etc., work on embrittlement reported by Sawatsky and I by Kawasaki, and later laboratory work by Cathcart, Leistikow, Biederman, I
and others on rates of oxidation of Zircaloy in steam and rates of diffusion of oxygen in the oxide, alpha, and beta phases. He then conducted laboratory studies of oxidation in steam and spray quenching of specimen tubes in a GLEEBLE, using SHWR fuel cladding material and 781220o310 Q
w , stb Files sizes. His principal findings were (1) that the 17% equivalent cladding oxidation and 2200*F peak clad temperature of the USNRC 10 CFR 50 ECCS criteria were adequate for the SGHWR % undergoing thermal stressing by quenching, (2) that analysis of the subsequent updating of embrittle-ment by pawel to propose a limit of 0.7 wt% oxygen in the beta phase was more applicable for a wider range of accident scenarios, and (3) that Sawatsky's recommendation of a limit of 0.7 wt% oxygen in not less than 1/2 the wall thickness would probably ensure safety in handling fuel bundles in the disassembly of a damaged core. He also pointed out the need for methods allowing calculation of the oxygen distribution in the wall of cladding that had undergone an arbitrary time-temperature sequence, and for a better knowledge of how oxygen modifies the mechanical response of the several phases present in oxidized fuel cladding. INNER SURFACE OXIDATION OF ZIRCALOY CLADDING IN A LOSS-0F-COOLANT ACCIDENT S. Kawasaki, JAERI Dr. S. Kawasaki, Japan Atomic Energy Research Institute (JAERI) reported on the work he and several co-workers have been conducting on the oxidation occurring on the inner surface of Zircaloy fuel element clad-ding ruptured in flowing steam (enclosure 4). Two types of experiments i were conducted. In the first, short lengths (rings) of Zircaloy pWR cladding were oxidized for times up to 20 minutes at temperatures rang-
- ing from 950 to 1100*C in an atmosphere of hydroger. plus steam at ratios from 0.04 to 2.5 (VH2 /VH 20). The oxidized specimens were then compressed in the radial direction in a tensile testing machine until fracture occurred to determine the loss of ductility in bending (called the ring compression test). When the volume ratio was approximately 0.25-0.4 (H2 / steam, and dependent somewhat on temperature), the specimen absorbed large amounts of hydrogen, and the deflection to fracture in the ring compression test decreased to smal: values. In the second experiment, short len pellets (gths (rings)fuel simulating of Zircaloy pellets),cladding internallywere filled with pressurized alumina with helium, and plunged into a muffle furnace at a preselected oxidation temperature and having a flowing steam atmosphere. During the temperature transient, the specimen burst at temperatures near 900*C to produce rupture open-ings of various sizes, and steam invaded the interior of the specimens.
After holding for the given time, the ruptured and oxidized specimens were withdrawn into a cold part of the apparatus. Ring compression test specimens were then cut from several locations on the oxidized fuel cladding. Again, ductilities were determined at 100*C. Correlation of deflections to fracture with position in the oxidized specimens showed the minimum deflection to occur most frequently at the start of the bulge of the ballooned section, where the gap between the pellets and the cladding was small. Chemical analyses of these sections found hydrogen contents up to about 1300 ppm H2 . Dr. Kawasaki found that the 4
i DEC s ;g3 Files reaction time to cause the deflection minimum decreased with increasing ' reaction temperature, and that the location of the section having the minimum deflection moved further away from the rupture opening as the flow rate of steam outside increased and the area of the rupture opening increased. He concluded that hydrogen as well as oxygen influences the embrittlement of the cladding (see " discussion" for additional data that l may modify this conclusion). He believes that the hydrogen was absorbed at the location on the inner surface of the specimen where the supply of steam was inadequate to maintain an oxide film on the surface of the cladding (hydrogen should diffuse through the helium gas in the gap . faster then steam, and should enter the metal in a region of laminar flow where mixing could not occur from the turbulence at the opening of the rupture). i ZIRCALOY CLADDING EMBRITTLEMENT, RECOMMENDED CRITERIA i T. F. Kassner, ANL i Dr. T. F. Kassner, Argonne National. Laboratory, then presented a summary of the experimental findings he has cbtained, with his co-workers, in the RES/NRC funded study on the effects of oxygen on the mechanical properties of Zircaloy (enclosure 5). The study has been conducted to allow a more quantitative statement on embrittlement than the present ECCS criteria permit, and to establish a suitable test method for evalu-ating the degree of embrittlement of the cladding. Earlier presenta-tions by him have reported the mechanical properties of Zircaloy as functions of oxygen content, oxygen distribution, temperature, micro-structure and test method. This report was primarily concerned with the establishment of failure boundaries on various types of data maps and a correlation of impact energy absorbed during fracture with other types of failure evaluation. For most of the data reported herein, short sections of the " master lot" of Zircaloy fuel element cladding (used in ' all of the FERB/RES cladding programs) were ruptured in steam during transient heating (Joule), oxidized at a maximum temperature between 1140 and 1770K for various times, cooled from the oxidation temperature to about 1100K at a rate of about SK/s, and rapidly quenched by bottom flooding with water at a rise rate of 0.05 m/s. Specimens were also quenched directly from the oxidized temperature, which resulted in a cooling rate of approximately 100 X/s through the beta-alpha phase . transformation. These experiments incorporated thinning of the cladding ! wall by the formation of multiple ballooned regions, oxidation of the inner and outer surfaces by steam, transformation of the central beta-phase region of the cladding to an alpha prime structure in which the cooling rate influenced the extent of oxygen redistribution in the microstructure, and thermal stress produced by bottom flooding with water. Those specimens which survived thermal quenching were subjected to an impact load, in situ, by a weighted pendulum losated to strike the specimen near its mid-point. The weight was increased incrementally until the specimen fractured or the maximum weight permissible had ceen l 1
Ob . 1978 Files used. Failure " maps" for fracture of the cladding by thermal shock were developed relative to the maximum oxidation temperature and various time-dependent oxidation parameters e.g., equivalent-clacding reacted to form Zr02 , fractional thickness of transformed beta layer, fractional
! saturation of the beta-phase by oxygen, and thickness of beta phase with less than a specified critical oxygen content. The principal results l are (1) if the cladding is cooled rapidly (about 100 K/s) through the beta-to-alpha phase transformation, the thermal shock failure boundary ; corresponds to about 20% of the wall thickness in equivalent oxidation for oxidation temperatures greater than about 1650 K, (2) if the clad-ding is cooled slowly through the phase transformation, the thermal shock fa11ure boundary corresponds to 28% equivalent wall thickness oxidized to Ir0 2, and (3) the best correlation of thermal shock failure boundary with parameters related to the degree of oxidation of the cladding was that of thickness of the beta phasa layer having less than 0.9 and 1.0 wt% oxygen for slow and fast-cooled cladding respectively, if the beta phase layer having less than this oxygen level was 0.1 mm in thickness, or more, the cladding did not fail irrespective of wall i
thickness, oxidation temperature, and total oxygen content. In-situ pendulum load impact tests at room temperature showed that the thermal shock boundary corresponded to approximately 0.03 J impact energy absorbed. Data were plotted as time of oxidation versus reciprocal temperature and evaluated as to (a) failed on thermal shock quenching, (b) failed at i 0.03 J impact, (c) survived 0.03 J but failed 0.3 J impact, and (d)
, survived 0.3 J impact load. These results allow a quantitative state-ment of energy absorbed by fracture, and a quantitative failure map to be drawn with any desired degree of conservatism. Finite-element model has been developed for crack growth in oxidized Zircaloy during thermal shock conditions, u:ing mechanical properties measured for homogeneous
- l. specimens of varied oxygen contents. The data suggest that the present ECCS embrittlement criteria are conservative both to failure by thermal shock and failure by impact loads likely to be encountered in disas-sembly of the core.
0XYGEN EMBRITTLEMENT OF ZIRCALOY-4 FUEL CLADDING A. Sawatsky, AECL The last presentation of the Symposium was by Mr. A. Sawasky, AECL-Whiteshell, who discussed the research he has conducted to determine mechanical properties of Zircaloy containing various amounts of oxygen as part of a data base for the computer program FAXMOD, which is used by AECL to predict the thermal-mechanical behavior of fuel elements. In the present work (see enclosure 6), he determined the effects of oxygen content and distribution, maximum temperature, and cooling rate on the tensile properties of Zircaley. Oxidation was in steam, and the tensile tests were conducted in air over the temperature range 24 to 500*C, and in argon at higher temperatures. Ring specimens were used, with "D" grips inserted to produce circumferential loads on 10 mm wide rings. For
DEC s W79 Files oxygen contents of 0.5 wt% and greater, the ultimate tensile strength had a maximum at a temperature which increased with increasing oxygen concentration, while the elongation generally increased with increasing temperature and decreased with increasing oxygen content, with no sudden change indicating a ductile-brittle transition. The major results are (1) a compilation of data on tensile properties as functions of temper-ature and oxygen content and distribution, and (2) the observation that 3 the maximum temperature and total oxygen content have little or no effect on the tensile properties of Zircaloy-4 cladding, but that the oxygen distribution controls, particularly that in the low-oxygen part of the cladding. In discussion not included in the handout, he endorsed the embrittlement criteria proposed by Pawel (based on Hobson and Rittenhouse data) of a maximum content of oxygen in the beta phase of 0.7 wt%, but modified it to specify "over at least half the wall thickness." DISCUSSION The presentations by Parsons, Kassner, and Sawatsky seem to agree that (1) the present ECCS embrittlement criteria are conservative relative to the thermal shock stresses produced by reflood quenching and probably tJ the loads placed on the damaged assembly during disassembly of the core. All concur on the need for more quantitative criteria, and that these should be related to the oxygen content of the beta phase present in the , cladding during high temperature oxidation. The observation by Kassner that a failure map can be developed based on a minimum wall thickness of 0.1 mm having less than 0.9-1.0 wt% oxygen and the recommendation by Sawatsky that at lesst one-half the wall have a maximum average oxygen content of 0.7 wt% are not really inconsisent, since Kassner's failure i boundary of 0.1 mm beta phase is a sharp boundary for failure by thermal i shock, and Sawatsky's reconnendation has some conservatism incorporated to ensure safe handling. If Kassner's failure boundary is made conser-vative, then the two should agree. While Parsons' presentation primarily concerned the applicability of the present ECCS embrittlement criteria of 17% equivalent oxide and 2200*F PCT to the SGHWR fuel, he recommended that more quantitative criteria based on measurable properties be devel-oped, along the lines of work presented and discussed by Kassner and Sawatsky. i l
l Files ^0EC 4 1973 awaskai's observation of decreased deflection of ring compression specimens due to hydrogen absorption through the inner surface of ruptured cladding has caused some concern that the present ECCS embrittlement criteria may not be conservative. The observation must be placed in proper perspective, and must be correlated with other data obtained by , Kawaski but not yet formally reported in those JAERI reports we have recaived. The points to be raised are these: (a) even the most brittle ring specimen tested by Kawasaki showed a failure load of more than 50 kg and a deflection of at least 0.3 mm, (b) the hydrogen absorption and " embrittlement" ocw ; at the edge of the ballooning bulge, where the gap between the fuel pellet and the cladding is still small, (c) in any loading situation for radial compression of this region in a fuel rod in a commercial power reactor, the fuel pellet would be encountered before the cladding could be deflected this amount and further deflection would require displacement or crushing of the fuel pellet, increasing the load required to produce failure, (d) the failure that is produced by the ring compression test is an axially oriented crack which grows through the wall while the most probable loading of a damaged fuel rod would be in bowing (with less concentration of stress), so that a circum-ferential crack would have to be initiated, grow through the wall in the transverse plane, and then proceed around the circumference of the fuel rod, (e) other data obtained by Kawasaki (see enclosure 7) shows that ring compression specimens containing up to 1200 ppm H 2 deflect less at all test temperatures (100, 200 and 300*C) when oxidized at 1000-1200'C , than when oxidized at 980-960*C, showing that the oxygen content is responsible for failure in such specimens oxidized at 1000*C and higher, (f) Kassner's impact data show that his ruptured specimens, oxidized in
< team to essentially the same condition as those by Kawasaki, survived not only reflood quenching stresses but 0.3 J impact load, (g) Kassner's ruptured specimens oxidized at temperatures below about 1250*C always failed at the edge of the ballooning bulge (where Kawasaki observed the maximum hydrogen aberption) whether the failure was produced by Mermal shock or by impact, while those oxidized at higher temperatura, v :?d in the ruptured part, (h) Kawasaki's specimen were cooled relaM5 slowly, leading to coarse hydride particles, while reflood que.ncW 3 would produce very small, dispersed hydride particles, as in Kasch's specimens, and (1) chemical analyses of the critical regions of Kassner's unfailed specirens, obtained too late to report at the Symposium, showed hydrogen contents up to 1200 ppm H2 comparable to those observed by Kawasa ki . While it is quite clear that Kawasaki's observations of hydrogen absorption from inner surface oxidation must be incorporated into any discussion on the establishment of new ambrittlement criteria, it appears to me that the combination of Kassner's and Kawaki's data indicate that the present ECCS embrittlement criteria are s:.fil conser-vative and the observation of decreased derlection in a ring compression test because of hydrogen absorption should not become an " embrittlement issue" at this time.
7
. Files DEC 1m Additional tests are being conoucted on the specimens of Kassner, et.al., I to determine more completely the impact energy required to fracture the specimens which had survived the 0.3 J pendulum impact test, the hydrogen contents and distributions in the microstructures of both failed and , unfailed specimens, ring compression tests on specimens similar to Kawasaki's to provide a quantitative tie between the two sets of data, and both impact and ring compression tests on hydrided specimens as functions of hydride particle size and dispersion. The data will be reported as quickly as possible. ,
/s/ede pl . Picklesimer M.
Fuel Behavior Research Branch Division of Reactor Safety Research
Enclosures:
- 1. Agenda for Symposium on Embrittlement of Zircaloy by Oxidation, Sixth WRSR Info. .
Meeting, NBS, November 6,1978
- 2. Handout-Zircaloy Cladding Embrittlement in a Loss of coolant Accident, P. D. Parsons, UKAEA
- 3. Report "Zircaloy Claddir1 Embrittleh nt in a Loss of Coolant Accident, P. D. Parsons, SNPDL, UKAdA DC07180(s), November,1978
- 4. Handout " Inner Surface Oxida-tion of Zircaloy Cladding in a Loss of Coolant :
Accident," S. Kawasaki, JAERI
- 5. Handout "Zircaloy Cladding .
Embrittlement, Recommended r Criteria," T. F. Kan ner, ANL
- 6. Handout "0xygen Embrittlement of Zircaloy-4 Fuel Cladding,"
A. Sawatsky, AECL
- 7. Untitled plot of data, Deflection of ring compression specimens versus absorbed hydrogen content, S. Kawasaki, private communication.
i 7 73
._~ _ . _ _ . , . _ . . . .. _ . .. _ _ _ , _ _ . , . . _ . _ . . _ , , . . - _ , _ _ -
ATTACHMENT 1 , e i i NOVEMBER 6, 1978 FUEL BEHAVIOR RESEARCH PROGRAM MORNING SESSION - GREEN AUDITOR!UM Symposium on Embrittlement of Zircaloy by Oxidation Chairman: M. L. Picklesimer, NRC 11:00 am - Invited Paper - Zircaloy Cladding P. D. Parsons, UKAEA Embrittlement in a Loss-of-Coolant Accident 11:30 am - Invited Paper - Embrittit.i,u.. of Zircatoy S. Kawasaki, JAERI Cladding: Japanese Program 12:00 pm - Zircaloy Cladding Embrittlement, T. F..Kassner, ANL Recommends;d Criteria 12:30 pm - Invited Paper: Embrittlement of Zircalcy A. Sawatzky, AECL Cladding 1:00 pm - Lunch AFTERNOON SESSION -- GREEN AUDITORIUM WORKSHOP - Multired Burst Tests 2:00 pm - 5:00 pm/3:15 pm - Coffee Break Chairman: M. L. Picklesimer, NRC Speakers: R. H. Chapman, ORNL Invited Speakers from England, Germany and Japan i-Ee
. . - . , , . . . . _ . , , . . . , _ . , - - ~r . .-,_ - _ - . . m. - - . _ , . .,,. .
w .
. - ATTACHMENT 2 ZIRCAICY CIATING DEPl~"TErff 3 A ICSS OF COC*A!C ACCIOCO P S Parsons UNITED CE)OM AT0fCC E!IRGY AI.*THCPlTY SPRINGFII:.OS NUCLFAR POWER DE"F._.DPMDf? IA3CPATOFlES PF.ESTON, PR4 CER, ClGIMD ABS"'?ACT In a postulated IWR loss of coolant accident (ICCA), the temperature of the c1Wng increases rapidly so that sissificant exidation takes place and cenecc:1 ant changes in 2 e mechanical properties of the cladding occur. The formation of nircor$a and high-oxygen-content stabilised alpha phase in the cladding wall section reduces the ability of the cladding to remain intact urrier the action of the thermal shock forces generatec by the rapid cooling of the emergency cooling system. Safety criteria are required which specify partmeters defining the amount of oxidation precluding sufficient embrittlement to induce failure of the cladding.
The currently specified US safety critarion of 1 5 equivalent metal reacted is based on the empirics.1 correlation of cladding fractured on cooling with the total exten: cf exidation. It cannot, in principle, be assumei to be applimb to cladding of thickness other 2an tha* used in the tests to define the criterien. The UK Steam Generating Heavy Water Reacter (SGEdR) has fuel rods clad in *.irci.lcy-2 which is thirner than that used to define the 15 total exidation criterion. Work has therefore been undertsken at SNL to date d e whether modification of the 15 criterion is required for the SGEdR 60 pin design of Pael (with wall thickness 0.0597-0.0697 cms) and ce higher rated 36 pin design (with wall thickness 0.06Al-0.C768 ces). Tubes 180 mm long were resistance heated in excess steam in a Gleeble test facility. After e. pre-determined isoth rmal or transient exidation exposure,*he specimen heatin; power was terminated and the tubes were quenched by a spray of cold water. Spray cooling was used as the SGHWR emergency water supply is delivered through one or more, many holed sparge tubes running the fall length of the Pael bundle. The specimens were assessed as either remaining intact er cracking after ex'datien, and meastrements of the total exidation were carried out by vacuum Pasion analysis for oxygen and also by using metallographic sections. An example of the limiting type of transient exposures considered for experimental purposes is shown in Fig. 1. The ability of 2e tubes to remain intac* in resisting the thermal stresses generated during a ICCA quench was regarded as an adequate property of 2e tube in defining an empirical failure criterion. The cracking of the tubes after the quench was readily apparent acd to detect possible incipient cracks a small pressure of arson was applied to the bore of the tube. In most experiments the outer oxide film. cracked and spalled from the tube leaving an inner very adherent oxide film. Only in very severe quenches from high temperatures did the complete oxide spall to reveal bright alpha 7.ircalcy which usually exhibited extensive cracking. In the SGHW reactor the Pael bundles are subject to oxidising water chemistry and exhibit nodular oxidation. Such cerrcaicn reaches significant thicknesses at high burn-ups and must 2erefore be taken into secou c with respect to embrittlement. Some Parthar 7.irealoy-2 tubes were pre-oxidised to evelop an out-of-reactor simulated nodular oxide before being subject to *.he IOCA type oxidation and quench tests in the Gleeble. The results of the oxidation and spray quench tests are presented in Figures 2 and 3 as empirical correlations of failure er survival after quenching, with 2e equivalent metal reacted to fe m zirconia. The figures show that for practical purposes the tests on virgin or lightly pre-czidised cle7 suggest a similar criterion to that initially preposed by I.iffengrea in that tubes with less than 15 equivalent metal reacted remain intact. The tests using nodularly pre-oxidised tubes were not done in 1 7
i i l sufficient numbers for unequivocal conclusions to be drawn but one tube with pre-oxidised nodular corrosion equivalent to 15 metal reacted was further oxidised and quenched from 13000C (23 LOOF), and did not crack on ecoling 1.e. tube 2 on Figures 2 and 3 The microstructure of failed and of intact tubes was e+ned and the spalled oxide was recovered for thickness measurements. The oxide was observed alwaya to spall at a boundary which delineated inner and outer layers of oxide by a plane of precipitatien usually around the mid section of the oxide. The precipitates were shown to be rich in tin and possibly zirconium but their crystallographic structure is as yet unreported. The inner exide structure is columnar and electron diffraction shows it to be twinned monoclinic zirconia at ambient temperature with many grain boundary cracks, probably due to the themal shock of the quench. The high o pgen stabilised alpha phase was always heavily cracked and in some cases where alpha incursions into the transfer =ed beta layer, were fcrmed, the cracks ran into the alpha incursiens. Since total oxygen uptake is an empirical basis for a criterien and does not take into consideration the individual effects of difhsed cugen in the different phases formed by exidaticn, it is desirable to have an improved criterion derived from the distribution of oxygen e m ughcut the clad wall and the effects of this distributien on mechanical propertiea of the cladding. The embrittlee -i of ::ircaloy depends upon ongen and ther=al induced changes in mechanical pro; _rties of the various material phases in the claddir4 and the nature and magnitude of the mard-tm forces en the Zirealey tube. Maximum stressing is calculated to occur at the Leidenfrost temperature when the clad is first rewetted. The mechanical properties of alpha and beta Zircaloy as a function of oxygen are currently under investig: Lion. Some criteria have already been suggested which pay particular attention to the amount and distribution of ongen in the transformed beta phase. Thus a more complete understanding of the embrittlement behaviour of ::ircaloy in ICCA
' type transients would be greatly aided by the availability cf methods for calculating the opgen distribution in the clad wall and knowledge of how the ongen ::cdified the mechanical response of the phases in 'he cladding.
A number of' finite difference computer codes for calculating the dithsien of exygen into Zircaloy during transient temperature oxidation are currently being used or developed in variout laboratories. Difficultice are enccuntered in modelling certein aspects of transient exidation sech as the extent of departure from equilibrium at phase bcundaries and also the =echanism of formation of rmlpha incursions into,and alpha precipitates within,the transfor=ed beta phase. Some comparisons between experimental oxidation exposures and calculations from an early version of SDffRAN published by S Malang have been made at SNL using tubes oxidised and quenched in the Gleeble. Previous Reports
- 1. PARSONS P D and . M W N. A Review of the Oxidatien Kinetics of Zirconium Alloys applicable to Ioss of Coolant Accidents. ND-R-7(S) 1977
- 2. PARSONS P D. A Review of the Oxygen Diffusion Coefficient in Alpha and Beta ::irealoys. TRG Report 2882(S) 1977 3 PARSCNS P D. Circalcy-2 Cladding Embrittlement in a lose of Coolant Accide.t. UKAEA Internal Repcrt Febmary 1978.
2
% EQUlVALENT METAL REACTED
, N ETA LLOG. VA C . SIM T R AN I FUSON
'I g.7 9.g 85 II 86 l 85 84 E 84 l 82 85 r
l 1200 - u
). ]?#f./3,t. -
fin. : ^ . e o - ..
- ~ . 9, pga 1100 U
E dg g !? { l
)
, 3. -
b ' . ). .vf '
~
X60 o .~ Nh'hIl l 1 1000 - I m i
~
-g. ;
T. l
$L*': ? l r ..
4 I r'- (* ,' X60 . 1 I I I 900 , 100 2 00 300 i TIME, s FIG.1. TIME-TEMPERATURE PRQFILE FORWATER QUENCHED TRANSIENT SPECIMENS l l a
,, n
~l i SNL TESTS INTACT FAILED 60 PIN SIZE VIRGIN METAL o
- 36 PIN SIZE VIRGIN METAL 0 60 PIN BLACK PRE 0XIDATION a &
60 PIN NODULAR PRECXIDATION O 5 60 PIN HIGH TEMP PRE 0XDN M 30 - i l' l 26 ~
$ TUBE 2 :S w*
c" a i W o 20 - l
$ 18 - - - - , a d
E _ 33 z O
-, 3 0 '
- 5 $
D c 0 < W o 10 - o o i "Q . 5 O o I i i i I I I 400 600 800 1000 1200 1400 1600 ZlRCALOY TEMPERATURE (PE AK) FIG.2. SNL EMBRITTLEMENT TESTS ,
* ..?~:
[ . . _ ~ _ _ . _ .. _. . _ _ . . _ _ _ . . . _
. . . _ . . - . _ _ _ . _ . ~ . -.
i .
. w . '. ', -
C 1000 1200 1400 1500
~
l ' ' l i ' I i
- SOURCE No 0F TESTS
.t l 0 ANL- INTACT QUENCH TEMP 8 ;- e ANL FAILED -
7 l 0 GE TTE INTACT a 24 E GE TTE FAILED = 1
; a FLECHT INTACT PEAK TEMP 22 T-lDAHO PBF FAILED "
2 O SNL 60 PIN SIZE IN TACT " - 21 30 - 4 SNL 60 PIN SIZE FAILED - 12
+ SNL 36 PIN SIZE
- 2 ,
i $ i . i O 25. j N
- i w
] TUBE 2 ; g e se 20 -
I t- 18 e' +
- e-I g
17 - - - - - - - - - - - - - - - a k -o - - - h 15 I . I a o O N CO z O w - 00 O
? %Q 5 to m
- O g
- C
}1h C 0 O d
- a. O 5 - o0 g o aa O a
O a OAu 0 i i Oc gg i t I en 7 1 F 1 500 1000 1500 2000 2500 3000 ZlRCALOY PEAK OR QUENCH TEMPERATURE, F FIG.3. SNL' AND US A EMBRITTLEMENT DATA _c .u., - , --m- - - .e - - ., - - - - , , , -- - - -av.,-,-.ru ,-,-my,.4-w-.wg.-.-.y y .-r , mv
.s 4
." - ATTACFDIENT 3 -
D007180(S) ZIRCALOY @AMmG EMBRITILEGE IN A LOSS OF COOLANT ACCIIENT '
? D Parsons SPRINGFIELDS NUC!2JLR LA30RL*0 RIES UKAEA PRESTON PR4'ORR ENGLAID N0YDGER 1978 i
To be presented at the US NRCSIITH WATER RIMCTOR SAFEFY RE3EARCH IT. !CRMATION MEETHIG N0YDGER 6-91978 ciITmSmG, warnan, USA i l j l i l I I
ZIRCAIOY CLADDI2iG DIBRITTLEENT IN A ICSS OP COOLANT ACCIIENT P D Parsons SPRINGFIELTE NUCLEAR LABORATORIES UEAEA PRE 5 TON PR4 ORR' ENGLAND j A3STRACT The data upon which the USA Nuclear Regulating Comission (NRC) base its
==v4=n= temperature and permissible oxidation criteria for precluding fragmentation of embrittled cladding are reviewed. Also described are, oxidation and quenching tests mainly on thin walled 60-pin fuel element.
sice Zircalcy-2 cladding. The total extent of oxidation has been correlated-with the ability to resist cr= *4ng during quenching in a loss of coolant accident (LOCA). Total oxidation is'a convenient parameter being easily calculated from the available kinetic data but the detailed understandir4 of the relation between clad embrittlement and oxygen uptake would be much enhanced by knowledge of the distribution of the diffused oxygen, and the role of the distribution, throughout the clad wall. A computer code SI2G'RAN-1 developed at GENL has been used te predict oxygen distribution in oxidised c1=M 45 aui the prediction is compared with some experimental resalta. s t
= -- - v-,- - . . ~ ., . ,
INTRODUCTION In the event of an LWR postulated less-of-coolant accident (LOCA), for the effective operation of the emergency core cooling system in removing the stored the=al energy, decay hea+:ag and the exothermic heat evolved from the steam-Zirealoy reaction, it is necessary for the fuel element to retain a coolable gecmetry. During a LOCA, cladding can deform by two mechanisms. Firstly, the rise in temperature and the loss of external coolant pressure can possibly balloon the cladding. Secondly, local embrittlement of the clad could fra6:ent the fuel element sheath with the fragments then blocking the cooling channels. To retain the efficiency of the emergency cooling system it is intended to define a criterion limiting the embrittlement and hence the possibility of fragmentation of the cladding. Since the cladding in the UK SGEWR was designed to preclude bursting, embrittlement studies have concentrated on embrittling due to oxidation on the external surface of the cladding.
'dhen cladding temperatures reach about 900 C the rate of the Zirealoy-steam reaction becomes significant. The effect of enhanced corrosjon is that the mechanical behavicur of the cladding is modified. If the time and temperature of exidation are sufficient, then the diffusien of oxygen into the metal produces a concentration gradient in the metal beneath the external onde film. A high oxygen concentration stabilises the & phase at high temperature thus creating a clad wall thickness of three distinct phases, oxide, ste.bilised er and the underlying S phase which, owing to a lower oxygen solubility, becomes enriched in oxygen to a lesser extent than the er phase. At ambient temperature the S phc.se has transformed to a Vidmanntntten fo m of a.
At low temperature the zircenia and high oxygen a phase are both ! brittle creating potential for fra6 mentation of the cladding during the l cooling phase of a LOCA, ; 1 The US Nuclear Regulatory Co= mission (NRC) has adopted fragmentation criteria based en exidation and quench testa on undeformed typical PWR and SWR cladding. Such cladding has wall thicknesses in the range 0.0780-0.0820 cm l and is thicker than the fuel sheathing used in the 36-pin or the 60-pin fcel : elements designed for the prototype Steam Generating Heavy Water Reacter at I Winfrith Heath. One of the US NRC criteria linits the oxidatien in a LOCA to an extent which effectively converts 17% of the wall thickness to zircenia. The validity of this criterion has not previously been evamined for the thinner walled cladding of UK designs. 1 To determine the relevant criterion in respect to thinner walled l cladding, exidation and quenching tests using Zirealoy-2 tubes of simila.: ' wall thickness to 60-pin fuel cladding have beer performed in a Gleeble rig. l THE USA DIERI'I'fLDENP CRITERIA l l The US NRC has adopted regulations dealing with the fectiveness of emergency core cooling systems. The acceptance criteria ( for emergency 1 l n .
1 l l core cooling systems for light water-cooled power reactors following a postulated range of breaks in reactor coolant systems, demand that a coolable geometry be maintained by the fuel elements during a IOCA. The criteria precluding embrittlement and subsequent fragmentation due to oxidation of i the cladding recommended by the regulatory staff are in precis, that ] (a) "the peak cladding temperatures the calculated m=v4 = = fuel I element cladding temperature shall not exceed 22000F (1204 40 0)" (b) "mmv4 = = cladding oxidation: the calculated total axidation of i the cladding shall nowhere exceed 0.17 times the total cladding 1 thickness before oxidation. A m=v4 =m cladding temperature limit was specified in order to preclude clad melting and limit energy release associated with the Zirealoy-steam exothermic reaction. The temperature limit also serves to preclude damage to cladding arising from eutectic formation at the 1 c9 9 tion of the spacer grid and fuel tube. Previously an interim criterion (27 specified a temperature limit of 12600 0 which is still regarded as adequate in respect 0 !,' of the foregoing considerations. A maximum temperature of 1204 was adopted solely in respect of limiting fragmentation of embrittled clam 43 The detailed arguments giving rise to the lowering of the mav4 mum temperature criterion are concerned with alleged anomalous loss of ductility resulting 0 from excursions above 1204 0 and are discussed in a later section. The basic data on which the USA 17)f oxidation embrittlement criterion is based are fr:m three principal sources and are briefly reviewed below. , THE ANL QUENCH TESTS soon et al } inductively heated simulated fuel rods in flowing steam and subsequently quenched the rods to simulate emergency cooling. The Zircaloy-2 tubing was 1.440 cm dia., 0.0787 cm wall thickness and centained UO2 pellets 1.270 cm dia. Thesteamflowwasmaintainedat2g/ min.and the rod heated to a predetermined temperature before being quenched by admitting room temperature water at 190 g/ min. either from the top or bottom. During the quench it was assumed that the heat less to the cooling water from a single pin was approximately equal to the heat loss of a fuel rod being quenched in a reactor core. Since it is difficult to simulate with precision the power input to the fuel element sheath during quenching, in some tests the power level was maintained constant throughout the quenching period whereas in other tests, power was terminated when quenching started. The hydrogen released during the oxidation reaction was measured and the average equivalent metal reacted was calculated assum4eg that 2 mol of hydrogen were liberated for one mol of oxygen consumed, i.e. 1 mol of zirconium reacted to foxm stoichiometric =irconia. Post-test metallographic measurements of the oxide film thickness were also made. Eighteen rods were testedatheatingratesfrom3toabout50C/sandthetemperaturecycle involved an almost li.3ar heating rate and a cooling rate dependent on when the power was turned off. Only the equivalent metal reacted and the zireenia thickness are quoted directly by ANL and the stabilised-r thickness and hence the combined oxide plus stabilised er phase thickness (!) have been calculated using the - 2 . _ . _ _ . . . _ _ _ _ _ _ _ _ . . . _ _ _ . _ _ . . .~ _ . . _ . .__- _
~
f b g. I a published' data, the results are listed in Table 2.. Also calculated are the -;
- values of the r===4ning prior S phase thickness as a fraction of the- original
~
and oxidised wall thMcnesses. The calculations of or thickness and C assume that the total oxygen uptake can be approximated to that contained in the oxide film plus stabilised a, i.e. that the oxygen diffused into_the $ phase can be neglected. A' relationship between the equivalent metal.reac andalphaphasethicknesshasbeenquotedbyScatena;je4,oxidethickness, \4) , 00
% equivalent cladding reacted (ER) = ,
x
'0xide thickness + - Alnha thickness' ***g)
L 1 54 - 1 54 a- - where w.t. is the wall thickness prior to oxidation, em and or is the ratio of the weight traction of oxygen in Zr0 'to'2 the aversge weight traction of oxygen in a pha zirconium. -The assumption is therefore also made that the oxygen concentration gradient is linear and bounded by the equilibrium values. If the average weight percentage of oxygen in the alpha phase is 0@ then,.since the weight percentage of oxygen in Zr02 is 25 97%, the **Pression can be written
~
100
% ER =
w.t.-_1 54 L 0xide thickness + Alrha thickness d; 25 97 ...(2) , Trowse(5)haspointedoutthatthisexpressionisanover-simplification and is more accurately written:
% ER = 100 70xide thicknes,s, , Aloha thicknessi v.t. u 1 54 1 32 a J-
,,, g) ,
Subsequent calculations of % ECR (or calculations of oxide and a thicknesses ' from quoted % ER's are derived from the expression given in equation (3). l
. This expression would give values of % ER slightly greater than those calculatedusingequation(1).
Thevaluesof(Dt)b listed in Table 1 are estimated from the calculated ) value of C a9d)the relationship between C and (Dt)? quoted by Hobson and Rittenhouse,(6 , 1 ( = 1 529 (Dt)E + 0.00168 ... (4) ! 1 THE FLEEP TETS (RTLL LENGTH TrDEiiumwi C00 LIT, N TRANSFER TETS) A series of tests called Zr-3, Zr-4 and Zr-5 to investigate the thermal response and cladding performance of 3WR fuel bundles raised to high temperatures and cooled by a simulated emergency core cooling system, were performed by General Electric. Duncan and Leonard (7) conducted tests with .
- full. scale Zircaloy clad cimulated 3WR bundles. The rods were heated by internally wound molybdenum filaments centred by filling the tube with vibro-ccmpacted alumina. The heater rods (Zircaloy-2) were 1 427 cm dia.
and 0.0813 cm wall thickness, 49 rods were held in a 7 x 7 array. The bundles were filled with steam and heated using a constant heat-up power of_240 kW until the bundle reached a predetazmined temperature. The mmHmm 3 u
. ,. .u--~... , .. , - .. - . . - . . . - . . - - - - . - - . . , - . - . - . . . - _ . . ~ . - - - - . . - - - - - - - - -
temperature in the bundle occurred at the mid-plane for run 6-4 and was I u 139000. Spray flow was initiated as the power was altered to simulate a decay transient. Meta 11ographic measurements on sectioned rod samples were made at a limited number of locations on the heater rods exposed in Run Zr-4 which had remained intact. The data from these rods are semnrised in Table 2. The values quoted for the percentage metal reacted are I calculated from equation (3) usipg of oxide,and a thicknesses published by Duncan and Leonard.\ f)the valuesThe values of (Dt)?quoted are the by Duncan and Leonard calculated using the caethod given by Mallett. There is some discrepancy between these values and those calculated using equation (4). I THE GE TTE TEST (TIME. TD4PERATURE. ENVIR0tMENIO The General Electric (4} company performed tests in which 12 inch lengths l of a 3 ft 1cng Zircaloy-2 tube were heated in steam and subsequently quenched I by bottom flooding. The Zircaloy-2 tubirs was typical of the 1967 and 1969 product line cladding and was 1.267 cm dia. and 0.0813 cm wall thickness. The tubes were heated by internally wound molybdenu:r. filaments packed in alumina. Steam was passed over the tubes at a constant rate of 0.17 g/s and the power supplied to the heaters was increased until the desired temperature was reached. The test temperature was maintained for a pre-determined time and cold water (at 2200) was admitted through the bottom of thetestchamberatarateof2in/s. The power was maintained to the tube until the temperature had dropped to about 32000. The tests were conducted in two series (TTE-1 and TTE-2) and after erposure each of the tubes was enmined metallographically and subjected to mechanical property tests. Only tube 1-7 in this series was cracked after testing. The next most severely oxidised tube was test 2-5-M which remainad intact after quenching. The metallographic measuremento gf oxide thickness and ar-phase thickness are published in tabular fom by gel 4/ and are reproduced in Table 3 S me tubes reacted with the alumina at the inside surface and whilst no exide was fo=ed as a result cf this reaction, some stabilised e-phase was fomed. The estimated equivalent metal reacted is published by GE separately in graphical fem and is plotted as a function of quench temperature. Since the temperatures to which the metallegraphic data pertain are not given, it is not possible to estimate directly the equivalent metal reacted from the metallographic measuram9nts. Table 3 shows the equivalent metal reacted as calculatedbyScatena(4/andpublishedgraphically. The order in which the data are listed in Table 3 are decreasing order of total extent of oxidation and are not in the same order as the metailographic measurements in the Jeft hand side of the table. As calculated using equation (1) the values of % ECR will be slightly lew. However, since the error does not materially affect the conclusions the values quoted by Scatena have been retained. SUMMAFY OF THE ANL. FLECHT AND GE DGRIT'?IIMENT TE?S One of the conclusions reached by Hesson on the basis of the ANL quench tests, was that when the equivalent reaction expressed as the percentage wall thickness of metal which reacts with the total or/ gen uptake to fom ZrC2 is a 18$ then the cladding bec aep 7ery brittle. The ANL tests were also considered by Hench and Liffen6:en\9/ who noticed that clad oxidation in tems of the total extent of reaction could be used parametrically as a measure of fragmentation propensity. They suegest that 4
z the ANL data could be used to define 15 equivalent metal reacted as a fragentation limit. The results obtained from the ANL, Flecht and GE-TTE tests are plotted in Fig.1 as the percentage equivalent metal reacted as a function of the quench temperature, (except for the Flecht tests which are plotted as a function of m*= temperature attained during the test). It is readily apparent that a design criterion can be established frem such a presentation of results defining the ability to remain intact after a LOCA and quench. The ability to remain intact after quenching is exhibited by tubes that have not been oxidieed to an extent equivalent to converting 15 of the wall thickness to stoichiometric ::irconia. Table 4 sum = arises the results derived from the ANL, Flecht and GE-TTE tests which are closest to the 15 criterion. The Table lists values of other parameters in addition to the percentage equivalent metal reacted, these. include the oxide thickness and the oxide thicknesses as a fraction of the original wall thicknesp, the oxide + cr-phase thickr.eas (;), the fraction of S-phase (Fw) and (Dt)2 Clearly the only parameter which consistently segregates cracked tubes from intact tubes is the 15 equivalent metal reacted parameter. IDAE0 p0VER BURST FACILITY As part of the US NRC reactor safety research progra==e, both irradiated end unirradiated fuel rods 1; ave been forced into film boiling under controlled test conditions. U O) The tests have been conducted in the Power Burst Facility at the Idaho National Engineering Laboratory. The fuel rods were Zircaloy clad containir4 UO2 pellets enriched to between 9 5-20 wt.% U2 35 and were tested singly or as a group of four, each element having a coolant shroud. The tests have been performed at fuel element powers in the range 51-a0 w /m and have resulted in film boiling occurring for periods up to 10 min. Fuel sheath temperatures up to 1400 C have been measured and the Zircaloy sheath has been oxidised by metal-water reaction en the cutside and metal-UO2 reaction following cladding collapse on the inside. The combined effects of internal and external cladding oxidation severely embrittled the cladding and have resulted in failure of some locations of fuel rod a few minutes following reactor shut down. As yet, only prelimin g results are available cence ming the myitudes of the oxide film and stabilised or fc=ed during the oxidation exposure. These values for one rod IE-ST-1, initial vall thickness 0.0509 cm, at the ends of a frag ented section are given in Table 5 The temperature of the claadi g was esti=ated usir4 the combined thickness of the ext 9rpal oxide film and the stabilised a phase and the methed given by Mallet. Wl The percentage equivalent metal reacted has been calculated for these twopositionsonthisfuelredusingtheapproximationofequation(g). At the fragented section the calculation gives values of 19.4% at 1346 C and 10 5% at 13360C. Rod IF,-ST-1 experienced 4.3 min continuous film boiling and molten fuel was present in the rod during the test. However, red failure only occurred after shut dcwn by fracture of embrittled cladding. The phase widths used in the calculation were obtained from small scale graphs ard are therefore only apprcIlmations. Ecwever, both results are consistent with the previcus US'frapertatien data upon which the 15 criterion is based, as is sh;wn in Fig. 1 even thcush significant ex11ation l l l
occurred en the inside, the sheaths were not cooled by cold water quenching and the wall thickness was much less than that of tubes exposed in the AE, Flecht and GE tests being 0.0509 cm Further more accurate data from this seriesoftestsinthePowerBur9thacilityshouldbecomeavailableasPIE is completed on other fuel rods.(111 OXII)ATION ANI) QUENCHING OF ZIRCALOY-2 TITBES AT SE Prior to oxidation and quench tests being carried out at SE, the available fragmentation data were related to cladding with wall thickness in the range 0.0787-0.0813 cm. Zircaloy-2 claMing for the 36-pin and the 60-pin Vinfrith fuel element have wall thickness in the range 0.0641-0.0768 cm and 0.0597-0.0697 cm respectively, i.e. thinner than tubes used in the AE, Flecht or GE tests, particularly so for the 60-pin element. To assess the affect of a thinner wall, apropos the 17% fragmentation criterien, a number of tubes with the 60-pin dimensions were oxidised, quenched and evaluated with respect to failure and total oxygen content after exposure. A small number of 36-pin size tubes were also included. In the work described below, fuel tube oxidation has been carried cut only on undeformed tubes so that particularly for virgin metal tubes, the original wall thickness can be used as a basis for calculating the equivalent metal reacted. Forpre-oxidised(butundeformed) tubes,theamountofpre-oxidation is included in the estimate of equivalent metal reacted. For in-reactor LOCA type oxidation where wall thinning T.ay occur before the oncet of significant defined oxidation, in 10 code the(wgli CFR 50-46 1 1 thickness prior to oxidation is as EXPERIMENTAL The Gleeble plastcmeter rig consists of a high speed time-temperature controller coupled with a high speed tensile tasting device. For the oxidation and quenching of Zircaloy tubes the specimen clamps were not employed in the straining mode, but whilst tightly gripping the specimen to . maintain good electrical conductivity for resistance heating rem 4n4eg free to slide in 0-rings in the entrances to the specimen chamber. A stainless steel specimen chamber has been adapted to allow specimens to be gripped in the Gleeble specimen clamps within the chamber and heated whilst steam is fed to the chamber. The steam is supplied from a 3 W boiler filled with tap water and enters the chamber through a diffuser. The steam supply rate is sufficient to precit de the metal steam reaction being gas phase diffusion limited. To si=ui .te the rapid cooling of the c emergency cooling water in a I4CA, a spray of cold tap water was rade to impinge on the specimen at the agpropriate time. The spray system consisted of three tubes arranged at = 120 around the specimen at a distance of = 3-4 cm, the tubes having numerous holes thrcush which water was directed at the epecimen over about the centre 6 cm of the specimen. The specimen chamber is fitted with a glaso cover so that the specimen can be observed during the oxidation. The schematic arrangement of the Gleeble rig is shown in Fig. 2 and the specimen oxidation chamber is shown in Fig. 3 The specimen temperature was monitored and controlled using two 0.010 in Pt/Pt-Rh thermoccuple vires each welded to a 50 pm thick 2 =m 6
square coupon of niobium, the coupons being welded = 1 cm apart, to the outside at the centre of the 18 cm long tube. The temperature at the weld between the thermoccuple leads and the niobium coupons only appror6tes to the temperature of the parent Zircalcy-2 metal. The temperature at the veld is locally depressed since (a) the junction is raised by about 50 pm frem the tube surface and (b) heat is conducted away from the junction down the thermocouple leads. The only way to eliminate the error completely is to maintain the temperature of the surrounrHngs at exactly the same temperature as the parent metal. The bore of the tube is an approximation to these evaditions and to calibrate the ther=occuple on the outside it was compared to the temperature obtained frcm a like themoccuple welded to the inside of the tube at the centre position. In this way all outside themoccuple settings in the Gleeble were adjusted to take account of the local heat loss. SPECIMENS Zircaley-2 tuber 180 mm long were used as specimens. The outside of the tubular specimen was held in water-cooled screwed-down cla=ps and prior to fixing the cla=ps, stainless steel end plugs fitted with G-rings were inserted into the end of the tube. The end plugs were connected to a line argon supply. The 60-pin size tubing was stress-relieved, 0.C660 cm v.t. 1.2200 cm o.d. The mean initial concentration of oxygen in the as-received tubing was 0.1136 wt.%. The 36-pin size tubing was annealed, 0.0695 cm v.t. and 1.6290 cm o.d. with mean initial oxygen concentration 0.1342 wt.%. A total of thirty-three 60-pin size tubes were oxidised and quenched, twenty four of these were in the virgin metal condition at the start of the oxidation exposuze, whilst nine tubes were in ene of three types of pre-oxidised condition. The two 36-pin size tubes tested were both virgin metal tubes. All the virgin metal tubes were degreased in acetone prior to loaC ng in the Gleeble rig. In the event of a postulated LOCA, the cladM ng throughout the whole of the reactor core is likely to be in various states cf oxidation depenMng on the irradiation hister of the particular fuel element and the water chemistry of the reacter. Fo determine the influence of prior in-service oxidation en the subsequent high temperature oxidation and possible embrittlement of einAM ng in a LOCA, seme of the tubes tested in the Gleeble were given pre-oxidation treatments intended to s h ute (a) normal black oxide of thickness corresponMng to approximately end of life (= 8 pm) and (b) nodular type cerrosion. Two tubes were also tested having been pre-oxidised to an undetermined extent, for a short time at high temperature (> 100000) in steam in the Gleeble rig cwing to a malfanctien of the thermoccuple weld. The normal black oxide was Sc:med on three tubes by exposing both the outside and inside to atmospheric pressure uteam for 130 h at 500 0.0 This treatment gave a weight gain of 1.205 mg/cm2 which is equivalent to 0.0009 cm I of oxide film on the outside and the inside or is equivalent to 1.6% metal , reacted. The nodular oxide was formed en four tubes by first expcsing both : the cutside and inside to atmospheric pressure steam at 500 C for 2d h. l This produced a nomal black exide film of = 3 pm en the surfaces. This l oxide was abraded off the outside and the tube was then expcsed to 1CC'O lb/in2 i steam at 500-550 C for 24 h. The nedules were not evenly distributed over l 1 I i ..
I the tube surface, tending to fem more nu=erously in the regions of most severe aorasion and also to be associated with the deepest secre marks. htular specimens pre-exidised in this . tanner have a think black exide film (*> 3 m) on the inside and a thick (50-170 pm) nodular oxide product on the cutside with variations in the amount of nodular corrosion along a tube length. In some instanew a small amount of nodular type oxide formed on parts of the inside. The oxygen uptake, exile thicknesses and ths percentage equivalent 'aetal reacted in the pra-oxidation treatment are listed for the feur pre-oxidised tubes in Tule 6. The figures quoted assume that the roughening did not significantly affect the tube wall thick-ness and that the nod'11ar type corrosion is evenly distributed along the j tube length, all on the cutside. TIME-TmpFRV"URE pARMffTERS OF OXIDATION D:p0SURE The majority of virgin metal tubes were exposed to steam isother= ally in the Gleeble at 1400 C and 13000 C up to times of 900s. The time-temperature parameters for all tha virgin metal tubes tested are listed in Table S. At 140000, tubes were exposed for 200 s (duplicate tests), 3C0 s (duplicate tests), 400 s (ono 60-pin and one 36-pin) and 500 s (single exposure), 550 s (cne 60-pin and one 36-pin) and 900 s (fcur exposures). Specimens 24 and 26 were first oxidised isother= ally at high temperatures (1300-14000 0) to an extent calculated to be less than or equal to 17% equivalent metal reacted and then the temperature was quickly lowered to 900 0 0 and the specimen allowed to dwell for an additional 600 s. It has been reported 'yo Graber(12) that cooling slowly thrcugh or dwelling in the er + S phase region (widened with increased oxygen concentration) promoted the pescipitation of platelets of ::irconia adjacent to needle-like Vid=anstatten grains of prior S phase, resulting in an extremely brittle structure. Tutes ?4 and 26 were held at 9000 0 to detemine the influence of a holding period en the metallurgical structure and subsequent embrittlement of the oxidised tube. l Table 7 also lists the virgin metal tubes exposed to transient tempera- , ture-time parameters. Duplicate specimens were given transient exposure l corresponding to the temperature-time parameters as shown by curves I, II l and III shown in Fig. 4. The curves peak at 1310-1340 00 between 60-13Cs and l 1 the cold water spray was initiated at 300 a as the power was te=inated. Single cpecimens were also exposed to the transient-temperature time parameters shown in Fig. 5 These differ from the other transients in being less severe with respect to total oxidation: they also peak in temperature earlier in the accident and, after falling frem the peak, the terperature rises again before being cooled by teminating the power at 300 s. These j soecimens were not water quenched but cooled in steam down to s 850 U C in 5 3 s. Table 3 lists the time-temperature parameters of exposure for the i pre-oxidised specimens. Of the three specimens pre-exidised with nc= al black oxide at 1cw temperarure, one was exposed isothe= ally at 1300cc for i 550 s (specimen 4) and two were exposed to high temperature exidation at 0 l 14000C or 1300 C before being held at 900 C as discussed previcusly for virgin metal opecimens. All the tubes pre-exidised with nodular type exide were exposed isothe m ally at 1400 C or 1300cc for up to 300 s. Of the two specimens pre-oxidised at high temperature (> 1C00 0), 0 One was exposed l 1sothemally and one was allcwed to dwell for 600 s at CCCC C after an I 1 initial exidation at 1400 00 for 200 s. S
=.- - . - - - . - . - -, - - .
EXPERIMEW AL D E ERMINATION OF EXTEE OF OXIDATION The or/ gen uptake during oxidation was measured using two methods. A
, direct chemical analysis for oxygen using staMmed vacuum fusion techniques '
at fusion temperatures between 2500 sna 2200 c was used for all specimens except transients IV and V. For comparison and cross reference, the thick-nesses of oxide and stabilised a-phase were seasured and used to calculate, by equation (3) and its associated assumpticas, the total oxygen uptake. Laboratory apparatus for routine vacuum fusion analysis of gases in metals is designed to analyse relatively small concentrations of gas. The specimens oxidised in the Gleetle contain an order of magM tude scre oxygen than as-received or lightly oxidised Zircaloy. The relatively large concentrations of oxygen in the oxidised samples required that very small
, samples be taken for analysis. The sample weight analysed was typically-about 100-200 mg which corresponds to a section through the complete wall section having an area on the ou % ide of = 2 mm 2 . The oxygon is not uniformly distributed through the wall section, there being a thick (up to %190 wm) oxide layer on the outside, and a high oxygen concentration (mean = 4 7 wt.%)
in the a-phase.being present over a thickness up to = 200 pm. Thus to obtain an accurate analysis of total oxygen with respect to the complete wall thick-ness of the tube it is'important to analyse a parallel-sided section of complete wall thickness. An additional factor increasing the difficulty of ; sampling was that oxide often spalled from a quenched tube after cooling. It is reported in a later section that the oxide film has a dual layer structure and the outer layer of oxide usually spalled from the tube, a tendency which increased with extent of oxidation. In a severely oxidised c tube the whole film spalled from some areas of the tube exposing bright a-Zircaloy. The spalled oxide films were recovered, analysed for osygen separately and found to be stoichiometric oxide within the limits of the analysis. The thickness of the spalled oxide was also measured microscopically to enable oxygen uptake to be estimated using equatior (3). The values of total oxygen obtained by chemical analysis are listed in Table 9 for all the tubos tested and all types of sample analysed. hbe P and Q were not analysed chemically and only values obtained from metallography are available for these specimens and these are listed in Table 10. Values are also listed for the percentage equivalent metal reacted calculated on the appropriate wall thickness prior to oxidation. The oxygen content of an oxidised tube can be estimated approximately if the oxide and ar-phase thicknesses are known, equation (3). In specimens which are more severely oxidised, the formation of a-incursions and a-precipitates precludes an easily obtainable estimate of the equivalent a-phase thickness. Thus specimens for which oxygen analyses are quoted in r l Tables 7,8and10,fallintooneoffourcategories,(a)non-incursed2/S bonMaries, (b) slightly incursed a/$ bonMaries, (c) significant incursions and precipitates of ar-phase, incursions and precipitation. For specimens of category (a)(d) very largeofaequation (3) is a good the application
- I approximation since a reliable esti:nate of a-thickness can be obtained.
Fig. 6(a) shows the section through tube 21 exposed for 300 s at 1300 CC and shows the thickness of the stabilised a-phase to be very uniform. Incursions of stabilised a within the prior $ have not formed in this specimen. For specimens in categor/ (b), typically tube 18 exposed for 300 e at 1.1000 0, an example of which is shown in Fig. 6(b), a nominal 5% has been added to the 9
I predominant 2-phase thickness. All the transient exposures shown in Fig. 4 are of this type. For categon (c) a visually estimated and therefore very approximate thickness for the mean 2-phase is quoted since the boundary is ; extremely uneven. This 0 is shown in Fig. 6(c) which is a section through ' tube E exposed at 1300 0 for 900 s. For specimens of category (d) no attempt has been made to estimate the extent of oxidation using this method, apart from t- 1 where the chemical analysis is in seme doubt. Tube 2, when exa=1. wallegraphically, was observed to have a very heterogenecus distriumn of oxide and in some areas, using equation (3), figures for equivalent metal reacted were in excess of 15 The values obtained for oxygen content converted to percentage equivalent metal reacted are given in Tables 7 and 8 as are the individual measurements of inner oxide, outer oxide and a-phase thickness. CORRELATION OF OXIDATION WITH TUBE FAILURE At the end of an oxidat4ep exposure the tubos were cooled quickly by a cold water spray. Scatena(4) has m 1ysed the ferees experienced by cladding during a LOCA. The forces censidered were these arising at varicus times through a LOCA ard the analysis showed that peak stresses varying from36300-38000lb/in occurred due to quenching for tubing oxidised in the range 2 5 to 1 5 equivalent =etal reacted. The calculation of the thermal quench stresses took into account the fact that the maxi =um stresses would not be generated until the vapour film (film boiling regime) collapsed and nucleate boiling ocuurred, resulting in a rapid quench. The temperature at which the vapour film collapses is known as the Leidenfrost point and is quoted by Scatena as 371-538 C. In the calculations, a constant cladding temperatures of 538oC was assumed. The peak quenching stress is not a strong function of either the extent of oxidation or of the initial clad & g wall thickness in the range 0.0610-0.0940 cm. ~ The abil1 y of an oxidised tube to remain intact and not crack or fragment in resisting the thermal stresses generated during a LOCA quench was regarded as an adequate preperty of the tube in defining a safety criterion for retaining fission products and a coolable geometry during a LOCA. The cracking of an oxidised tube was readily apparent at the end of a test and some examples of failed and intact tubes are shown in Fig. 7 The tubes shown in Figs 7(a) and 7(b) were initially virgin metal and were exposed isothemally for 300 s at 1400 C (tube 18) and for 550 s at 1300 C (tube 22) respectively. Tube 18 failed with a circumferential crack and tube 22 remained intact. The outer layer of oxide spalled ccpicusly from both tubes. The outer oxide layer was black and lustrous whilst the inner adherent exide, when exposed, exhibited a matt bitch appearance. The inner oxide has also spalled from tube 18 exposing bright stabilised a metal. Figs 7(c) and 7(d) show tubes which were pre-oxidised with as S pm of black lustrous oxide. Tube 4 in Fig. 7(c) remained intact after 550 s at 1300 C and seme outer oxide spelled frem the tube. Tube 6 in Fig. 7(d) failed with numerous cracks after a dual exposure at 1400 0 C for 200 s ari 10000 C for a further 600 s. Tubes 1 and 3 in Fiss 7(e) and 7(f) were pre-exidised with nedular type oxide ud Fig. 7(e) clearly shows the uneven distribution of the oxide. Tube 1 in hg. 7 failed and retained all the oxide; tube 3 in F1 . 7 remained intact and some of the cuter oxide was spalled. Fig. 7(g) shows tube 2 which was pre-oxidised with nodular exide to an extent equivalent to 10
17 4% metal reacted. This tube was further exposed for 60 s at 1300 C'and
. survived the quench. Meta 11ographic' enmination of parts of this tube suggest an equivalent metal reacted greater than 15 in some sections. -
.; Although cracking was readily apparent, a pressure of argon was applied to the. bore of the tube at the'end of the test to detect the presence'of very , y fine cracks. However, no fine cracking vas observed: a tube was either ! obviously cracked or.it remainad intact. -
. Fig. 8 shows the equivalent metal reacted (percentage of pre-oxidised .j wall thickness) as a function of the peak temperature of the exposure for all the thirty five tubes tested in the Gleeble. The equivalent metal reacted.
plott.3d in Fig. 8 is the most pessimistic value obtained for a tube where more than one chemical analysis was performed, i.e. for failed tubes the lowest chemical analysis was plotted whilst for tubes which remained intact the highest chemical analysis was plotted.; For tubes initially virgin metal, or lightly oxidised, the equivalent metal reacted can be used parametrically to divide failed from intact tubes,.in the same way.as previcusly described for the USA data. The limit of oxidation for survival of such tubes tested at SNL is = 13% equivalent metal reacted, which is in very good agreement with the USA data and-it is clear that' an oxidation limit of 15 equivalent metal-for non-fragmentation of'a tube is'compatibio with both sets of data irrespective of the initial wall thickness of the tubing within the limits tested. i.e. 0.0660-0.0813 cm wall thickness (both sets of data are shown in Fig. 9) . o The nodularly pre-oxidised tube 2, snrvived the quench intact. The pre-oxidation of this tube was' equivalent to 1 $ metal reacted (distributed heterogeneouslywithrespecttoaxial'lengthoutsideandinside)~andthe
, total extent of oxidation after Gleeble exposure was estimated to be about 20-24% ECR 1.e. in excess of the 18% ECR criterion. This observation does not conflict with the generalisation that failure does not occur unless the 3CR exceeds 15, i.e. any. tube oxidised to less than 1$ ECR will not fragment but under certain circumstances (e.g. heavy nodular pre-oxidation) y tubes oxidised to greater than 15 ECR may not fragment.
FRACTOGRAPFf AND OXIDE FIIM STRUCTURE
-The high oxygen content, brittle, stabilised a-phase.was extensively cracked in all the tubes exposed in the Gleeble, the cracks being numerous enough to allow small volumes of a-phase to be lost during specimen prepara-tion. For tubes oxidised to the lowest levels, i.e. = 8-9% equivalent metal reacted (= 3-4% for steam cooled transients IV and V) the e-phase was of uniform thickness with a large blocky grain structure and the cracks appeared to be wholly confined to the a-phase width. This is illustrated in Fig.10(a) which shows the e-phase in tube 9, a tube exposed to transient-temperaturetimeoxidationandoxidisedto}10%equivalentmetalreacted.
For tubes in which the oxidation has been severe enough to form incursiens of a, the cracks are able to propagate along the brittle a-phase incursions as in tube 23 which is illustrated in Fig. 10(b). In this particular tube the cracks did not propagate to the inside and the tube remained intact. Fig. 10(c) shows a scanv %g electron mieregraph of the oxide and stabilised a-phases at the fracture surface of a tube which cracked after an exposure of 400 s at 140000 (tube 17). The a-phase contains many brittle cracks which are not restricted to the radial planes.- The oxide 11'
film is also cracked nomal to the columnar grain structure although this was an infrequent observation. Fig. 10(d) is a scanning electron. micrograph of the surface of tube 5 where all the oxide had spalled exposing bright stabilised a metal and the brittle crackd v is readily apparent as is the imprint of the oxide columnar grain strucL.re. The oxide film fomed exhibited a dual layer structure which is delineatedinopticalmicrographsbyadarklineparalleltotheoxide/o boundary and near the middle of the oxide film. This line is clearly visible in Figs 6(a) and 6(c), and in all the 60-pin size tubes oxidised whether isothemally or transiently the outer oxide layer was larger, the outer-to-inner ratio varying from about 1.1-1 5 The two 36-pin size tubes oxidised had an inner oxide film marginally bigger than the outer film. The radial lines in the optical nicrographs of the oxide structure are shown more clearly in the scanning electron micrograph in Fig.10(c) and illustrate the columnar structure of the oxide. The surfaces at the boundary of the outer and inner oxide layers were en mined in the scanning electron microscope and an mnlysis was made using the energy dispersive X-ray analyser. Compared to the outer surface of the outer layer of oxide there was a significant concentration of tin at the boundary of the two oxide layers. A specimen mounted in section through the complete wall thickness was also e m ired in an electron microprobe and a concentration of tin was readily apparent at the interface between the inner and er oxides. This observation has been reported elsewhere.U 3,14,15) 31ederman has made the added observation that the tin is in particulate form and t t the size of the particles increases with oxidation temperature. Yurek et al 14/ have made a detailed examination of the oxide scales formed by isothermal oxidation of Zircalcy-4 at temperatures between 1000-130C%. The authors report that the oxide scale consisted of a =etallic phase rich in tin within a matrix of zirconia. The tin rich phase was concentrated in a line of particles near the centre of the scale. The cuter oxide scale contained little tin but the inner oxide contained a fine dispersion of the tin rich particles probably located at the grain beundaries of the colen'- Zr02 . The authors tentatively believe that the metal phase exists at the reaction terperature and that if the tin centent were high enough the phase could exist in liquid form. It is also believed that the particles move tewards the alloy, with respect to the stean/ scale surface, as the reaction proceeds. The structure of the oxide film has also been examined at SNL using a stnMa-i 200 kV transmission eleed.ron microsecpe and also an ZPEA-I7 electron microscope fitted with an energy dispersive X-ray analyser. Samples of spalled oxide film, both inner and outer were thinned to electron trans-parency using an ion beam thinning machine. Fig. 11 shews the structure of the spalled oxide from tube 18 and is typical of the outer and inner oxides in that the zirconia has a twinned monoclinic crystal structure (at room temperature) and precipitates, rainly at the grain boundaries. The grain boundaries are also cracked which may be the result of the ther=al shock during the quench. Attempts to determine the crystal structure of the precipitate phase using electren diffraction techniques have not been successful. Ecwever, the precipitates have been analysed qualitatively as individual particles 12
l l l in D!MA-IV and zirconium and tin were identified as the major elements present in the precipitates and were present in the ratio of about two zirconium to one tine by weight. However, the zirconium in the matrix probably contributed greatly to the zizeenium analysis. CA1CUIATION OF OXYGEN IISTRI30 TION AND TOTAL OrfGEN UPTAKE The effect of oxidation on Zircaloy is to embrittle the material and, i in general terms, the greater the extent of oxidation the greater the l embrittlement. The extent of oxidation per se can be used as a measure of I the threshold of embrittlement severe enough to cause cracking of the clad I in a LOCA. The propensity to embrittle depends upon changes in mechanical ph;,arties in the clad. These changes are brought about by the for=ation of a thick oxide film, a thickness of stabilised a-phase, (possibly with incursions into the prior S-phase) and an increase in the average oxygen content of the residual S-phase. As oxidation proceeds en the outside, the oxide and a boundaries move towards the inside of the clad and since l Zirealo) clad is relatively thin (< 0.0770 cm) the oxygen centent of the progressively decreasing thickness of S-phase, is increased. The =echanical i properties of the oxide and high oxygen stabilised 2-phase at temperatr e l are not yet well known, but both exide and e-phase are brittle at low j temperatures. The S-phase saturates with oxygen at a much lower level than I the 2-phase and the approach to saturation in the S-phase ::.ay ma:kedly affect the preperties of the transfor=ed B-phase. Thus a more cceplete understanding of the mechanical properties of the clad would be aided by i knowledge of the detailed distribution of oxygen through the clad wall. I l The high tegp ratyre oxidation of Zircaloy has been censidered by ( various authors. '17) A model based en the ideal diffusion of oxygen ' through the three phases oxide, alpha and beta Zircaloy has been used for the calculation of the distribution of oxygen in oxidised Zirealey by Malang(18) and has been assembled into a cceputer code SIERAN-1. The code solves the simultaneous transport equation for both mass and heat flow for ene-dimensional cultiphase, moving beundary, transient-temperatr e transport. The distributien of oxygen is calculated oy applying the basic laws of l diffusion to the three different layers in an exidised material. The ideal i diffusien model requires uniferm grew-Q of all phases and invekes the basic assumption that the=::odyne:c::1c equilibria are maintained at all boundaries ) I whatever their rate of movement or rate of temperature changes if super- - saturation of a phase occurs owir4 to the temperature charges, the excess l oxygen is dissipated by diffusion and uniform phase movement, not by precipitation. It is also assumed that each mass transport equation is characterised by a single kinetic paraceter, the coefficient of diffusien, which cui be a function of the particular phase, temperatre md if necessary oxygen concentratien. Thus by knowing the oxvgen diffusien coefficient in each of the three phcees and the equilibrium eeneentrations of orvgen at the oxide, and a bouncaries, the oxygen profiles and the movement of the boundaries can be calculated. The integration of the profile within each phase gives the total oxygen in each phase, the sum being a measure of total exygen uptake. l Malang and Schanz (19) have published a descriptien and preldr.inary verification of SIE 9A.W1. g Isothe:=al high temperatr e steam cxidation of Zircalcy-4 tubing (20/ has been used to cenpare the measured weight gain and 13
.-- _.- . - . - - - - - . . _ . . ~- ..-
the metallegraphically mee.sured layer thicknesses of oxide and a-phase to the code predictions. Data from double sided oxidation in the range 130000 and 2-15 min, which also shows good agreement with ORNL data,(90g- U/ t compares quite well with code predictions. The values of the diffusion. coefficients used for the three phases are not given but the authors indicate that further work is in progress to optimise the diffusion data. Transient-temperaturetimeoxidationparameters(22)havealsobeenused to obtain code predictions and compared to the experimental results the prediction of the SDERAN-1 code were reported to fit quite well. The opde has been used at SNL taking the listings as published by Malang.(18) A systematic programme, validating the code against a comprehen-sive data base has not been attempted, but comparisons of code predictions have been made with a number of the Zircalcy-2 exposures carried out in the Gleeble for embrittlement testing. Thicknesses of oxide and a-phases and total oxygen contents have been calculated with the code for a number of specimens where significant a incursions and precipitation did not occur. The values of the oxygen diffusion coefficients used in the code (or any similar code) for the three phases are of crucial i=pertance. The oxygen diffusion coefficisnt in E-Zircaloy his been intensiv9 17 at ORNL and the values obtained formed the basis of a review which(22) researched
. yielded a temperature dependence of the oxygen diffusien coefficient in S-Zircaloy ofs D,=0.0286exp(-28423M) ...(5)
The review also recommended an expression for the temperature dependence of the exygen diffusion coefficient in 2-Zircalcy of: Da =387exp(-50700M) ...(6) The above values in equations (5) and (6) have been used as ir,m2t to SDERAN-1 along with a value for D oxide publishedbyCathcartd4)atORNL. Doxide = 0.137 exp (- 34700/P2) ... (7) ORNL have used the accurately measured value of Dg together with experi-mentally obtained values of oxide and 2-phase thicknesses to back calculate the values of Dox and Ly which need to be used in SDERAN-1 to fit experi-mental results. The expressien obtained for Doxide is shown in equation (7) whilst the expression obtained for Dyusing this method ist D =41exp(-512005) ... (a) The values of Dy obtained frem equations (6) and (S) are in ver/ close agreement. Table 10 lists the tubes exposed in the Gleeble for which predictions have been made using SDERAN-1. As well as the eight tubes exposed to transient-temperature time profiles, isothermal 17 exposed tubes at 140000 and 13000C are included. The Table ecmpares the directly measured oxide and a-phase thicknesses with those predicted by SDERAN-1. The Table shews the oxide and alpha phase thicknesses calculated by the code at (a) the end of the isothermal ex.idatien just prior to the ap.11 cation of tne 14 i , - , - , , , _ . _ - . _ , , , _ , _ , . . , , , , . , - . . - . , _ _ . , .- ~ , . , , - . , , . .. . . . . . . , , - _ . ~ . , - . , , .
^
quench and (b) the end ef the quench. In the case of transient temperature oxidation (a) corresponds to the temperature of the transient at 300 e, bely the time at which power is te=inated and the quench water activated. The oxide and alpha thicknessee predicted by the code at a time and temperature just prior to rapid cooling agree quite well with the experi-mentally observed values. However, after rapid cooling, the alpha phase is consistently overpredicted by the code. This is believed to be due to the i adherence to equilibrium .alues of oxygen concentration at phase boundaries l used in the early model for SIyTRAN-1 and =odifications to this model are ! currently under consideration.s25) The Table also lists values of the percentage equivalent metal reacted obtained using four different methods, two experimentally observed values and two by calculation. The values obtained by s m W the integrals of the SIMPRAN-1 oxygen profiles in the exide, e and S phases is compared to the experimentally detemined values using vacuum fusien and the agreement is very close. The values obtained by =etallegraphic measurement and equation (3) are lower than the others as expected, since the oxygen in the S-phase is neglected. Other codes are oxygen profile concentration e.g. Zoro(avgi1able or bai- developed fcr 26/andOxwex.(2D DISCUSSION AN DGRI'1"LDGTP CRITERION The oxidation and quenching tests on 60-pin sized cladding pemit a design limit to be set en the total oxidation of cladding which will preclude fracture of the tube en cool-down after a LOCA. The criterion relates only to the ability of cladding to remain intact when subject to the the=al quenching stresses generated by re-wetting Py the emergency core cooling supply. The retention of post-quench ductility is not an i=plicit feature of such a criterion and the criterien does not relate to the ability of the cladding to resist handling forces due to subsequent discharging operations. Although the wall thickness of the 60-pin size cladding is s= aller than cladding used for the US tests, the frag =entatien linit in te=s of the percentage equivalent cladding reacted is cc= parable and the data obtained on 60-pin size te Fig. 6 suggest that 1 3 equivalent metal reacted is an appropriate iterien limiting the enset of frssmentation for initially unexidised or lightly oxidised-undeformed cladding. For nedularly pre-oxidised cladding the thick oxide centributes largely to the 1 5 ECH so that this criterion can easily be exceeded. But in such cases the total oxidation has no obvious relation to the oxygen diffused into the S-phase which may be the centrolling facter. The total extent of exidation of cladding in a LOCA is not uniquely descriptive of the distributien of the oxygen taken up during exidation. If it is accepted that the oxide and C e high crygen stabilised 2-phase are inherently brittle, exidation =odifies the =echanical properties of cladding by (a) effectively reducing the non-brittle cross section of the wall to that defined by the extent of the S-phase, (b) by increasing tFe strength and decreasing the ductility of the residual S-phase (by increasing the concentration of dissclved oxygen) and (c) by the possible fc=ation of cr-incursions and precipitates in the prior S-phase. The precise fem of the transient-temperature, time paraneters, particularly in a eccl-down 15
period, determine the relative magnitudes of the oxygen concentration related phase re-distributions described above and hence detemine the resultant modifications to the mechanical properties. Since the individual effects of these parameters are not explicitly accounted for in a total oxygen uptake criterion the acceptab".lity of the 1 5 ECR is thus limited to the types of transient-temperature profiles that have been used in the oxidation and quench tests. The desirability of a criterien explicitly - related to the distribution of oxygen in the cladding rather than the total oxygen uptake is discussed in a later section. APPLICATION OF A 17% ECR CRITERION TO SGFG CLADDING The time to oxidise SGE4R 60-pin size cladding (of wall thickness of the minimum of the spe.:ification range 0.0597 cm) to 15 equivalent cladding reacted to form stoichiometric zirconlahas been calculated using the isothemal rate data of Cathcart et al.(2e) Times have been calculated for both single sided and double sided oxidation (at equivalent rates) although only single sided oxidation is envisegcd in a hypothetical LOCA in SGHWR. The times for isother=al exposures are plotted in Fig.12 as a function of te=perature and demonstrate the censiderahle censervatism of the 15 ECH criterion for realistic transients for SGEWR. ALTERNATIVE DERITTLDT P CRITERIA, Other embrittlement criteria have nsuggestedbasedon(Lt)k29) cero ductility temperature (ZDT) and Bf and more recently the oxygen concentration in the residual S-phase.x30,31 / Apart from the criteria based on the oxygen in the residual S-physe, the other criteria have been exhaustively reviewed by Scatena(4) who established the equivalence of
.iese critaria to the 15 ECR criterion if the actual loads used in deriving the data ,ere extrapolated to estimated the =al shock loads experienced in a LOCA. The approxi=cte equivalent criteria were stated to be (Dt)? = 0.02 cm er Fw = 0 5, However, test No. 3 frem the ANL results for which Scatena quotes (Dt)! = 0.01e em failed in quenching with an extent of reaction
= 25% equivalent metal (i.e. > 15 ECR). This anomalously Icw (Dt)? can be attributed to a lor 4 cool-down in the (a + S) phase field prcmoting the fc=ation of embrittling 2-incursions and 2 precipitates in the prior S-phase. Scatena dismisses the.criterien based en zero ductility temperature broadly on the grounds that such a criterien is only related to residual ductility for specific loads and totally ignores the residual ultimate tensile strergth which must be an important property in resisting the loads imposed by LOCA quench stresses. The US NRC(l) acknowledged that an embrittling criterion which took account of the specific effects of oxygen in the prior S-phase was desirable but, at the time the criterien was defined, the finite difference =ethods for calculating oxygen in S-phase on finite thickness specimens were not available. Subsequently Pawel(17,30) has derived approx'w.te analytical and approximate finite difference solutions to the diffusion equations for ene dimensional diffusion in semi-infinite media and has calculated the crygen concentration profiles and mean oggen centent in the prior S-phase fer tubes which y ge cxidised en both sides and ecmpression tested by Hobson and Rittenhouse.(o) The fraction saturation was also calculated and Pawel correlated the fe =atien of e-incursiens with the approach to saturation. In specimens where the mean oxygen ecucentration of the S-phase was at least 16
1 l l l 95% of the saturation value, 2-incursions tended to form en cooling. It was observed that all speci= ens, except two exposed at 106600, which formed ry-incursions were brittle at room temperature. In this instance brittleness is defined by failure with no plastic deformation at the test load. Oxida-tion for 2 min at 2400 F (131600) hewever, produced a brittle specimen which did not have e-incursions. Pawel suggests that the mean oxygen concentra-tien in solution in the S-phase of this specimen is sufficient to cause a ; degree of hardening leading to failuye with nil ductility in the slev ' compression test performed by Hobsen\32) and thus suggests this level of erfgen in S-phase (as 0.7 wt.%) tegether with a 95% saturatica condition as an embrittlement criterien. Sawatzky(31) further developed this theme but suggests a lower limit should be set en the strength of the fuel clad as well as its ductility. Sawatsky proposed as an interis criterien for avoiding embrittlement that I the average concentration of the oxygen in the S-phase =ust be less than ) 0.7 wt.% through at least half the wall thickness. This, it is claimed l gives the clad not only some ductility but it gives it a miM m strength of half that of unexidised sheathing. An additional feature is introduced by the werk of hruta(33) who ! suggests that as well as the parameter of mean oxygen concentration, the ductility in a tube wall can be a function of the shape of the erfgen concentration profile when ring specimens are tested in compression at temperatures in the range 100-700 CC. These conclusions were based on experiments carried out on Zi caloy-4 specimens oxidised at 90000 and heat treated at 850 C so that only e-phase was present. The results of ring compression tests after various lengths of anneals, i.e. for different i orfgen concentration profiles, showed that certain annealing times gave a MM n l in ductility which was recovered at lenger annealing times. The authors concluded that the shape of the oxygen concentration profile in , e-phase can influence the ductility in a ring ec=pression test. Ecweve" I the interpretation of these results encountered difficulties due to the ! influence of high erfgen concentrations on the grain gevth in 2-Zircalcy. l The development of a criterien based on the precise distributien of i cxygen, particularly with respect to the cencentration profile and quantity dissolved in prior S-phase, requires that reliable calculational technin,ues I are available for the dete=ination of or/gep concentratien profiles thrcush the tube wall. Pawel's calculatiens(30) were the first atte=pt to calculate oxygen in S and the values obtained were used by Fawel to correlate with mechanical properties. Clearly, further woric needs to be done when the finite difference codes have been verified to correlate an agreed definition of frag =entation with the parameters of, (a) S-phase width and structure, (b) a-incursions and precipitates and (c) average cr/ gen concentration (and possibly profile shape) in pric: S-phase. SEPARATE EFFEC
- OF TIMp:RA*URE CN Da.R'?"LE O C As well as a frag =entation criterien based on time ac te=pers*ure phenomena, i.e. total or/ gen uptake, the US NRC(1) have adepted a criterien limiting the clad temperature to 2200 0F (12040 0) to preclude embrittle=ent of the cladding. There are other consideratiens which require te=peratce limitations, e.g. clad melting, Zirealcy-steam exother=ic self heating and '
Zircaloy-stainless steel pid eutectic for=atien. Ecwever, the limit of 17
0 2200 F (1204 0) is solely in respect of e= brittle =ent and a te=peratre criterien in ex9ess of 2200 0F (12040 0) would be appropriate for the other considerations.(1) The baeis for defining the above te=perature li=it in respect of e= brittle =ent is the apparent anc=alous behaviour of six specinens exposed to double sided oxidation for 2 min at 24000F (1316 C) and tested by Hobson(32) to failure in slow strain rate ec=pression at te=peratures below 3000 F (14900). For specimens exposed up to 22000F (1204cC) there was a correlation between "zero ductility te=perature" and Fw, the fractional S-phase relative to the original wall thick ess. Ecwever, specimens exposed at 2400 F (13160C) for 2 min (and 4 min) sere found not to fit this correlatien, being brittle at all values of Fw. No specimens exposed between 22000F (1204 0 0) and 24000F (131600) were tested, so that the icwer te=perature w9s defined as the te=perature limit. Also, Ecbsen and Rittenhouse(6 had determined the rate of penetration of ( (as Br/ Vt) as a
/
function of te=perature and stated that at te=peratures up to about 2200 0F (12040 0) the rate increased linearly with te=perature but at te=peratures above this, the slope of B fjVt versus temperature increased processively indicating that the embrittlement lirJced with increase in C acc91erated at te=peratures greater than 22000F (12040 0). Recently, 31eder=an(13) presented =easure=ents of g as a function of ti=e at tsethe==al te=peratures made with a high degree of precision and showed that the growth of g was c smooth function of time up to 27000F (1471cC). In explaining the ncn-correlation of Fw with deformation te=perature, Hobsen points cut snat for a given Fw, the exygen centent of the S-phase increases with temperature due to more rapid diffusion and an increased oxygen solubility limit in the S-phase. The fractien saturation, however, for the same Fw, decreases with increase in te=peratr e due to the relative kinetics of oxygen diffusion and C penetration. So, for a given total oxidation, i.e. a given Fw, at the higher te=perature, although e-incursiens , are less likely o for=, the higher concentratien of oxygen in the prior S-phase could be st (ficient to prc=cte the cbserved embrittlement, i.e. the same total exidation at the higher te=perature produces e= brittle =ent due to an increase in the oxygen content of the S-phase to a level at which the S-phase itself is inherently brittle. Sawatzky(31) ec==ents on Hobson's speci= ens oxidised at 24000 F (13160 0) for 2 min and points out that the =ean oxygen concentratien of the exygen in the S-phase would be greater than the 0.7 wt.% criterien suggested by Favel and would therefore be expected to be brittle due to time-temperature phenomena rather than solely a temperature effect. In fact specimens exidised for shorter times than 2 min at 2400 F0 (131600) may well have had less than 0.7 wt.% cxygen in the S-phase and therefere not been characterised as brittle when tested in ec=pressien at the sa=e strain rate as the previous specimens. Fawel(30) in suggesting an alternative e= brittle =ent criterien of a =ean oxygen concentration in the S-phase of 0.7 wt.% coupled with less than 9c% oxygen saturation of the S-phase, has calculated the time required fer the exidation reaction to reach either of these li:W:ing values as a fanction of te=perature. The critical times were plotted as a fanction of te=perature and are reproduced in Fig. 13 3elew abcut 1260 C the =ean oxygen cencentration in the S-phase is limited to less than 0.7 vt.% as that is the
- um solubility at that te=perature, (the solubility limits are derived 18
from data published by Mallet (B) and Gebhardt.(34) Thus below = 126000 the critical time is that required to reach 95% saturation which would cause 2-incursions to form and promote room-temperature embrittlement. Above
= 1260 C the mean oxygen concentratien in the S-phase is reached first which causes inherent brittleness in the S-phase. Fawel extends the analysis to show that if the locus of critical times is plotted as the locus of critical amounts of oxidation (expressed as Fw and obtained from experimental measurements of Fw at isothemal temperatures and the relevant critical time) the behaviour is as shown in Fig. 14. The critical amount of oxidation (expressed as Fw) is not constant with temperature but the critical extent of oxidation (Fw) to produce 95% saturation, decreases with increasirg temperature (C increases) up to about 12600 0 but at higher temperatures the average oxygen concentration in the S-phase reaches the critical 0.7 vt.% at lower extens C of exidation than the critical 95%
saturation. Thus at about 1260 C the mechanism of embrittling changes frem one specific effect of oxygen concentration in the S-phass (95% saturation), to another (an average of 0.7 vt.% in the S-phase).
"'he above arguments are atrictly only valid for isothe=al exidatien and if the equilibrium or/ gen cencentration at phase boundaries during transient-temperature time exposures is not mainta % d, the critical values would be difficult to predict. Also, the critical values do not necessarily represent the best data in te =s of diffusion coefficients, themedym mic equilibria etc. However, whatever li=iting values are appropriate, the general principles reeM n the same. Additionally, critical values may also depend to some extent on the actual loads chosen to define an embrittlement criterien.
Sawatzky(31) used Pawel's expression for the tedal solubi'dt crygen in S-phase, the diffusien coefficient published by MalletN and y of his own estimates of the 2/3 interface position, and has calculated the time to embrittlement fer double sided exidation of a 0.04020 cm wall thickness Zirealcy-4 tube based on the criterien of an average of less than 0.7 vt.% oxygen in at least b if the tube wall and also shows a discentinuity in embrittlement at = 126000. The plot of the times to reach the criterien as a function of temperature shows a marked discontinuity at this temperature, giving shorter times to embrittlement than the data at less than 12600C would indicate were they extrapolated. Further evidence of an accelera forequivalentamountsofoxidationh'icntowardsanembrittlementcondition tt C comes from the work of Kematsu et a1\35)e=peratures who showedgreater that thethan lead= required 1260 0 i to cause an initial crack in specimens ecmpressed at reos temperature decreases as the oxidation temperature increases but decreases more rapidly above 126000. The authers agree with earlier work in that the acceleration to an embrittlement cendition at temperatures greater than 126000 could be attributed to the more rapid filling of the S-phase with czygen at the higher temperatures. Evidence was presented, however, to shew that the onset of brittleness dees not always occur at a specific har:iness in the S-phase mid wall (taken as a measure of oxygen cencentration). Scme specimens oxidised at te=peratures greater than 1280 0 0 exhibited brittleness at lever hardness values (lower amounts of exidatien) than specimens exposed at lower temperatures. The authers suggest that phenomena other than the mean concentration of oxygen in the S-phase influence the mechanical 19
properties of such claddirg and tentatively suggest that the removal of the cold worked structure or the prior S-phase grain growth may be factors contributirg to this effect. In this connection it is perhaps pertinent to note that Farata et al(33) suggest that the shape of an oxygen concentration profile as well as the mean oxygen concentration may influence the =echanical response of oxygenated Zircaloy, albeit their results were obtained exclusively on 2-phase =aterial. In sm m 7 there is ample published experimental evidence to suggest that the onset of an embrittlement condition, as defined by crackir4 with nil ductility in rirg compression tests at temperatures between a=bient and = 10000 0, is arcelerated ence the material is oxidised at temperatr es greater than = 126000. Although the mean concentration of ovygen in the S-phase and the formation of a-incursions are significant fr Qcrs in prcmotir4 embrittlement, the detailed understandirs of all Ine factors that contribute to the embrittlement of Zircaloy claddirg exidised up to temperatures of about 14000C is not yet available. Ecwever, the specimens oxidised in the SNL Gleeble tests (and = cst of the ANL testo) have all been exposed to steam at temperatr es greater than 22000F (1204 0 0), (except transients I7 and 7) and have exhibited censistent behaviour, remining unbroken when subject to the leads caused by quenching as lors as the total oxidation did not exceed 17% equivalent =etal reacted. Scatena(4) has shown that the ther=al quenching stresses in 2WR cladding of wall thickness 0.0813 cm are = 38000 lb/in2 and relatively independent of the extent of oxidation. Se US NRC do not accept this calculation as being entirely accurate since it does not take into acccunt the different properties (as yet not ec=pletely available) of the high and lcw exygen Zirealey phases. However, if it is accepted that the stress calculated by Scatena is a reasonable approxi=ation, Scatena has also shown that the calculated stress is equivalent to a load of = 70 lb/in of specimen, in the radial direction. Whilst the load causing failure in the specimens tested by Hobsen and Rittenhouse is not reported, Scatena has shewn that the loads required to cause fracture in the specimens of Meservey and Her::e1 or the GE TIE tests were significantly higher than 70 lb/in unless, as in a very few cases, the oxidation of the speci= ens was such that (Ot)! 2 0.02 cm (i.e. 2 17% equivalent metal reacted). Fig.15, taken from rematsu(33) also shows that in this work no specimenswerefracturedatloadslessthanthequenchingloads(70lb/in= = 6.4 kg/5 =m) unless the specimen was oxidised in excess of 17% ECR. (This is approximately the same as = 20
- cxide based on a tube wall tnickness of 0.0600 cm). Also, the pest exida~ co=pression tests all defor: Zircalcy tubing with oxide, high oxygen e-phee and los c ygen (prior 3) a-phase, at room temperature, whereas in actual quenching tests the potential fracture stresses are generated at the Leidenfrcst te=peratre 370-53800, where the effect of oxygen on the mechanical properties of prier beta phase may be different to those at room temperature.
Thus even in oxidation exposures where the temperatures exceed = 1260 0C, the fuel tubes should be capable of resisting the thermal quenching stresses in a LOCA provided the tot:L1 oxidatien does not exceed 17% ECR. 20
1 CONCLUSIONS For fuel element cladding in the range 0.0660-0.0813 cm wall thickness, the ability of previously unexidised or lightly oxidised claddirg to remain intact after oxidation and quenching can be expressed in tems of the total extent of reaction. Such fuel tubes will not crack or fragment providing the total orfgen uptake is not greater than that required to convert 15 of 1 the original metal wall to stoichiometric circenia. Thus, providing the 15 ECR criterion is not exceeded, a coolable geometry should be maintained for a fuel element in a pressure tube. Tubes pre-oxidised with thin oxide fi hs simulating normal black oxide confo= to the 15 ECR criterion in the sa=e way as virgin metal tubes. For tubes pre-exidised with si=ulated nodular type oxide the subsequent behaviour on quenching ir lependent on the amount of pre-exidatien and the subsequent high temperature oxidation. If the pre-oxidised nodular film is sufficiently thick so that the pre-oxidation is in the region of 15 ECR i I such a tube will remain intact after quenching from a further period of high te=perature oxidation (the limits of which are as yet undefined). This result e=phasises the limitation of the total oxidation criterien and confi ns the role of the distribution of oxygen in dete =ining the resistance of the cladding to leads generated by quenching. hrther work needs to be l done ce quantify the effects of nodular type pre-oxidation en embrittlement j behaviour in a LOCA. The oxidation and quench tests in the Gleeble demonstrate a considerable margin in tems of total extent of oxidation between the 15 ECH embrittle- l ment criterion and the three types of extremely pessi=istic transient i simulated in the Gleeble. l The 1 5 ECR embrittlement criterion does not relate to the ability of post LOCA quenched fuel tubing to resiat handling forces. Further work needs to be done to measure the resistance of exidised tubir4 to imtact type loads chosen to simulate discharge ham 11ng forces. The validity of the 1 5 ECR embrittlement criterien has not been established from a theoretical standpoint. The criterien is a measure of total exidation and does not appear to account for the detailed distributien of or/ gen thrcughout the clad wall. As well as, as yet undefined effects from the oxide film and the stabilised a-phase, the ability to resist the = al shock leading of an oxidised fuel tube depends en (1) the resii. extent of prior S-phase, (ii) the ascunt of oxygen in the prior S-phase (iii) the fo= ation of 2-phase incursions and precipitates in the pric: E-phase and (iv) pessibly other effects such as the shape of the cr/ gen profile. 1 More detailed work needs to be done before an improved alte=ativa 1 embrittlement criterien relating the mechanical properties of oxidised fuel i tubes to the parsmeters (1) to (iv) above can be reliably established. ; 1 Specifically, in relation to the 1 5 ECH embrittlement criterien l further werk needs to be carried out to dete =ine the effects cf slew cooling through the (2 + S) phase field en the fc=ation of 2-incursiens l and the effects of incursions en the 1 5 ECH criterien. 21 1
There is substantial experimental evidence in the literature to support the view that the approach to embrittlement at ambient (or icw) temperatures is accentuated after oxidation at temperatures above 1260o0. This is ascribed to the increased solubility and enhanced filling of the prior S-phase with oxygen at the higher temperatures with the subsequent associated inherent embrittlement of the prior S-phase. However, the experimental evidence published to date, is related to compression tests usually employing higher loads than those experienced by claddirs in a LOCA quench and en specimens cooled after oxidatien and reheated to temperatures in the range ambient to = 15000. Erperiments on tubes oxidised at temperatures greater than 12000 0 (up to 140000 for SNL tests, up to 166300 for ANL tests) both isothermally and transiently and stressed directly by quenching shew that tubes remain intact if the total extent of oxidation is less than 1'l% ECR. Methods of calculating the oxygen concentration profiles in exidised Zircaloy tubing are necessary for the detailed understanding of the effects of oxygen profiles on the mechanical properties of the oxidised tube. Finite difference computer codes have been developed and have shown promising results with respect to calculating phase widths and exygen concentration profiles. FLirther detailed validation work en these codes is necessary to establish confidence in the oxidation model, basic physical properties and kinetic data, particularly as applied to transient-temperature time oxidation. ACKNOWLEDGDEUTS The author gratefully acknowledges the centributions of Mr A Marshall in carrying out the Chemical Analyses for oxygen and Mr D Crichton for performing the Gleeble oxidation and quench experiments. l 1 l l 22
REFERENCE
- 1. Acceptance criteria for emergency core cooling systems for light water power reactors. Paragraph (b) of 10 Code of Federal Regulations 50.46,
- 2. Citeria for emergency core coolirg systems for light water power reactors. Interim policy statement. USAEC Federal Register Vol. 36 p.12247, 29 June 1971 3 HESSON J C, IVIDS R 0, VILSON R E, NISHIO K and 3ARNIS C. Laboratory si=ulations of claddirg steam reactions following loss of coolant accidents in water cooled reactors. AEL-7609 January 1970.
4 SCATENA G J. Fuel claddirg embrittlement during a less of coolant accident. NEDO-10674, October 1972. 5 TROVS3 F V. Private co:m::unication.
- 6. HO3 SON D 0 and RI?I'ENEOUSE P L. Embrittlement of Zireale'/ clad fuel ,
rods by steam during LOCA transients. ORNL-4758, January 1972.
- 7. DUNCAN J D and LD3 NARD J E. Thermal response and cladding performance of Zirealoy clad si=ulated fuel bundles under high temperature loss of coolant conditions. GEAP-13174 May 1976.
G. MALL 21' M V, ANTW V M and VILSON P R. The diffusion of crygen in a and S Zirealoy-2 and Zircalcy-3 at high temperature. Jul mectrochem Soc. 106 1959, p.161. 9 HENCH J E and LIFFENGREN D J. The effect of fuel red failure on emergency core cooling system design for boiling water reacters. Nuclear Technology Vol. 11, August 1971, p.544.
- 10. G.UAFP W J and McCARDELL R K. Behaviour of Zircaloy clad UO2 fuel rods during film boilirg in a FVR environment. Presented at the OECD l cec:mittee on the Safety cf Nuclear Installation Specialists Meeting en l the 3ehaviour of Water Reactor Fuel Elements under accident cenditions. '
Spatind, Norway, September 1976. , 1
- 11. MEENER A S et al. Post-irradiation e m N tion results for the i i:Tadiation effects scoping test-1. ANCR-1336 (to be published). j l
- 12. GRA3ER M J and ZU N I V F. Metallurgical evaluation of Zircaloy )
exposed to emergency core cooling conditions. Trans. Am. Nucl. Soc. l 12 p,356, 1969. I l 13 3IEDERMAN R R, 3ALLINGER R G and DOESON V G. A study of Ziresloy-4 l steam oxidatien reaction kinetics. IFRI-NP-225 September 1976. l
- 14. YURIE G J et al. Microstructures of the scales formed on Zircaloy-4 1 in steam at elevated te..:peratures. Oxidatien of Metals 7el. 10, No. 4,1976, p.255 23
15 3RAEURST D E and HEUER P M. The effects of deformation on the high temperature steam oxidation of Zircalcy-2. Jnl. Nuc. Mats 55 (1975) 311-326.
- 16. TUCKER M 0 et al. Partitionir4 behaviour of oxygen in Steam Oxidised ZirealoyRD/3/N3877,1977 17 PAVEL R E. Diffusion in a finite system with a moving bomb 7 Jnl. Nuc. Mats 49, p.281-290, 1973-74
- 18. MALANG S. SIMERAN-1 a computer code for the si=ultaneous calculation of oxygen distributions and temperature profiles in Zircaloy during exposure to hi 6h temperature oxidising environments. OENL-5083, November 1975 19 MALANG S and SCEANZ G. Descriptica and verification of SIMERAN-1 A computer code for the calculation of the hi h6 temperature steam cxidation of Zircaley. Specialist meeting on 3ehavicur of Water Reactor Puel Elements under reactor conditions. OECD/Nue. Eng. Agency Spatind, Norway, September 1976.
- 20. LEinmW S et al. Projekt Nukleare Sicherheit Halbjahrsberich 1/75 E E -2195, 1975
- 21. CATHCART J V et al. ORNL-TM-5148, 1975
- 22. LEISTIICW S et al. ProjektNukleareSicherheitEalbjahrsbericht2/75 ER 2262,1976.
23 PARSONS P D. A review of the oxygen diffusien coefficient in :r and S Zirealeys. TR3 Report 2882(S),1977.
- 24. CATHCART J 7, McIEE R A, PAWEL R E and PEEKINS R A. Su= mary of the irconium metal-water oxidaticn kinetics programme. Pourth Water Reacter Safety Research Information meeting. Gaithersburg, USA, 27-30 September 1976.
25 MALANG S. Private ecmmunication, 1978.
- 26. D03SCN V G and n N.AN R R. ZORO-1. A finite difference computer model for Zirealoy-4 cxidatien in steam. EPRI-NP-347, December 1976.
SAWATZKr A. A proposed criterien ser the oxygen embrittlement of 27 Zircalcy-4 fuel cladding. 4th Int. Conf. Zircenium in the Nuclear Industry, Stratford-on-Avon, June 1979.
- 28. CATECARP J V et al. Zirconium metal-water oxidation kinetics IV.
Reaction rate studies, On'7L-Nureg-17, August 1977. l 29 MEEERTY R H and 3ERZEL R. 3rittle behaviour of Zirealey in an emergency core cooling environment. I3-1389 September 1970.
- 30. PAWEL 3 E. Oxygen diffusion in beta Zircalcy during steam cxidation.
Jnl. Nuc. Mats. 50, p.247-58, 1974 2A
4
- 31. Sf(ATZKT A. Sheath oxidation and embrittlement. Paper presented at the JUICE meeting 1974.
- 32. HOBSON D 0. Ductile-brittle behaviour of Zircaloy fuel. cladding.
Water Reactor Safety Meeting. 26 March, Salt Lake City UTAH, Conf. 730304, 1973 l 33 RTRUTA T, HASHIMDr0 M and KAWASAKI S. Effect of oxygen content and distribution on the embrittlement of Zirealoy cladding. JAERI-6182, July 1975
- 34. GMARTYf E, S]UEEZZI H D and DDRRSCHNA3EL V. J. Nucl. Mats 4, p.255 35 ECMATSU K, TAKADA Y, MIZUTA M and TAKAHASHI S. The effects of oxidation temperature and slow cool-down on ductile-brittle behaviour of Zircaloy fuel cladding. Presented at Water Reactor Safety Specialists meeting. Spatind, Norway, September 1976.
i i l ! l I
)
i l 25 Page 26 blank
I 1 i 1 i
, TAB 12 1 oxide, e-phase and 8-phase thicknesses calculated for ANL tests oxide thick-Test ness em oxide 1hickness Thickness of 8 x 100 - 8 x 100 Equi';alent. (Dt)3i
- g, % orig, of o ** remaining 3 orig. w.t. oxidised w.t. metal reacted cm i "*11 ** "01 I Av. Hux.
i d 12 0.0019 0.0039 2.4 - 0.0019 0.0774 98.3 97.6 - - 1 16 0.0046 0.0065 5.8 0.0170 0.0216 - 0.0587 74.6 73 1 6.59. 0.013 4 0.0042 0.0082 53 0.0192 0.0234' O.0567 69.4 70.8 6.87 0.142 15 0.0048 0.0097 6.1 0.0161 0.0210 - 0.0595 75 5 73.9 6.87 0.126 7 0.0045 0.0074 57 0.01 % 0.0241 0.0562 71.4 70.0 7.14 0.147 8 0.0058 0.0097 5.8 0.0178 0.0236 0.0571 72.6 70.8 7 97 0.143 9 0.0085 0.0145 10.8 0.0257 0.0342 0.0475 60.4 58.1 11. 9 0.213
~
5 0.0149 0.0256 18.9 0.0260 0.0409 0.0430 54.6 51.3 17.03 0.257 17 0.0170 0.0267 21.6 0.0238 0.0408 0.0439 $5.8 51.8 18.4 0.256 18 0.0173 0.0270 22.0 0.0245 0.0418- 0.0429 54.5 50.6 18.68 0.262 10 0.0213 0.0352 27.0 0.0242 0.0455 0.0406 51.6 47 2 21 98 0.287
.11 0.0194 0.0269 24.7 0.03 % 0.0528 0.0329 41.8 38.4 22.25 0 334 6 0.0211 0.0418 26.8 0.0256 0.0467 0.0394 50.1- 45.8 22.25 0.294 3 0.0158 0.0303 20.1 0.0694 0.0852tt - it - -
25.55 - 13 0.0249 0.0412 31.6 0.0414 0.of/33 0.0212 26.9 24.2 28.02 0.423 ] 14 0.0364 0.0497 46.2 0.0441 0.0805 0.0110 .14.0 12.0 38.46 0 516 ! Mn1 % Zro Calculated from 00 * *#I8* "*L'
- 1*5 Calculated from the relation between % metal reacted and oxide and a-phase thickness, i.e. 'egn (3).
T. Calculated from 11obson-Hittenhouse formula.(6) it Imrgo extent of reaction but anomolously small oxide thickness, calculation yields oxide + ar greater than w.t. I i
TAIIE 2 Remnits of Flecht testa for Zircaloy-2 fuel rod exposed in Run Zr 4 All locations remained intact "I'"" oxide thicknees Remaining Test " No. '8"P* thicknees E % metal thickness 8 x 100 $ x 100 (Dt)$ C % orig. cm rsected well of prior orig. w.t. Oxidised w.t. cm p, cm 17-6 1389 0.0089 10.9 0.0147 0.0236
~
9.6 0.0608 74.8 72.0 0.0120 25 6- 1388 0.0011 1.4 0.0124 0.0234 2.9 0.0682 83 9 83 5 0.0121 18-6 1386 0.0089 10 9 0.0246 0.0157 97 0.0598 73.6 70.9 1122 24-6 1383 0.0030* 3.7 0.0036' O.0066* 3.0 0.0758 93.2 92.0 0.0120 24-7 1287 0.0056 6.9
- 16-6
)4.5 - - -
0.0056 1273 0.0048 59 0.0089 '_o.0137 53 0.0693 85.2 83 5 0.0071 11-6 1264 0.0058 71 0.0157 'O.0216 0.0613 7.2 76.0 74.2 0.0072 g 11-2
- Q2 0.0137 16.9 0.0335 0.0472 16.4 0.0389 47.8 45.2 0.0298t 19-6 1257 0.0064 7.9 0.0048* o.0048 59 0.0723 88 9 86.6 0.0075 11 4 1244 0.0074 9.1 0.0173 0.0246 8.7 0.0592 72.8 70.6 0.0070 32-6 1233 0.0053 6.5 0.0064 0.0117 0.0715 53 87.9 85.9 0.0064 12-6 1229 0.0079 9.7 0.0109 0.0188 8.1 0.0653 80 3 77.6 0.0060 23 6 1221 0.0036 4.4 0.0071 0.0107 4.0 0.0719 88.4 87.0 0.0060 l 10-6 1195 0.0053 6.5 0.0104 0.0157 59 0.0675 83.0 81.1 0.0053 l 9-6 1191 0.0036 4.4 0.0079 0.0114 4.1 0.0711 87 5 86.1 0.0046 i 26-7 1182 0.0048 59 0.0058 0.0107 i
4.8 0.0724 89 1 87 2 0.0053 27-6 1151 0.0025 31 0.0074 0.0099 ( 32 0.0723 88.9 88.0 - 33 6 1149 0.0048 59 0.0084 0.0132 o.0698 5.g 85 9 84.1 - 17-8 1121 0.0025 3.1 0.0036 0.0061 25 0.0761 93.6 92.6 0.0036 8-7 1047 0.0010 1.2 0.0026* o.0036 t.2 0.0771 94.8 94.4 0.0015 l 2-6 996 0.0028 3.4 0.0026* o.0056 2.6 0.0739 90 9 89.8 0.0010 17-10 781 0.0 0 0.0 0.0 0.0813 100.0 100 - l l l
- Sampled by Idaho Nuclear Corporation.
{ INo value quoted by Duncan and Leonard; listed figure la calculated from equation (4).
5 2 de }{2)*5gARRCy*0**j$2Ey"?gRR"1 >t= . a' r - =- upsc 9 : uu:- x-fgj m.eeeeeeeeeeeeeeeeeeeeeece a- d a 2 1I 21 e sa m s o es ,- ~ ~ w n n t, e e e c o m m m e n $ a f f f f f
- fid sd A A A A AJ J d AJ Jc'di 2 5 ..
l u l
~'
%"$$s *E39 *2 $$20l OS9:" 5;; TS ,
s- e e1$^1 s s s 'i seeed 8588sss8 = l
~
&&daddddd assidadiddsaddd i U
l
~. e
. k k $$$
a sascessnaassasasssssss**s e ; . E t U
.
- i
* \
. g O.
Y @@
. . . . . M. @. N.
. .....N. . . N. D. . ......
"U '
3 RE88$87&MRSS$28$853%SR$$$ \ m u 8 l 2 8 .* C
- t 8 s"tess 451l,yW CR E'g*s's e s ear essssse 210 ss s s%286ssas 3"1
= adadadideieidadddaidi&&&& l
% 0% 1 a l l
^
. co,e c e ne e - e s c ee ,e e l
~
2 3. .. e $ c D O ; n "s"ee ee=ocoes
- X g8ee ^$ 2-3.ys*s* eee.ee m'J e - cETO I Sj ecoceeeeeeeeooceeeeeeeoco l
"s e i I ha m m' i 18 8 8885555$s$ne88855$55858 i a
- 2 d'
- d$edadidadiddidaddididd -
3 ' S $
$" $ s* C 8$$ C.* 8 E
' 5 3$"*8eS$$5855$sak"$C*$%%ROOC e$d8855888 $ ,
8 adidaddddddddddddiddddddi A 4 1 f 3 2- es-omoemecceneemce--commm ! _a 5} J AJJ4 Afdf j diddt3Jddddf d d;; a 2
$ss E EO
- 8e%h8e1fes E s$3#4es ee8esoee!$ se s didddddsadddddddidisiddds
"**"a***t**tE2XER33%RR$55 1
2 - - ~ 1 O h g E 4 E t # 29 1 i 1
i!t i ! , L '
. 4 e
T
'T f
N M d t e c %
=
r l a
- oh cp . i " " " "
a t e
" " " " " " " " r l
f k nmC F I e '%
'e t
aee a N eut 8 8 9 0 6 P q 8 9 6 5 4 3 8 2 9 6 e -
- - - - 6 9 6 8 8 r 5 7 5 6 6 6 4 2 3 3 g ,
1 1 1 1 1 1 1 1 1 1 1 4 ( - +
=
A s 4 t 2 2 9 3 0 g 3 5 8 5 1 8 3 0 u e
- 4 3 8 1 6 3 4 1 8 4 9 4 5 o l
) m 2 2 2 3 1 2 1
- 2 2 2 2 1 1 l b r
'v t c 0 0 0 0 0 0 0 0 0 0 0 0 0 o a e
D m T V ( 0 0 0 0 0 0 0 0 0 0 0 0 0 o
' n m a o e
, r %
s f u h'
- t. e~
v e
+
f w i r o - g n n 8 8 6 2 2 3 4 1 3 1 2 0 9 '- -
%oi a h 9 1 0 7 0 9 0 0 1 8 5 2 0 c
- st 3 5 54 4 4 5 6 5 54 7 7 d i -
aa e h T d t w 4 - 8i a t x l E o u P -. l c T- -
"1r a E c G a f t. g r t
a ow n o i f d %l a 2 8 5 6 2 7 3 1 6 4 8 8 6 s g sn u t 2 54 1 1 0 2 3 4 0 7 4 3 p n ai 4 5 5 5 4 5 5 6 5 6 4 7 7 ( e i g c -u h 8i f x ', c r o e r
. n o g
e , n 4 u o q i E 6 t L d " a , @ B n 3e 9 9 6 5 2 5 3 01 5 9 8 8 u n n A a " " f 3 2 0 3 1 2 3 7 8 0 9 i o = T 0" 3 4 4 4 3 4 4 5 4 4 3 6 5 t i r n
- 0 0 0 0 - 0 0 0 0 0 0 0 0 0 a s i
o
- e s i 0 0 0 0 0 0 0 0 0 0 0 0 0 e t , r a s p d '- s x i .
e e x 8 8 8 5 2' 8 4 9 0 9 2 2 6 6 n o 1 0 1 5 5 9 2 1 4 0 4 7 3 4 k e m 5 4 4 4 8 4 4 4 3 4 3 4 2 2 c s + A t e 0 0 0 0 0 0 0 0 0 0 0 0 0 0 i u T S I h o I 0 0 0 0 0 0 0 0 0 0 0 0 0 0 t h n e e d t e 7 0 3 3 8 6 6 9 4 9 5 7 9 9 i i d 3 7 7 1 5 5 6 8 1 4 18 3 8 8 x 'i i s xs m 1 1 1 2 1 0 0 0 1 0 1 0 0 o I t c 0 0 0 0 0 0 0 0 0 0 0 0 0 0, - oen l n 0 0 0 0 0 0 0 0 0 0 0 0 0 0 l o , ek gc a s m b 7_ ai s o 3, rh 9 6 0 0 1 9 1 9 0 9 8 9 9 9 H w et t p v % 6 1 2 7 0 6 8 0 4 81 0 6 0 0 u e ty A 1 2 2 2 2 1 1 1 1 1 1 b h I t 4 n ( m o ) a t n i 4 n t ( e e d c t r ll e% 8 8 5 a u a aat 6 4 6 9 5 3 8 0 5 2 5 7 e . q c = vt ct r ieao 7 8 8 1 5 2 2 2 2 7 1 6 9 9 t. e S e umem q r 1 1 1 2 2 1 1 1 1 1 1 1 f w m y p o o b
+
E l r t a f d nn e w 2 6 6 ei d t 7 o
- - - t g e w
* - 1 7 8 xi t n w
" 1 T B T M 1 1 1 er a g i
o
- - - - l E
P 7 8 0 4 4 1 5 t e u e - 1 1 1 3 - - - - 5 9 h gn c u e
" T- .
L L L 2 2 2 2 . c ra l l
" I I L e a. h a a E N N H N E N N l l t C V G A A A A O A A F
- t t + +
+ .i 3ji ; , i lIfj t!
il;i;%lli ;l jlti if :ii
TABLE 5 Equivalent metal reacted calculated from published phase widths for IDAHO PBF(10) Tg. Mean xygen oxide on a on or on Speclaen in tr phase outside outside analde
% equivalent C , p ,
wt.% cm cm cm 4 IE-ST-1/A 13f6 i 4.7 0.0099 0.0168 0.0089 19.4 IE-ST-1/B 1336 4.7 0.0096 0.0142 0.0089 18.5 4 J l i I 4
TABLE 6 Extent of simulated nodular pre-oxidation, Tubes 1, 3, 2 and 5 oxygen gain Thickness of % equivalent Oxygen gain Thickness of % equivalent metal Total % Tubo on inoide black oxide metal reacted on outside nodular oxide reacted to form equivalent metal flo. Llack oxide on inside to form black nodular oxide on outslue nodular oxide reacted in g pm oxide on inside g m on outside pre-oxidation ! 1 0.05829 2.8 0 31 1.02237 97 3 9 75 10.06 ~ 3 0.0582'e 2.7 0 31 0.58655 55.8 5.6 5.81 2 0.05335 25 0.29 1.79752 171 17 1 17 39 5 0.06072 2.8 0 32 0.76360 73 7.28 7.6 6 I 4
.. - -- _~ . - . - - _ . . . . - - . . - - . . . _ - . ~ -
TA8IE 7 0xidation exposure pareseters and chase thicicnesses fer virris metal tubes oxidised in the Gleeble Is theM M m at Outer on de Inner oxide a thicicness c9"I'"1*** Specimen fail or a incursed Specimen type No. Or $ Ct**P. W. s Survive into prior 8 thic h es ce thicanoes es utal reacted 5 60-pin size 20 % 00 200 3 None 0.0050 0.0043 0.0126 11. !E *'
# A *Ii" ***N C " " S None 0.0048 0.0043 0.0124 11.6(*
IDI 18 " 300 F Fingers of a into 0.0064 0.0058 0.0179 15.8 prior 8 D' D " " S - do - 0.0062 0.0059 0.01 a5 7 17 " 400 F Iarge o arsa.s in 0.0071 0.0064 . - prior 3. cr on inside 16 "
$0 F Iarge e in prior n. 0.0071 0.0060 - -
e on inside 21 1300 300 S Occasional very 0.0044 0.0033 0.0100 9.6 "' amall a finger. A " " S Very infrequent 0.0041 0.0038 0.0109 +0.0(* saml1 a finger. B " 500 S - do - 0.0051 0.0046 0.0136 12.2 "' I 22
- 550 S Small a fingers and C.0056 C.0050 0.01 6 13 4 ")
e pptes. 23 900 S Fingers of a. 0.0078 0.0063 0.0196 18.0 I " " F !arge areas of s ftiddle 0.0091 0.0069 0.0264 2*.1f in prior p. End totaJ exide = 0.0307 0.0450' 39 3' E " " F !arge e incursicas Middle 0.0083 0.0054 0.0200 17 3 End 0.0083 0.0061 0.0295' 20.1 M " " F !arge e incursions Middle 0.0083 0.0056 0.0217 #' 18.1 M 1 End 0.0081 C.0061 0.C339 20.3 24 1400 200 S !arge incursions 0.0072 0.0058 - -(d) ) 900 600 26 1400 400 F Very lang incursions 0.0005 0.0088 - -(d) 900 600 ! 0.0057 0.0065 0.0210' '5.5 33-pin size 30 1400 400 S Fingers of a
'i#818 ****A 31 1300 350 S Infrequent very 0.0050 0.0054 0.0%9 *2. l small incursions 60-pin size F Transient I 8 Small incursions 0.0034 0.0031 0.0110' 3.7
'i'Ei" ***"1 0.0106 IN 0 " S Samil incursions 0.0035 0.0033 3.9 i I Trenaient II 8 Semil incursions 0.0037 0.0031 0.Z99 3.7 J " S Very call incursion 0.0036 0.0026 0.0072 7.6
ID K Transient In 3 Incursions 0.0036 0.0030 0.0092 ' ?.4 IDI l L " 8 Incursions 0.0032 0.0C30 0.0094" 9.I P Transient IV S Small incursions 0.0017 0.0012 0.0033 3 5(*' 4 "W asient V 8 Small incursions 0.0017 0.0012 0.0038 36" I (a) Non or practically non-incurced e/B boundaries. s (b) A rough estimate of 5 of the e-phase thictcneas has been added to account for o incursions. (c) Specimens E. B. M which were badly incursed and were visually assessed separately, the quoted a width inc1:ades an estimate for incursions. (d) Not estimated. I 33 ;. I w r -Mr-+r+, e--<- e,. ,n-pw et - - --* w w.e- s cv. w,wr-p---e--,.m..e-._--%,.-a--p++v,--,-r-e ,1e <,w.ge- p w-vp-te - w m w e. r--n eo -c . ewe- we,- -- w a w wp ee e. or m e se'
_ . - . -.. . _. ~ . _ _ . . , _ _ . . TASLE 6 Oxidation exposure parameters and h thickneeses for ero-exidised tubes oxidioed in the Oleeble Specimen Specimen Isothersal Time at F C or e-Mou Outer oxide Inner oxide e thicuness M '"'
***P* E* * ' *'
type No. '# I '7P0 Servive with prior a cs ** ' Pro.oxiitsed 4 1300 550 8 Fingers of e 0.0056 0.0053 0.0218 15 1 " 3 ** U"*" 6 % 00 200 F !arge fingers of e 0.0099 0.0088 . .(d) onde 1000 600 and precipitates 7 1300 300 8 Fingers of a *otal oxide = 0.00M 0.0188(c) 11 3(c) 1000 600 and precipitates Pre.oxidised 1 1300 300 F Fingers of e and . '. . .(d) nodular e precipitates, oxide Oxideao1einside. 2 1}00 60 8 . do . . . . .(d) I J 1300 150 8 . do . . . . .(d) 5 % 00 300 F Massive e incursions . . . -(d) oxide and
- on outside and inside
!!igh tosp. % 1300 900 F Mneaive e incursione 0.0133 0.0066 . .(d) stees e on outside and inside pre.o t dation % 00 200 F None Triple oxide 0.0242 0.0210 15 29.2 900 600 layer NOTE 8s as Table 8 l
.l t l
- I l
a. 5 34
i 1 TAM E 9 01eeble enecteener total omten contents by chemical analysis tube sospio Oxygen " '9j,"1l*"* Fail or Tube sospi. Oxygen Fei or l No. type wt.5 84F717' No. tTPe wt.5 5 '9M*1'"* "" td reacted ) Virgin setal tutes 20 M + IA0 2.8 12.6 S F M + TO 3.2 9.4 5 0 M 1.6 13 3 3 G M iM 10.0 5 , M + IA0 3.5 12.7 M + 10 32 9.4 18 M 2.4 18 1 F I M + TO 2.8 8.2 3 x + IA0 4.4 18.3 J M 1.0 8.9 3 D M 23 17.8 'S M + IAC 2.5 10.6 M + TO 55 16.6 M + TO 2.5 10 3 17 M 29 20.5 F K M + to 2.8 8.2 3 4 M + IA0 4.8 20.1 I, M + TO 2.8 8.2 8 16 M + IA0 4.6 19 8 F Pre-oxidised 91s black eitide M + 20 91 28.5 4 M 19 15 8 8 21 M + IA0 2.8 12 5 s M + TO 5.8 17.6 I A M+TO 29 8.5 3 6 M + TO 6.0 18.2 F 22 M + IA0 36 15.6 8 7 M + TO 37 11.0 3 5 M + IA0 32 14.2 8 Nodttlar ure-oridised P- -even. 23 M + IA0 36 17 5 8 1 M + TO 7.2 22.1 (10.06) F E M 1.8 20 3 F 2 M + 20 43 *0-24 (17.39) 8 l M + IA0 4.9 22.4 3 M + 10 4.2 12.5 (5.81) s H M 34 22.2 F 5 M + IA0 53 20.4 (7.6) F M + IA0 43 20.0 Einb tennerature nre-oxidised M M + IA0 4.0 19.8 F 14 M 30 26.8 F 30 M + IA0 43 17.5 5 M + IA0 4.7 23 4 l 31 M + IA0 27 12 3 3 15 M + 10 7.0 21 5 F , 1 24 M + IAC 39 17.9 8 1 M + 10 57 17.2 26 M + IA0 k.) 21.1 F ' 1 1 Sample types M - ansple cut from region where all oxide had spalled exposing bright l stabilised a metal. M + !AO - metal plus inner adherent oxide, only outer oxide spalled. M + TO - complete section through total 41 thiccesa no oxide spallation. I
i I ! TASLR 10 Phase thicknessee ard estent 2f reaction by vzberiment and by STER &M-1 t SIMPS &M-1 Emperimental % ER posumi Spoolmen alpha Ta m Mah u ne m M e ., oxide allana nalde gggggAg,, i um pm pa pse fusion graphy Ban.t { fo-pin 1400 C 200 e 20, C At 200 socorde 99 135 93 126 13.6 12.6 11.6-11.8 13 8 200 e + quench * ! hplicate specimene '65 60-pin 1400*C 300 e D at 200 secorde 122 113 121 170(180)* 16.6 16.6-17.8 15.7 16.8 ' Single specimen 200 e + guanch
- 210
- 60-pin 1300"C 300 e 21. A at 200 seconste 86 104 80 110 11 5 0.5-12 5 9.6-10 11 5 hplicate specimene 200 e + quench
- 129 l 60-pto 1300*c $00 e B At 200 eeoonde 111 140 97 136 64 4 14.2 12.2 14.8 Single specia.n 200 e + quench
- 169 60-pin 1300*C 5$0 e 22 at 200 secommie 117 149 116 146 15 4 15.6 13 4 15.6 single specimen 200 e + quench *
- 174 36-pin 1300*C $$0 e 31 at 200 e conde - 117 147 104 149 14.8 12.3 12.9 14.8
+ Single specinea 200 e + 200 e + quench
- 179 60-pin Transtant I At 300 seconde 77 98 60-68 101-105 10.2 9.4-10.0 8.7-8.9a -
b plicate specimens F&G 300 e + guench = 143 60-pin Transient II At 300 eeoonds 72 92 62-68 72-94 9.5 8.9-10.6 7.6-e.7e - hylicate specimens I&J 300 e + quench = 100 Transtant III At 300 seconde il 96 62-66 88-90 9.4 8.2 8.1-8 4* - bylicate specia.no K&L 300 e + quench
- 112
} Transient IV P At 300 seconde 28 27 29 33 35 - 35 - l 300 e + quench 30 j Transtant Y Q At 300 seconde 30 26 32 32 38 - 3.6 - q 300 e + guench
- 32 I.
i btes (*) Indicates the inoluelon of a nominal *W to include a contribution from cr incursione (1) Calculated frue parabello rate constante published by ORML i 5 i 4 4 [ I ! I
C 1400 1500 40 e , , , , n u s SOURCE No. OF TE5TS 35 - 0ANL INTACT 8 eANL FAILED 7 o TTE GE TESTS-INTACT 24 e TTE GE TESTS-FAILED 1 a FLECHT PROGRAM-INTACT' 22 8 0 -
' PEAK 7 IDAHO PBF FAILED - 'jTEMP 2
_E O t; 25 - w e Q Z # e 8 j 20 - a , 2 FAILURE T
- l W 17 % g g --o- -- --
g INTACT g w 15 - p _; I d z o I g 2 a w C U I 10 -
= R l a O g e a
O a Co O .l c 0 5 - C 00 a a a 4 o a' I I "^ l'
- U So l
O cl I I O 500 1000 1500 2000 2500 3000' 3500 ZlRCALOY QUENCH TEMPERATURE, F ' FIG.1 USA EMBRITTLEMENT CRITERl0N _ . - . . _ - _ . . _ . - _ . . _ , . _ . _ . . . . _ . . . - . . . . . . . . . . ~ . - . . - . . _ , . . , _ . . - , _ . . . _ - , ,
40 3 kW STEAM ARGON B_ OILER TEMP. RECORO b STEAM O COLD WATER SPRAY TUBE OUTLET TEMP. TEMP. . PROFILE -COMPARATOR j T/C. -N b P AD
'g j GENERATOR & CONTROL ZlRCALOY-2 TUBE : l COLO l 7
TIME THYRATRON N ' i fg#p
^ PRESSURE LATOR FFUSER STAINLESS STEEL STEAM BOX Fig. 2. Schematic arrangement of Gleeble test facility
.. m--
N, . .
~
b
- 'Tb.lw.. . ,.. m ..
' W ;d .,,3 .
f ,'..'. . ] l>. ? k. f ,
'? % -};$,,.Y
.'*L.
^ a,l '
,7 g y .8 ~.
*t ,
,. ph s%
*L ;
S q'.; .y t v a;
.f
*"'(k; A
Y,4
?"!l .
;y .-. .. - 8 - ,4. ;.'. . .-
5 F-
. ri f 35 4
, j .
~
~ C -g= ,
. -{
.i.h
~
k r 1
,, u , 4,4
,e <
~
%) .) :?
y J. ' 1 Pp
.-n.-- -.-
's'1
..[. ~ . . }
~
^
-~..
~
)_l - .
- . . . -~ -.....,. .c; . . . . .
, , , 2W .C } 7-} t,b-
$r' ,, ,.j #
RFL 7609
. Fig. 3. Arrangement for specimen oxidising and quenching in the Gleeble stainless steel box.
i , -- - ~.~ . . . . . - . . . _ . , , _ _ , _ _ _ , , _ , , _ . _ _ , _ _ , _ _ . _ _ _ ____ _. _ _
1
.l i
l
% EQUIVALENT METAL REACTED METALLOG. VA C. SIMT R AN FUSON I g.7 95 85 II 86 85 84 E 84 82 86 1300 -
1200 -k i o u ,I
, ' b! , y "
- { "* , MA . ' ~ ' Ip * . ,. j b ? .- ? .N k, g #' A -*~,.. ,,_ ,,
f - *
- x60 n z I w Ce . . , -- m 3.s .re h l
,,00 -
\W h ' gg .. :$ . "s l
($ ( h,', ' ,*. w9 , %,.,3
'[ .b' , .' - .I
. $, i'
,_l)
* ' ,, . $ *** l '
p, : ; . . , ' , ' . , , 9 l yd.6 1.,. .1 - X60 ;
.~ 4' u:. .m,,.,.
.[+1. . ,
t,. '> t A*$* ,' ., l
. . . j l
1000 -- j y
-m
..~
4 C* 4(9. 4C
,e
, , C..T,.
-.g. . . . .
~* 7 m,y
'. Ey} ..
o f '/ ; g, =.s (.
;-{./-. , '(;
- 4.
i
'.-) we '
h* ,g-1,
- v i .. .-
X60
. , , , a . ' ', - -
. ~.
H : 1 I t ! : 900 ' 100 200 300 TIME, s l FIG.4. TIME-TEMPERATURE PROFILE FOR WATER QUENCHED TRANSIENT SPECIMENS
- . . . . . . . , , - . . , _ . . , e,.,~ .. ., ..-. ..-.,~.,...,,,,, ..., , _ ,-, __ ._ .,.-. ,. _,..._.. . ..~.,_ _ .-...._... ,. ..,_,- ..
1200 - TR ANSIENT E SPECIMEN Q 1100 U o TR ANSIENT I g
~
b # SPECIMEN P E
< 1000 -
5 a. I W
.s .
,2.
.Y .h%
.kh
, . .e d
""1 900 -
A'.' *.. g;^[^ -; y;# ., rg-lq~;; f
. :- _ p g.^
l v
.6 ~. L. . t ,:, 4. . .r.. . .y . g ,
l 1
- ~
ci &{% ; 1 i ass /scssa x60 eas/ a ts71 x 50 SPECIMEN P SPECIMEN O 800 100 200 3C0 TIM E, s FIG.S. TIME TEMPERATURE PROFILES'FOR TR ANSIENTS H AND E NOT WATER QUENCHED. POWER TERMIN ATED AT 300 s
a . ,p. . ~.m ' x $ 4,7 < t' g g
%. x -
s,.. ..\M M.E.DS.' '< $
. f- *d
~ .l*: ,if y '
5. W '5 .
~ ;y.m .
a
.f'[.dl{>
.*w. m r s .,
- h. #'
[,
- N M '". y,
^
5 4l,?kf., " " *C.W. l "3b' @ 4!Q
. i: ,
ifMY TO' ,\ m 5,, f Q y
..),3 1,?8& n- E m g
. .t. 'e . pp+yc,4f o ' -f. ,.
g 2 j
~ 'i .e hf$0 g 6' ') u s
.& w.y [t*/j ',Y ,,,'y a e i
f ' i 4) , a" l
%$.- , e . M.E e" (d.s) c ; 4c e N* 6 . & ,.. ** f ) ', y .Y C D
'1,
'8 i ,t,,;GN';l,ih .
D. ' 1 c sfq'g y,c , y $
,.. )Shjyp, j ysh.$'~' J-. a'(. g
. . .. ., _. .. a. - .. #u .
,e ,
. ;' c; .pff'"^'j'[h .. e ' th() { )
. %l.f.
> 2
" *iwr. > ;
w <. %
- , h m:
uQ[,p[t . eg p ; : N/d ' a ['r% c $4[Qld.h +s eSq s n;.._'; .= , 1 O l D
) -
O
<, : b ' c0 -, p 33 g y .
.- x , . .
. x ;
p: -
't.* * ' * ' " *
@ - & &'ll,D k? %
4 .$, d h~.*... .,.
, ~
R, '** iry- . h h
- e yg a ny A.
\\y} 2
~
7.,- b .,+b*' { 7 5) ' -
'# f $ ;^rkS'U:
f - y } :' '$.
.. .._$ y x QtWN.M.'.. m.-
Q~.
~=
5 #;-
#% . F. ~ . .
. . . .QA.q, *k ,,. , Ct ,1 '..
/ *N- ~*+ i 7 ,
,= *
.. w. . ..w,u . ,. . , w'. ,
Q N"
.. T.nl f2 Q. :~ rQ ?-
o
..J-o S.' '. o'h.'.s.. a,
.- s
[ r E4
' x.,
5 *m QQ ya s ,Wi'. , - . ,' .
,,s, 4 r:$r-i O % 4"' - .
A r n a syd C _ . . _ . _ _ , _ . _ - . _ ._ .._._ _ _ _. _ _ _ ...__ _.__ _ ...-_. _ ___ _ _.._ __. _ _.-. _ __.-.. _ . .. . . . _.. _._. I J
i i k i 1
*-----rc . . --r--- m7 , ,
i . b RFL 9428 a) Tube 18 1400"C, f 300 s failed s_ . _ . _ . . . _ . . . _ . _ . _ g- - l
' t A .
't. -
RFL ! b) Tube 22 13000C, 550 s intact
. _ . . __.y.,,___._ __ _ _ _ , . . . . . . . . . . _ , , _ , , , ,_ _ . _ ,
~
. ,. .M 2
a .t % . 0 , p % ; l M v ' 't t 1 RFL 6fa27 4 c) Tube f+ Black preoxidation 1300 C 550 s intact
.y n g . . ....,;,-,- . .. . .. ., m
, .,-: y g , , .
. v. , g
- - - - - ' - - ' ' ~
? .L'
'*)'h-
, , . - , (g , -
1 N x ,, c , . . ::a,c n,c ;mhwa '54: Hec b' r.N.%2 a liittyhM,55$' r'U:(G;c572.c3 Lv4)fki,W5M%$%'/Zhk0iM5 *;Ae RFL 6f36t d) Tube 6 Black preoxidation 14000C, 200 s - 1000 C 600 a failed 1 Fig. 7(a)-(d) Post test condition of oxidised and querched Zr-2 tubes i 6 I i __ - 4 - - - - - - , . . . , - .- _ ,- .-, ,- - - - - , , , , . . . - - - , , --
, . . , . - - g.--...,-
L : M.. h f
- r. . s _
*r* g g .
e _ b Y;.Lb
' u _
. t _
$;,Yi
.W j . d
' J .
e
. h W,
1 . c
,, n c .,$ f $ i e
u
* "s q
- y ..
i; d q . n a:. E g ... a
' . d kw ..
e f
. i s
. p d i
' d t x e c .
. t o l
i [' '> k t a , c a d a n , t e n
- f i ,
s t . i i s t a d 0
, h s i i-
. 0 3
b* d
%1 0 x
o
$ e 6 e 5 h r 0
e 0
- t. C
, p
' .. 0 0 w P y 0
0 3 1 ,$ W 0 0 3 1 0 0 3 1 l r a
' l
' u 1
e k $i 3 e
. , 2 e
d o
?[,., $ ,
b T u fh e , b T u b u f n f$
, ' T o fs n
,h ) $Y ) ) o e f - g i Wk i
( f$ h ( ( t i
*. .v [L h 4- d n
o c V f
.h M,.
Y ^ t
- I s
- k. "
e Y lE*.
$ h'S e s
t t
- h. .,- s W">
f; $ h Y b T Y-P o t b .
)
fk f h. a ( (
)
g 9 8 0 e 2e 3 36 ( 4 . G 9 0 7 . ig E R E F R k t l R i j j i i l;!, {!! t !.,1iIjsl!!ii k!: !
\
SNL TESTS INTACT FAILED 60 PIN SIZE VIRGIN METAL 0 8 36 PIN SIZE VIRGIN METAL 0 60 PIN BLACK PRE 0XIDATION O & 60 PIN NODULAR PRE 0X10AT10N O 'J N 60 PIN HIGH TEMP PRE 0XON 30 - 25 - M
# TUBE 2 r5 C
g
$ 20 l
2 e N is - gf- -- s W is -
! O o
E o $ O O 10 - 0 O h 5 - O 1 1 1 I I I I 400 600 800 100C 1200 1400 1600 ZlRCALOY TEMPERATURE (PEAK) F(G.B. SNL EM BRITTLEMENT TESTS
C 1000 -1200 1400 9600 1 i i i l SOURCE No.0F TESTS O ANL INTACT GUENCH TEMR 8 e ANL FAILED - 7 O GE TTE INTACT a 24 5 GE TTE FAILED " 1-a FLECHT INTACT PEAK TEMP. 22 T IDAHO PBF FAILED " 2 O SNL 60 PIN SIZE IN FACT
- 21 30 -
4 SNL 60 PIN SIZE FAILED - 12
-l- SNL 36 PIN SIZE a. 2 ,
i < h25 - u
< e w
TUBE 2 ' ** w 20 - y I
- 18 O' * * *- ,
17 - - - - - - - - - - - - - - - - - -C - - -' s ^ 5 - O e 15
' O O
w @O Oo O I C) O O
@ 10 -
sh0 a 0 o o cccp a O 5 - o'O g ao O a a 1 0 t i 1 mn 50 i er e 500 ?.000 1500 2000 2500 2000 ZlRCALOY PEAK OR QUENCH TEMPERATURE,*F FIG.9. SNL AND USA EMBRITTLEMENT DATA v -. - , - - , , - , - - .- - ~.. ..-.,,,.-~..u...,,, , , , , - - .....,v,, ,e, ,e.,,, , . ,, - - . ~ , - - - , - , - , . ,
; .c: 7:': :. ' - t A[* * '
. , lf ?%'Y{M' 7 ,
l *5 - Y '
, ' , j, ; - 'f ;)$[kl?, g, l-Q : ~ '
Biel+58 x f+15 D3998 x 83 a) Tube G cracks in a phase tube survived b) Tube 23 cracks in av incursionn tube survived
' 'n s p E1~ k:.='
~~~~~ t
,DT*W35 347//..W- Ar,4, - i g .,"
sp J y
*p g ((hk h- ,4,5 i
- r-v gg.. a
- agv *- ~ ' -
. A. .
; . - - . g.. T M-g34~Q, .. ,. 'N.
- d *~
.~'
se . s, ,
. ' , ~.
l
'-% g ' '
f' -t , h ,'f,;' ~ ..i %g.y> u.. . - Q d} ,, e
- f N'..xPJif>r..@<
~
i
&,ypCh4M..
e i pg.C_.;,. s
'4 sw? h y. j1 .
. . , . .y ., - w. me
, I E- 1 x 900 x 's10 c) Tube 17 fractured oxide and a nurface d) Tube 5 exponed c, nurface tube failed Fi g. 10. Morphology of cracking in stabilised a phase __ __ ___j
1
~MF 1 1 ,
I ' f4 . ' u s , , sg.r mg . . . . e; , s . ,a N
,- y --
;ps - *
~ , . . .
. j,
,s ,
't.. '
.Qs, ,, ..v * -
,% $ - 4, 3 .-
f-d * [,- h# . ,, ,, % ,' '% * *-" f:^ V : ppqh
~
~. _,..
.j;Qf. ,,',,.;L
. g .. . .
/. ' 'h.
s
** $f ' '
?, i:: Y: ' "
b 4.i{nL;.z.- a.ww, < , y,.:.. ?. l. . . ,.
. .. .^ .
~ ., .
.. h,Ej.y> .,V. ,Wy, *l *;; ? ?. ? ^h'.... : . I ' ' ', . .
- J :.
-c g gy -
4' 7.: . _y 4, : i ..r - ,. _, ar".
... s
., , . . - -. s 1
._ .:' . {
,./
s
.. .. .) _ '
. \
. 7 -
107 TEM x 86000 Fig. 11. Transmissien electron micr graph showing the structure of the outer oxide film from Tube 18 4 l l l I i
ns
+
1400 - SINGLE S10ED OXtDATION DOUBLE S10E0 OXIDATION 1300 o 1200 1200*C o , 1100 - u B 1 E 1000 1 u, 900 - 800 1 10 100 1000 10000 TIME, s FI G.12. IS0 THERMAL OXIDATION 60 PIN MINIMUM WALL THICKNESS (0 0597 cm) TIME T017 % OXIDATION, BARE METAL 1
, , , - - --.y_, , -n.
700 - TWO-SIDED REAC90N: 0 0686 cm ORIGINAL WALL i AL ONDITIONS: 600 -
% FILLING (jg ) r 95 % SATURATION AVERAGE 02 CONCENTRATION (4) 2 0 7wt %
500 -
% FILLING CRITICAL
( (9 5 %) fE L00 -
\
6 300 - I 200 - AVERAGE 02 100 - CONCENTRATION CRITICAL 0 ' ' ' ' 1000 1100 1200 1300 1400 1500 T EMPER ATUR E,"C FIG.13. CRITICAL ISOTHERMAL OXfDATION TIMES AS A FUNCTION O? REACTION TEMPERATURE. TWO SIDED REACTION OF 0 0686 cm SPECIMENS IN STEAM. CRITICAL CONDITIONS FOR OXYGEN IN REMAINING BETA STIPULATED AS EITHER 95% FILLED OR 0 7 wt % FROM PAWEL(30) AVERAGE 02 0 60 CONCEN TR ATION CRITICAL (0 7wt%) qp 0 65 -
% FILLING v CRITIC AL
~7 0 70 - (95%)
d p 0 75 - E 0 0686 cm ORIGINAL THIC NESS
~
/. F L LI /. SATURAil0N l 1
0B5 AVERAGE Os CONCENTWAT10N (4 ) 2 0 7wt % S00 1000 1100 1200 1300 1400 1500 TEMPERATURE,C FIG.14. CRITICAL Fw(FRACTION REM AINING SETA. BETA I THICKNESS DIVIDED BY ORIGINAL SPECIMEN THICKNESS) . AS A FUNCTION F THE REACTION TEMPERATURE FROM PAWEL I30
50 EXPOSURE
' \ TEMPERATURE a1050 C en eTHE LOWEST
- 40 - o1150
~ \ LIMIT 0F5-PHASE O g ,
INCURSIONS j J , @ C1280 . u 01320 U ' j 30 - N -l-a- PH ASE
;- \C INCURSIONS gi O \\ '
BLACK P0lNTS
'ZERO DUCTILITY
$ 20 O \ SPECIMENS sn 4
5 C u \ a 30 - i gg OUENCH LOAO ~ 70lb/in N 17 % FRAGMENTATION LlM1T
\g UCTlLITY 0
' , SPECIMENS 0 10 20 30 17% EQ.M.R.
FIG.15. RATIO OF OX10E THICKNESS TO ORIGINAL WALL THICKNESS (%) RESIOUAL STRENGTH AGAINST AMOUNT OF OXIDATION FROM KOMATSU G5)
1-
. ATTACHMENT 4 .,.'
, 1 I
JAPAN ATOMIC ENERGY RESEARCH INSTITUTE TCKAl RESEARCH ESTA8USHMENT TOKAFMUR A, NAKA*QUN. ISARAKbMEN i i Inner surface oxidation of zircaloy cladding ' in a loss-of-coolant accident. i 3-
~
i l S. Kawasaki I g T. Furuta l H. Uezuka I I Fuel Reliability Laboratory 3 2
)
.I 4 4 i-i 9 3
In the current safety analysis for a LOCA of a LWR, the embrittlement of zircaloy claddings are treated to be caused by absorbed oxygen exclusively. This treatment is based on the experimental results of zircaloy-steam reaction performed in flowing steam. However, the atmospnere of the inner surface of cladding after rupture seems to deviate from flowing steam condition. In the burst test of a simulated fuel red in flowing steam, we found 3 I abnormal oxide formation on the inner surface of cladding which had been ruptured and subsequently oxidized; that was thick and porous exide famation. I We assumed that this pnenomenon had occured in stagnant steam atmosphere, so ! l performed an oxidation experiment of zircaloy tubing in stagnant steam condi ti on . By the experiment, we deduced that hydrogen generated by the I i zircaloy-steam reaction played an important role in the abnormal inner j surface oxidation. Then, we conducted 'following two experiments. $ i i i
- 1) Oxidation of zircaloy tubing in flowing hydrogen-steam mixed gas. t Since the abnormal exidation phenomenon in stagnant steam was assumed l R
to be caused by hydrogen produced by the reaction and not to be taken off from the atmosphere, it can be deduced that the phenomenon may also occur in 2 flowing hydrogen-steam mixed gas. Then, we performed the following k experiment. , Zircaloy tubing with length of 15mm was heated in flowing hydrogen-steam , mixed gas. Af ter the reaction, ductility of the reacted specimen was measured by ring compression test at 100*C. When hydrogen fraction in the mixed gas was less than a certain value, oxidation behavice of the specimen was almost the same as i.n steam alone. , At a certain volume ratio (i.e. H / steam 2 was about 0.4 at 1000*C and 0.25 at 1100*C), the oxidation behaviors enanged; fluctuated weigne gain, severe I t i
}
4
3 hydrogen absorption and embrittlement, change in crystal structure of the oxide and porous and thick exide formation. 1 l
- 2) Oxidation of burst rod in flowing steam. i 1
From the experiment mentioned above, it became clear that the ratio, hydrogen / steam, in the atmosphere had important role on the behavior of , inner surface oxidation. Then, the degree of abnormal oxidation must depend on how easily the ratio. hydrogen / steam, in the inner surface atmosphere i reaches to the specific value. So, hydrogen generation rate (zircaloy-steam -, ] reaction rate) and the exchanging rate of atmospheric gas by steam flowing outside influence to the phenomenon. The former factor is controlled by ' the reaction temperature, and the later seems to depend on the steam flow 1 rate of outside, the rupture opening area and the distance fecm the rupture opening. Then, we studied these effects by the experiment which was burst ; and subsequent oxidation test of a simulated fuel rod in flowing steam. l i The simulated fuel rod contains alumina pellets and internally pressur- L f i:ed by helium gas was heated in flowing steam. During the temperature 8: transient , the specimen was burst, and steam invaded into inside of the 3 ; - cladding through the rupture cpening. After holding a given time , several i specimens for ring compression test were cut from the fuel rod. Ductili ty . of the specimens were measured by ring compression test at 100'C. The reaction time which caused severe embrittlement of the cladding became . i short with increasing reaction temperature. The position which became brittle separate from the ructure opening with increasing of outside steam , flow rate and the area of rupture opening. Also in this experiment, main cause of the embrittlement was found to be absorbed hydregen. From these experimental results, it seems to be able to concluce tnat hydrogen as well as oxygen influences to the embrittlement of cladding.
!l '
2 i l i i I
i 4 g
-,.,_,...,,,,.mm._,,,,,,.m_,, , , ,1 .
t 4 f i } t Dense orade
^
. ' . . .r; - ,,1 ..
i Stat). tired .r." - , 0 - d Transformed ! phase-1 Porous Omide . j ' i,i g -
- ,.v c..
6 d..:.y>
......s 1
fJ 1_ g #.j.!!:ty:idiN
- !; F- - ~ :-
4-r- ...
' !.-Y2 < ~
i" pQ .. . . 7 Fic. I cross section of a simulated fuel rod burst and oxidized in flowing steam. (burst at 870*C and held at 1000 *C for 7 min. steam flow rate : 0.f89/cm' min) - t e e
3_.- _-. . .... _ ._ _ _ _m . . . _. . . _ . . _ . _ . Source C SCR "I" '"- i Controller y 5, 5 I T. C . ' Sample 1
- - qu rit tube Water _ -
Tonk "#~ t u~ 3-y a O Rotary Pump Recorder Y $ T.C. J r Volume of fest section : about 280 mL Fic,2 Apporolus for zircoloy-stagnoni sieam teoclion 1 I 4 O
. _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ __ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - m - -. ._ __. _ _
l 1 Arnount of oxidation calculated by Saker-Just's equation (',') i l l I 5 10 15 20 l 1 I l 1 I i 4 - 3 min. i 115 C '"~ I ficwing i i 2 - stagnant : ,1 i l10 min. , I. I
\ df 1 0 ' '
S 10 15 20 - i ' l 3 min. l i 4 _ flewin$ 11100*C -' 1 10 m.in. i stagnant i 15 min. 2 - i 20 min.
! I l 5 10 15 20 4
10 min, j i i ) mc _ i i e. min. 1,0 5 C . ;
+
4 l I I 20 min. - - 2 - stagnant 1
, 2 s i
. !_____- ! . I I
c0 l. l 3 5 10 15 20 G i 8
)
3
$, flewing i i
* - l _ Om'in. 1000'C '
O 3 i l
, /20 min., 36 min. t
- l 2 _ stagnant i 4 1
l I I l I l l 1 g .#u.m. . .. 5 to 15 20 i 10 min- i i 9 5 0 *C
;{
4 - i f!cwingl I !
; cxidation time'
, i i 2 - 20 min.I : o S min.
stagnant l :
- 10 min. i
- o ,. . . t i . r i 0 1 l
0 5 10 15 Weight Gain (mg/cm ) 2 l Frc,3 Ouctititi's of SS*cimens oxidiz'c in stasrant and ficwing steam as a function of weight gain ! ( Ring compression test temp. : 100 *C ) l 1 I I I
- )
J l
i l' I l
< Mid Wall Hardness 1 1
e o o o e ' o o o o m O o e e m N - 4 4 i & 4 e i e H
@ 8
@ ydrogen Content 8 (ppm) m e o
. . . . . . . 4 i . . . . 4 4 , 6 6 m P
k
. O ,o
- o
- z
,' ,/ - t
, o
./
d- * " I l u 5*
~
l 5 L p - ,/ / j '- .O.- .C y
/ f - X *
. ./ .' o o
.t o i ~G m / A *e l.
E, S i (
- *"'--D---..--.-.h,# g. Og [,_,
/ k h
- . n w .
i
/
) 7
' c e o -m m u<
t r ' [ = < e ( - ,* o i 6 > w mc
/ -
E N w
/ i =m x 7
j '
- i i
l 3o >
- Z E e-o o ;;; r
*/o a: D >
i
/ i - + o<
i _. o < e e- w e-
- i o, -
> m
, - w os;z i
- i c i i o <w e 'e o< a v e o ao r - o-t .? w el Ii oc
> -e N Q F
P - 5.5a s e o
~
W
- o
- u. .t.
8eg t e E 3
- u. m z o e-5 n. v -
- 5 D@O u > I w < u.
- u. =
- u. w z I t i e i i i , ,. _ w m-i i ,
m o e e o o e a _
'(gwo/Sw) v!ec tyS!ag o .
s
.e I f . . t I i 1 t 1 o c? o. C. o. O o e m 4 m. ..m a
. s
~
( w w) uoltoeU+0 *- -
~
i i . t e
,- , ,-,,,e o.--,~,,. e , . , - , -, ,,-..,-w , , , . . . , -,a .w--,--an- , r, ,
4 . l r i o o o 6 b o
- .=
'o o.
o o b Y .d d$ O bo $ _;EEEE TE o o. spmog ag 9 * . o E E
- d a 3 E E c j 2 II
- v en ,
_b c o e 2
; a s
*J I
7
=
e o-
= < 4i ,a
.g t
x
.=
o - o T o -
- o. n o \
oN C l f l O x ws y q _ n >= E o
' N o s
) ,,<42 l
.% [
l s$ U
.E 1 ,,
x .c <nx .3 3 7 x Av e { < c .? e x <> u>x : . E _Le p Sn d '
; e o <,y g mx<3 o a
[ n - p E
,.f lx T o o y f ?Y s 'fif d j c 8
> :o
= o
.e, r O M o O d
e e oe e o e oe o , (,w o/Bw ) u!o5 JyB ! aM . d e 4
___.-...,..-,,..,--;-2... ,
., _ . __ __ ._ -a o oo 110 O*C 4 -
o a-a a o 3nin
-x a 5dn n '
=-" x tomin .
C
-o o- o 15 min O ' I ob B '
'4-E ^ ^ ^
E4 - 6- *- ' o to s o*C ^ *'" x iomin o g- o 15m,m p 2 -
-8 - ,*
, 20 min 8O e
- I M=# '
i e O oo - -Q c_ _a
.7 a tO O O*C 4 ~ a S min 8
-' ; 8Ma
]2
- ~$-* 8 8 t
- 20 min c0 ' '
- 1 Ew@ e' Q
~^ O ^ --- a wa 4 -" 9 5 0*C a Smin
-5 x "
x to min o a 015 min 2 _ zg- _ g- . N ' 20 min O 0.01 0.05 0.1 I
=b-05 1.0 50 Volume rafio (Vuz/VHzo )
Fm. 6 Variation of ductility of oxidized specimens with volume ratio (Vnz/Vg ,o) in steam hydrogen atmosphere. i
-= _ _ _ _ _ __-_____-______________-_____ma____ _ _-_____ ___
l I I I 10 0 - I 1 1 n 7 E
/
e k I a 50 - ){ 1 2
! 4 ]
2 5 11 cot 4 min. i
/ steam flow rate : i 3 0.66s/cmimin. !
r s , t r i O' 1 2 3 4 5 f Ceflection (mm) ! lower upper s
+15mm-*
f ff ff 0[ 0I O
<=::> i t d( )( Vl Vi ,d
' {
1 2 3 4 5 6 7 Fro.7 scNmatic itiustration of specimens used fer the ring ccrrpressicn test at 100'C
d f _ I i.. [* t fn - g 8 i o- , 0 6
- t *
- I i
I
,' /. #
/ '
, , / P' P
* ;- 4 5 U "7 ,' s s
6s . n o s a\
~
. f ., j 0 3
i
.i n t m nu c
s 4 f r a r ( o r s
\ a d
s e . _
' z n
. S A
i i d m
\ e) i _
Oa t . 5 ) x 'm _ an o c/ w Ri m wi 0 0 0 0 0 1 ( mdn 0 9 _ om 6 6 6 6 6 g a 6 _ x-l c n F/ 1 1 1 1 1 d 1 q i n e g _ m( a e r f u o n p _ e o t i _
. t S
p e n u t e 1 e, e r a p _
. e 0 r r o _
) u s _
p'.C 7 6 1 9 6 4 - _ t p g w e n ol r 1 5 0 6 u u _ m( e 1 1 1 1 1 0 0 9 1 r i d f t T d p
. a a h u
, i t r m c i
.- o w a
- r o
= f t
e m m _ fa N e 4 3 2 5 y 5 1 e c u a o r et fr 1 n r s
- n. i p 9 9 9 9 9 a o g e .
m a
" a o
- a t s y n n c
S - - - i D t i a
- - - t w t N
- i o s _
i l i
- - t f d c
, " o o e a u n f
,d t oi o _
0 g,I g, t e4 1 I i 8 _
/., ,
o i F _
.,\V'i
,g
\.
5 4
\
- l _ 1
.\
f , 0 . _ 6
\* -
s 4 3 2 i , t ii, , i1 ; 1: ,4 : ; ' !.il { i l i !! , . t1'
,1 -s --
m_a a ~_.. _
]s .
I,
!- _ g
%.w.N %
e5 k
'N*N s -
p ! N W s N
==-
OI M , w wm
- e. -
\
e a Eg
" c
,c
\/ 'm u-i "
Ch 5 C @ m \ C
" G
$ .~. k To .C r s=-e o
; * .; - T- e I I ;*
l - et I c
\ D Te i . c a
- C
. T.
p [ t=- l @b b '
^ 33 C kJ = c c O W w Q ,
cs - e-
- n-e .c 6 <
I * = = - - l 3 =w b l ed e e m O &
'c 4'
= am e
= t b @@ 1 1 sc 1 I gw 3 O , e .== en I i l r
t I E m
"* = l I b W . e g l C
i .
.[j 7 @
v o o i
=
C m o a / C m w$ D l y
- O "
/ $w o= C
{
- OA / \ D .e 1
/ A c
-(
I e 6.- em 04-u . 1
/
/
SUS '
/
/</' &
i
. o G- <. -
- p. [u
/
e 3 - e 1 I i / , e i e , e ~ _ o aO G v .-,,.-e, , -,- - - - ,.- , -,.-,,,w, - - , - - - , r-. - - - . +,,,w,,--,r .--,-,1- . . , . . ,,.e-. ,.- .--.,ev, +. . - - - ~ . -
f l c 8
~u
\4 E I
\
i . 4 -
\ .
e 4 E I e E 3a
<: - en e
*- 9 """" 0 4 E g
.i b
?
e M. .
=
e c e0 - c /o '\ e - F ~ 5 l
'a<
'r ~ ~ ~
<g 5 2 e
m e f ~* E a & E 3 *
- T yN-e e
e o ge- e o
.e
> - ,u a u / - =
< o e = .
+ c u t c c c 5 m 2
o E E E o < .. - c. t
- ~ ~ ~ - . 2 i : ~5 5 5 o 5
. 2 e ~.
e
~
m E ., c e en e- - o e '8 a e en w c:* e = - - c J == - o o o * = c
< v ~
e v v v v
- , e &
o o o 4 e .e., e, e c> c> c
- o .e a, u 3 m e w e -
4 . o w a C N N N g. '
~ 3 e v v
v 4 % I =e e e w o o .n o w g g o < - a - 5 u a o e e ap
< u a o
4 5
. o
- o t 1 I i e t o .
c* c u, e o. o. o. u. ,e,,,
. e = m N - e u (m) D 00t at uopoetjeg a
a I , ~ - - , . - , . - .
,r ,e., , +---,.,-e ,,-e .--n, , , . , - - - . e , -,.--n-c-,-e ,- , . - - - , . -
'V z
,o I '
$ e I m m
I % 1w L
' l I, i e l
E , of 8 2 . t c E [ l l I w n l' I s ~ + l b 2 m
* ~ s c
I ,g *
- I g a
- g a
! s l' e'* mO u
os l > e d? .o . I ~ 2
! i g I >
l . e o ' , l , 2 m
' I
.I: l
, 1 $E4
~ h*
R y i l I
? ;
, ' e p
- 8
- ess.
d l o $ .3-Cg l ' i. u ~ e s ' l f'
. f, e
L *D i el 8' 4 ~ I I I p o& o i - 8, e t .
~ w ! l l $sI ' fe 5 l ,
f ) t l
.t
- g;,* .:
?u t' 3 o 'g
*, e
- E-1 X '
e r. $ 5 fq, e y I ~
,1
// -l-i I
R I'
~
i
?, yt*
c
./ o ,,
r *. e : 8 a [j ,i t o
./ /
o p
- *. *I N
3u g
. , Se .o M. _. e* .
e
- I.. -
1 e 1 ~ g # - . l I ,t 9.e e y o o* O
- o. a a
; w m ~ ~
(uu) 0*001 2P uo'238 u eo
. . , 1
-[ ,
ATTACHMENT 5 ZIRCALOY- CLADDING EMBRITTLEENT, RECOMMENDED CRITERIA
- by T. F. Xassner, H. M. Ch:mg, A. M. Garde, and S. Ma;*wndar Materials Science Division ARGONNE NATIONAL LABORATORY Argonne, Illinois 60439 FOR PRESENTATION AT THE SIXnl WATER REACTOR SAFETY RESEARCH INFORMATION MEETING Nat.ional ~ Bureau of Standards .
Gaithersburg, MD November 6-9, 1978
- Work supported by the Division of Water Reactor Safety Research, U. S. Nuclear Regulatory Coc=ission.
ZIRCALOY CLADDING EMBRITTLEMENT, RECOMMENDED CRITERIA
- T. F. Kasaner, n. M. Ch:mg, A. M. Gc.rde, and S. M:,iwndar Materials Science Division ARGONNE NATIONAL LA3 ORATORY Argonne, Illinois 60439
SUMMARY
The mechanical response of Zirealoy cladding has been evaluated under thermal-shock conditions typical of hypothetical loss-of-coolant accident (LOCA) situations in ligh t-water reactors (LWRs) . Integral tube-burst / thermal-shock tests were performed in which cladding specimens were ruptured in steam during transient heating (10 K/s), oxidized at maximum temperatures between 1140 and 1770 K for various times, cooled from the oxidation te=perature to s1100 K at a race of N3 K/s, and rapidly quenched by bottom flooding with water at a rise rate of 0.05 m/s. Specimens were also quenched directly from the oxidation temperature, which resulted in a cooling rate of N100 K/s through the 3+s' phase transforma-tion. These experiments incorporated thinning of the cladding wall by the for=a-tion of multiple ballooned regions, oxidation of the inner and outer surfaces by steam, transformation of the central S-phase region of the cladding to an s' structure in which the cooling rate influences the extent of oxygen redistribution and the microstructure, and thermal stresses produced by flooding with water. Failure " maps" for fracture of the cladding by thermal shock were developed relative to the maximum oxidation temperature and various time-dependent exi-dation parameters, e.g., equivalent-cladding reacted to form Zr0 2 , fractional thickness of the transformed 8 layer, f ractional saturation of the S-phase layer by oxygen, and thickness of the transformed S layer with less than a specified critical oxygen concentration. For cladding that was slow cooled (5 K/s) to s1100 K, to simulate the effect of radioactive decay heat of the fuel on the cladding temperature during the transient, failure of the cladding by thermal shock occurs when the equivalent-cladding reacted exceeds 28% .of the cladding wall thickness, and the amount of allowable oxidation increases as the temperature decreases. For cladding that was rapidly cooled (%100 K/s) through the S-c' phase transformation, the thermal-shock failure boundary corresponds to 20% of the wall chickness for oxidation tem-peratures (1650 K; in good agreement with previous data which for=ed the basis for the present oxidation limit of 17% of the cladding wall thickness in the acceptance I criteria for emergency-core-cooling systems in LVRs. l l The best correlation of the thermal-shock failure data with parameters that l characterize the degree of oxidation of the cladding was obtained relative to the l thickness of the 8-phase layer with less than 0.9 and 1.0 we % oxygen for fast- l and slow-cooled cladding, respectively. All cladding in which the chickness of the S-phase layer, with less than the critical oxygen levels, exceeds 0.1 mm remained intact irrespective of the wall thickness, oxidation temperature, and 1
- Work supported by the Division of Water Reactor Saf ety Research, U. S. Nuclear l Regulatory Commission.
i l 1 l
total oxygen content of the cladding. The difference in the thermal-shock failure properties of slow- and fast-cooled cladding can be attributed to the extent of oxygen redistribution and microstructure of the material since no systematic de-pendence of the wetting temperature on the maximum temperature of the cladding at the moment of flooding was observed, i.e. , the magnitude of the thermal shock was equivalent for cladding that was cooled through the S-.c' phase transformation at different races. 1 In-situ pendulum-load impact tests were performed at room temperature on ' oxidized Zircaloy cladding that did not f racture during the thermal quench. ( Information on the total absorbed energy f rom these tes ts was correlated with more extensive results f rom instrumented drop-weight impact tests on homogeneous subsize charpy speciseas, which contained 0.13 to 1.2 we : oxygen. A finite-element model was developed for crack growth in oxidized Zircaloy l during thermal-shock conditions. The model has been used to calculate the maximum l axial- and circumferential-crack depths in cladding from the integral tube-burst / l thermal-shock experiments. Measured oxide , 2 , and transformed 3-layer thick- l nesses in the tube, computad oxygen-concentration profiles across the tube wall, I experimental data for the dependence of the crack-initiation energy on tempera-ture and oxygen concentration, and other material-property data for Zircaloy were used in the analysis of the thermal-shock behavior of the cladding. The thermal-shock results indicate that the present Zircaloy embrittlement criterion (i.e., a total oxidation limit of 17*. of the wall thickness and a maximum cladding te=perature of 1477 K) is conservative and that a more quantitative cri- ! terion, based upon the mechanical behavior of the oxidized material, can be for- I mulated with a specified degree of conservatism consistent with the =echanical l loads imposed on the cladding during reflood and the max 1=um amount of oxidation l set by the margin of performance of emergency-core-cooling systems in LWRs. Recent Publications
- 1. A. M. Garde, H. M. Chung, and T. 7. Kassner, "}iicrograin Superplasticity l in Zircaloy at 850*C," Acta Met. 16, pp. 153-166 (1978). l
\
- 2. H. M. Chung, A. M. Garde, and T. . Kassner, " Development of an Oxygen l E=brittlement Criterion for Zircaloy Cladding Applicable to LOCA Condi-tions in L1ght-Water Reactors," Preceedings of the 4th International l Conference on Zirconium in the Nuclear Induserv, ASTM, T. P. Papagoglou, Ed., 1978, in press. ,
l
- 3. S. Majumdar, "Modeling of Crack Growth in Oxidized Zircaloy Cladding during Thermal-shock Conditions," Argonne National Laboratory, ANL-78-44 )
(NUREG/CR-0136) May 1978.
- 4. Argonne National Laboratory Light-Water-Reactor Safety Research Program: l Quarterly Progress Reports, ANL-78-3 (July-September 1977); ANL-78-25, NUREG/CR-0089 (October-December 1977) ; ANL-78-49, NUREC/CR-0201 (January- j March 1978) ; and AML-78-77, NUREG/CR-0423 (April-June 1978) . I l
- 5. H. M. Chung and T. F. Kassner, " Deformation Characteristics of Zircaloy l Cladding in Vacuum and Steam Under Transient Heating Conditions: Su==ary l Report," ANL-7 7-31, NUREG/CR-0344, November 1978.
. ZIRCALOY CLADDING EMBRITTLEMENT I INTEGRAL IUBE-3URST/IHERMAL-SHOCK IESTS.
O CORRELATION OF IHERMAL-SHOCK FAILURE DATA WITH SEVERAL OXIDATION PARAMETERS. 8 IN-SITU IMPACT TESTS ON 0xtDIzsa ZlRCALOY CLADDING AT ROOM TEMPERATURE. O INSTRUMENTED DROP-WEIGHT IMPACT IESTS ON 0xtDIzED-HOMOGENIZED 3AR & SHEET SPECIMENS AT 373 TO 823 K. 8 MODELING OF CRACK GROWTH IN 0xtDIzED ZIRCALOY CLADDING DURING IHERMAL-SHOCK CONDITIONS. 1 I .
- - . . - . - . . . . x. .
i . L l ( i I I I I I I*I I I I i I I I - ac0 ' OtART SPEE9 M s OWQ y\
~
33 q 2~ ' 1500 - 5 ! O-*I E U i L ~ 80ITOM-FL000fts . [ 11 TC # 2 VALVE OPEN 1400 g ko f, TC # 6 ~ i m ti m-g i m 9 -
~
DIRECT. I, g* ILOG0!rEj g'd
.y TC # 4 FAST-C01~ ~
~ ~ - - ~ -
~" ~ '-
IM o O
~ N CLADCING BURST 000 KA) SLOW-COOL
\
_ g$
., (5KAl Q N $ HEATING RATE 10 Wh T1uE W d
, INITIAL PRESSUNE 6.9 MPo SCALE i TC# 5 OiAEE Gl l l- l T M ,i 3. l l. l l l lg l l ! ! . 0 12 0 240 360 480 600 720 960 1080 ll20 ' l14 4 116 8 1l92 1216 1240 1264 - TiuE (S) . Fig. 1. Temperature vs Time Curves f rom Tube-burst / Thermal-shock Tests That Show an Abrupt i Change in the Cooling Rate at Temperatures between s750 and 900 K. -Cooling rates through the S-+a' phase transformation were s100 K/s for tubes that were flooded at.the oxidation temperature and S K/s for tubes that were slow cooled to simulate the ef fect of radio-i act Ive decay heat on the temperature' transient. - e l
l 464 i 4 4 4 i . A g6ei*i 6 jii e a 4 4 6 . iii6 6 3 , , .
~~
o o Oo o - k5b eo-E g 9 k .*, yu I o we l uu 6 3 o6 E
-8 o o@o 8 o
e
<xa M o0-2 4- H "3 4 -
o e e - . i u-u o===
=ao <
oo8 co o6 l
'01 o >.o m a = = oe !
o=o
. 9oWoe 9 8 - =="~
-mc: < e y _
y v g1 o 4g o og c oo - . c..
-w.w_e=
u -4 A = cm.
- w> w.
I - os oo o e oo Omwg!qEu 3.
!! g ,l e+o .
g _ e=1 . *.o 8 ng s yo
-o o
2
. . . wom eu == e. o o - .
ose== =-2sou o= 0
. g
. .g;;a .
- 2c -
>,,=-u uo-
.I
- g. .. ee. , < &. e e,o c s-o 35] . . =ce C o o
- .c ,.> I e. o =c, 2
I - o..
. oo
~~
- =wo.mu-,
uw
= mm -
o e o= = c e l we @ 8 . A o e ** O : C - w '"l:' w a-n m ==m
. . . . 2 .0 ,- g. -: uM3<t 3=~ou C W % =
e c a .
- a. . u
.. * -u u un
,,,. , . . l,....
. . l...., . ., l,,,,,., ,
uo
,, . c o m _m, ts o .
, .sztg ,~g
- ! on -a e 3 e o e n
.. no... ..u... . g 8
I I I I I l 4
. o g -
u,
<=
O og0 o o O x ow
~
o o Cg ~ O g m -c e - e > W o O N oo o dW
-a E
o oo o o o o o o hx o
~
-5 n,
o o o T o g w u-
~
o o o ~ ww h. o T O
$=
- o k
o o A ** w o 0 - fW M
~
W o. e2 A 3 o o C z 00 0 2 *$ 5 O 2 *W 22*
~
3 o o C 4
- co o O _Z_ w$
-d o 00 e oo
- 9 o
~
O oy e= 0 wa o cp 3 o$4 Q Jk =3
- oo og - U
&H D *o o o co o =,, .
- o O o oo 0 -
b o $$3 p-w
=a e O
O YNO 25-m a A u .=.s e s e a a O vn i l l l l I 3 m O an O O O O O O O Os m3 O O O O O O O O +
~ O o e r. e o v N 9jl
.: o (L) 3Mn1V83dW31 DN!113M " ' "
w
OX t0ATION TEMPE'1ATURC ( Cl 5 1500 1400 1300 1200 IK)O 1000 900 10 . g i g gq g g i
~
g jq , q tJUMDER (i CL AD0ltJG PtECES - s -- AFTEH IRACTURE . 4 2 3 4 5 OR 120f tC
- ritACTuf tE IN NON- , , ,, , _
DAL t (X)Nt D ItEGION _ g 2 - INUHE pi , , , , 8 -
-4 UALLOOtJED ltEGION i! t
- J
- c. =
ac o 8 ' .a. 4 _ _- its ~ e3 _ i.2 TC #5 s2 m
.+ * -
8): F 10 3 o
.o
*e. -
hD ,.,l. ' *~ 7 o _- p : u ee : . i!E cn 9 4 - 8 - a- % ,t/n1 g3 -
.T -.
4
* -.' ti 2 ,l -
#3 e w . 5 s:
a ,J e
#4 :- . d s -
,i : > -
- , 4 - . .
3 - , _.
-: - #2 lgi v +
2 - - 3 :- 8
.~
i 'O'1b5 __1___L___L s.O ss 7.o l . ._.._. I 7s 80 l__ os
-s m s / T a lO* (*Kl '
ss Fig. 4. Zircaloy-4 Cladding af ter Thermal - Fig. S. Representation of Thermal-shock shock Failure. Fracture (Ballooned vs Nonballooned Region of the Tube) Relative to the Time and Tem-l perature of Isothermal Oxidation af ter Rupture in Steam ($1040 K) under Transient-
; heating conditions. At .cmperatures (1570 K, fracture occura in the ballooned region.
i
~+
Omit >AflQN TEDAP(HAftsNE l*Cl 1400 1.500 1200 t 800 0000 I I u {l"--{1-~{1] ] 6 000 g . 30* Co
/
\.
g e . . Co
\ '
o Co n - i2
- 4 -
3 3 - BEST EST* MATE - , - poemtv.nv ttwe o 2 ! 2 - SLO *-COOLED o - I 1 CL ADDwaG % o e AeAREA { $ e At 4 A2 = A3 + A4 + A5 + A6 + A7 ! h to' ~ - e A6 + A7 s A8 - q e - I. N. s g 4 h 4 . kO _. Ca I - e ECR* ICO -X O 3 + IX T -I I I - i 10 0 h t-3 .. - X g -X O ! -N 2 - * - O a 5 -
*f .
o r, . ( * -(A .
, i
*i
- D' Xg -X o 10 a uu * ,
a us o - C o M
- s a ,
e as - Ca -- -
- FBS a l '~ - ~
Cs j 6 - e O o o FAST-rOOLEO AND INTACT '
4 . O < 00o e FAST-COOLED AND FAtt0 _ a t.CF
- II~IO
- i 3 -or Cs in o -
3 3 j 4 _ _ i = 2 - o --
- .- - .G..
- h. ,,N-
? 8 ,' ,,A ,.i / ,7 - - - --
C, I
..J.." .--_ -
3 C r es e~ 60 1 I I I I .. l 6.0 E5 IP 15 8.0 0 -" ' . o I 1 1 4 in . io. .c's x>x. t ,o t~ La Yo Y1 LaL1sx i x x,1 i i Fig. 6 Capah!Lity of the Zircaloy-4 Clad-- Fig. 7. Schematic of the Oxygen Distribution in l ding Tuhes to Withstand the Thermal Shock the Tube Wall and Definitions of Several Oxida- ) after Rupture, Isothermal Oxidation, a_nd Lion Parameters. i Bottom Flooding from the Isothermal Oxida- - tion Temperature. The best-estimate failure boundary for cladding cooled to st100 K at a rate of %5 K/s before bot tom flooding with I water is shown for comparison. I }
ISOTHECMAL OXIDATION TEMPERATUXE FC) 900 1000 110 0 1200 1300 1400 1500 10 0 i i ; 3 g g 3
; 3 1 o CLACOING INTACT g
90 -
.
- FAILED CN CUENCHING 80 -
. SURVIVED CUENCHING BUT" FAILED QURING HANDL!NG
{z 70 - e 60 - o E *
*o oo e, 8
8 50 - 3 . W 40 ~~17% LIMIT , # ,* e o , **.,*, T z I p e oo
- So w 30 -
o -
- E' 'g? *
- 9 h o oo 0 6' ,o , o a o 5;
20 gg o, g, , S ql_f_g3__,_o,,o,9g_d o* ' o o?'E , =9 8 o* o io _ o $ g '% f See o o 1477 K LIMif o'o n I* . I , I , l1 . ! , I , 1200 1300 1400 1500 1600 1700 ISOTHERMAL CXICATICN TEMPERATURE ('K) Fig. 8. Failure Map for Zircaloy-4 Cladding by Ther=al l Shock or Normal Hand 11rg Relative to the Equivalent-Cladding-Reacted Parameter and Maximum Isothermal Oxida-cica Te=perature after Rupture in Steam. CLADOING MAXIMUM TEMPERATURE ('C) 1000 (200 1400 1600 1800 2000 2200 10 0 , , ; ; 3 i , g l l l l l msT-ccett: l
- 90 -
FAILURE INTACT FAILED - O SOUNCARY
# go _ gcR stew _ THIS STUDY o e ._
y COCLED CLACCING HESSON ET AL o + g 70 - ~ SCATENA A A g 60 - b 50 - s 3e 40 - * + - 30 -
~-
o $ 8 ,- - - 7% LIMIT - g + s o n . . g 20 _ __ _ +t_o__ __ _ __ L __ , _ _ _ - 5 to _ _ . p; __ '__ ]o < o a - o W o - db 0 3'8 I I I 1200 1400 1600 1800 2000 2200 2400 CLA00 LNG MAXIMUM TEMPERATURE ('K) Fig. 9. Thermal-shock Failure Map for Zircaloy-4 Cladding (Bottom - Flooded with Water at the Oxidation Te perature) Relative to the Equivalent-Cladding-Reacted Parameter and Maximu= 0xidation Tem-perature af ter Rupture in Steam. The best-estimate failure boundary for cladding that was slow cooled through the B-s' trans fo r=a tion before flooding with water and the data of Hesson et al. and Scatena are shown for comparison.
-l l
- l d
i i ~ - - -
.: ,.g 5 -:. ..
m a- ,
.. PRIOR BETA V ..
. ., w *
}, w.. .
~
T.L , .
,200 m, l
l n j - PRIOR BETA 1 **
- (B)
Fig. 10. Microstructures of 21rcaloy-4 Cladding afcar Oxidation in Steam at 1278 K for 5.4 ks and Cooling through the S ~x' Phase Transformation at a Rate of %5 (A) and 100 (3) K/s. Preferential etching of the stabilized a phase is shown by the arrows. ll
k % i-
- n. :h * 'y~~ _
jf.R)X[ a3% e
- PRIOR 3 ETA bd .
P.- c' a
~
' .y
=,
(A) - 200 prrt , 7 5.* i t l l PRIOR BETA l .n
,.* +
h l fd l . i
+
ll1%('*'?WC &f$5 & "'
- i. s - M M iib 3
l 1 ' Fig. 11. Iticrostructures of 21rcaloy-4 Cladding after 0:cidation in Steam at 1603 K for 100 s and Cooling through the 5-/ Phase Trans-formation at a Rate of %5 (A) and 100 (3) K/s.
------------__-_---_________,__..______m ___...____,,s._,____,,m , , _ _ _ , . _ , _ , , _ , , , , _ , _ , , , _ , _ , _ _ _ , . ,. , _
l l l 1 i I 8 ~ c 1 1 I I I i s o i =
- o o i o mme l O M m m w.
o lO = c-w d 8 l *=M mt g-o o o 4,2 - saose-mu l, o I g vctE-~yv l l 1 W oo e o W F: w e 4 G4 5 -e 3 Q N2 g Oo o I WM 322 g 2 U-X=- g$ ee o o o o o.
. cm l
I e.f Q 2 f I e=+ 4J 20 gga c3ca o y l 2 g 1 - eau ,
" o " 5 S~ '
"
- eo w w o o I e
- u .=
J k u . j 5 gg l - R8 - k M k. 3
. s aELL
.a-x O ,,,.
v v mm o oj I l
.. - E e
x O u oo w > = 3"s"x a
.=
.J ue 20 c
..6 l
i y J *UM =? j 3o I ~ 8 19 23w s e z l v $ SM A d 2 Z G w _
- 3 ~
E$ og e l
~
w g 4 "C--C yyC 3 C
- 4 < . W% *U
- a e l .=. W H ~3 =
$ - v v , l- o 33 o o o , o o
3 e -ua
*=d C 3-Gw C l n m .= A 6 a O o an o e
~
C~- o % M 7. -5 w ilw
~
c o 8 wc=mo
-l l ol l l l m6 M V/ 4 sa WW N. @ O T M N Q M b .= = C. =
c i o o' d o o 0 0 C E m
." w aw (W) N30Ax0 % 6* 6*O F DNINIV1NC3 S$3NWlH13$VHd-V138 m A
8 - e i t. i i i
- > l = wai o -
2 - - e e c = I ; 8 .x 2. __ = = y :0 - o*2
- o e o o
oI i (s-W
- 0 "
um e. w
.- j =
=-2 sd e.o ,
s ' s -
.W W e o o o ,. . . .$ g i.
se,12 f WWwog fa ol .u . t e- =
=m-w z- ,g ooo == o e
o o i sm - .
= =.=
um-x g q r 5 w c oo o o o
'o o w6 ex c
i s* we >-2: s eta al ..g@
~. o :$ M .Hau *C
< gi-<d=d o
o .N o ooao o eo3 of .d- c < w 1 o<mw J w 24 e ooo o o o o o oo o o i p e .f u ea
.: 3 i
o - o . . g o, e o ,,m, I o3 o,Lo o o m oo o o ooI f.
* -,* o I"3 .-.x =* w m - g -- oo s 8 6 o
- I - -
Asoao ,
- oco o o 00 : ,$, 1 i I i l i .3 = -
N o o o e v.
"o N
o c' o U~2t!: - o
. O e .= .=
30 v / Hw (m) N3DAXO %WC13 0NIN171NCC SS3N)ClH13S79-713E E i,. .-- - - n . - . . , , , . , , - - ,, . , . . , , -- ~.,wom e es.y. . - , - . . -e. . , - . . .- ,.m.
ISOTHERM At. Oxl0ATION ILMfiHAf t#tE l*Cl 8400 1300 9200 880 0 0000 I l I I lI l I ( l 4 . 80 ._ _ a 5 f -
. . 0 0 t
0 e
'e'ek
- 0 I I I I I I I TRANSVERSE ZlRCALOY-4
** IMPACT VELOCITY O.75 m/s 3 2 -
eee ee O O _ u^3
- TEMPERATURE .-
- k O 373 K
* 'l v. A 473 K
'd Q
80 3 -- e, ,booo, - 9 - O 673 M - e 00 - O 823 M h p 6 - 00 0 - g E 4 _ e _ $2 - _- b $ i
. Y.e -
~ -
2 -
,
- FAILED ON OUENClHNG -
4 6 e SURVIVf D OUENCHtt4G BUT 9 f Att.ED DURif4G HAf4DL.ING $ ,
,,, o g3o 4 3 to -
e FAltEO HY IMPACT - o I ~ - F 8 - , ENEIF2V O.03 J AT 300 M - I 6 - e
- O SUHVIVED O.03 J IMPACT 190T _ O o.
# All LD AT O.3 J AT 300 M _
4 e e$ O SOHVIVED O.3 J IMPACT AT - y 300 M 3
- e.- [
2 - * - Q0 a-
- 3 -
t
'1 I l i I I I 80' O O.2 0.4 0.6 0.8 8.0 8.2 1.4 1.6 6.0 6.5 7.0 7.5 8.0 OXYGEN CONCENTRATION (wl %I 8# T a 10* l*lt)
Fig.'14. Capability of Zircaloy-4 Clad- Fig. 15. Total Absorbed Energy Per Unit Area ding Tubes, Which Survived the Tliermal of Cross Section as a Function of Oxygen Con-Shock, to Withstand Impact Energies of , centration for llomogeneous Zircaloy-4 Speci-0.03 and 0.3 J at 300 K. mens at Several Tennperatures between 373 and , 823 K.
-i <
__ c . . .e-
{ I
'U i i i I I i I
- TRANSVERSE ZlRCALOY-4 l 900 -
IMPACT VELOCITY 0.75 m/s - 800 - - TOTAL ABSORBED 700 - ENERGY al.25 x 104 J/m 2 - x L y 600 - - 0 . k A g 500 - c. 3 A
.# 400 300 - - -
i I 200 - - I I I I I I I 10 0 O O.2 0.4 0.6 0.8 1.0 f.2 1.4 1.6 OXYGEN CONCENTRATION (wt %) Fig. 16. Failure Criterion Relative to the Tempera- l
. ture and Oxygen Concentration of Hemogeneous I Zircaloy-4 Alloys for a Total Absorbed Energy Per Unit Cross Section of 1.25 x 104 J/m 2 . Failure will occur at an energy of 41 x 104 J/m2 for conditions of temperature and oxygen concentration to the right of the line.
i
1 stamaman.w e P[# 1 -
,20 pm ,
h
.x, 'c- ,
, i
s,' ' & ,, -
W ,
?> &__ : l ,, ., ! ,
o-Yq w '- 4
. ^
_y,
/.
1
~
., ,QW y l
-' ry..
m
.~ - .
_g 7 b 1
' f;gx* .' '
1'gi.., % .-
- IJAM '
=
G;r&y O . g!'.;.~ y O!B
- M,Q, g
A
- 20 pm 9#- wm
.. j v
. NM utug g, .
i !
$s#
s a M . daO;I - L (a) (b) l i l
,G'
~~~ ~
' 5 fh &Ef.~,~ -
t .
, Q'C'X. 4S ,
j I' *
;'~d. ~. ~
>:$.r -i:s .h-. m
, ~ -
h' hk:,~~s.s , f-
~
1 7"'-4'&=ffE74TN3[;f4-m&p-a ww E^ L-si*y M 2Ce.y- .
~
$' $t~ D .M',M % W Q - g p W*b h L%g YsGbY&hD2-w:g;3-F25 1 ,% n - -.'MY.23:sC :"Zfu ** v~.nM g
(:) (d) I Fig. 17. SD! ? actographs of 1:ocagenecus ir:alcy-4 Specimens sith 0.75 a: Or/ gen Fractured a: (a) 373. (b) 473, (c) 673, and (d) 323 K. l l
4 - - 8 e g T, # *a i v t neui d . o n 5 5. 2 aon t 1 1 g a, s L l iiec et al e 0 5. e, 5. R at gl 9 ) rb nf l P-N % nt OAe on d sl '
- t w i esn _
5 P ( t cnoe 7 5. i neih
/s 0 N somt t 4- m 9 0 O nCi c lY 5 9 5. $. I 8-I T a cen A r nel e O7 0 R T e pf h L0 9 T gS ew A N ey D CY 0 $. l x4 d 5 E I lRT I 9 I 8 C t O - ee ZC 0- N t yh r id ot r E O 0 E 0 O r nl u S L E 0 9 ER 1 C B aaf c R Lo coc E V 9 0 4- E N 1 I T T
1 oer o C,C I V G t ri s S T uZt e NC 0 Un 5 Y nr AA P 9 DE 2 X D et l aseu . R A - f rumt* 1 Tff I 2- t eoec5 0 0 0 0 cper a4 9 7 6 5 umnur D eesf s Tga a 0 0 0 0 7 0 0 0 0
)lo 9h o euHl neew t e
- 0 0 0 7
0 0 0 0 0 o 1 t l mtl 9 8 e 5 4 3 2 i. I oig
. of rr n gt ofB a
, - =ph wm 2~M i F
h t
- s. ) i s i aswe en r E rosu Aint ER 4 t ea tU TT l
eamr l uie TC pl cp 0 0 2 iA nR maem i vpe i *6 eF O g z )% t DEST ( l
* *2 (
w epca yt t 0 0 r oa
! /s 7 5 l 0 3 t u c pd 4m 5 O t sme 0 1 I
T coI t Y 5 *0 .3 A ar s O7 0 H rcf e 5 8 Fi oT L0 1 9 l 0 N T M A CY
- 85 0 E e- en 0 o 2 C l nce .
HTI l ZC O 0 1
.o g 0 3* l s.
N O i oagK t rf y ct r x3 ES L 0 0 C ucuO2 E 0 N D eS 8 R E V E 1 0* 8 5 E l % H 0 4 G tE e d VT SC cU
,T 0
1 3 l 0 Y n eguua rt n NA AP ,C . 5 . X cnt 8* O RM Tl cA oH oF 0 0 5 7, 2 ric ena 2. 7 3 1 0 O P nr1 3 9* . aF c on -
.S et e
- _ - '- - _ - 8 h e _
O 1 mt 3 1 tw 0 0 o o 0 0 0 o 0 o _ 0 0 o g o 7 0 0 0 4 o 3 0 2 ' .rf . e . 0 9 6 5 gf o0b _ 1 i _ F _[ =s6Nm$mt - - i~ . j, , ! ,; ! ,i1 ,
MODELING OF CRACK GROWTH IN OXIDIZED ZIRCALOY CLADDING DURING THERMAL-SHOCK CONDITIONS O GE0 METRY e CONCENTRIC CYLINDRICAL SHELLS e ZR02 / /S / /ZR02.(0D-ID OXIDATION) t MATERIAL PROPERTIES . e ANIS 0 TROP:C ELASTIC PROPERTIES e FRACTURE INITIATION ENERGY (IEMP. AND OXYGEN CONTENT) e IHERMAL CONDUCTIVITY AND SPECIFIC NEAT 6 HEAT-TRANSFER ANALYSIS s RADI AL HEAT FLOW ; e BOUNDARY CONDITIONS (IIME DEPENDENT) , i
- 1. HEAT-IRANSFER COEFF. AND COOLANT IEMP., OR
- 2. CLADDING SURFACE TEMPS. (ID AND OD) e STRESS ANALYSIS e LINEAR-ELASTIC MATERI AL ,
e PLANE-STRAIN DEFORMATION MODE I FRACIURE ANALYSIS e INITIAL AXIAL AND CIRCUMFERENTIAL CRACK DISTRIBUTION IN IHE OXIDE LAYER
SUMMARY
- 1. EMBRITTLEMENT CRITERION 8
EQUIVALENT-CLADDING REACTED TO FORM 2R02 e SLOW COOL, 28% e FAST COOL, 20% (PRESENT LIMIT'17%) (VALUES INCREASE AS THE TEMPERATURE DECREASES) 8 B-PHASE IHICKNESS WITH A CRITICAL OXYGEN CONTENT e SLOW COOL, L(1.0%) = 0.1 MM e FAST COOL, L(0.9%) = 0.1 MM (VALID IRRESPECTIVE OF CLADDING IHICKNESS, OXIDATION TEMPERATURE,'AND IOTAL OXIDATION) l 8 FAILURE LOCATION IN IHE CLADDING e NONBALLOONED REGION AT T s 1520 K i
-e BALLOONED REGION AT T 2 1570 .( (FRAGMENTATION) l l
l l l i
,r r r , , - - . , , . , - e .-, , , - - - , - ,s--a.-., - . , - - ,- ~ -- n, --ue,.
SUMMARY
- 2. IMPACT RES_U_LJS 0 INCIPIENT IHERMAL-SHOCK FAILURE OF 0xtDIZED ZIRCALOY CLADDING CORRESPONDS TO AN ABSORBED ENERGY OF ~0.03 'J AT 300 K.
8 THE VALUE, WHEN NORMALIZED 3Y IHE CROSS SECTIONAL AREA 0F THE TRANSFORMED S LAYER (0.1-MM THICK), YlELDS AN ENERGY OF ~1 X 104 J/M2 , 8 A NINIMUM ABSORBED ENERGY OF ~1.3 X 104 J/M2 PRODUCES l FRACTURE IN HOMOGENEOUS SUBSIZE CHARPY SPECIMENS WITH 0.6 TO 1.2 WT % 0XYGEN AT. TEMPERATURES EEL 0w 373 TO 823 K, RESPECTIvELY. 8 IMPACT AND IHERMAL-SHOCK RESISTANCE OF OXIDIZED ZIRCALOY CLADDING IS DETERMINED BY IHE PROPERTIES OF IHE TRANSFORMED S LAYER. 1 i l l I J
1
- l j
CONCLUSIONS
$ IF IHERMAL SHOCK PRODUCES THE MOST SEVERE LOADING ON THE CLADDING DURING HYPOTHETICAL LOCA TRANSIENTS, THE RESULTS INDICATE THAT IHE PRESENT ZIRCALOY EMBRITTLEMENT CRITERION IS CONSERVATIVE.
8 A MORE OUANTITATIVE CRITERION, BASED UPON IHE MECHANICAL 3EHAVIOR OF 0x!DIZED ZIRCALOY, CAN BE FORMULATED WITH A SPECIFIED DEGREE OF CONSERVATISM CONSISTENT WITH: A. MECHANICAL LOADS ON.THE CLADDING DURING REFLOOD. B. MAXIMUM OXIDATION DETERMINED BY IHE PERFORMANCE OF IHE ECCS. l
ATTACMXENT 6 OXYGEN DiBRITTLEMENT OF ZIRCALOY-4 FUEL CLADDING by A. Sawatzky Atomic Energy of Canada Limited. Whiteshell Nuclear Research Establishment Pinawa, Manitoba Development of analytical procedures to predict fuel behaviour during hypotheticalaccidentsconstitutesanimportantcontinuingresearchp{ogramat Whiteshell Nuclear Research Establishment. A computer program FAXMOD 1 has been developed to predict the thermal-mechanical behaviour of fuel elements. One phase of the work is to provide mechanical properties data for the fuel cladding including that for cladding exposed to the conditions postulated for a loss-of-ccolant accident (LOCA)(2-51 In a LOCA, the cladding would experience a camperature transient during which it might absorb an appreciable amount of oxygen from the coolant and possibly some from the UO2 fuel. In the present investigation, the ef fects of oxygen concentration and distribution, maximum temperature and cooling rate on the tensile ~ properties of Zircaloy-4 have been investigated. Specimens were 10 mm lengths of cladding having a wall-thickness of 0.43 mm or 0.71 mm. Oxidation in all but one test was carried out in steam. Tensile tests were carried out in air over the temperature range 240C (RT) to 5000C and in argon at higher temperatures. A tangential load was provided by means of D grips inserted into the specimen and attached to the pull rods of the ! tensile machine. Cross head speed was 0.043 mm/s for all tests. The specimen circumference was taken as the gauge length. Oxygen distributions were deter-mined by microhardness. The tensile properties of oxidized cladding exposed to conditions simulating a LOCA were determined. Zircaloy-4 specimens containing 0.11 to 1.70 l
, wt. : oxygen were homogenized in argon at 14850C, cooled through the s+3 region l to RT and tensile-tested over the temperature range 240 to 800 C. The results ;
l are given in Figure 1. For oxygen concentrations of 0.5 we.: and greater, the ultimate tensile strength (UTS) has a maximum at a temperature which increases with increasing oxygen concentration. The elongation is very scattered but can be divided roughly into two bands as shown. It increases with increasing tempera-ture and decreases with increasing oxygen concentration, but with no sudden change indicative of a ductile / brittle transition. Figure 2 shows the region near the fracture of a specimen containing 1.1" oxygen and tensile tested at RT. The , l lamellay2 grain structure is typical of oxidized 21 caloy-4 cooled from the S-regich '. The short cracks across the central portion of the grains suggest alternate layers of ductile and brittle material. This structure may account for the shape of UTS curves in Figure 1 with ductile-brittle material to the lef t of the maxima and completely ductile material to the right. The effect of cooling rate on the tensile properties of Zircaley-41s l 0 shown in Figure 3. The specimens were oxidized to 0.8 wt.% at 900 C, homogenized at 14850C and cooled to RT at races ranging from 0.3 to 160 0 C/s. Although scatter is.large, no effect of cooling rate is seen on either UTS or elongation. l To determine the effect of maximum temperature on the tensile properties, Zircaloy-4 specimens containing 0.3 to 1.1 we.: oxygen were homogenized in argon
0 at 1200 C, 1400 , 1485 and 1600 C and cooled rapidly to 24 C. The UTS and elongation at 400 C are given in Figure 4. The UTS is independent of homo-genizing temperature as is the elongation except for two points. The enhanced elongation of the specimens homogenized at 12000C and containing 0.1 and 0.3 wt.% oxygen is probably due to a smaller grain size. Two series of tests show the effect of the oxygen distribution on the tensile properties. One set of specimens was exposed to a mixture of steam (at low partial pressure) and helium for 1200 or 1800 s and rapidly cooled. The results for six specimens tensile-tested at RT and two ccepression-tested specimens are given in Table 1. All tensile specimens exhibited a low UTS and elonge tion. Both compressed specimens showed little displacement before cracking although the one heated for 1200 s showed considerably more than the one heated for 1800 s. Another set of tensile 0 specimens were exposed in steam to 12 s thermal spikes ranging from 1000 to 1600 C followed by oxidation at 1000 C for a total of 300 s. As seen in Figure 5, the room-temperature UTS was well above 100 MPa for thin-0 walled (0.43 =m) specimens exposed to 1400 C or less and for thick-walled (0.71 mm) specimens exposed to 15000C or less. A comparison of the results of Table 1 and Figure 5 shows that neither maximum temperature nor percent exidation are important factors in determining the tensile properties of oxidized zirconium. In the low- , l strength specimens for which the oxygen distribution was determined, the oxygen , concentration exceeded 0.8 we.% through most of the specimen, whereas in the 1 higher strength specimens, the oxygen was concentrated near the surface leaving the oxygen concentration less than 0.7 we.: over at least half the thickness. The 1 oxygen distribution, particularly over the low-oxygen portien, is therefore an important f actor in determining the tensile properties of Zirealcy-4 cladding. An attempt was made to pred1ct the UTS of Zircaloy-4 cladding for a known oxygen distribution. The measured oxygen distribution of each of the specimens in Figure 5 was divided into narrow elements and the strength of each element calculated using the UTS from Figure 1. These were then added together to cbtain the UTS of the entire specimen. These calculated values are compared , with the exper.imental ones in Table II. The agreement is very good for the thin- ' walled specimens but not quite so good for the thick-walled specimens. The results of the present investigation show that the maximum temper-ature and total orygen content have little or no effect on the tensile properties of Zircaloy-4 cladding. What does determine the tensile properties is the oxygen distribution, particularly over the low-oxygen part of the- cladding. References l (1) Too, J.J.M. and Tamm, H. , "FAXMOD and its Application on the Prediction of High Temperature Creep and Sheath 3allooning 3ehaviour". Submitted to Nuclear Engineering and Design. (2) Sawatzky, A. , Ledoux, G.A., and Jones, S., " Oxidation of Zirconium during a High-Temperature Transient". Zirconium in the Nuclear Industry, ASTM STP 633, A.L. Lowe, Jr. and G.W. Parry Eds., American Society for Testing and Materials, 1977, pp. 134-149. (3) Sawetzky, A., "A Proposed Criterion for the Oxygen Embrittlement of Zircaloy-4 Fuel Cladding". Presented at 4th Inc. Conf. on Zirconium in the Nuclear Industry. Submitted for publication. (4) Rosinger, H.E. , Bera, P.C. .and Clendening, W.R. , "The Steady-S tate Creep of 1
.. . . -. . , . _.- . - ~ . . ... - - . - . - - ..
* * ?e
-. 3- s Zirealoy-4 from 940 to 1873 K", Atomic Energy of Canada Limited Report, AECL-6193 (1978) .
- (5)' ' Rosinger, H.E. and Bera, P.C., "The Steady-State Creep of Zircaloy-4 Containing 0.5 - 0.6 wt.: Oxygen", Atomic Energy of Canada Limited Report.
To be published. I o 4 r s I i p y
- ww1r w .
++.yw-w ge-- -=-,-gwt- ,r* e*--Wwyn e e- r- w e r+--we" r wy,- -e+ = w e --- ,r re r *mw=m,me.-e-mve--- ww ,
g
- -- = 4 +.+
TABLE l 4 Room Temperature Tensile Properties of Zircoloy-4 Oxidized in Low Partial Pressure Steam-l l Diametrol No. l UTS Elong. Compression of Time ! % Oxidation (MPa) (%) (%) Cracks (s) 15.3 29.8 1.4 1800 7.5 42.7 1.6 1800 , 10.0 59.7 1.7 1800 8.8 23.2 !.I 1800 i I 3.9 4.6 0.5 I800
- 3. 5 69.0 2.0 18OO 1.0 1 1800 9.4 Some specimen 3.0 2 1800 1
9.0 4 1800 10.9 Some specimen 6.0 ? 1200 1 7.0 2 1200 ___-____m___ . . . _ _ _ _ . _ _ . _ _ . _ . _ . _ _ _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ . . _ _ _ . _ _ . . _ _ _ ___ _ _ _ ._ ._ _ _ _ _ _ _ _ _ _ _ __________a_.-___._ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ __
TABLE II Room Temperature UTS of Zircaloy-4 Sheathino Havino an Ox.ycen Distribution Thickness Soike Teneerature UTS (MPa) (m) (DC ) Calculated Measured 0.43 1000 504 601 0.43 1100 554 556 0.43 1200 482 535 0.43 1300 395 3% 0.43 1400 276 300 0.43 1500 16 26 0.43 1600 13 11 0 .71 1000 531 587 0.71 1100 548 594 0.71 1200 521 579 4 0.71 1300 511 427 0.71 1400 482 300 0.71 1500 375 173 0.71 1600 130 49
~
700 , , , 0 2 CONC. (WT %) 600 - 0 - 0.1 - a - O.3 o-0.5 7 - 0.7 50 0 - ~
* - 0.9 '
^ - 1.1
.. a - 1.3 l l
E 400 - v - i.7 2
- \
W3 H 3 300 - - O 200 _ -
. A 10 0 -
O
, [ '
_v v o l d
//
0.1 - 0.9 4 z 10 -
- 9
$ O.9 - 1.7 E 5 - -
o a W l f f I 1 1 i O 10 0 200 300 400 500 600 700 800 TEST Th.MPERATURE l*C) FIGURE 1: The UT5 and clones Lion .in . funition of temp.r..ture nd oavgc. concentr.ation fe.r Zircaloy-4 tuosentred at l'AS*C in rgon .tna cooled at 20*Cle.
- .- -- _ - . - . - _ - . ~ . _ . - - - - - - - , - . . - . . - . - , . _ . . . . . . - - - . - - - - - - - - -
. -_ . _...-.- =. -- . . . .-- .- - . . .
1 , l i l'
+:.,
s.- 1
'i g
-?
k-
\
~#
l .- ib\ g
\) \
^
.f t -
,\ ts ,\
(", q
!:1 4M
.tq \
1 h3 . t (U{ ,[\t\ Nidi"s- . i l FIGURE 2: Specimen containing 1.1% oxygen cooled from 1485 C at 20 C/s and fractured at room temperature. l 1
500 i i i i , , , COOLING RATE ('C /S) 400 - o 16 0 E b O 21 a 10.5 g 300 D@ o 9 5.3 a
\\\N O l.8 -
I e i. i , N l 5 N\ l
$ 200 -
{ NN\\ - l b N\\ 10 0 N - g N - O i i i i i i i i 20 g 15 ,
,,-9 z - i
$ s' j s
4 10 LD V 4 l
/
s# a LaJ 4.. - - 5 - E
.O l i l I l 1 l t O 10 0 200 300 400 500 600 700 800 TEST TEMPERATURE ('C)
F1 CURE 3: The UTS and clonu.ition an a function of rootten, rate and temp.r-ature for Zircaloy-4 having an omygen concentatlos of 0.8 wt.2. +t'+4 - 3-sys* but-et e 1*
- r- g wm.,.----e g,ym.- upg,---y -.y y ,we,owwa,.y wy w,tr-*vwne ww- t e-se ar m-I- eee == er wriew've-,,--we9
800 , , , , , HOMOGENIZING 700 - TEMPER ATURE PC) - o 1200 0 1400 600 - o 1485 - O 1600 500 - o i n. 3 400 - - F O-
= 300 -
een sg - A 200 - - loo _. . . i i i i i 20 o 9 15 - - z o O E o y10 w p A - z y_ g cn_ o a " 9-w 5 - -
! i 1 1 1 O O.2 0.4 0.6 0.8 1.0 1.2 OXYGEN CONCENTRATON (WT%)
F1CUltE 4: The UTS .and t It'nc.it i.en of Zl rr.alov-4 .it 400*C an a function of hoeCEenl2 inlit I PS PP f *8 %M r P .arki om .c.vn gone'f nt r;st ion. I
+ _ _ - - - - - - _ _ _ . .-.___.,__-,._m-,._-.e-- .,e ,-n.,-.....- ..,,,__.,.--..w,cx.,.--.....-,--,s-.--- ,.,c- . ,g-.m , - - -
-3 , ,
700 , , , , , i WALL THICKNESS l 600 g(6.3) ,
@l) (5.2) o 0.43 mm -
(4.1) O O O.71 mm o (8.6) o ( )% OXIDIZED 500 - 4) - 5.9)
}400 -
o - 3 (11.3) N (12.6) l 3 30 0 - (6.7) 200 - (9.2) - 100 - (15.7) O (16.1)o 4.3
, , , , 1 o 15 O ^
O O
.N 10 -
5 m %
$5 -
N N-
$ c 1 1 1 1 1000 110 0 1200 1300 1400 1500 16dC MAXIMUM TEMPERATURE (*C)
T! CURE 5: ilie CTS .nd ci..e. itser . .$ i ist .ix i r- i- t . nti. r.i t u ra f. 7.tro..lov-4 explosid t.. . .. .' a d.;+ i.. t r.a:ier i n es=;t i' re- f o l liwd by .'wt
- aL 1000*C.
ATThCfDENT 7 4
. . . ~ . . . .
Ring compression test temp.100'C 5
- 890---964*C 4 .
o 1010---1194* C i 3 - 1 2 .*
- s* %
1
~, ,
t u 9-- 4 $ W W : W^'=0% ~
- 1500 2000 >.500 500 1000 Ring compression test temo.200'C 5 s.
*N
- 908---962*C
\ I T4 -
N s 01012---1159'C . f_, N : g3 - s
- s N
U. 2 _ s C s e N S1 - e N
, .c o . O o , co 0 1500 2000 2500 500 1000 1
1 Ring compression test teme.300*C 5 s
* '~,
- 897---961*C 4
c1019---l167'C 3 - '%~ ,* e 2
- ~,
e 1 o
-o e - o-e-
, e ,
O 2500 1000 , 1500 2000 500 AbsorbedHydrogenCentent(wt. ppm) l l l , , _ _ . . . - . , _ . . _ . , - - . . - , - . . . . ~ . , - . , _ , - . , , , . - , - , _ _ .m ... ._ , , - , . , , , . . _ _ ~ , . , . . _ . . . -u .m..--,_.-,- . . . - _ . . . . ~ , - ,..m. -
,.,}}