US3170788A - Uranium base alloy - Google Patents

Description

United States Patent 3,17tl,788 URANTUM BASE ALLUY Martin H. Binstock, Canoga Paris, and Harold E. Kline,

Chatsworth, Calif., assignors, by mesne assignments, to the United States of America as represented by the United States Atomic Energy Qommission No Drawing. Filed July 16, 1962, Ser. No. 210,191

6 Claims. (Cl. 75-1227) This application is a continuation in part of our application Serial No. 806,231, filed April 14, 1959, for Uranium Base Alloy, now abandoned.

Our invention relates to a uranium base alloy, and more particularly to a uranium base alloy of improved metallurgical characteristics for nuclear reactor application.

Dimensional changes in a reactor fuel element material undergoing nuclear irradiation cause reduced heat transfer efficiency, hot spots, or eventual failure, and in any event reduce the useful life of the fuel element. Since fuel cycle costs are a major factor in the economics of nuclear power generation, it is essential that longer-lived fuel elements be developed.

Uranium metal undergoes deformation when subjected to nuclear irradiation. In the fission process, gases such as xenon and krypton are released which remain ontrapped in the uranium leading to swelling (dimensional change with change of density). Thermal cycling in a nuclear reactor, that is, change of temperature over a wide range associated with startup and shutdown of a reactor, also contributes to plastic deformation or growth of uranium metal (dimensional change without change of density). The physical properties of uranium which lead to plastic deformation include relatively low creep resistance and yield strength at elevated temperatures. Higher creep resistance and yield strength would resist the stresses induced by entrapped fission gases, and resist the effects of thermal cycling. Due to the metallurgical limitations of uranium metal, the usefulness of uranium metal as a power reactor fuel material is severely limited.

The prior art has established that alloying of uranium metal increases creep resistance and yield strength at elevated temperatures. For example, molybdenum adds solid solution strengthening to uranium. The binary uranium molybdenum alloy exhibits higher resistance to plastic deformation than does uranium metal. However, since alloying elements are parasitic neutron absorbers, economic aspects limit the total alloying contents to under, say, approximately 10% of the fuel weight, and preferably even lower. Therefore, a high stren th alloy must be developed with relatively small alloy additions.

The criteria for an improved uranium alloy for reactor application are many and severe, and the meeting of these criteria represents the objects of our present invention. The fuel must resist dimensional changes under irradiation for a long lifetime. It must Withstand stresses induced by thermal cycling in reactor shutdown and startup. It must have sumcient strength and creep resistance to support its own weight and that of its cladding in the fuel element at reactor temperatures. (When aluminum is used as cladding, which is preferred over stainless steel because of its lower neutron absorption, the weight of the element must be borne by the fuel rather than the non-load-bearing aluminum.) Alloy additions enabling the fuel to resist deformation should be low to minimize neutron absorption. Finally, the process for manufacturing the alloy should be low in cost.

In accordance with our present invention we have developed a uranium base alloy which satisfactorily meets the stringent requirements outlined abve. This alloy consists essentially of approximately 1.5-6 wt. percent molybdenum, one addition selected from the class consisting of (a) approximately 0.1-1 Wt. percent aluminum,

3,170,788 Patented Feb. 23, 1965 and (12) approximately 0.2-1 wt. percent aluminum and silicon, wherein the concentration of each of aluminum and silicon is at least approximately 0.1 wt. percent, and the remainder uranium.

The present improved uranium alloy is manifested in yield and tensile strengths and creep resistance which are factors of 3-5 greater than uranium metal. The addition of aluminum: and/ or silicon considerably and unexpectedly improves the characteristics of the alloy, and imparts additional resistance to plastic deformation. The alloy contains the required low amounts of alloying constituents. The alloy can be easily prepared by a number of different methods, including economical direct casting into the desired fuel element shape. The improved characteristics of our alloy have been confirmed by irradiation testing.

The concentration of molybdenum in the alloy may satisfactorily vary between approximately 1.5 and 6 wt. percent, with approximately 3.5 wt. percent being preferred. Under approximately 1.5 wt. percent the uranium-molybdenum alloy undergoes phase transformation above 1250 F. into a gamma (body centered cube) phase and beta (tetragonal) phase. The beta phase has undesirable mechanical properties and should be avoided. Above approximately 6 Wt. percent molybdenum, a single phase may be reached at about 28 atomic percent molybdenum. Grain boundary movements permit fission gas pockets to accumulate, and such pockets act as stress points in the alloy. Quite surprisingly, our low molybdenum alloy exhibits greater tensile and creep strength below 932 F. than does the 10% molybdenum binary alloy (higher Mo contents, without the neutron absorber limitation, would be expected to give the'higher value), although above about 932 F. the tensile and creep strength of the alloy falls off as compared with the 10% binary alloy. Since in many reactor applications the maximum center temperature of a fuel element would be below 932 F., the aluminum ternary or the quaternary alloy is actually preferred, even apart from the economic necessity of mini mum amount of neutron absorbing molybdenum in the alloy. For example, the maximum average center temperature of a fuel element in an organic-cooled reactor, for which our alloy is especially suitable, is about 850 F., with the organic coolant at a typically mixed mean outlet temperature of about 650 F.

Small amounts of aluminum and/ or silicon are added to produce a ternary or quaternary alloy. The concentration of these additives may satisfactorily vary between approximately 0.1 wt. percent to 1.0 wt. percent, with approximately 0.5 wt. percent being preferred. The aluminum and/ or silicon additive forms a very fine, submicroscopic phase within and/or along U-Mo grain boundaries, and imports considerable additional resistance to plastic deformation. Also, the additive produces grain refinement which is desirable in an anisotropic material such as uranium and its alloys. This phase increases hardness of the alloy and increases creep strength. Furthermore, this phase does not show age coalescence, and does not dissolve nor precipitate further with thermal cycling; in other words, the precipitated phase remains very fine and stable. The dispersed phase is suggested to be compounds of uranium and aluminum or silicon, whose identification may be, for example, UAl UAl USi or U Si. The small aluminum and/or silicon addition in the alloy, which produces the dispersed phase in uraniummolybdenum, does not significantly alter the molybdenurn-uranium phase diagram, which has been determined and published for uranium-rich uranium-molybdenum alloys.

In compming the aluminum and silicon additives to the alloy, the 0.1 wt. percent tertiary aluminum alloy has the highest tensile strength, 110,000 p.s.i., and the highest yield strength, 80,600 psi. However, all alloys exceed the tensile and yield strength of unalloyed uranium (16,700 and 11,900 psi, respectively) by a factor of 4 or greater. Tensile and creep strengths of our alloys also exceeded those of the binary molybdenum alloy. Ductility is lowest for the aluminum alloy (less than 0.2% elongation for U3.5Mo0.5Al). The silicon additive ternary alloy displays the greatest stability to thermal cycling. Thermal cycling of the ternary alloys between 200 F. and 900l000 F. did not produce any significant distortion or growth after 300 cycles, while reference specimens of unalloy uranium were severely warped.

Combining aluminum and silicon in a quaternary alloy containing a total aluminum and silicon addition of approximately 1% will give the very high tensile and yield strength of aluminum arid relatively better ductility and resistance to thermal cycling. The concentration of each of aluminum and silicon is at least approximately 0.1 wt. percent. Since considerable strengthening is irnparted even by small concentrations of aluminum, the concentration of aluminum can generally be in the lower portion of the range. Typical quaternary alloy-s may contain either equal amounts of aluminum and silicon, or relatively higher concentrations of silicon than of aluminum. For example, a preferred quaternary alloy consists essentially of approximately: 3.5 wt. percent molybdenum, 0.1 wt. percent aluminum, 0.3 wt. percent silicon, and the remainder uranium. Another satisfactory alloy consists essentially of approximately, by weight percent: 3.5 mo-0.1Al0.lSi, and the remainder uranium.

vacuum induction heating and cast into a MgZrO -washed graphite mold as fuel plates. The mold, which is also under vacuum, is normally heated slightly, say to 200 F. to prevent thermal shock between it and the melt. The alloy melts at approximately 1950 F., is cast into the graphite mold, and the melt is cooled to room temperature. Melting and casting may also be conducted under an inert gas atmosphere provided the impurity content of the gas is sufficiently low.

In the preparation of as-cast reactor fuel plates, the mold is typically gang-type, md each heat may result in six or eight plates of either 0.130 in. or 0.150 in. nominal thickness, 2 /2 in. nominal width, and 10 in. to 14 in. length depending upon the mold design. The melting and casting of fuel plates, as well as casting to desired thickness for direct use in fuel elements without surface machining or rolling, involve conventional metallurgical methods, present no unusual problems, and may be done by various methods since no one is critical.

The following tests were performed on the ternary alloys prepared as just described by vacuum induction melting and casting; mechanical, thermal cycling, and nuclear irradiation. The results are tabulated below and summarized, beginning first with mechanical testing.

The mechanical test included tensile and creep tests at 900 P1, plus hardness tests at room temperature. In the tensile test, all specimens were heated for one hour at temperature prior to testing. Temperature was maintained within *-3 F. A strain rate of 0.0005 in. per in. per min. was used for all tests. The results of the tensile properties at 900 F. are summarized in Table I.

TABLE I Results of tensile tests at 900 F.

No. of Yield Stren th .s.i. at Elon ation Alloy Specimens Ultimate Strength g p in g in.

Tested (p.s.1.) (percent) 0.2% Oflsct 0.1% Offset 0.02% Ofiset Maximum 79,300. 42, 000 32, 000 22, 500 11. 0 U-3.5M0 2 Minimum 73,500--- 40, 800 31, 600 15, 700 3. 5 Average 70,400- 41, 400 32, 100 19, (S00 7. 0 Maximum 74,700 :15, 000 44, 000 32, 200 0. (1 U3.5M0-0.5S1 2 Minimum 66,400- 54, 600 43, 900 30, 200 0. 6 Average 70,600 54, 800 44', 200 51, 200 0. 0 Maximum 53,600. 52, 600 0. 2 U-3.5Mo-0.5Al- 3 Minimum 49,400--. 30, 500 0.2 Average 51,900. (0 45, 300 0.2 Unalloyed U 1 Average 16,700. 11, 900 10, 500 0, 000 14. 5

1 Not measurable because of low ductility.

Notes:

Specimens preheated for 1 hour in helium. All tests at strain rate of 0.0005 in. per in. per min. Specimens machined from as-cast 0.130111. or 0.150 in. thick plates.

A number of methods are available for the preparation of our alloy and for the fabrication of fuel element plates with the dimensions and tolerances required, and no single method is critical. Among the satisfactory techniques for fabrication of fuel plates are static casting of plates directly to design thickness; casting of slabs, hot rolling and cold rolling to thickness; and centrifugal casting of plates to thickness. Since static casting of plates to design thickness requires only machining of edges and ends prior to assembling as a fuel element plate, the use of as-cast fuel plates of design thickness is the most economically attractive method of alloy preparation. Furthermore, casting Without subsequent working gives the desirable random orientation of grain structure.

One suitable method of melting and casting the alloy comprises placing weighed portions of the alloy constitueuts in a MgZrO -washed graphite crucible. The MgZrO wash prevents any carbonizing reaction between the alloy and the crucible. The metals are melted by It is seen that the U3.5Mo-0.5Si alloy has the highest yield strength followed by the tcrnary aluminum and binary alloys. The 3.5Mo-0.5Al alloy, with its high yield-to-ultimate ratio, has the highest resistance to plastic deformation with respect to tensile properties followed by the ternary silicon and binary alloys. The tensile strength of the ternary alloys greatly exceeded that of unalloyed uranium, which was also tested for reference, by factors.

TABLE 11 Results of creep tests at 900 F.

Test

Stress (p.s.i.)

Alloy Specimen Number Rate 0.19 3. 3

U-3.5M-0.5Si

U-3.5M0-0.5A1

1 Literature reference. 2 Secondary creep rate in percent per 100 hours for hours mdlcated. 3 Data erroneous because of severe oxidation.

The results show that U-Mo-Al has the highest creep strength at 900 F. followed by U-Mo-Si, and both ternary alloys were better than the binary alloy and considerably better than compared with literature values for beta-quenched uranium. U-Mo-Al exhibited a secondary creep rate of about 0.3%/100 hrs. for stresses up to 20,000 p.s.i. U-Mo-Si also showed good creep resistance at loads up to 15,000 psi, the creep rate being 0.6%/100 hrs. for this load. At 20,000 p.s.i., however, the creep rate increased sharply to more than 2%/ 100 hrs.

Thermal cycling was conducted between 200 F. and .1000" F. in alpha phase cycling tests, and the differences between the various alloys with respect to thermal stability were not significant, although the relative instability of unalloyed uranium was again demonstrated. With re- 200 F. and 1300 F. during which the low temperature orthorhombic alpha phase transformed to the body-centered cubic gamma phase. The stability of the ternary alloys as compared with unalloyed uranium is clearly evident, as summarized in Table III, which indicates the actual changes in dimensions and in volume.

The unalloyed uranium specimens grew more than 50% in thickness, shrunk by some 20% in length, and had a total increase of some 34% in volume. The U-Mo-Si specimens were the most stable, with a maximum change for any dimensions of less than and a volume change of +0.73%. The U-Mo-Al and U-Mo specimens had total volume changes of 1.9% and 7.7%, respectively, with correspondingly larger increases in individual dimensions.

TABLE 111 Dimensional changes* due to thermal cycling between 200 and 1300 F.

Test No. of Percent Percent Percent Percent Alloy Specimen Cycles A L A T AW AV 1 Number (i0.1%) (i0.4%) (510.2%) (i0.8%)

95 +0. 80 +0. 81 +0. 04 +1. 72 1T1 2 +1.80 +3. +1. 09 +6. 43 1T3 95 +0. 80 +0. 65 +0. 05 +1. 46 1T4 95 +1. +0. 48 0. 14 +1. 1L2 95 +0. +0, 16 +0. 41 +0. 97 240 +2. 13 +3. 20 +3. 47 +9. 03 1L7 05 +1. 30 0. +0. 23 +1. 22 1L8 95 +1. 10 +0. 16 0 +1. 46 U-3.5M0O.5Sl 2T1 95 +0. 40 +0. 24 0. 14 +0. 73 2T2 95 +0. 30 0 -0. 09 +0. 24 2T3 95 +0. 20 +0. 24 0. 13 +0. 24 240 +0. C0 +0. 24 -0. 09 +0. 97 2L5 95 0. 07 0. +0. 05 0. 24 2L6 95 0. 60 +0. 24 +0. 13 +0. 24 240 +0. 20 +0. 24 +0. 18 +0. 49 223 32338 +3 36 iii? Unauoyed U 3L3 240 2 25 21 4 2 +91 2 I31 3T1 95 2 10 2 +24 2 +2. 3 2 +13 240 2 18 2 +48 2 +20. 4 +37 *Reierence to uncycled specimen dimensions, nominally 1.500 in. x 0.125 in. x 0.220 in.

1 Calculated from measured specimen dimensions. 2 Approximate values because of specimen distortion.

spect to dimension changes, the unalloyed uranium specimen grew about +2.4% in volume after 95 cycles, while as-cast plates were as follows:

the change for the ternary molybdenum-bearing alloy was less than +0.25%. In appearance, however, the molybdenum ternary alloys were essentially unchanged, while incipient surface roughening was visible for the unalloyed uranium specimen, along with slight distortion.

Hardness measurements were made on as-east plates with a Rockwell tester. Representative hardnesses on the U-3.0 to 3.5 Mo0.4 w 0.5 Si 40 to 42 RC. U3.0 to 3.5 Mo-0.4 to 0.5 A1 42 to 46 RC. U-3.0 to 3.5 Mo as m 40 RC.

Thermal cycling tests were also conducted between Unalloyed uranium 88 to 93 R 7 The U-Mo-Al alloy possessed very high hardness, while the U-Mo-Si alloy was also hard but less brittle than U-Mo-Al. The uranium metal and binary alloys were relatively softer. No significant changes in hardness due 8 TABLE v Minimum creep rates at 550 C. (1022 F.) and 6000 p.s.i.

to thermal treatments (thermal cycling, or elevated temperature tensile and creep tests) were noted using stand- Alloy MGR (percent per hr.) ard indentors.

A fuel element containing uranium-3.5 wt. percent U-3.5Mo-0.lAl-0.3Sl 1.3 10=; Mo0.5 wt. percent Si was fabricated into the finned, gjg gigjg fitg, fiat plate configuration described in the co-pending ap- U-3.5Mo 2 10 plication S.N. 711,595 of Wheelock et al., now US. 10 2,999,058. The fuel plate cladding and structural material of the element was aluminum. This fuel element TABLE VI was placed in the Organic Moderated Reactor Experi- Density changes as a result of irradiation to approximatement reactor (OMRE) and irradiated to a burn-up of aply 15,000 mwd./M Lu. burn-up at temperatures of proximately 3000 megawatt days per ton. The fuel ele- 400 and 500 C. (752 and 932 F.) ment was then removed and examined. Fuel plates showed no significant swelling or growth due to nuclear irradiation or thermal cycling. The dimensions of a Alloy Apfiment) typical fuel plate (including the 35 mil aluminum clado a ding) before and after irradiation were as follows: 400 500 4.5 10.9 7.0 30.0 Before Irradiation After Irradiation 5 8 5.6 11.8 Extreme Extreme Width 2.330111.... 2.380 in. to 2.383 in. Thickness .470 in. .467111. to .470 in. Length 12.175 in. 12.174 in. to 12.177 in.

The foregoing examples and physical data of our alloys are by way of illustration only, and are not intended The aluminum alloy displays similar fine behavior under as restrictive. Our invention is understood to be limited irradiation. only as indicated by the appended claims.

The following Table IV shows the tensile properties of We claim: various ternary and quaternary uranium alloys. 1. A uranium base alloy consisting essentially of ap- TABLE IV Tensile properties of U-alloys Ultimate Tensile Strength Yield .Strength Percent Elongation in 2 in. Modulus of Elasticity (p.s.i. 10 (p.s.1.X10 lengths (X100 Composition U-1.5Mo 59. 0 44. 4 27. 5 13. 3 9. 0 12. 0 13. 0 5. 3 U-l.5M0-0.5Si 75. 8 48. 0 27. 4 52. 24. 8 11. 0 1. 5 5. 0 17. 0 3. 3 U-1.5Mo-0.1Al 68. 2 52. 0 34. 8 51. 38.6 1 20.6 6. 5 7. 5 1 11.0 11.9 12.5 1 5. 7 U 1.5Mo 0.1A1 79. 3 61. 5 37. 0 63. 41.8 19. 7 4. 0 4. 0 12. 0 14. s 12. 0 4. 3 U-3.5M0 110.4 52.3 23.0 66. 18.3 4.0 6.0 15.5 68.0 10.9 4.3 1.0 U-3.5Mo-0.5Si 88. 2 01. 3 29.8 70. 29. 3 13. 2 0. 5 2. 0 11.0 11.5 8. 7 2. 3 03514001111 127.5 77.7 33.3 71. 37.0 15.0 4.0 0.0 24.0 11.0 7.0 2.9 U 3.51\1o-0.1A1-0.3si 127. 5 s3. 4 1 45. 0 87. 44. 2 1 22. 4 2. 0 2. 5 1 9. 5 1 4. 4 07.0 54.3 59.9 04 59.3 44.5 1.5 2.5 8.0 8.5 7.9 7.9 59.0 54.0 54.0 42.2 0 0 1.5 8.9 7.3 8.4 50.5 52.8 59.9 52.4 1.0 1.5 2.5 9.4 8.3 8.7

1 Results of one test only.

It is noted that the ultimate tensile and yield strengths of the quaternary U-Mo-Al-Si alloys are better than those of the ternary Si alloys containing the same amount of molybdenum, even though the total weight of Si and A1 (0.4 wt. percent) is lower than the Weight of Si in the ternary allow (0.5 wt. percent). The data also confirms the discussion above relative to the surprisingly greater tensile strength of the lower Mo alloys than of the 10% Mo alloys at temperatures to about 932 F. It is noted that the ultimate tensil strength of the 10% M0 alloys with the ternary additions of Si and A1 are below that of the U-10 Mo binary, and that the corresponding 3.5 Mo ternary alloys have greater ultimate tensile strengths.

The following tables compare the creep behavior and radiation behavior of several uranium base alloys. It is noted that our quaternary allow was superior in these respects to the other alloys tested.

proximately l.S-6 wt. percent molybdenum, one addition selected from the class consisting of:

(a) approximately 0.1-1 Wt. percent aluminum, and

(b) approximately 0.2-1 wt. percent aluminum and silicon,

wherein the concentration of each of aluminum 3. A uranium base alloy consisting essentially of:

approximately 3.5 Wt. percent molybdenum,

4. A uranium base alloy consisting essentially of approximately 3.5 Wt. percent molybdenum, approximately 0.3 Wt. percent silicon, approximately 0.1 Wt. percent aluminum,

and the remainder uranium.

5. A uranium base alloy consisting essentially of approximately 1.5-6 wt. percent molybdenum, approximately 0.1-1 wt. percent aluminum,

and the remainder uranium.

6. A uranium base alloy consisting essentially of 7 cent, 1 and the remainder uranium.

approximately 3.5 Wt. percent molybdenum, approximately 0.1 wt. percent aluminum, and the remainder uranium.

References Cited by the Examiner UNITED STATES PATENTS 2,886,430 5/59 Allen et al. 75122.7 2,919,186 12/59 Colbeck 75122.7

OTHER REFERENCES Nuclear Science Abstracts, vol. 13, September-October 1959, p. 2060, Abstract No. 15355.

Nuclear Science Abstracts, vol. 14, July-August 1960, p. 1786, Abstract No. 14126.

15 CARL D. QUARFORTH, Primary Examiner.

OSCAR R. VERTIZ, Examiner.

Claims (1)

1. A URANIUM BASE ALLOY CONSISTING ESSENTIALLY OF APPROXIMATELY 1.5-6 WT. PERCENT MOLYBDENUM, ONE ADDITION SELECTED FROM THE CLASS CONSISTING OF: (A) APPROXIMATELY 0.1-1 WT. PERCENT ALUMINUM, AND (B) APPROXIMATELY 0.2-1 WT. PERCENT ALUMINUM AND SILICON, WHEREIN THE CONCENTRATION OF EACH OF ALUMINUM AND SILICON IS AT LEAST APPROXIMATELY 0.1 WT. PERCENT, AND THE REMAINDER URANIUM.