US3067030A - Nickel base alloy - Google Patents

Description

Dec. 4, 1962 E. l.. DUNN ETAL 3,057,030

- NICKEL BASE ALLOY Filed Aug. 13, 1958 2 Sheets-Sheet l -Lfll 'EE Mw IrrAGx/f/ Dec. 4, 1962 E. L. DUNN ETAL NICKEL BASE ALLOY 2 Sheets-Sheet 2 Filed Aug. 13, 1958 e N m v Y instaat latented Dec. 4l, 1952 3,067,030 NItCliEuL EASE ALLEY Eugene landwirt: Dunn, Batavia, and James Eliis Wilson,

Loveland, hic, Marion Edward Ciesliclri, Worcester, and .lioseph Burton Moore, Weather-o, Mass., assigners, hy mesne assignments, to the United States et America as represented hy the Secretary ci the Air Force Filed Aug. i3, 1958, Ser. N 754,304 it Claim. (Ci. 75--1'731- rThis invention relates to precipitation hardening type, vacuum melted nickel base alloys and to their heat treatment, the alloys being especially suitable for use in the casting oi articles and for elevated temperature applications under stressed conditions.

The best nickel base alloys developed heretofore for high temperature, high stressed conditions such as may be experienced by cast blades, vanes, buckets or struts in gas turbine apparatus, tend to lose some of their strength properties at a substantially increasing rate in the 1600" F, and above range. At l800 F. and above, articles such as blades made from the best of such known alloys have failed in relatively short periods of time under only moderately stressed conditions such as 15,000 pounds per square inch.

it is well known that the structure of many alloys, iucluding nickel base alloys, is composed of a group of grains joined at grain boundary areas between the grains. At low temperatures, the grain boundary is more resistant to stress than the grains; therefore failure occurs through the grains or transgranularly. However, at levated temperatures the grain boundaries can become more plastic and tend to allow the grains to slip over one another causing failure between the grains or intergranularly. Therefore, in order to strengthen nickel base alloys for elevated temperature applications, it is necessary to place hard, blocking particles in the grain boundaries to prevent such slipping.

it is believed that nickel base alloys for high temperature applications are hardened through the precipitation of a phase called gamma prime sometimes referred to as Ni3(Al, Ti) in which it is seen that the elements aluminum and titanium are significant. in addition, hard particles of carbides and borides strategically precipitated or located in the grain boundaries atleet the high ternperature strength of such alloys. When particles such as these are deposited or precipitated, the alloy is called a precipitation hardening" type of alloy.

Besides the elements boron and carbon which cooperate to form the borides and carbides, the element molybdenum cooperates to add solution strengthening to nickel base alloys. Therefore, although the elements aluminum, titanium, boron and molybdenum have been included in other known nickel base alloys, their interactions in certain narrow ranges, coupled with the eiects of heat treatment and size of metal grains, has not previously been clearly dened.

An object of this invention is to provide an improved, precipitation hardening type, vacuum melted nickel base alloy and heat treatment for that alloy so as to produce a material suitable for use in the casting of articles and characterized by the fact that it has improved strength properties under stress conditions at elevated temperatures.

r'edy stated in accordance with one aspect of our invention, we provide an improved, precipitation hardening type vacuum melted nickel base alloy including the iinportant elements titanium, aluminum, boron and molybdenum in percent by weight of 3 4 titanium, 0.05-0.l0 boron, 3.6-5.0 aluminum and 4-7 molybdenum. ln general our alloy comprises in percent by weight up to about 0.08 carbon, about Lriron, about 14-17 chromium, about 3-4 titanium, about 0.05-0.10 boron, about 3.6- 5.0 aluminum, about 4-7 molybdenum, with the balance essentially nickel and impurities. in cast form our alloy, though relatively strong in the as-cast condition, may be Luther strengthened through heat treatment by first heating within the range of 2075-220 F. and then air cooling. We have found that the average grain diameter of our cast alloy is a signicant factor in determining strength properties. We prefer the average grain diameter oi our alloy to be in excess of 0.015 inch.

The composition range oi our alloy and its heat treatment will be better understood from our description taken in conjunction with the accompanying drawings and the oilowing examples which are given by way of illustration only and not in any way by limitation. The scope of our invention will be pointed out in the claims.

in the drawings:

FIGURE l is a graph of the curves which compare stress rupture strength of our alloy in cast and heat treated form with other currently available elevated temperature cast alloys;

FlGURE 2 is a graph showing the tensile and yield strengths of our alloy in cast and heat treated form as well as percent reduction of area and elongation;

iti-URE 3 is a graph comparing stress rupture life to average grain diameter.

The chemical compositions of the best available alloys suitable for use at elevated temperatures are shown in Table l along with the chemical composition of our alloy in our preferred range. included in these alloys as irnpurities may be, in percent by weight, a maximum of about 0.l manganese, about `0.1 silicon, about 0.01 sulphur and about 0.1 copper.

Table 1 [Composition in percent by weight] Our Alloy Alloy B Alloy B Alloy C Alloy D 08 max .10. 20 .02 .20 mar. .l mar. 4-6 3. 5-5. 0 0 5 2 1 -16 14-17 15 11-14 18 3. 2543. 65 2. 0-3. 0 0 25-1. 5 3 .O7-. 09 0 o. 1 0 0 el. 4%. S 3. 25-4. (l 6. 6 5. 5-6. 5 3 4. 5-6. 0 4. 5-6. 0 5 3. 5-5. 5 4 O 0 0 0 18 0 0 0 2 O Bal. Bal. Bal. Bal. Bal.

Two of the tests generally conducted on alloys to determine their physical properties are called stress rupture and tensile tests. The results or these tests are reported as stress rupture strength and tensile strength respectively and are generally deiined as follows:

Stress rupture strength is the value of stress in pounds per square inch obtained by dividing the amount of an applied load by the area supporting that load initially while the test specimen is maintained at a selected temperature. The strength of material of a test specimen is also measured by the length of time it resists failure at a given stress and temperature. Such a length of time is called stress rupture life. By incorporating such values as temperature, initially applied load and stress rupture life into a special graph, later to be described, that graph may be used to predict the life of a material or the time a given material can be expected to be operable prior to failure under a speciiied stress and at a specified temperature.

Tensile strength, somes called ultimate strength, is the value in pounds per square inch obtained when the maximum load recorded during the straining of a speci- 3 men is divided by the cross-sectional area of the specimen before straining.

Referring to FIGURE l stress rupture strengths are represented by the comparison of stress with a time-temperai propertie etallurgical technology aiong these lines has advanced to the point where it is no longer possible to broadly designate the composition range of such elements as aluminum, titanium, molydcnum and boron and still ure parameter shown at the horizontal coordinate. This be able to predict the physical properties which will result. parameter, known as the Larson-Miller Parameter, has In addition, new important heat treatment cycles must be been calculated from the formula P=T(+log i) l0*3 specified in order to control theprecipitation ot phases in which P equals the time-temperature parameter nu'n'iwhich are developed in alloys having aluminum, titanium, ber, T eqilialsh absolute l.emperature in degrees P anlririe lo molybdenum and boron as part of1 their lornposition. and tequast e time in ours. The curves of FIGUREI As an illustration, Table 3 inc tides t e composition have been prepared from a large amount of stress rupranges and the average stress rupture lite of a number of ture test results and represent a compact summary of a heats of material which we have prepared and tested. The Wide range of data. Using this special graph it is poS- stress rupture properties were obtained at l800 F. with sible to predict the stress rupture life of a material under the samples under an axial load of 15,000 pounds per a given load at a given temperature. 15 square inch:

Table 3 [Composition (percent by weighty] Avg. i S'CI'OSS Ex. Rrptfture C Fe Cr B Ti A1 Mo Ni si Mn s l e (Hours) 94' .05 4.8 15.6 .09 2.1 4.6 4.0 Bai. .10 .05 .005 213l .o5 4.8 15.6 .69 2.8 4.6 4.0 B51. .10 .05 .005 290y .05 4.8 15.0 .o9 3.5 4.6 4.o B211. .10 .05 .605 313 .0s 4.8 15.6 .0a 4.0 3.6 6.9 Bai. .06 .05 .066 543 .o5 4.7 15.2 .0s 3.5 4.4 5.3 Bai. .i0 .05 .007 59 .07 4.7 15.8 .03 3.5 2,5 5.3 Bai. .io .o5 .007 446 .07 4.8 14.8 .0s 2.7 4.5 5.4 Bai. .05 .05 .007

it is to be noted that our alloy in vacuum cast form lt is to be noted that the alloys of Examples A-4, G-3, when heat treated according to Example shown in C-3 and D-lf, which lie within our unusual composition Table 6hherei(i)i,5and Ihaii/ing an average grain dilameter range, all have unusually high stress1 ripture lil/ies at greater t an l inc as stress rupture strengt propl800 F. under an axial load suci as 5, 00 poun s per erties Superior to any other available hiUh temperature square inch. All the alloys shown in Table 3 were heat nickel base alloy. Villen heat treated according to EX- treated according to Example Aof Table 6. ample C of Table 6, our alloy in vacuum cast form is in Table 3 it can be seen that as the titanium content shown to have even greater stress rupture strength. These increases over a very narrow range, the stress rupture increased strengths are attributed riot only to the special life increases over a very widg range thus Showing the heat treatments as later herein described, but also to the importance of such range. Similarly, small differences `Special ranges of titanium, boron, aluminum and molybin aluminum content may result in vast differences in denui'n as Well as the specicatiori of average grain diamstress rupture life. This is easily Seen by comparison of eter in excess of about 0.015 inch. Examples C-S and C-l, From the testing of a group 0f Although AHOY A 0f Table l has a range 0f elements alloys of which the examples of Table 3 are a part, we Which might at fusi appea 'i0 be Similar li0 that Of 0ur found that the most signiicant range of molyderium, tialloy, reference to FIGURE l shows that the differences V15 ianium and aluminum in an alloy such as ours was in in the amounts of aluminum, boron, titanium and carbon percent by weight molybdenum, 4 7, aluminum 3.6-5 and combine with other elements and phases to result in a titanium 3 4, 0f this range We prefer the range in permuch weaker alloy than is ours. cent by Weight of molybdenum 4.5-6.0, aluminum 4.4-4.8 Figure 2 represents the average of results of tensile testand titanium 325-355 Th@ Specic compositions which ing 0f Ouf alu 0f the OHOWug llfld CumptOSuOu lul', 50 gave the most outstanding results are shown by Examples vacuum mete vacuum cast an 'ieat trea e accor ing C 3 and D 4 of Table 3 to Example Aoi Table 6: Although we had determined that a certain specific Table 2 range for titanium, molybdenum and aluminum, coupled [Composition in percent by weight] with the other elements in percent by weight of carbon 1 0 O8 p 0 O5 0 10 .08 max., iron 4 6, chromium lli-17, boron about :05110 C J 3'6 o with the balance essentially nickel and impurities, We Fe 4""6 A1 rrrrrrrrr r 4 '7 recognized in our testing that the average grain diameter Cf 1441 M) B l P of our alloy was another important factor in determin- Tll 3 N1 rrrrrrr r a am ing strength properties of alloys such as ours in cast form. Jax' L 60 In a property evaluation program of this alloy, we noted In addlton t0 howmg the utlmflte length of Ourlm that there was a variation in stress rupture lives which PTO"ed alloy FUURELSUmmUZeS a lrg? mou' could not be explained either by chemical analysis or da? tt? NPO Lheadchgooolfafg slmslema dure metallographic studies. However, ve noted that there reucioninarea n ogai i L v. nu Q hg testing The term o 2% yield Strennf, Shown in 65 was a relationship between tne average grain diameter C and the stress rupture life in that increasing average grain FIGURE 2 1s the stress at which a material exhibits 0.2% y M diametei gave increasing stress rupture life. Our theory deviation from the proportionality of stress to stiain. was statistican, corrmed usm th four uadr Ft cof This ligure, sometimes called 0.2% odset is commonly 1 t; t 1 5, LT1: thi s Ahe ltq f lih used to measure the tensile strength and, indirectly, the fi a *011. eglcl 311s me. 0 s L fru S l Wlic are tensile ductility of alloys such as ours. s *Own m cnsms O f d1 mg e p Otmto As was stated before, the elements aluminum, titanium, four quadlaus by vefglug ubSClSSa Values and Ordinate boron and molybdenum are significant elements in deter- Values ludelleudeuly. rl-ll@ quadrants are then numbered mining the strengih pmperties of nickel base alloys, consecutively counter-clockwise' beginning with the upper Siight Variations in percentages of these elements in alloys right quadrant. Points falling in the quadrants I and III such as ours produce signilcant dilerences in physical are designated positive and ll and lV negative. FTGURE sharpen 3 shows 35 positive values as opposed to only 2 negative values indicating a strong positive correlation.

This correlation was further substantiated by the fact that four different heats of material were vacuum cast into different shapes and tested by different sources. The results of such varied materials show the correlation.

ln Order to produce grains of a controlled size, we first melted our alloy in a Crucible and then cooled while still in a Crucible to a temperature just above the solidication point, a point at which both solid and liquid are in equilibrium. Thus we formed small solid particles or nuclei in the melt. Then we applied full power to our equipment to raise the temperature of the melt as fast as possible. After a predetermined time, we poured or cast the metal from the Crucible into a mold. The lapse of time between this latter increase in temperature of The chemical analyses of the examples of Table 4 above are shown in the following Table 5:

Table 5 [Composition in percent by Weight] Example C S Al Ti Mo Cr Fc B Si Mn 2- 0.003 4. 45 3. 60 5. 30 14. 4 4. 80 0Y 075 0. l0 0. 05 X 0. 004 4. 40 3. 65 5. 55 14. 6 4.55 0. 081 0.10 0.10 2- 0. 006 4. 35 3. 40 5.35 14. 3 4. 00 0. 080 0.10 0, 05 X 0.005 4. 3. 30 5. 40 15. 0 4. l0 OSO 0. l0 0. l0 X 0. .001i 4. 37 3. 33 5.25 14. 7 4. 50 0Y 076 0. 07 0. 05 X 0. 004 4. 30 3. 40 5. 45 14. 5 4. l0 0. 050 O. 10 0. 10 X 0. 005 4. 30 3. 45 5. 40 14. 5 4. 55 088 0. 10 0. 10

the melt and the time of pouring determined the size of grains produced since that period is the initial factor which controls the growth of the grains. Other factors such as the size of the section being poured or the temperature of the mold bear a relationship to the ultimate grain size. The time required for melting equipment to produce grains of a given size will vary from unit to unit depending on its power capacity and the like. Therefore, the period of application of the last surge of power must be calibrated for each unit to produce grains of a given size.

The resultant grains tend to be regularly shaped, sometimes called equi-axed grains rather than elongated, rod-like grains sometimes referred to as columnar grains.

The method which we used to measure average grain diameter and one which is highly reliable and reproducible is as follows: we first cut and polished a crosssection of a test bar normal to its longitudinal direction. After etching to distinguish grain structure, we then inscribed two perpendicular lines on that surface. We counted the number of grains in a measured length along one of the lines and divided that length by the number of grains counted. We then repeated the count for the other l'me and averaged the values together. For example, if five grains are found along one inch of the line, the average grain diameter would be 1/5 inch or 0.2 inch. The value thus obtained is the actual average grain diameter and not a reference number.

In FIGURE 3, the American Society for Testing Materials (ASTM) values are shown for comparison purposes along with our actual average grain diameter values. Therefore, the values for average grain diameter for our alloy which we have found produces increased stress rupture strength properties is greater in size than about ASTM macro-grain size number N-l3 or about 0.015 inch. Preferably the average grain diameter of our alloy should lie within the range between about ASTM macrograin size number lVl--6 and ivi-l2 or about 002-025 inch.

As shown in Table 4, as average `grain diameter decreases, stress rupture life decreases.

Although our alloy in the as-cast form exhibits unusual stress rupture properties, some of the heat treatment cycles which we studied led to our improving the stress rupture strength of our alloy even more over that of other currently available alloys. Table 6 gives examples of such cycles:

Table 6 Example: Heat treatment cycle A heat at 2l00 F. for l hour, air cool,

heat at 1950 F. for 2 hours, air cool, heat at l550 F. for 4 hours, air cool.

B heat at 2l00 F. for l hour, air cool,

heat at 1500c F. for l2 hours, air cool.

C heat at 2100" F. for l hour, air cool.

D heat at 2200 F. for 1 hour, air cool,

heat at l500 F. for l2 hours, air cool.

E heat at l500 F. for l2 hours, air cool.

The lengths of time referred to in Table 6 are the times at which the material being heat treated is in equilibrium with the temperature specified.

Although in lgeneral stress rupture strength decreases rapidly in excess of 15300 F., it has not been practical to conduct stress rupture testing at higher temperatures under an axial load of 15,000 p.s.i. However, the stress rupture properties of our alloy were so great that we were able to conduct stress rupture testings at l850 F. under an axial load of 15,000 p.s.i. The results of stress rupture testing of a number of samples heat treated accor ing to the heat treatments shown in Table 6 are shown in Table 7.

Table 7 [Stress rupture properties at 1850" F. and 15,000 p.s.1.]

Average stress rupture life (hours) Heat treatment:

Referring to Table 7, it is to be noted that an initial heating or solution-heat treating at about 2l00-2200 F.

will produce relatively high stress rupture properties. The heat treatment which we found best is shown as heat treatment C. lt involves only the heating at about 2100 F. for one hour and then air cooling. Most other cornmercially available high temperature alloys will require the use of a commercial furnace for approximately one day or longer in order to complete their heat treatment cycles sometimes referred to as double aging cycles. The introduction of a heat treatment cycle such as shown in Example C of Table 6 tto give exceedingly high stress rupture properties reduces the cost of the manufacture of articles by a substantial amount.

Although in the foregoing description We have disclosed in connection with specific examples an improved precipitation hardening ytype nickel base alloy and heat treatment for such alloy to produce stress rupture strength properties in excess of any nickel base alloys heretofore produced, the examples are to be construed as illustrative of rather than limitations on our invention. Those skilled in the art of metallurgy and heat treatment Will readily understand the modiications and variations of which our invention is capable. We intend in the appended claim to oever modiications and variations which come within the true spirit and scope of our invention.

What we claim as new and desire to secure by Letters Patent of the yUnited States is:

A precipitation hardening type nickel base alloy suitable for use at elevated temperatures consisting of, by weight, 0.05% carbon, 4f.1-4.8% iron, 15% chromium, 3.5% titanium, 0.08% boron, 4.5% aluminum, 5.3% molybdenum, with the balance essentially nickel and impurities; the alloy having an average grain diameter of 0.024025 inch.

References Cited in the tile of this patent FOREIGN PATENTS taly Dec. 1, 1954 Great Britain Mar. 13, 1957 OTHER REFERENCES

Images