UDC 669.018.472:629.73

PRECISION CASTABLE ALLOY OF INVAR CLASS

FOR OPERATING TEMPERATURES OF UP TO 500C

V. I. Chermenskii,1 I. V. Konchakovskii,1 S. V. Grachev,1 A. V. Maiorov,2 and P. S. Kuchin2

Translated from Metallovedenie i Termicheskaya Obrabotka Metallov, No. 10, pp. 50 53, October, 2010.

A castable invar alloy based on the Fe Ni Co C system is developed. The highest average values of the

coefficient of linear thermal expansion of the alloy are 5.2 10 6

K 1

, 6.3 10 6

K 1

, and 7.4 10 6

K 1

in the temperature ranges of 20 400, 20 450, and 20 500C, respectively. The alloy is designed for mold

casting of parts with elevated heat resistance including those for operation in a couple with ceramics, carbon

plastics, and other polymer composite materials (PCM).

Key words: castable invar alloys, temperature coefficient of linear expansion, thermal size stability.

INTRODUCTION

Precision alloys with specified thermal expansion

[termed precision alloys with specified coefficient of linear

thermal expansion (CLTE) in GOST 10994] are out of com-

petition from the standpoint of lowering of stresses in

metal-nonmetal coupling units both under stationary and

nonstationary operating conditions [1]. This is explainable

by the anomaly of their thermal properties known in scien-

tific literature as the invar effect. This means that in specific

temperature ranges invar, and especially superinvar, alloys

have a CLTE an order of magnitude lower than any other me-

tallic structural material, which makes their sizes stable un-

der conditions of heating or cooling. Invar alloys have a

CLTE close to that of ceramic, polymer, and composite mate-

rials and this makes it possible to solve problems of tempera-

ture-and-size compatibility of materials of different kinds in

high-technology articles.

In the middle 1970s researchers of the Ural Polytechnic

Institute managed to prove the possibility of creation of pre-

cision cast alloys with minimum temperature coefficient of

linear expansion. It has been assumed earlier that invar alloys

can only be deformable, whereas in a cast structure the CLTE

cannot be low due to segregation of elements. A study of the

problem of optimization of physicochemical and process fac-

tors has made it possible to find conditions for ensuring a

specified level of invar effect and other required properties of

cast invars and superinvars.

In the recent years the development of cast invars and

superinvars was directed at solving problems connected with

the operating capacity of intermediate (temperature compen-

sating) frames for quartz and glass ceramic panels of the

head part of aircrafts. This research gave alloys with low

CLTE in the ranges of 20 200 to 20 350C and solved the

problem of temperature-size compatibility of metallic frames

and nonmetallic shell. This is a result of maximum matching

between the CLTE of specific alloys and quartz and glass ce-

ramics under the heat load mentioned. In future the required

range of operating temperatures may reach 20 500C, and

this is en incentive for further research.

In this connection high-technology cast invar alloys with

hydrogen additives attract special interest. Among the deve-

loped invars and superinvars [2] they are the most promising

for making large and complex-configuration load-carrying

construction members operating in a wide range of negative

and high temperatures. Hydrogen alloying of invar and

superinvar alloys improves substantially their casting, weld-

ing, and cutting process properties and makes it possible to

cast shapes virtually without restricting the mass or the size.

Foreign researches study actively such carbon-bearing alloys

with a low CLTE (up to 5 10 8 K 1 ) in a temperature

range of 0 200C [ISO 2892-73(A)] and in a range of up to

400C [3].

Our earlier invars 31NKUL-1 and 31NKUL-2 [4] are no

inferior to such high-temperature alloys. Their mean CLTE

does not exceed 3.5 10 6 K 1 at a temperature ranging

from 60 to 300C and from 20 to 300C and

Metal Science and Heat Treatment Vol. 52, Nos. 9 10, 2010

504

0026-0673/10/0910-0504 2010 Springer Science + Business Media, Inc.

1Ural State Federal University in the Name of the First President

of Russia B. N. Eltsyn, Ekaterinburg, Russia (e-mail:

mvd@mtf.ustu.ru).2

Linvar Research and Production Center, Ekaterinburg, Russia

(e-mail: linvar@mail.ru).

4.5 10 6 K 1 at a temperature ranging from 60 to 350C

and from 20 to 350C. The chemical compositions of these

alloys are presented in Table 1.

On the whole, the invar, physical, mechanical, and

thermophysical properties of alloys 31NKUL-1,2 meet the

new aim of the study. But the experience of production of

castings from these alloys has shown that their crack resis-

tance is insufficient for large (> 0.5 m) aircraft frames. This

is connected with the high content of cobalt (8.5 ; 0.5%),

which widens the interval of the invar property due to raising

the Curie point and simultaneously causes formation of

cracks, especially in thin-walled and large-size castings. In

addition, cobalt increases the cost of the castings by 8 10%

per 1% of its content on the average.

In this connection, we decided to study the following

composition (in wt.%): Fe (32 34) Ni (6.0 8.0) Co

(0.75 1.5) C. Such an alloy contains less cobalt than the

31NKUL grades, which raises the crack resistance of the

castings to the required level. The nickel content is elevated

in order to preserve (and event to raise somewhat) the Curie

point. Homogeneity of the invar austenite in cooling to nega-

tive temperatures is guaranteed. The total content of Ni + Co

is the most favorable for minimizing the CLTE. Carbon

and the other elements are contained in an amount ensuring

the possibility of obtaining low CLTE, which we have

shown in [4].

METHODS OF STUDY

Solution of the problems posed in the present work re-

quires the use of informative and experimental methods.

Theoretical analysis of the possibility of obtaining minimum

possible values of CLTE is performed by simulating the de-

pendence of the value of linear expansion as a function of the

carbon content on the basis of the principle of additivity of

CLTE in heterogeneous systems. The reliability of analysis

of the dependence of CLTE on the temperature and on the

content of carbon and of the minimization factors was

checked under laboratory and industrial conditions.

We melted blend materials of commercial purity that

have been used for many years in the production of carbon-

less cast invars and superinvars, i.e., nickel of grade N1

(GOST 849), low-carbon steels 10895 and 10880 (GOST

11036), ferromanganese FMn80 (GOST 1415), ferrosilicon

FS75 (GOST 1415), and mish metal Mts40 (TU 48-4-28073)

containing REM. Carbon was introduced into the alloys in

the form of pure iron (semiproduct of the production of

ferroalloys). The possibility of obtaining the required values

of CLTE in the alloy was studied for various heats differing

in the content of the main components. In order to work in

the range of the most serviceable compositions we varied the

content of the elements in quite wide ranges, namely, 0 9%

cobalt, 31 37% nickel, and from < 0.05 to 2.0% carbon.

The alloys were melted in open induction furnaces under

laboratory and industrial conditions. Cast specimens for dila-

tometric and metallographic analysis were fabricated by cut-

ting from the castings or by vacuum pumping into quartz

tubes. Heat treatment was performed by the earlier devel-

oped two-stage regime, i.e., 4-h hold at 950 1000C and air

cooling followed by high-temperature tempering at 600

650C for 2 h. The first stage of such treatment is performed

for homogenizing the cast ingot, the second stage is aimed at

removing stresses and stabilizing the structure (this regime

was later used for production of cast frames).

RESULTS AND DISCUSSION

The chemical composition of the alloys and the values of

their CLTE are presented in Table 2.

We started the tests with alloy 10. The content of Ni and

Co in this alloy was equal to that of the chosen composition

but the carbon additive was absent. After casting, alloy 10

had CLTE20 100 = 1 10 6 K 1 in the range of 20 100C,

which met the minimum specified values of the GOST 10944

Standard. The carbon alloys in cast state had quite high

CLTE (& 8 10 6 K 1 ). This was explainable by the high

carbon content in the matrix and rapidly developing crystalli-

zation of the specimens. However, the effect of the heat treat-

ment on the carbon alloys was quite favorable.

The CLTE of the carbonless alloy 10 increased abruptly

after the heat treatment starting with 400C. The carbon-

bearing alloys had CLTE 2 3 times inferior to that alloy 10

at the temperatures of up to 100 200C and were superior

to it at 400C and higher temperatures. Physical explanation

of the positive role of carbon in our case involves the fact

that it widens the temperature range of the invar effect, be-

cause it is known [5, 6] to raise the Curie point. As for the ef-

fect of cobalt (Table 2), we obtained quite acceptable results

in alloys 50U and 60U with 6.4 and 8.0% Co respectively.

With allowance for the cost of the alloys, the upper boundary

(8% Co, alloy 60U) is not preferable from the standpoint of

the required values of CLTE20 400, 450, 500, because at 6.4%

Precision Castable Alloy of Invar Class for Operating Temperatures of up to 500C 505

TABLE 1. Chemical Composition of Earlier Developed Alloys

Alloy

Content of elements, wt.%

Ni Co C Ce Fe Mn Si S P

31NKUL-1 30.0 31.5 8.3 9.3 0.75 1.5 0.05 0.30 Res./ 0.40 / 0.50 / 0.02 / 0.02

31NKUL-2 30.0 31.5 8.3 9.3 0.75 1.5 Res./ 0.40 / 0.50 / 0.02 / 0.02

Co (alloy 50U) we obtained virtually equal values of these

coefficients.

When studying the influence of nickel, we established

that the most acceptable concentration of this element was

the one chosen in the beginning of our study. In order to ob-

tain the required CLTE the nickel content should neither be

increased no decreased with respect to the chosen range of

32 34%. Additives of carbon and silicon exceeding the ini-

tially chosen concentrations would be useful for the casting

properties of the alloy but influence negatively the

CLTE20 400, 450, 500 (alloy 70U in Table 2). In this case the

unfavorable growth in the CLTE cannot be prevented even

by raising the cobalt content to 9%.

It was important to study the relation between the

microstructure of the alloys and the values of CLTE obtained

after heat treatment and meeting our task (alloys 50U and

60U). We used the results of the study of the conditions of

minimization of CLTE in invar alloys with carbon [7, 8] and

chose the volume fraction of graphite formed in these alloys

as an appropriate criterion. We established that this quantity

has an optimum value depending on the composition and on

the conditions of structure formation; in one and the same al-

loy a minimum possible value of CLTE is hard to obtain both

at too low and at too high volume fractions of graphite.

The matrix of the alloys is represented by invar austenite

with dendritic structure. Results obtained for laboratory and

industrial heats show that at the used rates of cooling of the

castings the dispersity of the matrix does not affect substan-

tially the CLTE. The graphite inclusions have a compact

globular or vermicular shape.

Control check of the CLTE of the alloys in ranges begin-

ning with negative temperatures showed that the values pre-

sented in Table 2 for the ranges beginning with 20C are vir-

tually the same as those for the ranges beginning with

60C. The structure of the alloys remains invariable and

contains no martensite phase.

Thus the compositions of 50U and 60U meet the

sought-for properties and we have taken them as the base

variant. As a result of the subsequent field tests we chose the

following proportion of the main components (in wt.%):

32.0 34.0 Ni, 6.0 8.0 Co, 0.75 1.5 C. In addition, the al-

loy should contain low amounts of Mn and Si, possess com-

mercial purity with respect to the other impurities, and con-

tain cerium and other REM in an amount of 0.05 0.30%.

After heat treatment the castings has a mean CLTE no

higher than 5.2 10 6 K 1 for the temperatures ranging

from 60 to + 400C or from 20 to 400C, no higher than

6.3 10 6 K 1 for the ranges from 60 to + 450C or from

20 to 450C, and no higher than 7.4 10 6 K 1 for the

ranges from 60 to + 500C or from 20 to 500C.

The Curie point of the alloy is 625 K (352C); the modu-

lus of longitudinal elasticity is 125 GPa; the ultimate rupture

strength (at room temperature) is 370C, the elongation is

8%, and the HB hardness is 130 kgfmm2.In addition to the scientific tasks we have solved a num-

ber of problems connected with the process of production of

castings from the alloy developed. The results of the tests

performed in the present work, the available production data,

and the results of shop tests allowed us to suggest a process

for casting preforms by a centrifugal method and for chill

casting and to fabricate a series of pilot castings.

The expedience of the principle of minimization of

CLTE in carbon-bearing invars at elevated temperatures

based on optimum combination of alloying elements raising

the Curie point has been confirmed. The purity of the blend

materials and the modes of melting, casting and heat treat-

ment are important factors of the technology.

CONCLUSIONS

We have studied a precision cast alloy of the Fe Ni

Co C system with low CLTE in a wide temperature range

ending with 500C and worked out a process for fabricating

castings from this alloy. The alloy contains 32 34% Ni and

6 8% Co, is alloyed with 0.75 1.5% C, bears low contents

of Mn and Si, and possesses commercial purity with respect

to the other impurities. The values of the CLTE in the tem-

perature ranges of 20 400, 20 450, and 20 500C en-

sure reliable enough operation of the alloy upon heating in a

couple with structural glass ceramics. Its physical, mechani-

cal, thermophysical, and process properties are comparable

on the whole with the properties of other cast invars.

506 V. I. Chermenskii et al.

TABLE 2. Chemical Composition and CLTE of the Alloys Studied

Alloy

Content of elements, wt.% Mean CLTE 106, K 1, in temperature ranges, C

Ni Co C Mn Si 20 100 20 200 20 300 20 400 20 450 20 500

10 32.9 6.2 < 0.05 0.32 0.29 1.55 1.70 3.22 6.31 7.50 8.63

20U 37.0 0.0 0.93 0.20 0.40 4.05 4.18 4.25 6.26 7.22 8.21

30U 36.8 2.0 0.98 0.18 0.36 4.53 4.50 4.48 5.86 6.81 7.86

40U 35.0 4.3 1.20 0.17 0.37 4.31 4.03 4.05 5.28 6.25 7.38

50U 33.2 6.4 1.00 0.18 0.32 3.76 3.61 3.4 4.91 5.91 7.08

60U 31.9 8.0 1.00 0.22 0.32 3.02 2.95 2.94 4.82 5.88 7.05

70U 31.0 9.0 2.00 0.30 1.00 3.00 3.26 3.71 5.81 6.93 7.90

Foreign experience proves good prospects of making

precision large-size articles such as substrates, mandrels, and

bulk press molds for vacuum, press, and autoclave forming

of polymer composite materials. Our experience of produc-

tion of thermally stable fittings from invar alloys with carbon

for making carbon plastic reflectors for satellite antennas [4]

and casing parts of helicopters allows us to infer that the ope-

rating and technological characteristics of the studied alloy

are suitable for using it in the production of composite struc-

tures from high-temperature PCM.

REFERENCES

1. S. V. Rabinovich, M. D. Kharchuk, V. I. Chermenskii, and

R. A. Sidorenko, Precision cast alloys with specified thermal

expansion, Liteinoe Proizvod., No. 6, 16 18 (2002).

2. S. V. Rabinovich, M. D. Kharchuk, V. I. Chermenskii, and

R. A. Sidorenko, New direction in the technology of precision

alloys with specified thermal expansion, in: Vestnik

UGTU-UPI, Design and Technology of Machine Production [in

Russian], GOU VPO UGTU-UPI, Ekaterinburg (2005), Part 1,

No. 18(70), pp. 206 217.

3. S. Enomoto, Cast iron with low coefficient of linear expansion

Nobinite, Sokeydzay, 29(9), 16 22 (1988).

4. V. I. Chermenskii, I. V. Konchakovskii, S. V. Rabinovich, et al.,

Carbon invar alloys 32NKUL and 33NKUL for casting ther-

mally stable parts, Aviats. Prom., No. 3, 37 39 (2008).

5. I. Ya. Georgieva and O. P. Maksimova, On the effect of carbon

on the Curie point of austenite in iron-nickel alloys, Fiz. Met.

Metalloved., 24(3), 574 576 (1967).

6. V. M. Kalinin and V. P. Beskachko, On the problem of the Curie

point of iron-nickel invar alloys with third component, Fiz. Met.

Metalloved., 36, 73 78 (1973).

7. V. I. Chermenskii, I. V. Konchakovskii, S. V. Rabinovich, et al.,

New high-technology invar alloys of the Fe Ni Co C sys-

tem, in: Novel Materials, Nondestructive Inspection, and High

Technologies in Machine Building, Mater. III Int. Sci.-Eng. Conf.

December 2005, Tyumen [in Russian], Felix, Tyumen (2005),

pp. 20 21.

8. V. I. Chermenskii, I. V. Konchakovskii, S. V. Grachev, et al.,

Cast alloys of the Fe Ni Co C invar class for operating

temperatures up to 200 300C, Izv. Vysh. Uchebn. Zaved.,

Neft Gaz, No. 4, 78 80 (2007).

Precision Castable Alloy of Invar Class for Operating Temperatures of up to 500C 507