US3650846A

US3650846A - Process for reconstituting the grain structure of metal surfaces

Info

Publication number: US3650846A
Application number: US773135A
Authority: US
Inventors: William P Holland; Robert E Hueschen
Original assignee: General Electric Co
Current assignee: General Electric Co
Priority date: 1968-11-04
Filing date: 1968-11-04
Publication date: 1972-03-21
Anticipated expiration: 1989-03-21
Also published as: FR2022546A1; GB1285203A; AT301291B; DE1952854A1

Abstract

Surface brittle fracture of metals is inhibited by impinging on the surface of the metal an intense pulsed beam of charged particles, neutral particles or radiation while the metal is maintained at a temperature at least above its nil ductility temperature. Energy from the pulsed beam creates high thermal gradients and concomitant high stresses which in turn effect high plastic strain in the surface region and produce a unique fine subgrain structure on and immediately underneath the surface. The treatment enhances the resistance of the metal surface to brittle fracture and results in decreased cracking, delaminating and grain lifting.

Description

United States Patent Holland et al. 1451 Mar. 21, 1972 [541 PROCESS FOR RECONSTITUTING THE 3,158,513 11/1964 Janssen et a1 ..14a/13 GRAIN STRUCTURE OF METAL 3,240,639 3/1966 Lihl ....148/143 SURFACES William P. Holland, West Redding, Conn.; Robert E. liueschcn, Hales Corners, Wis.

General Electric Company Nov. 4, 1968 inventors:

Assignee:

Filed:

Appl. No.:

References Cited UNITED STATES PATENTS 1/1961 Steigerwald 148/13 Primary Examiner-Richard 0. Dean Attorney-Jon Carl Gealow, Arthur V. Puccini, Frank L. Neuhauser, Oscar B. Waddell and Joseph B. Forman [57] ABSTRACT Surface brittle fracture of metals is inhibited by impinging on the surface of the metal an intense pulsed beam of charged particles, neutral particles or radiation while the metal is maintained at a temperature at least above its nil ductility temperature. Energy from the pulsed beam creates high thermal gradients and concomitant high stresses which in turn effect high plastic strain in the surface region and produce a unique fine subgrain structure on and immediately underneath the surface. The treatment enhances the resistance of the metal surface to brittle fracture and results in decreased cracking, delaminating and grain lifting.

9 Claims, 8 Drawing Figures Patented March 21, 1972 r 3,650,845

2 sheet sheet 1 @160 rzag Patented March 21, 1-972 2 Sheets-Sheet z PROCESS FOR RECONSTITUTING THE GRAIN STRUCTURE OF METAL SURFACES BACKGROUND OF THE INVENTION There are many instances in industry where metal surfaces are suddenly subjected to large temperature changes. These temperature changes often occur in a cyclic manner resulting in extremely high thermal gradients being periodically created in the surface region of the metal. The accompanying mechanical stresses are often sufficiently large to cause initiation and propagation of brittle intragranular or intergranular fracture of the metal surface. As this fracture process is continued, microscopic surface cracks, lifting of surface grains, delamination of the surface material and other distortions of the surface result. cross section.

Missile shield materials, aircraft turbine blades, metal molds, electric switch contacts andX-ray tube targets are but a few of the many examples in which functional degradation results from surface fracturing. The X-ray tube target is a case of unusual severity and serves as a good example. At the initiation of an X-ray exposure, a highly energetic beam of electrons suddenly impinges on the surface of a relatively cool metal target. Electron beam power densities at the target surface are typically from 10 to watts per square centimeter. Thermal gradients at the surface may be in the range of 50,000to 100,000 Kelvin per millimeter. The high mechanical stresses created by the high thermal gradients may exceed the strength of the target material and ultimately cause plastic flow and/or brittle fracture at or near the target surface.

Distortion of the target surface caused by the fracturing results in a marked decrease of X-radiation output intensity, since the probability of an X-ray photon escaping from the target is markedly less for a rough target surface than for a smooth surface. After the target surface is distorted and roughened, the power density in the electron beam must be increased to maintain the necessary X-ray output intensity. The higher power densities yield higher thermal gradients which create greater mechanical stresses which in turn cause more fractures to initiate and propagate in the target surface. This cyclic process leads to rapid failure of the X-ray tube. Thus, the inhibition of brittle fracture which leads to surface disruption is greatly to be desired in a high power X-ray tube target. Moreover, brittle fractures can propagate into the subsurface regions of the target and eventually cause the separation of the target into two or more pieces. Prevention of this catastrophic type of failure of an X-ray tube is also greatly to be desired, especially in those targets which are rotating at speeds up to 11,000 r.p.m.

Terminology used in this specification will now be defined in reference to FIG. 8 of the drawings. This figure shows a typical plot of ductility versus temperature of a metal. Below temperature A, the metal is quite brittle; this temperature region is called the brittle range. Above temperature C, the metal is quite ductile; this temperature region is called the ductile range. Between temperatures A and C there is a region in which the ductility of the metal increases rapidly; this temperature region is called the transition range. The transition range is also called the ductile-to-brittle transition range in the literature and sometimes the temperature that corresponds with fifty percent ductility on the linear part of the curve is arbitrarily taken as the ductile-to-brittle transition temperature (DBTT).

Plots of ductility versus temperature are generally S-shaped with an almost linear, high positive slope portion in the transition range. Nil ductility temperature (NDT) can be defined as that temperature at which a linear extrapolation of the linear, high slope portion of the ductility temperature curve intersects the temperature axis at a point which is marked B.

The position of the curve with respect to the temperature axis and the slope of the linear portion depends on several factors such as the type of metal, its degree of cold work, its impurity content, the rate at which stress is applied during testing and others. Ductility may be expressed in several ways. In the curve shown, ductility may be considered as being expressed in terms of percent reduction in cross-sectional area at the fracture interface when fracture occurs. Greater ductility is then indicated by greater reduction of area. Ductility may be expressed in terms of bending angle if a bending test is used.

If the shape of the curve is different because of one or more of the factors mentioned in the preceding paragraph, the NDT or point B would shift along the temperature axis, but would, nevertheless, be a rather definite temperature regardless of the magnitude of the curve.

Metallurgists have attempted to solve the surface fracturing problem in high temperature applications of metals by increasing the purity of the metal, decreasing the grain size, redistributing impurities, controlling grain orientation, cold working and alloying. All of these techniques tend to lower the nil ductility temperature (NDT), above which metals are increasingly ductile and below which they are brittle. When a metal is in a more ductile state and is subjected to high thermal gradients at its surface, stresses are relieved by slip rather than by brittle fracture. When slip occurs, the metal deforms plastically and the stress is relieved before brittle fracture can occur. in general, the tendency for brittle fracture is reduced by lowering the NDT of a metal because the metal exists in a brittle state for a shorter period during a given heating cycle. The ideal condition is to have the NDT of the metal below any temperature in its operating range.

The surfaces of X-ray tube targets are usually made of high atomic weight refractory metals such as molybdenum and tungsten. These metals are particularly susceptible to brittle fracture because they have relatively high NDTs as conventionally processed. Moreover, in the course of normal usage the grain structure of a tungsten X-ray tube target is subject to grain growth, thereby increasing the NDT even further. For this reason, surface recrystallization by a technique such as that suggested in US. Pat. No. 3,158,513 does little to diminish the tendency of the target surface to undergo brittle fracture.

Another approach to reducing brittle fracture in diagnostic X-ray tubes which has been put into practical use is to alloy elements such as rhenium with tungsten. Reduced brittle fracture obtained in this way is at the expense of a somewhat lowered melting point and higher vapor pressure for the alloy. Thus, when the X-ray tube is operated at the high voltage and current levels which radiologists often desire, the target surface melts where the electron beam is focused or concentrated. Even without surface melting, evaporation of the alloy from the focal spot region may degrade the electrical insulating qualities of the X-ray tube. One less than satisfactory solution, of course, is to derate the X-ray tube and sacrifice output for extended life.

Other known cases of metals undergoing brittle fracture under high thermal stress will not be discussed for the sake of brevity and because they are known to those practicing the metallurgical arts. X-ray tube targets are used herein for exemplifying the problem of brittle fracture of metals and as a basis for describing the solution achieved by the present invention.

SUMMARY OF THE INVENTION A primary object of the present invention is to provide a method for treating metal surfaces that results in inhibition of brittle fracture when the metal is subjected to high thermal stress and that imparts other desirable properties to the metal surface. Stated more specifically, an object of the invention is to increase the number of subgrains on and near the surface, to reduce impurities, to redistribute residual impurities and to impart a prestressed condition in the surface region. A smaller grain size is shown in the literature to result in a lower NDT, perhaps by as much as 200C. for each tenfold reduction in grain size. Better purity is believed to result in a more orderly atomic lattice arrangement within the subgrains and also to contribute to reduce NDT and improved ductility and toughness. The increased number of subgrains may afford a greater surface area and intergranular volume for distributing those impurities that remain after the surface treatment is carried out. The prestressed condition, which is believed to result from the increased number of subgrains, is compressive in nature and produces a shear force between the surface layer and the base metal. Upon heating the surface, much of the thermally inducted stress is used up in overcoming the prestressed condition so that the total shear stress is lower than that which occurs in the absence of prestressing. The lesser net final stress means a reduction in the forces that tend to delaminate, lift and fracture the surface grains.

Briefly stated, the new method is characterized by slowly raising the temperature of a metal object to be treated to its NDT or higher and preferably to a temperature in its ductile range. This is usually done while the metal is in a vacuum ambient although the metal may be in a gas in some cases. Then,

high thermal energy pulses are induced in the surface of the metal at an appropriate repetition rate and for a controlled period. The energy may be radiant or particulate as long as it is or converts to thermal energy in reacting with the surface. For instance, the energy may be derived from a laser although it is preferable to employ a concentrated electron beam. It is important that the metal be at least above its NDT and preferably in its ductile range before thermal pulses are initiated in order to be sure that brittle fracture will be avoided.

In the electron beam method there is a cathode spaced from the work piece in the vacuum and the work piece is at a positive potential to accelerate the electrons. The electron beam may be deflected over the surface if desired or the work piece can be moved in the beam or a very powerful beam may cover the whole area to be treated at one time. Beam current intensity may be controlled by modulating the cathode temperature or by use ofa control grid.

When the object is in an adequately ductile state, a pulse or a short series of concentrated electron pulses are impinged on the metal surface. The pulses are followed by an interpulse period after which pulsing is repeated. These alternate periods of high concentrated heating and cooling may be repeated for different lengths of time, depending on the nature of the metal, and generally run from 15 minutes to about 2 hours, as an illustration, although shorter or longer treatment periods are dictated in some cases. After treatment, the metal is used in its normally expected way and it has the desirable properties which are discussed above.

A further discussion of steps involved in a preferred way of practicing the new process in respect to various metals, the results achieved thereby, the apparatus required therefor and what appears to be a plausible theory for the unexpected results obtained by the process, based on the present state of knowledge, will now be presented in reference to the drawings.

DESCRIPTION OF THE DRAWINGS FIG. 1 is a schematic diagram of one type of apparatus that may be used to practice the new metal surface treatment method;

FIG. 2 is an alternative type of apparatus, with parts omitted, for practicing the new method;

FIG. 3 is a photomicrograph ofa cross section ofa tungsten object which shows the grain and subgrain structure magnified 100 times and has been treated according to the invention;

FIG. 4 is a photomicrograph ofa cross section ofa tungsten X-ray tube target showing its structure at and immediately under its surface after having been treated according to the invention, the structure being magnified 500 times;

FIG. 5 is a photomicrograph ofa cross section ofa tungsten target magnified 500 times and illustrating the appearance of the structure of the object shown in FIG. 4 before the new surface treatment and as it exists at a depth in the object which is beyond the penetration range of the surface treatment;

FIG. 6 is a photomicrograph at a magnification of 500 times, of a surface region cross section of molybdenum object that has been treated according to the invention;

FIG. 7 is a photomicrograph, at a magnification of 500 times illustrating the structure of the molybdenum object of the preceding figure as it appears at a depth beyond the penetration range of the surface treatment; and,

FIG. 8 is a plot of ductility versus temperature for a representative metal.

DESCRIPTION OF THE PREFERRED EMBODIMENT FIG. 1 shows apparatus set up for treating the upper surface of an X-ray tube target 1. The treatment is intended in this case to cover at least the circular focal track region of the tapered surface on which an electron beam impinges to produce X-rays when the target is installed in a rotary target X-ray tube.

In this example, the metal target 1 is adapted for being rotated while its surface is being subjected to pulses of electrons in connection with the new treatment. For this purpose, the target 1 is held with a nut on a metal stem 2 which extends from an induction rotor 3. The rotor 3 has internal bearings 4 which journal it to a stationary shaft 5. Target 1, stem 2 and rotor 3 rotate together.

The rotor is encircled in a vacuum-tight casing that includes a nonmagnetic cylinder 6 which is sealed at its lower end by a cap 7 and at its upper end by a thin ring 8. The latter is grazed in the bore of a heavy ring 9 which is gasketed and bolted to base plate 10. The base plate has a bell-jar-like enclosure 11 bearing on it with a vacuum-tight gasket 12 intervening. Since X-rays are emitted from the target during the surface treatment, it is desirable to preclude radiation by making enclosure 11 of sufficiently thick metal. Ifenclosure 11 is made of an X- ray transmissive material such as glass, an additional X-ray shield, not shown, should be installed around the apparatus. Refractory metal radiation shield 42 may also be used to surround the target to prevent vacuum gasket 12, and other system components from overheating.

The enclosure 11 may also be equipped with devices such as 13 for admitting a sensor, not shown, that measures the temperature of target 1 during treatment and that permits viewing of the target, if desired, when no X-rays are being produced. An X-ray absorbing lead glass window may also be installed in device 13 for making observations during the treatment.

The interiors of cylindrical enclosure 11 and the casing around the rotor 3 are evacuable through a pipe 14 which leads to a conventional vacuum system, not shown. The rotor 3 and target are rotated by induction with a stator assembly 15 which is analogous to that used in connection with rotating target X-ray tubes. With 60 Hz. to Hz. applied to the stator, the rotor 3 is generally rotated at about 3,000 to l0,000 r.p.m., accounting for slip during treatment. Other rotational speeds may be used depending on the choice of some other parameters as will appear later.

In accordance with the invention, it is necessary to induce cyclic high thermal gradients in the surface region of the metal object being treated while the metal is at least above its NDT and preferably in its ductile range in order to create subgrains and bring about the desirable resistance to surface fracturing.

In the FIG. 1 example, an electron beam is used to preheat and induce high thermally caused stresses in the target surface. The electron beam 16 emanates from a heated filament 17 which is supported from an insulating disk 18 that is mounted on a beam focusing cup 19. This type of electron gun is functionally the same as those that are commonly used in X- ray tubes. Metal focusing cup 19 is supported on metal rods 20 which attach to a support bracket 21. The bracket is held with a nut on an insulator feed-through 22 which is sealed in the base 10 of the apparatus.

A bare nickel wire 23 for one side of filament 17 extends through insulator 24 in base 10. The other side of the filament 17 is connected at a point marked 40 to support bracket 21.

One side of the filament 17 is connected to the secondary of the stepdown transformer 25 by the use of wire 26 from the atmospheric end of feed-through insulator 24. The other side of filament 17 is connected to the secondary of transformer 25 by use of wire 41 through feed-through insulator 22. The electron beam current 16 is set by the temperature and emissivity of the filament l7 and this is controlled in this example by adjusting a variable resistor 27 in the primary of filament transformer 25. The primary is supplied from an autotransformer 28 which may be energized from a 60Hz. power line through a disconnect switch 29.

A relatively high full wave rectified electron accelerating voltage is applied between focusing cup 19 and lower cap 7, and hence, target 1 through the agency of a pair of conductors 30 and 31 extending from the DC output terminals of a rectifier bridge 32. The bridge is supplied with AC from the secondary winding of a step-up transformer 33. The secondary is split at the center to include a milliameter 34 which effectively reads the electron beam current 16. The primary of transformer 33 is supplied from autotransformer 28 through a fastacting relay switch 35. Accelerating voltage may be set by adjusting a tap 36 on autotransformer 28. Switch 35 may be a semiconductor type, such as a silicon-controlled rectifier, which is gated on and off to produce electron pulses for the surface treatment which are of any desired pulse duration and repetition rate. Applied voltage may be read on a voltmeter 37 which is scaled in terms of peak voltage that appears across the secondary of transformer 33.

Relay switch 35 is symbolized as being operable by a solenoid coil 38 which is energized intermittently in accordance with the setting of an interval timer 39. Knobs and scales on the interval timer 39 are intended to symbolize that the timer can be set to produce pulses of any desired duration and repetition rate. The interval timer is a type that is well-known to those who are involved in manufacturing X-ray exposure interval timers and need not be explained further. In fact, the whole apparatus is analogous in function to an X-ray tube and its power supply which suggests that the new method of surface treating X-ray targets may be carried on in the X-ray tube itself as an adjunct to the usual degassing and seasoning process. It is probably more desirable, however, to perform the surface treating method in the vacuum chamber 11 of FIG. 1 to avoid reduction of tube filament life and to avoid condensation of filament vapors and extracted nongaseous target impurities on the glass X-ray tube envelope,

A tungsten surface, prior to treatment in accordance with the new method, will usually have an NDT around 200C. to 500C. Most grades of tungsten are extremely brittle at room temperature which accounts for the bad surface fracture that occurs in use when untreated targets are subjected to rapid heating and cooling cycles which produce damaging high thermal stresses ordinarily. Pure, single crystals of tungsten may have an NDT near room temperature. Molybdenum, which has a body centered cubic crystalline structure like tungsten, has an NDT usually between room temperature and about 150C. Beryllium, a close-packed hexagonal metal usually has an NDT between 300C. and 500C. The NDT and, more commonly, the DBTT of various metals are given in the literature. Generally, a reduction of either or both grain size and impurities results in a lower NDT. The face centered cubic metals such as nickel, are usually ductile and above their NDT at room temperature or lower so that preheating while carrying out the new surface treatment is unnecessary.

Target 1 may be raised to above its NDT and preferably in its ductile range with the apparatus in FIG. 1 by heating the target with a low energy electron beam. To illustrate, the accelerating voltage may be set at about 100 kilovolts peak and the current in beam 16 may be set at about milliamperes while target 1 is rotating at 3,000 to 10,000 r.m.p. This relatively small power input will heat rather massive tungsten targets to as much as 500C. in five minutes. Gradual heating in this phase of the process with low beam current avoids producing high thermal gradients and concomitant stresses in the target and avoids surface melting in the focal region of the electron beam. The electron beam current and voltage for preheating will depend in practice on several variables such as the metallurgical characteristics of the object being prepared for treatment and on its mass and elemental constituency. Other methods of preheating may occur to anyone desiring to perform the new method with other types of apparatus.

After raising the temperature of the object to NDT and preferably to its ductile range, its surface is treated with higher energy electron beam pulses for a relatively long period of time and preferably in a vacuum. As an example, a certain type of tungsten is subjected to beam pulses of kilovolts peak and average current of 600 milliamperes. This produces subsurface temperatures of about 3,000C. in about 200 microseconds in the region of the beam spot. Each pulse in this case consists of four half waves resulting from rectification of two full wave cycles. With a 60 Hz. power supply, pulse duration is one-thirtieth of a second. With a pulse repetition rate of 30 pulses per minute (r.p.m.) there is one pulse every 2 seconds. As a first approximation, if the target 1 is rotating at near synchronous speed of 3,600 rpm, or 60 revolutions per second, each revolution occurs in one-sixtieth of a second and the pulse is on for one-thirtieth of a second or during two revolutions after which it is off for nearly 2 seconds. Pulsing was continued for 3,000 exposures or about 100 minutes in one example. The foregoing data is primarily for acquainting the reader generally with the method that is applicable to one type of tungsten and is by no means an exclusive example. In general, the energy input is adjusted to provide as high thermal gradients in the metal surface as is possible without melting or otherwise damaging the surface.

As a general rule, proper beam energy for carrying out the surface treatment can be determined by adjusting the beam current and voltage until a throwaway sample piece exhibits slight surface melting and then reducing the beam energy to about percent of that energy so melting of the pieces in a production run will be avoided. After the pulsing procedure is underway, the target or other object is maintained adequately ductile by the electron pulses.

It is important to recognize that mere indiscriminate heating of a metal surface to a high temperature will not effect the desired grain structure. According to the invention, it is necessary to heat the surface transiently and cyclically, that is, the method involves heating intensely for a very short time and then discontinuing application of the energy that creates the heat for a short time. It is the short heating intervals which create the high thermal gradients and the high stress and plastic strain which effect the change in structure progressively in small volume increments near the surface. Despite the fact that temperature gradients are extremely high, the temperature of the bulk of the metal near the surface does not reach the point where melting occurs. Moreover, even though total energy in any heating interval is great enough to recrystallize the metal, recrystallization should be and is inhibited in the new method as a result of the short heating intervals. This is because the rate at which recrystallization proceeds and the resulting crystal size depends not only on exceeding a certain temperature, but also on the time during which the temperature is maintained. It is the product of time and temperature that governs recrystallization. A convenient way of saying it is that the time averaged temperature of the object being treated must not exceed its recrystallization temperature.

If high energy input causes recrystallization when practicing the method, and such is not desired, either the electron current density or the duration of the pulses or both may be reduced. External cooling may also be applied. Other alternatives are to defocus the beam or to increase the rate of relative movement between the beam and the object being treated.

When a metal is exposed to energetic pulses of radiation under conditions outlined in general terms above, a new structure is formed in the surface layers of the metal as manifested by photomicrographs and other evidence to be discussed later.

Although the crystal reconstitution mechanism is not susceptible to positive proof, one hypothesis is that the unique structure is produced in the following manner. The radiation beam is converted at the surface to thermal energy having a power density approximately 10 watts per cubic centimeter or a power flux of 10 watts per square centimeter. Resulting high temperatures concentrated in small regions of the surface produce huge thermal gradients in the surface. Under these conditions, vacant atom sites are created and there is a prolific production, motion and annihilation of dislocations in the crystals.

The concentration (C) of vacant atom sites in a crystal increases with increasing temperature by the relation C A exp (E,/kT) where A is an entropy factor, E, the formation energy of a vacancy, (approximately equal to l ev.), and k is Boltzmans constant 1.38054 X10 ergs/deg. Kelvin. One may see that large vacancy concentrations can be produced near the melting point of the metal as is the case with the new method. That these point defects grow into clusters and dislocation loops is well-known.

The number of dislocations N that can be produced by the cyclic input of electrons in the metal are directly related to the shear stress developed in the metal. The plastic strain 2, is proportional to the shear stress and depends on thermal gradients and the thermal expansion of the metal:

N w e, w (a) (AT) where AT is the temperature gradient and a is the coefficient ofthermal expansion in the same units.

Thus, the higher the temperature and the higher the thermally induced stress at the surface, the greater will be the number of dislocations produced. The dislocations are free to move, subject to local stress, and aggregate at elevated temperatures to create pronounced new boundaries within an existing grain or crystal. As thermal cycling continues, new vacancies are created which can again diffuse to dislocations, or to the surface or fall into new dislocation loops or other defects. If the metal were not held at least above its NDT, the crystals would not deform plastically or ductilely to any significant degree under the rapidly applied high thermal stresses and they would undergo brittle fracture. Brittle fracture delaminates the grains in the surface of an X-ray tube target during treatment just as it does when an untreated target is subjected to cyclic exposure.

An important aspect of the new method is to obtain sufficiently high power density and high thermal gradients between the surface and substrate layers. For the case of electron heating, depth of penetration is inversely proportional to the atomic number of the metal and directly proportional to the accelerating voltage. Greater penetration means lesser power density, but one cannot compensate in most cases by increasing beam current or intensity because the correspondingly higher power that results may melt the metal. For instance, molybdenum has a lower density than tungsten and is, therefore, penetrated more deeply. If the same power density were used for surface treating both, molybdenum might melt. Thus, a reduced power density must be used for treating molybdenum. Both types of metals are usually treated at a level to maximize the plastic strain without causing brittle fracture and without melting.

An alternative form of apparatus for practicing the new method is shown in FIG. 2. Here, the upper end of the belljar enclosure 11 has an electron gun 46 mounted in it. The gun assembly has horizontal and vertical deflecting plates for sweeping the electron beam 16 over the stationary object 49 being surface treated in the manner that a raster is produced on a television picture tube screen. The beam 16' may be a continuous DC beam because the cyclic heating and cooling of surface regions is effected by instantaneous exposure followed by movement of the sharply focused beam. One may also sweep the beam in one direction and translate the object to distribute the beam spot over the surface. The beam may be swept over the object until the grain structure is converted to a depth that is satisfactory for the intended use of the object.

In FIG. 2, base 10' has a cavity 50 through which cooling water may be circulated by means of inlet and outlet pipes 51 and 52. A connection 14' is provided for evacuating the chamber before and during the surface treatment method. The beam sweeping voltage generators and the power supply are omitted from FIG. 2 because they are conventional.

The rates of relative motion between the beam and the treated object and the beam current voltage and current values may be adjusted in both the FIGS. 1 and 2 embodiments so that the power density and exposure intervals are equivalent to each other, in which case, the same results can be produced with each. In either case, the thickness of the layer of subgrains depends on the duration of the treatment and is not limited by electron penetration. Generally, sharper rise time of surface temperature in a zone results in more rapid development of subgrains.

EXAMPLES 1. A recrystallized tungsten X-ray tube target was seasoned conventionally by heating in a vacuum ambient for several hours and then cooled to near room temperature. It was then raised to 800C, which is in its ductile range, and maintained at that temperature by an external power source. It was then exposed to kilovolts peak (KVP), 1,000 milliampere pulses of 0.0 l 2 seconds duration at a repetition rate of two exposures per minute until 7,613 exposures were accrued. The target was rotating at about 3,600 r.p.m. A full wave rectified, threephase, 60 Hz. power supply was used. The electron beam size was about 2 millimeters by 9 millimeters in cross section. A section through the electron focal track area was removed from the target and prepared for a photomicrograph. There is an average of 9,770 subgrains per square millimeter over an average depth of 0.0025 to 0.0035 inch from the surface. The unconverted substrate grain size is 990 grains per square millimeter, almost 10 times as large as the converted layer. The photomicrograph is shown in FIG. 3 herein.

2. A recrystallized tungsten X-ray target was seasoned in the manner of example number one. It was raised to about 800C, which is in its ductile range, and exposed to 80 KVP, 600 milliampere pulses of one-thirtieth second duration at a repetition rate of 20 exposures per minute until 3,000 exposures were accrued. The target was rotating at about 3,600 r.p.m. and a full wave rectified, 60 Hz. single-phase power supply was used. The electron beam was 2 millimeters by 9 millimeters in cross section. A section was taken from the target and a photomicrograph prepared. To an average depth of 0.0045 inch from the surface there is an average of 9.020 subgrains per square millimeter and as many as 10,070 subgrains per square millimeter in some incremental areas. The surface layer photomicrograph is shown in FIG. 4. The unconverted substrate grain count averages 990 grains per square millimeter as shown in the FIG. 5 photomicrograph.

3. A molybdenum target was seasoned by heating it in a vacuum with a kilovolts peak, 4 milliampere beam for one-half hour. Being at a temperature in its ductile range it was exposed to 55 KVP, 600 milliampere pulses of one-thirtieth second duration at 20 exposures per minute for 1,000 exposures followed by 2,000 exposures at 55 KVP, 575 ma., one-thirtieth second pulses, 20 per minute. Treatment continued at 55 KVP, 550 ma., one-thirtieth second pulses, 20 per minute until 12,000 exposures had accrued. Beam spot size was 2 millimeters by 9 millimeters. A single-phase, 60 Hz., full wave rectified power supply was used. The target was rotated at about 3,600 r.p.m. At an average depth from the surface of 0.007 inch and ranging from 0.0061 to 0.009 inch the count is about 7,300 subgrains per square millimeter as illustrated by the photomicrograph in FIG. 6. The grain count in the unconverted substrate beneath the surface layer is about 1,260 grains per square millimeter as shown in FIG. 7.

4. A nickel sample was placed in the chamber of FIG. 2 which was then evacuated. The DC electron beam was adjusted to 35 KV and 65 ma. The circular beam cross section was 0.020 inches in diameter. The sample was moved at a speed of 0.33 inch per second as the beam was swept over it at 109 Hz. and blanked out on the back sweep. The relative motion between the beam and sample was such that the beam spot exposed any point on the sample about seven times as the object passed under the beam one time. The sample was passed 100 times, thus exposing each point on the sample 700 times. Nickel has a DBTl below room temperature and, therefore, required no preheating. The bottom of the sample was water-cooled to preclude melting and enhance the thermal gradient. A photomicrograph showed that the fine structure was produced in a layer 0.002 inch thick measured from the surface.

5. Another nickel sample was treated in the manner of preceding example No. 4 except that it was passed through the region swept by the beam 50 instead of 100 times. A photomicrograph showed that the time converted structure was produced to a depth of 0.0017 inch measured from the surface.

DISCUSSION OF EXAMPLES The foregoing examples and other experiments indicate that no exclusive set of parameters need be stated for surface treating any metal. Variations in one or more parameters are allowable without sacrificing results. An underlying principle, however, is to produce repetitive, intense and transient heating and cooling cycles with high power density in regions of the surface being treated while the metal is at least above its NDT and preferably in its ductile range. If electron heating is used, the required power density may be obtained with relatively low current if the beam has a small diameter and with a proportionately higher beam current if the beam has a large diameter. The volume power density, which is the beam power per unit area divided by the depth of penetration by the beam,

' may also be varied in dependence on the choice of other parameters. If volume power density is so great as to melt the metal in the beam spot, the accelerating voltage may be reduced to compensate.

In the examples given above, the power densities are great enough to melt the metal being treated were it not for relative movement between the beam and the sample. Usually, in

'order to permit use use of high power density beam, and

thereby shorten the treatment time for whatever depth of grain conversion is desired, the beam or sample is moved at speeds ranging from 1,000 to 21,000 inches per minute. Electron pulses having short rise and decay times and higher repetition rates seem to be most effective.

The photomicrograph designated FIG. 3 typifies tungsten that has been surface treated by the new method. This is a vertical section taken through an X-ray tube target that has been treated in the circular focal track region. Note the fine subgrains which have been formed at the top of the picture as compared with the coarse grains of the original base metal underneath. The many fine subgrains and additional boundaries occupy a little more lateral space than the fewer original grains which is believed to cause a compressive stress or negative prestressed condition in the surface when the target is cool. When heated, this prestress is first relieved before shear stress is developed between the surface and substrate grains in which case the net final shear is less and there is less tendency for the surface grains to uplift. The new grain structure survives multiple heating and cooling cycles under normal X-ray tube exposure factors.

The photomicrograph designated FIG. 4 shows just the surface layer of a treated tungsten target with the subgrains magnified 500 times. The large number of subgrains that are formed from the coarser original grains are evident and one may also see that additional subgrains are beginning to form as revealed by discontinuous dark lines within larger grains near the bottom of the photomicrograph. Grain reconstitution at this greater depth could be completed with additional exposures or pulses under the original conditions. FIG. 5 shows the large substrate grains, at the same magnification of 500, which underlie the surface subgrains of the preceding figure.

FIG. 6 is a photomicrograph of the surface layer of a molybdenum target which has been treated by the new method. One

may see the marked increase in the number of subgrains by comparing FIG. 6 with FIG. 7 which is a photomicrograph of a substrate layer in the same target and at the same magnification of 500 times. Additional subgrains may be formed by continuing the treatment as evidenced by the new discontinuous subgrain boundary lines which are just beginning to appear.

Although in the examples given above the ratio of the number of apparent grains in the surface layer to the number in the substrate ranges between about 6 or 10 to l it should be recognized that if the new surface treatment is only carried on until a 2 to 1 ratio is achieved, the benefits of the treatment will be realized to some degree. Moreover, the method can be continued until the ratio is much greater such as 25 or more to l, to obtain even greater benefits in most cases. There are, of course, economic disadvantages to going beyond what is necessary to mitigate brittle fracture to the desired degree in a particular case.

Studies using electron diffraction techniques show that very low angles exist between the crystal lattice planes of adjacent grains separated by grain boundaries in the new fine grain structure. FIGS. 4 and 6 indicate that the small grains are formed from larger original grains. Polygonization is the metallurigical term used for the production of subgrains of this nature. As is known in the art, sometimes when a cold-worked metal is heated under suitable conditions, a subgrain structure of this type forms. It is generally characterized by many regions of perfect lattice which are slightly misoriented and at low angles with each other. Polygonization has also been observed during creep testing, which is a low stress method of testing metals at high temperature. Insofar as is known, polygonization has not been produced with thermal pulses heretofore.

The terms apparent grain size and subgrain used herein may require further definition in relation to the present invention. Photomicrographs of cross sections of metal objects which have been surface treated according to the new method exhibit what appear to be small grains near the surface and large grains underlying the surface. One cannot be certain whether a cluster of small grains is formed from a single original large grain or from parts of one or more grains so the term subgrain is adopted to cover both situations. Then, too, there is a question as to whether one is observing small grains or subgrain sizes at the surface, so the term apparent grain size is adopted for both situations.

Although polygonization is believed to be the mechanism in producing this unique fine-grained structure, the present state of knowledge indicates that renucleation of grains under these extreme conditions of temperature and stress should not be discarded as an alternate mechanism.

The new surface treatment also reduces the impurities in the surface layer. Intense thermal energy in the beam spot liberates more of the gaseous impurities such as hydrogen and oxygen which may cause poor ductility in metals. Carbon and oxygen, both especially harmful to ductility, are removed by diffusion to the surface and are desorbed. Other impurities are more effectively diffused to the subgrain boundaries by the high thermal energy and they are distributed over a greater volume which results in higher purity grain boundaries, more toughness, greater ductility and a higher melting point in the surface layer.

X-ray tube targets which were treated in accordance with the invention have been compared under actual use conditions in X-ray tubes with untreated targets from the same lot. The tube factors, that is, the current, voltage and exposure times during the tests were high such as to simulate the abusive destructive conditions which sometimes occur in practice. Some of the untreated targets exhibited surface roughening and marked radiation fall-off after few exposures and others showed significant radiation fall-off as the number of exposures increased. Treated targets maintained their X-ray output at 97 percent of original value after 10,000 exposures with tube factors set at KVP, 800 ma., 0.012 second exposure intervals while the targets were rotating at about 3,600 r.p.m.

Untreated targets generally did not survive 4,000 exposures and even the best of them had decreased X-ray output to less than 85 percent of original value at that time. Thus, X-ray tube targets treated in accordance with this invention have markedly superior performance to those not so treated.

We claim:

1. A process for treating a refractory metal surface to make it more ductile and thereby enhance its resistance to brittle fracture, comprising:

a. intermittently applying a series of intense heat-producing energy pulses to the metal surface to produce high temperature gradients in the surface layer coincident with each pulse after the metal surface region has been heated to a temperature at least above its nil ductility temperature while simultaneously restricting the energy of each pulse and the time during which the pulses are applied, so that during each pulse the surface temperature reached is near but below that at which any melting would occur and so that a temperature at which recrystallization and grain growth would occur is not reached, and

b. continuing said intermittent series of pulses until said surface layer has a substantially continuous polygonized subgrain structure which is finer than the grain structure of the metal below the surface layer.

2. A process for treating a refractory metal surface to make it more ductile and thereby enhance its resistance to brittle fracture, comprising:

a. applying a series ofintense electron beam pulses at a rate of at least two per minute to the metal surface to produce a high transient temperature gradient coincident with each pulse in a layer bounded by the surface after the metal surface region has been heated to a temperature at least above its nil ductility temperature and controlling the pulse duration and electron current so that the pulse energy is near but below that which would raise the surface temperature enough to cause melting and is below that temperature at which grain growth would occur, and

b. maintaining the surface region of the object at a temperature at least above its nil ductility temperature while the pulsed electron beam is being applied, and

c. continuing said intermittent series of pulses until said surface layer has a substantially continuous polygonized subgrain structure which is finer than the grain structure of the metal below the surface layer.

3. A process for treating a refractory metal surface of a metal object to make the surface more ductile and to thereby enhance resistance of the surface to brittle fracture, comprismg:

a. establishing the object surface at a temperature at least above its nil ductility temperature,

b. applying an intense electron beam repetitively to a surface of the object to heat the surface transiently at least 350 times at a high rate and simultaneously restricting the time of each beam application below the time during which substantial grain growth would occur while maintaining the object surface at a temperature near but below its melting temperature and at least above its nil ductility temperature, thereby creating repetitive high temperature gradients and resultant stresses in the surface layer which reconstitute the grain structure thereof, and

c. continuing said repetitive application of said electron beam until said surface layer has a substantially continuous polygonized subgrain structure which is finer than the grain structure of the metal below the surface layer.

4. A process for treating a surface of a refractory metal object to make the surface more ductile and thereby enhance resistance of at least a part of the surface to brittle fracture, comprising:

a. directing onto the surface of an object an electron beam which has sufficient energy to slowly elevate at least the surface region which is to be treated to a temperature at least above its nil ductility temperature and then b. increasing the energy of the electron beam and concentrating it on a small area increment of the surface while maintaining relative motion between the beam and the surface, the increased beam energy being sufficient to both maintain the region being treated at a temperature at least above its nil ductility temperature and to heat the surface during intermittent short intervals at a high rate while the time-average temperature of the metal is maintained below a temperature at which grain growth may occur, and below but near the melting temperature of the surface, thereby creating repetitive high temperature gradients and resulting stresses in the surface layer which reconstitute the grain structure thereof, and

c. continuing the application of said electron beam until said surface layer has a substantially continuous polygonized subgrain structure which is finer than the grain structure of the metal below the surface.

5. The process set forth in claim 3 wherein relative motion is obtained between the electron beam and the surface of the object by the steps of:

between the object and the electron beam is obtained by the steps of:

a. directing the beam in a fixed path whereby it impinges on the surface of the object, and

b. rotating the object.

7. The process set forth in claim 3 including the steps of:

a. moving the object repetitively through the path of the electron beam, and

b. pulsing the electron beam on and off sequentially with the on time, the off time, the energy density of the beam and the rate of object movement being so adjusted and related that there is development of optimum heat in the surface for creating the requisite high thermal gradients and stresses.

8. The process set forth in claim 3 including the steps of:

a. moving the object along a line extending through the beam path, and

b. sweeping the electron beam at a relatively higher rate of movement along lines that substantially parallel each other on the surface and are substantially normal to the line of movement of the object with the sweeping velocity of the beam, the energy density of the beam and the velocity of the object being so adjusted and related that there is development of optimum heat in the volume increments for creating the requisite high thermal gradients and stresses.

9. A process for treating a refractory metal target surface which is intended for production of X-rays in an X-ray tube for increased ductility in the surface so as to increase its resistance to brittle fracture, comprising the steps of:

a. disposing the target in an evacuable enclosure and in the path of an electron beam that may be generated in the enclosure and directed onto the target surface,

rotating the target,

c. gradually preheating the target to a temperature at least above its nil ductility temperature by impingement thereon of an electron beam which is adjusted to a suitably low energy level,

d. adjusting the accelerating voltage and current of the electron beam to a higher combination of values that results in introducing enough heat in the surface to melt the metal but for the target being rotated, thereby producing repetitive high temperature gradients and stresses in the surface which reconstitutes 'the grain structure in a surface layer of the target, and

e. continuing the application of said electron beam to said target surface until its surface layer has a substantially continuous polygonized subgrain structure which is finer than the grain structure of the metal below the surface.

Claims

2. A process for treating a refractory metal surface to make it more ductile and thereby enhance its resistance to brittle fracture, comprising: a. applying a series of intense electron beam pulses at a rate of at least two per minute to the metal surface to produce a high transient temperature gradient coincident with each pulse in a layer bounded by the surface after the metal surface region has been heated to a temperature at least above its nil ductility temperature and controlling the pulse duration and electron current so that the pulse energy is near but below that which would raise the surface temperature enough to cause melting and is below that temperature at which grain growth would occur, and b. maintaining the surface region of the object at a temperature at least above its nil ductility temperature while the pulsed electron beam is being applied, and c. continuing said intermittent series of pulses until said surface layer has a substantially continuous polygonized subgrain structure which is finer than the grain structure of the metal below the surface layer.
3. A process for treating a refractory metal surface of a metal object to make the surface more ductile and to thereby enhance resistance of the surface to brittle fracture, comprising: a. establishing the object surface at a temperature at least above its nil ductility temperature, b. applying an intense electron beam repetitively to a surface of the object to heat the surface transiently at least 350 times at a high rate and simultaneously restricting the time of each beam application below the time during which substantial grain growth would occur while maintaining the object surface at a temperature near but below its melting temperature and at least above its nil ductility temperature, thereby creating repetitive high temperature gradients and resultant stresses in the surface layer which reconstitute the grain structure thereof, and c. continuing said repetitive application of said electron beam until said surface layer has a substantially continuous polygonized subgrain structure which is finer than the grain structure of the metal below the surface layer.
4. A process for treating a surface of a refractory metal object to make the surface more ductile and thereby enhance resistance of at least a part of the surface to brittle fracture, comprising: a. directing onto the surface of an object an electron beam which has sufficient energy to slowly elevate at least the surface region which is to be treated to a temperature at least above its nil ductility temperature and then b. increasing the energy of the electron beam and concentrating it on a small area increment of the surface while maintaining relative motion between the beam and the surface, the increased beam energy being sufficient to both maintain the region being treated at a temperature at least above its nil ductility temperature and to heat the surface during intermittent short intervals at a high rate while the time-average temperature of the metal is maintained below a temperature at which grain growth may occur, and below but near the melting temperature of the surface, thereby creating repetitive high temperature gradients and resulting stresses in the surface layer which reconstitute the grain structure thereof, and c. continuing the application of said electron beam until said surface layer has a substantially continuous polygonized subgrain structure which is finer than the grain structure of the metal below the surface.
5. The process set forth in claim 3 wherein relative motion is obtained between the electron beam and the surface of the object by the steps of: a. moving the object in a direction extending through the beam path, and b. sweeping the beam at a relatively higher rate of movement along a straight line that is substantially normal to the line of movement of the object.
6. The process set forth in claim 3 wherein relative motion between the object and the electron beam is obtained by the steps of: a. directing the beam in a fixed path whereby it impinges on the surface of the object, and b. rotating the object.
7. The process set forth in claim 3 including the steps of: a. moving the object repetitively through the path of the electron beam, and b. pulsing the electron beam on and off sequentially with the on time, the off time, the energy density of the beam and the rate of object movement being so adjusted and related that there is development of optimum heat in the surface for creating the requisite high thermal gradients and stresses.
8. The process set forth in claim 3 including the steps of: a. moving the object along a line extending through the beam path, and b. sweeping the electron beam at a relatively higher rate of movement along lines that substantially parallel each other on the surface and are substantially normal to the line of movement of the object with the sweeping velocity of the beam, the energy density of the beam and the velocity of the object being so adjusted and related that there is development of optimum heat in the volume increments for creating the requisite high thermal gradients and stresses.
9. A process for treating a refractory metal target surface which is intended for production of X-rays in an X-ray tube for increased ductility in the surface so as to increase its resistance to brittle fracture, comprising the steps of: a. disposing the target in an evacuable enclosure and in the path of an electron beam that may be generated in the enclosure and directed onto the target surface, b. rotating the target, c. gradually preheating the target to a temperature at least above its nil ductility temperature by impingement thereon of an electron beam which is adjusted to a suitably low energy level, d. adjusting the accelerating voltage and current of the electron beam to a higher combination of values that results in introducIng enough heat in the surface to melt the metal but for the target being rotated, thereby producing repetitive high temperature gradients and stresses in the surface which reconstitutes the grain structure in a surface layer of the target, and e. continuing the application of said electron beam to said target surface until its surface layer has a substantially continuous polygonized subgrain structure which is finer than the grain structure of the metal below the surface.