Powder Components Heat Treatment PDF

13
CONTENTS
13.1 13.2
Heat Treatment of Powder Metallurgy Steel Components

Joseph W. Newkirk and Sanjay N. Thakur
Introduction ........................................................................................................... 741 Overview of P=M Processing .................................................................................. 742 13.2.1 Press and Sintering................................................................................... 743 13.2.2 Metal Injection Molding .......................................................................... 748 13.2.3 Powder Forging ....................................................................................... 749 13.3 Designation System for P=M Steels ....................................................................... 750 13.4 Overview of Heat Treatment.................................................................................. 751 13.5 Effect of Porosity on the Heat Treatment of P=M Steels ...................................... 754 13.6 Effect of Alloy Content on P=M Hardenability..................................................... 758 13.6.1 Copper Content ....................................................................................... 760 13.6.2 Nickel Content ......................................................................................... 760 13.6.3 NickelCopper Content ........................................................................... 760 13.6.4 Molybdenum Content.............................................................................. 762 13.7 Effect of Starting Material on Homogenization .................................................... 764 13.8 Quench and Tempering .......................................................................................... 769 13.9 Sinter Hardening .................................................................................................... 772 13.10 Warm Compaction................................................................................................. 774 13.11 Powder Forging...................................................................................................... 775 13.12 Case Hardening ...................................................................................................... 778 13.12.1 Carburizing .............................................................................................. 778 13.12.2 Carbonitriding ......................................................................................... 780 13.12.3 Induction Hardening ............................................................................... 781 13.12.4 Nitrocarburizing ...................................................................................... 781 13.12.5 Nitriding .................................................................................................. 783 13.12.6 Steam Treating ........................................................................................ 784 13.12.7 Black Oxiding .......................................................................................... 785 References .......................................................................................................................... 786
13.1 INTRODUCTION
While the heat treatment of powder metallurgy (P=M) parts can be exactly the same as other parts produced by other methods, majority of time, there are significant differences that must be taken into account. Porosity, composition, and homogeneity are usually the causes for these differences. Understanding the process of P=M is necessary to allow the heat-treatment professional to anticipate these differences and adjust for them. In addition, many P=M compositions exist that do not have wrought analogs, and hence can only be properly treated
2006 by Taylor & Francis Group, LLC.
with a knowledge of the specific system. Finally, there are also differences in the measurement of some mechanical properties and the manner of proof tests. P=M is increasingly used to produce near-net-shaped parts inexpensively. In 2000, almost 80% of all powders produced were of iron or steel [1]. Iron and steel powder shipments increased 55% from 1993 to 2000. This amounted to 445,000 short tons; 93.5% is made into P=M parts. While shipments are relatively unchanged in the last few years due to the economic conditions, this growth is expected to continue over the long run as more and more P=M parts are specified in automotives, appliances, lawn and garden equipments, recreational products, power tools and hardware, business equipments, and electronics. The typical U.S. passenger car contains more than 41 lb of P=M parts, a figure that will go higher within the next several years. More than 500 million P=M hot forged connecting rods have been made for cars produced in the United States, Europe, and Japan. Commercial aircraft engines contain 1,5004,400 lb of P=M superalloy extruded forgings per engine. It is likely that the heat-treatment supplier will continue to see more P=M parts in the future. P=M itself is not a single process but an umbrella of processes that includes, but is not limited to press and sinter, metal injection molding (MIM), powder forging, hot isostatic pressing, powder rolling, spray forming, and more. The common theme among these processes is the use of metal powders as the feedstock for the manufacture of parts. The most likely processes to be encountered in the heat treatment of manufactured steel components are press and sinter, MIM, and powder forging. As an example of just how different a P=M part can be to apply a required heat treatment, let us consider a case where a part is received that is made from wrought 4620 steel. A check of the standards for 4620 shows the composition limits for this material, including the carbon, which varies from 0.17 to 0.22%. For many heat treatments, this along with part dimensions may be the only information necessary to determine the proper parameters. Future experiences with 4620 will usually require the same conditions. Combining different parts of 4620 into a single furnace load can often be done without any adjustments. A contrast to P=M parts made of a close equivalent is FL-4605. Not only has the nomenclature changed with 05 indicating a nominal 0.5% C but many other things as well. A check of the standards indicates that the FL-4605 may have anywhere from 0.4 to 0.7% C. This means that the part can have over 40% greater or lesser carbon content than nominal. While the Ni contents have similar ranges, the Mo can range from 0.4 to 1.10% as compared to 0.2 to 0.3% for wrought. Adding to this is the possibility that the original powders used for creating the P=M part may have used admixed Ni powder, which may lead to inhomogeneities in composition and altered hardening response. Finally, the physical properties of the part are dependent on density and will change depending on the requirements for that part. Many times, the lower densities of P=M parts can be such that the surface has connected pores, which absorb quenchant or lead to larger than expected case depths. All of these things means that different batches of P=M parts can behave very differently and special furnace loads may be required to meet the unique parameters of the parts. No wonder some commercial heat-treatment shops shudder when asked to heat-treat P=M parts. Despite all of these problems, P=M parts can be properly heat-treated, given an understanding of the differences in P=M steel and prior knowledge of the parts fabrication details.
13.2 OVERVIEW OF P=M PROCESSING

Different P=M processing methods will cause different effects during heat treatment. The high porosity of press and sinter parts has a large effect on the heat-treatment response. Metal injection molded parts have less porosity than sand-cast parts, but a very fine grain size.
Powder-forged parts have a greater homogeneity than many wrought parts. In order to properly heat-treat P=M parts, greater knowledge of the specifics of the fabrication of the part is needed than for traditional steel parts. Therefore, an overview of the different methods is helpful when discussing the differences in various P=M parts. A side-by-side comparison of these methods is given in Table 13.1.
13.2.1 PRESS
AND
SINTERING
Press and sintering is the P=M process that is most commonly encountered by heat treaters. Parts that have two-dimensional complexity, but have a limited number of flat surfaces in the third dimension, are especially suitable to be produced by this method. A power transmission gear is an ideal part to be made by press and sinter with its complicated outline, but simple shaped along the axis of the gear. The most important attribute of a press and sinter part is its dimensions. If the dimensions of each part can be controlled, then press and sintering is an extremely cost-effective way to manufacture metallic parts. The general steps in the press and sinter process are shown in Figure 13.1. Compaction is probably the most critical step in the P=M process since it sets both the density of the powder and the uniformity of that density throughout the product. It is also where almost all of the shaping of the part occurs. A typical compaction sequence using a mechanical press would start with the bottom punch in its fully raised position with the powder feed in position over the die. The bottom punch then descends to a preset fill depth and the powder-feed retracts, leveling the powder. Next, the upper punch descends into the die and compacts the powder. The upper punch then retracts, and the bottom punch rises to eject the powder compact. As the powder feed advances to the next cycle, its forward edge clears the compacted part from the press, and the cycle repeats. The objective of compaction is to produce a high-density compact with sufficient green strength for in-process handling and transport to the sintering furnace. Since final properties strongly depend on density, uniform properties require uniform density. Most compaction is performed with mechanical or hydraulic presses and rigid tooling. Compacting pressures range from 3 to 120 t=in.2, with 10 to 50 t=in.2 as the most common. Because of press capacity limitations, most P=M products have pressing areas less than 10 in.2.
TABLE 13.1 Comparison of Powder Metallurgy Techniques Discussed

Conventional Press and Sinter Intermediate (<5 lb) Good Excellent >5,000 Excellent (+0.001 in.=in.) Fair 8090% of wrought Low $0.505.00=lb Metal Injection Molding Smallest (<1=4 lb) Excellent Good >5,000 Good (+0.003 in.=in.) Very good 9095% of wrought Intermediate $1.0010.00=lb Hot Isostatic Pressing Largest (11,000 lb) Very good Poor 11,000 Poor (+0.020 in.=in.) Excellent Greater than wrought High >$100.00=lb
Characteristic Size of workpiece Shape complexity Production rate Production quantity Dimensional precision Density Mechanical properties Cost
P=M Forging Intermediate (<5 lb) Good Excellent >10,000 Very good (+0.0015 in.=in.) Excellent Equal to wrought Somewhat low $1.005.00=lb
Elemental or alloy metal powders
Additives (lubricants or binders)
Blending
Elemental or alloy metal
Sintering
Optional secondary manufacturing
Optional secondary finishing
Finished P/M product

FIGURE 13.1 Simplified flowchart of the press and sinter process.
Another important consideration in the compaction of P=M parts is the ability of the green part to survive the ejection process and subsequent handling prior to sintering. Since the greatest stress is usually seen during ejection from the die, and die wall friction also leads to greater bending moments on the part, surviving ejection typically means the part will survive until sintering. The green strength of the part is critical to this step and impacts design considerations for compaction, even at the expense of final properties. Ejection forces are reduced by adding a dry lubricant to the powder mix. This also aids in compaction, but reduces green strength to a small extent. The amount of admixed lubricant is balanced to provide the best combination of attributes. The properties of P=M parts are strongly dependent upon the density of the part, and part density in turn depends upon the amount of pressure that is applied and the characteristics of the powder (including size, shape, surface texture, and mechanical properties). When the powder particles are pressurized, density can increase for a variety of reasons, including the bulk movement and rearranging of particles, the deformation of individual particles, and particle fracture or fragmentation. The characteristics of the powdered material controls which of these mechanisms dominates. For examples, Elemental metal powders are usually highly deformable and a great deal of deformation can occur during compaction. Ceramic powders are usually brittle and will fracture. Metal powders that are cold-worked or highly alloyed will be between the two extremes. Densities are usually reported as a pure value (such as grams per cubic centimeter) or as a percentage of the theoretical value, where the difference between this number and 100% is the amount of void space left within the compact. Adjectives before the word density define the conditions of the part when the density was determined. Apparent density refers to the density of the loose, uncompacted powder. Green density provides the value after compaction, but before sintering. Final density is determined after sintering.
Since uniform, high density is a common objective of compaction, various means have been developed to assist the densification. As mentioned above, lubricants are frequently added to reduce die wall friction and promote transmittance of the compacting pressure. Unfortunately, lubricants also reduce the green strength, and too much lubricant can actually reduce mechanical properties because of the voids produced when the lubricant is removed from the compact. Another technique that is having an expanded use, particularly in automotive applications, is warm compaction. The powder is preheated prior to pressing, i.e., softening the metal for better response to the applied pressure. Heated dies are typically used to prevent cooling of the powder during application of the pressure. The advantages of warm compaction are higher green and sintered densities, higher green and final mechanical properties, and greater uniformity of density in the sintered part. With good mechanical compaction practice, the density of loose powder can be raised to about 80% of an equivalent cast or forged material. At this stage, the parts generally have sufficient strength (green strength) to hold the shape and permit a reasonable amount of handling. In addition, the compaction operation determines both the nature and distribution of the remaining porosity. In the sintering operation, the pressed compacts are placed in a controlled atmosphere environment and heated. A burn off or purge is first conducted to combust any air, volatilize and remove lubricants or binders, and raise the temperature of the compacts in a controlled manner. The temperature is then increased to a level where solid-state diffusion occurs (typically 70 to 80% of the melting point for metals and near 90% for refractory materials), and sufficient time is provided to form bridging bonds between the particles. The mechanical bonds of compaction become metallurgical bonds. With additional time, diffusion acts to decrease the size of the remaining pores, with a parallel improvement in properties. However, pore reduction is strongly affected by the driving force for densification, the reduction in surface area, and the ability to transport atoms to pores. As densification reduces the driving force, the process slows down. In addition, at densities exceeding 90%, grain growth leads to the separation of grain boundaries from the pores and the transport of atoms slows significantly. The closer to full density, the harder it is to fully densify by sintering alone. So, the additional cost of maintaining the temperature and protective atmosphere is weighed against the ever-increasing level of properties, and an acceptable compromise is determined. The protective atmosphere is then maintained while the product is cooled to room temperature. Atmosphere control is critical to successful sintering. Compacted powder typically contains between 10 and 25% residual porosity, and some of these voids are interconnected to exposed surfaces. If the material were heated in air, rapid oxidation would occur, and this would severely impair the bond formation and resultant properties. Reducing atmospheres break down any oxide that may be present on particle surfaces, and combust any harmful gases that may enter the furnace or be generated during the sintering. Inert gases cannot perform these tasks, but serve to prevent the formation of any additional contaminants. Vacuum sintering is often performed and nitrogen atmospheres are also common. Ferrous alloys are often produced by mixing graphite with iron powder to form steel or stainless steel. While the high mobility of carbon allows for good homogeneity of the interstitial element, this mobility also means that carbon is free to interact with the atmosphere. This situation is similar to that found in heat-treating wrought or cast steel parts and has similar answers. The control of the carbon potential of the atmosphere may be necessary to maintain correct carbon levels. A typical microstructure of a press and sintered plain carbon steel is shown in Figure 13.2. During the sintering operation, metallurgical bonds form between the powder particles. With the increase in density, strength, ductility, toughness, and electrical and thermal
FIGURE 13.2 F-0008, compacted 50 tsi and sintered 9808C in Argon. Etched with 2=5 Nital.
conductivities also increase. If different materials were blended together, interdiffusion may promote the formation of alloys or intermetallic phases. As a consequence of the density increase, size decreases. To meet the desired tolerances, the compacted shape must be appropriately oversize. Not all of the porosity is removed, however, final-pressed and -sintered P=M products generally retain between 5 and 25% residual porosity. A typical pore structure from a press and sintered part is shown in Figure 13.3. Higher densities along with greater homogenization can also be achieved by the use of high-temperature sintering. High-temperature sintering will result in improved pore rounding and uniformity in alloying, which allows better heat-treating response for the same alloy content. Most of the oxides are reduced during the conventional sintering at 11508C (21008F) but oxides of certain elements like manganese and chromium are reduced at higher temperatures in a reducing atmosphere. These oxides not only deteriorate the mechanical properties but also interfere with the heat-treatment response of the alloy. High-temperature sintering not only results in better mechanical properties, but also the benefits, which extend even in the heat-treatment stage of P=M part. Sintering of P=M steel compacts is usually performed in mesh belt furnaces with a maximum operating temperature of about 11508C (21008F). Furnaces are now readily available, which overcome the temperature limitations of a mesh belt. Pusher furnaces are widely used when the sintering temperature is greater than 21008F. One of the issues while using higher temperatures 1150 to 13158C (2100 to 24008F) is the increased cost of maintenance. An alternative route for achieving higher densities with press and sintered parts is with double pressing and sintering (DPDS). The difficulty here is that normally, after the sintering operation, die P=M steel is too hard to compress, primarily because of the dissolved carbon. By lowering the initial sintering temperature until the solution of graphite is minimized, the presintered compacts can be more easily densified by repressing. After a second sintering treatment at or above the normal temperature, an alloyed steel can be produced with densities up to 95% of the theoretical value. In a conventional press and sinter operation, the compaction operation sets dimensions, but these dimensions then change as the part shrinks during the densification of sintering. In
FIGURE 13.3 737SH sinter-hardened. Unetched.
addition, warping or distortion may occur as a result of nonuniform heating and cooling. Repressing, coining, or sizing may be employed to restore or improve dimensional precision. The accompanying cold working and further increase in density can combine to provide a 25 to 50% increase in product strength. Conventional finishing operations, such as heat treatment, machining, and surface treatment, can also be performed. High-density parts, those with less than 10% porosity, can be treated as conventional parts. Parts with lower density usually require some form of special precautions, such as protective atmospheres, special coolants or quenchants, or adjusted processing conditions. This is due to the fact that the porosity will most likely be connected throughout the part when the porosity exceeds 10%, allowing the atmosphere to penetrate to the internal surfaces of the part. P=M steel parts can be quenched and tempered to achieve high strength or hardness as required. However, the hardenability of steels depends in part on the density of the parts since the thermal conductivity of the metals is affected by the density. In addition, the homogeneity of the parts, prepared from mixed powders, will be affected by sintering conditions and can result in changes in hardenability due to composition variations. In addition to these problems, hardness measurements on a P=M sample will result in a lower hardness value due to residual porosity. If a certain heat-treated hardness is expected and used to measure the formation of martensite, the resulting lower measured hardness, called the effective hardness, due to porosity will make the steel appear not to be fully hardened. Microhardness tests, which do
not interact with the pores in the material, do give a true reading of the state of the part. These factors must be taken into account when specifying the heat treatment of the sintered steel parts. Sinter hardening is a process where normal sintering provides the austenization treatment of steel compacts. Then quenching must be done in the furnace before removal of the part and the subsequent tempering. This process provides a significant economic benefit by the elimination of reheating to the austenite temperature. However, materials with very good hardenability are required to allow for sufficient depth of martensite formation at relatively slow cooling rates possible in the sintering furnace. The presence of porosity and permeability also opens up the possibility of unique secondary processing, such as impregnation and infiltration. During impregnation, oil or other liquid, such as polymeric resin, is introduced into the porous network. The most common application is the oil-impregnated bearing, which contains between 10 and 40% oil by volume. Parts impregnated with Teflon offer the combined properties of high strength and low friction. If the pores are filled with a solid material, such as a lower melting point metal, the process is called infiltration. Strength and toughness improve due to the absence of pores. Components can be made gas- or liquid-tight, machinability and corrosion resistance improve, and smooth surface platings are now possible.
13.2.2 METAL INJECTION MOLDING

MIM is similar to conventional P=M process in several ways and different in other important areas. This leads to new considerations for the design of MIM parts when compared to conventional P=M. It also offers many new advantages that can be utilized to produce costeffective, high-performance parts. Like conventional P=M, MIM uses powdered metals as the starting material, and properties are controlled by the density produced during sintering. Every other aspect of the process is different. An overview of the process is shown in Figure 13.4. The metal powder is mixed with a specially designed plastic binder, often with thermoplastic properties. This mixture is often produced in the form of pellets, which makes for ease of handling and uniformity of composition. The mixture is called the feedstock and can contain up to 50 vol% binder. The feedstock is heated to the plastic range and injected into a mold under pressure. The mold temperature is such that the part becomes harder and can then be ejected from the mold with excellent green strength. The part is now treated to remove the plastic binder. Techniques such as solvent extraction, controlled heating to volatilize the binder, or catalytic debinding using acids are used to remove the binder. The resulting part is very fragile and must be treated carefully before sintering. The part is sintered similarly to conventional P=M parts, except that the lower density of the part before sintering leads to large amounts of shrinkage, often up to 25% or more. In conventional P=M, the powder is often chosen to flow well during filling and compaction, as well as to result in low shrinkages so that dimensions are closely held. Since in MIM the powder is carried by the plastic binder, flow characteristics are not the major attributes of the powder chosen, and instead the powders are often chosen on the basis of sinterability instead. The powders are very fine, usually in the range of 2 to 20 mm in diameter. The fine powder sinters to high densities. Densities of 99% of the theoretical value are common. This has a significant effect on the properties of MIM parts usually making them superior to other fabrication techniques. While the size of conventional P=M products is generally limited by press capacity, the size of MIM is more limited by two factors: the cost of the fine powders and the ability to completely remove the binder from all areas of the part. Typically, MIM parts are complex-shaped metal parts with thicknesses as large as 6.3 mm (1=4 in.) and weights under 60 g (2 oz). Section thicknesses as small as 0.25 mm (0.010 in.) are possible because of the fineness of the powder. Special binders are now available, which allows much larger parts to be successfully debound [2].
Fine metal powder
Binder
Mixing and pelletizing
Injection molding
Binder removal
Sintering
Optional finishing operations
Finished product
FIGURE 13.4 Simplified flowchart of the metal injection molding process.
The heat treatment of MIM parts is much like that of wrought parts. Their small size means that they heat up fast and cool down quicker. The only difference is that there is usually a smaller grain size, sometimes the composition does not have an exact wrought equivalent, and sometimes admixed additives, such as nickel, are used. The heat treatment of MIM will not be covered separately as there are no special treatments involved.
13.2.3 POWDER FORGING

Forging can be used to form complex shapes from canned powder or simple-shaped sintered preforms. One approach for making a complex-shaped product that would be difficult to compact is to first produce a simple-shaped preform by conventional press and sinter manufacture. Subsequent hot forging converts this material to the desired final shape, while
adding the benefits of both metal flow and further densification as a result of the shear under compression that occurs during forging. By starting with P=M, forged products that can be produced and are free of segregation, have uniform fine grain size throughout, and utilize the novel alloys and unique combinations that are possible through the blending of powders. Because of the initial porosity and permeability, protective atmospheres may be required during heating and forging operation. Powder-forged properties can be exceptionally good.
13.3 DESIGNATION SYSTEM FOR P=M STEELS

P=M steels are designated according to a system adopted by the Metal Powder Industry Federation (MPIF) [3]. This system uses a combination of letters indicating alloying elements and numbers, which give the nominal compositions. In addition, another set of numbers is attached, which indicates a minimum strength value to which the part was made. When the powder is prealloyed, there is a modified system that is used. This occurs for a number of alloyed steels or stainless steels. Table 13.2 gives the entire list of letter codes used in designation system for P=M steels. The first letter in the code gives the base metal for the alloys, so C in the first position is for copper-based alloys such as brass and N would be for nickel-based alloys, etc. Since the topic is P=M steels, then the operative letter is F for iron. If the P=M steel contains any alloying elements of significance besides carbon, they are then added after the iron designation. The fact that an alloying element is not represented in the designation does not mean that it is not included in the alloy, so the standard composition needs to be checked. Still this system does identify the more important elements, making it easier to keep track off. The letter code is usually followed by four digits. Unfortunately, there are differences in the significance of these four digits. For nonferrous alloys, the first two digits are the percentage by weight of the major alloying element and the second two digits are the percentage of the minor or second alloying element. For ferrous alloys, the first two digits have the same meaning, but the second two digits represent the carbon content in tens of percent. Typical American Iron and Steel Institute (AISI) designations for wrought steel would use 80 to represent 0.8% C, while the Metal Powder Industries Federation (MPIF) designation would use 08 for the same amount of carbon. The last digits following the four alloying digits is the minimum tensile strength of the material. Once again, there are two different significances to these numbers depending on
TABLE 13.2 Letter Code for Designating P=M Alloys

A C CT CNZ CZ F FC FD FL FN FX Aluminum Copper Bronze Nickel silver Brass Iron Ironcopper or copper steel Diffusion-alloyed steel Prealloyed ferrous (not SS) Ironnickel or nickel steel Copper-infiltrated iron or steel G M N P S SS T U Y Z Free graphite Manganese Nickel Lead Silicon Stainless steel (prealloyed) Tin Sulfur Phosphorous Zinc
whether the steel was heat-treated or not. Parts that have not been heat-treated will always give the minimum yield strength here. Heat-treated parts are given with the ultimate tensile strength. The reason for this is that heat-treated (hardened) P=M steels usually do not have sufficient ductility to measure with a standard tensile test, resulting in the yield strength and the ultimate tensile strength, which are essentially the same. Exceptions to each of these sets of digits occur. One set of exceptions is for soft magnetic materials and steels with phosphorus additions. These materials are not usually subjected to heat treatments other than annealing and are not included here. The reader should consult the MPIF standards for the codes for these materials. Another exception is for the use of prealloyed powders. Powder blends, which have a prealloyed steel powder as one of the components, have a designation that begins with FL. If the steel powder is mixed with graphite only the designation is usually modified from the AISI alloy codes. FL-4605 is a 4600 steel powder or near equivalent with 0.5% C nominally. This designation may or may not be followed by more digits giving the strength, as discussed above, and the letters HT, if it is heat-treated. Elemental alloying additions to the prealloyed powder other than graphite are designated with one or two more letters following the FL. If additional Ni is blended with the FL-4605, it becomes FLN-4605. There may also be Ni and Cu additives for alloys designated as FLNC4405. When Ni is added to a steel normally not containing Ni, the amount is sometimes given directly after the letter code, such as FLN2-4408. Diffusion-alloyed powders lead to designations of FD, followed by the Ni amount and graphite, such as FD-0208, which contains nominally 2% Ni. A check of the compositions for this material reminds us that there is also 1.5% Cu. Finally, copper-infiltrated steels will have the FX designation. The first digit is the approximate amount of copper, which has been infiltrated into the open porosity, not the copper added to the powder blend or prealloyed powder. Thus FX-1005 is approximately 10% copper-infiltrated and 0.5% graphite. Table 13.3 lists many commonly used P=M steels and their composition ranges. Depending on the starting powders used, there may also be up to 0.5% Mn. Other possible tramp elements include P, S, and O. The O2 content of the powder is typically reduced during sintering in a reducing atmosphere and can vary depending on processing.
13.4 OVERVIEW OF HEAT TREATMENT

There are a number of considerations when heat-treating P=M parts depending upon the chosen heat treatment. The following suggestions act as a guide to general considerations for heat treatment of P=M parts: 1. Since the cost benefit of P=M often involves avoiding machining, most parts are not machined after heat treatment. This means that any type of surface damage, dent, or scratch created during handling can reduce the functionality of the parts. Therefore, extra care must be exercised in loading parts for heat treatment and any subsequent handling. 2. Parts that have seen some sort of secondary operation (sizing, coining, machining, or rolling) prior to heat treatment probably contain residual oils. These substances, partially entrapped inside the open porosity, may adversely affect the properties that could be achieved on heat-treating and may also reduce the life of some part of the furnace. In this case, a suitable washing operation, with a proper choice of solvent, should be carried out prior to heat treatment. 3. Loading patterns of parts to be heat-treated must permit a free flow of a carburizing gas or quenching fluid. When parts are small, closely spaced parts may have little
TABLE 13.3 Compositions of Common P=M Steels

MPIF Designation F-0000 F-0005 F-0008 FC-0200 FC-0205 FC-0208 FC-0505 FC-0508 FC-0808 FC-1000 FN-0200 FN-0205 FN-0208 FN-0405 FN-0408 FL-4205 FL-4405 FL-4605 FLN-4205 FLN2-4405 FLN4-4405 FLN6-4405 FLNC-4405 FLN2-4408 FLN4-4408 FLN6-4408 FLN-4608 FLC-4608 FLC-4908 FLNC-4408 FD-0200 FD-0205 FD-0208 FD-0405 FD-0408 FX-1000 FX-1005 FX-1008 FX-2000 FX-2005 FX-2008 C 0.00.3 0.30.6 0.60.9 0.00.3 0.30.6 0.60.9 0.30.6 0.60.9 0.60.9 0.00.3 0.00.3 0.30.6 0.60.9 0.30.6 0.60.9 0.40.7 0.40.7 0.40.7 0.40.7 0.40.7 0.40.7 0.40.7 0.40.7 0.60.9 0.60.9 0.60.9 0.60.9 0.60.9 0.60.9 0.60.9 0.00.3 0.30.6 0.60.9 0.30.6 0.60.9 0.00.3 0.30.6 0.60.9 0.00.3 0.30.6 0.60.9 Ni Cu Mo
1.53.9 1.53.9 1.53.9 4.06.0 4.06.0 7.09.0 9.510.5 1.03.0 1.03.0 1.03.0 3.05.5 3.05.5 0.350.55 1.702.00 1.352.50 1.003.00 3.005.00 5.007.00 1.003.00 1.003.00 3.005.00 5.007.00 3.65.0 1.602.00 1.003.00 1.551.95 1.551.95 1.551.95
1.003.00
1.003.00 1.003.00 1.003.00 1.31.7 1.31.7 1.31.7 1.31.7 1.31.7 8.014.9 8.014.9 8.014.9 15.025.0 15.025.0 15.025.0
0.500.85 0.750.95 0.401.10 0.490.85 0.650.95 0.650.95 0.650.95 0.650.95 0.650.95 0.650.95 0.650.95 0.391.10 0.391.10 1.301.70 0.650.95 0.40.6 0.40.6 0.40.6 0.40.6 0.40.6
Up to 0.5% Mn may be present depending on starting powder.
permeability of the gas or fluid. This low permeability may cause uneven flow of active form of both carburizing gas and quenching fluid. A controlled pattern loading, using wide mesh supports and choice of suitable spacers between layers of parts, decrease the unevenness of active fluids. For best results, parts should be single-spaced in layers.
4. Care should be exercised in stacking parts with flat faces in an orderly way so that there are no dead zones, where the carburizing gas and the quenching medium cannot reach, leading to insufficient carbon enrichment and low-quenching speed. The final result is typically a low hardness. 5. As mentioned above, P=M parts do not typically undergo finishing operations after heat treatment, so exposing the parts to air before oil quenching, even for very short times, should be avoided. 6. Due to open porosity in the P=M parts, the amount of quenching oil, which penetrates into the open and interconnected porosity, cannot be neglected. Also, if the part is going to be immersed in oil during normal service, all of the quenching oil should be removed, to avoid any dangerous pollution of the suitable lubricant. A specific solvent extraction operation may become unavoidable after quenching and before stress relieving. More than one washing step is likely to be required to completely remove the oil from the pores. 7. In the case of induction hardening, a preliminary washing operation may be needed. This cleaning operation is aimed at removing residuals of cutting fluids or sizing lubricants. 8. Also in the case of induction hardening, the use of water as a quenchant may lead to rust formation on the interior of the parts, if the density is low enough to still have significant open porosity. A mild stress relieving, carried out immediately after water quenching, can be used to avoid this problem. 9. The heat-treatment conditions for a P=M part, which is replacing a full density part must be altered from those previously used, especially when case hardening. In the case where the chemical compositions are comparable, there does not usually need to be a change in temperature. However, there is almost always a need to use a faster quenching medium to account for the reduced hardenability. 10. Unless the part density exceeds 7.4 g=cm, gaseous or liquid nitriding must be preceded by a suitable pore closure treatment. Steam treatment is an effective treatment that gives consistent results. On the contrary, plasma (or ion) nitriding does not require any pore closure pretreatment. However, if the parts are not adequately cleaned prior to plasma nitriding, the remaining contaminants may deleteriously affect the results. While these general rules do not cover all of the differences in the heat treatment of P=M steels, they do give an introduction to many of them. Many of these rules are covered in this chapter. The two basic differences are the large amount of porosity and the typically inhomogeneous microstructure. Each of these has several effects on the heat treatment of P=M steels. Also, there are often subtle to considerable differences in the starting compositions picked for a given part. The most common problem in heat-treating P=M parts is overloading. Most commercial heat treaters base their pricing on the amount of furnace hours required to process the load and the capacity of their furnace baskets. This tendency is contrary to good practice in the heat treatment of P=M parts. Once a critical loading is reached, adding more parts is going to increase the variability in hardening. The critical load for typical P=M parts is lower than for comparable wrought materials. Heat treatment of P=M parts is most widely done in endothermic gas with additions of methane and air to control the carbon potential. Heat treaters of P=M parts have the liberty to select from a variety of atmospheres. Endothermic gas, nitrogenmethanol blends, or any nitrogen-based atmosphere can be used if the carbon potential of the atmosphere is controlled to prevent oxidation or reduction. During case hardening, additives like methane or propane and ammonia are added to the atmosphere. The amount of additions added depends on the part density and surface
porosity. For example, a low-density part or a part having open-surface morphology will result in greater nonuniformity of case depth. This variation can be minimized by close control of the atmosphere gas composition and case hardening at lower austenizing temperatures.
13.5 EFFECT OF POROSITY ON THE HEAT TREATMENT OF P=M STEELS

As mentioned, several factors change the response of P=M steels to heat treatment when compared to conventional steels, however, porosity probably has the greatest single effect. While it is possible to achieve very high densities in P=M steel materials, the increased difficulty, and hence cost, combined with the decreased dimensional control mean that few P=M steels are made to be fully dense. The majority of parts have porosities between 5 and 15 vol%, far more than is found in a typical casting. Every mechanism in press and sinter is subjected to diminishing returns. Compact density has a decreasing dependence on compaction pressure, as does sintering and homogenization. Porosity can influence the heat-treatment process through its effect on:
. . . . .
Density Thermal conductivity Proof testing (particularly hardness) Permeability (to gases and liquids) Electrical resistivity (in induction hardening)
The first two factors have an effect on the hardenability of P=M steels. The third affects the perception of the results of heat treatment, potentially leading to an unexpectedly low result. These factors are talked about next in greater detail. Maximum hardness in steels is achieved when the available austenite fully transforms into martensite. Sufficiently rapid cooling is required to suppress the eutectoid reaction and subsequently to form martensite. Hardenability for a wrought steel can then be defined as the depth below the surface, where the cooling rate is just adequate to transform the austenite into martensite. Hardenability depends directly on the thermal conductivity of the steel and the composition. Since the hardenability of steel is much more sensitive to changes in composition than it is to the relatively small changes in the thermal conductivity, we generally consider only the composition changes when designing a heat treatment. The effect of porosity on the hardenability of a P=M steel is given in Figure 13.5, where the results of a series of Jominy tests on sintered carbon steel test pieces are compared with those for wrought steel [8]. The P=M specimens were prepared from iron powder, mixed with 0.8% graphite, pressed, and sintered to various densities from 6.0 to 7.1 g=cm3. These corresponded to porosity levels ranging from 9 to 24%. The P=M test specimens and wrought C1080 steel specimen were austenitized at 8708C (16008F) for 30 min in a protective neutral atmosphere. The test specimens were then quenched in a Jominy end quench and the hardness was measured. Figure 13.5 graphically points out two important differences of P=M heat treatment: Not only hardenability is reduced by porosity through its influence on thermal conductivity, but the porosity also reduced the apparent hardness of the P=M materials. This is seen in the lower hardness readings at the first Jominy distance with succeeding lower densities. As evidenced above, the hardenability of P=M steels depends on both the composition and the porosity of part. The thermal conductivity of P=M parts has been shown to be strongly influenced by porosity, which acts as an insulator in slowing heat transfer. One model of the effect of porosity on the thermal conductivity is: l lm (1 2)
90
80 1080
Apparent hardness (HRA)
70 7.1 g/cm3 60 6.8 g/cm3 6.4 g/cm3 50
6.0 g/cm3 40
30 0 1 2 3 4 5 Distance (mm) 6 7 8 9 10
FIGURE 13.5 Hardenability curves for F-0008 compared to wrought 1080 steel. The wrought alloy has
higher manganese content. The plus signs indicate the depth at which at least 50% martensite was observed in the microstructure. (Data from H. Ferguson, Metal Progress, ASM, 107(6), 8183, June 1975; 108(2), 6669, July 1975.)
where l is the thermal conductivity of the P=M part, lm is the theoretical conductivity of the fully dense material, and is the fractional porosity (% porosity=100) of the part. A study using simulation and experimental data resulted in a confirmation of the above equation for porosity levels between 0.1 and 10% [4]. For porosity levels greater than 12% the water-quenched samples showed higher cooling rate than expected. This was attributed to water entering the pores and cooling the samples faster. The flow of heat from the center of a P=M part has been estimated by Bocchini [11]. The heat quantity to be removed on quenching can be given by the following relation: H M (1 )cm T (13:1)
where H is the heat quantity to be removed, M is the mass of a fully dense body having the same shape of the sintered component, cm is the average heat capacity of the constituting metal, and DT is the temperature difference between the part and the quenching medium. One estimate of the relative cooling speed of a porous P=M part to a fully dense part is given by Relative cooling speed lm 1 3 cm 1 2
0.95 Relative cooling speed
0.9
0.85
0.8
0.75 6.6
6.8
7.2 Density (g/cm3)
7.4
7.6
7.8
FIGURE 13.6 The relative cooling speeds of a porous steel part compared to a fully dense as predicted by
two different theoretical treatments. The lower curve ignores fluid flow inside the pores. (From Bocchini., G.F., Advances in Powder Metallurgy and Particulate Materials2001, Vol. 6, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 5686.)
This approximated formula indicates, as expected, that the internal cooling speed of a part, made of porous steel, is lower than the internal cooling speed of a comparable part, made of fully dense steel. The deviation from the cooling speed of an equivalent fully dense steel increases as the porosity increases. The curve showing the decrease of internal cooling speed on quenching versus density, calculated by means of this formula, is plotted in Figure 13.6. The flow of fluids inside pores may influence the heat exchange. A second relation gives the relative cooling speed taking this factor into account: Relative cooling speed lm (1 ) cm
The straight line corresponding to this formula is also plotted in Figure 13.6. The difference between the curves is insignificant above 7.4 g=cm3, but nearly attains 6% at 6.8 g=cm3. Unfortunately, no experimental results are available to eliminate the curve (or the line), which does not correspond to the practical results. Considering the heat release at the part surfaces, two different factors should be taken into account:
.
The surface porosity, which may be thought of as an increase of the roughness, consequently slowing down the speed of the cooling medium, which is removing heat from the part surfaces The suction effect exerted by the hot gas present inside porosity, which, by contracting itself during cooling, tends to stop and suction the vapor bubbles, originating from the quenching liquid
Since the coefficient of external heat exchangesometimes defined as the coefficient of adductionis linked with the state or microgeometry of the surfaces through a constant coefficient, usually between 1 and 2.5, the stage of the whole cooling process corresponding to heat removal from the external surfaces of a porous part may be considered to be slowed down to about half the typical speed of fully dense materials. In other words, the cooling of a PM component during quenching, internally and from the surface, may be visualized as a slowmotion version of the same succession of mechanisms, occurring during quenching of a nonporous body of material. Soak time is as much as 50% longer for P=M steels [5]. Usually the porosity of sintered materials is completely closed when the relative density exceeds 97%. On the contrary, porosity is completely opened, when the relative density is lower than 75 to 80%. Altogether, between porosity levels of 3 and 20%, which is typical of most mechanical components, there is a combination of opened and closed porosity. The relative amount is difficult to measure and depends on the starting material and processing conditions. Additionally the porosity in most compacts is not uniform. This is particularly important on the upper and lower faces, which were in contact with the press punches. On these faces, the porosity is higher at the surface in press and sinter parts [6]. Figure 13.7 illustrates this effect for a F-0000 compact, which had either admixed lubricant or the die wall was lubricated. The overall porosity was approximately 8.5%, but rises to almost 15% within 0.2 mm of the surface. This creates a higher degree of open porosity on these surfaces. While the effect of porosity on the measured hardness, called the apparent hardness, is easy to observe, it is hard to take into account. Hardness deviations from the expected values caused by porosity are generally greater, the lower the density. The relationship is generally exponential, but varies with the pore shape and dimensions, which change with the processing
16 Die wall lubrication Admixed lubricant 14
15
Local porosity (%)
13
12
11
10
8 0 0.1 0.2 0.3 0.4 0.5 0.6 Distance from front face (mm) 0.7 0.8 0.9
FIGURE 13.7 Local porosity variation with distance from the punch face of press and sintered F-0000
compact with eithor die wall or admixed lubricant. (From Bocchini, G.F., Fontanari, V., Molinari, A., International Conference on Powder Metallurgy and Particulate Materials, Metals Powder Industries Federation, Princeton, NJ, 1995.)
85 80 75 70 65 Hardness 60 55 50 45 40 35 30 6.6 6.7 6.8 6.9 7 7.1 7.2 Density (g/cm3) 7.3 7.4 7.5 7.6 HRC range HRA range
FIGURE 13.8 Dependence of hardness on density derived from experimental data. Variations in pore
morphology and microstructure result in large bands. (From Bocchini, G.F., Advances in Powder Metallurgy and Particulate Materials2001, Vol. 6, Metals Powder Industries Federation, Princeton, NJ, 2001. pp. 5686.)
and starting powder. One general trend derived from experimental data is shown in Figure 13.8. The trend is shown as linear, but the resulting dependence is shown as bands, rather than exact numbers. The difference between apparent hardness and actual hardness is clearly seen in Figure 13.9. The figure shows the standard Vickers hardness for loads greater than 1 kg compared to the microhardness, also measured by Vickers using loads less than 100 g. In the lower set of data, the Vickers indenter is producing an indent that covers large regions of the microstructure, essentially averaging the hardness of the material with the hardness of air (the pores). The upper set of data shows that the microhardness is essentially unchanged with density. The difference that can occur from simply changing the load is obvious. A plot of the change in Vickers hardness number with density is shown in Figure 13.10, which emphasizes the difference.
13.6 EFFECT OF ALLOY CONTENT ON P=M HARDENABILITY

The most important function of alloying element in heat-treatable steel is to increase hardenability, whether for quenching and tempering or sinter hardening. This increase in hardenability makes possible the hardening of larger sections and the use of oil rather than water quench to minimize distortion and to avoid quench cracking. It is known that increasing the carbon content of steel raises the tensile strength and hardness levels in proportion to the added carbon. In wrought steel, this ratio is maintained to about 1.2% C. In sintered P=M steels, the maximum tensile strength is reached at the eutectoid composition of 0.8% C. When carbon content is increased above this level, carbide networks begin to form at grain
900
800
<100 g load
700
VHN
600
500 110 kg load 400
300 6.6 6.8 7 Density (g/cm3) 7.2 7.4
FIGURE 13.9 Difference in Vickers hardness of a P=M steel as a function of load and density. Low load corresponds to micro hardness and high load to apparent hardness. (Data from Bocchini, G.F., Advances in Powder Metallurgy and Particulate Materials2001, Vol. 6, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 5686.)
850 800 750 700 650 600 550 500 450 400 0.01 0.1 1 Test load (kg) 10 100
FIGURE 13.10 Dependence of Vickers hardness on applied load of heat treated P=M steels with densities between 6.8 and 7.0 g=cc. (Data from Ferguson, H., Metal Progress, ASM, 107(6), 8183, June 1975; 108(2), 6669, July 1975.)
VHN
boundaries and along porosity channels, which cause embrittlement and loss of strength. As alloying elements such as nickel, molybdenum, chromium, and copper are added, the optimum carbon level content is lowered. In sintered steels, the most common alloying elements are copper and nickel.
13.6.1 COPPER CONTENT

The addition of copper increases both hardness and tensile strength in the sintered condition. On heat treatment, they increase depth of hardness but reduce toughness and elongation. As copper content increases, rupture strength rises to an optimum level and then declines. In the plain ironcarbon compositions, peak strength occurs near the euctectoid composition in the as-sintered condition and at approximately 0.65% C in the heat-treated condition. FeC systems soften progressively as the tempering temperature is raised whereas FeCCu systems show significant temper resistance of up to 3718C (7008F) [7]. Figure 13.11 shows the effect of tempering temperature and composition on hardness for FeC and FeCCu compositions. The variations in the apparent hardness can be due to the effect of composition on the martensite percentage formed along with the effect of tempering. Figure 13.12 and Figure 13.13 illustrate plots showing variation of hardness and transverse rupture strength in the assintered and heat-treated condition. From the above figures, it is clear that for FeCCu composition, the best heat-treated properties are obtained at around 2% copper. Heat-treated strength overlaps with the sintered strength at higher copper levels as shown in Figure 13.13.
13.6.2 NICKEL CONTENT

Nickel increases tensile properties of as-sintered P=M steels approximately one half as much as copper but provides significantly higher strengths after heat treatment. This is because of the heterogeneous nature of nickel alloy steels made from blended elemental powders. Copper melts at 10838C (19808F) and forms a liquid phase on sintering at 11208C (20508F), creating a more homogeneous ironcopper alloy. Nickel has a higher melting point than the sintering temperature, and it alloys by solid-state diffusion. This produces a duplex microstructure consisting of partially alloyed iron-surrounding nickel-rich islands. On quenching, the matrix transforms to martensite but the nickel-rich areas remain austenitic. This combination results in a stronger matrix with some toughness and ductility retained.
13.6.3 NICKELCOPPER CONTENT

The effect of increasing nickel and copper contents on the hardenability of P=M steels is illustrated in Figure 13.14. Jominy bars were pressed to a density of 6.7 g=cm3 with increasing levels of copper and nickel, while carbon content was maintained at 0.5% C. After sintering, all the Jominy bars were austenitized at 8508C (15608F) for 2 h and quenching. The addition of 2.5% copper produced only a relatively small increase in hardenability compared to the ironcarbon alloy. However, the surface hardness increased dramatically. Adding 1% nickel to the ironcoppercarbon alloy produced little change in the surface hardness but a significant improvement in hardenability, while a 2.5% nickel addition improved surface hardness slightly, but dramatically increased the hardenability. Many P=M components that require optimum heat-treated properties contain both nickel and copper. Figure 13.15 shows Jominy curves for FLC-4608 at different density values [8]. It shows that there is no significant drop in the hardness through Jominy position 36, which indicates its relatively high hardenability.
45 0.3 0.5 40 0.7 0.9 35
30
25
20
15
0% Copper
1% Copper
Hardness (HRC)
10 45
40
35
30
25
20
15 2% Copper 10 150 200 250 300 350 400 450 500 150 200 250 300 350 400 450 500 Tempering temperature (C)
FIGURE 13.11 Variation of hardness with tempering temperature and composition. (Data from Capus, J.M. and Maaref, A., Proceedings of the 1973 International Powder Metallurgy Conference, Vol. 8, 1974, pp. 6173.)
3% Copper
50
40
Heat-treated hardness (Rc)
30
Heat-treated Sintered hardness (Rb) 90
20
10
Sintered
80
70
60 50 40 0 1 2 Copper (%)
FIGURE 13.12 Variation of apparent hardness with copper content in as-sintered and heat-treated condition. (From Capus, J.M. and Maaref, A., Proceedings of the 1973 International Powder Metallurgy Conference, Vol. 8, 1974, pp. 6173.)
13.6.4 MOLYBDENUM CONTENT

Molybdenum is a more effective alloying element than nickel for enhancing hardenability. Also it has lower affinity for oxygen and lower effect on compressibility as compared to nickel. Typical additions range from 0.5 to 1.5%. The addition of molybdenum promotes
160
Transverse rupture strength (1000 psi)
140
Heat-treated
120
100
Sin
tere
80
60
40
1 Copper (%)
FIGURE 13.13 Variation of transverse rupture strength with copper content in as-sintered and heat-
treated condition. (From Capus, J.M. and Maaref, A., Proceedings of the 1973 International Powder Metallurgy Conference, Vol. 8, 1974, pp. 6173.)
78 76 74 72 HRA 70 68 66 64 62 60 7.15 g/cm3 7.02 g/cm3 6.73 g/cm3
10
20
30 Distance (mm)
40
50
60
FIGURE 13.14 Jominey end-quench results on FLC-4608. (From Pershing, M.A. and Nandi, H., Advances in Powder Metallurgy and Particulate Materials2001, Vol. 5, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 2630.)
800 Fe + 2.5Cu + 0.5C 700
600 Fe + 2.5Cu + 2.5Ni + 0.5C 500 Fe + 2.5Cu + 1Ni + `0.5C VHN 400
300
200 Fe + 0.5C 100
0 0 0.5 1 1.5 2 Distance (mm) 2.5 3 3.5 4
FIGURE 13.15 Effect of copper and nickel addition on the hardenability of P=M steels. Austenitizing
condition was 2 h at 8508C (15608F) and the parts were oil quenched. (From Ferguson, H., International Journal of Powder Metallurgy, 39(7), 3338, 2003.)
response to heat treatment, and increases hardness, wear resistance, and strength to the required level for highly loaded P=M applications such as helical gears. A comparison of the effect of cooling rate on properties for P=M steels containing nickel, copper, and molybdenum is shown in Figure 13.15 through Figure 13.18. Figure 13.16 shows the increase in hardenability with higher nickel and copper of up to 4% Ni and 2.25% Cu. The effect of these alloying additions is particularly apparent in the tensile strength results (Figure 13.17). The ductility is also improved significantly at all cooling rates tested (Figure 13.18). Figure 13.19 shows the effect of alloying and cooling rate on the dimensional change in the alloys. The highest addition likewise results in the smallest change in dimensions, which is the optimum result. Also, the small changes in dimensions at all cooling rates should result in lower quenching stresses and residual stresses.
13.7 EFFECT OF STARTING MATERIAL ON HOMOGENIZATION

Another difference between P=M steels and wrought steels is the degree of inhomogeneity, which is common in P=M parts. The alloying additions, discussed above, typically are added by blending powders, called admixed, of the alloying elements with iron powder. The most important reason that the alloying is done this way, rather than melting the elements together and producing the powder, called prealloyed, is to have a soft powder mixture that can be easily compacted. It is possible that a given P=M steel part could have one of four different powders as its starting material, which affects the properties and heat treatment of the steel.
350
300
250 Hardness (HV20)
200
150 SA 100 SE SF 50
0 0.1
1 Cooling rate (8C/s)
10
FIGURE 13.16 Effect of cooling rate of the hardness of three diffusion-alloyed P=M steels with 0.45% C. Sintered for 30 min at 11008C in endogas. SA, Fe1.75Ni1.5Cu0.5Mo; SE, Fe4Ni1.5Cu0.5Mo; and SF, Fe4Ni2.25Cu0.5Mo. (From Lindskog, P. and Tornblad, O., 5th European Symposium on Powder Metallurgy, Vol. 1, 1978, p. 98; Lindskog, P. and Thornblad, O., Powder Metallurgy International, No. 1, 1979, p. 10; and Exner, H.E. and Danninger, H., Powder Metallurgy of Steel, Metallurgy of Iron, Vol. 10a,b, Springer-Verlag, Heidelberg, 1993, pp. 129a209a, 72b141b.)
The four types are as follows:

.
AdmixedElemental or master alloy powders are added to a base iron powder [9]. This method is used widely and has the lowest cost. Due to the high fraction of the mix exists as compressible (easy to compact) iron powder, it usually retains most of the compressibility. Elements that form intermetallic alloys with a low-diffusion rate are not included and the degree of alloying is limited by the mutual diffusivity of the other components. This precursor material typically has the higher degree of inhomogeneity and is particularly prone to powder segregation during handling. Diffusion alloyedThe alloying elements are diffusion-bonded to the base iron powder particles such that the core iron retains its high compressibility. Segregation of the powder during handling is essentially eliminated, but the powder still produces a very inhomogeneous microstructure with an alloying gradient varying between the center of the former powder particles and the highly alloyed interparticle boundaries. PrealloyedAll alloying elements, except carbon, are added to the melt, which is atomized to produce the powder. The sintered powder has homogeneous microstructures and compositions. However, the compressibility suffers with most alloying elements compared to admixed- and diffusion-alloyed powders and hence prealloyed powders are often considered undesirable for many turnkey operations. Hybrid alloysMolybdenum is one alloying element, which can be in solution in iron and still be very compressible. Prealloyed- or diffusion-alloyed powders of molybdenum
900
800 Tensile strength (MPa)
700
600
SA SE SF
500 0.1 1 Cooling rate (8C/s) 10
FIGURE 13.17 Effect of cooling rate on the tensile strength of three diffusion-alloyed P=M steels with
0.45% C. Sintered for 30 min at 11008C in endogas. SA, Fe1.75Ni1.5Cu0.5Mo; SE, Fe4Ni1.5Cu 0.5Mo; and SF, Fe4Ni2.25Cu0.5Mo. (From Lindskog, P. and Tornblad, O., 5th European Symposium on Powder Metallurgy, Vol. 1, 1978, p. 98; Lindskog, P. and Thornblad, O., Powder Metallurgy International, No. 1, 1979, p. 10; and Exner, H.E. and Danninger, H., Powder Metallurgy of Steel, Metallurgy of Iron, Vol. 10a,b, Springer-Verlag, Heidelberg, 1993, pp. 129a209a, 72b141b.)
5 SA SE 4 SF
Elongation (%)
0 0.1 1 Cooling rate (8C/s) 10
FIGURE 13.18 Effect of cooling rate on the elongation (%) of three diffusion-alloyed P=M steels with
0.45% C. Sintered for 30 min at 11008C in endogas. SA, Fe1.75Ni1.5Cu0.5Mo; SE, Fe4Ni1.5Cu 0.5Mo; and SF, Fe4Ni2.25Cu0.5Mo. (From Lindskog, P. and Tornblad, O., 5th European Symposium on Powder Metallurgy, Vol. 1, 1978, p. 98; Lindskog, P. and Thornblad, O., Powder Metallurgy International, No. 1, 1979, p. 10; and Exner, H.E. and Danninger, H., Powder Metallurgy of Steel, Metallurgy of Iron, Vol. 10a,b, Springer-Verlag, Heidelberg, 1993, pp. 129a209a, 72b141b.)
0.1 0.08 0.06 Dimensional change (%) 0.04 0.02 0 0.02 0.04 0.06 0.08 0.1 0.1 1 Cooling rate (8C/s) 10 SA SE SF
FIGURE 13.19 Effect of cooling rate on the dimensional change of three diffusion-alloyed P=M steels with 0.45% C. Sintered for 30 min at 11008C in endogas. SA, Fe1.75Ni1.5Cu0.5Mo; SE, Fe4Ni 1.5Cu0.5Mo; and SF, Fe4Ni2.25Cu0.5Mo. (From Lindskog, P. and Tornblad, O., 5th European Symposium on Powder Metallurgy, Vol. 1, 1978, p. 98; Lindskog, P. and Thornblad, O., Powder Metallurgy International, No. 1, 1979, p. 10; and Exner, H.E. and Danninger, H., Powder Metallurgy of Steel, Metallurgy of Iron, Vol. 10a, b, Springer-Verlag, Heidelberg 1993, pp. 129a209a, 72b141b.)
in iron have master alloy additions to create a series of hybrid alloy steel powders. The sintered powder typically has a less heterogeneous microstructure but retains most of the compressibility. Carbon is also typically added in the form of graphite powder. However, carbon diffusivity is high at the sintering temperatures. The easy diffusion of carbon leads to homogenous distributions and uniform microhardness values when no other alloying additions are present. The beneficial effects of nickel additions have been shown above. However, nickel diffuses slower than other elements, making homogenization more difficult. Consequently, it is common for the nickel composition to have a greater degree of nonuniformity than other elements [11]. Figure 13.20 illustrates a plot of the diffusivity of nickel in austenite compared to those for molybdenum and chromium [10]. The diffusion of nickel in iron between 900 and 9508C (1650 to 17908F), common austenization temperatures, is only 1% of the rate at sintering temperatures. A rise from 1100 to 11508C (2010 to 21008F) produces a doubling of the diffusivity. In this temperature range, an error of only 258C can cause a 50% difference in nickel diffusion rate. This makes the control of temperature uniformity in the furnace very important and affects loading of parts in the furnace. Figure 13.21 shows the effect of composition and starting powder on the hardenability. The prealloyed FL-4605 is similar in composition to the FN-0205, but has a much greater hardenability, although the surface hardness is only slightly better. The addition of nickel to F-0005 to make F-0205 also increases hardenability and surface hardness. As a consequence of the variety of local chemical compositions, which are present after sintering, unless completely prealloyed powders are used, different microstructures will be formed on cooling. As the degree of difference in compositions locally can be very significant, it is better to look at the concept of critical cooling speed in terms of these local variations [11]. Figure 13.22 shows the conditions of maximum and minimum hardenability, and assuming that some carbon is present, it can be pointed out that [12]
1E11 Cr 1E12 Mo 1E13 Diffusivity (m2/s) 1E14 Ni 1E15 1E16 1E17 1E18 850
900
950
1000
1050
1100
1150
1200
Temperature (C)
FIGURE 13.20 Difficulty of Cr, Mo, and Ni in Fe at sintering temperatures calculated from data from Smithels Metals Reference Book, 7th ed., Brandes, E.A., and Brook, G.B., Butterworth-Heinemann, Oxford, 1999, pp. 1370 to 1397
When cooling speeds are lower than Vp, only pearlitic structures are formed. The V p would be the critical speed to form only pearlite in the regions richest in alloying elements. When cooling speeds are higher than Vm, the structure becomes completely martensitic on cooling. Likewise, V m would be the critical speed to form only martensite in the regions, richest in alloying elements.
0 The cooling speeds used in typical industrial practice after sintering are generally between Vp and Vm. This situation typically occurs when materials based on admixed- or diffusionbonded powders containing Ni and Mo are used. Consequently, even without a specific
85 80 75 70 HRA 65 60
4140
55 50 45 40 0 2 4 6 8 10 12 Distance (mm) 14 16
1045 8620 FL-4605 FN-0205 F-0005
18
20
FIGURE 13.21 The effect of composition and starting powder on hardenability. See text for explanation.
(From Ferguson, H., International Journal of Powder Metallurgy, 39(7), 3338, 2003.)
1000 900 800 Temperature (C) 700 600 500 400 300 200 100 0
Local alloy content Low High
V p
V p
Vm
Time, logarithmic scale
V m
FIGURE 13.22 The effect of local variations in alloying elements on transformations in P=M steels. The inhomogenears composition leads to a variety of responses during heat treatment. (From Bocchini, G.F., Advances in Powder Metallurgy and Particulate Materials2001 Vol. 6, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 5686.)
heat treatment, it is possible to observe some martensite (or possibly bainite) islands in the as-sintered state. In sintered steels, many small areas of hard phasesregularly distributed are surrounded by softer microconstituents, which can plastically deform, through a self-adapting mechanism in case of high-applied stresses, by a cushion or damper effect [11]. Moreover, the slow diffusion of nickel results in high local concentrations in many points. When the local concentration exceeds a given threshold, austenite will survive independently of the cooling speed. This austenite should be clearly differentiated with respect to retained or residual austenite, because it is stable or incoercible. The areas of soft austenite increase the resistance to stresses when the part is put in operation by means of local plastic-yielding and self-adapting. Lindskog and Boccini [13] explained in more detail the favorable combination of properties, typical of steels obtained from diffusion-bonded powders. According to Boyer [14], in many instances, finely dispersed residual austenite of amount up to 30% is not detrimental to pitting fatigue strength of fully dense steels. This is partly because small amounts of retained austenite apparently allow mating surfaces to conform slightly faster and to spread the load more evenly, thus reducing the local area of high stress. This is similar to running-in bearings at low load to prolong their life at high loads. As demonstrated by repeated experiences, P=M steels from Ni-containing diffusion-bonded powders frequently exhibit randomly distributed (or finely dispersed) austenite areas. Following Boyer, this peculiarity should be seen as a favorable factor explaining several positive results of P=M parts, which have undergone demanding stress conditions. However, much lower amounts of austenite can be harmful if the austenite is not finely dispersed.
13.8 QUENCH AND TEMPERING

The as-quench properties of P=M steels depend on the cooling medium (severity of quench) and the P=M steels hardenability. Quenching in water, salt, brine, or water-based polymer solution improves the rate of heat transfer, leading to a faster quench, but in many cases, it accelerates corrosion due to the residual quenchant trapped in surface pores. Oil quenching, although less severe than cooling in water or brine, is favored due to minimized distortion and cracking. Control of oil temperature is a must for achieving consistent results from
load-to-load. A fast oil (9 to 11 s in the magnetic quenchometer test) can be used to improve the heat transfer characteristics and is therefore recommended. The P=M parts can absorb nearly 2 to 3% oil (by weight), which can cause problems during postquench treatment. If the oil is not completely removed, it results in heavy smoke during the tempering operation. This can not only lead to health hazard but also poses safety concerns due to the large amount of oil in the tempering furnace. Also, the much higher amount of drag out of the oil needs to be considered during processing. Most sealed quench furnaces are designed to quench out 0.5 kg (1 lb) of parts per gallon of quench oil. For P=M parts, it is recommended that 11 to 15 l (3 to 4 gal) of quench oil be used per 0.5 kg (1 lb) of parts. Usually parts having densities greater than 6.7 g=cm3 should be tempered after hardening. Recommended tempering temperatures for P=M parts range from 150 to 2008C (300 to 3928F) with 1778C as very common. The effect of tempering temperature on impact properties is shown in Figure 13.23 for FL-4205 at various densities. Tempering above 2008C (3908F) results in improved toughness and fatigue properties of the hardened part, while making a compromise in tensile strength and impact resistance. However, when tempering at temperatures higher than 2008C (3928F), care should be exercised as entrapped oil can ignite. Tempering is also done to reduce the effects of retained austenite formed in low-alloy steels when rapidly quenched. Freezing by cryogenic treatment (below 1008C or 1488F) transforms the retained austenite (not the stable austenite in some nickel steels) to martensite. Following the cryogenic treatment, a second temper is done at 2008C (3908F) to relieve the stresses in newly formed martensite. Table 13.4 shows the heat-treat properties of four different low-alloy steels tempered at 1768C (3508F) for 1 h [15]. The sintered carbon contents for the nominal 0.5% carbon level alloys varied from 0.47 to 0.53% carbon for all the listed alloys. The effect of varying carbon content on the mechanical properties of FL-42XX and FL-46XX alloy system is listed in Table 13.5. Table 13.6 lists some recommended quench and temper parameters to achieve good wear resistance and core strength based on different ranges of porosity. These are of course general conditions and should be used only for a starting point in deciding how to heat-treat P=M steels.
60
4xxx
55 50 HRC 45 40 35 30 0 100 200 300 Temperature (C) 400 500
0.45% C 0.55% C 0.75% C
600
FIGURE 13.23 True hardness (VHN100 converted to Rockwell C) of several P=M steels vs. tempering
temperature. With porosity effect removed the response during tempering is close to wrought. (From Moyer, K.H. and Jones, W.R., Heat Treating Progress, 2(4), 6569, 2002.)
Heat Treatment of Powder Metallurgy Steel Components
TABLE 13.4 Heat-Treated Properties of Low-Alloy Steels

Density (g=cm3) Alloy Grade FL-4205 TRS 6.72 6.95 7.17 6.76 7.00 7.29 6.61 6.95 7.17 6.70 7.06 7.32 Tensile 6.78 6.97 7.15 6.75 6.94 7.31 6.73 6.96 7.19 6.72 7.01 7.33 Impact 6.72 6.95 7.17 6.75 6.94 7.31 6.61 6.95 7.17 6.72 7.01 7.33 TRS (psi) 190,100 (232,000) 270,100 185,900 227,100 299,700 168,700 (240,000) 280,600 177,100 255,200 336,000 UTS (psi) 129,700 150,700 179,600 121,900 155,900 202,200 125,000 147,900 175,300 115,100 159,900 202,600 YS (psi) 159,400 146,700 179,900 Elongation (%) <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 1 Impact Energy (ft-lbf) 6.0 (9.0) 10.5 6.3 8.7 13.2 5.7 (8.5) 11.5 7.0 10.2 17.3 Hardness (HRC) 28.9 34.8 40.5 28.2 35.0 41.4 27.6 33.5 37.8 31.6 39.7 45.2
FL-4405
FL-4605
FL-4205 1.5 Ni
Note: Values in parentheses estimated from density carbon content data. Source: From Sanderow, H.I. and Prucher, T., Advances in Powder Metallurgy and Particulate Materials1994, Vol. 7, Metals Powder Industries Federation, Princeton, NJ, 1994, pp. 355366.
771
TABLE 13.5 The Effect of Varying the Carbon Content on Heat-Treated Properties of Low-Alloy Steels
Alloy Grade FL-42XX % Carbon 0.12 0.38 0.50 0.67 0.16 0.44 0.50 0.68 UTS (psi) 88,200 180,300 179,600 151,300 121,200 175,600 175,300 141,700 TRS (psi) 147,600 270,100 255,400 202,800 280,600 249,400 Impact Energy (ft-lbf) 18.2 10.5 9.7 13.5 11.5 12.2 Hardness (HRC) 8.5 39.8 40.5 45.5 19.8 40.3 37.8 43.3
FL-46XX
Note: All samples tempered at 1768C (3508F) for 1 h with sintered density range of 7.157.20 g=cm3. Source: From Sanderow, H.I. and Prucher, T., Advances in Powder Metallurgy and Particulate Materials1994, Vol. 7, Metals Powder Industries Federation, Princeton, NJ, 1994, pp. 355366.
Other special consideration for quench and tempering P=M parts is the large amount of surface area associated with the porosity. The significance of this effect can be seen in Table 13.7. A steel P=M part with a density of 6.81 g=cm3 could have approximately 100 times the active surface area than expected based on part size and loading. This has important effects on the uniformity of the load, as the atmosphere composition will have a greater change from the inlet to the exhaust. Very good circulation of gases must be maintained to achieve a higher degree of uniformity. Table 13.8 lists a number of P=M steel alloys and their hardenabilities. The hardenabilities are in terms of the depth in Jominy distance (1=16 in.) to which the Jominy bar was hardened to a minimum of 65 HRA (~29.5 HRC). The effects of density and alloy content can be readily seen.
13.9 SINTER HARDENING

It is common to have a separate, postsintering heat-treatment unit, but in some cases, it is possible to incorporate the heat treatment in the cooling portion of the sintering cycle, as done in sinter hardening. The term sinter hardening refers to a process in which the cooling rate, experienced in the cooling zone of the sintering furnace, is fast enough to transform a
TABLE 13.6 Recommended Heat Treatment Conditions for Best Wear Resistance and Core Strength at Different Densities
Quenching Density (g=cm3) 6.46.8 6.87.2 >7.2 Austenizing Temperature (8C [8F]) 870890 (16001635) 850870 (15601600) 820850 (15101560) Soak Time (min) 3045 4560 6075 Transfer Time (s) <8 <12 <25 Transfer Medium Fast oil Fast oil Medium to fast oil Tempering (8C [8F]) 150180 (300355) 170220 (340430)
Source: From Ferguson, H.A., Metals Handbook, 10th ed., Heat Treating, 10th ed., Vol 4, ASM, Materials Park, OH, 1991, p. 229.
TABLE 13.7 Extra Surface Area Associated with Porosity in P=M Parts
Density (g=cm3) Total porosity Open porosity Ratio opentotal porosity Total pore surface (mm2) Total surface of open pores (mm2) Ratio open porosity=external area 5.71 27.4% 26.8% 97.8% 451,000 440,000 220x 6.81 13.5% 10.4% 77% 274,000 211,000 105x 7.22 8.3% 4% 48% 209,000 100,000 50x
Source: From Rudnayova, E., Salak, A., and Zabavnik, V., Zbornik Vedeckych Prac VST, Kosice, No. 1, 1975, p. 177.
TABLE 13.8 Hardenability of P=M Steels vs. Density

Depth to 65 HRA Material Designation Density (g=cm3) 6.65 6.87 7.03 6.78 6.91 7.06 6.50 6.82 6.96 6.40 6.81 7.15 6.90 7.10 7.38 6.88 6.97 7.37 6.75 7.00 7.20 6.64 6.94 7.20 6.76 6.99 7.12 6.68 7.00 7.29 6.71 7.11 7.22 Depth (1=16 in.) <1.0 1.0 1.0 1.5 2.0 2.0 <1.0 1.5 1.5 1.5 2.0 2.5 1.5 1.5 2.5 2.0 2.0 3.0 2.5 3.5 3.5 2.0 3.0 4.5 2.5 5.0 7.0 2.0 5.0 6.0 7.5 10.5 10.5 Material Designation Density (g=cm3) 6.72 7.10 7.23 6.79 7.15 7.30 6.63 7.06 7.14 6.72 7.08 7.16 6.82 7.08 7.27 6.65 7.06 7.22 6.98 7.24 7.32 6.78 6.97 7.29 6.70 7.13 7.26 6.70 7.08 7.21 7.40 7.39 7.38 Depth (1=16 in.) 8.5 14.5 17.5 13.0 18.0 26.0 26.0 36.0 36.0 8.5 9.5 10.5 22.5 36.0 36.0 9.0 11.0 15.5 2.5 2.5 4.5 4.0 9.5 12.0 2.0 4.0 10.0 3.0 8.0 15.0 2.0 2.5 <1.0
F-0005
FLN4-4405
F-0008
FLN6-4405
FC-0205
FLC-4608
FC-0208 FN-0205
FLC-4908 FLN-4608
FN-0208
FLNC-4408
FL-4205
FD-0205
FL-4405
FD-0208
FL-4605
FD-0405
FLN-4205
FD-0408 FX-1005 FX-1008 FX-2005
FLN2-4405
significant portion of the material matrix into martensite. The sinter-hardening route has become increasingly popular due to its cost-effectiveness. Advantages of sinter hardening:
. . . .
Secondary quench-hardening treatment is eliminated. Less severe quench leads to better dimensional control. Retention of quench oil in pores eliminated is leading to fewer steps during tempering. Oil does not have to be removed to perform secondary operations such as plating
Disadvantages of sinter hardening:

. .
Sinter hardened parts may not be able to be machined due to high hardness. Sinter hardened parts need tempering to relieve stresses and improve dynamic mechanical properties. High apparent hardness of sinter-hardened parts may make sizing or other secondary operations impractical
Cooling rate during sinter hardening can have a significant effect on the final properties. While there are compositions, which do not require any form of accelerated cooling to harden, there is still a gain in properties with faster cooling rates. Fans, which circulate the atmosphere in the cooling zone of the furnace, can increase the cooling rate, while even faster cooling rates can be achieved with a pressure gas quench. Figure 13.17 through Figure 13.20 show the effect of cooling rate during sinter hardening on the mechanical properties and dimensional change of three 0.45% C diffusion prealloyed steels. As discussed earlier, composition control can lead to very good hardenability upon cooling from the sintering temperature. Pressurized gas quenching is also effective in producing higher hardness when sinter hardening. However, this requires that the sintering furnace be equipped with an integral gas quench chamber. Since this process is of more batches than continuous, there is usually a reduction in throughput. Vacuum sinter hardening and gas quenching show some promise for another method to achieve higher hardness. A comparison of vacuum sinter-hardened materials to wrought 4600 steel is shown in Figure 13.24. The P=M steels were sintered for 2 h at 12608C (23008F) in H2, followed by a 2-bar helium gas quench. The time maintained at each temperature is 1 h. As shown, the hardness is very comparable with the wrought equivalent.
13.10 WARM COMPACTION

A major advance in P=M technology has been the warm compaction process, which yields higher density levels along with uniform density distribution. The process is applicable to both ferrous and stainless steels (still in developmental stage). The raw material consists of conventional powder mixed with a special lubricant. The lubricant system has excellent flowability and filling behavior at increased temperatures. The equipment consists of a slot heater filled with powder and fitted with oil-heated elements. The heated powder from the slot heater flows freely through electrically heated insulated hose to the filling shoe and the die assembly, which are also heated. The die assembly consists of the die, upper punch, and the adaptor table. The temperature range is 100 to 1508C (212 to 3028F). The lubricant begins to
25 7.25 g/cm3 20 7.05 g/cm3 6.75 g/cm3
Impact energy (J)
15
10
0 0 50 100 150 200 250 300 Tempering temperature (C)

FIGURE 13.24 Effect of tempering temperature and density on the impact properties of FL-4205. (From
ASM Handbook, Powder Metal Technologies and Applications, Vol. 7, ASM International, Metals Park, OH, 1998, p. 650.)
break down at temperatures higher than 1508C (3028F). The apparent density is decreased and the flow rate increases, which limits the working temperatures for warm compaction. Green machining of warm compacted parts is possible partly due to the increased green strength of the compact. The mechanical properties for warm-compacted, heat-treated single press single sinter (SPSS) FLN2-4405 are given in Table 13.9 [16]. The specimens were sintered in a belt furnace at 11208C (20508F) for 20 min in a 9010 N2H2 atmosphere. Hardening was performed in an integral quench furnace at 8708C (16008F) at a 0.8% carbon potential with a subsequent oil quench. Tempering was done at 1758C (3508F) for 1.5 h.
13.11 POWDER FORGING

The heat treatment of powder-forged parts is simpler in that there is a relatively small amount of porosity. Therefore concerns with low thermal conductivity and open surface porosity are practically nonexistent. The preform can be made from prealloyed powders or diffusionbonded powders. Given sufficient forging strain and temperature, the chemical inhomogeneity of the diffusion-bonded preform can be reduced significantly. Properties depend on the forging design, as it controls porosity and homogeneity. Part surfaces tend to have somewhat higher porosity than the bulk, however, proper design eliminates most of these from consideration when the parts are heat-treated. The difference of forged parts from typical wrought parts is the composition used. Powders tend to have lower Mn contents than wrought steels and a few other differences in minor alloying elements. These changes are relatively small. Figures 13.25 through Figure 13.27 show Jominy curves for several powder-forged steels.
776
TABLE 13.9 Heat-Treat Mechanical Properties for SPSS Warm-Compacted FLN2-4405 at Different Compaction Pressures
Compaction Pressure (tsi) (MPa) 30 (415) 40 (550) 50 (690) 60 (825) Green Density (g=cm3) 6.92 7.17 7.29 7.32 Quench and Tempering Density (g=cm3) 6.95 7.19 7.33 7.39 TRS (103 psi) (Mpa) 184 (1270) 242 (1670) 270 (1860) 287 (1980) Apparent Hardness HRA 72 75 76 77 Tensile Strength (103 psi) (MPa) 102 (572) 145 (690) 165 (779) 163 (759) 0.2% Yield Strength (103 psi) (MPa) 143 (490) 160 (538) 161 (587) Elongation (%) 0.7 0.9 0.9 0.9 Impact Energy (ft-lbf) (J) 5 (7) 7 (9) 8 (10) 9 (12)
Steel Heat Treatment: Metallurgy and Technologies
90 P/ F-1020 80 P/F-11C40 P/ F-1060 50 HRC 70
HRA
60
50
40
30 0 5 10 15 20 25 30 35
Jominy distance (1/16 in.)

FIGURE 13.25 Jominy curves for various powder-forged P=M steels.
90 P/ F-422 0 P/ F-424 0 80 P/ F-426 0 50 HRC 70
HRA
60
50
40
30 0 5 10 15 20 25 30 35

90 P/F-4620 P/F-4640 80 P/F-4660 P/F-4680 50 HR C 70
HRA
60
50
40
30 0 5 10 15 20 25 30 35
13.12 CASE HARDENING

As mentioned earlier, when porosity levels are such that there is significant open porosity, the pores are interconnected by small channels. This allows gases to penetrate with little hindrance. When case-hardening P=M parts contain gases with a high-carbon potential, this interconnected porosity allows diffusion of carbon into the internal pore surfaces as well as to the external surfaces of the part. This can result in carburization to a significant depth and consequent loss of any sharply defined case-hardening effect. The same condition can also be applied to other gaseous surface-hardening procedures such as nitriding or carbonitriding. In fact, the swelling, which occurs during nitriding, can be very detrimental if it occurs internally due to open porosity. Figure 13.28 shows the effect of density on the case depth during vacuum carburizing of an ironcopper alloy. Lower densities produce much higher case depth. The line for 50 HRC indicates the effective case. P=M parts are case-hardened to obtain higher hardness, wear, fatigue, or impact properties. The primary purpose of case hardening low-carbon steels is to provide a hard, wearresistant surface while maintaining a soft-ductile core. This combination imparts optimum wear resistance and toughness properties to these materials.
13.12.1 CARBURIZING
The material usually specified for carburizing contains allowing elements such as nickel, molybdenum, and copper to increase hardenability, but with relatively low-carbon content. To develop optimum dynamic properties at porosity levels between 10 and 15%, a combined carbon level of 0.30 to 0.35% is recommended. As porosity is reduced below 10%, combined carbon can be reduced to 0.15 to 0.25%. As improved dynamic properties are also associated with high densities, it is recommended that combined carbon can be adjusted to a level best suited to repressing after sintering. Parts with porosity levels much higher than 15% are not
800 750 700 650 600 VHN 7.8 g/cm3 550 500 450 400 350 300 0 0.1 0.2 0.3 0.4 Depth (mm)
FIGURE 13.28 Effect of porosity on carbon penetration and effective case depth during vacuum Carburizing. Ferguson, H.A., Metals Progress, 7(6), 1975, 8183.
7.0 g/cm3
6.8 g/cm3
50 HRC
0.5
0.6
0.7
0.8
recommended for carburizing because of the penetration of the pores by the carburizing gas. As already mentioned, liquid carburizing is not suitable for P=M parts because of the problem of washing the parts free of salt after the treatment. Table 13.10 gives some expected results from carburizing for different ranges of porosity. In wrought steel, carburization is normally characterized by a surface-hardness range and an effective case depth. Microhardness measurements can accurately show the hardness profile in a cross section of wrought steel but can be erratic when used on P=M steels, where subsurface porosity can influence the hardness readings. It is recommended that at
TABLE 13.10 Effect of Density on Features of Carburized Layers on P=M Steels

Density, g=cc >7.1 7.1 to >6.9 6.9 to >6.6 <6.6 Typical Properties of Carburized Layers Definite thickness, repeatable with precision, analogously to the one, which occurs when carburizing fully dense steels The thickness may exhibit a certain scattering, which is fairly controllable through careful control of the carburizing process, carried out in a reliable equipment The thickness may exhibit nonnegligible scattering, not withstanding the care and the precision in conducting the carburizing process The thickness may exhibit very high variations, from lot to lot, from part to part, and below differently oriented part surfaces
Source: From Bocchini, G.F., Advances in Powder Metallurgy and Particulate Materials2001, Vol. 6, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 5686.
least three readings be taken at each level below the surface and averaged to determine effective case depth (a detailed review of this procedure can be found in Ref. [13]). P=M steels are usually gas-carburized at temperatures between 900 and 9308C (1650 and 17058F). Time cycles are normally short because of the rapid diffusion of carbon through the interconnected porosity. Therefore, atmospheric carbon potentials for P=M steels need to be somewhat higher than those required for wrought steels of similar composition. The rate of carbon penetration into steel is controlled by the diffusion of carbon in austenite and not by the kinetics of the chemical reactions at the surface [17,18]. As a general guideline, the carburizing of ferrous P=M components can be achieved with a lower carbon potential and in a shorter time compared to their wrought and cast counterparts. For an 86% dense P=M alloy, the carbon can be raised from 0.04 to 0.5% at a depth of 2 mm (0.08 in.) in just 2 h at 0.90% carbon potential. Figure 13.29 shows the effects of density and atmosphere on the case depth achieved at 8708C (16008F).
13.12.2 CARBONITRIDING
Carbonitriding is a modified form of gas carburizing, which uses additions of up to 10% ammonia into the gas-carburizing atmosphere, and results in the addition of nitrogen to the carburized case. Dissociated nitrogen forms at the workpiece surface from the ammonia in the furnace atmosphere. Nitrogen diffuses into the steel surface simultaneously with the carbon, retarding the critical cooling rate on quenching and leading to a more consistent martensite transformation. It also produces a more consistent surface hardness profile, which improves wear resistance and toughness of the P=M steel. Process temperatures for carbonitriding are lower (800 to 8508C (1470 to 15608F)), providing better control of distortion than carburizing. Carbonitriding is a shallow case-hardening treatment. Case depths greater than 0.5 mm (0.02 in.) are seldom specified. For this reason, cycle times are relatively short, usually on the
2 Vacuum - - - Atmosphere 1.5 Case depth (mm)
7.0 g/cm3 1
7.4 g/cm3 0.5
0 0 10 20 30 Carburizing time (h)

FIGURE 13.29 Dependence of case depth after carburizing P=M parts at 8708C to microhardness of 700
40
50
60
HV (10 g) on carburizing time and density. (From Jansson, P., Powder Metallurgy, No. 1, 1992, p. 63.)
order of 30 to 60 min. As in neutral hardening, carbon control is a critical aspect of the treatment. Normally, carbon potentials of 1.0 to 1.2% are specified to maintain the carbon profile in the part.
13.12.3 INDUCTION HARDENING

Spur gears, bevel gears, splined hubs, and cams are ideal components for the production using P=M. These parts usually require hard, wear-resistant surfaces at some location, and a more tougher ductile matrix. Locally hardening the piece helps to maintain dimensions. Induction hardening is commonly specified for these applications. The inductance of P=M materials is typically less than that of a wrought material of similar composition due to porosity; a higher power setting is normally required to reach a given depth of hardening. Furthermore, as the heat is rapidly dissipated, a rapid transfer to the quench is mandatory. As with wrought steels, the response to hardening by induction is dependent on combined carbon content, alloy content, and surface decarburization. This latter variable can be a major concern with P=M parts. With todays conventional belt-type sintering furnaces using an endogas atmosphere, decarburization can occur as the parts leave the hot zone and cool slowly through the 1100 to 8008C (2010 to 14708F) temperature range. In most instances, induction-heated P=M parts are quenched in a water-based solution, containing some type of rust preventative to forestall internal corrosion. In those applications, where induction hardening is considered, densities above 90% should be specified. With a decrease in density, the resistivity of the steel increases and permeability decreases. For this reason, integral quench coils using a high-velocity spray quench are generally used to attain maximum surface hardness in the P=M part. Drag out of the quenchant will depend on the open surface porosity as discussed with oil quenchants. The effect of porosity on the induction-hardening process is dependent on the material, as mentioned above. The magnetic properties of P=M steels are not all known, so predicting behavior is somewhat difficult. One study considered some measured parameters and calculated the relative penetration depth as a function of porosity and is shown in Figure 13.30 [11]. Recently, progress has been made in using induction-hardening techniques, which combine two frequencies at the same time. The high frequency helps to achieve the desired level of surface hardness, while a mid-frequency helps to provide the desired depth of hardening.
13.12.4 NITROCARBURIZING
Nitrocarburizing is increasingly performed on P=M parts. Gaseous nitrocarburizing is a diffusion heat treatment that involves the addition of nitrogen and carbon to the surfaces of steel parts. It differs from carbonitriding mainly in that the temperatures used are lower, within the ferritic phase region, typically 570 to 6008C (1060 to 11108F). Since no phase transformations occur, the dimensional changes and distortion are significantly reduced. Nitrogen is diffused into the surfaces of the steel in sufficient concentration to form a thin layer of e-iron nitride on the surface of the part, producing a thin, hard, wear-resistant case. The effect of nitrocarburization on improving the fatigue strength of two P=M steels is shown in Figure 13.31. The nitrocarburizing process can be carried out in a conventional integral quench atmosphere furnace. The atmosphere usually consists of a 50:50 mixture of endothermic gas and anhydrous ammonia. The control of the nitrided layer thickness, as with other treatments, is dependent on density. If the nitrided layer is allowed to form on the internal pore surfaces to any significant extent, a volume expansion can occur. For this reason, density of the P=M part should be above 90% of the pore-free density. This nitrided layer can reduce the coefficient of friction and improve the wear resistance better than conventional quench-hardening treatments. The process is best used in applications where sliding wear and fretting are involved.
3 2. 8 2. 6 2. 4 Relative penetration depth 2. 2 2 1. 8 1. 6 1. 4 1. 2 1 0 0.02 0.04 0.06 0.08 Porosity (%) 0.1 0.12 0.14
FIGURE 13.30 Porosity can increase penetration depth during induction hardening. One calculation of the
relative penetration depth with porosity is shown. (From Bocchini, G.F., Advances in Powder Metallurgy and Particulate Materials2001, Vol. 6, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 5686; Bocchini, G.F., International Seminar on Principles and Methods of Engineering Design, Naples, Italy, 13 October 1997, pp. 239266.)
300
250
Nitrocarburized
200 Fatigue strength (MPa)
150 As-sintered 100
50
F-0000 FC-0205
0 1E + 05
1E + 06 Cycles
1E + 07
1E + 08
FIGURE 13.31 Effect of nitrocarburization on the notched fatigue strength of two low-carbon P=M steels
(7.0 g=cm3). (From Ferguson, H.A., Metals Progress, 7(6), 1975, 8183.)
As the hard-nitrided layer is relatively thin, the process should not be applied where high indentation or impact loading is involved. The e-nitride layer that is formed can attain a surface hardness in excess of HRC 60, depending on the alloy content of the steel. Indentation hardness testing is also not recommended when evaluating this process. Since no transformation occurs, the P=M parts can be air-cooled without loss of surface hardness. Also no oil absorption occurs, which leaves the porosity open for impregnation if desired.
13.12.5 NITRIDING
P=M parts can be nitrided using alkali salts, gases, or plasmas, with the majority of applications using gases. Alkali salts can be used in some applications, as the trapped salts may not be detrimental if the part is constantly immersed in oil during service. This disadvantage can be negated and increase in fatigue strength of 100% can be achieved [19]. Gas nitriding is most often done in ammonia or nitrogen-containing ammonia. Dissociation of ammonia in the presence of iron at temperatures above 4508C (8408F) can be used to nitride up to ~9508C (~17408F). However, normally gas nitriding is performed at temperatures below 5908C (10958F)as higher temperatures can lead to complete nitriding of the iron. Depths of 0.5 to 1 mm are achieved in the range of 540 to 5608C (10058F to 10408F) [20]. The change from closed porosity to open porosity that occurs when density drops below 92% dramatically changes the nitriding behavior of P=M parts. The e-phase can form internally if open porosity is present. Figure 13.32 [23] shows the fraction of e-phase formed with nitriding time at 5408C (10058F) with different densities. Open porosity can lead to nearly complete transformation to e-phase at these temperatures. Typically, lower density
100 5.2 g/cm3 80 6.1 g/cm3
Percentage of -phase
6.6 g/cm3 60
40
6.95 g/cm3 20
0 0 2 4 6 8 Nitriding time (h) 10 12 14 16
FIGURE 13.32 Effect of nitriding time of sintered sponge iron specimens of different porosity in ammonia at 5408C on the percentage of e-phase. (From Rudnayova, E., Salak, A., and Zabavnik, V., Zbornik Vedeckych Prac VST, Kosice, No. 1, 1975, p. 177.)
800
700 5.7 g/cm3 600 Microhardness (HV0.02)
500 6.8 g/cm3 400
300 7.2 g/cm3 200
100
0 0 1 2 3 4 5 6 7 8 9 10
Ammonia content (%)

FIGURE 13.33 Effect of density on gas nitrided hardness of sponge iron specimens in ammonia at 5408C.
(From Rudnayova, E., Salak, A., and Zabavniks, V., Zbornik Vedeckych Prac VST, Kosice, No.1, 1975, p. 177.)
parts will have higher nitrided hardness, as shown in Figure 13.33. The formation of e-phase inside pores can cause significant swelling, with the effect increasing with time and porosity (Figure 13.34). In order to nitride parts with densities below 92%, some form of pore-closing treatment is recommended.
13.12.6 STEAM TREATING

Steam treating is often done on P=M parts to both seal surface porosity and to increase the hardness of the surface. This tends to improve the corrosion resistance, pressure tightness, and wear resistance. Unfortunately, it also tends to reduce the tensile strength and ductility from 10 to 20%, depending on conditions and the steel chosen. Steam treatment usually consists of heating P=M parts in a steam atmosphere at elevated temperatures to form a layer of iron oxide, most often Fe3O4 at temperatures between 510 and 5708C (950 to 10608F). The oxide is formed inside the surface porosity as well, pinching off the opening due to the volume expansion associated with the conversion of the iron. In addition to sealing the surface the hardness of the surface is raised, as magnetite has a hardness equivalent to HRC 50. The recommended procedure for steam treating is as follows [21]: 1. Wash parts to remove any fluids that may have been absorbed from prior operations. 2. Load onto a fixture and place in a furnace preheated to 3158C (6008F). 3. Heat parts in air until the center of the load has stabilized as the set temperature
8 5.2 g/cm3 7
6 Dimensional change (%) 6.1 g/cm3
4 6.6 g/cm3 3
6.95 g/cm3
1 7.2 g/cm3 0 0 4 8 Nitriding time (h)

FIGURE 13.34 Dimensional change of sponge iron specimens exposed to Ammonia at 5408C (10008F).
12
16
Swelling increases with increasing porosity. (From Rudnayova, E. Salak, A., and Zabavnik, V., Zbornik Vedeckych Prac VST, Kosice, No.1, 1975, p. 177.)
4. Introduce superheated steam at a line pressure of 35 to 105 kPa (5 to 15 psi) and purge for at least 15 min 5. Increase furnace temperature to desired steam treatment temperature (between 510 and 5958C (950 to 11008F) and hold for no longer than 4 h at heat. 6. On completion of treatment, reduce furnace temperature to 3158C (6008F). When parts reach this temperature, the steam can be shutoff and the parts unloaded Caution should be exercised when opening the furnace after the treatment, as hydrogen gas, produced during the process, can ignite. It is recommended that a nitrogen purge be applied prior to unloading. The increase in density and apparent hardness produced by steam treating on various P=M steels is shown in Table 13.11 [10,22]. In steam-treated P=M steels, the ductility is significantly reduced due to the internal stresses created by the formation of the iron oxide. It is not recommended to steam treat steels with more than 0.5% carbon, due to the high internal stresses, initiating microcracks and causing extremely low ductility.
13.12.7 BLACK OXIDING

Black oxiding is used to improve the hardness of steel parts and provide an oxidationresistant coating that improves the appearance of the steel. It is used with parts for firearms, auto parts, turbines, bearings, and electrical parts. Black oxidizing is a chemical process, which uses a heated alkaline nitrate solution, typically 1408C. The black-oxidizing solution can become trapped in the pores in the steel
TABLE 13.11 Effects of Steam Treatment on Density and Hardness of P=M Steels
Density (g=cm3) Material F-0000-N F-0000-P F-0000-R F-0008-M F-0008-P F-0008-R FC-0700-N FC-0700-P FC-0700-R FC-0708-N FC-0708-P FC-0708-R Sintered 5.8 6.2 6.5 5.8 6.2 6.5 5.7 6.35 6.6 5.7 6.3 6.6 Steam-Treated 6.2 6.4 6.6 6.1 6.4 6.6 6.0 6.5 6.6 6.0 6.4 6.6 Sintered 7HRF 32HRF 45HRF 44HRB 58HRB 60HRB 14HRB 49HRB 58HRB 52HRB 72HRB 79HRB Apparent Hardness Steam-Treated 75HRB 61HRB 51HRB 100HRB 98HRB 97HRB 73HRB 78HRB 77HRB 97HRB 94HRB 93HRB
Source: From Pease, III, L.F., Collette, J.P., and Pease, D.A., Modern Developments in P=M, Vol. 21, Metals Powder Industries Federation, Princeton, NJ, 1988, p. 275.
part if the density is low enough. In addition, copper-containing steel parts do not black oxide well, leading to poor-quality surfaces.
REFERENCES
1. D.G. White, State of the North American P=M Industry2002, in Advances in Powder Metallurgy & Particulate Materials2002, Metals Powder Industries Federation, Princeton, NJ, 2002, part 1, pp. 112. 2. K. Barton, S. Das, J. Lu, M. Goldenberg, D. Olsen, and J. LaSalle, Advances in Powder Metallurgy and Particulate Materials2001, Vol. 4, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 159165. 3. MPIF Standard 35, Materials Standards for P=M Structural Parts, Metals Powder Industries Federation, Princeton, NJ, 2000. 4. S. Saritas, R.D. Doherty, and A. Lawley, Advances in Powder Metallurgy and Particulate Materials2001, Vol. 10, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 112 130. 5. D.H. Herring and P.T. Hansen, Key consideration in the heat treatment of ferrous P=M materials, in Proceedings of the 17th Heat Treating Conference, ASM International, Metals Park, OH, 1998, p. 213. 6. G.F. Bocchini, V. Fontanari, and A. Molinari, Friction effects in metal powder compaction; part one: theoretical aspects, in International Conference on Powder Metallurgy and Particulate Materials, Metals Powder Industries Federation, Princeton, NJ, 1995. 7. J.M. Capus and A. Maaref, Modern developments in powder metallurgy, in Proceedings of the 1973 International Powder Metallurgy Conference, Vol. 8, 1974, pp. 6173. 8. M.A. Pershing and H. Nandi, Advances in Powder Metallurgy and Particulate Materials2001, Vol. 5, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 2630. 9. W.B. James, Advances in Powder Metallurgy and Particulate Materials1998, Metals Powder Industries Federation, Princeton, NJ, pp. 125, 1998. 10. E.A. Brandes and G.B. Brook, Eds., Smithells Metals Reference Book, 7th ed., ButterworthHeinemann, Oxford, 1999.
11. G.F. Bocchini, Overview of surface treatment methods for PM parts, in Advances in Powder Metallurgy and Particulate Materials2001, Vol. 6, Metals Powder Industries Federation, Princeton, NJ, 2001, pp. 5686. 12. G.F. Bocchini, How to design sintered steels for advanced applications, in International Seminar on Principles and Methods of Engineering Design, Naples, Italy, 13 October 1997, pp. 239266. 13. P.F. Lindskog and G.F. Bocchini, Development of high strength P=M precision components in Europe, International Journal of Powder Metallurgy & Powder Technology, 15(3), 199, 1979. 14. H.E. Boyer, Ed., Case Hardening of Steels, ASM International, Metals Park, OH, 1985. 15. H.I. Sanderow and T. Prucher, Advances in Powder Metallurgy and Particulate Materials1994, Vol. 7, Metals Powder Industries Federation, Princeton, NJ, 1994, pp. 355366. 13. I.W. Donaldson and M.L. Marucci, Advances in Powder Metallurgy and Particulate Materials 2003, Vol. 7, Metals Powder Industries Federation, Princeton, NJ, 2003, pp. 230245. 17. G. Krauss, Microstructure, Residual Stresses, and Fatigue of Carburized Steels, 3 rd International Seminar on Quenching and Carburizing IFHTSE and IMM, 1991, pp. 201217. 18. C.R. Brooks, Principles of the Surface Treatment of Steels, Technomic Publications, Lancaster, PA, 1984, pp. 7880. 19. H. Krzyminski, Harterei Technische Mitteiluugen., No. 1, 1971, p. 2. 20. A. Salek, Ferrous Powder Metallurgy, Cambridge International, Cambridge, 1995, p. 300. 21. H. Ferguson, Heat treatment of ferrous powder metallurgy parts, in Powder Metal Technologies and Applications, ASM International, Metals Park, OH, 1998, pp. 645655. 22. L.F. Pease, III, J.P. Collette, and D.A. Pease, Mechanical properties of steam-blackened P=M materials, in Modern Developments in P=M, Vol. 21, Metals Powder Industries Federation, Princeton, NJ, 1988, p. 275. 23. E. Rudnayova, A. Salak, and V. Zabavnik, Zbornik Vedeckych Prac VST, Kosice, No. 1, 1975, p. 177. 24. H. Ferguson, Heat treatment of P=M parts, Metal Progress, ASM, 107(6), 8183, June 1975; 108(2), 6669, July 1975.

Powder Components Heat Treatment PDF

Uploaded by

Copyright:

Available Formats

Powder Components Heat Treatment PDF

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Powder Components Heat Treatment PDF

Uploaded by

Copyright:

Available Formats

13

Heat Treatment of Powder Metallurgy Steel Components

2006 by Taylor & Francis Group, LLC.

13.2 OVERVIEW OF P=M PROCESSING

2006 by Taylor & Francis Group, LLC.

TABLE 13.1 Comparison of Powder Metallurgy Techniques Discussed

2006 by Taylor & Francis Group, LLC.

Elemental or alloy metal powders

Additives (lubricants or binders)

Elemental or alloy metal

Optional secondary manufacturing

Optional secondary finishing

Finished P/M product

2006 by Taylor & Francis Group, LLC.

2006 by Taylor & Francis Group, LLC.

2006 by Taylor & Francis Group, LLC.

FIGURE 13.3 737SH sinter-hardened. Unetched.

2006 by Taylor & Francis Group, LLC.

13.2.2 METAL INJECTION MOLDING

2006 by Taylor & Francis Group, LLC.

Fine metal powder

Mixing and pelletizing

Optional finishing operations

13.2.3 POWDER FORGING

2006 by Taylor & Francis Group, LLC.

13.3 DESIGNATION SYSTEM FOR P=M STEELS

TABLE 13.2 Letter Code for Designating P=M Alloys

2006 by Taylor & Francis Group, LLC.

13.4 OVERVIEW OF HEAT TREATMENT

2006 by Taylor & Francis Group, LLC.

TABLE 13.3 Compositions of Common P=M Steels

Up to 0.5% Mn may be present depending on starting powder.

2006 by Taylor & Francis Group, LLC.

2006 by Taylor & Francis Group, LLC.

13.5 EFFECT OF POROSITY ON THE HEAT TREATMENT OF P=M STEELS

2006 by Taylor & Francis Group, LLC.

Apparent hardness (HRA)

70 7.1 g/cm3 60 6.8 g/cm3 6.4 g/cm3 50

2006 by Taylor & Francis Group, LLC.

0.95 Relative cooling speed

7.2 Density (g/cm3)

2006 by Taylor & Francis Group, LLC.

Local porosity (%)

2006 by Taylor & Francis Group, LLC.

13.6 EFFECT OF ALLOY CONTENT ON P=M HARDENABILITY

2006 by Taylor & Francis Group, LLC.

500 110 kg load 400

300 6.6 6.8 7 Density (g/cm3) 7.2 7.4

2006 by Taylor & Francis Group, LLC.

13.6.1 COPPER CONTENT

13.6.2 NICKEL CONTENT

13.6.3 NICKELCOPPER CONTENT

2006 by Taylor & Francis Group, LLC.

45 0.3 0.5 40 0.7 0.9 35

2006 by Taylor & Francis Group, LLC.

Heat-treated hardness (Rc)

Heat-treated Sintered hardness (Rb) 90

13.6.4 MOLYBDENUM CONTENT