of the available volume inside the container, a considerable amount of shrinkage occurs The science/art of designing a container must account for the packing density and the symmetry or lack thereof to achieve an acceptable part During the initial production of a particular component, preproduction trials and/or iterations of the full-size shape may be necessary to determine the shrinkages empirically However, this iterative approach is frequently costly and time consuming Through the years, HIP P/M part manufacturers have employed engineering intuition and previous experience to develop the starting can design At this time, other approaches are being developed and used, namely, empirical and continuum mechanics/finite element modeling Some of these are briefly described here Empirical Models A large percentage of the HIP P/M compacts produced are either simple or hollow cylinders An empirical model for these shapes was developed several years ago (Ref 26) by analyzing dimensional data from before and after HIP for a variety of cylindrical shapes To eliminate the effects of the HIP cycle, alloy systems, and container thickness, the analysis focuses on cylinders made from nickel-base alloys consolidated in similar HIP cycles with a certain can thickness range Best fit curves were generated for axial and radial shrinkage as a function of aspect ratio (length to diameter) and surface area ratio (area of cylindrical component to area of the lateral ends) as shown in Fig 13 Based on these data, a computer program was generated to provide either the starting container dimensions to make a finished near-net shape or predict post-HIP dimensions given a specified starting container (Ref 26) Fig 13 Normalized shrinkage on solid cylinders (normalized shrinkage = actual shrinkage/isotropic shrinkage) Source: Ref 20 Engineering Models One promising approach to perform a direct process simulation via the use of computer models is to combine a constitutive model of continuum mechanics equations solved by a finite element computational method Several approaches ranging from simple plastic to compressible, viscoplastic constitutive models have been investigated as described in the article "Principles and Process Modeling of Higher-Density Consolidation" in this Volume Even though some of these mathematical models are utilized in production, none has matured into a reliable modeling system for arbitrary geometries and HIP cycles Container Fabrication Tooling and Container Component Fabrication Once the design has been established, the metal container is fabricated This is not a trivial step because the container must be producible in an economical fashion or the finished part cannot be manufactured The most economical and easily formed container material is low-carbon steel; however, other materials (e.g., stainless steel, nickel alloys, titanium, etc.) can also be used The process is constrained by existing metalforming techniques (e.g., metal spinning, hydroforming, stamping, hand forming, casting, machining, etc.) with each having its inherent advantages and limitations Tooling for the HIP P/M process refers to that which may be used to fabricate the container components Usually, the quantity of parts to be made dictates the precision and cost committed to the tooling for HIP P/M containers Large numbers of parts (1000 or more) would employ stamped containers Anything less than this would be determined on a case-by-case basis Cost of tooling must be amortized over the quantity of parts produced, so more expensive and more precise tooling can be cost effective only for sizable production runs Because HIP P/M is most often used as a near-net shape process with small quantities of parts, container tooling is not made to be as precise as other processes where net shape is important Cleaning Contamination of encapsulated powders will result, unless dirt, oxides, metalworking lubricants, and rust preventatives used to fabricate container components are removed See Surface Engineering, Volume 5 of the ASM Handbook for cleaning procedures applicable to various metals Proper cleaning, storage, and handling procedures immediately prior to any welding operation are necessary to prevent dirt entrapment or contamination on can surfaces Powder metallurgy alloys that are particularly sensitive to contamination (titanium, nickel-base alloys, and refractory metals) require controlled humidity and stringent cleanliness for final can preparation, assembly, welding, and filling Electropolishing of stainless steel container components and nonchlorinated solvent cleaning (usually acetone, methylethylketone, or methanol) of titanium container components represent typical cleaning processes for specialized applications Carbon steel sheet metal containers should be supported carefully to prevent distortion during welding Similar precautions are recommended for the outer sheet metal container used in a ceramic mold process Welding A matched-weld-lip container configuration is designed to promote directional plane front solidification with good liquid metal feeding in the solidifying weld metal Certain oxides (iron, nickel, and copper) can be reduced at high temperature in a high-pressure argon environment This process may produce leaky containers during HIP if oxides extend through weld metal or container wall materials Use of stainless steel filler metal for carbon steel container repairs is recommended because chromium oxides essentially are stable under processing conditions up to 1200 °C (2200 °F) in argon Gas tungsten arc welding of nickel containers with stainless steel filler metal is also advised Containers for loose powder are assembled, welded, leak tested, and filled in sequence Containers with interior spacers (mandrels), powder/solid composites (e.g., clad components), or precompacted and sintered P/M compacts can be filled with at least one cover removed This procedure results in an extensive assembly weld area that cannot be leak tested in the vacuum mode because of the slow response time of helium through the interior of the filled container Consequently, careful removal of loose powder from the weld area is necessary and use of precision and reproducible (preferably machine) weld techniques is required to prevent leakage Leak tightness of HIP containers is a major process consideration Electron beam, gas tungsten arc welding, and stick welding are used for final container assembly Argon dry box and electron beam welding are used for titanium alloy containers because nonoxidizing conditions are required Gas tungsten arc welding with and without filler is used for carbon steel and stainless steel containers Carbon steel containers may require a final reduction anneal after weld assembly, and some clad parts may need to be preheated prior to welding because of substrate material considerations or section size differences Because weld metal is essentially a solidified casting, shrinkage and gas porosity are the fundamental causes of leakage at welds Leak Testing Containerized HIP of metal powders can only be achieved successfully with leak-free containers Location of leaks in a fully assembled container by use of valid leak-testing procedures and subsequent repair are fundamental requirements of HIP P/M technology Leak detection is based on characteristics of helium and argon flow through small capillaries when compared at 1 atm (0.1 MPa) and 1000 atm (100 MPa) total pressure Flow characteristics of a cylindrical capillary have been described by Guthrie and Wakerling (Ref 27): (Eq 1) Q = 1/L [C1P2 + C2P + C3 ln where Q is the flow rate, cgs units; P1 and P2 are the exterior and interior pressure, cgs units; C1, C2, C3, and C4 are constants; and L is the capillary length, cm For P2 = 0 (evacuated container interior) and P1 large: Q= = L (Eq 2) where is the gas viscosity and D is capillary diameter, both in cgs units This applies strictly in the viscous flow region, when Reynolds number (Re) is 445 kN (50 tons) usually require pits Top-drive and bottom-drive presses are comparable in terms of partmaking capability, reliability, and equipment cost Press Tonnage and Stroke Capacity Required press capacity to produce compacts in rigid dies at a given pressure depends on the size of the part to be pressed and the desired green density of the part, which in turn is determined by requirements for mechanical and physical properties of the sintered part Compacting pressures can be as low as 70 to 140 MPa (5 to 10 tsi) for tungsten powder compacts or as high as 550 to 830 MPa (40 to 60 tsi) for high-density steel parts When a part is pressed from the top and bottom simultaneously, the press should apply the required load to the upper and lower ram of the press To eject the pressed compact, an ejection capacity must be available that is sometimes divided into the load for the breakaway stroke (which is the first 1 to 12 mm ( to in.) of the ejection stroke and the load for a sustained stroke) The load for a sustained stroke is generally one-fourth to one-half of the breakaway load The stroke capacity of a press, or the maximum ram travel, determines the length of a part that can be pressed and ejected In presses used for automatic compacting, the stroke capacity is related to the length available for die fill and ejection stroke Load Requirements The total load required for a part is determined by the product of the pressure needed to compact the part to the required density and the projected area of the part Compaction curves relate pressure, P, to the required density, q, and are usually obtained from compacting tests on cylindrical shapes with the height, L, equal to the diameter, D For thicker parts the load must be increased, by as much as 25% for a length to diameter ratio of 4 to 1, to give the required density Required compacting pressures can be estimated with a correction factor, k, such that (Ref 1): P = P1 (1 + k) where P is the compaction pressure for a larger part and P1 is the compaction pressure for a "standard" part (i.e., L = D) The correction factor is: k = (0.25/3)(L/D - 1) for L/D >1 k = 0 for L/D < 1 For parts that are not cylindrical, an equivalent L/D ratio can be used: Le/De = (V · p)/(2 · A2) where V is the part volume and A is the projected area The press load required is then obtained by multiplying the required compaction pressure by the projected area of the part Reference cited in this section 1 W.A Knight, Design for Manufacture Analysis: Early Estimates of Tool Costs for Sintered Parts, Annals of the CIRP, Vol 40 (No 1), 1991, p 131-134 Powder Metallurgy Presses and Tooling Revised by John Porter, Cincinnati Incorporated Mechanical Presses In most mechanical P/M compacting presses, electric motor-driven flywheels supply the main source of energy used for compacting and ejecting the part The flywheel normally is mounted on a high-speed shaft and rotates continuously A clutch and a brake mounted on the flywheel shaft initiate and stop the press stroke To initiate a press stroke, the brake is disengaged and the clutch is engaged, causing the energy stored in the rotating flywheel to transmit torque through the press gearing to the final drive or press ram Clutch and brake systems should be of the partial revolution type that can be engaged and disengaged at any point in the pressing cycle The clutch usually is pneumatically engaged with a spring release, and the brake is pneumatically released with a spring set, thereby providing full stopping ability in the event of loss of air pressure An adjustable speed device normally is supplied with electric drive motor, providing production rate adjustment as indicated by pressing and ejection conditions On presses that have main motor capacities up to 19 kW (25 hp), the adjustable speed drive is usually of the variablepitch pulley or traction-drive type Above 19 kW (25 hp), direct-current or eddy-current control devices are preferred The motor and drive must be totally enclosed to prevent contamination by metal powder dust Gearing systems usually are either single-reduction (Fig 1) or double-reduction (Fig 2) arrangements Single-reduction gearing frequently is used in lower tonnage presses ( 445 kN, or 50 tons) that have stroking rates of 50 strokes/min Higher tonnage presses use double-reduction gearing and commonly have maximum stroking rates of 30 strokes/min Fig 1 Single-reduction gearing systems for P/M compacting press Fig 2 Double-reduction gearing systems for P/M compacting press The low-speed shaft of the press, normally called the main shaft, is linked to the press ram, causing motion of the tooling for the compacting and ejection cycles Ram driving mechanisms can be either cam- or eccentric-driven arrangements Cam-driven presses generally are limited to pressing capacities 890 kN (100 tons) The main shaft of the press has two cams one cam operates the upper ram, and the other cam operates the lower ram for compacting the part The cam that operates the lower ram also controls the powder feed into the die and ejects the part from the die after compacting Cams normally operate linkages that convert the main shaft rotary motion into the linear motion of the tooling Figure 3 shows a schematic of a cam-driven press The cams in this type of press can be adjusted or arranged with removable sections, thus allowing cam motion to be varied to produce special motions to compact the part Pressure can be applied either simultaneously or sequentially to the top and bottom of the compact Anvil and rotary presses are types of cam-driven machines These presses are described in more detail later in this article Fig 3 Schematic of cam-driven compacting press Eccentric-Driven Presses Presses that have a final drive mechanism consisting of an eccentric or crank on the main shaft are the most widely used type of mechanical press A connecting rod is used to convert the rotary motion of the main shaft into the reciprocating motion of the press ram Generally, an adjustment mechanism is built into the connecting rod or press ram assembly, thus permitting the height position of the press ram to be changed with respect to the main shaft or press frame, thereby controlling the final pressing position of the ram This adjustment mechanism can be used to control the length of the compacted part Standard eccentric-driven presses have pressing capacities ranging from 6.7 to 7830 kN (0.75 to 880 tons) Powder Metallurgy Presses and Tooling Revised by John Porter, Cincinnati Incorporated Hydraulic Presses Hydraulically driven compacting presses are available with pressing capacities ranging from 445 to 11,100 kN (50 to 1250 tons) as standard production machines, although special machines with capacities 44,500 kN (5000 tons) have been used in production Hydraulic presses normally can produce longer parts in the direction of pressing than mechanical presses, and longer stroke hydraulic machines are less expensive compared to an equivalent stroke produced in a mechanical press The maximum depth of powder fill in mechanical presses is 180 mm (7 in.), while 380 mm (15 in.) of powder fill is common in hydraulic presses The maximum production rate for hydraulic presses producing a single part per stroke is 650 pieces per hour The slower speed of a hydraulic press when pressing long parts is preferable, because the longer time during pressing permits trapped air within the powder to escape through the tooling clearances Most hydraulic presses are considered top-drive machines because the main operating cylinder is centrally located in the top of the press This main cylinder provides the force for compacting the part Hydraulic presses have three distinct downward speeds: • • Rapid advance: Produces minimal pressing force, used for rapid closing of the die cavity Medium speed: Pressing capacities 50% of full-rated capacity, used during initial compaction when lower pressing force is required • Slow speed: Maximum capacity available for final compaction Two types of hydraulic pumping systems are commonly found in P/M presses: the high-low system and the filling circuit system The high-low system has a double-acting main cylinder A regenerative circuit is used for rapid approach Initially, the piston of the cylinder is activated by a high-volume, low-pressure pump; the fluid from the bottom of the cylinder is directed into the top of the cylinder in addition to the low-pressure pump volume At medium speed, the regenerative circuit is deactivated, while the piston remains activated by the low-pressure pump In full-tonnage press, the low-pressure pump is deactivated, and a high-pressure pump activates the piston The filling circuit hydraulic pumping system has a single-acting main cylinder, and ram motion is controlled by small double-acting cylinders The ram control cylinders are smaller than the main cylinder, so only a low flow rate of fluid is needed to cause rapid movement of the ram During approach and return cycles, however, the fluid flow rate into and out of the main cylinder is high The main cylinder is fitted with a large two-way valve that allows fluid to flow at low pressures (usually gravity feed) During pressing, the two-way valve is closed, and high pressure from the pump is applied to the main cylinder piston Ejection of the part usually is accomplished by a cylinder that is centrally located in the bed of the press The cylinder either upwardly ejects the part or pulls the die downward from the part, depending on the type of tooling used When pressing parts to a given thickness, positive mechanical stops are used on hydraulic presses to control downward ram movement When pressing parts to a desired density, downward ram movement is controlled by adjustment of the pressure to the cylinder When the part is pressed to the desired unit pressure, the press ram stops and returns to the retracted position Some types of P/M materials, such as P/M friction materials, are always pressed to density rather than size, because uniform density provides uniform friction and wear properties The drive-motor horsepower on a hydraulic press is considerably larger than on an equivalent mechanical press A mechanical press has a flywheel from which energy is taken during the pressing and ejection of the part Energy is restored to the flywheel during the die feeding portion of the cycle The motor on a hydraulic press must supply energy directly during the pressing and ejection portion of the cycle Powder Metallurgy Presses and Tooling Revised by John Porter, Cincinnati Incorporated Comparison of Mechanical and Hydraulic Presses In terms of partmaking capability, no distinct advantage is gained by using either a mechanical press or a hydraulic press Any part can be produced to the same quality on either type of machine However, the following parameters influence press drive selection Production Rate A mechanical press produces parts at a rate one and one-half to five times that of a hydraulic press as a result of inherent design of the energy transfer systems and stroke length Operating cost of a hydraulic press is higher, because the total connected horsepower of a hydraulic press is one and one-half to two times that of an equivalent mechanical machine Theoretically, the required energy to compact and eject a part is the same for a hydraulic or a mechanical press, except that the overall efficiency of a mechanical press is slightly higher than that of a hydraulic press Also, the kilowatt usage of the larger motor on a hydraulic press is greater than that of a mechanical press during the idle portion of the machine cycle Machine overload protection is an inherent feature of a hydraulic press If the hydraulic system is operating properly, the machine cannot create a force greater than the rated capacity Consequently, overload of the machine frame is not possible, even if a double hit or operator error occurs in adjusting the machine Misadjustment or double hits can cause a mechanical press to overload, can damage the machine, or may cause tooling overload and failure if the tooling cannot withstand full machine capacity Some new mechanical presses are equipped with hydraulic overload protection systems Equipment cost of a hydraulic press generally is one-half to three-quarters that of an equivalent mechanical press Facility, foundation, installation, and floor space costs generally are comparable Die Sets The mounting into which the tooling is installed is known as the die set Generally, the die set must be well guided because of the close tooling clearances used Guide bearings must be protected with boots or wipers to prevent powder particles from entering guiding surfaces Tooling support team members should have high stiffness to minimize deflection The die set must be free of residual magnetism The maximum acceptable level is 2 G To ensure press operator safety, die sets should be adequately guarded In a complex tooling arrangement, as many as seven independent tooling members and supports are moving relative to one another during the pressing and ejection cycles Die sets can be classified as removable or nonremovable Both types are used in mechanical and hydraulic presses Nonremovable die sets are used throughout the entire tonnage requirements of available presses Manually removable die sets are used primarily in presses with pressing capacities up to 2670 kN (300 tons) Above this press size, the die set assembly is moved by a powered system, and removable die set presses with capacities of 17,800 kN (2000 tons) are available The major advantage offered by nonremovable die sets is flexibility in setup and operation Presses equipped with nonremovable die sets usually have all adjustments required for setup and operation built into the press and die set, including: • • • Part length adjustment: Any dimensions of the part in the direction of pressing can be quickly changed during production Part weight: Material weight in any level of the part can be changed easily during production Tooling length adjustment: Adjustments are provided to accommodate shortening of punch length due to sharpening or refacing Another advantage of nonremovable die sets is the greater space available for tooling, compared to the removable type This space provides more freedom in tooling design However, presses incorporating nonremovable die sets must be shut down during tooling changes or maintenance Tooling change and setup time generally is from 1 to 4 hours but sometimes substantially longer, depending on the complexity of tooling Nonremovable die sets are well suited for developing new P/M parts, because press and tooling adjustments can be made quickly to achieve the desired weight, density, and part dimension Adjustment features of nonremovable die sets make them desirable on long production runs, where changes in powder quality among lots require frequent tooling adjustment to maintain part quality Users of removable die sets normally have two or more die sets per press Tooling can be set up in a spare die outside the press Removable die sets normally can be changed in less than 30 min, so loss of production time is minimal On small presses where the die set is also small, the die set is restricted to a given set of tools and is considered semidurable tooling One disadvantage of many removable die sets is that pressing is controlled by pairs of pressing blocks made of hardened tool steel, such as D-2 The height of the pressing block controls the height of the part If the part length dimension is changed due to design, or if the tooling length is changed due to repair, the pressing blocks must be changed accordingly Removable die sets are ideally suited for shorter production runs On newer presses with removable die sets, complete powder adjustment is available, even when the die set is outside the machine Powder Metallurgy Presses and Tooling Revised by John Porter, Cincinnati Incorporated Part Classification The Metal Powder Industries Federation has classified P/M parts according to complexity Class I parts are the least complex, and class IV parts are the most complex To better understand the types of commercially available P/M compacting presses, and their advantages and limitations, an understanding of P/M part classification and tooling systems used to produce parts is necessary Part thickness and number of distinct levels perpendicular to the direction of powder pressing determine classification not the contour of the part Class I parts are single-level parts that are pressed from one direction, top or bottom, and that have a slight density variation within the part in the direction of pressing (Fig 4a) The highest part density is at the surface in contact with the moving punch, and the lowest density is at the opposite surface Parts with a finished thickness of 7.5 mm (0.3 in.) can be produced by this method without significant density variation Fig 4 Basic geometries of (a) MPIF class I (simple) and (b) MPIF class IV (complex) parts Class II parts are single-level parts of any thickness pressed from both top and bottom The lowest density region of these parts is near the center, with higher density at the top and bottom surfaces Class III parts have two levels, are of any thickness, and are pressed from both top and bottom Individual punches are required for each of the levels to control powder fill and density Class IV parts are multilevel parts of any thickness, pressed from both top and bottom (Fig 4b) Individual punches are required for each level to control powder fill and density Powder Metallurgy Presses and Tooling Revised by John Porter, Cincinnati Incorporated Shape of Rigid Tooling Rigid tool compaction differs from roll compaction, isostatic compaction, hot isostatic pressing, and injection molding in that a quantity of powder (fill) is confined in a rigid die cavity at ambient temperature The die cavity is entered by one or more punches, which apply compaction pressure to the fill powder As a result of the compaction pressure, the fill powder densifies, develops green strength, and assumes the exact shape of the die cavity and punch faces Following the pressure cycle, the shaped powder fill, now a piece part, is ejected (stripped) from the die cavity The physical size of parts made in rigid tool compaction systems is a function of press tonnage capacity, fill depth, and also the length of a green powder fill that can be effectively compacted in terms of a maximum density variance Parts vary in size from those weighing 1 g (0.035 oz) that are made in presses with capacities as small as 35 kN (4 tons) to those weighing 10 kg (22 lb) that are made in presses with capacities of 8900 kN (100 tons) Rigid tools must also be constructed oversize, with exact linear dimensions, to compensate for the final volume change Although theoretical computations are useful, most successful rigid tool sets are based on shrinkage allowances developed from existing tooling and the dimensional histograms developed for particular powders However, shrinkage allowances can be complex depending on subsequent sintering and binder additives For example, some metallic powders, such as the carbide and tool steel types, and some gas and centrifugally atomized specialty powders, such as spray-dried tungsten carbide, do not develop significant green strength, because their individual particles are predominantly spherical or they lack plasticity To compact such powders in rigid tool systems, wax or wax-stearate binders are added, which can occupy up to 20 vol% of the green compacted shape The development of full metallic properties during sintering also requires a volume shrinkage Powder Metallurgy Presses and Tooling Revised by John Porter, Cincinnati Incorporated Powder Fill The important consideration in P/M part production is the fill ratio required to produce parts to a density that is compatible with end use requirements The fill ratios must remain constant for a given part to maintain dimensional reproducibility Parts can be of single-level or multilevel design Single-level parts, designated as class I by the Metal Powder Industries Federation (MPIF), present the least difficulty to the tool designer, regardless of the size or part configuration The main consideration is designing a die that is long enough guidance for the lower punch (usually 25 mm, or 1 in.) and providing adequate fill depth for compacting the powder to the required density This challenge, coupled with the primary mechanical consideration of locating the center of mass in the press center, provides the best potential for producing a uniform quality part Figure 4(a) shows basic geometries of MPIF class I parts Multilevel parts, with industry classifications II through IV, present two additional complications to the tool designer: powder fill and part ejection Because metal powders tend to compact in vertical columns and generate little hydraulic flow, the tool designer must create fill levels in the tools that compensate for the thickness variations present in the final part configuration Uniform density, neutral axis of compaction, and part ejection should be considered to determine the need to vary fill levels and the manner in which these variations are achieved Excessive density variations contribute to green cracks and sintered distortion A common method of varying fill levels is by using multiple lower punches, which are timed to react to one another either through the use of springs or air, or by mounting on separate press platens Other methods are less effective, because punches are not adjustable and are fixed on one of the tool members, such as the die or core rod Fixed levels are commonly referred to as die chokes, core rod steps, or splash pockets (Fig 5) Fixed fills are sensitive to the apparent density of the material being compacted In operations that control compacting pressure, such as in hydraulic pressing, fixed fills cause dimensional variations in part thickness Because mechanical presses are set to operate to a fixed position relative to the die, the variation created by the apparent density of the powder causes overdensification or underdensification, resulting in a corresponding oversize or undersize peripheral area on the part Green expansion occurs as a part is stripped from the die Ideally, the part returns to die size through shrinkage during sintering Fig 5 Methods of achieving fixed fill levels (a) Fixed fill on an upper level using a step die (b) Fixed fill using a splash pocket to permit a projection feature on an upper punch (c) Stepped core rod forming an internal shoulder When a part has more than one level in the compacting direction, the step height should be limited to one-quarter of the overall height for a single punch (Fig 6a) If a larger step is required, multiple punches should be considered (Fig 6b) Fig 6 Two-level compaction (a) Single lower punch when h H/4 (b) Double lower punches when h > H/4 Fill Height The fill height is the depth of the loose powder required to give the required part thickness after compaction The value is determined by the compressibility of the loose powder at the required density The fill height, hf, is obtained by multiplying the finished part height by the compression ratio of the powder: ... (100 0 atm) 0.001 0.01 0.1 1.0 3.8 × 1 0-1 6 3.8 × 1 0-1 2 3.8 × 1 0-8 3.8 × 1 0-4 3.8 × 1 0-1 4 3.8 × 1 0-1 0 3.8 × 1 0-6 3.8 × 1 0-2 3.3 × 1 0-1 1 3.3 × 1 0-7 3.3 × 1 0-3 3.3 × 1 0-1 Note: Leakage rate is inversely... flow, cm3 diameter (D), m at a leakage time of: 100 0 s 10, 000 s 0.001 0.01 0.1 3.8 × 1 0-8 3.8 × 1 0-4 3.8 × 1 0-0 3.8 × 1 0-7 3.8 × 1 0-3 3.8 × 1 0-1 Note: Leakage flow is inversely proportional to... on solid-to-solid, powder- to-solid, and in some cases, powder- topowder surfaces As with powder/ metal container combinations, material compatibility must be evaluated to ensure no low-temperature