Volume 07 - Powder Metal Technologies and Applications Part 12 pdf

Sutherland, Statistical Quality Design and Control, Macmillan, 1992 Planning and Quality Control of Powder Metallurgy Parts Production Jack R.. Planning and Quality Control of Powder M

capability index The first measure, Cp, is based on the relationship between the upper and lower specification limits and the standard deviation process as follows:

The minimum acceptable value for Cp is considered to be one

However, because it is possible to achieve a Cp of greater than 1.0 while still producing a significant proportion of parts that

do not meet specification limits, a revised version of the process capability index Cpk, has been widely adopted This index, sometimes called the "actual process capability index," imposes a penalty for deviation of the process mean from the nominal

value of the specification The minimum value considered acceptable for Cpk is usually 1.33, although some industries are now beginning to demand 2.0 or better

The Cpk index is defined as the difference between the process mean and its closest specification limit, divided by 3 To calculate this index, first the relationship between the process mean and the specification limits in the units of standard deviations is determined:

Then the minimum of these two values is selected:

Zmin = min[ZUSL, -ZLSL]

The Cpk index is then defined by dividing this minimum value by 3:

Commonly, Cpk must be 1.33, but higher limits are also specified

For example, Fig 4 shows a process capability study of the overall length of a fully processed bevel gear (as heat-treated)

The Cp of 1.75 shows that the inherent process variation is less than the maximum acceptable range for the product The Cpk

of 1.71 demonstrates that the process is properly targeted to produce an adequately centered finished product dimensional distribution

Fig 4 Process capability for a heat treated bevel gear

Rational Sampling

Perhaps the most crucial issue to the successful use of the Shewhart control chart concept is the definition and collection of the samples or subgroups This section discusses the concept of rational sampling, sample size, sampling frequency, and sample collection methods and reviews some classic misapplications of rational sampling

Rational subgroups or samples are collections of individual measurements whose variation is attributable only to one unique constant system of common causes.In the development and continuing use of control charts, subgroups or samples should be chosen in a way that provides the maximum opportunity for the measurements within each subgroup to be alike and the maximum chance for the subgroups to differ from one another if special causes arise between subgroups

Sample Size and Sampling Frequency Considerations. The size of the rational sample is governed by the following considerations:

• Subgroups should be subject to common cause variation The sample size should be small to minimize the chance of mixing data within one sample from a controlled process and one that is out of control This generally means that consecutive sample selection should be used rather than distributing the sample selection over a period of time There area, however, certain situations where distributed sampling may be preferred

• Subgroups should ensure the presence of a normal distribution for the sample means In general, the larger the sample size, the better the distribution is represented by the normal curve In practice, sample sizes of three or more ensure a good approximation to normality

• Subgroups should ensure good sensitivity to the detection of assignable causes The larger the sample size, the more likely that a shift of a given magnitude will be detected

When the above factors are taken into consideration, a sample/subgroup size of three to six is likely to emerge Five is the most commonly used number because of the relative ease of further computation

Sampling Frequency. The question of how frequently samples should be collected is one that requires careful thought In many applications of and R control charts, samples are selected too infrequently to be of much use in identifying and

solving problems Some considerations in sample frequency determination are the following:

• If the process under study has not been charted before and appears to exhibit somewhat erratic behavior, samples should be taken quite frequently to increase the opportunity to quickly identify improvement opportunities As the process exhibits less and less erratic behavior, the sample interval can be lengthened

• It is important to identify and consider the frequency with which occurrences are taking place in the process This might include, for example, ambient condition fluctuations, raw material changes, and process adjustments such as tool changes or wheel dressings If the opportunity for special causes to occur over a 15-min period is good, sampling twice a shift is likely to be of little value

• Although it is dangerous to overemphasize the cost of sampling in the short term, clearly it cannot be neglected

Common Pitfalls in Subgroup Selection. In many situations, it is inviting to combine the output of several parallel and assumed-to-be-identical machines into a single sample to be used in maintaining a single control chart for the process Two variations of this approach can be particularly troublesome: stratification and mixing

Stratification of the Sample. Here each machine contributes equally to the composition of the sample For example, one

measurement each from four parallel machines yields a sample/ subgroup of n = 4 In this case, there will be a tremendous

opportunity for special causes (true differences among the machine) to occur within subgroups

When serious problems do arise, for example, for one or more of the machines, they will be very difficult to detect because of the use of stratified samples This problem can be detected, however, because of the unusual nature of the -chart pattern (recall the previous pattern analysis) and can be rectified provided the concepts of rational sampling are understood

The R-charts developed from such data will usually show very good control The corresponding control chart will show very wide limits relative to the plotted values, and their control will therefore appear almost too good The wide limits result from the fact that the variability within subgroups is likely to be subject to more than merely common causes

Mixing Production from Several Machines. Often it is inviting to combine the output of several parallel machines/lines into a single stream of well-mixed product that is then sampled for the purposes of maintaining control charts

If every sample has exactly one data point from each machine, the result would be the same as that of stratified sampling If the sample size is smaller than the number of machines with different means or if most samples do not include data from all machines, the within-sample variability will be too low, and the between-sample differences in the means tend to be large Thus, the -chart would given an appearance that the values are too far away from the centerline

References cited in this section

1 W.A Levinson, Make the Most of Control Charts, Chem Eng Prog., Vol 88 (No 3), March 1992, p 86-91

2 R.E DeVor, T.H Chang, and J.W Sutherland, Statistical Quality Design and Control, Macmillan, 1992

Planning and Quality Control of Powder Metallurgy Parts Production

Jack R Bonsky, Cleveland State University, Advanced Manufacturing Center

P/M Process Planning

The first step in quality planning is a dialogue between engineering teams of the P/M part producer and the customer At this stage, finished product (generally an assembled component) can be evaluated, and preliminary part prints may be exchanged While the product is still "on paper" relatively inexpensive design changes should be considered to accommodate the P/M process and lower overall cost Design factors in P/M manufacturing are addressed elsewhere in this Volume

After final part prints have been agreed to, the P/M manufacturing process can be engineered Several of the key factors include tooling dimensions, sintering parameters, and secondary operations such as heat treating and machining

Dimensional and Tool Size Determination

When determining tooling sizes and process control limits, one must begin with a consideration of the final product, then work backward through the process As parts move from compaction, to sintering, to secondary operations, the sizes usually change significantly Additionally, the variation generally increases with each step in the process (Fig 5) Notable exceptions are sized or machined parts, as these secondary processes are intended to improve tolerances and thus reduce variation

Fig 5 Typical change in size distribution after sintering and heat treatment of a P/M compact The distribution

widens after additional processing steps

The tool designer must draw upon material and processing knowledge to determine the size of the compaction tools Commonly, the tooling designer begins with the finished part print and then "factors" the dimensions of all tooling members For example, for a part with a 1.000 in outside diameter and 0.800 in inside diameter that shrinks through processing from compacting tooling size, an appropriate factor may be 1.005 The designer thus would size the die at 1.005 in (1.000 in × 1.005) and the core 0.804 in (0.800 × 1.005) as shown schematically in Fig 6

Fig 6 Example of using tooling factor for die sizing Factor used: 1.005; tooling not shown: upper punch, lower

punch, various adapters, and so forth

The determination of the appropriate factor depends on the material, the processes used, and the parameters of the processes used Powder manufacturers offer a great deal of baseline data for tooling size determination, but the best method is to draw upon previously tooled parts of similar processing and geometric configuration When this previously generated data are not available, a pilot run using prototype tooling is an appropriate method of determining sizes This reduces the likelihood of having to retool production tooling and can decrease the lead time for the first production run

Complicated geometries can require the use of different factors for different portions of the tooling For example, raised hubs formed by the top punch tend to have lower densities than the balance of the part Lower densities have a tendency to shrink more, which means that the factor used should be higher than for the rest of the part geometry

Sintering Parameters

Densification for a metallurgically sound part by sintering is determined by sintering temperature, time at temperature, and atmosphere composition To ensure that metallurgical properties are not compromised, limits on the process inputs are normally established General part categories, based on material and finished part application, are determined, and limits put

on each of these categories For example, the degree of sinter must be significantly better for structural parts such as gears compared to lightly stressed spacers

When engineering the process for a new part, baseline, or trial, parameters are normally established These settings are comfortably above the established minimums for the part category This allows for minor adjustments during the initial sample process Adjustments are sometimes needed to ensure that the required mechanical and physical properties are achieved

Once the production parameters are established, it is best if they are kept nearly constant Trying to use sintering to compensate for material or compacting variations leads to a high degree of run-to-run metallurgical variation and can contribute to part failures

Secondary Operations

Secondary operations, discussed in more detail elsewhere in this Volume, must also be considered in process planning and control Their relation with quality process control are briefly discussed below

Heat Treat. As with sintering, baseline heat treating parameters are generally established based on the part material and the end use Parameters are established during the initial sampling procedure, and adjustments from these settings are kept to a minimum to ensure consistent material properties

Restrike. Whether the part is being re-pressed (densified), coined, or sized, the primary factor to producing good parts is the tool design As with the design of compacting tools, the sizes may have to be factored to take into account subsequent processing that may change dimensions

Plating. With powdered metal parts, virtually any of the commonly applied platings can be utilized The primary difference between plating noncopper-infiltrated P/M versus wrought materials is that typically the parts first must be resin impregnated

to fill the porosity before the application of the plating Copper infiltration generally fills enough of the porosity to eliminate the need for resin impregnation

Machining. A great many powdered metal alloys are readily machinable The primary difference between P/M versus wrought is that P/M machining is actually a series of interrupted cuts, due to its inherent porosity Speeds, feeds, and coolant parameters are different, but overall throughput and process control philosophy is similar to the machining of non-P/M metals

Planning and Quality Control of Powder Metallurgy Parts Production

Jack R Bonsky, Cleveland State University, Advanced Manufacturing Center

Quality Control and Inspection

Lot Traceability

The sophistication of a system for part traceability depends on customer requirements and the risk that the producer is willing

to assume Should a problem develop, discrete traceability to a specific production operation can help reduce the quantity that might be involved in a rejection or product recall

As a minimum, most P/M suppliers maintain traceability to the material used Each raw material (metal powder, lubricant, additive, etc.) is assigned a unique material lot number by its supplier If the material is blended in-house, then it is common

to assign a lot number based on the blended batch This lot number is marked on all in-house processing containers and clearly designated on the finished goods containers as well When the powder is received as a preblend, a new lot number may be designated, or the lot number created by the powder supplier may be used throughout processing

Unlike the plastics molding and metal castings industries, easily changed lot designators cannot be inserted into the P/M compaction tools High molding pressures make this sort of designator impractical If as-molded designators are desired, they are high-strength portions of the tooling that require significant setup to change Because of this expensive setup, generally as-molded designators are changed only for each run, or up to once per week

Individual serialization of parts can be achieved with mechanical engravers These engravers can be used after any of the operations, including on green compacts

Powder Inspection

Consistency of powder characteristics is key to producing a quality finished part Chemistry, cleanliness, particle shape, and size distribution are the primary drivers of powder performance

A few, relatively simple checks are generally sufficient to ensure the consistency of incoming powder the first check being flow rate

Flow Rate Check. The determination of flow rate is generally measured according to MPIF Standard 03 (Ref 3) The procedure is summarized:

1 Obtain a sample of the powder A good way to get a representative sample from a bulk pack or drum of powder is to use a Keystone Sampler (Fig 7) according to MPIF Standard 01 (Ref 4) This manually operated device augers its way to the bottom of the container, then opens up to retrieve powder fro m throughout the container

2 Load the flowmeter funnel with the powder while keeping a dry finger over the discharge orifice

3 Start a stopwatch simultaneously with the removal of the finger from the orifice Stop timing when the last

of the powder leaves the flowmeter The flow rate is recorded in elapsed time in seconds

Flow rate is critical to press setup Faster-flowing (shorter rate in seconds) powders can fill the die cavity faster and can allow for faster press speeds

Fig 7 Keystone sampler for measuring powder flow Source: Ref 4, used with permission

Apparent Density Check. Another simple method to ensure that the incoming powder is consistent lot to lot is to check apparent density The most commonly followed method is MPIF Standard 04 (Ref 5), which is summarized:

1 A test specimen of metal powder is obtained Again, the Keystone Sampler is a good tool for getting a representative sample from a container

2 This entire specimen is loaded into the Hall flowmeter, and the powder should flow into the density cup

3 Level the heaped powder in the density cup with a nonmagnetic spatula with the blade held perpendicular

to the top of the cup Do not jar the cup

4 Tape the side of the cup slightly so that the powder settles, allowing the cup to be easily moved without spilling powder

5 Transfer the powder to a balance and check its mass

6 Apparent density is calculated as the ratio M/V, where M is the mass of powder from the density cup in grams, and V is the volume of the cup

Apparent density determines the amount (weight) of the powder that fills the die It is critical that setup personnel be aware of the apparent density value for a powder It is especially critical that they know when a new container of powder has a large change in apparent density The relative position of the die and lower punch(es) determines the amount of fill If the press is set for a powder of apparent density, then loading a powder with higher apparent density into the press can have catastrophic results during pressing The higher apparent density powder can increase the green density of the part to the point where the elastic limit of the tooling is exceeded, and the tools break

Microscopic Evaluation. A simple check of the powder under a microscope is also a good safeguard against sending defective powder to the press Powder discrepancies that can be detected vary from rust to gross anomalies such as large agglomerates of lubricant

References cited in this section

3 "Determination of Flow Rate of Free-Flowing Metal Powders Using the Hall Apparatus," Standard 03, Metal Powder Industries Federation

4 "Method for Sampling Finished Lots of Metal Powders," Standard 01, Metal Powder Industries Federation

5 "Method for Determination of Apparent Density of Free-Flowing Metal Powders Using the Hall Apparatus," Standard 04, Metal Powder Industries Federation

Planning and Quality Control of Powder Metallurgy Parts Production

Jack R Bonsky, Cleveland State University, Advanced Manufacturing Center

Process Control

A key aspect of process control philosophy is that shop floor operators be given as much "ownership" of their process as possible Ideally, the operator gages the process, charts it, takes corrective action when it goes out of control, and makes the appropriate adjustments to the process when necessary

Certain process control procedures, however, cannot be practically conducted by the production operators Some measurements cannot be accurately determined on the production floor Examples include green density checks and measurements requiring a coordinate measuring machine Both of these checks are best left to the laboratory specialists

Additionally, some adjustments, most compaction press adjustments for example, are beyond the level of technical expertise

of most operators These adjustments require specially trained setup personnel

Compacting Process Control. A variety of statistical process controls are used across the industry for the compaction Charts successfully employed include X-bar, range, and median This section briefly describes key variables that influence the implementation of statistical process control for powder compaction

For conventional rigid die compaction, generally vertical-direction dimensions and weight are statistically charted direction dimensions are those that run in the same direction as the compacting motion (Fig 8) These features vary significantly as a result of the various press inputs such as voltage, temperature, and mechanical movement Radial-direction dimensions, those that run horizontally or not in the direction of the press stroke generally do not vary to a large degree throughout an individual production run Therefore, monitoring of radial-direction dimensions is not adequate to show if the pressing action is in statistical control For longer runs, typically in the 50,000-piece or longer range, a periodic spot check is

Vertical-a good ideVertical-a to ensure thVertical-at the tooling hVertical-as not worn excessively The exVertical-act frequency required for the rVertical-adiVertical-al dimension checks depend on the abrasiveness of the powder, part configuration, and the wear resistance of the tool steel

Fig 8 Radial dimensions of part relative to typical die configuration and tool motion

Many newer presses are equipped with electronics that monitor and chart the various press characteristics These characteristics include total press tonnage, loads at various press locations, air pressures, hydraulic pressures, and hydraulic temperatures These outputs are sometimes used for true statistical control, but the majority of manufacturers use these outputs for simple, mechanically based, processing limits

Green density is the density of the P/M component after compaction A check of green density is a good verification method at the start of each production run and is checked periodically throughout the run for most parts However, for single-level parts of less than 6.35 mm ( in.) length, an in-process density check is generally not required Assuming that the control limits established will always produce correct density parts, so long as the process does not stray from these limits, the density will always be correct

Generally, these density checks are performed once or twice per shift The reason for the checks is that it is possible for the weight and lengths to be within control limits and the density in a portion of the part to be either over or under specification

For example, the process may shift, causing the flange density to increase and the body density to reduce The overall density

is correct, but portions of the part are high density and portions are low density High degrees of density variations can cause cracks, break tools, and cause part service failures

To ensure that the density is relatively homogenous, the green compact is sectioned and the density of each section is determined The difference between the density sections is calculated and compared against established limits The limits established may vary depending on a number of factors including tooling fragility and end-product use For example, a density split limit of 0.2 g/cm3 may be established on a two-level gear If the flange is running 7.25 g/cm3 and the body is running 7.00 g/cm3 (for a 0.25 g/cm3 split), the density should be adjusted at the compaction press Parts run since the last density split check must be carefully evaluated to ensure that they are mechanically sound and will produce finished parts that are within print tolerances

MPIF Standard 42 (Ref 6) is the industry standard Application of this procedure for green density is summarized:

1 A test specimen is obtained

2 Determine the mass of the specimen using a balance of adequate precision (Mass A)

3 Impregnate the specimen with oil

4 Remove excessive oil

5 Determine the mass of the impregnated test piece (Mass B)

6 Suspend the oil-impregnated specimen from the balance hook into water Completely submerge the piece

and be sure that all air bubbles are removed (Mass C)

7 Remove the specimen from the water and measure the mass of the balance hook as it is suspended in the

same manner as it was when holding the test piece (Mass E)

Green density is calculated as:

where w is a correction factor for water temperature

Sintering Process Control. Unlike compaction, which usually has a cycle time of well under 1 min, the sintering process may take hours from beginning to end (furnace exit) On shorter runs, all of the green parts may be loaded into the furnace before a part is removed from the furnace exit Because thousands of parts can be loaded before a part measurement can be made, a sample batch is usually run to ensure that the processing parameters are having their desired effect on the parts

Throughout the production runs, parts are usually measured and charted to statistical control limits Because the radial (from compaction stroke) dimensions are typically highly consistent in green compacts, these are normally good indicators of sintering performance The variation noted in these measurements is largely from the sintering process; thus, they can be reliably used for assessing the state of sintering process control

As with some of the newer compaction presses, many newer sintering furnaces have the capability to run true process control

on the processing characteristics With sintering, the outputs are temperature, speed, atmosphere flow rate, and so forth

In addition to dimensional checks, part macrohardness is also generally monitored throughout processing Control limits may

be established based on the material type, the density being pressed, and the desired as-sintered properties

Restrike Process Control. The movement of the material being repressed dictates the type of process control required Control charts may be appropriate for vertical dimensions or radial dimensions, and sometimes both are monitored

Heat Treat Process Control. Heat treating is similar to sintering in that radial dimensions are usually checked to ensure that the process is running well For continuous-belt furnaces, in-process statistical charting is appropriate The process can

be adjusted throughout the run to maintain control

With batch furnaces, all part-based verification for process control is after-the-fact inspection The best way to ensure that the process is running properly is to have good, tight controls of the inputs (temperature, carbon content, oxygen, etc.) As with sintering, part macrohardness is usually checked Control limits may be established based on the material type, density, sintering, and heat treating parameters

Secondary Operations. For most other secondary operations, process control is conducted on P/M parts similarly to that

of other metal fabrication techniques

Reference cited in this section

6 "Method for Determination of Density or Compacted or Sintered Metal Powder Products," Standard 42, Metal Powder Industries Federation

Planning and Quality Control of Powder Metallurgy Parts Production

Jack R Bonsky, Cleveland State University, Advanced Manufacturing Center

Inspection Considerations

Inspection techniques for finished parts should be based on concurrence between the customer and the supplier There are many ways to gage most part print specifications, and the specific technique used can have a large effect on the results reported A great many P/M part attributes may be measured with the same techniques used in other industries There are, however, significant exceptions as briefly described below

Hardness. In powder metallurgy there are generally two types of hardness specified apparent hardness (macrohardness) and microhardness The microhardness is the hardness of each particle of material, and the apparent hardness is the hardness

of the surface bridging across many particles and the porosity, too

Apparent hardness is typically measured according to MPIF Standard 43 (Ref 7) The procedure is relatively straightforward and quick The basics are:

1 Obtain a sample part of adequate thickness and parallel configuration (or, for cylindrical parts, a correction factor may be used)

2 The sample must be large enough so that the indenter marks from the hardness tester are at least three indenter diameters from any edge or previous impression

3 Sand each face of the sample so that no burrs are present (burrs will cause erroneous readings), or be sure to use a holding fixture that avoids the burrs

4 Take readings with a properly calibrated hardness tester

5 Reject obvious outliers and report the average of at least five nonoutliers

Typically, the outliers are on the low side The cause of these occasional low readings is a chance happening that the hardness indenter falls right into a pore

Microhardness is usually measured according to MPIF Standard 51 (Ref 8) The determination of microhardness is significantly more difficult than measuring apparent hardness and requires specialized equipment that many P/M users do not have on-site The procedure involves:

1 Sectioning the part and making a polished mount for the evaluation

2 Placing the mount in a special microhardness testing machine

3 Under magnification, orienting the mount and making a diamond indenter mark precisely over a particle of the material

4 Measuring the length of the penetration on the particle and converting this length to a hardness reading

Microfinish. The porosity of P/M causes debate over the proper method of measuring microfinish When using a standard cone stylus, unmachined P/M can give relatively high microfinish values The individual particles are very smooth, but the probe path is interrupted by pores of varying sizes (Fig 9) Because most parts are running against a mate that is much larger than the standard probe, many feel that the chisel probe provides a better gage as to the actual serviceability of the parts The chisel probe bridges across the porosity and usually produces significantly lower microfinish values

Fig 9 Surface measurement with chisel stylus (a) (b) Effect of cone and chisel styli on an as-sized P/M surface

Source: Ref 9, used with permission

Physical Tests. Performing a physical test on finished parts is a great method of verifying that all processes ran properly, and the final parts will perform adequately Crush, torque, impact, and tensile tests are commonly called out on P/M part

prints These tests greatly reduce the need for expensive, time-consuming, and sometimes ambiguous, microstructure analysis

References cited in this section

7 "Method for Determination of Apparent Hardness of Powder Metallurgy Products," Standard 43, Metal Powder Industries Federation

8 "Method for Determination of Microhardness of Powder Metallurgy Materials," Standard 51, Metal Powder Industries Federation

9 P/M Design Guidebook, Metal Powder Industries Federation, 1983, p 15

Planning and Quality Control of Powder Metallurgy Parts Production

Jack R Bonsky, Cleveland State University, Advanced Manufacturing Center

Defect Detection

The problem of forming defects in green parts during compaction and ejection has become more prevalent as parts producers have begun to use higher compaction pressures in an effort to achieve high-density, high-performance P/M steels Several nondestructive inspection methods are practical for detecting defects as early as possible in the production sequence The most promising nondestructive testing methods for P/M applications include electrical resistivity testing, eddy current and magnetic bridge testing, magnetic particle inspection, ultrasonic testing, x-ray radiography, gas permeability testing, and -ray density determination The capabilities and limitations of each of the techniques are briefly summarized in Table 2

Table 2 Comparison of the applicability of various nondestructive evaluation methods of flaw detection in P/M parts

Applicability to P/M parts(a)

pinpoint defect location

Extremely high initial cost; highly trained operator required; radiation hazard

Gamma-ray density

determination

Density variations A A High resolution and

accuracy; relatively fast

High initial cost; radiation hazard

Thermal wave imaging Subsurface cracks, density

variations

D C No coupling agent

required

Fat or convex surfaces only

Electrical resistivity Subsurface cracks, density of

variations, degree of sinter

A A Low cost, portable, high

potential for use on green compacts

Sensitive to edge effects

Eddy current/magnetic Cracks, overall density, C A Low cost, fast, can be Under development

bridge hardness, chemistry automated; used on P/M

valve seat inserts

Slow; operator sensitive

Liquid dye penetrant

A A Low cost, simple, fast Gas-tight fixture required;

cracks in green parts must intersect surface

Source: Ref 10 and 11 (reprinted with permission)

(a) A, has been used in the production of commercial P/M parts; B, under development for use in P/M; C, could be

developed for use in P/M, but no published trials yet; D, low probability of successful application to P/M

The four most common types of defects in P/M parts are ejection cracks, density variations, microlaminations, and poor sintering These defects are briefly described in this section, and SPC techniques related to defects are described in the next section of this article

Ejection Cracks. When a part has been pressed, there is a large residual stress in the part due to the constraint of the die and punches, which is relieved as the part is ejected from the die The strain associated with this stress relief depends on the compacting pressure, the green expansion of the material being compacted, and the rigidity of the die Green expansion, also known as spring out, is the difference between the ejected-part size and the die size A typical value of green expansion for a powder mix based on atomized iron powder pressed at relatively high pressure (600 to 700 MPa, or 45 to 50 tsi) is 0.20% In

a partially ejected compact, for example, the portion that is out of the die expands to relieve the residual stress, while the constrained portion remains die size and a shear stress is imposed on the compact When the ability of the powder compact to accommodate the shear stress is exceeded, ejection cracks are formed

The radial strain can be alleviated to a degree by increasing the die rigidity and designing some release into the die cavity However, assuming that the ejection punch motions are properly coordinated, the successful ejection of multilevel parts depends to a large degree on the use of a high-quality powder that combines high green strength with low green expansion and low stripping pressure

Density Variations. Even in the simplest tool geometry possible a solid circular cylinder conventional pressing of a part

to an overall relative density of 80% will result in a distribution of density within the part ranging from 72 to 82% (Ref 12) The addition of simple features such as a central hole and gear teeth presents minor problems compared with the introduction

of a step or second level in the part Depending on the severity of the step, a separate, independently actuated punch can be required for each level of the part During the very early stage of compaction, the powder redistributes itself by flowing between sections of the die cavity However, when the pressure increases and the powder movement is restricted, shearing of the compact in planes parallel to the punch axis can only be avoided by proper coordination of punch motions When such shear exists, a density gradient results

The density gradient is not always severe enough for an associated crack to form upon ejection However, a low-density area around an internal corner can be a fatal flaw because this corner is usually a point of stress concentration when the part is loaded in service

Microlaminations. In photomicrographs of unetched part cross sections, microlaminations appear as layers of unsintered interparticle boundaries that are oriented in planes normal to the punch axis They can be the result of fine microcracks associated with shear stresses upon ejection; such microcracks fail to heal during sintering Because of their orientation parallel to the tensile axis of standard test bars, they have little influence on the measured tensile properties of the bars, but are presumed to be a cause of severe anisotropy of tensile properties

Poor Sintering. When unsintered particle boundaries results from a cause other than shear stresses, they are usually present because of insufficient sintering time or sintering temperature, a nonreducing atmosphere, poor lubricant burn-off, inhibition

of graphite dissolution, or a combination of these Unlike microlaminations, defects associated with a poor degree of sintering are not oriented in planes

References cited in this section

10 Prevention and Detection of Cracks in Ferrous P/M Parts, Metal Powder Industries Federation, 1988

11 R.C O'Brien and W.B James, Powder Metallurgy Parts, Nondestructive Evaluation and Quality Control, Vol

17, ASM Handbook, 1989, p 537

12 G Taguchi, On-Line Quality Control During Production, Japanese Standards Association, 1981

Planning and Quality Control of Powder Metallurgy Parts Production

Jack R Bonsky, Cleveland State University, Advanced Manufacturing Center

Shewhart Control Charts for Defect Data

Many quality assessment criteria for manufactured goods are not of the variable measurement type Rather, some quality characteristics are more logically defined in a presence-of or absence-of sense In P/M, examples include chipped gear teeth, discolored plating, scratched surfaces, and powder accumulation on parts

Such nonconformities or defects are often observed visually or according to some sensory criteria and cause a part to be defined simply as a defective part In these cases, quality assessment is referred to as being made by attributes

Many quality characteristics that could be made by measurements (variables) are often not done as such in the interest of economy A go/no-go gage can be used to determine whether or not a variable characteristic falls within the part specification Parts that fail such a test are simply labeled defective Attribute measurements can be used to identify the presence of problems, which can then be attacked by the use of and R control charts The following definitions are

required in working with attribute data:

• Defect: A fault that causes an article or an item to fail to meet specification requirements Each instance of

the lack of conformity of an article to specification is a defect or nonconformity

• Defective: An item or article with one or more defects is a defective item

• Number of defects: In a sample of n items, c is the number of defects in the sample An item may be subject

to many different types of defects, each of which may occur several times

• Number of defectives: In a sample of n items, d is the number of defective items in the sample

• Fractional defective: The fractional defective, p, of a sample is the ratio of the number of defectives in a sample of the total number of items in the sample Therefore, p = d/n

Operational Definitions

The most difficult aspect of quality characterization by attributes is the precise determination of what constitutes the presence

of a particular defect This is so because many attribute defects are visual in nature and therefore require a certain degree of judgment and because of the failure to discard the product control mentality For example, a scratch that is barely observable

by the naked eye may not be considered a defect, but one that is readily seen is Furthermore, human variation is generally considerably larger in attribute characterization (for example, three different caliper readings of a workpiece dimension by three inspectors and visual inspection of a part by these same individuals yield anywhere from zero to ten defects) It is therefore important that precise and quantitative operational definitions be laid down for all to observe uniformly when attribute quality characterization is being used The length or depth of a scratch, the diameter of a surface blemish, or the height of a burr

The issue of the product control versus process control way of thinking about defects is a crucial one From a product control point of view, scratches on a magnetic catch plate should be counted as defects only if they appear on visual surfaces, which would directly influence part function From a process control point of view, however, scratches on a catch plate should be counted as defects regardless of where they appear because the mechanism creating these scratches does not differentiate between visual and concealed surfaces By counting all scratches, the sensitivity of the statistical charting instrument used to identify the presence of defects and to lead to their diagnosis will be considerably increased

A major problem with the product control way of thinking about part inspection is that when attribute quality characterization

is being used, not all defects are observed and noted The first occurrence of a defect that is detected immediately causes the part to be scrapped Often, such data are recorded in scrap logs, which then present a biased view of what the problem may really be One inspector may concentrate on scratch defects and will therefore tend to see these first Another may think brightness is more critical, so his data tend to reflect this type of defect more frequently The net result is that often such data may then mislead those who may be using it for process control purposes Therefore, it is essential from a process control standpoint to carefully observe and note each occurrence of each type of defect

p-Chart and c-Chart Analyses

Example 1: p-Chart Analysis for Fraction Defective during Tapping

Consider a tapping operation on a P/M bracket Suppose the measures of quality conformance of interest are the presence of threads (generally absent due to parts being inadvertently transferred into the finished parts bin without processing) and the size of the tapped threads as measured with a go/no-go plug (generally defective due to cutting tool wear)

To establish the control chart, rational samples of size n = 50 parts are drawn from production periodically (perhaps, each

shift), and the sampled parts are inspected and classified as either defective (from either or both possible defects) or

nondefective The number of defectives, d, is recorded for each sample The process characteristic of interest is the true process fraction defective p' Each sample result is converted to a fraction defective:

(Eq 1)

The data (fraction defective p) are plotted for at least 25 successive samples of size n = 50 The individual values for the sample fraction defective, p, vary considerably, and it is difficult to determine from the plot at this point if the variation about

the average fraction defective, , is solely due to the forces of common causes or special causes

Control Limits for the p-Chart. It can be shown that for random sampling, under certain assumptions, the occurrence of

the number of defectives, d, in the sample of size n is explained probabilistically by the binominal distribution Because the sample fraction defective, p, is simply the number of defectives, d, divided by the sample size, n, the occurrence of values for

p also follows the binominal distribution Given k rational samples of size n, the true fraction defective, p', can be estimated

by:

(Eq 2)

or

(Eq 3)

Equation 3 is more general because it is valid whether or not the sample size is the same for all samples Equation 2 should be

used only if the sample size, n, is the same for all k samples

Therefore, given , the control limits for the p-chart are then given by:

Thus, only has to be calculated for at least 25 samples of size n to set up a p-chart The binomial distribution is generally not symmetric in quality control applications and has a lower bound of p = 0 Sometimes the calculation for the lower control

limit may yield a value of less than 0 In this case, a lower control limit of 0 is used

c-Chart: Analysis for Number of Defects. The p-chart deals with the notion of a defective part or item where defective

means that the part has at least one nonconformity or disqualifying defect It must be recognized, however, that the incidence

of any one of several possible nonconformities would qualify a part for defective status A part with ten defects, any one of which makes it defective, is on equal footing with a part with only one defect in terms of being defective

Often it is of interest to note every occurrence of every type of defect on a part and to chart the number of defects per sample

(c) A sample may only be one part, particularly if interest is focusing on final inspection of an assembled product, or

inspection may focus on one type of defect or multiple defects

The probability law that governs the incidence of defects is known as the Poisson law or Poisson probability distribution,

where c is the number of defects per sample It is important that the opportunity space for defects to occur be constant from sample to sample The Poisson distribution defines the probability of observing c defects in a sample where c' is the average

rate of occurrence of defects per sample

Construction of c-Charts from Sample Data. The number of defects, c, arises probabilistically according to the

Poisson distribution One important property of the Poisson distribution is that the mean and variance are the same value

Then given c', the true average number of defects per sample, the 3 limits for the c-chart are given by:

Planning and Quality Control of Powder Metallurgy Parts Production

Jack R Bonsky, Cleveland State University, Advanced Manufacturing Center

Tolerance Control

The issue of part tolerancing and, in particular, the statistical assignment and assessment of tolerances are excellent examples

of the need for design and manufacturing to understand what each other is doing and why The best intentions of the design process can go unmet if the manufacturing process is not operated in a manner totally consistent with design intent To more clearly appreciate the relationship that must exist between the design and manufacturing operations, some of the basic assumptions of the tolerancing activity and their relationship to the manufacturing process are examined The following sections clearly point to the importance of statistical process control relative to the issue of process capability

The key concepts in statistical tolerancing are:

• The use of a statistical distribution to represent the design characteristic and therefore the process output for the product/part in question relative to the design specifications

• The notion of random assembly, that is, random part selection from these part process distributions when more than one part is being considered in an assembly

• The additive law of variances as a means to determine the relationship between the variability in individual parts and that for the assembly

To assume that the parts can be represented by a statistical distribution of measurements (and for the assumption to hold in reality), the part processes must be in a state of statistical control The following example illustrates the importance of statistical process control in achieving design intent in a tolerancing problem

Example 2: Statistical Tolerance Model for Optimal Fit of a Pin Assembly into a Molded Hole of a P/M Plate

Figure 10 shows two simple parts: a plate with a hole and a pin that will ultimately be assembled to a third part but must pass through the hole in the plate For the assembly, it is desired for function that the clearance between the plate hole and the pin

be at least 0.015 in but no more than 0.055 in

Fig 10 Pin and hole components statistically analyzed in Example 2 (a) Plate with hole (b) Pin assembly

Dimensions given in inches

To achieve the design requirement stated above, the nominal values and tolerance for the plate hole and pin were statistically derived and are shown in Fig 10 To arrive at these tolerances, it was assumed that:

• The parts would be manufactured by processes that behave according to the normal distribution

• The process capabilities would be at least 6 , the processes would be centered at the nominal values given

in Fig 10, and the processes would be maintained in a state of statistical control

• Random assembly would prevail

If these assumptions are met, the processes for the two parts, and therefore the clearance associated with assembled parts, would be as shown in Fig 11, and the design intent would be met

Fig 11 Statistical basis for satisfying design intent for the hole/pin assembly clearance in Fig 10 (a) Distribution

of hole (b) Distribution of pin (c) Distribution of clearance

Suppose that despite the assumptions made and the tolerances derived, the processes manufacturing the pin and plate hole were not maintained in good statistical control As a result, the parts actually more nearly follow a uniform/rectangular distribution within the specifications, as shown in Fig 12 Such could have arisen as a result of sorting or rework of a more variable process(es), in which case the results are doubly distressing, that is, poorly fitting assemblies and increased cost to the system

Fig 12 Clearance implications of poor process control of plate hole and pin dimensions for components of Fig 10

(a) Distribution of hole (b) Distribution of pin (c) Distribution of clearance

Figure 12 shows the distribution of the clearance if the hole and pin dimensions follow the uniform distribution within the specifications The additive law of variances has been used to derive the variation in the clearance distributions but assuming the uniform distribution for the individual part processes Some assemblies may not go together at all, some will fit quite tightly and may later bind if foreign matter gets into the gap, and others will fit together with a much larger clearance than desired

The problem here is not a design problem The plate hole and pin tolerances have been derived using sound statistical methods However, if the processes are not in good statistical control and therefore not capable of meeting the assumptions made during design, poor-quality assemblies will follow It should be noted that the altogether too common process appearance of a uniform distribution of measurements within the specifications can arise in several different ways:

• From processes that have good potential with regard to variation, but are not kept in good statistical control

• From unstable and/or large variation processes that require sorting/rework to meet the specifications

• From processes that are intentionally allowed to vary over the full range of the specifications to take advantage locally of wide specifications relative to the process variation

In all of the three cases mentioned above, additional costs will be incurred and product quality will be eroded Clearly, SPC is crucial to the tolerancing issue in engineering design

Planning and Quality Control of Powder Metallurgy Parts Production

Jack R Bonsky, Cleveland State University, Advanced Manufacturing Center

References

1 W.A Levinson, Make the Most of Control Charts, Chem Eng Prog., Vol 88 (No 3), March 1992, p 86-91

2 R.E DeVor, T.H Chang, and J.W Sutherland, Statistical Quality Design and Control, Macmillan, 1992

3 "Determination of Flow Rate of Free-Flowing Metal Powders Using the Hall Apparatus," Standard 03, Metal Powder Industries Federation

4 "Method for Sampling Finished Lots of Metal Powders," Standard 01, Metal Powder Industries Federation

5 "Method for Determination of Apparent Density of Free-Flowing Metal Powders Using the Hall Apparatus," Standard 04, Metal Powder Industries Federation

6 "Method for Determination of Density or Compacted or Sintered Metal Powder Products," Standard 42, Metal Powder Industries Federation

7 "Method for Determination of Apparent Hardness of Powder Metallurgy Products," Standard 43, Metal Powder Industries Federation

8 "Method for Determination of Microhardness of Powder Metallurgy Materials," Standard 51, Metal Powder Industries Federation

9 P/M Design Guidebook, Metal Powder Industries Federation, 1983, p 15

10 Prevention and Detection of Cracks in Ferrous P/M Parts, Metal Powder Industries Federation, 1988

11 R.C O'Brien and W.B James, Powder Metallurgy Parts, Nondestructive Evaluation and Quality Control, Vol

17, ASM Handbook, 1989, p 537

12 G Taguchi, On-Line Quality Control During Production, Japanese Standards Association, 1981

Quality Control and Inspection of

It is extremely useful to have mutual (customer/supplier) agreement on gaging concepts during the quality planning process Quality requirements must also be discussed at this time to avoid capability and processing issues later on Classification of characteristics (C of C) and the associated capability indices should be thoroughly understood to properly process the part Secondary operations are often added to satisfy statistical process control requirements

Another important step in the quality planning process is to plan and develop the process(es) before the first parts are

produced it is much less desirable to backtrack (reactive) in the quality planning activities than to act (proactive) on a well thought out plan with failure modes defined and methods provided to prevent and/or detect those potential failure modes Two useful tools to help in the planning process are a failure mode and effects analysis (FMEA) and a control plan

A FMEA process is a very good engineering tool that is used to detect problem areas in the process and to mistake-proof and/or assign the correct process controls It should be noted that the FMEA process is used to flag potential causes of failure

in the process: if the potential for failure is high, efforts should first be made to minimize the potential effects of the failure Once the risk has been minimized, the appropriate process control can be assigned Mistake-proofing should be used to reduce risk and ensure quality

The control plan assigns control methods to the process in an effort to help minimize process and product variation Features

to be checked, method of inspection, control method, frequency of inspection, and responsibility are assigned for each operation The control plan takes into consideration the FMEA results as well as any classification of characteristics that the customer has defined Depending on the characteristic, the amount of control may vary, from simple first piece layouts (for tool generated dimensions) to 100% inspection on characteristics generated using incapable machinery

Quality Control and Inspection of Powder Metallurgy Secondary Operations

Pat Kenkel, John Engquist, and Mike Blanton, Burgess-Norton Manufacturing Company

Secondary Operations

The purpose of this article is not to discuss technology gains made in secondary operations for powder metallurgy, but to provide useful information in controlling and improving secondary processes using quality control and inspection methodologies However, development of the following secondary operations is briefly discussed:

• Restrike densification

• Restrike sizing

• Heat treat

• Machining and grinding

Examples of FMEA and control plans are shown in Fig 1 and 2, respectively, for the restrike densification process to illustrate the use of such tools when assigning process control methods and inspection instructions to a process Additional information on statistical process control concepts and terminology are also discussed in the article "Planning and Quality Control of Powder Metallurgy Parts Production" in this Volume

Fig 1 Brief example of a FMEA for restrike operation

Fig 2 Brief example of a control plan for restrike operation

Restrike Densification. Restrike operations are frequently used to densify parts above the density attainable at compacting In these instances, more emphasis is placed on attaining the desired material property through densification than

on achieving tight dimensional limits (although dimensional control is also important) Adequate control must be placed on the process to ensure density requirements (and dimensions) are met Because density checks are not readily performed in a production environment, dimensional controls may be used to control density if correlation can be proved to be sufficient When densifying thin parts, greater dimensional and weight control may be required at compacting to avoid overdensification and damage to the restrike press and/or tooling Press and tool load monitors are effective process control tools

Restrike Sizing. Sizing using a restrike operation is used to enhance dimensional characteristics on P/M parts In this instance, the density of the part has already been met in the compacting and sintering operations Sizing is used to meet tighter dimensional limits on diameters, lengths, tapers, and so forth The quality control methods assigned for sizing operations depend on the characteristic sized and the capability of the machine in achieving the desired limits of size; highly capable processes require very little in-process or post-process inspection while marginally capable processes require significant and frequent inspection

Heat Treat. The properties from heat treatment (such as hardness and tensile strength) are dependent on many process variables Carbon potential, atmosphere, heat treat temperature, quench rate and temperature, and draw temperature affect the properties of parts A FMEA of the heat treatment should be done to list potential failures caused by incorrect settings of the above inputs To minimize potential process failures, process control of the heat treat operation should include monitoring the inputs as well as the outputs Minimal variation of these parameters results in significantly reduced variation of product characteristics Again, depending on the stability of the process, more or less dimensional and metallurgical inspection is required as post-process inspection

Machining/Grinding. Although machining and grinding are different technologies, process control issues are similar for each Variation in the process can and should be minimized in the selection of capable machinery, the design of the fixturing, and the selection of the tooling (inserts, cutters, grinding wheels, etc.) Process control and inspection techniques vary depending on the characteristics being generated, the capability of the process, and the quality requirements specified on the blueprint (classification of characteristics) Machine controls have allowed technologies such as automatic compensation and redundant tooling to significantly reduce process variation due to tool wear and other time-related variations The selection of the machine to be used is a big factor when analyzing the FMEA and developing control plans

Quality Control and Inspection of Powder Metallurgy Secondary Operations

Pat Kenkel, John Engquist, and Mike Blanton, Burgess-Norton Manufacturing Company

Process Control Methods

Many methods are used to control part conformance in manufacturing processes Control plans should be selected to minimize the financial impact to the process (either in the cost of the gage and/or the time required to gage the part), but also

to allow for optimal control of the process (100% inspection controls the outgoing product, but typically is not a effective-inspection method) As stated above, the control plan must include the gaging method, the control method, the frequency, and the responsibility A discussion of each area follows

cost-Gaging Method. The two types of commonly used gaging in the manufacturing environment are attribute gages and variable gages Attribute gaging is used to check conformance or nonconformance for a specific dimension/tolerance Variable gaging is used to record actual dimensional values for comparison to the tolerance specification Both types of gages can be used to effectively control processes

Typically, attribute gaging is used to control relational features (GDT features such as true positions, etc.) and dimensional

features that are either incidental characteristics or produced by an extremely capable and stable process (i.e., Cpk values over

2.00) (see the article "Planning and Quality Control Powder Metallurgy Parts Production" in this Volume for definition of Cpkand other statistical control indices) Attribute gages range from inexpensive go/no-go plug gages to complex, part specific

true position gages that can be very expensive For well-centered processes, these gages can be useful in helping the operator control the process It should also be noted that restricted tolerance attribute gaging, also called "narrow limit" gaging (50%

or 75% of the tolerance band), can be used to monitor and compensate the process before nonconforming parts are produced

One of the biggest advantages of powder metallurgy is the ability to press features that are complex The secondary operations mentioned above refine pressed features and/or generate new features that often have a geometric relationship to pressed features With proper GDT, functional attribute gaging is a cost-effective way to control the relationship of features generated by secondary operations to those generated at the press

Variable gaging allows the operator to record specific dimensional data that can then be analyzed and used to make decisions

on process adjustments Variable gages are used to control processes that are incapable and/or unstable over time (tool wear, etc.) Variable gaging is frequently used on significant characteristics that require a high degree of control Variable gages can

be standard components such as digital micrometers or complex gages that are part specific In order for the variable gage to

be effective for product acceptance decisions, the gage must have a minimum resolution of 10 divisions over the tolerance band being measured (0.010 in tolerance band needs a gage capable of measuring in 0.001 in increments)

Variable data is generally required by customers to evaluate the capability of a process Gage resolution used to estimate process capability may be greater than the gage resolution required for product acceptance It should also be noted that variable gaging is often used to establish process capability, but the process can then be monitored using attribute gages (if capability is high, for example) In these instances, gaging costs may still be high, but the cost impact to the process has been minimized by using attribute gages

Control Method. The most common control method for manufacturing processes is statistical process control (SPC) Three common statistical tools are widely used in manufacturing to control processes For ease of understanding their use and effectiveness, they are discussed below (in order of effectiveness, least to greatest)

Sampling tables were used in industry long before the current statistical approaches were made popular Typical sampling tables, such as MIL-STD-105E (current standard ANSI/ASQC Z1.4 1993), were developed for use with attribute gaging and mandated some acceptable quality level (AQL) limit as a decision criteria for accepting or rejecting a lot of product Statistical methods were used to determine what the sample sizes should be for a specific lot of material Sampling tables can still be used as a quality control tool, but are usually used to control lots of product (post-process inspection plans), not the process itself

A histogram/frequency distribution can be used in process control to visualize the pattern of variation being generated

by an operation This picture of the data is very useful in easily determining (1) the approximate central value or central tendency (average), (2) the spread of the values (variation), and (3) the relationship of the values to the process specifications (capability index) A histogram can turn machine-recorded values (Fig 3) into a beneficial tool to monitor a process (Fig 4)

Fig 3 Data generated at a machine

Fig 4 Histogram of the same data as Fig 3 A histogram gives a "picture" of the process that is easily analyzed by

the operator

Control Charts. Time-ordered data are an effective way to view process information Control charts vary in the

presentation of the data collected (individuals chart, x-bar and R, group charts, etc.), but are useful tools to control processes

over time For example, if the same data used in the histogram in Fig 4 are displayed as an individual chart (or average chart) (Fig 5), the same information (average, variation, process performance) can be approximated Because the control chart is time ordered, trends (decreasing or increasing values) can also be depicted The range is determined as the difference in subsequent values and plotted on a range chart (Fig 6) The range chart shows how stable the process is (the variation from one reading to the next)

Fig 5 x or x-bar chart for 49 diameter measurements

Fig 6 Range chart of variations for measurements in Fig 5

Other control methods (nonstatistical) can be used to control manufacturing processes If the features to be controlled are generated by the compacting or sizing tooling, an "audit" of the characteristic may be sufficient The audit is usually done at setup (first piece) and/or at the end of the production run (last piece) This ensures that no dramatic change occurred during the process

Another control method is 100% inspection Inspection of all parts may be necessary if the process is incapable or out of control If possible, 100% inspection should be avoided it adds cost to the operation and is not an effective way to control the process (100% inspection is not 100% effective) The best control methods for capable and stable processes are statistically based Mistake-proofing is a cost-effective approach to 100% inspection in some cases

Frequency. The sampling frequency for a manufacturing process varies based on the classification of the characteristic, the capability of the process, and the stability of the process:

Process parameter Inspection Frequency

Classification of characteristics Increases as characteristics become more critical

Process capability Decreases as capability increases

Process stability Decreases as stability increases

A stable and capable process may only use tool control as a control method The opposite extreme may require 100% inspection Most process parameters can be controlled using statistical methods with either attribute or variable gaging and varying frequencies of inspection (statistical methods are discussed above)

Responsibility. Organizations assign quality and inspection responsibilities differently The most effective inspections are done at the point of manufacture by the person(s) performing the manufacturing task Real-time data collection and instant process feedback is the best control in any manufacturing environment Obviously, this is not realistic in some instances (heat treating for example), so the responsibility in those situations must be assigned appropriately (to Inspection, for example) The main goal of collecting data should be to control the processes It is useless to generate data that means nothing after the fact Every effort should be made to utilize prevention activities, not detection activities, when developing control plans

Quality Control and Inspection of Powder Metallurgy Secondary Operations

Pat Kenkel, John Engquist, and Mike Blanton, Burgess-Norton Manufacturing Company

Continuous Improvement and Examples

The tools used in quality control and inspection can also be used to continually improve the processes mentioned above The use of statistical analysis can be used to make small incremental improvements to processes Design of experiments (DOEs) also utilize statistics to review data and make significant changes to improve existing processes

Technology has dramatically improved information gathering, and consequently better, faster feedback can be made to the process to improve control Improvements to equipment, machine tools, programmable logic controllers (PLCs) and computer numerical controllers (CNCs) have enhanced capabilities Electronic data collection and analysis allows for communication between the machine and the data collector, thereby providing automatic compensation and reduced variation

in the process In-process gaging techniques can be used to measure the part as it is being processed (machined, ground, etc.) Below are examples where quality control and inspection techniques have been used to improve existing (or new) processes

Case Study No 1: Tool Life Study on CNC Lathe. A tool life study was conducted on a new CNC lathe The lathe was purchased with an automatic tool compensation option To implement automatic compensation, a study was done using statistical methods A design of experiment was conducted to select the optimal tool insert and feed rate to be used (Fig 7)

Fig 7 Parts produced versus feed rate from lathe test

Once the best tool insert was selected, parts were 100% inspected and recorded using real-time data collection Adjustments were made only when the measured dimension reached the high limit of conformance The manual adjustment targeted the low conformance limit of conformance The machine continued to produce parts until the measured dimension was again at high limit where another manual adjustment was made This mode of operation continued over several "manual" adjustments

in order to determine the tool life of the insert and the size degradation of each cutting cycle The study showed that tool life had a predictable trend (Fig 8) With this information, the amount of compensation and the insert life span were entered into the lathe program The lathe now runs with no operator input and less time-ordered variation than previously observed Variation related to tool wear and operator "hunting" was virtually eliminated Ongoing verification of the effectiveness of automatic compensation continues with real-time control charts

Fig 8 Trend of dimensional change from tool wear

Case Study No 2: Capability Study New Equipment Purchase. A runoff of a new CNC machining center was performed and all data recorded The runoff criteria was met, but a pronounced pattern existed in the collected data (Fig 9) The pattern consisted of four repeating readings that drifted over time Automatic compensation was originally set at four pieces It was determined that the amount of compensation was set too high, and an adjustment was made The machine tool vendor was contacted about the repeating pattern, and upon investigation adjustments were made to the tooling and machine controls (thermal adjustments, etc.) The second run of parts revealed the results shown in Fig 10 Again, a slight drift was observed over time After one more adjustment to the automatic compensation, the process became capable and very stable

Fig 9 Variation in runoff from startup test of a new CNC center

Fig 10 Runoff variation after thermal adjustments and adjustments with tooling and machine controls

Case Study No 3: Capability Study Existing Outside Diameter (OD) Grinder. An existing process on an OD grinder was set to run with a 0.0005 in tolerance band on the OD The process was stable, but marginally capable (Fig 11a) The product was measured and charted on every machine cycle The process drift was analyzed to determine wheel wear and offsets (and timing) were set Using real-time data collection, a central tendency within 0.000005 in was achieved (Fig 11b) Ongoing process improvements with this process have led to dramatic quality improvements over time (Fig 12)

Fig 11 Histogram of tolerance with OD grinder a) Before process improvement b) After process improvement

Fig 12 Process quality index (Cpk ) of improved grinding operation

Quality Control and Inspection of Powder Metallurgy Secondary Operations

Pat Kenkel, John Engquist, and Mike Blanton, Burgess-Norton Manufacturing Company

Summary

The manufacture of P/M components represents focused technologies and creative innovations When product tolerances and/or product features cannot be attained by conventional P/M processing (press-sinter), secondary operations are required These operations may include repressing, heat treating, machining, grinding, and others Quality requirements mandate control of operations throughout the manufacturing process Although processes vary in technology, the quality control and inspection methodology is consistent

A good quality planning process includes:

• Preliminary discussions between the customer and supplier (understanding of the product and the form, fit, and function of the component)

• Thorough understanding of the quality requirements (i.e., classification of characteristics, quality index requirements, etc.)

• The use of engineering tools such as FMEAs to plan manufacturing processes

• A control plan that includes gaging methods, control methods, inspection frequency, and responsibility

Consistent use of these tools and the methodology minimize the risks and failure modes in any manufacturing process Process control methods are also used for continuous improvement

Quality Control and Inspection of Powder Metallurgy Secondary Operations

Pat Kenkel, John Engquist, and Mike Blanton, Burgess-Norton Manufacturing Company

Selected References

• Advanced Product Quality Planning (APQP) and Control Plan reference manual, AIAG, 1995

• A.V Feigenbaum, Total Quality Control, 3rd ed., McGraw-Hill, 1983

• E.R Ott, Process Quality Control, McGraw-Hill, 1975

Testing and Evaluation of Powder Metallurgy Parts

This article is adapted from L Pease III, Inspection and Quality Control for P/M Materials, Powder Metallurgy, Vol 7, ASM

Handbook, 1985, p 480-491 and R.C O'Brien and W.B James, Powder Metallurgy Parts, Nondestructive Evaluation and Quality Control, Vol 17, ASM Handbook, 1989, p 536-545

Testing and Evaluation of Powder Metallurgy Parts

Dimensional Evaluation

Dimensional accuracy of P/M sintered parts is determined with the same measurement techniques that are used for wrought materials Other testing methods for P/M materials are specialized, however, such as determination of surface finish For sintered parts, a chisel-pointed stylus is used to de-emphasize the effects of porosity A conical stylus tends to engage porosity, thus giving an exaggerated measurement of roughness Pores do not interfere mechanically with mating parts (Ref 1)

During the manufacture of sintered parts, dimensional change must be accommodated for during each processing step Causes of these changes include:

• Elastic springback during ejection from tooling used for cold pressing

• Growth or shrinkage during delubrication, presintering, and sintering

• Elastic springback from tooling during cold repressing or sizing

• Thermal contraction from the tools used in hot forging or hot repressing

• Tool wear in cold or hot compacting

• Machining tolerances at secondary machining and associated tool wear

• Distortion during annealing

• Growth or shrinkage during carburizing, nitriding, or neutral hardening

• Shrinkage during tempering

• Growth during steam blackening

Parts manufacturers must be familiar with the amount of dimensional change to expect for the materials and equipment in use

so that tooling can be produced that accommodates these changes and produces accurate parts Understanding and controlling these factors is essential to commercial P/M parts manufacturing, as discussed in the following paragraphs

Springback at Molding. Metal powders have varying yield points, and green compacts have varying elastic moduli Thus, even if a tool set is perfectly rigid, the size of the ejected compact is larger than the tool cavity This amount also varies, depending on molding pressure and powder characteristics Furthermore, expansion is not uniform or isotropic, except for right circular cylinders Dies exhibit some compliance, and split dies distort slightly in use, resulting in further changes in green dimensions Thus, green dimensions depend on die design and construction

Sintering Dimensional Change. Compacted elemental powders generally shrink during sintering The compacts begin

as unsintered objects that are larger than die size and eventually shrink below die size This phenomenon is illustrated in Fig

1 for pure iron (F-0000), based on the 89.61 mm (3.528 in.) dimension of a Metal Powder Industries Federation (MPIF) standard 10 tensile bar Prealloyed powders generally shrink during sintering (see the curve for AISI 4680 steel in Fig 2) Graphite additions to iron do not inhibit shrinkage from green dimensions, but they do cause additional elastic springback from the mold dimension Such materials thus begin sintering 0.5% larger than die size (see Fig 2)

Fig 1 Dimensional change on sintering of F-0000, F-0008, and 45P irons Sintered at 1120 °C (2050 °F) in

dissociated ammonia Green density: 6.8 g/cm 3 45P: atomized iron plus 0.45% phosphorus added as ferrophosphorus master alloy F-0000: pure atomized iron F-0008: atomized iron plus 0.9% graphite

Fig 2 Dimensional change on sintering of FC-0208, 4680, and FN-0308 irons Sintered at 1120 °C (2050 °F) in

dissociated ammonia Green density: 6.8 g/cm 3 FC-0208: sponge iron plus 2% Cu plus 0.97% graphite 4680: prealloyed AISI type 4600 steel plus 0.9% graphite FN-0208: atomized iron plus 2% Ni plus 0.9% graphite

Elements such as carbonyl nickel accelerate the shrinkage of iron mixes (see FN-0208 in Fig 2) In iron mixes, copper tends

to cause growth The higher the density of an iron particle (low surface area), the greater its dimensional growth with copper

additions Carbon dissolved in iron prior to copper diffusion inhibits growth (see Fig 3 and 4) Growth of copper-tin premixes is shown in Fig 5

Fig 3 Dimensional change on sintering of MH100 iron Values inside the graph indicate the percentage change

from die size MH100 iron sintered 30 min at 1120 °C (2050 °F) in dissociated ammonia Green density: 6.4 g/cm 3 Source: Ref 2

Fig 4 Dimensional change on sintering of ATOMET 28 iron ATOMET 28 powder plus 0.9% graphite plus 0.75%

zinc stearate plus copper as shown Green density ranges from 6.2 to 7.0 g/cm 3 Sintered 30 min in dissociated ammonia at 1120 °C (2050 °F) Source: Ref 3

Fig 5 Dimensional change on sintering of 90-10 bronze PMB 18 premixed bronze sintered 15 min in hydrogen at

various temperatures Green density: 6.3 g/cm 3 Source: Ref 4

Growth of these bronze premixes is strongly density dependent, with densities near 7.0 g/cm3 providing greater growth and densities near 6.0 g/cm3 providing significantly less Dimensional change data during sintering are complicated by the fact that dimensional change parallel to the pressing direction is not the same as dimensional change occurring perpendicular to the pressing direction

Annealing Dimensional Change. Annealing dimensional change is similar to change occurring during extra sintering or stress relaxation, as compacts continue to shrink Lower density compacts experience the most shrinkage With pure copper that has been pressed, sintered, and repressed to high density, annealing in a hydrogen atmosphere that contains gases can cause growth or blistering The hydrogen diffuses into the pores that are isolated from the surface If hydrogen encounters residual oxygen, water vapor is formed, which causes expansion and growth This phenomenon also can be caused by residual sulfur or the presence of lubricant Consequently, the initial sintering of pure copper should remove as many impurities as possible while the pores are still open

Dimensional Change during Heat Treating. Carbon and nitrogen absorbed during carburizing and nitriding of steel cause growth in areas in which they dissolve During quenching, regions that form martensite experience growth Thus, if a small part is through-hardened to all martensite, all regions (inside and outside diameter) experience outward expansion Many sintered steels have modest hardenability, and only the outer 3.2 mm (0.125 in.) adjacent to surfaces forms martensite and expands

On quenching a medium- or large-sized part (with a section >9.5 mm, or 0.4 in.), the interior remains as ferrite and fine pearlite, experiencing neither shrinkage nor growth The outer surfaces expand outward, and the inner surfaces shrink inward This phenomenon also is evident in case-hardened wrought parts Thus, prediction of exact size change during heat treatment

can be measured metallographically with a polishing procedure described in the section of this article on density measurement

Evaluation of Dimensional Change in Incoming Powder. New lots of blended or raw powder are checked against internal standard lots to ensure consistent sintered dimensional change Transverse-rupture bars 31.8 by 12.7 by 6.4 mm (1.25

by 0.50 by 0.25 in.) are molded at a fixed density or pressure from both the standard and test lot of powder The two sets of bars are sintered simultaneously in a laboratory or production furnace Dimensional change in the 31.8 mm (1.25 in.) length are checked against the requirements of American Society for Testing and Materials (ASTM) standard B 610

Although dimensional change from sintering a bar made from the standard powder can differ from previous tests, comparable dimensional changes in the test bar made from incoming powder demonstrate the difference in the performance of the powders Dimensional change in test and standard lots must agree to within a specified range (±0.1% of the bar length) These bars also can be used to evaluate sintered strength and hardness

Dimensional Control. Table 1 illustrates typical dimensional tolerances of P/M materials Separate tolerances apply to sintered, as-sized, and as-heat treated conditions For concentricity between an inside diameter and an outside diameter, a total indicator reading of 0.075 mm (0.003 in.) is permitted The distance between holes can be as great as 0.075 mm + 0.013 mm/mm (0.003 in + 0.0005 in./in.) Gears can be molded to American Gear Manufacturers Association (AGMA) class

as-7, which is limited primarily by the concentricity of the bore to pitch line If gears are held on the pitch line and bored more concentrically, AGMA class 10 or 11 is achieved

Table 1 Typical P/M tolerances (other than length)

Copper alloy steel ±0.038 ±0.0015 ±0.025 ±0.001 ±0.038 ±0.0015

Nickel alloy steel ±0.038 ±0.0015 ±0.025 ±0.001 ±0.038 ±0.0015

Stainless steel ±0.025 ±0.001 ±0.013 ±0.0005

Note: Up to 12.7 mm (0.500 in.) Length tolerance, ±0.102 mm (±0.004 in.), unless machined or ground Source: Ref 1

Other processes, such as P/M hot forging, injection molding, and high-temperature sintering, produce wider tolerances than presented in Table 1 Powder metallurgy forged dimensional tolerances are given in Table 2 High-temperature sintering tolerances are given in Table 3 Injection-molded tolerances range from 0.075 to 0.10 mm/mm (0.003 to 0.004 in./in.), even though parts have experienced 12 to 15% linear shrinkage (Ref 5)

Table 2 Tolerances on P/M forged parts

Nominal dimension Tolerance Parameter

Table 3 Dimensional tolerances of parts in the as-high-temperature sintered condition

Nominal dimension Tolerance Material

(b) Inside diameter sintered against a mandrel

References cited in this section

1 P/M Design Guidebook, Metal Powder Industries Federation, 183, p 15

2 "Anchor MH100 Standard Molding Powder," Hoeganaes Corp., Riverton, NJ

3 "Aromet 28, Sintered Properties of P/M Copper Steels," Quebec Metal Powders Ltd., Sorel, Quebec, Canada

4 "Controlled Dimensional Change," SCM Metal Products, Cleveland

5 L Pease III, "An Assessment of Powder Metallurgy Today, and Its Future Potential," Paper No 831042, Passenger Car Meeting, Society of Automotive Engineers, Warrendale, PA, 1983

Testing and Evaluation of Powder Metallurgy Parts

Measurement of Density

Density is the ratio of mass to volume For a given material, degree of sintering, and heat treatment, density determines mechanical and physical properties For example, higher density in sintered steels results in higher tensile strength, elongation, and impact resistance values As-pressed, or green, density also influences growth or shrinkage that occurs during sintering With nonuniform green density, parts grow or shrink nonuniformly, as in a thin-walled bronze bearing with a low-density region equidistant from the ends This results in a significantly smaller diameter at midlength than at the ends and necessitates repressing or sizing for close dimensional control

If cubes or right cylinders could be extracted from actual parts, linear dimensions could be measured and volume could be calculated easily From the weight of a part, density can be easily calculated This yields a value that, under ideal conditions, differs by 0.04 g/cm3 (0.5%) from a reference (Ref 6) Unless the sintered part is directly molded to an easily measured shape,

such as a transverse-rupture bar (31.8 by 12.7 by 6.4 mm, or 1.25 by 0.50 by 0.25 in.), this method of measuring linear dimensions is used infrequently

Methods Based on Archimedes' Principle. Typical methods of measuring density depend on Archimedes' principle, in which hydrostatic forces in liquids exert buoyant forces proportional to the part volume This measurement is standardized in ASTM B 328 (Ref 7), MPIF test method 42 (Ref 8), and International Standards Organization test method ISO 2738 (Ref 9) When an object is immersed in a liquid, the liquid exerts an upward buoyant force that is equal to the product of the object volume and the density of the liquid The difference in weight between an object weighed in air and its weight when suspended in water is equal to the object volume in cubic centimeters times the density of water Approximating the density

of water as unity:

V = Wair - Wwater

where V is the volume, cm3; Wair is the weight in air, g; and Wwater is the weight of object suspended in water less the weight

of the suspending wire in water (tare), g Density in g/cm3 is then:

Density = Wair/(Wair - Wwater)

For unsintered materials molded with 0.75% lubricant, pores are well sealed, and water cannot penetrate For such parts, the above calculation is suitable It is also suitable for materials with pores that are sealed off from the surface (materials close to theoretical density) For most sintered materials that are 70 to 95% dense, water tends to infiltrate the pores during weighing

in water This minimizes the buoyancy effect of the water (that is, the liquid is acting on a smaller volume) and results in an erroneous calculation of low volume

This low volume then causes an erroneously high density value Infiltration of water into pores usually is accompanied by air bubbles escaping from the part If the part is blotted to remove surface water and reweighed in air after weighing in water, any weight gain indicates that water has entered the pores Although not a standard procedure, volume can be approximated

as the weight in air after removing the part from the water, minus the weight in water

To prevent infiltration of water, all three standard test methods require that the pores of the part be filled with oil Oil impregnation is done after the part is weighed in air; this is carried out under vacuum or by immersion in hot oil Oil prevents the water from entering the pores The volume of the part is then determined as the part weight in air with oil in the pores, minus the weight of the oiled part suspended in water Care should be taken to select an oil that is not soluble in water or not soluble in water plus wetting agent Such oils also must exhibit superior demulsibility

The precision of the ISO method is ±0.25%, regardless of sample density, and assumes a water density of 0.997 g/cm3 Moyer (Ref 6) has reviewed the literature on precision methods of density determination (Ref 10, 11, 12, 13, 14, 15, 16) and has devised a method that provides accuracy to two or three decimal places, depending on sample porosity The basic measuring apparatus is shown in Fig 6 Requirements of precision density measurement include:

• Balance capable of measuring to the nearest 0.0001 g

• Vibration- and draft-free atmosphere

• Measurement of the density of the immersing liquid (water) by checking the density of a substance of accurately known density (four decimal places)

• Conversion of all densities back to 20 °C (68 °F) by compensating for thermal expansion of the sintered part

• Maintenance of liquid level at a constant height on the suspending wire

• Careful brushing of all bubbles from the test object

Fig 6 Density measurement apparatus

Using the above procedures, Moyer reports standard deviations of 0.0130 to 0.0005 g/cm3 on 17 g parts with densities ranging from 5.12 to 7.85 g/cm3, respectively

To determine density variation from one point to another in a complex part, the available samples must be considerably <17

g According to ASTM B 328, a minimum sample of 2 g is recommended, because a relatively high error rate results from measuring small samples Table 4 shows density errors that can occur because of weighing errors, if all weighing errors are assumed to occur so as to maximize density errors A 1 g sample on a balance accurate to 0.01 g (per ASTM B 328) could result in a density range from ±1.05 g/cm3 to a mean of 6.5 g/cm3

Table 4 Effect of sample size and weighing errors on density measurement

Density range, g/cm3, for a

sample weight (at 6.50 g/cm3) of:

• Remove oil and cutting fluids from the pores by Soxhlet extraction or heating in air or atmosphere to 315

°C (600 °F)

• Impregnate the pores with epoxy resin; generally does not fill all the pores, and polishing is required

• Mount the sample in epoxy, bakelite, or other suitable medium

• Wet grind through 600 mesh silicon carbide-coated paper