ASM Metals HandBook Vol. 17 - Nondestructive Evaluation and Quality Control

Подождите немного. Документ загружается.

Fig. 5

Statistical summary report showing the mean of measured observations, the blueprint specification for

the mean, the differe

nce between specified and measured means, the tolerance, the standard deviation of the

measured dimensions, and the capability of the process. These calculations are performed for all measured

dimensions on the part.

Histograms. An alternative method of assessing capability involves the use of a histogram, or frequency plot. This is a

graph that plots the number of occurrences within successive, equally spaced ranges of a given measured dimension.

Figure 6 shows an example output report of this type. A graph such as this, which has superimposed upon it the tolerance

limits for the dimension being analyzed, allows a quick, qualitative evaluation of the variation and capability of the

process. It also allows the normality of the process to be judged through comparison with the expected bell-shaped curve

of a normal process.

Fig. 6 Frequency plot for one measured dimension showing the distribution of the measurements

Other Applications for Computer-Aided Inspection. The general sequence described for semiautomatic

dimensional inspection can be applied to a number of other inspection criteria. Examples would include pressure testing

or defect detection by electronic vision systems. The statistical analysis of scrap by defect types is also very helpful in

identifying problem areas. In some cases, direct data input to a computer may not be feasible, but the benefits of entering

data manually into a statistical analysis program should not be overlooked. The computer allows rapid analysis of large

amounts of data so that statistically significant trends can be detected and proper attention paid to appropriate areas for

improvement. The benefits and costs of each anticipated application of automation to a particular situation, as well as the

feasibility of applying state-of-the-art equipment, need to be studied as thoroughly as possible prior to implementation.

Nondestructive Inspection of Castings

By the ASM Committee on Nondestructive Inspection of Castings

Liquid Penetrant Inspection

Liquid penetrant inspection essentially involves a liquid wetting the surface of a workpiece, flowing over that surface to

form a continuous and uniform coating, and migrating into cracks or cavities that are open to the surface. After a few

minutes, the liquid coating is washed off the surface of the casting, and a developer is placed on the surface. The

developer is stained by the liquid penetrant as it is drawn out of the cracks and cavities. Liquid penetrants will highlight

surface defects so that detection is more certain. Details of this method can be found in the article "Liquid Penetrant

Inspection" in this Volume.

Liquid penetrant inspection should not be confined to as-cast surfaces. For example, it is not unusual for castings of

various alloys to exhibit cracks (frequently intergranular) on machined surfaces. A pattern of cracks of this type may be

the result of intergranular cracking throughout the material because of an error in composition or heat treatment, or the

cracks may be on the surface only as a result of machining or grinding. Surface cracking may result from insufficient

machining allowance, which does not allow for complete removal of imperfections produced on the as-cast surface, or it

may result from faulty machining techniques. If imperfections of this type are detected by visual inspection, liquid

penetrant inspection will show the full extent of such imperfections, will give some indication of the depth and size of the

defect below the surface by the amount of penetrant absorbed, and will indicate whether cracking is present throughout

the section.

Liquid penetrant inspection is sometimes used in conjunction with another nondestructive method. Such is the case in the

following example.

Example 1: Tail Rotor Gearbox Failure on the H-53 Helicopter.

The H-53 tail rotor gearbox mounting lugs experienced failures that resulted in at least two aircraft accidents. The

gearbox is a magnesium casting, either AZ-91 or ZE-41. Initially, only visual inspection using a 10× magnifying glass

was required. Initiation sites for the cracking were found to be localized in a few areas adjacent to the attached lugs (Fig.

7). A manual eddy current inspection was developed that concentrated on the potential initiation sites. If an indication was

found, then a fluorescent penetrant inspection was performed as a backup inspection.

Fig. 7 Helicopter transmission mounting lug. Note: bolt is shown removed for clarity.

Bolt is not removed for

eddy current inspection.

Inspection Procedure. An eddy current tester with a flawgate and shielded probes was used to perform the

inspection. The instrument is calibrated using standard magnesium alloy (either AZ-91 or ZE-41) calibration blocks. The

tester is calibrated on a 0.50 mm (0.020 in.) slot to achieve a minimum 150 A deflection. The flawgate is calibrated to

activate for meter deflection greater than 150 A. The part is scanned very slowly over all critical areas. As an aid for the

inspection of the lug edges, a wooden toothpick is taped to the probe to maintain a constant probe-to-edge distance. A

positive indication--meter deflection greater than 150 A--requires a fluorescent penetrant inspection.

In preparation for the fluorescent penetrant inspection, the paint in the area to be inspected is removed. The required

penetrant material sensitivity is level II or III and is used with a halogenated solvent remover and non-aqueous developer.

Clean cloths moistened with solvent cleaner are required for the cleaning step. The inspection area is never flushed with

solvent. An initial inspection is required before the developer is applied. If no indications are noted, developer is applied.

When an indication is noted, it is wiped with a cotton swab moistened with solvent, and the area is inspected for the

indication bleed-back. A 30-min penetrant dwell time is required. The minimum required dwell time between cleaning

and application of the developer is 10 min before inspection. The minimum developer dwell time is half the penetrant

dwell time, and a maximum developer dwell time is not to exceed the penetrant dwell time.

After the inspection, all developer residue is removed from the mounting lug. Areas of greatest concern include the mount

pad radius areas, lug flanges along edges, and the mounting pad. After completion of penetrant inspection, the finishes on

the mounting lug are replaced in such a way that the eddy current inspection is not hindered and the visual inspection is

enhanced. The visual inspection is enhanced by painting the lugs light gray or white.

Conclusions. These inspections were performed during scheduled inspection intervals until magnesium alloy castings

were replaced with A356.0 aluminum alloy castings.

Nondestructive Inspection of Castings

By the ASM Committee on Nondestructive Inspection of Castings

Magnetic Particle Inspection

Magnetic particle inspection is a highly effective and sensitive technique for revealing cracks and similar defects at or just

beneath the surface of castings made of ferromagnetic metals. The capability of detecting discontinuities just beneath the

surface is important because such cleaning methods as shot or abrasive blasting tend to close a surface break that might

go undetected in visual or liquid penetrant inspection.

When a magnetic field is generated in and around a casting made of a ferromagnetic metal and the lines of magnetic flux

are intersected by a defect such as a crack, magnetic poles are induced on either side of the defect. The resulting local flux

disturbance can be detected by its effect on the particles of a ferromagnetic material, which become attracted to the region

of the defect as they are dusted on the casting. Maximum sensitivity of indication is obtained when a defect is oriented in

a direction perpendicular to the applied magnetic field and when the strength of this field is sufficient to saturate the

casting being inspected.

Equipment for magnetic particle inspection uses direct or alternating current to generate the necessary magnetic fields.

The current can be applied in a variety of ways to control the direction and magnitude of the magnetic field.

In one method of magnetization, a heavy current is passed directly through the casting placed between two solid contacts.

The induced magnetic field then runs in the transverse or circumferential direction, producing conditions favorable to the

detection of longitudinally oriented defects. A coil encircling the casting will induce a magnetic field that runs in the

longitudinal direction, producing conditions favorable to the detection of circumferentially (or transversely) oriented

defects. Alternatively, a longitudinal magnetic field can be conveniently generated by passing current through a flexible

cable conductor, which can be coiled around any metal section. This method is particularly adaptable to castings of

irregular shape. Circumferential magnetic fields can be induced in hollow cylindrical castings by using an axially

disposed central conductor threaded through the casting.

Small castings can be magnetic particle inspected directly on bench-type equipment that incorporates both coils and solid

contacts. Critical regions of larger castings can be inspected by the use of yokes, coils, or contact probes carried on

flexible cables connected to the source of current; this setup enables most regions of castings to be inspected. Equipment

and techniques associated with this method can be found in the article "Magnetic Particle Inspection" in this Volume.

Nondestructive Inspection of Castings

By the ASM Committee on Nondestructive Inspection of Castings

Eddy Current Inspection

Eddy current inspection consists of observing the interaction between electromagnetic fields and metals. In a basic

system, currents are induced to flow in the testpiece by a coil of wire that carries an alternating current. As the part enters

the coil, or as the coil in the form of a probe or yoke is placed on the testpiece, electromagnetic energy produced by the

coils is partly absorbed and converted into heat by the effects of resistivity and hysteresis. Part of the remaining energy is

reflected back to the test coil, its electrical characteristics having been changed in a manner determined by the properties

of the testpiece. Consequently, the currents flowing through the probe coil are the source of information describing the

characteristics of the testpiece. These currents can be analyzed and compared with currents flowing through a reference

specimen.

Eddy current methods of inspection are effective with both ferromagnetic and nonferromagnetic metals. Eddy current

methods are not as sensitive to small, open defects as liquid penetrant or magnetic particle methods are. Because of the

skin effect, eddy current inspection is generally restricted to depths less than 6 mm ( in.). The results of inspecting

ferromagnetic materials can be obscured by changes in the magnetic permeability of the testpiece. Changes in temperature

must be avoided to prevent erroneous results if electrical conductivity or other properties, including metallurgical

properties, are being determined.

Applications of eddy current and electromagnetic methods of inspection to castings can be divided into the following

three categories:

• Detecting near-

surface flaws such as cracks, voids, inclusions, blowholes, and pinholes (eddy current

inspection)

•

Sorting according to alloy, temper, electrical conductivity, hardness, and other metallurgical factors

(primarily electromagnetic inspection)

•

Gaging according to size, shape, plating thickness, or insulation thickness (eddy current or

electromagnetic inspection)

The following example demonstrates the use of eddy current inspection for the evaluation of an induction-hardened

surface on a gray iron casting.

Example 2: Eddy Current Inspection of Hardened Valve Seats on Gray Iron

Automotive Cylinder Heads.

Exhaust-valve seats on gray iron automotive engine cylinder heads were induction hardened to withstand use with lead-

free fuels. The seats were specified to be hardened to 50 HRC to a depth of 1.27 to 2.03 mm (0.050 to 0.080 in.). Failure

to achieve these depths could lead to excessive recession of the seat, causing premature failure.

To establish the feasibility of, and the conditions for, eddy current inspection of cylinder heads having valve seats of

different diameters, an adjustable probe was made using a 10 mm (0.400 in.) diam standard probe. The end was machined

to a 120° conical point, and the probe was mounted in a plastic holder. A scrap valve stem mounted in the holder was

used to position the probe on the valve seat. The probe was centered on the valve seat by moving the probe in its holder

and by the adjusting screw. After the inspection conditions were established, rugged units for use by production

inspectors were made of stainless steel with the protected probe in a fixed position to suit the cylinder-head seat.

In operation, the probe was brought into intimate contact with the hardened surface, and a meter reading was taken.

Positioning of the probe on the valve seat is shown in Fig. 8. The meter reading, taken from an instrument calibrated from

-100 to +100, was compared with a developed chart for case depth versus meter reading (a linear relationship). A meter

reading of 0 was established for 50 HRC at a depth of 1.27 mm (0.050 in.) and -40 for 2.03 mm (0.080 in.). The

instrumentation was set up to respond to A

(resistive component). Inspection was performed at a frequency of 2 kHz.

Extensive laboratory and production inspection has shown a standard deviation, of ±0.8 mm (±0.03 in.) and a

confidence limit of 3 for the total data generated.

Fig. 8 Eddy current inspection using an adjustable probe for determining case depth of an induction-

hardened

gray iron valve seat.

Nondestructive Inspection of Castings

By the ASM Committee on Nondestructive Inspection of Castings

Radiographic Inspection

Radiographic inspection is a process of testing materials using penetrating radiation from an x-ray generator or a

radioactive source and an imaging medium, such as x-ray film or an electronic device. In passing through the material,

some of the radiation is attenuated, depending on the thickness and the radiographic density of the material, while the

radiation that passes through the material forms an image. The radiographic image is generated by variations in the

intensity of the emerging beam.

Internal flaws, such as gas entrapment or nonmetallic inclusions, have a direct effect on the attenuation. These flaws

create variations in material thickness, resulting in localized dark or light spots on the image.

The term radiography usually implies a radiographic process that produces a permanent image on film (conventional

radiography) or paper (paper radiography or xero radiography), although in a broad sense it refers to all forms of

radiographic inspection. When inspection involves viewing an image on a fluorescent screen or image intensifier, the

radiographic process is termed filmless or real-time inspection. When electronic nonimaging instruments are used to

measure the intensity of radiation, the process is termed radiation gaging. Tomography, a radiation inspection method

adapted from the medical computerized axial tomography scanner, provides a cross-sectional view of a testpiece. All of

the above terms are primarily used in connection with inspection that involves penetrating electromagnetic radiation in

the form of x-rays or -rays. Neutron radiography refers to radiographic inspection using neutrons rather than

electromagnetic radiation.

The sensitivity, or the ability to detect flaws, of radiographic inspection depends on close control of the inspection

technique, including the geometric relationships among the point of x-ray emission, the casting, and the x-ray imaging

plane. The smallest detectable variation in metal thickness lies between 0.5 and 2.0% of the total section thickness.

Narrow flaws, such as cracks, must lie in a plane approximately parallel to the emergent x-ray beam to be imaged; this

requires multiple exposures for x-ray film techniques and a remote-control parts manipulator for a real-time system.

Real-time systems have eliminated the need for multiple exposures of the same casting by dynamically inspecting parts

on a manipulator, with the capability of changing the x-ray energy for changes in total material thickness. These

capabilities have significantly improved productivity and have reduced costs, thus enabling higher percentages of castings

to be inspected and providing instant feedback after repair procedures. Figures 9 and 10 show real-time digital

radiography images of automotive components.

Fig. 9

Digital radiography image of a die cast aluminum carburetor. Porosity appears as dark spots in the area

of the center bore, through the vertical center of the image. Courtesy of B.G. Isaacson, Bio-

Imaging Research,

Inc.

Fig. 10

Evaluation of cast transmission housing assembly. (a) Photograph of cast part. (b) Digital radiography

image used to verify the steel spring pin

and shuttle valve assembly through material thicknesses ranging from

3 mm ( in.) in the channels to 25 mm (1 in.) in the rib sections of the casting. Courtesy of B.G. Isaacson, Bio-

Imaging Research, Inc.

Advances. Several advances have been made to assist the industrial radiographer. These include the computerization of

the radiographic standard shooting sketch, which graphically shows areas to be x-rayed and the viewing direction or angle

at which the shot is to be taken, and the development of microprocessor-controlled x-ray systems capable of storing

different x-ray exposure parameters for rapid retrieval and automatic warm-up of the system prior to use. The advent of

digital image-processing systems and microfocus x-ray sources (near point source), producing energies capable of

penetrating thick material sections, have made real-time inspection capable of producing images equal to, and in some

cases superior to, x-ray film images by employing geometric relations previously unattainable with macrofocus x-ray

systems. The near point source of the microfocus x-ray system virtually eliminates the edge unsharpness associated with

larger focus devices.

Digital image processing can be used to enhance imagery by multiple video frame integration and averaging techniques

that improve the signal-to-noise ratio of the image. This enables the radiographer to digitally adjust the contrast of the

image and to perform various edge enhancements to increase the conspicuity of many linear indications (Fig. 11).

Fig. 11

Digital radiography images of an investment cast jet engine turbine blade showing detail through a

wide range in material thickness. The

trailing edge of the blade (along the top of the image) is 2 mm (0.080

in.) thick, the root section of the blade (to the far left in the image) is 19 mm (0.75 in.) thick, and the shelf

area (to the right of the root section) is 25 mm (1 in.) thick. The ima

ge shown in (a) is unprocessed; the image

in (b) is processed to subdue the background and to enhance edges and internal features.

Courtesy of B.G.

Isaacson, Bio-Imaging Research, Inc.

Interpretation of the radiographic image requires a skilled specialist who can establish the correct method of exposing the

castings with regard to x-ray energies, geometric relationships, and casting orientation and can take all of these factors

into account to achieve an acceptable, interpretable image. Interpretation of the image must be performed to establish

standards in the form of written or photographic instructions. The inspector must also be capable of determining if the

localized indication is a spurious indication, a film artifact, a video aberration, or a surface irregularity. The article

"Radiographic Inspection" in this Volume provides additional information on real-time digital systems.

Computed tomography, also known as computerized axial tomography (or CAT scanning), is a more sophisticated x-

ray imaging technique originally developed for medical diagnostic use (Ref 2). It is the complete reconstruction by

computer of a tomographic plane, or slice of an object. A collimated (fan-shaped) x-ray beam is passed through a section

of the part and is intercepted by a detector on the other side. The part is rotated slightly, and a new set of measurements is

made; this process is repeated until the part has been rotated 180°.

The resulting image of the slice (tomogram) is formed by computer calculations based on electronic measurement (digital

sampling) of the radiation transmitted through the object along different paths during the rotating scan. The data thus

accumulated are used to compute the densities of each point in the cross section, enabling the computer to reconstruct a

two-dimensional visual image of the slice.

The shapes of internal features are determined by their computed densities. After one slice is produced through a

complete rotation, either the part or the radiation source and detector can be moved and a three dimensional image built

up through the scanning of successive slices.

Compared to electronic radiography, computed tomography provides increased sensitivity and detection capabilities. The

contrast resolution of a good-quality tomographic image is 0.1 to 0.2%, which is approximately two orders of magnitude

better than with x-ray film.

In addition, images are produced in a quantitative, ready-to-use digital format (Ref 3). They provide detailed physical

information, such as size, density, and composition, to aid in evaluating defects. Methods are being developed to use this

information to predict failure modes or system performance under operating loads. The data can also be easily

manipulated to obtain various types of images, to develop automated flaw detection techniques, and to promote efficient

archiving. Figures 12 and 13 illustrate the uses of computed tomography for examining castings. More detailed

information can be found in the article "Industrial Computed Tomography" in this Volume.

Fig. 12 Computed tomographic images of a die cast alumi

num automotive piston. (a) Photograph of cast part.

(b) Vertical slice through the piston shows porosity as dark spots in the crown area (point A) and

counterbalance area (point B). (c) Transverse slice through the crown of the piston verifies the porosity

; the

smallest void that is visible is 0.4 mm (0.016 in.) in diameter. (d) Transverse slice through the counterbalance

area also verifies porosity. Dimensional analysis of the piston walls is possible to an accuracy of ±50

(±0.002 in.). Courtesy of B.G. Isaacson, Bio-Imaging Research, Inc.

Fig. 13

Use of computed tomography for examining automotive components. (a) Photograph of a cast

aluminum transmission case with (b) corresponding tomographic image. (c) Two three-

dimensional images of a

cast aluminum cylinder head generated from a set of continuous tomographic scans used to view water-

cooling

chambers where a leak had been detected. Courtesy of R.A. Armistead, Advanced Research and Applications

Corporation.