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

Подождите немного. Документ загружается.

Fig. 8

Voxel reconstruction of a turbine blade derived from computed tomography slices showing internal

cooling air passage surfaces outlined in blue (T.N. Claytor, Los Alamos National Laboratory)

Fig. 9 Three-dimensional tomographic end view of a failed nuclear fuel rod

bundle showing white insulation

surrounding rods cut away (with threshold operation) at an angle to expose relative positions of orange rods

(T.N. Claytor, Los Alamos National Laboratory)

Fig. 10 Representations of two-dimensional data in a three-

dimensional format. (a) Ultrasonic image of defects

in circuit board pads. (b) C-scan of a filament-wound vessel with defects. (T.N.

Claytor, Los Alamos National

Laboratory)

Fig. 11 30-MHz SLAM acoustic image of a Kevlar-reinforced composite test coupon with

delamination and

cracks. The damage was induced by cutting the bar perpendicular to fibers. The extent of the damage is shown

by the ingression of color, with red/green indicating the most extensive damage. The colors are associated with

levels of acousti

c transmission. A high level (white) indicates a homogeneous material, while a lower level

(red/green) indicates discontinuities. (L.W. Kessler, Sonoscan, Inc.)

Fig. 12 15-MHz C-SAM reflection mode image of a plastic-

encapsulated integrated circuit showing defects in

the lead frame. In this image, the echo polarity information is presented by color scale changes in which

disbond

s are red (negative polarity) and good bonds (positive echoes) are blue. Without this polarity

information, it is difficult to make a correct interpretation of the bonding condition in this sample (see Fig. 32 in

the article "Acoustic Microscopy"). (L.W. Kessler, Sonoscan, Inc.)

Fig. 13 50-MHz C-SAM reflection mode image of the same die-

attach sample shown in Fig. 29 and 30 in the

article "Acoustic Microscopy." In this image, areas of disbond are indicated by the color red.

(L.W. Kessler,

Sonoscan, Inc.)

Fig. 14 30-

MHz SLAM acoustic micrograph of the carbon fiber reinforced polymer component shown in Fig. 13

of "Acoustic Microscopy." Long, irregular defect (arrow) corresponds to blistering of one of the laminates.

(L.W.

Kessler, Sonoscan, Inc.)

Fig. 15 100-MHz SLAM micrograph of the alumina panel shown in Fig. 16 in the article "Acoustic Microscopy."

Color is used to delineate a crack with a specific acoustic brightness level. (L.W. Kessler, Sonoscan, Inc.)

Fig. 16 15-MHz reflection mode C-

SAM images of internal defects in the alumina test disk shown in Fig. 18 in

the article "Acoustic Microscopy." In (a), the darker features correspond to higher-

magnitude echo levels than

the yellow. In (b), the focus and gate levels were changed slightly. Because the defects are not parallel to the

surface of the part, echoes will not return at different times; therefore, a fixed-

gate position will show echoes

only at the depth selected. (L.W. Kessler, Sonoscan, Inc.)

Fig. 17 Thermal images of graphite-

epoxy sheets with 20 × 20 mm implanted Teflon defects at 0.3 mm depth

(a) and 2.25 mm depth (b). Image processing (c) improved the detectability of the 2.25 mm depth defect.

(P.

Cielo, National Research Council of Canada)

Fig. 18 Thermal analysis of a silicon mass airflow sensor. (a) Heater resisto

r (arrow) under analysis, 50×. (b)

and (c) Thermal maps indicating a 100 °C rise in temperature of an exposed heat resistor chip under no-

flow

conditions and normal voltage (b) and under 150% voltage (c). (R.W. Gehman, Micro Switch).

Fig. 19

Emissivity display of a GaAs FET semiconductor chip. Color scale at right represents 16 emissivity levels

between lowest (black) and highest (white) values. (J.O. Brown, EDO Corp.)

Fig. 20 Temperature display of a GaAs FET device showing a substrate bonding flaw as an oval-

shaped

anomaly (yellow) across center. Area displayed is 0.6 mm across. (J.O. Brown, EDO Corp.)

Fig. 21 50-MHz surface wave ultrasonic image of scratch and lapping marks in Si

. The smallest visible

scratches are 25 to 50 m wide. (T. Nelligan, Panametrics Inc.)

Fig. 22 Thermal imaging of two simulated defects (Teflon inserts) in a graphite-

epoxy laminate (D. Mauro,

Techmarketing Inc./AGEMA Infrared Systems)

Fig. 23 Ultrasonic image of an impacted region in a 1.4 mm thick graphite-reinforced

polymer sheet. The

damaged region is 6 mm long. (M.C. Bhardwaj, Ultron Laboratories, Inc.)

Fig. 24

Photoelastic fringe patterns produced when a photoelastic material is strained and viewed under

polarized light (a) and the photoelastic color sequence viewed with a circular polariscope (b). N

, fringe order

phase shift. (L.D. Lineback, Measurements Group, Inc.)

Nondestructive Inspection of Forgings

Introduction

IN FORGINGS of both ferrous and nonferrous metals, the flaws that occur most often are caused by conditions that exist

in the ingot, by subsequent hot working of the ingot or the billet, and by hot or cold working during forging. The

nondestructive inspection (NDI) methods most commonly used to detect these flaws include visual, magnetic particle,

liquid penetrant, ultrasonic, eddy current, and radiographic inspection. This article discusses the applications of these

methods to forgings. Information on the equipment and techniques used in these inspection methods is available in the

articles so titled in this Volume.

Nondestructive Inspection of Forgings

Flaws Originating in the Ingot

Many large open-die forgings are forged directly from ingots. Most closed-die forgings and upset forgings are produced

from billets, rolled bar stock, or preforms. Many, though by no means all, of the imperfections found in forgings can be

attributed to conditions that existed in the ingot, sometimes even when the ingot has undergone primary reduction prior to

the forging operation. Some, but again by no means all, of the service problems that occur with forgings can be traced to

imperfections originating in the ingot.

Chemical Segregation. The elements in a cast alloy are seldom distributed uniformly. Even unalloyed metals contain

random amounts of various types of impurities in the form of tramp elements or dissolved gases; these impurities are also

seldom distributed uniformly. Therefore, the composition of the metal or alloy will vary from location to location.

Deviation from the mean composition at a particular location in a forging is termed segregation. In general, segregation is

the result of solute rejection at the solidification interface during casting. For example, the gradation of composition with

respect to the individual alloying elements exists from cores of dendrites to interdendritic regions. Segregation therefore

produces a material having a range of compositions that do not have identical properties.

Forging can partially correct the results of segregation by recrystallizing or breaking up the grain structure to promote a

more homogeneous substructure. However, the effects of a badly segregated ingot cannot be totally eliminated by

forging; rather, the segregated regions tend to be altered by the working operation, as shown in Fig. 1.

In metals, the presence of localized regions that deviate from the

nominal composition can affect corrosion resistance, forging and

joining (welding) characteristics, mechanical properties, fracture

toughness, and fatigue resistance. In heat-treatable alloys,

variations in composition can produce unexpected responses to

heat treatments, which result in hard or soft spots, quench cracks,

or other flaws. The degree of degradation depends on the alloy and

on process variables. Most metallurgical processes are based on an

assumption that the metal being processed is of a nominal and

reasonably uniform composition.

Ingot Pipe and Centerline Shrinkage. A common

imperfection in ingots is the shrinkage cavity, commonly known as

pipe, often found in the upper portion of the ingot. Shrinkage

occurs during freezing of the metal, and eventually there is

insufficient liquid metal near the top end to feed the ingot. As a

result, a cavity forms, usually approximating the shape of a

cylinder or cone--hence the term pipe. Piping is illustrated in Fig.

Fig. 1

Microstructural bonding due to chemical

segregation and mechanical working

2. In addition to the primary pipe near the top of the ingot, secondary regions of piping and centerline shrinkage may

extend deeper into an ingot (Fig. 3).

Fig. 2 Schematic showing piping in top-poured ingots

Primary piping is generally an economic concern, but if it extends sufficiently deep

into the ingot body and goes undetected, it can eventually result in a defective

forging. Detection of the pipe can be obscured in some cases if bridging has

occurred.

Piping can be minimized by pouring ingots with the big end up, by providing risers

in the ingot top, and by applying sufficient hot-top material (insulating refractories

or exothermic materials) immediately after pouring. These techniques extend the

time that the metal in the top regions of the ingot remains liquid, thus minimizing

the shrinkage cavity produced in this portion of the ingot.

On the other hand, secondary piping and centerline shrinkage can be very

detrimental because they are harder to detect in the mill and may subsequently

produce centerline defects in bar and wrought products. Such a material condition

may indeed provide the flaw or stress concentrator for a forging burst in some later

processing operation or for a future product failure.

High Hydrogen Content. A major source of hydrogen in certain metals and

alloys is the reaction of water vapor with the liquid metal at high temperatures. The

water vapor may originate from the charge materials, slag ingredients and alloy

additions, refractory linings, ingot molds, or even the atmosphere itself if steps are

not taken to prevent such contamination. The resulting hydrogen goes into solution

at elevated temperatures; but as the metal solidifies after pouring, the solubility of

hydrogen decreases, and it becomes entrapped in the metal lattice.

Hydrogen concentration in excess of about 5 ppm has been associated with flaking,

especially in heavy sections and high-carbon steels. Hydrogen flakes (Fig. 4) are

small cracks produced by hydrogen that has diffused to grain boundaries and other

preferred sites, for example, inclusion/matrix interfaces. However, hydrogen

concentrations in excess of only 1 ppm have been related to the degradation of

mechanical properties in high-strength steels, especially ductility, impact behavior,

and fracture toughness.

Fig. 3

Longitudinal section

rough an ingot showing

extensive centerline shrinkage

Fig. 4

Hydrogen flaking in an alloy steel bar. (a) Polished cross section showing cracks due to flaking. (b)

Fracture surface containing hydroge

n flakes. Note the reflective, faceted nature of the fracture. (c) SEM

micrograph showing the intergranular appearance of the flakes in this material

Metals can also possess a high hydrogen content without the presence of flakes or voids. In this case, the hydrogen may

cause embrittlement of the material along selective paths, which can drastically reduce the resistance of a forged part to

crack propagation resulting from impact loading, fatigue, or stress corrosion.

In cases where hydrogen-related defects can serve as the initiation site for cracking and thus increase the likelihood of

future failures, it is advisable to use a thermal treatment that can alleviate this condition. For example, slow cooling

immediately following a hot-working operation or a separate annealing cycle will relieve residual stresses in addition to

allowing hydrogen to diffuse to a more uniform distribution throughout the lattice and, more important, to diffuse out of

the material.

Nonmetallic inclusions, which originate in the ingot, are likely to be carried over to the forgings, even though several

intermediate hot-working operations may be involved. Also, additional inclusions may develop in the billet or in

subsequent forging stages.

Most nonmetallic inclusions originate during solidification from the initial melting operation. If no further consumable-

remelting cycles follow, as in air-melted or vacuum-induction products (with no remelting cycle to follow), the size,

frequency, and distribution of the nonmetallic inclusions will not be altered or reduced in size or frequency during further

processing. If a subsequent vacuum-remelting operation is used, the inclusions will be lessened in size and frequency and

will become more random in nature. If an electroslag-remelting cycle is used, a more random distribution of inclusions

will result.

Two kinds of nonmetallic inclusions are generally distinguished in metals:

•

Those that are entrapped in the metal inadvertently and originate almost exclusively from particles of

matter that are occluded in the metal while it is molten or being cast

• Those that separate from the metal because of a change in temperature or composition

Inclusions of the latter type are produced by separation from the metal when it is in either the liquid or the solid state.

Oxides, sulfides, nitrides, or other nonmetallic compounds form droplets or particles when these compounds are produced

in such amounts that their solubility in the matrix is exceeded.

Air-melted alloys commonly contain inclusions mainly of these chemical characteristics. Vacuum- or electroslag-

remelted alloys more commonly contain conglomerates of any of these types, frequently combined with carbon or the

hardening element or elements that precipitate during stabilization and aging cycles to form inclusions such as titanium

carbonitrides or carbides. Homogenizing cycles are normally used for the ingot prior to conversion or at an early stage of

conversion.

Because these compounds are products of reactions within the metal, they are normal constituents of the metal, and

conventional melting practices cannot completely eliminate such inclusions. However, it is desirable to keep the type and

amount of inclusions to a minimum so that the metal is relatively free from those inclusions that cause the most problems.

Of the numerous types of flaws found in forgings, nonmetallic inclusions appear to contribute significantly to service

failures, particularly in high-integrity forgings such as those used in aerospace applications. In many applications, the

presence of these inclusions decreases the ability of a metal to withstand high static loads, impact forces, cyclical or

fatigue loading, and sometimes corrosion and stress corrosion. Nonmetallic inclusions can easily become stress

concentrators because of their discontinuous nature and incompatibility with the surrounding composition. This

combination may very well yield flaws of critical size that, under appropriate loading conditions, result in complete

fracture of the forged part.

Unmelted electrodes and shelf are two other types of ingot flaws that can impair forgeability. Unmelted electrodes

(Fig. 5a) are caused by chunks of electrodes being eroded away during consumable melting and dropping down into the

molten material as a solid. Shelf (Fig. 5b) is a condition resulting from uneven solidification or cooling rates at the ingot

surfaces.

Fig. 5 Sections through two heat-

resistant alloy ingots showing flaws that can impair forgeability. (a) Piece of

unmelted consumable electrode (white spot near center). (b) Shelf (black line along edge) resulting from

uneven solidification of the ingot

The consumable-melting operation has occasionally been continued to a point where a portion of the stinger rod is melted

into the ingot, which may be undesirable because the composition of the stinger rod may differ from that of the alloy

being melted. To prevent this occurrence, one practice is to weld a wire to the stinger rod, bend the wire down in tension,

and weld the other end of the wire to the surface of the electrode a few inches below the junction of the stinger rod and

the electrode. When the electrode has been consumed to where the wire is attached to it, the wire is released and springs

out against the side of the crucible, thus serving as an alarm to stop the melting. A disadvantage of this practice is that the

wire may become detached and contaminate the melt.

Nondestructive Inspection of Forgings

Flaws Caused by the Forging Operation