welding inspection

Trang 1

150 A SURVEY OF THE WELDING PROCESSES

2 American Welding Society, Welding Handbook, 3rd Ed New

York: 1949

3 The Linde Air Products Co., The Oxy-Acetylene Handbook, 3rd

Ed New York: 1943

4 Lincoln Electric Co., Procedure Handbook of Arc Welding, Design

and Practice, %h Ed Cleveland: 1950

5 The Welding Engincer Publishing Co., The Welding Encyclopedia,

12th Ed Chicago: 1947,

6 Chute, George M., Electronic Control of Resistance Welding, 1st

Ed New York: McGraw-Hill Book Company, Inc., 1943

7 General Electric Co., Arc Welding Manual, Schenectady: 1940

8 Conway, M J., “Inert-Gas Welding of Hard-to-Weld Metals,”

Welding Arcs, 15, 3 (June, 1949)

9 Pilia, F J., “Inert-Gas Shielded-Arc Spot Welding,” reprint from

The Welding Journal, 28, 5 { January, 1949)

10 Anderson, R J., “Heli-Arc Welding,” reprint from Canadian

Metals and Metallurgical Industries, 10, 14 (January, 1947)

11 Singleton, Robert C., “Electric-Arc Stud Welding,” reprint from

The Welding Journal, 26, 1095 (December, 1947)

12 Harris, H., Metal Arc Welding New York: Longmans, Green and

Co., 1935

—

TESTING AND INSPECTION

OF WELDS

Gencrally speaking, satisfactory use of a weld will usually depend on the ability of the associated personnel to predict, with reasonable accuracy, its final quality This can be done by making certain suitable tests prior to application Also it is paramount that the results of the application be checked, qualified, or otherwise judged

as to quality

After exhaustive tests have been made to establish the suitability

of the purposed constructional metals, tests must be made to qualify the process or processes and the operators Inspection usually involves the examination of completed welds to establish their quality and their conformation to specifications

Thus, testing and inspection determines whether or not the quality standards of materials and workmanship are being met Testing and inspection are closcly related and, in some instances, may considerably overlap

In this chapter an effort is made to describe the common tests applicable to welded joints and to point out some of the more important methods of maintaining quality standards

Societies concerned with the fabrication of structures or manu- facture of products, in most cases, have formulated testing and inspection standards applicable to their particular fields of interest

It is important to understand that the tests discussed herein and their an~“cations are of general nature To meet specific standards, appli odes should be followed

) ive tests are those tests in which the materials being

151

Trang 2

152 TESTING AND INSPECTION OF WELDS

examined are of such shape or condition after testing that they are

of no further constructional value Nondestructive tests have no

tendencies to reduce the usefulness of the test piece

The specimens for destructive testing may be taken either from

a completed joint in a welded structure or from a test: piece welded

under conditions closely simulating field or shop practice Such tests

are designed to represent, in some instances, the various types of

anticipated service loads

DESTRUCTIVE TESTS

Tensile Test

The tensile testing of welded specimens or coupons differs from

other tensile tests only in the manner in which they are taken In

TEST PLaTe `

Ce ` số

mm

+

—⁄ tt WELO

/ Fic 83 Test coupon should be taken from a representative locality

im the weld In the finished cougon the weld should be at the approxi-

mate center

order to obtain the tensile properties of a deposited weld the speci-

mens must be taken transversely to the joint, from a representative

region within the joint, and with the weld located in the approxi-

mate center of the coupon as in Fig 83

Although testing specifications vary widely, a common procedure

is to use a % in thick parent metal plate where the test involves

steel being welded with either the oxy-acetylene or metal are process

The plates welded together are machined into strips 144 in wide to

comp'ete the coupon Whether or not the reinforcement is removed

will depend upon the information sought Where strength compari-

sons between the parent metal and the reinforced weld are required,

no machining is done on the coupon other than that of reducing it

to the correct width Where strength comparisons are to be made

TESTING AND INSPECTION OF WELDS 153 between the parent metal and a similar cross section of weld metal, all reinforcements are removed In testing procedures requiring unit tensile strengths of the weld metal rather than comparisons, the coupon is reduced in width at the welding zone to assure its failure

at that point

‘Tensile specimens selected from a joint in a fabricated structure usually require machining to reduce them to standard dimensions since the thickness of the section in question may be other than that

of a standard test specimen

—

uLTmwate

STRENGTH ane amis

YIELO STRENGTH

STRENGTH

UNIT STRENGTH

ELASTIC ZONE

UNIT DEFORMATION

Fro 8+ Typical stress-strain curve

In addition to tensile strength, other information is obtained from the tensile test such as clongation, reduction in area, and yield strength

The tensile test essentially consists of subjecting the prepared coupon to a measured tensile load; deformation data are taken as the loading progresses By plotting unit load as the ordinate against unit deformation as the abscissa a curve is obtained Figure 84 is typical of such a curve

It is important to note that in tensile tests made on specimens in which the measured elongation includes that of both parent metal

and weld metal, the total unit deformation obtained is not likely to

be typical of either To obtain the true tensile properties of weld metal the procedure is to deposit beads or layers into the open side

of a section of angle shape (or other suitable shapes) to a depth

Trang 3

sufficient to allow the machining of a coupon of all-weld metal Thus

any tensile properties taken are not influenced by adjacent parent

metal

Tensile testing may be adapted to the determination of fillet-wel 4

shear strengths by fillet welding two butt straps to opposite sides of

two abutting test bars and loading as in the standard tensile tests

(see Fig 85) Properly proportiuned, failure will occur along the

fillet welds with a shearing action The total load divided by the

«2s & ;ˆ«

~ + - -— ©

` LES sumer _a* _

POR AU WELOS T7”

American Welding Society

Fic 85 Transverse fillet-weld shearing specimen

total length of the welds will give the unit shear strength in pounds

per linear inch of fillet weld

A modern tensile testing machine, fully mechanized, is illus-

trated in Fig 86 Some machines, suitable for loading small speci-

mens, are hand-operated The coupon is gripped by self-tightening

jaws on cither end or by screwing the threaded ends of the specimen

into threaded jaws To obtain deformation data, an extensometer

is mounted directly upon the coupon to indicate its extension unde:

the progressive load Before failure of the specimen the instrument

should be removed since it is obvious that the violence of the final

failure would cause damage

The tensile strength is calculated by dividing the maximum

total load by the original cross-sectional area in square inches Per-

cent elongation between gage marks is calculated by use of the fol-

lowing formula:

TESTING AND INSPECTION OF WELDS 155

final length — original length x 100 = percent elongation of the gage length

original length

Reduction in area is calculated by the following relationship:

original cross-sectional area final cross-sectional area

orginal cross-sectional area

100 = percent reduction in area

Tiwler Olsen Testing Machine Co

Fra 86 Tensile testing machine

A simple method for finding the yield strength is to locate it directly on a carefully plotted stress-strain curve similar to that

Trang 4

shown in Fig 84, The unit load at which extension occurs without

appreciable increase im the load is called the yield strength In low-

carbon steels it is identified by a hook or a flattened section in the

curve at the upper limit of the elastic range

Nick-Break Test

For preliminary testing, the nick-break test affords combined

convenience and economy The technique of making this test is

Iaponox Pase e0ces wa BE FLAME Ct

[P| ~

Loao

k) fw

Fin 87 (a) Nick-break specimen (b) Method of rupturing nick-breat

specimens,

comparatively simple A coupon representing the weld to be tested

is notched or nicked in the weld metal in such a manner that the

fracture is sharp and in the weld as shown in Fig 87 Breaking of

the notched coupon is obtained by securing one end in a jig or vise

and striking the end with a hammer Probably a safer practice would

be to support the two ends and conéentrate a steady press load in

the center, thus preventing possible injury to personnel by flying

test pieces,

The nick-break test is a continuity and homogencity test that

shows the number of gas pockets, their location and arrangement,

TESTING AND INSPECTION OF WELDS 157 degree of penetration, slag inclusions, and overlap, as well as some approximate information in regard to number of passes used and the grain size

On the basis of density being a rather dependable criterion of weld quality, the gas- and slag-pocket count and their size determination is probably the most significant information obtained from the nick-break test Although policies vary some, a standard often used

is that of approving a fractured coupon if the number of voids does

—

"5 MINIMUM LENGTH

et OF WELO+L.=t”

a

Ị all a

Le APPROX,

“te

(b

i

oo mage |

Fr 88 (a) Fillet-weld break specimen (b) Method of rupturing

Allet-weld break specimen

not exceed eight per square inch, if no single void has a dimension greater than Ye in., and if the largest combined dimensions of all voids do not exceed ¥ in

Fillet welds may be given a fracture test similar to the nick-break test, the principle difference being the manner in which they are broken The test is usually made by fillet welding two pieces of plate

of suitable length and thickness to form a T joint The weld is ruptured by pressing or hammering, the load being applied from a direction so as to cause the failure to be initiated at the weld root

(Fig 88).

Trang 5

The quality of the weld is revealed by visual examination similar

to that used in the nick-break test

Bend Tests

Bend tests are becoming increasingly popular since they do not

require expensive equipment for the rapid testing of welds, Also the

information derived from bend tests is generally thought to be more

indicative of the quality of the weld than tensile testing

Bend tests are of two general types—the guided bend and the

free bend

The guided-bend test is a bending test in which the coupon IS

forced to conform to the shape of a jig (Figs 89a und b) Guide

bend test disclose defects in the welded joint such as voids and lack

of penetration The preparation of these coupons is giv cn in relating

codes

Bend tests may be made either by bending against the face of

the weld or against the root side or both Root bends are particularly

valuable for testing the welded joint for root penetration and

porosity

Free-bend tests are methods by which the shape of the speci-

men is obtained without constraint (see Fig 90) Their most valu-

able use is in the determination of the plastic properties of the

welded joint

Preparation of the free-bend coupons is also given in relating

codes They are usually given some initial bend in a press to provide

columnar eccentricity, and all bending thereafter is obtained vy the

coupon acting as a column, Bending is continued until a surface

crack occurs between the gage marks If failure does not occur, the

coupon is bent until its ends meet Generally a failure is considered

to be a surface opening of Yo in in length or greater resulting from

the test Corner cracks, if not extensive, are not counted as failures

Ductility may be mathematically computed by dividing the

deformation between gage marks by the origina] gage distance and

multiplying by 100 The result is percent elongation These measure-

ments should be carefully made with a flexible scale, graduated in

hundredths of an inch, and a suitable magnifying glass

Ceorgie Inmitete of Tech eology

Fico 89a Guided-bend test pig and press

MARDENEO ROLLERS, 1Ý OIlAM

TAPPED MOLE TO SUIT MAY BE SUBSTITUTED FOR

TESTING MACHINE - > JIG SHOULDERS, =

'

fo RECURRED Sy” \ F AS REQUIRED —

\ SMOULDERS HARDENED

_— > jy”

PLUNGER 4 MEM0e® “ - 4%

3U

a

~- ar -

oe

wewrrere

— ®ˆ——>*

American Welding Soriety

Fico 89b, Guided-bend test sig.

Trang 7

Hardness Tests

Hardness testing methods usually embrace the surface indenta-

tion principle, whereby a penetrator of given size and shape is forced

into the surface of the materials being tested by a given load The

degree of hardness is indicated by the depth penetrated or by the

Frio 93 Shore scleroscope

area of the impression The Brinel] and Rockwell hardness testers

operate on this principle

The Shore scleroscope (Fig 93) operates on the principle of a

bouncing hammer or shaft When the shaft is lifted and allowed to

fall through a fixed distance agaist the surface of the metal being

tested, it rebounds to a height, the extent of which depends upon

the hardness of the tested metal Measurement of the rebound of the

hammer gives the hardness value

TESTING AND INSPECTION OF WELDS 163 The Brinell tester consists of a suitable loading press and the ball penctrator The 10-mm diameter hardened ball is forced into the surface of the test metal under a load of 3000 kg, allowed to remain

i) sec and released Hardness values are cak ulated by the folk winec

formula:

Brincll number = Vi

where P = load,

A = areca of indentation, and

D = diameter of bali

Brincll hardness numbers are usually taken directly from con

1 «

vern:cnt chart!A

Thc Reckwcl Incas tester makes use of both a diamond cone

; f “| balls: the former is used for the hardest

and various sizcs ol eccl balls: the former is use r the

é

f 94 Pockweil Aardaess tester

materials A minor load of 10 kg is applied for the purpose of pene trating any surface covering such as paint and scale A major load of

150 ke (C scale) is then applied for the actual hardness test The measured depth of penetration induced by the major load is mechanically transmitted to an indicator dial, calibrated in Rockwell values (Fig 94

Trang 8

For testing soft materials, such as nonferrous metals, a 4¢-in di-

ameter steel ball with 100 kg (B scale) of major load is used Other

penetrators and loads are provided to encompass all hardness values

encountered in the testing of commercial metals

Generally speaking, hardness testing is confined to the laboratory,

although in recent years some excellent portable instruments have

been offered to the market

Fatigue Tests

This test is designed to determine the suitability of a metal for

transmitting cyclic, repetitious, or intermittent types of loads

Failures resulting from this type of load are of a progressive na-

ture and give little warning of their ultimate failure Generally,

cracks of this type emanate from sharp notches where stresses tend

to concentrate The most common example of a stress riser or of

a notch in welding practice is an undercut alongside a weld

To simulate service loads, three common types of tests are used

They are

1 The rotating beam supported on the two ends

2 The rotating cantilever beam i

3 The reciprocating flexure beam i!

In the rotating-beam types the test procedure is ta load the

beams to a predetermined skin fiber unit stress By starting from a

given position, if the beam is rotated through 180 deg) the stress

will be reversed; that is, the side that was formerly in tension will

change to compression, Thus cyclic stresses are imposed Where the

beam is mechanically rotated, many reversals per minute may be

imposed Cyclic stresses may also be imposed by use of reciprocating

testing machines

For a given unit stress intensity, other factors being equal, a

material will sustain a given number of reversals before failure

To plot S-N curves similar to those in Fig 95 a number of indi-

vidual tests are necessary Curve A shows that the number of cycles

of loading sustained will increase as the stress intensity drops At

TESTING AND INSPECTION OF WELDS 165 point x the curve flattens and seems to indicate that the member would continue to operate indefinitely at that load intensity This load is the endurance limit or fatigue strength of the material tested under these conditions For steel this is roughly 40 to 50% of the tensile strength but less than the yield strength

For many types of materials the S-N curve has no “knee” and no flattened section in the curve but continues to increase in number of sustained cycles as the load intensity drops Nonferrous metals are rather noted for this characteristic Therefore, where materials of this type are used and if their failure would seriously endanger lives

v _NONTERROUS

“wT As

-

x “sag

LOG N (NO CYCLES)

Fic 95 Typical S-N curves

or property, their useful life may be anticipated and the members re- tired from service at the expiration of that time

Figure 96 illustrates welds made in various forms and their endurance strengths Discontinuities and sharp changes in section tend

to reduce the endurance stre.gth

Figure 97a illustrates a rotating-beam fatigue-testing machine The prepared specimen is shown in Fig 97b

Corrosion Tests Corrosion tests are for the purpose of determining the suitability

of a metal for job application in certain corrosive environments These tests should apply to both the parent metal and the weld metal by taking specimens from a joint in a transverse direction, with the welded area not to exceed 10% of the total specimen area After the suitability of a metal has been established for a given condition, it is desirable that similar tests be imposed on it after

Trang 9

welding to show whether or not the original properties have been

retained The tests used may not resemble service conditions, but

this fact alone does not reduce their usefulness for corrosion deter-

minations The important thought here is that data assembled under

laboratory conditions should be carefully interpreted in order that

degree and taode of corrosion can be anticipated for metals operating

under industrial conditions

si +

‹ 5 2

TYPE I SPECIMEN TYPE I SPECIMEN

BASC METAL ONLY TEE -WELOED wis, TCE ~WELOES wes,

BOTH SIDES

OnE S10e

Average Fatigue Strength, T1 Thousand Psi

ype of Steel Specimen

N = 100,000° | N = 2,000,000°

Plain carbon I 25.8f 22.8

ll 25.4 18.9

Low alloy I 35.3 26.4

* N denotes the number of cycles of stress for failure fi

¢ Each value is the average of three tests ]

(From AWS Handbook, 1942 Ed.) |

Fro 96 Fatigue strength of plates with transeerse fillet

u The aims of corrosion testing of welds are, according to the

American Welding Society Handbook (1942 edition): |

1

1 To find a welded joint to withstand a given environment

2 To determine whether a given weld will stand certain corro-

3 To determine whether the weld metal is more or less resistant

than the parts welded together

4 To check the corrosion-resisting quality of the welded metal

in a suitable test

Although much time is gained by using accelerated laboratory tests, some applications and metals require that long-time tests be conducted under actual job conditions

Corrosion tests are usually of three types: laboratory, field, and

Laboratory tests are intended to determine, in a comparatively short period of time, the probable extent of corrosion of metal over

Ya-20 AT HAND TAPS 4

(9) |aW°{ ges” $04" G57 °F

nh —

-

¬" . -.- -

au

Loa] Loan]

Liecols Electric Compeny

Fro 97 Attachment to rotating-beam testing machine for fatigue

testing

a normal service period These tests are not necessarily accelerated but are usually short time exposures The results obtained are ex- trapolated to predict long time-exposure effects

Field tests are those in which the specimens are exposed to field conditions but are not working members

Service tests are those in which the specimens become working members This type of test usually requires a longer period of time

to complete than a laboratory test

Trang 10

Corroding media consist of the atmosphere, water, and chemical

solutions The appropriateness of each depends mainly upon the

time alloted to the test, materials used, and the intended service con-

ditions Atmospheric exposure is not rapid in comparison with other

media and is generally used for testing the suitability of welded joints

for applications of protective coatings Water is frequently used as

the corroding medium, especially in testing ferrous metals Chemical

solutions are generally used for investigating special properties of

metals

The preparation of specimens is usually done by cutting trans-

verse sections from the welded joint, leaving parent-metal portions

on either side of the weld sufficient to include a portion which has

been unaffected by heat If corrosive attack of the joint is to be

observed, the coupon may be used without removal of reinforcement

Where weight losses are to be computed, careful machining of the

specimen to size is required The test pieces are usually supported in

the medium with glass rods

Under actual conditions of testing, factors such as time, solution,

composition, and agitation should be held constant if dependable

results are expected

Results are usually evaluated by visual inspection, computing

losses in weight, and considering depth of pits Visual inspection will

often give information as to uniformity of attack and appearance of

corrosion products Pits may be macro- or microscopically examined

to determine their size, depth, shape, etc

Photographs make valuable records of degree and nature of cor-

rosion In recent years photography has further extended its useful-

ness in this connection by recording corroded surfaces in their true

colors

Specific Gravity Test

The operation of this test is confined to the laboratory! The test

piece is accurately ground to a cylindrical shape % in injdiaimeter

an! 2 in long Its volume and weight are carefully determined in

metric units, The specific gravity is the result of dividing jts weight

by its volume, and is given as grams per cubic centimeter.’

“

Ị

TESTING AND INSPECTION OF WELDS 169 Other Tests

Where a shape or product does not lend itself readily to the standard forms of the tests described, special tests may be set up

to qualify the welds made or to qualify the welded structure This

is perfectly acceptable practice if the test conditions set up are closely adhered to If the tests are haphazardly conducted, they will have little value

The testing of tubing is typical of these special tests A common practice in this case is to force a section of welded tubing over a tapered pin or drift, thus applying circumferential stress which leads

to its ultimate failure The principle use of this test is to qualify the weld by comparing the weld strength to the strength of the tube wall Another variation of tube tests is obtained by placing a short section of welded tubing in compression Load is applied to cause failure by buckling Stresses thus derived are severe because of their complexity Welds must be of good quality to survive the compression test

NONDESTRUCTIVE TESTS

In welded structures or joints where, for any reason, it is un-

desirable to cut out samples for destructive testing, a type of test must _be applied to the welds that will leave them intact Several of these tests will be described Each has its own advantages; none gives complete information regarding the soundness and mechanical properties of the welded joint

Visual Examination The great majority of all welds made are passed or approved by visual examination This test is far from infallible, but if the general appearance of the finished weld meets certain standards (see Fig 98), the product is usually considered to be satisfactory, except in those cases where danger to life and property or other serious losses may be incurred by failure of a welded joint Fortunately, much of the welding done in industry permits observation of incipient failures

or in case of failure, probably creates no particular hazard

Tiêu đề	Welding Inspection
Trường học	University of Welding Technology
Chuyên ngành	Welding Engineering
Thể loại	Thesis
Năm xuất bản	2023
Thành phố	Hanoi

Định dạng
Số trang	18
Dung lượng	2,8 MB