JOURNAL

OF T H E
FRANKLIN INSTITUTE.
OF THE STATE OF PEN~SYLVANIA~
FOR THE PROMOTI ON OF THE MECHA: NI C ARTS.
VOL. CXV. MARCH, 1883. No. 3.
THE Franklin Institute is not responsible for the statements and opinions
advanced by contributors to the JOURNAL.
B R I D G E I N S P E C T I O N .
By W. S. THOMPSON.
[Apaper read before the Engineers' Society of Western Pennsylvania, December 19,1882.]
I have been asked by members of this society to give t hem some of
the results of my experience as an inspector of bridge material. To
oblige them I offer this paper, otherwise I would not have trespassed
upon Your time, for so much has been written upon every depart ment
of bridge bui l di ng and in every style, t hat I very much doubt my
ability to say anyt hi ng t hat is new on the subject.
For convenience I will divide the paper into the following heads :
Inspectors, their duties, etc. ; Drawi ngs ; Tests.
INSPECTORS.
Ever y shop t hat buys any article, on receiving it, gives it to some
one to exami ne to see i f it is of the proper material, size and qual i t y.
This person, it is clear, must have a knowledge of what t hi s article
should be and what deviations axe allowable from the order for it.
You will reply, of course, he must have such knowledge to be able to
decide whet her the article is suitable. Perhaps so, but I will vent ure
to say, t hat i f you made such an assertion to the foreman or manager
of a bridge shop, he would tell you i t mi ght be so in a general way,
but t hat it was not always the case in t he appoi nt ment of those who
WHOLE No. VOL. CXV.--(TRIRD SERInS, Vol. lxxxv.) 11
162 Bridge Inspection. [ J o u r . Frank. Inst.
are to look after the inspection of bridge material dur i ng manuihc-
ture. Al l of them, at least those of the eight or ten shops I have
visited recently, as well as some inspectors of my acquaintance, could
cite you cases of inspectors who had little or no knowledge of their
duties and a very indistinct idea of any sort of manufacture.
What are the requirements fbr an inspector? As he must decide
questions, to solve which require a knowledge more or less great, of
strains, designing, drawing, materials and shop work, i t is plain he
should be fami l i ar wi t h each of these subjects, part i cul arl y the last
three, as the first two are generally looked after by his employer.
Hi s duties bring him in contact with men of all classes, so I may add
t hat he should have some knowledge of business.
Next, what are his duties ? He is appointed by his employer to see for
hi m t hat an agreement is fai rl y carried out, neither for the ~urpose of
showing his (inspector's) knowledge and aut hori t y, nor for t he purpose
of annoyi ng and del ayi ng the contractor, on the contrary, he should
rat her assist hi m i f an opport uni t y offered itself. I do not mean that
he is to consider hi msel f in any way an employ6 of t he contractor,
but occasions sometimes arise where some such exertion on his part as
can reasonably be asked for, will hasten the completion of his em-
ployer' s work. I n such cases it seems to me to be his dut y to make
t hat exertion.
At times there will be instances where there is an honest difference
of opinion between t he inspector and the" contractor. Then the former
must hear what the latter has to say, but make up his own mi nd as to
what is ri ght , and to hol d to t hat opinion.
Should he t hi nk any question t hat may arise too i mport ant for him
to decide, he should refer it immediately to his employer and not vas-
cillate nor delay, for by acting in this way he will weary and annoy
every one and possibly stop his work.
Whi l e any of his work is being done he should make it his dut y to
be at the shop dur i ng the usual office hours, unless he wishes others to
decide for him such questions as may arise dur i ng his absence. Of
course it is not necessary for hi m to ask permission from the shop au-
thoritie~ when he wishes to be absent, but it is j ust as well to let them
know when and how long he intends to be absent, even i f i t is onl y
fbr an afternoon. I f not hi ng else, this is at least a courtesy, and I do
not t hi t & courtesy is disliked by any one.
The shop is ahvavs to suppl y hi m with the necessary labor for
~ V I a r . , 1 : S 8 3 . ] B~qdge Inspection. 163
handling work. So far as my experience goes, there is never any diffi-
culty about t hi s matter, provided he will be reasonable, except t hat
there are times when the labor cannot be gi ven at the moment i t is
wanted, the exigencies of the shop not always permi t t i ng of i t ; but,
as a rule, I have found t hat the superintendents and foremen were
always wi l l i ng to help an inspector in every way, as long as the latter
was not all for hi msel f and t hought of no one else. The shop has to
pay for the labor t hey gi ve hi m, therefore it is onl y proper t hat he
should, as much as possible, arrange his work and his time for doing
it in a way t hat will put t he shop to as little expense as possible.
He" shoul d avoi d being di st rust ful of every one. I f he has cause
to believe t hat people are dealing unfai rl y wi t h him, there are many
ways of cor r ect i ng the matter, or, at least, of relieving hi msel f of
the responsibility resulting from such practices.
I have been asked what errors an inspector should look for, etc. I t
is difficult to answer general questions in a definite manner, but
the following may t)e answer to some I have had asked : Inspection
can be superficial or it can be minute, or somewhere between the two.
As an instance I may cite the following : An inspector had some 200
pieces about 15' long and weighing 800 lbs. each to examine. He had
to make at least four careful measurements on each piece before he
could say t hey were correct. To do this took hi m about two hours a
day for some four or five weeks. Anot her inspector went over t hat
same lot, and gave it what he called a t horough inspection ; the time
he devoted to it, all told, was not over t went y minutes.
I n exami ni ng work it is har dl y possible to say what one t hi ng to
look for. Pr oper l y speaking, however, and inspector shoul d expect to
find everyt hi ng exactly as it ought to be. Wi t h many inspectors ri v-
eting is the onl y mat t er t hat is looked after, but there are also, and of
equal importance, the size and qual i t y of the material of the different
parts; the straightness of the finished piece ; the accuracy of the work
done upon it, such, for instance, as the size and position of the pi n or
other holes, or slots, and how these will compare with t he other parts
with which t hey are to fit. To sum .up in a fhw words, he must see
that the work in all essentials is exactly what the drawi ng calls f or ;
also, t hat the material and the work done on it is of the qual i t y called
for, for any or all of these may be wrong. :Now, he can examine one
piece for all these requirements; he can examine all pieces for one
thing, as, say ri vet i ng ; he can see t hat each piece is correct in every
164 Bridqe It~spection. [Jour. Frank. Inst.,.
detail ; or, as I have seen done, he may be able to cast a glance over a
l ot of work and say whet her it is r i ght or wrong. Of how this was
done I am i gnorant , but of one t hi ng i am sure, t he opinion as to its
condition was not wort h much.
Ther e are times when some points in exami nat i on may be omitted ;
j udgment and experience must decide when this can be done, but there
should be no mere t r ust i ng to luck.
The manufact urer is most undoubt edl y reponsible for all these
things. Inspect i on does not relieve hi m f r om responsibility, nor does
his responsibility relieve t he inspector. The l at t er' s empl oyer, and the
manufact urer. al so, provi ded t he inspector is competent, t hi nk t he mat-
t er is of so much i mport ance t hat it is well to have some one to see
t hat t he agreement is fai rl y carri ed out.
I t mi ght not be amiss for t he inspector to remember "t hat he has
a reput at i on to sustain, " and t hat i f he is careless, others wi l l l i kel y be
so to, , ; also, t hat in t he of t - quot ed Ta y disaster t her e was a strong
presumpt i ve evidence t hat t he inspectors had neglected t hei r duties, or
were i gnorant of t hem.
Wha t allowance can be made in t he way of deviation from dr aw-
ings ? For definite answers special cases woul d here also have to be
cited, but general l y t he following may be made As to size, shape irons
are general l y accepted i f t hey are not more t han 3 per cent. l i ght .
Rods, flat bars, and plates are more easily rol l ed to a size, t herefore
t hey are seldom accepted i f not up to t hat size, t hough under special
circumstances i f t he l i ght pieces are but comparat i vel y a small propor-
tion of t he whole number, t he above percentage may be allowed.
Ri vet ed members shoul d be peri bct l y st rai ght except top chords and
inclined posts, which may perhaps be bet t er fi~r havi ng a slight camber~
but I often find it necessary to aeeept posts say 30 r long out of line
by 1- . Some few men work closer t han this, and it nmst be acknowl -
edged it is a ver y vi t al point. I inspected t he mat eri al for a ver y
large bridge, ever y riveted member of which, except one, was perfect l y
st r ai ght ; and the error in t hat one was, i f I recollect ri ght , 41" in
35q
Punchi ng is not always accurate. I f i t is not out more th~n { " in
20 ~ or 30 r it is considered t ol er abl y good work. I n t he ease of angles
used as brackets, if, aft er being ri vet ed on, t hey are wi t hi n " of posi-
tion, it is considered correct. Thi s is usual l y near enough.
When work of the above ki nd must be more exact, it should be so
Mar., 1883.] Bridge Inspection. 165
stated, as it calls for an ext ra class of work. I n the wi dt h of chords,
i. e., the distance apart of side plates or channels, i t is usually con-
sidered very fair work when t hey are wi t hi n 1-16 '~ of the distance
called for, t hough t hey must be nearer t han this when the side plates
are t hi n, say ~ ' .
Top chords and posts should never have t hei r pin holes bored less
than the distance apart called for. 1-32" excess in former and 1- 16"
in latter will never do harm, except in some few special designs.
Bottom chord and tie bars should, i f t hey var y at all, be less t han the
length called for. 1-32" error in these will not be at all serious, pro-
vided all the bars are bored to exactly the same length. Struts i f
bored wi t hi n 1-16 ~t are near enough. Shop foremen generally follow
the above rules.
Ill maki ng eye-bars, the smi t h is allowed ~" over or under length
from out to out. I n thickness of head he is allowed 1-32" under or
1-16" over required thickness, which he sometimes exceeds, but not h-
i ng is allowed at the neck, as this is apt to be the weakest part of t he
bar. [ u maki ng lateral or other similar rods he is allowed to var y ~
ei t her way from required length. Ot her parts must be made closer
than this, but smiths rarely forge very closely.
Pi ns should never be taken i f shorter t han called for unless t hei r
lengths are calculated for washers, and t hey should be examined
closely to see i f t hey are of uni form diameter and not gouged out at
the corners or ends.
Rollers may var y a little from required diameter, but t hey must be
al l of the same diameter.
Web plates of girders of all classes should be flat, but it is custom-
ary to allow the web of large girders to be sl i ght l y buckled or dished
to the amount of say ~pr versed sine to a 6 p chord.
Mills do not always deliver plates perfectly straight. Have t hem
straightened i f necessary, but otherwise, when possible, use t hem as
they are, for there is no way of st rai ght eni ng t hem except by the sledge
or drop hammer, neither of which are very good for the iron. For
other similar matters I can onl y refer to common sense and t hat intui-
tive knowledge which is the restllt of experience.
As to inspection at mill. An inspector can onl y seldom see the actual
rolling of his material unless the mill men would roll onl y to suit his
convenience, and even t hen it would l i kel y be impracticable.
Tevlsion and bending tests of the material can be made at the mill,
166 Bridge l~spection. [Jour. Frank. Inst.
"though where it is practicable, it is very much better to cut off and
test the test pieces at the shop where the material is delivered ; this
can easily be done by ordering ext ra pieces or a few pieces of extra
length. The inspection of the rolled iron, however, is best done at the
shop, for there one can usual l y see it to better advant age, to say noth-
ing of the fact t hat the most serious flaws are generally made apparent
duri ng manufacture.
My experience has been t hat the inspection of material at mill, ex-
~ p t testing, generally amount s to not hi ng, for often times t hey have
to load material (at which t i me the exami nat i on is expected to be made)
when the inspector cannot be there, or after dar k when he cannot see;
this is not from any desire to be contrary, for t hough t hey are generally
hard worked I have found shi ppi ng clerks very accommodating, but
at an iron mill i t requires them to use their wits to keep the yard
clear.
Inspectors necessarily lose a great deal of time wai t i ng on others.
Thi s is of dai l y occurrence and cannot be avoided, and such time, as a
rule, must be a dead loss.
I have been asked what are the fkults of some of the bri dge shops.
To answer t hat I fear mi ght be considered as get t i ng personal, though
undoubt edl y to my t hi nki ng all of t hem have t hei r faults, as, with a
few exceptions, t hey all have one faul t in common, perhaps I can speak
of t hat , viz., painting. A bridge-shop pai nt i ng gang usually consists
of a lot of hal f ; gr own boys led by a man who has not a very great
reputation for either force or thoroughness. The chief idea of the
whole crowd is to see how fast t hey can get over work. I t seems to
be of no consequence to t hem whet her the whole piece is painted or
not, for t hey sometimes seem to t hi nk t hat it is unnecessary to paint
corners or unseen parts, so long as the piece looks as i f it was painted.
Possibly this may be a result of t hei r being usual l y hurried, to say
nothing of bridge shop pai nt -work havi ng to be done out of doors and
in all sorts of weather. I have no doubt you will smile, but I can as-
sure you t hat one of t he most satisihctory pai nt i ng gangs I ever came
across was led by a boy fourteen years old ; most satisfactory, because
the boy was a good foreman and saw t hat his gang di d their work
fai t hful l y.
DRAWI NGS.
The inspector seldom has anyt hi ng to do wi t h the maki ng of draw-
ings, but he has a great deal to do with t hem after t hey are made, anct
Mar., 1883.] B r i d g e I n s p e c t i o n . 167
this generally under unfavorabl e circumstances, consequently he is
much interested in t hei r clearness and fullness of detail.
Tile bridge draught smen should bear in mind t hat drawings are a
species of plcture-writing, used to convey to those employed in shops a
clear idea of the wishes and intentions of those in charge of design-
ing. Some individuals when wri t i ng a letter will write a good, clear,
readable hand wi t hout crossing their lines, or maki ng fantastic additions
to their letters ; t hey also express themselves fullv and clearly, conse-
quently one can read their letters rapi dl y yet have a complete idea of
their contents. Ot hei s write a cramped, illegible hand, add fantastic
scrolls to their letters, omi t i mport ant words and i~il generally to con-
vey their ideas clearly. Wi t h writers of this latter style all of us
know what a time it requires to find out the exact i mport of their
letter. I t is the same with drawings. Some men make their drawings
so clear t hat t hey can be read at sight ; others again draw in such a
manner t hat in at t empt i ng to read t hei r drawings you fancy t hat t hey
tried to make t hem as much like puzzles as possible. The draughl s-
men should also bear in mi nd t hat there is a difference between t he
style of a drawi ng to be made for shop use and one t hat is made ibr a
periodical or other similar purpose ; also, t hat when he uses a drawing,
he has clean hands, a dry, quiet, and well-llghted place to lay his
drawing and weights to keep the corners down, and paper and pencil
to supply missing dat a; but not unfrequent l y the cont rary of this is
the case wi t h those in the shop, and part i cul arl y so wi t h inspectors,
who often times have to hol d drawi ngs in high winds and at the same
time use rule and callipers, etc. I n other words, he should not make
sheets too large, nor crowd too much on one sheet; nor put so many
lines on top of each other t hat i t is difficult to follow any, nor omi t
any figures t hat are required to be known at a given point because t hey
can be found on bill of material or another dr awi ng; t hus necessi-
tating a bill of material and a couple of drawings to know, say, what
a post end should be, and he should not imagine t hat pieces of un-
laced material are always the exact length called for.
Some may say t hat the draught sman' s time is expensive, and the
ibremau would underst and this or that. True, the foreman generally
<!an do so, but the foreman does not do the work nor st and over the
men doing it, consequently he has to supply the omission of this picture
writing to his men, i . e., do draught smen' s work to the neglect of his
own, which, according to an old-fashioned notion, is t hat he should
168 Bridge Inspechon. [ Jonr. Frank. Inst.
look af t er t he qual i t y of t he material being used, what sort of work
t he men were doing, and mat t ers of si mi l ar ki nd, all of which was
t hought enough f or hi m to do. Thi s l at t er may be an old-fashioned
custom, but I believe it is still adhered to in locomotive and ot her ma-
chi ne building shops.
For emen do not always cheek t he drawings to see i f t hey are full
and compl et e; unf or t unat el y this results sometimes in wor k being
spoiled, si mpl y because t he dr awi ng did not state what was wanted.
Tr ue, t he foreman could have seen what t he dr aught sman had omitted
i f he had been doi ng t he work himself, but his men do not necessarily
have t he same education and abi l i t y as he has, and drawi ngs are as
much for t he men' s use as for any one else.
The errors of draughksmen, who can be called such, are not, as a
rul% in t he large matters, but in ttle small ones, which t hey t hi nk are
uni mpor t ant and can be left to the shop. Thi s is a f r ui t f ul cause of
del ay and expense, general l y small, t hough not always so, and, in my
experience, is ~)ne of almost daily occurrence; in fact, I doubt whether
I have had a single span under my charge for years on whi ch there
could not have been saved a few dollars and upwards, by t he draughts-
men payi ng a little more at t ent i on to clearness and those little things
whi ch ever y one knows.
Great accuracy is unnecessary in some of t he parts of a bridge
which it will be well to remember, as t he more aeecurate t he work is
requi red to be t he nmre expensi ve it will prove, also t he f e we r the
part s t he cheaper will be t he work. Shor t pieces of angl e iron or any
small pieces fastened to a member weigh little but cost great l y.
A dr awi ng shoul d be clear and to the scale, with little or no shad-
ing, and when possible should be complete in itself wi t hout reference
to anyt hi ng else. Ti me saved in the drawi ng room is not always an
economy. I n maki ng a drawi ng, a dr aught sman should use as few
different sizes and lengths of material as possible, for si mi l ari t y in
sizes of pins, bars, rods, and in l engt hs generally, saves in t he cost of
wor kmanshi p and reduces t he chances of error. A theoretical saving
of material sometimes necessitates an out l ay of several times the
amount saved to do t he ext ra work, also, usually, each different form
necessitates a different set of rolls fi)r its maki ng, whi ch may cause de-
l ay in del i very of material. The designer should have some considera-
tion for what is obtainable, as for instance, a vari at i on of lb. per foot
in a 12" channel onl y makes a difference in thickness of web of l -
Mar., 1883.] Bridge Inspection. 169
100" closer t han it is possible to roll, unless t here are a large number
of same weight.
Plates and bars ordered by 1- 32", or squares and rounds ordered by
1-16", general l y come ei t her full or scant, more l i kel y t he former.
1-32 fl is looked upon in bri dge shops as somewhat of an i magi nary
quantity, and part i cul arl y so to ask it of a bl acksmi t h. My own ex-
perience has been t hat by those wor ki ng on tools it is considered to
mean scant or ful l of a gi ven 1-16 <t, but a smi t h scarcely seems to
realize what it is. Tel l hi m a piece is to be so many si xt eent hs t hi ck
and he will come ver y near it, but ask hi m to make i t so many t hi r t y-
seconds t hi ck and you will be l ucky i f you get it wi t hi n an ei ght h.
The office says t he man shoul d be educated up to this close work, to
which the shop replies t hat t he office shoul d be educated up to t hei r
wants. I say, t hat i f t he office wants such close work, t he men are to
be found, but t hey are onl y to be f ound in gover nment arsenals and
tool maki ng and ot her establishments t hat requi re ver y close work.
Of late days i t seems to be considered i mpor t ant onl y to be ver y
careful to gi ve distances of r i vet spacing. Thi s is no doubt of pri me
necessity to t he layer-off, but not so to t he machinist and inspector, the
former onl y wants to know t he distances he has to work to, and t he one
governing poi nt from which he has to st ar t ; besides this t he inspector
usually onl y cares to know i f t he number of rivets is correct, and t hat
theh" spacing is appr oxi mat el y so. I n maki ng drawings, distances
from centre lines and ot her gover ni ng points should be given, and all
parts that arc to be planed, bored, or drilled~ shoul d be so marked. A
hole or slot should always be mar ked with t he size of t he pi n or t enon
that it is to t ake; and stating what sort of fit is required. Angles are
best defined by t hei r base and perpendi cul ar, and it should be remembered
that sheared part s are not perfect l y st rai ght or smooth ; t hat a punched
hole has not parallel sides nor is it always exact l y where it ought to be,
and t hat when two or more plates are ri vet ed t oget her t hei r uni t ed
thickness is general l y great er t han t he stun of t hei r several thicknesses ;
that ri ght and l eft pieces or ot her complications are to be avoided when
possible , as t hey are t roubl esome and liable to cause errors; and t hat
rivet holes should nei t her be too large a proportio,~ of t he wi dt h of a
channel flange, nor too near its edge.
As few rivets should be lef~ to be dri ven in t he field as possible, i br
they are more expensi ve and less reliable t han those dri ven in t he shop.
As to the latter, shop hands have ever y convenience possible, and are
1 70 Bridge I~81oeotio~L [' Jour. Frank. lnst,,
dai l y occupied in dri vi ng rivets, whereas the men in the field have ol~
the other hand almost every inconvenience to contend with, and driv-
ing rivets is but a very occasional part of t hei r work.
TESTS.
Looki ng over the results of a large number of specimen tests of
wrought iron made by myself, I find t hat t he results are about the
same for bars, rods, plates (edge rolled), angles and channel bars, viz.,
an ul t i mat e strength of from 50,000 to 53,000 ]bs. per square inch, gen-
eral l y 51,000, and an elongation in 6" of from 16 per cent. to 25 per
cent. channel bar webs, and plates rangi ng chiefly from 16 to 20 per
cent. Thi s I t hi nk is due very much to the fact t hat in the test
pieces made from these, it was necessary to use a rect angul ar ~'ection.
I have tested sheared plates up to 72t r x~" wi t h the same ultimate
strength as the above, but the elongation var yi ng from 9 per cent to
16 per cent.
Bending Tests.-- Bars, rods, and channel flanges wi t h few exceptions
bent unt i l flat or t hei r ends touched with a curvature, whose diameter
wr i ed from one-hal f to three times thickness of piece, generally one and a
half. Plates and channel webs would bend 150 degrees, or unt i l their
ends touched wi t h a curve whose diameter ranged from one tq four
times thickness. Al l bending pieces were about one foot long, and
from 2" to 3" wide. Here it may be well to say t hat bending pieces
must not be sheared but planed, and also have t hei r edges well rounded
on a grindstone or emery wheel, so as to remove any incipient cracks, i
These are tests of such wrought iron, wi t h but few exceptions, as has
}een used in the work t hat I have had charge of dur i ng the last si x
years. Of course all will understand, t hat these tests were applied in
line of fibre, as is the ordi nary usage.* Wher e more t han this is
expected~ viz., a strain, crosswise to fibre, it is always specially men-
tioned, as witness British Admi r al t y and other specifications The
reduction of area has been omitted, as t he measurements t aken for it
arc generally unreliable, and ms Stoney says, the elongation is much
more reliable for any question as to quail U ot iron.
I have other tests, which are of condemned lots, but presume they
would be of no interest, ibr surely no one is going to t ry for the worst
material he can get. I have had but little to do with cast iron, since
See t abl e f or t r a ns ve r s e s t r e n g t h of pl at es.
Mar., 1883.] Bridge Inspection. 171
testing has become common, but have about 12 or 15 tests f~om as.
many heats. I n these, t he bars 1" sq., 5 r long, 4 ~ 6" between sup-
ports, it took from 650 to 700 lbs. resting on dul l kni fe edge at cent re
to break them, which was a better result t han asked for. Of course
the foundry hands took good care to place the bar on supports, same
side up, as when cast. Such east iron should onl y be used in very
special cases, for it is certainly much too good to be used in such places
as masonry plates, and washers of bridges.
Thus filr I cave onl y tested steel from plates rolled in t hi s count ry
from Engl i sh Bessemer blooms. I made a large number of tests from
these plate% both tension and bending ; of the latter there were over
100. The plates ranged in size from 12rrx ~" to 24 t1 x ~ ' . The results
were very uni form, gi vi ng an ul t i mat e st rengt h averagi ng 70,000 lbs.
per square inch, wi t h an average elongation of 30 per cent in 6 II.
The bending pieces were planed from plate shcarings very l i t t l e
broader t han the test piece. Al l these were well rounded on gri nd-
stone, heated red hot, and dropped into water of a temperature from
70 to 100 degrees. :Nearly all bent 180 degrees flat, some few o f
those it to ~ cracked at edges when 90 degrees; these latter were
then punched wi t hi n ~" of edges wi t hout cracking, probably there
were invisible edge cracks.
Some of yon, no doubt, would like to know what strain full size
pieces, j ust as t hey go into a bridge, will stand. I have enough o f
one kind, I t hi nk, to satisfy you, viz., of tension members. ] have
made a few tests of compression pieces which have been published
elsewhere.
I will first give you the result of four lots of bars, rangi ng from
1" fo 2" diameter, and from 151 to 251 long wi t h enlarged ends for
thread.
1st lot consisted of 60 iron rods. 14 of these broke in thread, the
rest in the body of the bars with 12 to 19 per cent. elongation. Al l
except four stood over 45,000 lbs. per square inch ; one of these broke
at 35,000 lbs., two at 40,000 lbs., and one at 43,000 lbs. per square
inch.
2d l ot ; 23 rods. 13 broke in screw, the others elongating 12 t o
16 per cent. Al l but one stood over 4,5,000 lbs. ; t hat one stood
44,000.
The above are onl y moderately good.
3d l ot ; 38 rods. 11 broke in screw, the others elongating 13 to 18~
172
per cent.
follows:
Br i dge I ns pe c t i on.
All but seven stood 50,000 lbs. or over;
1--38,000 lbs. per square inch.
1 - - 4 2 , 0 0 0 " " " "
1--43,000 " " " "
1--48,000 " " " "
3--49,000 " " " "
4th lot; 30 bars. Two broke in screw.
[Jour. Frank. Inst.,
these broke as
All stood 50,000 lbk per
Of 26 6" bars, 13 broke in bar with tbllowing strains:
5 between 40,000 and 45,000 lbs. per square inch.
6 " 45,000 and 47,000 " " " ~ "
1 " 47,000 and 49,000 " " " "
1 " 49,000 and upwards " " " "
13 broke in head.
11 between 42,000 and 45,000 lbs. per square ilmh.
2 " 45,000 and 48,000 " " " "
square inch or over. Elongation of those that broke in bar, was from
10 to 18 per cent. generally 16 per cent ; nearly all these broke within
2 feet of end of rod, as did those in the other lots.
Another lot of 20--1~-" round rods with 1~" threads (all of course
in thread) broke with 37,000 to 40,000 lbs. per square inch of rod.
Before proceeding further, I may as well state here, that the above
lots of rods were made by different shops, as were also the eyebars,
the test results of which I am about to give you. These I have
classed by width of bar, as it will give you as good an idea of
the average eyebar as any other classification.
I will first give you the results of tests of some heads made entirely
by upsetting, merely saying that they have been given to me, and that
I did not see the tests made :
Of 52 bars about ,,~21'P wide, 22 broke in bar, elongating 7 to 15 per
cent., and 30 broke in the head. The ultimate breaking strains seemed
too high, consequently I omit them, excepting to say that one broke in
the head at 36,000 Ibs. per square inch, and two others at 42,500 lbs.
Next I will give you result~ of tests of eyebars made by piling. With
the exception of some 10 or 12 bars, the iron in all of them showed a
good fibrous fracture, and with very few exceptions, perhaps six, the
heads were proportioned by usual tbrnmla.
}lar., 1883.] B r i d g e . I n s p e c t i o n .
Of 159 5 ' I bars, 80 br oke in bar :
2 bet ween 35, 000 and 40, 000 lbs. per square inch.
10 " 40, 000 and 45, 000 " " " "
17 " 45, 000 and 47, 000 " " " "
27 " 47, 000 and 49, 000 " " " "
24 " 49, 000 and upwards " " " "
79 broke in he a d- - one at 33, 000 lbs.
6 bet ween 35, 000 and 40, 000 lbs. per square i nch.
36 " 40, 000 and 45, 000 " " " "
22 " 45, 000 and 47, 000 " " " "
11 " 47, 000 and 49, 000 " " " "
3 " 49, 000 and upwar ds " " " "
Of 46 41t bars, 26 br oke in bar :
3 bet ween 40, 000 and 45, 000 lbs. per square i nch.
11 " 47, 000 and 49, 000 " " " "
12 " 49, 000 and upwar ds . . . . . ' "
18 broke in h e a d - - o n e at 37, 000 lbs.
10 bet ween 40, 000 and 45, 000 lbs. per square inch.
1 " 45, 000 and 47, 000 " " " "
4 " 47, 000 and 49, 000 " " " "
2 " 49, 000 and upwar ds " " " "
Of 19 3 ' f bars, 14 br oke in bar :
5 between 45, 000 and 47, 000 lbs. per square inch.
3 " 47, 000 a~d 49, 000 " " " "
6 " 49, 000 and upwar ds " " " "
5 broke in head :
1 between 40, 000 and 45, 000 lbs. per square inch.
2 " 47, 000 and 49, 000 " " " "
2 " 49, 000 and upwar ds " " ' " "
173
Fr om t he foregoi ng it seems to me t hat these resul t s shoul d be
called good when
6 II bars br eak at 45, 000 lbs. and upwards.
51f " " " 46, 000 " " "
4 II " " " 47, 000 " " "
3 I I " " " 48, 000 " " "
Wi t h an elongation of not less t han 10 per cent. nor mor e t han 25,
to be measured in not less t han 51 . Bar s of same rol l i ng even when
broken in t he body wi l l gi ve differences in elongation ; 14 to 16 per
174 Bridge Inspection. [Jour. Frank. inst.,
-cent, however, is the most common elongation in the above tests. They
will also break at different strains ; e. g. 6' rx 3" bar broke at 46,800 lbs.
per square inch, elongating 12 per cent., when a 6" x 1~" bar broke at
49,500 lbs., elongating 15 per cent. One 3 x l ~ " bar broke at 46,000
lbs. per square inch, elongating 17 per cent., anot her 3 x l ~ " bar broke
at 50,000 lbs., elongating 17 per cent., all of which bars, I believe,
were rolled from exactly the same stock, and seemed to me to be purely
fibrous and wi t hout flaw. Al l who have done much testing, know that
results will sometimes var y in test pieces cut fl'om t he same piece of
iron, to wit, a ~,I round ; but the above variations, seem to be greater
t han is due to t hi s cause. Some say i t is the result of unknown factors
in the manufacture ; perhaps the effect of heating onl y a portion of the
bar. Is it so?
Some tests made at Wat ert own on 5" and 3" flats 10 f long showed
even better t han this, but all those bars were rolled specially for the
test, had no rough handling, and were not heated or worked on after
leaving the rolls. Fr om these and other tests it seems probable that
all double-rolled iron bars will, i f tested as t hey leave the rolls, stand
fi-om 50,000 to 52,000 !bs. per square inch ultimate strength, with an
el ongat i on of 15 to 23 per cent.
Some lay stress upon havi ng a bar break in the shank. I do not
t hi nk it so essential, t hough I prefer it, for when the bars have elong-
ated 8 per cent., the bottom chords and ties will have lengthened so
much t hat the bridge will, most likely, fal l between its abutments.
I append a table of some tests of eyebars, but wish to give results of
three here :
5"x1~ '1 bars had an elastic limit of 27,000 with an ultimate, strength
o f 37,800, broke 6' from pin, little or no weld. Moral rol l i ng mills do
not al~vays pile fidl length piles. I am very sorry to say, I could show
others like this, and testing to destruction is often the onl y way of dis-
covering this defect.
A 5' ~x 1~ '~ bar, 12' centers had an elastic l i mi t of 33,000 lbs., but
broke wi t h 33,500 lbs.
A 5" x 1 7-16 r' bar, 15' 3 " center, pins 4s 3' ' , had accidentally been
strained beyond it~ elastic limit, being lengthened t hereby ~"; pin-
holes had a sligbt set, say 1-100". Elastic l i mi t supposed to have
taken place at 31,000 lbs. Thi s bar was loaned to me, provided I did
not stretch it any more. I put the bar into press with 4~ r' pins, and
applied a strain of 15,000 lbs. per square inch, hol di ng it 5 mi nut es;
)Iar., 1883.] Bridge Inspection. 175
released this, and applied a strain of 20,000 lbs., which was likewise
held 5 minutes and released. Fi nal l y a strain of 24,500 lbs. was
applied for 5 minutes. The pin-holes were t hen measured; both
measured the same, viz., 4 25-64" transverse to bar, and 4 13-32 ill
line of bar.
The test of 5" x l~-" bar is given to show t hat rol l i ng mills are not
always careful, even when paid for extra good iron. That of 5" x
1 7-16" bar is given to show t hat with properl y proportioned heads
it is not necessary to make pins fit holes so very accurately as is often-
times required. I do not, however, advocate such a difference as this
between pins and pin-holes, but I do not t hi nk t hat i t is allowable to
make pin-holes 1 32" larger t han pins for diameters of 4, 4", and
upwards.
Neither 5 " x 1~ nor 5 " x 1~" bars gave any i nt i mat i on of their
yielding unt i l t hei r elastic l i mi t had been very nearl y reached.
By some, great stress is laid upon appl yi ng a proof-strain. I have
personally tested several t housand bars and rods of all styles of make,
but never found one defect by this proof-strain. Indeed, I have sub-
jected eyebars wi t h visible defects to this strain ; and al t hough I have
closely watched them, not hi ng more could be seen t han before the
hal' was under stress.
I have found proof-strains-serviceable onl y wi t h small things t hat
were considered of no importance by the smith.
Whenever the mat t er is left to me I never use a proof-strain, ex-
cept ~vhere the ends are welded to main body of bar or rod, i. e., fin-
ished bar louger t han original bar, and then not t hat I expect it to be
of any service, but because I considered it to be a safe plan to examine
a weld i n every possible way. I n place of proof-strains, I should
advise, in addition to this, t hat a number of extra tension pieces should
be ordered; and when the whole lot are finished, t hat as many be
chosen as were ordered extra, and then break t hem; but not till
all are finished. The cost will not be great, and I am sure it will
have a good effect, for havi ng tried it on a small scale, I have found
it to work wonderfiflly well. So I may say I speak to some extent
from experience.
Some rai l way companies require the modulus of each bar to be
taken. I n my experience t hi s cannot be done wi t h any accuracy wi t h
such appliances as are to be found in shops. To be accurate, the bar
ought to be counterbalanced at say every ten feet wi t h movable
176 Bridge fnspeetion. [Jour. Frank. inst.,
weights, so as to eliminate the eftbets of any sagging whi ch is l)ro-
dueed by weight. The pressure-gauges shoul d read accurat el y and
not approxi mat el y, and much more t i me gi ven to it t han ordinary
shop testing will admi t of.
I have tested over 5,000 bars for t hei r modul i , and all t he results I
have obtained lie between a modul us of 24, 000, 000 and 30,000,000, or
say extensions of . 008" to . 01" per foot for 20, 000 lbs. strain.
Gener al l y at any one testing t he modul i di d not var y more than
3, 000, 000, of whi ch t he equi val ent vari at i on in extension would be
. 001" per fbot.*
I n my opinion, even this is great er t han it would be i f all sources
of er r or were eliminated, fbr I am inclined to believe, on l ooki ng over
t he tests I have made, t hat t he modul us of" t he iron (from t hree differ-
ent mills) I have tested lies between 26, 000, 000 to 28, 000, 000, or the
equi val ent extensions of f r om . 0092" to . 0086" per foot to 20, 000 lbs.,
and t hat t he ext remes are perhaps even less t han here given.
Since wri t i ng t he foregoing, I find t hat in Wat ert own-t est s of 2510 r
bars of two different classes of iron from t wo different mills, maki ng the
l ot referred to above, t he extension of 21 for 20, 000 lbs. st rai n was
wi t hi n t he above l i mi t ; t he others had t he fol l owi ng: . 0093", . 0084",
and . 008" per foot. I feel inclined to t hi nk t hat this last ought to
be omitted. Taki ng t hem as t hey are for 10, 000 lbs. strain per square
inch, t he variations in bars 3(i)' long woul d be . 021", or one and a
hal f si xt y-fourt hs of an inch. I n six bars rolled at one mill, the
difference of extension fbr 20, 000 lbs. strain was . 00048" per foot,
which woul d be for a 30' bar with 10,000 lbs. strain, 1- 128" nearly.
As so much testing of bars has been done, perhaps it would be well
to state t he different ways of maki ng iron eyebars. As to design, the
heads may be square, octagonal, circular, pear-shaped, or an oval,
made by st ri ki ng the curves formi ng eye, f r om t wo centres wi t h same
radius, and connecting t hem by tangents. Thi s is done to give a
great er proport i on of metal back of eye, whi ch, I t hi nk, is ver y j udi -
cious, because i t makes some allowance for t he inaccuracy of t he smith,
and experience convinces me t hat it aids ver y great l y in prevent i ng
* These elongations for moduli were t aken between strains of 10,000 lbs.
and 20,000 lbs. per sq.H; previous to adopting this method the elongations
were taken between strains of zero to 20,000 lbs. per s q. ' , and I t hen ob-
tained as great variations as you can find published, for t he unsupported
length of bar was t he greatest factor in the result.
Mar., 1883. ! Bridge Inspection. 177
ally deformation to the eye; for when so made, and other proporti,ms
as usual, it is extremely rare for t hem to break at pin-hole. Ol d
EngliSh tests, and also those of Chi ef Engi neer Sprague, of the navy,
confirm this. General l y the t hi r d and last shaped heads are used ;
they may be either same or greater thickness t han the bar, and are
connected to the bar by the neck, which is formed of curves of greater
or less radii.
As to manufact ure :
1st. These heads may be cut out of a rolled plate, or may be forged
from billets, and welded to the bar.
2d. The body of t he bar may be upset to form the head, wi t hout
the addition of any extra metal.
3d. Pi l i ng pieces are put on end of bar, and the whole part i al l y
upset, aft er which t hey are pressed or hammered into shape in a die.
4th. Same as last, omi t t i ng upsetting.
5th. I s the Springer patent. Thi s is made by bending a piece of
iron of proper size and same thickness as a bar, requiring head ilm)
somewhat the shape exteriorly of a horse-shoe, wi t h a rectangular slot
on the inside. Thi s is slipped oil to end of bar, and cover-plates, cut
to shape, placed on top and bottom. The whole is then heated, and
pressed or hammered into a die, as above.
6th. Is the Kl oman bar, wi t h Which you are all acquainted.
7th. I s the loop eye, which is formed by bendi ng the bar edgewise
around a pin and wel di ng edges together.
No. 1, or welded bars, are nowadays considered inadmissible.
No. 2, or upset bar, makes a very good bar, and the manufacture of
it will probably be improved, but I am inclined to t hi nk it would be
better i f t hey used a little piling.
Nos. 3 and 4, or piled head, is t he one in most general use, and
makes a very good head; but care should be t aken to keep the diam-
eter of the eye and radius of neck as small as possible, the difficulty of
manufacture increasing wi t h diameter of eye, and the bar should not
be too t hi n to hol d its heat, nor shoul d the head be piled crosswise.
5th. The Springer bar gives any desired latitude to size of head.
All bars of this make t hat I have seen tested broke entirely clear of
the head, except one, which broke at pin-hole at 49,600 lbs., the head
having an excess of onl y 31 per cent. I have broken four heads
longitudinally and transversely, all of which showed solid welding.
6th, or Kl onl an. Thi s makes a very good bar; but the onl y tests I
WIt0LE NO. VOL. CXV.--(TttIRD SERIES, Vol. lxxxv.) 12
178 Bridge Inspection. [Jour Frank. Inst,
have of these were made on bars Mr. Kl oman tested for his private
us e .
The 7th, i f made by a good smith, will, I believe, always break in
body of bar.
For steel bars, heads are designed aher same pattern as those of
iron, but wi t h t hem welding or pi l i ng is not permissible, so t hat either
a Kl oman head or one made by upsetting is here necessary. The
Edge Moor I r on Co. has a pl ant for upsetting steel bars. At the
time I examined it I was f ul l y convinced t hat the process could not
ii~il to make a good head; I have since been told t hat all t hat had
been tested broke near centre of bar when properl y annealed---/, e.,
whole l engt h at one t i me; i f onl y annealed at ends, t hey broke near
where effects of annealing ceased.
The anneal i ng of steel eyebars appears from tests to be a matter of
considerable importance.
As to round and square rods, t hey have eyes and enlarged ends for
screws, made after tile same modes as eyebars, except t hat t he diameter
of the ends of iron bars arc sometimes increased by piling or splitting
the bar, put t i ng in a wedge, and welding all up.
I intended to have embraced specifications as a part of my subject,
but I have detained you long enough. I hope some one will give us
a paper on this subject, and in it show t hat some specifications ask for
impossibilities; t hat others have costly manufact uri ng requirements,
which are omitted after the work is let, and t hat much also could be
done to relieve the labor of estimating by adopt i ng a st andard column
formul a and by adopting rolling loads at so much per foot, or, at least,
using standard engines.
)Iar., 1883. ] Br i dge I~v2ef t i on. 1 7 9
Results of Tests of Eye .Bars wi t h He ads ma d e wi t h Pi l i ng Pi eces--Test s
made wi t h Hydr aul i c Machi nes f i t t ed wi t h Mercury Gauges. .Bars vari ed
in JLength f rom 1 0 / - 0 / / t o 2 5 / - 0 / / .
i
i
Si ze o f B a r !
i n i n e b e s .
6 x l ~
6x
6 x l
6 x l
6 x l ~
6 x l
6x
6 x l ~
6xl~/~s
5 x l ~
5 x l
5 x l
5 x l ~
5 x l ~
5 x l ~
5 x l
5 x l
5x13/s
5 x l
5 x 1 ~
5xl
o
24, ~ 0
30, 000
27,000
'27, 000
25, 600
29,900
30, 600
28, 800
28, 000
28, 300
31, 000
28, 000
29,000
25, 800
26, 500
31,000
28,600
28, 000
29, 000
30, 000
30, 000
=
"7~
43, 700
47, 000
40, 500
42~ 200
45, 800
46, 500
46, 800
47, 400
49, 500
43, 500
46, (DO
47, 000
48,700
37, 000
,37, 800
42, 000
45, 000
46, 500
49, 000
49, 000
50, 000

=
3
R E MA R K S .
5
7
21
13
12
15
12 G r a n u l a r .
24 "
32 F i b r o u s .
17 "
15
32 "
15 21 "
10
8"5
9
11"5
4
5
4
14
. : . . . . . . . . . . . . . . . . . . B r o k e a t b e a d .
i
. . . . . . . . . . . . . . . . . . . . . i
= I
24 G r a n u l a r .
I
66 ) F i b r o u s . B a r n o t w e l d -
e d .
. . . . . . . . . . . . . . . . . . . . . f i n b a r , d i s t a n c e a n d
c h a r a c t e r o f f r a c t u r e
n o t n o t e d .
85 F i b r o u s .
9 ' 5 24 "
16 24 "
14 20 "
16 20 "
. . . . . . . . . . . . . . . . . . . . B r o k e a t h e a d . - - 11
13 . . . . . . . . . . . . . . . . . . . . . [
1 8 0 Bri dge I nspecti on. [ J o u r . F r a n k . l n s t . ~
Results of Tests of .Eye Bars, .Eta., Continued.
S i z e o f B a r
i n i n c h e s .
H
5 X1 ~
5xl ? ~
5 x l
4 x l Y4
4 x l ~
4 x l ~
4 x l ~
4 x l ~
4x
4 x ] ~
4x1~/~
4 x l ~
4 x l
4 x l ~
4x
4x
4 x l
3xl~/~s
3x
3 x l
3 x l ~
: ~ x ~
= = . - ' ~
0
..~ ~ = ~
30, 000 50, 000
30, 400 51,000
31,000. 51,000
28,000 43, (DO
30, 000 48, 6n0
'29, 000 45, 000
27, 300 47, 500
29,000 47,800
34, 000 48, 000
27,400 48, 000
28,300 48, 5(~)
28, 700 48,700
30, 500 49, 290
30, 000 49,600
32, 000 49", 800
29,700 50,000
31,800 I 50, 000
30, (DO ] 46, 000
30, 600 46, 400
. . . . . . . . . . . . . . . . . . . . . 50, 000
31,700 5~, 000
i
32, 0 0 0 I 50, o ~ )
h~
O
: 2
og..a
16
] 5"2
15
10
12
13
18
15"5
6' 8
13"3
16"6
14
13
16
15
16
15
17
14
12
17
15"3
9-
;2
R E MA R K S .
20 i F i b r o u s .
60 I "
136 [ "
. . . . . . . . . . . . . . . . . . . . . . i B r o k e a t h e a d .
! .
80 : F i b r o u s .
i
26 I
i
I
24 I G r a n u l a r ~
16 F i b r o u s .
20 "
13
24
2O
24
" B r o k e i n b a r ,
p o s i t i o n o f f r a c t u r e
n o t n o t e d .
i F i b r o u s .
24 i " B r o k e i n b a r
p o s i t i o n o f f r a c t u r e
i n o t n o t e d .
. . . . . . . . . . . . . . . . . . . . . l F l b r o u s .
1
22 , "
2~ I t,
Mar., 1883.] Bridge Inspection. 181
" q ~ u o I
e o t l o u ! u ! p o a n
- s ~ o u ~ u o i ~ u o I ~
" u o I l
' ~ U O l O " ~ o o a e ~ I
$
s p u n o d u t
q a u I a . ~ ' a b s a e d
q ~ u e . ~ I s o ~ t a ! ] t 1
. o
i
" q l ~ u o I
s o t o u ! u I p e a u ~ ~ o ~ : ~ ~ ~ ~
- s ~ o t u u o ! ~ U O l ~ [
+
, ~
~ g
+
+ p ~ o ~ , ~ . ~ ~ $ $ ~ ~ ~$ $~
q o u t o J ' ~ n b s . l o d
q g t ~ t l o . t ~ s o ~ m ~ t ~ i / - 1 ~ - ~ - ~ - ~ , ~ - - ~ - _ . - ~ - ~ - - ~ - ~ - . . - ~ - t ~ . - t~ ~
- s o q o u . t
u t s s o u ~ I o ! q J ~
_ +
2
c