generally are found to suffer under the excessive wheel loads.
There is no remedy for this except in the gradual improvement of the
road-beds of the outlying lines.
There will be in the future, as there always has been in the past, a
constant struggle between Track and Load, which will only terminate
128 DISCUSSION ON DESTRUCTION OF RAILS.
wlien "we arrive at a standard weight and section of rail, which may not
be further modified witliout unfavorably affecting its economical em-
I think, therefore, it is entirely useless to look for a present remedy
to suffering road-beds in a reduction in the weight of cars. The lines
which have found the increased capacity of their equipment both neces-
sary and economical will make no change in this direction, and it only
remains for those who feel the burthen to come gradually up to the
The Special Committee of the Society now considering the relation
to each other of the Sections of Railway Wheels and Eails may, and it is
hoped 'will, throw light on this question in its report.
D. J. "Whittemoke, Past President Am. Soc. C. E. — In this increased
weight of the rail, from 56 to 60 or 70 pounds to the yard, has the quality
of the material been kept up? On the line with which I am connected
we have a rail of 65 or 70 pounds to the yard, that, when new, will stand
less traffic than one of 56 or 57, made under the same specifications as
the former. Does not this imi^ly that as the mass is increased, more
thorough working of the material, in the process of manufacture, is
James B. Feancis, Past President Am. Soc. C. E. — I suppose this
question of flow is a case of exceeding the elastic strength of the metal.
By this I mean a strain, so that the metal does not return after the load
is lifted. The remedy would be to reduce the strain ; there is no other
way to accomplish it. I don't see that it depends upon the strength of
the rail, but upon the strain that is applied to the u^jper point in the
rail. If the strain there is beyond the elastic limit this flow will take
Max J. Becker, M. Am. Soc. C. E. — It seems to me that there ought
to be no difficulty in providing a rail capable of resistingthe force which
is said to cause a flow in the metal. The trouble is, that the increase in
the carrying capacity of the rail has not been kept up in the same ratio
with the loads imposed. In the earlier fifties the general weight of rail
was about 50 pounds per yard. At that time the freight cars carried 10
tons, the locomotives weighed 20 or 25 tons. To-day we are carrying
30 tons in our cars and our engines weigh 60 tons, which shows an in-
crease of 300 j)er cent, in the loads. At the same time, the largest rail
sections in this country now rarely exceed 74 pounds, so that the in-
crease in the sectional area of the rail would be about 50 jjer cent.,
while the loads have been fully trebled. This is disproportionate.
When you remember that the head of the rail has been but very little
increased in bearing surface, you will readily see the utter disin'oijortion
between the provision made for carrying and the amounts to be carried.
DISCUSSIOX OK DESTRUCTION" OF RAILS. 129
Still, I am uot prepared to accept Mr. Dodge's tlaeory, that any con-
siderable portion of rail failure is due to the yielding or flowing of the
metal under the large concentrated wheel loads. Is it not more likely
that the effect observed by Mr. Dodge is caused by the sliding of wheels
"while passing over the rails with brakes set?
It is the general opinion of observing railroad men that the increased
speed of the trains, added to the greater weight — in other words, the
increased momentum — causes the destruction of the metal in the rails.
If this opinion is wrong, have we not been at fault in making the large
customary allowances for impact in computing the members of our
J. T. Dodge, M. Am. Soc. C. E. — I think Mr. Francis has presented
the real point. The difficulty is not that the rail is uot strong enough
now to carry this load from tie to tie, but that the surface of the rail is
uot strong enough to support the weight which disturbs the fiber. The
elasticity of the rail is overcome and the metal takes a set, it simply
flows. Widening the head of the rail will not be a sufficient remedy ;
the wider it is, the greater difficulty to make the wheels fit that head.
The drivers will crush the outer edges of the head, because they will
bear heavier on the outer edges than the car wheels will. It is a question
of the elastic limit of metal.
J. B. Johnson, M. Am. Soc. C. E. — I don't think there is any dis-
crepancy in theory between the increased flow of rails under slow loads,
and the increased load on a bridge under a fast train. We all under-
stand that iron and steel are plastic bodies which flow the same as any
other plastic body. We know that if we take a cake of wax and strike
it with a hammer it flies to pieces; but if a very small weight be left
upon that was a few hours it will settle itself into it. The distortion of
a plastic body results more from the length of time during which the
force is applied, than from the amount of force. If the force applied to
this iron beyond the elastic limit is increased, the amount of flow will
be almost directly proportioned to the time during which the force is
applied. In the case of a bridge resisting the force of impact, the sud-
denness of the load is in proportion to the speed of the train. If the
load be instantly imposed, the effect of the load is twice the static load,
so that both these results are according to theory and they are not at all
The wheels of locomotives and of cars are not and will not be of the
same shape, since locomotive drivers will grind themselves badly. The
average condition of the car wheel will be somewhat conical and a worn
driver will always be grooved. It is difficult to say what the surface of
the rail should be; but we may say this, that the shape which the loco-
motive driver finally attains as the effect of grinding on the track
130 DISCUSSION" ON DESTRUCTIO]Sr OF RAILS.
depends on tlie original shape of the rail. If you have a rail that was
originally sloped inward, and the driver grinds upon such a rail, it
-would grind to such a slope. If the rail is fiat on top and the driver
grinds it, it will grind to a horizontal shape. I have wondered whether
the rail head could not ba rolled to a slight slope, so that the final slope
of the rail would be that of its original slope. "We know that when a
new rail is put down, the bearing will be really on a quarter of an inch
of the width of the rail, so that the rail is but a quarter of an inch in
width, in fact, but when the rail has Avorn down, it always has a slope
inward; this inward slope depending, in form, on its original form, and
on those of the wheels running upon it. Why should not the rail be
rolled to that shape in the beginning, so that it would bear over a large
surface, just as when the rail becomes worn? The final shape we find
is the result of the original; if we would give to the rail this final shaj^e
originally, then there is no reason why the shape should change. I
would like to ask the practical men what is the objection to rolling rails
slightly sloping to the inside?
Mr. Gottlieb. — It could be done, but it would be poor economy to
do it; but there is a very simple way to obtain the same result. On the'
European roads, or at least on the German roads, they give the rail the
proper inclination by adzing or notching the cross-tie at the rail seat,
which gives the same effect.
Mr. J. T. Dodge. — Embracing the opportunity offered by a view of
the foregoing discussion, I would remark that my subsequent observa-
tions confirm the facts stated in my original paper, viz., that a large-
amount of the destruction of rails is due to a flow of the metal.
Mr. Francis stated the deduction from the facts very concisely and
correotly. Mr. Becker expressed a doubt of the facts and at the same
time suggested the sliding of the wheels as an exidanation of them.
This being a question of fact, it is better to appeal to observation
than to argument. If observation does not confirm the statements-
made, then the evil does not exist as stated.
While the whole of the discussion is highly interesting and instruc-
tive, much of it is not strictly pertinent, and hence is out of the range of
AMERICAN SOCIETY OF CIVIL ENGINEERS.
Note. — This Society is not responsible, as a body, for the facts and opinions aclvauced io
any of its publications.
(Vol. XX.— April, 1889.)
ENGLISH AND AMERICAN RAILEOADS COM-
PARED IN OPERATING EXPENSES.
By Edwakd Bates Dorsey, M. Am. Soc. C. E.
For tlie information of a very prominent official and large share-
liolder in English railways, the author made some of the following
comparisons. The result was so astonishing that he decided to enlarge
upon it, and make a short pajoer, in hope of getting the subject freely
discussed. The English and American practice differ so widely that
both cannot be right; engineers and railroad officials should discuss it
freely, and adopt the principle that is the best and most economical,
regardless of its nationality.
"When the author commenced this investigation over five years ago,
it was with the intention of showing the American engineers that it
was the best policy to build good roads at first, as the English do, and
save in the cost of operating expenses. The author then thought that it
would be possible, on roads with large traffic, to save more than the
additional interest charge, but as the investigation progressed, he was
greatly astonished to find that the Americans were operating their roads
much cheai^er than the English, though built probably at less than one-
fifth of the average cost of English railways.
This so surprised him that he thought there must be some mistake
in his data and investigations. Acting upon this belief, he postponed
132 DORSEY ON ENGLISH AXD AMERICAN RAILROADS.
tlie reading of the first paper for six months, until he coukl again visit
England and re-examiue the subject and data.
It is now more than three years since his first paper entitled "English
and American Kailroads Compared," was read before this Society.
Since then the subject has been much discussed by the joress and rail-
road officials. About the only thing that has been said against it is,
that it was based upon assumed data; in reality all the important
iigures given in it were from official sources, except the tonnage and
passenger mileage; unfortunately the English roads do not give these.
In reference to this, the author will quote from his first paper or book,
page 16:* "No return of these all-important items for comparison is
made in England. The author, after careful inquiry and investigation,
decided that the average freight charge on all freight moved in the
United Kingdom was about li pence, or 2.5 cents per mile per ton; but
la order to be conservative he has taken the average charge at 1 penny,
or 2 cents i)er ton per mile."
"In order to get the ton mileage, the receipts from freight,
as reported by the companies, in pounds sterling has been multiplied
In the supplementary paper a year later, and after more thorough
investigation had been made, the author said on page 790:1
" He hoiked that the English railways which are so largely interested
in this question will promptly replace his estimate of ton and i:>assenger
mileage by their official fig ares. Until this is done he claims that
these figures should be accepted as correct."
Since this was written it has been investigated for two years more,
iuUy confirming the preceding opinion and data.
All the data and figures in this paper are taken, for the English
roads, from the "Board of Trade Railway Returns," or "General
Report," and for the American roads from "Poors' Manual," or the
annual reports of the companies.
In traffic expenses the Pennsylvania Railroad includes "rates and
taxes," and "compensation for personal injury " and "compensation
for damage and loss of goods," amounting to $322 370 for 1887, which
should be deducted to make it correspond to the English roads, as on
them they are not included in traffic expenses.
* Transactions, Vol. XV, No. 318, January, 188G.
t Transactions, Vol. XVI, No. 313, October, 1886.
DORSEY OX ENGLISH AND AMERICAN RAILROADS. 133
In the accompanying table tlie five principal English railways have
been selected. Among the American railroads the Pennsylvania Railroad
Division has been selected as representing the best constructed American
road, and the Knoxville Branch of the Louisville and Nashville system
as a sample of the cheaply constructed class, both running through a
comjiaratively rough country. Both of these roads have most of their
traffic in one direction. This reason is very frequently given by English
officials for the high freight rates on their railways.
At first the author included in the Pennsylvania system, the Phila-
delphia and Erie, and the united railroads of New Jersey divisions, but
on reflection, he thought best to exclude them, as it might be said that
the level or easy grades on these roads were particularly favorable to
cheap transportation. The Pennsylvania Eailroad Division was retained,
as it traverses a broken country with many heavy grades, and one sum-
mit of 2 154: feet above tide.
In absence of official data as to the average charges for trans-
porting one ton of freight one mile on the English railways, the
author has assumed that the average charge is one penny, or two
cente, per ton per mile (one pound sterling = ^4.80), being the same
figures as given in his book and papers on pages 16 and 789, which
figures have not been contradicted in any way, biit have been reiseatedly
These comparisons speak very strongly in favor of the American
system. The high terminal charges (including receiving and delivering
freights by carts), are generally given as an excuse for the very high
freight charges in England; the last two paragraj)hs show that there must
be other reasons than these. In absence of any data as to what this ex-
pense amounts to, the author has in paragraph, as noted, deducted from
the operating expenses of the English railways everything included under
the name of " Traffic Expenses " (coaching and merchandise); this must
certainly include all terminal charges and cartings. On the American
railroads no deductions of any kind from the total operating expenses
have been made; notwithstanding this, there is still a very large per-
centage in favor of the American roads, on the Pennsylvania it being 56
per cent, in its favor, compared with the London and North "Western.*
*Being Hi per cent, in favor of the Pennsylvania Kailroad when compared to the Lon-
don and North Western.
134 UORSEY ON" ENGLISH AND AMERICAN" RAILROADS.
In paragraph 11 all traffic expenses (coaching and merchandise) have
been deducted from the operating expenses of all roads; this shows a
very large percentage in favor of the American railroads.
Paragraph 6 shows the error in the statement usually made that the
English railways are operated more economically than the American, on
account of the operating expenses of the English roads being about 52
per cent, of the earnings, while the American roads are al:)out ten per
cent, higher. If the earnings on both roads were derived from the
same charge for moving freight, the operating expenses of the Penn-
sylvania road would be only 21 per cent, of the earnings against 52 on
the London and North Western, or 148 per cent, in favor of the American
It is frequently stated that the English railways are operated more
cheaply than the American, because the train mile on the former costs
somewhat less than it does on the latter, which is entirely wrong, if
jn-oper allowance is made for the difference in the average number of
tons in the train load, which average, as per paragraph 3, is 79 tons on
the London and North Western, against 207 on the Pennsylvania. Para-
graph 13 shows that if proper equation is made for the heavier load, the
cost per train mile will stand on the London and North Western ^.6588,
against ^.3365 for equal load on the Pennsylvania, or 96 per cent, in
favor of the latter.
There must be some cause for this great difference in the operating
expenses of these roads ; in the author's opinion the following are the
Ijrincipal reasons :
First. — Too light loads to the English trains, both freight and pas-
Second. — The universal use on American roads of rolling stock with
bogie-trucks, which move with much less friction and wear and tear
than the rigid wheel base rolling stock used by the English railways.
Third. — The use on the American roads of freight cars carrying a
much greater percentage of paying load to dead weight than the wagons
used on the English railways.
Fourth. — The great speed at which the English freight trains run.
FiftJi. — The use on the English roads of light locomotives, with in-
side connections that cannot be as cheaply or as easily repaired as the
American locomotive, with its outside connections.
Sixth.— The custom on the English railways of giving each indi-
TABLE No. 50.
ENGLISH AND AMERICAN RAILROADS COMPARED IN OPERATING EXPENSES.
Vol. XS, page 134.
Loudon and North
all Railways of
h. & N. W.
£1=$4.80. Id. = 2 cents.
All figures representing cost are in cents, except
the average cost per mile of railway operated.
and in favor
and in favor
and in favor
and in favor
and in favor
and in favor
which is in dollars.
1. Percentage of total tonnage mileage transported
2. Percentage of total tonnage mileage transported
4. Average charge for transporting one ton one
5. Percentage of operating expenses to earnings. .
6. Percentage of operating expenses to earnings
on the Pennsylvania Eailroad Division, pro-
vided it charged the same rates for freight as
are charged on the English railways, viz..
7. Percentage of increased earnings that the Penn-
sylvania Eailroad Division would receive if it
worked with their present operating expense
and received the same freight rates (Id. per
ton per mile), as the English railways now
8. Percentage of traffic expenses (coaching and
merchandise) to total operating expenses
9. Average cost of transporting one ton one mile.
The cost tor the English railways is found by
multiplying the estimated charge of Id. per
ton per mile by the percentage of operating
expenses to the gross earnings, as per para-
graph 5 above ; cents
10. Average cost of transporting one ton one mile.
deducting on the English railways all ■• Traffic
Expenses (coaching and merchandise)," and
deducting nothing from the total operating
expenses of the American railroads ; cents. . .
11. Average cost ol transporting one ton one mile
deducting all "Traffic Kxpenses " (coaching
13. Average cost of train mile— freight and pas-
senger : cents
13. Equated cost of train mile on the Engiish rail-
ways, provided they transported their pretent
train loads at the same cost per ton per mile
as is done by the American railroads, to which
Per cent. 112
DORSET ON ENGLISH AND AMERICAN RAILROADS. 135
vidual shipper the exclusive use of a freight wagon, even for very small
quantities of goods.
The Pennsylvania Kailroad Division is one of the few American
railroads that approximate iu thoroughness of construction to the aver-
age English roads. Even on this line there are ihany cheap station
buildings, islatforms, etc., that require constant repairs and renewals.
Consequently, their maintenance is more expensive than the permanent
masonry structures on the English roads. This appHes still more
forcibly to the inferiorly constriicted American roads.
Of the 150 000 miles of railroads iu the United States, it is safe to
say that there is not 5 per cent, of them ballasted up to the English
average standard, and perhaps more than one-half of the remainder
without any pretense whatever to ballast.
The large English railways are very similar to each other in the
general details of their operations; this is shown by the preceding table.
Erom all the data that are available, it appears that the cost of trans-
porting one mile, one ton or one passenger, is about the same on the
jjrincipal English railways. The question has been frequently asked,
"How can the English railways i>ay the interest on their great cost, un-
less they continue to charge their i?resent high freight rates?" This is
entu-ely foreign to the object of this paper, which proposes only to dis-
cuss the cost of operating expenses in the two countries. The author
is, however, thoroughly convinced, that on the English railways large
reductions could be made iu the i^reseut freight rates and cost of
ofterating expenses, without diminishing the present dividends in the
On the English railways there is a great variety of rolling stock,
differing very widely; this is especially true regarding the locomotive.
It is not possible that these different varieties, differing so very much,
can all be equally good. It is strange that the best type has not been
selected, and generally adopted, especially on the same ro*id. During a