American Society of Civil Engineers.

Transactions of the American Society of Civil Engineers (Volume 81) online

. (page 86 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 86 of 167)
Font size
QR-code for this ebook

it ^vill be observed that there is considerable difference in the frictional
resistance between wet and moist samples of the same materials, and
therefore the question arises: Might not the frictional resistance be
still lower than is indicated by the tests when the clays are subjected
to the conditions that will exist in the foundation materials of the dam ?

At first thought, the coefficients of frictional resistance for shales
(0.40 and 0.50), assumed by Mr. Scheidenhelm in the design for the
reconstruction of the dam, may seem to be rather low. However, it
should be remembered that the shales in question have "pronounced
horizontal laminations" and are somewhat disintegrated, and that clay
seams exist in places.

The lamination planes in the fire-clay shales of the coal measures
often show a slick, smooth surface, sometimes of a more or less slip-
pery nature. It is very evident that the cohesion between the sub-
strata at these planes must be exceedingly slight, and, unless the planes
of contact are very much warped or of limited extent, the resistance of
the materials to sliding must be rather low.

Mr. H. L. CoBURN,* M. Am. Soc. C. E. (by letter). — The Engineering

Coburn. profession is to be congratulated that there are engineers who have
the time, the disposition, and, more especially, the ability, to prepare
such a paper as this. One rarely sees a more complete exposition of
an engineering problem from inception to finish than is here presented.
The writer says from inception, for, to all practical purposes, this
was an entirely new problem, as few or no data were available for
the author's use.

As originally built the Stony River Dam was designed to meet
certain stated conditions of foundation, which, subsequent events
proved, did not apply, and the failure of the dam is attributable to
this fact alone and not in any way to the type of structure. 'No
dam designed to fulfill the conditions given could have withstood
the treatment to which this one was subjected, and the writer thinks
it an evidence of sound fundamental principle that the dam failed
only in part and that part so gradually.

This is neither the time nor the place to discuss the conditions
that led to the designing and building of this dam orj data that were
wholly inadequate and erroneous. The great lesson to be learned
from the failure is the advisability, not to say necessity, of complete
preliminary investigations on which to base design, and adequate and

' '•* ' * New York City.


competent engineering supervision of construction — money thus spent Mr.
• a m 4. » Coburn.

IS 'well spent.

As to the reconstruction, though the writer doubts that all the
safeguards against a possible second failure, which Mr. Scheidenhelm
has taken, were necessary, yet he feels that, in the circumstances, they
may be justified. If more accurate data were available as to "passive
thrust" or resisting power of the local soils, and more real information
as to coefficient of friction between these soils and the concrete cast
thereon, engineers might criticize his details in some respects, but
with the limited knowledge of actual values, and particularly in view
of the fact that this was a case of rebuilding a structure which had
partly failed, and about which very sensational stories had been cir-
culated, the writer considers that if the dam was to be rebuilt at all,
it was good judgment to assume the "worst possible conditions" and
to make assurance doubly sure.

H. F. Dunham,* M. Am. Soc. C. E. (by letter). — In the discussion Mr.
of a paper on the repairs made to a structure, it may not be in order '^"°'^*™-
to refer to the reasons for erecting that structure; but, if such refer-
ence is allowable, replies to two or three questions may be of interest.

An article published in a technical journalf at about the date of
the construction of the Stony River Dam contained a few words relat-
ing to the object to be secured by the improvement. At some point on
the North Branch of the Potomac below its Stony River tributary
there was a paper mill with an insufficient supply of wash-water in
dry-weather periods. There was no mention of the use of an increased
flow of the river for power purposes. There were no figures for the
dry-weather discharge of the main stream at the mill, nor of its
tributary. The additional quantity of wash-water needed -^jas not
given, and, furthermore, there was no longitudinal profile of either
the North Branch or the Stony River, and no description of the river
valleys or reference to any other method of providing the unestimated
but desired supply.

It should be generally admitted that if, within a few years, the
supply of wash-water has approached the desired quantity there would
certainly be enough at all times, if the natural or original conditions
pertaining to those streams could be restored. When the entire dis-
trict was heavily wooded, the winter snows were longer in melting,
the ground did not freeze to such depth as now, and more of the
spring rainfall remained as ground-water for a longer period. More-
over, when floods were less frequent, fallen tree trunks and branches
lodged in narrow channels, holding the water back. The beaver built
his dams, with similar results, and when they were old he, too, repaired

* New York City.
t Engineering News.

1040 DISCUSSION : reconstruction of stony river dam
Mr. them. All of this was favorable to a longer and more uniform flow

Dunham. . . , . • i i i •

01 water in dry-weather periods than obtains at the present time.
It may be within reason to inquire how far Nature's regulation of
flow in such streams can now be imitated. The forest and frost are
items that cannot be changed, but, given favorable topography and
soil, the fallen timber dams and those made by the beaver with the
higher ground-water table could be copied successfully and possibly

The photographs in the paper, and the dam itself, 1 000 ft. long
at the most favorable location, indicate a valley of some width. The
sandstone mentioned, and the fact that the water found its way
under the cut-off wall, show the existence of a somewhat porous soil.
It is evident that the structure described was expensive. Had a part
of that expense been used to construct a number of low but permanent
dams at favorable places for low dams, each to be provided with a
cheap low-water sluice-gate that would require a minimum of attention,
a considerable change in the dry-weather flow of the stream would
have been secured. The author refers to the existence of ground-water
in and about the foundations of the dam. To raise the level of
such ground-water over considerable areas gives available storage,
and water thus stored is not subject to such rapid evaporation as
surface water.

The dams, of course, would hold back surface water as well as
ground-water. The extent to which the flow of a stream is regulated
by higher ground-water level is often surprising. This is well illus-
trated by rivers flowing through wide flood-plains which, sponge-like,
retain large quantities of flood water which are gradually returned
to the river in its lower stages.

At certain seasons, especially in the spring, streams of moderate or
sluggish current may carry too much silt to be suitable for a supply
of wash-water. In the article previously referred to no mention was
made of any filtration process, but, if that method is used, a
reservoir below the mills suitable for the purpose of sedimentation
and refiltration should have substantial value.

The writer has no wish to "go behind the returns" or to ask for
any reference to the business affairs of business men that need not be
disclosed. . The first studies pertaining to any improvement are always
interesting, and they are quite apt to be overlooked in the description
of a finished work.

How the cost of the dam as it is now compares with the estimated
cost of a gravity-section dam of the same general dimensions would
be of interest, as would also the question of a curved versus a straight
dam for structures of either type in that place.


Orrin L. Brodie,* M. Am. Soc. C. E. (by letter).— The author is Mr.
to be thanked for the lucid and well-arranged presentation of the ^^°^^'^-
subject matter relating to the reconstruction of the Stony River Dam,
and the thorough manner in which he has treated the various phases of
that work.

One of the most striking features of his design, to the writer at
least, was the unique method of anchoring the up-stream heel and
cut-off. Relative to this matter, certain statements of the author in
his brief outline of the previous failure of the structure were impres-
sive. These were:

"Failure occurred where the up-stream cut-off wall extended only
a short depth (5 to 7 ft.) into the over-burden."

"Failure was caused by undermining due to leakage."

"The type of dam was not in any way a cause of failure."

Thoughtful consideration of these three statements taken together
forces on the writer the following query:

Would not a dam subject to considerable head of water, such as
the one described, but containing sufficient mass that could contribute
by its dead weight a resistance per se to sliding and overturning ten-
dencies, be more satisfactory in every respect than one of the type
here considered? Even though the hollow concrete type, such as the
one described in the paper, may possess decided advantages in the
matter of stability against overturning, there is always the possibility
of the admission of water below it, which, besides increasing its
tendency to slide, destroys any advantage inherent in its form in the
matter of stability. Besides, as shown by the author, an up-stream
cut-off thoroughly anchored to the heel seems to be a vital necessity.
It is also significant that the reconstruction involved a quantity of
masonry equal to that of the initial construction, which might better
have been utilized in the first place.

The method of computing the storage effect of the reservoir area
with respect to the spillway run-off was interesting to the writer, as
was the detail of the spillway shape, as depicted on Plate X and noted
by the author as being a more advantageous crest than that of the old

An investigation as to spillway run-off with respect to two basins,
the waste weirs of which were at different crest elevations, was made
by the writer, and involved an interconnecting tunnel between these
two reservoirs (constituting the substitute supply works, connected
with the new Kensico Reservoir for New York City).

The understanding of the method will be facilitated by the
following nomenclature :

* New York City.


1042 DISCUSSION : reconstkuction of stony river dam

Mr. Volumes are in cubic feet ; flows in cubic feet per second ; and times,

in seconds.

H^' , H^, and Si", S^", are elevations of water surfaces in the

respective reservoirs, Nos. 1 and 2;
i = time for Reservoir No. 1 to rise through a small distance
' ,- 'iF^ = increment of capacity due to {B." — -H^i') for Reservoir
No. 1;
Fg = increment of capacity due to {B." — Ho') for Reservoir

No. 2 ;
1^ = average inflow during time, t, from water-shed of Reservoir

No. 1;
1 2 = average inflow during time, t, from water-shed of Reservoir

No. 2;
i = average flow, during time, t, through tunnel and from Reser-
voir No. 1 to Reservoir No. 2 ;
Q-^' and Q^' =flow over respective weirs at the beginning of time,

t; and
Qj" and Q^" = flow over respective weirs at the end of time, t.

For Reservoir No. 1

t- Y^ 0)

For Reservoir No. 2 :

t=^ ^ (2)

Eliminating i between Equations (1) and (2) :
from Equation (1) t f/, — i — ^ (Q/ + Q/')] = J\ (3)

from Equation (2) ^ Tj, + i _ (Q^' + Qo")! = Y.^ (4)

from Equations (3) and (4)

. = , ''' + ''■' (S)

A + A — 2 (^1' + ^1" + ^2' + Q2')

Method, by successive approximations, of applying the foregoing
expressions :

Fj is a direct function of t and i, that is, T^i = / (i, i) and must
be so considered.


Begin with Equation (5) : First, assume values for V^ and Vr, and Mr.
corresponding values for the other terms in the right-hand member. ^° '^'
Substitute the value of t thus found in the left-hand members of
Equations (3) and (4), with the corresponding new value of i, until
Equations (1) and (2) are satisfied. The second value of t can thus
be tested by Equations (1) and (2). The values of V^ and V^ of
Equations (3) and (4), respectively, must agree with the values of F^
and Fj ^^ Equation (5).

The calculations, as briefly outlined in the foregoing, may be made
in connection with the use of chosen run-off diagrams, together with
capacity curves prepared both for the weirs, tunnel, and reservoirs.
The results may be shown by appropriate curves of reservoir rise with
time intervals from the beginning of the storm, or with respect to
any stage desired.

In connection with the consideration of the spillway shape, the
writer has determined the applicability of a parabola, the origin of
which may be taken at the actual crest of the weir and the parameter
at about 1.8 times the head of full reservoir level on the so-called
"theoretical crest" (that is, the actual crest of a corresponding thin-
edged weir). This parabola, with axis vertical and through the crest,
will determine approximately the lower nappe or sheet of overfall
for a weir.

The study leading to this conclusion was based on weir experiments
by M. Bazin. This parameter may be increased for practical consider-
ations to as much as 2| times the head.

The writer is not aware of the detailed method by which the shape
of the new spillway crest of the Stony River Dam was fixed, but,
by the foregoing method, he calculated the crest for the given head
and was interested to find that, as far as could be ascertained by
careful scaling of the spillway of Plate X, a practical identity resulted.
The advantage of a larger value than 1.8 for the parameter factor
used by the writer in this instance is that it permits of flattening
the curve just down stream from the crest, thus improving the flow
conditions. For a solid masonry weir, a larger parameter enables the
parabolic section to be extended to a lower elevation than otherwise.
In the case under consideration it was obvious that the comparatively
steep apron slope required a minimum value for this particular para-

The author intimates that the preliminary soundings were made
by test pits and auger borings, that failure occurred where the up-
stream cut-off wall extended only from 5 to 7 ft. into the over-burden,
and that the technical advisers consulted on the original project
were not familiar with the local conditions. Later, it appears, the
results of core borings driven for reconstruction investigations disclosed

1044 discussion: reconstruction of stony river dam

Mr. lentils and alternations of pervious with impervious strata, and that
""^ '^' sandstone boulders or false cliffs had been mistaken for bed-rock be-
neath the original cut-off construction.

These facts cannot emphasize too strongly the great importance
of extensive sub-surface investigation preliminary to construction such
as this, and the author's words in regard to foundation conditions as
uncovered in the subsequent explorations, that they "required the most
serious consideration and care", are conclusive.

Mr. William Cain,* M. Am. Soc. C. E. (by letter). — The writer is
' especially interested in the author's use of "anchoring walls" to increase
the margin of security against the sliding of the dam. Such walls,
projecting below the foundation, have been used repeatedly in dam
and retaining wall design, but the projections were generally of small
depth, so that little attempt has been made to state the principles
affecting their design. The author's distinct contribution consists in
the use of anchoring walls of such depth that the weight of the soil,
from the foundation to the level of the lower end of the anchoring
wall, is quite appreciable, and is utilized in increasing the resistance
of the dam to sliding by lowering the possible plane of sliding from
the foundation level to one passing through the bottom of the anchor-
ing wall. He has entered into great detail regarding the principles
affecting this design, and, in connection with it, has discussed the
passive resistance of earth placed below the dam to sliding up a possible
plane of rupture.

It is to be regretted that the author, in his experiments to deter-
mine the combined "adhesion" and friction of the various soils, did
not apply increasing weights to the box containing the soil, and like-
wise measure the pull for each weight in turn, corresponding to
'•impending motion"; meaning by that term, that no actual sliding
occurs, but that any increase in the pull, however small, would cause
actual motion. If this had been done, the actual values of the coeffi-
cients of friction and cohesion could have been ascertained and used
in the analysis. As the pulls were only recorded when the box con-
taining the clay was in actual motion, the results give neither the
friction coefficient alone nor the combined full cohesion and friction,
such as would be actually exerted near the foundation of a stable
dam, or perhaps on some lower plane or curved surface.

The results refer only to a combined friction and cohesion ("ad-
hesion", as the author characterizes it, for this case), when the box
was in motion. The numerical values of the "coefficients" in Table 2
are so much larger than the coefficients of friction of various clays,
as given by Bell, as quoted in Table 7, that evidently a very appreciable
amount of cohesion ("adhesion") was exerted during the motion. This

* Chapel Hill, N. C.



negatives the idea, suggested by the writer in a previous paper, that, Mr.
after motion began, possibly all cohesion was destroyed, so that only *'°'
friction remained to resist the pull. Evidently, from the results,
although a large part of the full cohesion of the solid clay was not
exerted during the motion, still an appreciable part of cohesion in
addition to the friction was exerted during the actual sliding of clay
on clay.

Coulomb's laws, concerning "impending" sliding of earth on earth,
are symbolized in the equation :

Q=f P,^ c A (1)

where Q = the total resistance to sliding in the plane of shear, in
pounds ;
/ = the coefficient of friction ;

c = the cohesion in the plane of shear, in jwunds per square
P„ = the normal pressure in the plane of shear, in pounds ;
A = the area, in square feet, of the plane where shear is im-

Also, if = 0.213), on substituting
in Equation (1), we find,

^ = 2 030 400 X 0.213 + 780 X 660 = 947 300 lb.,

which is but little below the water pressure tending to cause sliding.
In fact, the resistance to sliding is exactly equal to the water pressure
for (^ = 13° 50', which is a possible value.

This computation involves the assumption that there exists a co-
hesion between the concrete and clay at the foundation of 660 lb. per
sq. ft., and the fact that the dam stood renders the assumption proba-
ble. No illustration can be given that more forcibly points to the need
of "comprehensive experimentation" to determine the coefficients of
cohesion and friction, not only for earth on earth, but likewise for
concrete or other masonry on earth; for here is a dam that stood,
which presumably should have failed if no cohesion was exerted at the
base. This supposed cohesion that engineers (the writer included)
have hitlierto ignored, in computations ailecting the sliding of dams
and retaining walls, is probably a vital element concerning stability,
and it doubtless has saved many walls from destruction by furnishing
an additional resistance to sliding over that due to friction alone.

The argument that the resistance to sliding can be computed in

Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 86 of 167)