American Society of Civil Engineers.

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

. (page 57 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 57 of 167)
Font size
QR-code for this ebook


In other words, thirteen specimens of the 1 :3 -.6 mixture were made,
and the one to which the integral water-proofing, clay, had been added,
ranked third in quality, the two ahead of it being special.

Now, as to No. 22, to which hydrated lime had been added, Mr.
Bakenhus states:

"This specimen is also poorer than eight of the other ten similar
specimens in series, showing conclusively, so far as one specimen can,
the deleterious effects of hydrated lime when used in concrete immersed
in sea water."

This fijiding is entirely in accord with those of other investigators,
and entirely in accord with the now generally accepted theory as to the
cause of the disintegration of concrete by sea water, but it happens
that, primarily, this has nothing to do with the theory of integral
xVater-proofing.

It has been shown* that the lime is the element in the cement
which is attacked by the sulphates and chlorides in the sea water, and
this fact has been substantiated by other investigators. The addition,
then, of more lime could not fail to increase this action, and, in the
specific case of Specimen No. 22, the lime added replaced an equal
percentage of cement, and this made a leaner mix.

As to Mr. Larsen's own tests, they certainly do not show that
"integral water-proofing has a detrimental effect on concrete." Such a
statement could only mean either that all integral . compounds were
the same, or that he had tested all known compounds, and there is
certainly nothing shown in his tests to substantiate either claim. As
a matter of fact, there are many of these materials on the market.
The Bureau of Standards has reportedf tests on more than twenty,
and the theories on which they are based are radically different.

In fact, the hazy comprehension with which the subject of integral
water-proofing continues to be discussed is astonishing, in view of the
simplicity of the rational theory on the subject, and seems to have
its foundation in a very general lack of knowledge of some of the
most important characteristics of the most widely used of engineering
materials, namely, Portland cement. Portland cement seems to con-

* Technologic Paper No. 12, Bureau of Standards.

t Technologic Paper No. 3.



692 DISCUSSION ON TESTS OF CONCRETE IN SEA WATER

Mr. tinue to be regarded as a more or less passive agent, whereas the
^^®"' experiments of Michaelis, Cabolet, the Bureau of Standards, Nathan
Johnson, and others have shown, on the contrary, that it contains
elements which, though latent under ordinary conditions, are capable,
under proper conditions, of a great increase in volume during hydra-
tion. The function which this characteristic may be made to play in
rendering concrete non-porous seems to be complacently overlooked.
It is possible to bring this action into play with a very great void-
closing effect, and it is possible, moreover, to reduce the voids by
lubrication of the aggregate; and the accompiishment of these two
perfectly comprehensible results is the rational theory of integral
water-proofing.

The elimination of the pores is acknowledged and has been proved
to be the secret of the preservation of concrete subject to salt water
tidal influences, and any elimination based on these two simple prin-
ciples would be, and has been proved to be, just as effective under salt
water as elsewhere. There are two conspicuous instances in the
vicinity of l^ew York City where integral water-proofing has been used
in concrete immersed in salt water, and the concrete has been unaf-
fected. These are Dry Dock No, 4 of the Brooklyn Navy Yard and
the pier on Mr. J. P. Morgan's place at Glen Cove, Long Island.

Mr. Marshall W. Brown," Assoc. M. Am. Soc. C. E. — The discussion

Brown. ^£ ^j^-^ p^pgj. j^^g drifted toward the failure of concrete structures,
due to the action of sea water. The speaker believes that great atten-
tion should be given to the nature of the stone used in the concrete.

If it is assumed that a structure has been adequately designed and
properly built to withstand the forces acting on it as a structure, then
its life depends mainly on its ability to resist the constant and all-
powerful force of sea water acting on it as a destructive agent.

This force is both chemical and mechanical. The chemical reac-
tion, being promoted by the presence of moisture, is facilitated by
the mechanical action of the water, which, carrying away particles
that have been disintegrated, presents new surfaces to be acted on.

The chemical destruction of any particular mass of concrete is
slow. The rate of the mechanical destruction varies greatly, depend-
ing on the intensity of the wave action and the nature of the surface
on which that action takes place.

The rate at which concrete disintegrates increases with the intensity
of the wave action, and with the increasing roughness of the surfaces
exposed.

In any volume of concrete there are three materials which are
acted on, namely, the cement, the sand, and the stone. As the cement
which binds the whole mass together is the one material which is

* New York City.



DISCUSSION ON TESTS OF CONCEETE IN SEA WATER 693

most readily acted on chemically, its rate of destruction must be ^r.
reduced to a minimum. This is done in a number of ways : Brown.

First. — By reducing the quantity of cement needed in a mass,
that is, by proportioning the sizes of the sand and stone in
such a way that the volume of the voids filled by the cement
in the mass is as small as possible.

Second. — By using sand and stone which are in themselves chem-
ically inert and impervious to moisture, and by placing the
mixture of these materials in such a manner that the mass,
as a whole, when set, is impervious to moisture.

This eliminates all chance of disintegration of the interior of the
mass, and confines it to the surface exposed directly to the action of
the water. Without moisture, chemical action cannot take place.

The first disintegration of the surface of concrete takes place very
slowly, and is mainly chemical. The surface is hard and smooth, and
the water has slight effect on it mechanically. Gradually, the cement
is eaten away, and a close examination shows that the area of chemical
action has been greatly increased around the particles of sand. These
are washed away, and, as the action continues, the stone is exposed.
After a time, the whole surface, exposed to the action of the water,
appears to be one mass of stone surrounded by small seams of mortar
composed of sand and cement.

Thus far, only the outer surface of the structure has been dis-
placed, and the integrity of the whole structure thereafter depends
mainly on the character of the stone used. On it comes the chemical
action of the sea water, the intensive mechanical action of the waves
dashing against it, as well as the effect of frost, all tending to loosen
and tear each individual particle of stone from the cement which
binds it to the structure.

For this reason the speaker believes that the character of the stone
has much to do with the life of a concrete structure exposed to the
action of sea water. The stone in the concrete should be chemically
inert, insoluble, non-absorbent, and should present a surface to which
the cement will adhere with the greatest possible strength.

As far as local structures are concerned, this would practically
limit such stone to crushed trap rock, granite, or washed quartz gravel.

The speaker has seen many failures of concrete structures in sea
water, and in each instance he believes it has been due to the use of
shale rock, sandstone, limestone, and other rocks of a sedimentary
nature.

Albert Larsen,* Assoc. M. Am. Soc. C. E. (by letter).— In 1912 Mr.

' 1 • j: j-i, Larsen.

the Grand Trunk Railway System recons tructed a portion o± me

* Providence, R. I.



694



DISCUSSION" ON" TESTS OF CONCRETE IN SEA WATER



Mr. Back Cove Bridge, at Portland, Me., building two rest piers and
Larsen. j-g.^^^ppij^g d^q center pier. The center pier, built in 1892, was of
granite ashlar, and was in such good condition that it is still used
as the center pier, even with the increased loading. During the
progress of the construction of the piers in 1912 Mr. Armour, Masonry
Engineer, Mr. Durham, of the J. S. Metcalf Company, and the writer
investigated the durability of concrete in sea water in order to deter-
mine its suitability for the new rest piers. Various concrete struc-
tures (including the test piles of the Aberthaw Construction Com-
pany), were visited in Boston Harbor, and from these observations
it was finally decided to face the rest piers with granite ashlar, backed
with concrete. The Grand Trunk Eailway System contemplated
future work in Portland Harbor, and therefore actual results of
concrete placed in sea water were desirable. Ten 12 by 16-in. test
piles, 16 ft. long, were made by various water-proofing methods, and
immersed in sea water. Six piles were treated with integral water-
proofing; two were plain, one of whicli was cured in the open for 3
months before being immersed; and two were treated with external
water-proofing. A 1:2:4 mixture was used, with sand and crushed
stone aggregates. The i^ercentage of voids in the sand was 33.

The proportions and different methods of water-proofing for the
piles are stated in Table 11.

TABLE 11.



Pile
No.



Cubic


Cubic


Cubic


feet of


feet of


feet of


cement.


stone.


sand.


5


20


10





20


10


5


20


10


o


20


10


5


20


10


5


20


10


5


20


10


5


20


10


5


20


10


5


20


10



Gallons

of
water.



Plain (cured in open, 3 months)

Ijye and alum (external i

Plain

Water glass (external)

Water-proof cement (" 4 " )

Regular cement plus the water-proofing

compound ( " B " )

Regular cement plus the waterproofing

compound (" C")

Regular cement plus the water-proofln

compound ('• D ")

Regular cement plus the water-proofing

compound (■'£")

Regular cement plus the watei'-prooflng

compound (" F")



25
25
25
25
25

25

29

24

23

25



Water-proof cement was used in Pile No. 5, and the others contained
"Knickerbocker" cement, either alone or' in combination with water-
proofing compound. A ^-yd. Smith mixer was used; each batch was
turned 26 times, and the time was 14 min. Portland City water was
used.



DISCUSSION- ON TESTS OF CONCRETE IN SEA WATER



695



The following results were obtained with the cement :



Mr.
Larsen.



Neat.



318 lb.

405 "



Oue-day test 390 lb.

Seven-day test * 628 "

Twenty-eight-day test 735 "

The chemical analysis of the cement was as follows :

Si 0., 23.46

Fe., 63 9.35

Ca"0 .' 61.24

Mg O 3.59

SO3 1.60

Loss 0.64

The piles were fastened to the pile trestle adjoining the spur-track
leading to the coal pockets, and were made and immersed in the
water on the following dates :



^o.



1.
2.
3.
4.
5.
6.
7.
8.
9.
10.





Moulded.




Placed.




October


10th,


1912.


January 23d,


1913.




3d,


u


October


23d,


1912.




10th,


u




24th,






3d,


!,(




24th,






3d,


!.<




18th,






10th,


a




24th,






4th,


a




18th,






4th,


a




18th,






4th,


"




18th,






4th,


i(




18th,





The following is the formula for the Sylvester wash used in Pile
No. 2: To 2 gal. of water add 1 lb. of concentrated lye and 5 lb. of
alum, and mix until completely dissolved. This is a concentrated
stock. When used, 1 pint of stock solution and 10 lb. of cement are
mixed with enough water to make a mixture that will lather freely
under the brush. Two coats are applied, the first as soon as the
forms are removed (the surface must be wet), and the second as
soon as the first is dry.

The formula for the water-glass or sodium silicate method of water-
proofing (Albert Meyer) is as follows : After the forms are removed,
grind with a carborundum stone any projections due to the concrete
seeping through the joints between the boards. Keep the surface
damp for 2 weeks after placing the concrete. Wash the surface thor-
oughly and allow it to dry. Mix a solution of 1 part water-glass
(sodium silicate) 40° Baume, with from 4 to 6 parts of water, total
5 to 7 parts, according to the density of the concrete surface treated.



696 DISCUSSION ON TESTS OF CONCEETE IN SEA WATER

Mr. The denser the surface the weaker should be the solution. Apply the
^''®^°- water-glass solution with a brush. After 4 hours and within 24 hours,
wash off the surface with clear water. Again allow the surface to dry.
When dry apply another coat of water-glass solution. After 4 hours
and within 24 hours, again wash off the surface with clear water and
allow to dry. Repeat this process for three or four coats, which should
be sufficient to close the pores.

The water-glass which penetrates the pores comes in contact with
the alkalies in the cement and concrete, and forms an insoluble hard
material, causing the surface to become very hard to a depth of
J to i in. according to the density of the concrete. The excess of
sodium silicate which remains on the surface, not having come in
contact with the alkalies, is soluble, and is easily washed off with water.
The reason for washing off the surface between each coat and allowing
the surface to dry is to obtain a more thorough penetration of the
sodium silicate.

The tops of the concrete piers were treated by the water-glass
method, and when examined on March 13th, 1917, were found to be
in as good condition as when treated. The tops were finished with
a 1 : 2 mortar with granite chips.

The piles were examined by the writer on March 13th, 1917, and
photographs then taken are reproduced as Figs. 2 and 3.

The fungous growth was not removed from the piles.

The following indicates the condition of the piles:
Perfect condition.

Nearly perfect condition.

il U C(

Yery marked deterioration at base of pile.

" " " from low to high tide.

ic . (( ic a (c (( c( u a

No. 6 " : Very decided deterioration from low to high tide;

very bad condition.
No. 8 " : Very decided deterioration from low to high tide;

very bad condition.
No. 10 " : Very decided deterioration from low to high tide;

very bad condition.

The results are very interesting, and confirm in a general way the
conclusion of Mr. Bakenhus that a 1:2:4 is better than a leaner
mixture; that external water-proofing does not help when the concrete
has been made as dense as possible; and that integral water-proofing
has a detrimental effect on concrete.

From the writer's observations he would say that, once the concrete
is affected by the action of the sea water, the rapidity of the disin-



No.


1


pi]


No.


3


it


No.


4


iC


No.


2


il


No.


5


ii


No.


7


ii


No.


9


ii




Fig. 2. — Specimen Piles Tested in Sea Water.




Fig. 3. — Specimen Piles Tested in Sea Water.



DISCUSSION ON TESTS OF CONCRETE IN SEA WATER 699

tegration is increased, as the water gets in between the stones and Mr.
the mortar faster and thus breaks up the concrete. The writer thinks L'^''^'^"-
this may be overcome by facing the concrete with a dense mortar
to a depth of from 1^ to 2 in. This may be done during the process
of building, or afterward by the cement gun. It would be interesting
to experiment with concrete made by either of these methods.

W. Watteks Pagon,* M. Am. Soc. C. E. (by letter).— The author Mr.
has summarized the most valuable series of tests that have been made ^^^°°-
in the United States; tests that are on a par with the several series
made in Germany, Norway, and Eussia some years ago ; and the writer
feels that he should be accorded the sincere appreciation of the Pro-
fession.

The destructive action of sea water on concrete has become in
recent years a very serious subject in all those cities on the Atlantic
Coast north of New York City, and in corresponding climatic regions
on other sea coasts. South of New York City, the writer is aware of
only two cases of disintegration, one at Atlantic Cityf and the other
in Chesapeake Bay, but both were special cases.

The writer's experience in a specific case on the Connecticut
shore of Long Island Sound may be of interest, and is quoted briefly
here from a paper written by him, and now on file in the Society's
Library. The concrete which suffered from the sea-water action
consisted of the copings of a series of bridge piers which were faced
with granite masonry below the coping. The facing extended up to
mean high tide, but the copings, which extended from this point
(Elevation 0) to Elevation + 1.5, were of 1:2:4 concrete without
facing.

Pier No. 4, which suffered most severely, was constructed on
December 19th, 1913, during weather sufficiently cold to require the
use of steam to warm the mixing water. After placing, the concrete
was covered with a tarpaulin, stretched from one side of the coffer-
dam to the other, and cold steam was allowed to escape beneath this.
During the week of January 19th to 24th, 1914, there was consider-
able freezing weather, accompanied by high tides. The result was
the formation of a thick ice coating over the concrete surfaces. When
this had melted, on January 26th, it was found that the surface of
Pier No. 4 had been loosened, and large pieces of scale could be
removed. Within a week the mortar between the stones of the gravel
had disintegrated and fallen out, leaving the stones projecting. The
loose portions were removed with a hammer, and behind them the
concrete was perfectly sound. At the end of two weeks, the ice

* Baltimore, Md.

t "Report on the Destructive Action of Sea Water on Concrete", reprinted from
the Monthly Journal of The Engineers' Club of Baltimore.



700 DISCUSSION ON TESTS OF CONCEETE IN SEA WATEE

Mr. coating formed again, and when this had disappeared the erosion
*^°°- was found to have increased greatly.

The concrete used throughout the work was of a wet "sloppy"
consistency. The forms were tight, so that there were no pockets.
Immediately after removing the forms, the surfaces were rubbed with
carborundum blocks. Cow Bay sand was used, and a mechanical
analysis showed it to be too fine; there was also some clay. Analysis
of the cement showed it to be excellent. Later, a coarser sand was
used, with the results described hereinafter.

In the paper quoted, the writer discussed in detail the various
possible causes of this disintegration, and he refers interested readers
thereto. He felt that undoubtedly the fineness of the sand was one of
the factors causing disintegration, and, therefore, drew up a specifica-
tion for the sand which was used in all later work, under which, for
acceptance, not more than Y0% should pass the No. 20 sieve. Long
Island sand testing as low as 40% could be obtained from one dealer,
but it was found that sand which tested below 60% was too coarse
for reinforced concrete work, where a smooth, easy flowing concrete
is necessary. To avoid delay in obtaining sand, 70% was adopted as
a criterion, and immediately a noticeable change in the strength of
the concrete was shown in all the field tests, in the time of setting
of the concrete, and also by observation of the concrete itself.

In the writer's opinion, however, the sand was by no means the
governing factor. From the detailed description which is given in
the paper quoted, it will be seen that the first failure of the surface
skin on the concrete always took place where a pebble of the gravel
aggregate was close to the surface, and where, therefore, the surface
skin had small adhesion to the backing. After this area of skin
had peeled off, it was a simple matter for the water, working its way
along the surface of the pebble, to force off (by freezing) the surface
skin between pebbles. When the skin was gone, the porosity of the
concrete, due to the fineness of the sand, caused the disintegration
of this mortar, as previously described. Photographs of the concrete
just after the beginning of disintegration revealed innumerable little
''pock-marks" where the pebbles showed through the broken skin.

In order to test this theory, three specimens were made for the
writer by the contractors, McHarg-Barton Company. These were
4 by 4-in. by 10 ft., reinforced axially with a |-in. round bar, hooked
at the upper end. The mixtures were as follows:

Specimen No. 1 1:2:0

No. 2 1:2:4

No. 3 l:li:3

These were removed from the forms in 24 hours (in order to simulate
as nearly as possible the temperature and greenness conditions of




Fig. 4. — Pier No. 3, March 13th, 1917.
Early Stages, Showing Pitting.



Fig. 5. — Pier No. 4, February 3d,

1914. Scale Gone ; Concrete

Pitted, Showing Stones

OF Gravel.




Fig. 6. — Pier No. 4, February 5th,
1914. SuEFACE Scale LiOOSEned
AND Falling off. Edges
Rounded off ey Dis-
integration.



Fig. 7. — Pier No. 4, April 4th, 1917.
Damaged Concrete Removed ; Re-
inforcement Placed, Ready
for Applying New Mor-
tar Facing.



DISCUSSION OX TESTS OF CONCRETE IN SEA WATEE 703

the pier coping). No. 2 was cracked in handling in one place; No. Mr.

P&STon
3 in three places. When 3 days old the specimens were hung in the

river so as to extend above high tide and below low tide.

After the first tide, they were all covered with a white coating,
except on the troweled face. At the end of 5 weeks, while they were
being shifted to another place, they were accidentally lost. No.
1, however, was in perfect condition; Nos. 2 and 3 were rather badly
eaten away, especially at the cracks, where, in fact. No. 2 was eaten
away so completely as to expose the rod.

The theory having proved satisfactory, so far as the test went,
it was decided to cut away the face of the coping and replace it with
1 : 2 mortar made of sand bought under the new specification. This
was done on April 4th, 1914, the mortar being mixed to such a
consistency that it could be poured and stirred like thick cream.
The forms were left on for a couple of weeks to protect the concrete
from the sea water as far as possible. When they were removed, the
mortar was in good condition, although there were numerous in-
stances of bubbles having formed while it was being handled and
placed. The surface was not rubbed for fear of causing injury to
the protecting skin.

When examined 2 years later, this mortar facing was in perfect
condition, although it had passed through two winters. This would
seem to confirm the theory, in so far as one test can confirm it.

If other members of the Society have had similar experiences, it
would certainly be valuable to hear from them, and the writer's only
regret has been that in the Aberthaw tests, described by Mr. Bakenhus,
none of the specimens was of mortar only. The writer feels strongly
that the relatively good condition of the 1:1:2 specimens is due not
only to the greater percentage of cement to aggregate, but also to the
greater percentage of cement and sand combined to stone, or, in
other words, that a 1 : 1 : 2 mixture approaches more nearly to a mortar.

R. J. WiG^' and Lewis E. FERGUSON.f Assoc. Members, Am. Soc. Messrs.

. . Wig

C. E. (by letter). — The increasing necessity for marine structures, and

due largely to the improvement of harbor facilities and the military
work along our sea coasts, gives an added value at the present time
to the paper presented by Mr. Bakenhus.

The points he brings out are largely confirmed by an extensive
investigation recently completed by the writers. During this investi-
gation practically all the concrete structures along the Pacific, Atlantic,
and Gulf Coasts, as well as a number in Canada and Cuba, were care-
fully examined. Their present condition was noted, and in most

* W^ashington, D. C.
t Philadelphia, Pa.



704 DISCUSSION" ON TESTS OF CONCEETE IN SEA WATEK

Messrs. cases a detailed description of the materials and methods used in

and construction were obtained.
Ferguson. -pj^g effect of the quantity of water used in mixing the concrete
is rightly emphasized by Mr. Bakenhus, for this factor undoubtedly
has a most important bearing on the ability of the structure to resist
sea-water action. He states that the specimens made of wet or very
wet mixtures are in much better condition than those made of dry
concrete, and this is substantiated by the data he presents.

Perhaps it would have been better if the terms "wet" and "very
wet" had not been chosen for describing the consistencies used. Ordi-
narily, in present practice, concrete which would be designated as



Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 57 of 167)