American Society of Civil Engineers.

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

. (page 56 of 167)
Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 56 of 167)
Font size
QR-code for this ebook


disintegration when exposed to the weather. There were some delays
in the shipment of the hard limestone regularly used, and it is very
likely that some of the poor local stone may have been used, in order
to avoid waiting.

The contractor on this work had rather a disappointing experience
with the local stone. In excavating for a reservoir, probably 4 000
or 5 000 cu. yd. of stone were piled up to be used later for concrete.
After this material had remained piled up nearly a year, it was found
to have weathered to almost a soft clay. When taken out it appeared
to be a good hard stone.

The other concrete structures at Clarksburg, including a filter plant
and a 1 000 000-gal. concrete tank, were substantially impervious to
water, as judged by leakage measurements, and it is puzzling to under-
stand just why the beams proved to be so porous as to allow the
entrance of water to cause their disintegration.

There is another factor which should be mentioned. The river
water receives a great deal of acid mine waste, and the sulphur com-
pounds may have had some effect on the concrete. This, however,
would not explain a selective action on some parts and not on others.
Probably the best explanation is that poor stone was used in a portion
of the beams, and good stone in the rest. The stone was of a nature
between a shale and a sandstone. The different strata change gradually
from a true shale into a sort of stone locally called sandstone.

• New York City.



DISCUSSION ON TESTS OF CONCKETE IN SEA WATER 683

S. B. Williamson,* M. Am. Soc. C. E. — Before the construction of Mr.
the locks on the Panama Canal, there were several independent investi- son.
g'ations as to the effect of sea water on concrete, and all the information
obtainable was gathered at that time. According to the speaker's
memory, the failures recorded were due largely to porous or pervious
concrete. On the strength of this information, the engineering authori-
ties on the canal agreed to use concrete in the lock construction ; and in
the lower parts of the locks that were subject to sea water a richer
mixture was used.

At Miraflores, where there is a 20-ft. tide, and the lock gates are
about 83 ft. high, an interesting condition developed: It was found
that the difference in the weight of the salt water on the lower side
and the fresh water on the inside of the gates made it possible to have
a head of about 18 in. on the lower side of the gates, tending to force
them open when the lock was full.

In order to discover a means of overcoming this, a number of experi-
ments were carried out, and it was finally ascertained that, by turning
up the lower ends of the culverts so as to discharge at a certain ele-
vation, the inter-mixture or diffusion took place in such a way as to
equalize the pressure heads.

Thus far, the concrete in the locks has shown no signs of disin-
tegration. There has been some disintegration, however, in the concrete
docks built at Balboa. These are reinforced docks, supported by rein-
forced concrete cylinders, and some of the floor girders are partly
submerged at extreme high tide. These show some disintegration,
especially where cracks have appeared, but as yet there is nothing
serious.

Great care was taken in the composition and placing of the concrete
in the locks, particularly where they will be subject to sea water, and
the speaker believes that the concrete is impervious and will not be
readily attacked.

Waldo C. BRiGGS,t M. Am. Soc. C. E. — There is one point which Mr.
the speaker would very much like to have developed. If "the addition
of clay to the cement had a slightly beneficial result" with the concrete
subjected to such unusual conditions, why not modify the customary
specification for sand, requiring it to be "clean and sharp", to read
"clean and sharp, free from dirt, loam, mica, and organic matter, and
shall contain not more than 5% by volume of clay" ?:j:

It happens sometimes that what appears to be a very good grade
of sand is encountered in excavation made in connection with con-
struction work; but if this sand contains clay, it necessarily is
rejected if the specifications require it to be clean.

* New York Cit}'.

t Long Island City, N. Y.

t Bureau of Sewers, New York City.



684 DISCUSSION ox TESTS OF CONCRETE IN SEA WATER

Mr. Charles S. Bilyeu,* Assoc. M. Am. Soc. C. E. — A curve of maxi-

' ^^"' mum density, or an ideal curve for the combination of sand and aggre-

, gate, has been plotted from the results of an extended series of tests by

William B. Fuller and Sanford E. Thompson, Members, Am. Soc. C. E.,

at Jerome Park Reservoir, New York City, in 1902-04.

The speaker's criticism of specifications refers particularly to those
for reinforced concrete buildings. It is true that such specifications are
frequently prepared in architects' offices, and that engineers are not
responsible for them. They generally cover adequately the testing of
the cement, but rarely contain any requirement for the analysis of the
sand and aggregate.

The speaker has been called on, from time to time, to make tests
of materials (particularly sand) used in large and important buildings.
In several instances the sand has been found to contain an abnormally
large percentage of very fine grains, and to be otherwise unsuitable for
use in work of such character. Tests of this nature are usually arranged
for only after the work is well under way and after it has been found
that the concrete has not developed proper strength.

In the speaker's opinion, careful sand and aggregate analyses are
very important investigations, and are absolutely necessary for safe and
economical concrete construction.

Mr. W. E. Day,* Assoc. M. Am. Soc. C. E.— In line with the discussion
■ of sand mixtures, an incident on some construction work during the
past season might be described. A sand pit was opened, about 2J
miles from the work, which contained a sand of excellent quality. The
testing laboratory reported that it was very clean, and that the per-
centages of various sizes were such as to make it an almost perfect
sand for concrete.

On account of the long haul, the experiment was made of mixing
this pit sand with a very fine beach sand, containing considerable
loam, which was right on the works. Test cubes and briquettes were
made with pit sand containing from 0% to 45% of beach sand. These
cubes and briquettes were tested after being stored 30 days, and the
results showed that adding up to 30% of beach sand improved the
strength of the concrete, as compared with concrete made from the
so-called perfect sand.

Mr. Robert Ridgway,* M. Am. Soc. C. E. — It is generally agreed that

the grading of sand is very important. A well-graded sand containing
a small quantity of clay is preferable to a poorly-graded sand which is
free from impurities, particularly if the sand is uniformly fine. It is
well, however, to reduce the quantity of clay and other impurities to a

* New York City.



Ridgway.



DISCUSSION ON TESTS OF CONCRETE IN SEA WATER 685

minimuin, and what one should strive to get is a clean, well-graded Mr.
sand. The speaker does not agree with those who have advocated the ' ^^*y-
use of a small quantity of clay to increase the density of the concrete.
It is better to omit the clay and substitute a little more cement, in
order to secure this density, thereby increasing the strength of the
concrete as well. If the clay appears as lumps in the sand, it is
obviously more objectionable, as Mr. Fuller states, than if in the form
of a powder. The holes often seen in the surfaces of cement walks
and in the faces of concrete walls are generally caused by the dis-
solving of lumps of clay which were incorporated in the mix.

Mention has been made of limestones as concrete aggregates. The
use of limestones was not permitted in the Catskill Aqueduct specifi-
cations for pressure-tunnel lining because of their solubility. This
was the result of a careful study of the subject by the Designing Divi-
sion of the Board of Water Supply before the specifications were pre-
pared. It was found, in a number of instances, that concrete had
disintegrated, due to the solubility of the coarse aggregate. A notable
example cited was that of the Thirlmere Aqueduct for the City of
Manchester, England. In this case, as the speaker understands it, an
almost pure limestone was used, and the soft water in the Aqueduct
dissolved it to such an extent that the concrete was badly disin-
tegrated. In most works, however, limestones are not prohibited; for
example, the Public Service Commission specifications for subway
construction in New York City permit the use of "Sound, hard,
broken limestone."

In driving the new rapid transit tunnel under the East River from
Old Slip, Manhattan, last year, the shield working under compressed
air encoimtered the concrete foundation of the wall at the river bulk-
head line. This concrete had been deposited in bags under the salt
water by the Department of Docks and Ferries of New York City
about 10 years before. The mix was said to be 1 : 2 : 5, and it was found
to be in excellent condition, as somid and dense as though it had been
placed behind forms in the dry. The burlap of the bags was still
there, but the concrete was bonded so well that it looked like good mass
concrete. There was no sign of any disintegration.

George W. Fuller,* M. Am. See. C. E. — This discussion brings out Mr.
four leading points : First, that imperviousness of concrete is a sine ^^'
qua non for structures subjected to a fluctuating water line; secondly,
laitance must be carefully guarded against; thirdly, there is the
question of temperatures, in relation to disintegration by alternate
freezing and thawing; and, fourthly, there is the question of what
may be brought about by organic matters and their decomposition
to form acids under putrefactive processes.

• New York City.



G86 DISCUSSION ON TESTS OF CONCEETE IN SEA WATER

Mr. Sanitary engineers know well, in connection with water-works and

" ^^' sewerage structures, what porous concrete may signify, especially as
to the stability and integrity of structures above and near the flow
line, as ordinarily seen.

Mr. McClintock has given some interesting observations on concrete
sedimentation basins of the water-works plant at Clarksburg, W. Va.
This shows what freezing and thawing may do, in the absence of any
acids or organic matter to decompose the structures, where concrete,
probably not as impervious as it should be, is subject to the various
influences of the elements, especially where there is a fluctuating flow
line.

Now for a few words in respect to the fourth element, and that is
the disintegration of concrete, mortar, or brick, under those circum-
stances where the decomposition of organic matters of sewage or
polluted water may bring about substances which produce acids.

That is probably an old story. It has been debated, year in and
year out, in relation to the question as to whether concrete is as good
as brick for large sewers.

It is a fact that the organic matters contained in sewage under
some circumstances will undergo bacterial decomposition, and that
the sulphur compounds in the organic matters will be changed to
sulphureted hydrogen, in the complete absence of dissolved atmospheric
oxygen, and in the presence of the right kinds of bacteria which ordi-
narily grow there. It is a further fact that, when bacterial agencies
bring about what are commonly called putrefactive changes, sulphureted
hydrogen and other products of putrefaction may be obtained from
mineralized sulphur compounds, such as sulphate of lime or magnesia,
and found in cement, fresh water, and sea water.

Sea water is so highly mineralized that we probably do not realize
what opportunities there are for the formation of compounds of a dis-
solving nature. ISTeither do we know what inter-reactions may produce,
in conjunction with putrefaction, as to solvents from magnesium
chloride and other neutral salts of a supposedly stable and non-corrosive
nature with respect to concrete.

The formation of sulphureted hydrogen does not cause an appreci-
able corrosion of fairly good concrete below the minimum water line
of a sewer, septic tank, or pier of concrete, so far as the speaker's
observations and studies go. Even in septic tanks, he has heard of only
one instance of serious disintegration of concrete in the lower part
of the structure, and that was in the case of a tank in the arid regions
of the Southwest.

Sulphureted hydrogen must be converted into sulphuric acid in
order to develop corrosive properties. This is effected when the gaseous
sulphureted hydrogen rises above the water, meets the oxygen of the



DTSCUSSIOK OX TESTS OF CONCRETE IX SEA WATER 687

air, and in the presence of moisture. This is why sewers, septic tanks, Mr.
and piers are disintegrated above and not below the flow line. Of " ^^'
course, this solvent action varies in intensity with the porosity of the
concrete, effect of freezing, etc.

The outfall sewer of Los Angeles showed disintegration of mortar
and signs of deterioration of brick in manholes beyond the main siphon,
but not below the flow line. When the atmosphere with its oxygen
was excluded, this disintegration was stopped to a surprisingly satis-
factory extent.

Much sulphureted hydrogen does not escape into the atmosphere,
as it combines with iron salts to form the black ferrous sulphide of
iron, which gives liquids in process of putrefaction their black charac-
teristic appearance. This is why the sludge of septic tanks is black,
and why the sediment in pump wells and settling basins is black, as
was reported in detail by the speaker some 20 years ago in his water
purification investigations at Cincinnati.

Laitance should certainly be kept within reasonable limits, other-
wise undesired results are to be expected. Undoubtedly, the pendulum
is swinging back from the very wet mixes of recent years toward the
dry mixes of 20 years ago. However, the mix should not be so dry
as to make it problematical as to removing entrained air from the pores
of the concrete.

It is feasible to use sand with a moderate quantity of silt or clay
without vitiating the integrity of a structure, if that fine material can
actually be used as filling uniformly a certain portion of all voids of
the sand as a substitute for cement; but if such silt or clay is in lumps
or is otherwise irregularly and improperly dispersed in the voids of
the sand, the programme becomes invalid.

Fine material is objectionable if it occurs as a coating on the sand
particles, as it weakens the concrete; or if it contains sufficient organic
matter to interfere with the setting of the concrete; or if it is flaky or
of such character as to produce lumps, on mixing the concrete.

On the other hand, flne material may add materially to the imper-
viousness and density of the concrete. This is a function of the mix of
concrete, the nature and quantity of the fine material; the percentage
of voids in the sand; and the relative size and shape of the sand
particles.

As Mr. Ridgway says, clean sand filled with an excess of cement has
marked advantages in the direction of safety; but this is not always
practicable in some localities, or worth its cost in others — as compared
with using sand with fine material in it.

At York, Pa., the local building sand is a crushed rock containing
10% or more of fine material passing a 100-mesh sieve. It was used
with satisfaction on several hundred thousand dollars worth of work



688 DISCUSSION ON TESTS OF CONCEETE IN SEA WATER

Mr. on outfall sewers, pumping station, and sewage treatment plant. Appar-

^^' ently, this material did not form Iximps. Its use saved the needless

added cost of obtaining other sand transported from a long distance.

However, this is a procedure which ought not to be adopted without

careful investigation in each local case.

Mr. A. H. Ehett,* Assoc. M. Am. Soc. C. E. (by letter). — The subject
■ of the disintegrating action of sea water on concrete has received
much discussion among engineers, and the results of the tests at the
Boston Navy Yard, as described by Mr. Bakenhus, seem to the
writer to throw some light on the subject and point to one primary
and basic cause of disintegration. This is the porosity of the concrete.

In Technologic Paper No. 12, of the Bureau of Standards, on the
"Action of the Salts in Alkali Water and Sea Water on Cements", it is
stated that "a hydraulic cement is readily decomposed if intimately
exposed to the chemical action of various sulphate and chloride solu-
tions", if these solutions are of sufficient strength.

The most conspicuous result of these tests is, that whatever dis-
integration did occur, occurred approximately between high and low
water. It is quite obvious that, if the concrete is porous, the pores
would be filled with water at high tide, that the salts of the water
would be deposited by evaporation during the recession of the tide,
and that this process would be repeated until the solution in the
pores would become of sufficient strength to give rise to this chemical
action. It must be, therefore, that if a concrete is porous, it is sub-
jected to a much stronger chemical action between high and low water
than elsewhere, due to this accumulative concentration; on the other
hand, if it is non-porous, no such concentration can occur. It is true
that evaporation is necessary for the precipitation of the crystals the
increase of which in volume causes the actual disruption or disinte-
gration of the concrete, but the unaffected pieces were subjected to
the same alternate wetting and drying as the affected ones, and this
alone cannot account for the erosion. It seems possible, also, that sea
water in its normal condition is not sufficiently strong in sulphates
and chlorides to produce chemical action.

The writer has re-stated the conclusions drawn by Mr. Bakenhus
on the results of the tests, which conclusions are substantiated by a
detailed examination of those tests, and it will be seen that in each
case these conclusions can be interpreted as confirmatory of the
proposition that the disintegration was more or less proportional to
the porosity of the concrete.

"(a) That the 1:1:2 mixture is superior to the 1:2^:4^, and
that the 1 : 2^ : 4^ is, in turn, superior to the 1:3: 6."

• New York City.



DISCUSSION ON TESTS OF CONCEETE IN SEA WATER 689

It requires no exposition to confirm the statement that the richer Mr.
the mix the less porous the concrete, with similar conditions of mixing ^^^"•
and placing.

"(h) That the wet mixtures are superior to the dry."

The tests made by the Bureau of Standards on the hydration of
Portland cement, as reported in Technologic Paper No. JfS, show that
there is a latent element which in many cements does not begin to
hydrate until 28 days or later, but if then continued in the presence
of water, it hydrates with a large increase in volume and with an
amorphous structure. It is obvious that this action tends to close
the pores of the concrete by filling them with this amorphous material,
and that this inert, but most important, action cannot occur except
in a wet concrete which is not permitted to dry out too quickly. In
the present instance, it was specified that the forms should be "matched"
and "as tight as possible" so that the water in the wet mix was retained.

In Technologic Paper No. 58, of the Bureau of Standards, on the
"Strength and Other Properties of Concrete as Affected by Materials
and Methods of Preparation", it is shown from the tests made that,
with a given mix, the resulting compression strength of the concrete
was about 100% greater with 10.5% of mixing water than with 6.5%;
that it was about 25% greater when cured in a moist atmosphere than
in the open air; and that the compressive strength is proportional to
the density.

"(c) That the effects of magnesia and alumina in varying pro-
portions are not very marked, and follow no apparent law, although
the two most durable specimens are those lowest in alumina content."

This conclusion is confirmed by Technologic Paper No. 12, pre-
viously referred to, which states that "contrary to the opinion of
many, there is no apparent relation between the chemical composition
of a cement and the rapidity with which it reacts with sea water
when brought into intimate contact"; and that "the quantity of
alumina, iron, or silica present in the cement does not affect its
solubility."

"(d) That extra care in mixing produced decidedly beneficial re-
sults."

The experiments of Michaelis, Cabolet, and others have showm
that when cement is continuously mixed with water an increase in
volume as much as seven hundred-fold may occur. Such an increase
in the volume of the cementitious material would tend to fill the
natural voids in the aggregates and make the concrete non-porous.

"(e) That hydrated lime was of no benefit, but rather a detriment."

To refer again to Technologic Paper No. 12, it is shown that lime
is the element of the cement that is attacked by the sulphates and



690 DISCUSSION ON TESTS OF CONCRETE IN SEA WATER

Mr. chlorides of the salt water. This action is shown to bo even more
pronounced with set than with unset cements, the reason for this being
that free lime is always liberated in the process of setting. The
addition of more lime to the cement of course will increase this action.
In this specimen, too, the 10% of lime replaced an equal quantity of
cement, and decreased to this degree any possible automatic pore-
closing effect from the full hydration of the cement as already described.

"(/) That the addition of Sylvester wash was harmful."

The theory of the addition of alum and soap is that they combine
and coat the pores with a water-repellant material. It is difficult to
see, however, how such an action could occur, and yet the particles
of the cement remain uncoated with the same water-repellant material
which would prevent their full hydration and the swelling and void-
filling action resulting therefrom.

"(fir) That the addition of clay to the cement had a slightly ben-
eficial result."

It is well known that clay is a colloidal material with respect to
water, that is, that it remains suspended in water or holds water in
suspension. When mixed with cement, it would seem that this action
could not but retard evaporation, and then tend to promote the further
hydration of the cement. It is also probable that the clay would have
some void-filling effect.

It is affirmed, quite confidently, sometimes, that this disintegration
of concrete between high and low water occurs only in northern lati-
tudes, though the writer has never seen any very convincing evidence
adduced to substantiate this claim, and that the disruption of concrete
is due to the freezing of water in the pores and not to crystallization,
neither action can occur, however, without the open pore to pocket the
water, and to the open pore must be attributed the primary cause of
trouble.

Mr. Larsen's discussion (jiages 693-699) reporting his tests on the
disintegration of concrete by salt water is interesting in its general
confirmation of the results reported by Mr. Bakenhus as to the no-
longer-to-be-questioned fact of such disintegration, but it is impossible
to pass his final conclusion unchallenged.

He states that his results "confirm in a general way the conclusion
of Mr. Bakenhus that * * * integral water-proofing has a detri-
mental effect on concrete." The writer has studied Mr. Bakenhus'
paper very carefully, and fails to find anything either in his
conclusions or in a detailed analysis of the tests to indicate even
in the remotest way that "integral water-proofing has a detrimental
effect on concrete." That paper states very explicitly that integral
water-proofing was used only in two specimens (Nos. 22 and 24) and
that both of these were lean mixes (1:3:6). In No. 22, 10% of



DISCUSSION ON TESTS OF CONCRETE IN SEA WATER 691

liydrated lime Avas "substituted for an equal quantity of cement;" in Mr.
Xo. 24, "to each bag of cement there was added 4| lb. of clay." Clay
is just as properly an integral water-proofing as lime, unless Mr. Larsen
confines his definition of integral water-proofing solely to hydrated
lime, and, if he does, his failures are easily explained, as will be
shown later. Mr. Bakenhus states that No. 24 "came through * * *
better than all the other specimens of a similar mixture, save only
No. 15 (low alumina cement) and No. 20 (extra well-mixed concrete)."



Online LibraryAmerican Society of Civil EngineersTransactions of the American Society of Civil Engineers (Volume 81) → online text (page 56 of 167)