.
Sept., 19x3
.l,
T H E J O U R N A L OF I N D U S T R I A L A i i D E Y G I S E E R I S G C H E M I S T R Y
It follows t h a t , given sufficient experience and a set of standards, the metallographic method will give information t h a t cannot be obtained conveniently b y
PIC. 23.
any other manner of test. It becomes, therefore, an additional safeguard i n the inspection of important castings. 1423 R Sr.. N. W. WASHINGTON, D. C.
LUBRICATION WITH
OILS, AND GRAPHITE
WITH
COLLOIDAL
By CBIRLBSF. M ~ B H R Y Received July 14. 1913
I n a paper‘ published three years ago, an account was given of some results o n the comparative efficiency in lubrication of oil lubricants, and oils carrying colloidal2 graphite. It appeared in all t h e tests therein described t h a t a lower coefficient of friction was given b y the use of graphite t h a n b y t h e use of oils alone, t h a t oils supported a much greater pressure with t h e aid of graphite, and especially t h a t the graphite film was capable of sustaining the friction of a heavy pressure f o r a long period after the supply of lubricant was s h u t off, Much attcntion has since been given t o various features of lubrication with colloidal graphite, especially in attempts t o ascertain t h e actual economy of its use in replacing oil lubricants. The observations t o be described in this paper present high economic cffcicncy and a remarkable durability of a graphoid surface. (For a definition of this term see page 7 2 2 . ) It has long been felt t h a t lubrication with oils under heavy pressures is an artificial system, for t h e friction is supported by a thin film of oil which must scparate completely and continuously the bearing surfaces. If this film he in. t h e least broken, even in minute places, there will be a catch between t h e metal surfaces with greatly increased friction, as shown by higher temperatures as well as by higher coefficients. This uncertainty in oil lubrication depends on an inherent weakness of thc hydrocarbons which constitute the main body of petroleum lubricants, assuming ‘ t h a t they have been separated from the crude oil without decomposition in the process of refining; they are few in number, and members of a limited series reprcsented mainly b y the general formulas CwHm, C,H,,.,, and C%Hsm-.. The limited number of hydrocarbons in these series is shown by the fact t h a t they may be collectcd within comparatively narrow limits of temperature during distillation, provided, of course, * T ~ i Joualin~. s P, 11s: Journal of the American Sarirlg of Mcchonirol Enginrrrs, January, 1910. At the time of the former publication the term “defloeeulatcd“ wused to deripnate this iorm of which had b-en iound to be altogether different in it. properties from the other well known forms. Further study demolistrated its colloidnl gudity.
7‘7
that decomposition is avoided. The stability of these hydrocarbons diminishes in a somewhat regular manner with the increase in complexity of composition, until a point is reached where the oils cannot be distilled without cracking even in YOCUO. This variation in stability appears in their use as lubricants, especially under irregular conditions of friction and temperature. But so long as the temperature is kept down and t h e bearings have a properly even surface, the hydrocarbons of suitable viscosity serve as durable luhricants. Under uneven conditions of friction they are liablc t o immediate decomposition even t o carbonization. This complete decomposition is frequently observed especially in the extremely variable conditions of automobile lubrication. I n fact the demands of modern locomotion with unprecedcnted high speeds, such as in automobile racing, uneven loads, and t h e variable changes of highway traffic, have reached a. burden of lubrication t h a t no oils, mineral, vegetable, or animal, are capable of supporting. Carbonization in automobile lubrication is an occurrence of common observation, and oils are rated on t h e basis of a socalled carbon test, which shows certain differences in stability depending on a difference in t h e composition of the oils, and also on the method of refining. N o oils can withstand the irregular operations of certain automobile practice without carboniaing t o a greater or less extent. Lubrication with oils is based on the quaiity of oiliness, or grcasiness t h a t is inherent in the hydrocarbons poorer in hydrogen mentioned above. It is not strictly defined by viscosity as ordinarily determined. Whilc the molecules have a certain freedom of motion within the body of the oil attended with a cansequcnt inherent friction, they have also an attraction for external surfaccs on which they may form an attachment, b u t preserving their continuity and freedom of motion even under high pressures and high speeds, thus forming, under constant conditions, a continuous and a durable film. Engler’ in referring t o oil lubricants statcd t h a t “ D a s Schmiermittel par Excellence” is not known, b u t t h a t for every special use a lubricating oil must be selected on the basis of its viscosity; t h a t since for variable combinations of pressure and speed, there is no definite standard, the viscosity must be dctermined for any set of conditions in practical operation. It may be said further t h a t viscosity as ordinarily determined is not always reliable for determining t h e quality of a lubricating oil for any stated condition, for the reason t h a t i t is possible t o prepare a n oil lubricant b y compounding a heavy .distillate with a lighter one, leaving out t h e middle fraction, in such a manner as t o give any viscosity desired as determined b y the viscosimeter. But in the use of such a lubricant there is a tendency of the lighter constituents t o creep and evaporate, leaving t h e heavy constituents between the hearing surfaces. It is evidently possible t o determine the presence of any considerable amount of the lighter constituents by determining t h e flashing point of t h e oil, b u t this test is scarcely sufficient to I
Das Erdod. Leiprig. 191% I). 83.
indicate with precision t h e presence of such pro- lubricant by the grooved bearing. It is evident t h a t portions of t h e lighter constituents as may be used t o observations of this nature are altogether relative, give t h e desired viscosity. A more accurate means especially as i t is practically impossible t h a t indeof control is distillation which, if conducted in vacuo, pendent bearing surfaces are precisely in the same should show t h e smallest proportion of lighter oils. condition. By t h e use of a standard lubricant it is of A further aid from a practical point of view, a n d one course possible t o compare tests made under different t h a t is highly desirable if properly conducted with conditions. I n t h e work herein described a n especially hard reference t o factory use a n d conditions, is t h e trial of a n oil on a suitable bearing provided with t h e means Babbitt was selected, a n d the bearing surfaces were for ascertaining t h e coefficient of viscosity a n d for milled down t o true contact in t h e beginning, a n d by long-continued use were worn t o as extreme a condireading temperatures. O n account of the inherent weakness in oil lubri- tion of fine smoothness as i t is possible t o obtain b y cants referred t o above, t h e need is evident of a solid ordinary mechanical operation. Comparing t h e exlubricant capable of equalizing t h e inequalities of the tremely low coefficients a n d temperatures recently metal surfaces a n d of such adequate lubricating obtained with those formerly presented, the difference quality as t o avoid intermolecular friction. Of all in the condition of the bearings is evident. While known materials t h e substance graphite alone possesses these conditions are necessary in showing small differt h e qualities of a normal lubricant. I n its ordinary ences of friction, evidently no such strict adherence is natural condition it is not possible t o mechanically necessary in factory operation, although the more subdivide i t so completely t h a t it can penetrate t h e closely they are applied in practice t h e less will be fine interstices of metallic surfaces a n d a t t h e same time t h e loss in power. Under a n y reasonable condiform a persistent coherent lubricated surface; but tions of operation t h e use of colloidal graphite as in a form of complete purity, free from t h e mineral lubricant is certain t o reduce t h e friction very maconstituents of natural graphite a n d in a condition of terially a n d t o serve as a n important economic element minute subdivision, such as is formed b y t h e con- in factory maintenance. For t h e purpose of ascertaining with greater preversion of Acheson’s electric furnace graphite into its colloidal condition, there is available a solid lubricant cision t h a n formerly the influence of colloidal graphite t h a t fulfills t h e requirements of economic lubrication. in reducing friction, observations were made under a It is so finely divided t h a t it readily permeates metals variety of conditions, more especially for t h e purpose a n d b y reason of its unctuous quality its own friction of comparing its superior economy over t h a t of oil is reduced t o a practically negligible quantity, thus lubricant. First in t h e series of tests one of t h e best escaping t h e internal friction of oil lubricants t h a t automobile lubricants was selected for a test of its is a n important factor in the losses of power in factory frictional capacity alone, a n d then with different percentages of graphite. T h e oil was allowed t o r u n operation. T h e action of colloidal graphite is two-fold: its from t h e cup a t the rate of eight drops per minute for permanent suspension in oil as oildag or in water as two hours, with a thermometer inserted in a hole in aquadag renders i t capable of convenient application the bearing for the purpose of reading the temperatures. a n d it invariably reduces t h e viscosity of t h e oil as a T h e pressure selected was zoo lbs. per sq. in. or a medium of application as shown b y many tests with total of 1500 lbs. T h e speed was q j o revolutions per a great variety of oils; its greatest value. however, minute (r. p. m.). Fig. I shows t h e coefficients of friction extending depends on its readiness t o form a graphoid condition o n bearing surfaces. It is only necessary t h a t i t be through the period of the test, two hours, a n d also suspended in a suitable medium free from a n y kind of t h a t t h e oil film broke seventeen minutes after t h e electrolyte t o give the finely divided particles freedom supply was shut off. It should also be mentioned in connection with this of motion. When evenly spread in a n oil medium over a bearing surface such, for example, as a Babbitt observation t h a t a supply of eight drops per minute bearing of the proper quality, t h e colloidal particles of t h e lubricant is the minimum amount of this oil immediately enter t h e fine metallic interstices a n d t h a t will support the friction of this pressure under accumulating form a combination somewhat analogous these conditions. This mas determined in another t o a n amalgamated surface, which needs only t o be experiment, wherein the flow of oil was reduced t o six properly renewed b y regular additions of t h e lubricant drops per minute; t h e oil film broke soon after t h e t o present a bearing surface capable of supporting a n y test was started showing t h a t this quantity of oil was reasonable pressure a n d with t h e lowest friction t h a t insufficient, a result precisely similar t o what was observed in t h e work three years ago with t h e same oil i t is possible t o obtain. I n t h e description of t h e Carpenter machine on which a n d with other oils. T h e low coefficient of friction in t h e tests presented in this paper mere made, in t h e this test is worthy of note, a n d also its evenness after former paper referred t o above, t h e necessity of using normal conditions were established, a n d until t h e oil a h a r d Babbitt bearing was mentioned. T h e con- film broke. Fig. I also presents the curve for t h e same oil carrydition of t h e journal a n d of t h e bearing have been more carefully considered in t h e recent work, especially ing 0.35 per cent graphite under t h e same elements with reference t o the hardness of t h e Babbitt, t h e of pressure, speed, a n d supply of oil t o t h e bearing. smoothness of surface, a n d t h e even distribution of t h e T h e low coefficient of friction is apparent which, with
Sept.,
I g I 3
T H E JOCR."\AL
OF I S D I - S T R I A L .4.\-D
ESGI-YEERI-YG C H E - I I I S T R I '
719
Chart f o r Motor Oil w i t h a n d without 0 . 3 5 % Colloidal Graphite
0
20
40
60
20
40
60
eo 46
60
20
40
60
20
40
60
Lo
40
60
T/me i n M i n u t e s . a corresponding lower temperature, indicates clearly t h e influence of t h e graphite in reducing t h e viscosity of t h e oil, b u t the observation of especial interest i n this test is t h e permanence of t h e graphoid surface after t h e supply of lubricant was discontinued. It appears t h a t this surface supported a pressure of 2 0 0 Ibs. per sq. in. with a n extremely low a n d even coefficient of friction t h a t continued unchanged during five hours a n d would probably have continued much longer. For t h e purpose of ascertaining whether larger or
Chart f o r Motor Oil w i t h
pressure of I j o lbs. per sq. in. The results of this test are not essentially different from those of Fig. 2 with 0.3j per cent graphite, although t h e coefficients are somewhat lower with t h e larger percentage. There is therefore little t o choose between these percent ages in establishing t h e initial graphoid surface; b u t as mill be shown later a suitable surface can be permanently maintained when i t is once established b y a much smaller addition of lubricant, whether i t be used as a smaller percentage of graphite, or by a diminished supply of oil carrying t h e normal proportion. 0.25%
Colloidal Graphite
77me in M i n u t e s . smaller percentages of graphite are advantageous, several runs were made with lubricants carrying 0 . j per cent a n d smaller percentages t o 0.1 per cent, b u t neither of these extremes were satisfactory. Fig. 2 presents t h e results as t o coefficient with t h e lubricant carrying 0 . 2 5 per cent of graphite under a
For t h e purpose of ascertaining t h e minimum amount of graphite t h a t will maintain a graphoid condition when once formed on t h e bearings, a series of tests were made gradually reducing t h e supply of lubricant all under the same conditions of pressure a n d speed. Fig. 3 gives t h e curves after t h e flow of
T H E J O U R S A L OF IiVDL-STRIAL Ah'D ESGIiZ'EERING CHEAVISTRY
720
Chart for Endurance T e s t s .
e
0
4
3
Time i n
Val. 5, No. 9
Oil a n d 0 . 3 5 7 ~Colloidal Graphite
7 Hours.
oil was reduced from eight t o four drops per minute, t h e oil containing 0.35 per cent graphite. It will be observed t h a t the oil ran for six hours with t h e coefficient of friction practically unchanged after normal conditions were established.' I t seems advisable t o give in detail the results of these tests in order t h a t their connection with t h e final result may appear. Fig. 3 also gives the results of a
8
/e
9
//
/2
/3 # / S
were allowed t o run on the same surface a n d under t h e same conditions as before, except a reduction in t h e supply of lubricant from two drops t o ,one drop per minute. Fig. 3 presents t h e results of this test with no change in the coefficient. Fig. 4 presents another test of the graphoid surface under the same conditions as t o pressure a n d speed, but with t h e flow of oil reduced t o one drop in
Chart for Endurance Tests. ,Oil a n d 0.35% Colloidal G r a p h i t e
0
I
2
3
4
5
6
7
Time
8
9
in
Hours.
further test of the same graphoid surface with t h e flow of oil reduced from four t o two drops per minute; as before, the coefficients remained practically the same during fifteen hours with slight breaks due t o stopping a n d starting. Still continuing the endurance tests, the bearings
/O
//
/Z
I3
/d
/ S I 6
two minutes. During this run of sixteen hours it will be observed t h a t the coefficient of friction remained constant a n d there was no change in temperature. Since t h e coefficient of friction in this test was unchanged a t the end of sixteen hours, even a greater reduction in the flow of lubricant would evidently
Sept., I 9 1 3
T H E J O U R N A L OF I i Y D U S T R I A L A-YD ELYGI+YEERIi\'G CHE-IJISTRY
721
Chart f o r Endurance Test on Colloidal Graphite A ---_hnP
*
Time
in H o u r s .
have maintained t h e graphoid surface, b u t this flow was practically at t h e lowest point where i t could be accurately measured from t h e oil cup. I t appears, therefore, t h a t the same result in lubrica-
i t was shown t h a t a flow of eight drops per minute is t h e minimum supply of oil alone t h a t wiH support t h e friction under these conditions. For t h e purpose of testing still further the qualitv of
tion is obtained by t h e use of one-sixteenth of the quantity of oil t h a t is necessary t o maintain the same lubrication without the use of colloidal graphite; for
the graphoid surface, at t h e end of the last test t h e flow of lubricant was suspended and t h e machine allowed t o run until the bearing caught. T h e
T H E J O U R - T A L OF IA’DUSTRIAL A S D E S G I L V E E R I S G C H E M I S T R Y
722
results of this test are shown in Fig. j where it appears t h a t t h e pressure was supported for nearly ten hours a n d with a- coefficient only slightly higher t h a n in the preceding tests. I t should be borne in mind t h a t a break in the continuity of t h e lubricated surface is indicated suddenly b y a great rise on the friction a r m a n d it is caused b y the first point or section however minute wherever the graphite becomes worn through, yet there may still be a large section of lubricated surface. This appeared in the next experiment. T h e manner in which colloidal graphite is able t o support such heavy pressures with low friction has already been explained. After the metallic surface becomes completely saturated with graphite evidently without renewal continued friction would be necessary t o remove i t completely. It therefore seemed of interest t o ascertain how readily i t could be removed. A series of runs were made on the same surface after it broke in t h e last test, with the addition of oil alone a t the rate of eight drops per minute, t o determine just the point where the graphoid surface could no longer assist in lubrication. I n each of the runs the oil was allowed t o flow for thirty minutes, a n d the bearings were then carefully wiped. Fig. 6 shows the effect of the graphite in assisting the oil lubrication without change during six r u s , and also t h a t i t became exhausted a n d broke in the seventh run showing t h a t altogether approximately three a n d one-half hours were required t o wear off the graphite until i t was no longer a n aid in lubrication. A
GRAPHOID
SURFACE
AKD
THE
CONDITIONS
OF
ITS
F 0R M A T I 0N
I n the former paper on this subject’ the effect of colloidal graphite (then referred to as a “deflocculated f o r m ” ) appears t h e following statement: “One of t h e most characteristic effects is t h a t of a surfaceevener, b y forming a veneer, equalizing t h e metallic depressions a n d projections on the surfaces of journal a n d bearing.” After four years experience with graphite lubrication i t appears t h a t t h e former explanation falls short of defining t h e intimate relation of colloidal graphite t o metallic bearings. N o doubt in building u p the graphoid surface t h e entire depressions a n d projections are saturated with graphite which doubtless enters into a closer state of combination with t h e metallic surface t h a n t h a t merely of a mechanical veneer. I n t h e use of colloidal graphite as a lubricant it appears t h a t b y reason of t h e tenuity or fineness of its particles, i t is capable under t h e conditions of lubrication of penetrating the porous surface of metals, a n d coming into such close contact in their intermolecular structure, of approaching a condition of graphitic combination. It then continues t o accumulate until a continuous saturated surface is formed, which extends equally over t h e depressions a n d protuberances if t h e bearing is not in the best condition of smoothness; b u t in its best condition t h e graphoid surface formed seems t o be nearly frictionless. Since this relation of metal a n d carbon is not de1 LOC.
cit.
Vol. 5 , No. 9
fined by a n y t e r m now in use, the word graphoid used above may serve t o distinguish it from t h e t e r m film t h a t expresses the state of a n oil lubricant on a bearing surf ace. AS t o t h e greater efficiency of the graphoid surface over a n oil film under a n y conditions of lubrication there seems t o be no question. It all depends on establishing t h e conditions whereby this surface may be readily formed. An oil medium in which t h e colloidal graphite is permanently suspended can evidently carry i t t o a n y form of bearing surface t h a t needs lubrication. Whatever t h e condition of t h e metallic bearing t h e graphite soon combines with it, a n d the smoother the surface t h e more readily will a continuous graphoid surface be formed. I n all tests i t has appeared t h a t t h e internal friction of a n oil is diminished by colloidal graphite, although its larger effect is altogether independent of oils except as a medium of application. After a graphoid surface is formed only a small continuous addition is necessary to replace the wear which, as shown in t h e tests, reduces the consumption of oil t o a small fraction of what is necessary in t h e use of oil alone. While eliminating the internal friction of oils a n d rendering viscosity of secondary importance, t h e graphoid surface is capable of taking care of light a n d heavy pressures equally well and with a minimum loss of power. RELATION O F COEFFICIENT O F FRICTION AND VISCOSITY TO TEMPERATURE
Since the curves for temperatures were found t o follow closely those of friction practically unchanged, i t was not thought necessary t o plot them. I n all t h e observations described in this paper, i t was observed t h a t the temperature gradually increases with t h e duration of the test until it reaches a practically constant value not exceeding 6 j oF., a n d t h a t for the most part t h e temperatures were considerably lower. It appears t h a t t h e friction generates a certain amount of heat until i t reaches a normal which is practically constant, a n d t h a t beyond this point t h e increase if a n y is so slight t h a t it is dissipated. I n general t h e lower the coefficient of friction, the lower will be t h e temperature. This is shown in Fig. I , where t h e temperature a n d the coefficient for oil alone were considerably higher t h a n those for oil carrying 0.35 per cent graphite. I t also appears in the other charts especially in the endurance tests, with a small supply of oil. This is doubtless explained, in part a t least, b y the internal viscosity of t h e oil which is of course less with the smaller supply of oil aided by t h e graphite, a n d i t demonstrates a superior quality of t h e graphoid condition over a n oil film, in eliminating practically internal viscosity. AUTOMOBILE LUBRICATION
There is probably no variety of lubrication in which colloidal graphite shows its economic value t o better advantage t h a n in reducing t h e friction on automobile bearings. On t h e Babbitt bearings of t h e cylinder shaft i t readily forms a graphoid surface t h a t wears indefinitely, a n d t h e self-lubricating quality of this surface reduces friction t o the smallest possible value.
Sept., 1913
T H E JOL7R.Y=1L OF I S D C S T R I r 3 L A S D EA\*GI."\'EERI.YG C H E J I I S T R Y
It eliminates also a n y possibility of heating due t o a n irregular flow of oil. I n t h e wide a n d sudden variations of highway automobile traffic t h e bearings are often subjected t o greater strain t h a n a n oil film can stand, b u t not a graphitized surface when once well formed. Such protection from undue wear a n d sudden strains t h a t cannot be avoided in highway locomotion a d d greatly t o t h e safety a n d length of service of the finely adjusted mechanism. Assuming properly selected materials in construction, no doubt t h e most uncertain element in t h e proper operation a n d in t h e economic durability of a n automobile is t h e friction of its moving parts. I t s sure control protects t h e mechanism of t h e moving parts a n d materially reduces the expense of operation.
THE ACTION O F VARIOUS SUBSTANCES ON CONCRETE1 By RICHARD K. MEADE
The following experiments on the action of various substances on concrete were begun some five or six years ago! about the time t h a t the agitation over t h e destruction of concrete b y t h e alkaline waters of the West was first started a n d was undertaken not only t o see i f such acTion was really likely t o take place b u t also to determine which of t h e salts ordinarily found in ground waters m-ere the cause of such destruction. m
TESTSOF
THE
All briquettes were made from a mixture of one part cement a n d three parts standard Ottawa sand. They were allowed t o harden 2 8 days in air a n d then immersed in a solution of t h e salt. T h e briquettes were piled in such manner t h a t t h e solution had access t o almost their entire surface. T h e solutions in all cases except t h a t of the calcium sulfate, which was a saturated solution, were made up of one part of t h e salt t o I O O parts of water, t o form practically a one per cent solution. At first t h e solutions were changed every few days, b u t after t h e first month t h e solutions were changed weekly a n d after t h e first year less often. T h e results obtained are given in the table below: ACTION OF VARIOUS SALTS ON CEMENT MORTARS Age in a i r . , , , . , . , , . 28 days Age in the solution.. 0 days 7 days 28 days 3 mas. 6 mos. 1 yr. 2 yr TENSILE STRENGTH I n MgSO4 . . . . . . . . . . . 219 268 272 28; 196 Disinte... grated Disinte. . . . . . . . . 219 245 300 315 202 115 grated
. .
CASE SCHOOL OF APPLxEn SCIEKCE CLEVELAND, OHIO
ANALYSISA
723
CEMENTEMPLOYED
Silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
............ .........................
Lime . . . . . . . . . . . . . . . . . . . . . . . . . . . ............ Magnesia.. . . . . ............ Sulfur dioxide. Loss on ignition.. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
20.20 2.50 6.96 62.40 3.01 1.60 2.38
PHYSICAL TESTS
Steam O.K. Cold water O.K. Boiling O.K. Air O.K. Fineness-Passing No. 100.. ................. 94.3% Passing No. 200.. . . . . . . . . . . . . . . . . . 97 .8% Setling Time-Initial set-2 hrs. and 15 min. Final set-6 hrs. and 30 min. Tensile Sfrengfh- 1 day neat. ............. 315 lbs. 7 days n e a t . . . . . . . . . . . . . 765 7 days s a n d . . . . . . . . . . . . . 245 '' 28 d a y s n e a t . . . . . . . . . . . . . 876 " 28 days sand.. . . . . . . . . . . . 340 I ' 3 mos. neat.. . . . . . . . . . . . 885 " 3 mas. s a n d . . . . . . . . . . . . . 415 '+ 6 days neat . . . . . . . . . . . . . 855 6 days sand.. . . . . . . . . . . . 435 '' 1 yr. neat ............... 890 '' 1 yr. s a n d . . 510 "
In In In In
Cas04 . . . . . . . . . . . . Na2SO4 . . . . . . . . . . . NaCl . . . . . . . . . . . . . NazCOa . . . . . . . . . . .
219 219 219 219
227 257 236 225
300 334 268 277
334 354 299 324
314 378 287 320
..
............
T h e salts usually found in the so-called "alkali waters" of t h e West are also those which occur in sea water a n d are those present in largest amounts in many spring a n d river waters. They are sodium chloride, magnesium sulfate, calcium sulfate, sodium sulfate a n d sodium carbonate. I n order t o test t h e effect of solutions of these substances o n cement mortars, a sample of normal Lehigh Valley cement was selected a n d f r o m it a large number of s a n d briquettes were made. Read a t the Sixteenth Annual Meeting of the American Society for Testing Materials, Atlantic City, June 26, 1913.
"
141 325 360
First, it should be remembered t h a t the 28-day strength of briquettes kept in air is much less t h a n t h a t of those kept in water. X s will be seen from the results given in t h e table, the sulfates have a marked action on concrete which seems t o be most apparent in t h e case of the magnesium salt. T h e action of magnesium sulfate on cement mortars has been discussed quite voluminously of late, and. I will not go into i t t o a n y length in this paper beyond t h e fact t h a t we carefully analyzed the affected portion a n d the unaffected portion of a sand briquette which has been stored in a solution of magnesium sulfate. These analyses follow:
Soundness-
..
209 271 310 337
After immersion 7 -
Percentages
Alumina.
Before immersion . . . . . . . . . . . . . . 75.12 . . . . . . . . . . . . . . 0.52
.........
.................. . ............
Magnesia.. Sulfur trioxide., Loss on ignition..
0.70 0.33 7.02
Unaffected portion 73.96 0.60 1.30 14.50 1.66 0.83 7.14
1
Affected portion 60.40 0.30 0.64 14.21 3.64 5.78 14.97
T h e large increase in t h e magnesia a n d sulfur trioxide a n d t h e decrease on the oxides of iron a n d alumina indicate t h e elements which react with each other. T h e loss in silica may be due t o chemical action also, b u t as the surface of t h e briquettes was very much attacked a n d t h e sand grains could be scraped away with t h e finger, I a m inclined t o think t h a t t h e lower silica in t h e disintegrated portion is probably due t o mechanical causes rather t h a n chemical action. It will be noted t h a t in almost all cases t h e first effect of t h e solution was t o increase t h e strength of t h e briquettes a n d t h a t signs of disintegration in no cases became evident until after a period of three months in t h e solution. Some of t h e briquettes were even boiled in a 5 per cent solution of magnesium sulfate for several days a n d in all cases t h e briquettes were much stronger