Popular Science Monthly/Volume 38/March 1891/The Relative Value of Cements






CLASS OF 1890.

IN The Popular Science Monthly of June, 1890, page 253, there appeared an article entitled Natural and Artificial Cements, by Prof. La Roy F. Griffin, in which theories were advanced in regard to the setting of cement which are at variance with the chemical reactions that are known to take place. There were also given the results of some cement tests, with deductions from the same, that not only are contrary to the results obtained by other experimenters, but are also contrary to the results obtained from the use of cements in works of construction. That there are so many points in Prof. Griffin's article to which exception must be taken, and the exceedingly false impression his article would leave upon the public as to the relative value of American cements, both natural and artificial, is the excuse of the writers for the following article. The diagrams and tables given have been compiled from results obtained in an extended course of cement tests now in progress in the Engineering Department of the State University of Iowa.

Cements, such as are used for constructive purposes, may be divided into two general classes, natural and artificial. The essential ingredients, carbonate of lime, silica, and alumina, are the same in both classes, the principal difference being the proportions in which they are present, and their purity.

In the manufacture of natural cement the raw material generally used is some stone in which the carbonate of lime, silica, and alumina are present in more or less correct proportions, while in the manufacture of the artificial cement the raw material used consists of the essential ingredients, each in a comparatively pure state, thoroughly mixed in theoretically the correct proportions. It is due to this fact that artificial or Portland cement is not only much superior to natural cements, but that it is much more uniform in its quality. This feature of uniformity is perhaps the most valuable possessed by Portland cements, and one which can never be attained in the manufacture of natural cements.

The term Portland cement is now generally used to designate artificial cement, from the fact that the first artificial cement made in England, when hardened, resembled the famous Portland building-stone.

Whether the mixture of the necessary ingredients is artificial or not, it is burned almost to the point of vitrification and then ground to an extreme fineness. The fineness to which cement is ground is one of the most important points in its manufacture, for the reason that, if not finely ground, its strength may be reduced fifty or seventy-five per cent. The theory advanced by Prof. Griffin, on page 254, in regard to the setting of cement, namely, the absorption of carbon dioxide, the uniting of this gas with the lime, and the reforming of lime-stone, is simply the old lime-mortar theory, and in no way applies to the setting of cement. In regard to the changes that do take place during the setting of cement, the following quotations from an article upon the subject by Dr. L. W. Andrews and F. W. Spanutius, in The Transit for December, afford the clearest explanations:

"The setting of a cement is, in general, a complex process, partly chemical in its nature, partly mechanical. Broadly stated, the chemical changes which occur may be said rather to afford opportunity for the mechanical changes which result in hardening than themselves to cause the hardening. The chemical changes are, therefore, susceptible of wide variation without materially influencing the result. . . . In some cements, of which plaster of Paris may be taken as the type, water simply combines with some constituent of the cement already present. In others, of which Portland cement is the most important example, certain chemical reactions must first take place. These reactions give rise to substances which, as soon as formed, combine with water and constitute the true cementaceous material. Portland cement contains as chief, sometimes as almost sole constituent, a lime peridote, and in addition a tricalcium aluminate, Ca2Al2O6 soluble in 3,000 parts of water, and a dark-brown fusible substance, Ca3Al2Fe909. In the act of setting, the tricalcium aluminate first dissolves in water and then begins to separate again as a mass of felted needles consisting of calcium aluminum hydrate, which extend in every direction and are directly the cause of the first setting of the cement. At the same time an action begins which requires a much longer time for its completion, and which probably consists in a combination of the first formed aluminum hydrate with the calcium peridote and the water, forming a mineral belonging to the zeolite class and possessing very probably the composition H10CaAl2Si4O17. This zeolite crystallizes out as it forms, and this continues, for long periods subsequent to the first setting of the cement, to add to its solidity and tenacity."

Following the reasoning of Prof. Griffin, we are unable to understand the meaning of "pure lime cement," as the two terms "pure lime" and "cement," when used in an engineering sense, are incompatible. The effect of the presence of magnesia upon the quality of cement is not perfectly understood; but that an increased hardening in cement for a long period of time is due alone to its presence, is not so, as cements that contain no magnesia have been known to improve constantly during a period of two years.

To quote from Prof. Griffin:

"So a Portland cement will develop its full strength in a few months, while our natural cements will not for years, and, so long as it (this chemical action) continues, the structure improves."

Unless Prof. Griffin classes Portland cement as a "pure lime cement" (which it is not), he has advanced no proof of the above quotation; and furthermore this statement itself is incorrect. It is a fact well known to all engineers and builders that as a class Portland cements are slower setting than the natural cements; and also that natural cements attain their full strength within a comparatively short time (within the first year as a maximum limit), and that, after the full strength has been attained, this strength may decrease, as time goes on, in some natural cements. There has, however, been found no limit of time beyond which Portland cement deteriorated, and for two or three years at least it improves its strength.

In speaking of cement testing, Prof. Griffin says:

"No one of these" (meaning tests for compression, tension, torsion, and cross-strain) "can be dispensed with, since material that will endure one satisfactorily will utterly fail in another; . . . but for general purposes the test of cement which is the most valuable is that which determines its tensile strength."

There are very few cases in practice where any tests other than for tension are made. The statement that "no one of these can be dispensed with," etc., is contradicted by what follows, that the most general test is for tensile strength.

From the very nature of cement, these necessary qualities are so dependent one upon the other that practice and experiment have shown that, where one of these physical tests is passed satisfactorily, the others within certain limits must also be satisfactory. It is due to this fact alone that tests for tensile strength are accepted as standard, as in construction cement should never be subjected to tension or cross-strain, but is usually subjected to compression, or possibly in some cases to torsion; but because the compressive strength in cement is generally proportional to its tensile strength, tension tests have, on account of the facility and accuracy with which they can be made, been adopted as standard.

The form of the test briquette given by Prof. Griffin is not that approved by the American Society of Civil Engineers and PSM V38 D684 Standard one inch thick briquet form.jpgFig. 1. Standard Form of Briquette, One Inch in Thickness. adopted in all standard cement tests. The approved form of briquette is that shown in Fig. 1.

These briquettes are usually made by hand, as described by Prof. Griffin. But unless a great deal of help is available, the process is much too slow for any very extended series of tests; the amount of mortar that can be mixed at once is small; and where different persons are employed it is impossible to obtain briquettes that give satisfactory comparative results, owing to the difference in the personal equations of the makers. This was soon found to be the case in the "State University of Iowa" cement tests, and a specially designed machine was built, having a capacity of making over three thousand briquettes per day, being run by two men. This made possible the making of a much greater number of briquettes under practically the same conditions. Owing to the greater amount of pressure machine-made briquettes are subjected to (about one hundred and fifty pounds per square inch), they are probably stronger than the hand-made; but, as this pressure is uniform for all the briquettes, which is not the case when they are made by hand, the comparative value of the tests is far superior to anything attainable by hand-made briquettes. The following table shows the difference in tensile strength between hand- and machine-made briquettes. Each result is the mean of ten briquettes broken at the end of six months:

NAME OF CEMENT. Hand-made. Machine-made.
Neat Milwakee (American) 333 346
Gibbs's Portland (English) 609 703
Buckeye Portland (American) 669 844

All the briquettes used in the tests from which the table and diagrams here given were taken were allowed to stand twenty-four hours in the air, and were then immersed, the time of immersion being the zero marked upon the diagrams, and all the periods of time being reckoned from this point in weeks, which are noted along the bottom of the diagram. A number of briquettes were broken each day for the first seven days; after this a number was broken every seven days, and the average of these results giving the ordinates to the line on the diagram. Besides these briquettes, ten extra ones were broken at the expiration of one week, one month, three months, and six months. The average of the tensile strength of these, and the time of breaking, are shown on the diagram by black dots, the letter showing the brand of the cement: M, Milwaukee; U, Utica; G, Gibbs English Portland; and B, Buckeye American Portland. This system of breaking briquettes shows the effect of time upon their strength. The testing-machine used in these tests was Riehle Brothers' "Standard Cement Tester," in which the strain upon the briquette is gradually increased by means of a screw-and-worm gear. Although the type mentioned by Prof. Griffin possesses accuracy, and is very satisfactory, still the Riehle machine gives equally satisfactory results, and allows of a much greater number of briquettes being broken within a given time.

Any comparison of the relative value of cements based upon their percentage of increase in strength, as made by Prof. Griffin, is of no value. A cement that attains a certain strength in seven days, even if it only increases one per cent during the following ninety days, is superior for constructive purposes to one that increases four hundred per cent during the same time, provided the ultimate strength of the latter is not greater than the former.

The strength of Milwaukee cement, of which Prof. Griffin has much to say, can be seen in the diagram, as compared with the other brands of cements given. The table given by Prof. Griffin, to illustrate the superiority of Milwaukee cement, shows that the strength of the Milwaukee was three hundred and eighty-two pounds per square inch at thirty days, and only three hundred and fifty pounds per square inch at sixty days. This hardly proves either the superiority of Milwaukee, or that natural

PSM V38 D686 Table of relative strengths of english and american cements.jpg

Fig. 2.—Diagram showing the Relative Tensile Strength of English and American Portland and Two Brands of American Natural,

cement increases in strength with age; and, as it is the average of seventy-five specimens, would rather seem to disapprove the points mentioned. It is true that briquettes made of the same cement, at the same time, kept under the same conditions, and broken at the same time, show a great variation in tensile
PSM V38 D687 Table of relative tensile strengths of natural and artifical cements.jpg

Fig. 3.—Diagram showing the Relative Strength of Natural and Artificial Cements.

strength. The statement of Prof. Griffin that Milwaukee cement has been shown to have the greatest crushing strength is rather too sweeping, to say the least, as all first-class Portland cements are superior to Milwaukee in this respect, and there are a number of brands of American natural cements that are in every way its equal. Although English Portland cements are among the best in the market, still they are equaled by both German and French Portland, while there are now manufactured in the United States Portland cements that in tensile strength exceed any imported cements. Briquettes made of American Portland have shown a tensile strength of eleven hundred pounds at the end of twenty-six weeks.

A careful study of the diagrams will give a correct idea of the relative values of typical English and American Portland cements, American natural cements, and Portland and natural cements. It is not intended to show in Fig. 2 that Milwaukee cement is inferior to all American natural cements, but simply that there are American natural cements that under the same treatment will give at least as good results. The numbers along the bottom of the diagrams indicate the age of the briquettes in weeks; the numbers at the side indicate pounds.