Along this same line in comparing laboratory work with what is done in the field, there is no record whatever made of the work required to make and perform laboratory tests such as beams or cylinders, but, when we get into the field, the quantity of work required to handle the concrete, is the controlling factor. This same factor also controls the cost to a great extent. If a contractor gets a mixture that is not workable yet has the required slump, he will make it workable regardless of what the slump may be. You can rest assured that such will be the case and the inspector will certainly have a very difficult time trying to make a contractor use a harsh, unworkable mix that may have the required slump. The contractor must have a workable mixture and we should specify the use of such materials and so design these mixtures that the desired workability will be secured without compelling the contractor to use an excess of water. MR. GEORGE C. WARREN: Mr. President.

PRESIDENT HATTON: Mr. Warren, we shall be pleased to hear from

you.

MR. GEORGE C. WARREN: I am indeed very sorry to have missed most of Mr. Hirst's valuable paper. A few words concerning proportions of Portland cement concrete-this takes me back twenty-five years, when the Company with which I am associated adopted a specification that no engineer of that time would entertain. The theory was, (1) for fine aggregate the coarser the sand the betterpreferably ranging from, say 1/4" to fine; determine the voids in the sand actually used and add enough Portland cement to fill the voids and overplus of ten per cent; (2), for the coarse aggregate, whether crushed stone or gravel, endeavor to secure supply ranging from 1/4" up to a size equal to about half the depth or smallest dimension of the concrete structure; determine the voids in the coarse aggregate and use enough mortar as above to fill the voids and overplus of five per cent.

I believe by getting down to some such rule as that we would make the best as well as the most inexpensive concrete for pavement foundations. With aggregates as most generally used, the proportions would be approximately the conventional 1-3-6, but with well graded aggregates the proportion of cement would be less and provide better concrete. On the other hand, if the fine and coarse aggregates are nearly uniform in size, the proportions of cement would be much more than 1-3-6.

Referring to proportions such as 1-3-6 and 1-2-4, I remember one time I astounded the members of this Society by saying that the most successful piece of Portland cement concrete I had ever made was proportioned about 1-3-22. Most of the coarse aggregate was rock handled by derrick up to as large as two cubic yards, the spaces between the large derrick handled stone not less than one foot and were "poured" with a rich, well-tamped concrete having a maximum coarse aggregate of about three inch size. The construction referred to was in 1901 on a dam at Shawinigan, Quebec, it being one of the

largest and most successful hydro-electric power developments of the world that of the Shawinigan Water and Power Company. That dam, which you will recognize as Cyclopean masonry, has now stood for twenty-five years and perfectly sustains the high pressure of sixty foot deep of water.

Shortly subsequent to the Shawinigan construction, the company, of which I was then General Superintendent, was constructing a Portland cement concrete dam at Trenton Falls, New York, for the Utica Electric Light and Power Company, the specifications for which called for proportions of 1-3-6 for the core and 1-2-4 for the facing. The consulting engineer was one of the very highest experts in hydroelectric power construction. We, however, could not "sell him" to the idea of Cyclopean masonry. The Trenton Falls dam is quite successful, but not as fully successful as the Shawinigan Falls dam, and on account of the much larger proportion of Portland cement the Trenton Falls dam cost much more per cubic yard.

Turning to the subject of proportions of Portland cement concrete in road construction, engineers still pretty generally adhere to the conventional proportions of 1-3-6, 1-2-4, 1-1-3, quite regardless of either the maximum sizes of or voids in the mineral aggregate. One of the most important ingredients in proportion of Portland cement construction, which is far too often entirely overlooked, is a reasonable proportion of common sense. A pretty general tendency of engineers is to say that Portland cement concrete roads should contain certain proportions of cement, sand, and stone. They find that for road surfaces a large proportion of cement is quite necessary to hold the ingredients together under direct abrasion of vehicular traffic. The present tendency is to apply the same rule of proportions to Portland cement concrete foundations for other types of wearing surface, although a half century's experience has proved that smaller proportions of cement-say, normally 1-3-6 or equivalent thereto under void. theory is ample to hold up any load as a foundation for an alignment wearing surface. Such practice of excessive proportions of cement in Portland cement concrete bases is not only a waste of a vast amount of money in cost of foundation, but, generally speaking, is an absolute detriment to the construction. This for the reason that greater rigidity resulting from the increased proportions of Portland cement increases the inherent cracking of rigid Portland cement concrete bases and correspondingly increases the tendency of cracking through the wearing surface, whether that surface be of brick, granite block, sheet asphalt, Warrenite-Bitulithic, or what not. I am sure it is a fair criticism to say that the tendency towards modern increased cracking of pavement surfaces generally is due to modern. and unnecessarily expensive increased rigidity of the base through increased proportion of cement used. I thank you. (Applause.)

A SUMMARY REPORT OF THE THIN BRICK PAVEMENT INVESTIGATION OF THE BUREAU OF PUBLIC ROADS

By J. T. Pauls, Associate Highway Engineer, U. S. Bureau of Public Roads.

One of the important investigations carried on during the last year was the study made of brick pavements. A final report of this investigation has been made and published in Public Roads.'

In view of this complete report so recently published on the brick investigation, this paper will be limited to a consideration of certain of the more important features of this study.

The primary object of this investigation was to determine the relative merit of the thinner brick. The increasing growth of the use of brick under 3 inches in thickness during the last few years, together with much favorable evidence as to their suitability, made it desirable to determine by a thorough investigation, the limitations, if any, which should be placed on the use of the thinner brick.

To obtain this information, two major studies were carried on; first, a field survey was made on existing brick pavements built of brick under 3 inches in thickness, and second, an accelerated traffic test was made on sections of pavement built of brick with thicknesses 2, 212, 3, 32 and 4 inches with supplementary physical tests. on these brick to determine their quality. The conclusions drawn from these studies were as follows:

1. That a 22-inch brick, when properly supported and of the quality used in the Arlington Traffic Tests, will prove satisfactory for surfacing on pavements carrying the heavier types of traffic.

2. That a brick of 2-inch thickness, when properly supported and of the quality used in the Arlington Traffic Tests, would prove satisfactory as a surfacing on streets carrying the lighter types of traffic.

3. That a bedding course of plain sand is more effective in reducing the breakage of brick than a cement sand bedding course, the breakage being much less on the former than the latter. The depth of this sand bedding course should not greatly exceed 3/4-in., as increasing the depth tends to produce roughness in the pave

ment.

4. That cobbling of the brick is greatly increased as the spacing between bricks is increased.

5. That the use of excessive quantities of asphalt filler is a common and serious fault in the construction, unnecessarily increasing

1 Report of Accelerated Traffic Tests and Field Studies by The Bureau of Public Roads, by L. W. Teller and J. T. Pauls, Public Roads, Vol. 7, No. 7, September, 1926.

T

the cost and resulting in a condition which impairs both the appearance and the serviceability of the pavement.

6. That base construction of other than the rigid type may in many cases prove entirely satisfactory. Macadam bases and those constructed of certain types of natural earth appear to be suitable when local conditions are such that these types of construction maintain their stability throughout the year.

7. That no difference in the base construction is necessary for the different thicknesses of brick.

FIG. 1

2="

CEMENT-SAND BEDDING COURSE

3"

Fig. 1-General layout of Test Pavement at Arlington, Va.

Description of the Field Survey.

In carrying on the field survey every effort was made through consultation with local engineers and highway officials, to obtain. accurate information regarding all conditions which might in

fluence the behavior of each pavement and also their views as to the suitability of the pavement to traffic requirements.

Special attention was paid to the effect of the first attempt at using thinner brick on the subsequent policy with regard to brick thickness, using this criterion as a measure of the sufficiency of the design. The first thin brick pavement laid in a community in many cases could be classed as an experiment; but later, similar construction may be taken as an expression of the satisfaction of the community with the type of construction.

This survey involved a detailed inspection of the condition of several million square yards of pavements in which brick of 21⁄2 inch and 24 inch thickness were used. Data on age, type of construction, type of traffic, maintenance and other influencing factors were obtained for each pavement inspected.

Several brick plants which are manufacturing paving brick of less than 3-inch thickness, were visited. These inspections were made primarily to determine the attitude of the industry toward the use of the thinner types of brick and further to find out if their manufacture presented any particular difficulties.

Tests at Arlington, Virginia.

A circular concrete base, which formerly had been used for two accelerated traffic tests on bituminous surfaces, was used for this traffic test on the brick. This pavement was 13 feet wide, and had a mean circumference of 540 feet. See Fig. 1.

In order to minimize the possibility of variation in the quality of the brick used, they were all obtained from one manufacturer. They were vertical fibre, plain wire-cut, lugless type, 81⁄2 inches long, 4 inches wide and 2, 22, 3, 32 and 4 inches thick.

The base was divided symmetrically into ten segments. On one-half, plain sand bedding course 3/4-inch in thickness was used, while on the other the bedding was a similar depth of (1-4) cementsand. On each type of bedding course five test sections were laid, one of each thickness of paving brick. An asphalt filler, having a penetration of 32, by laboratory test, was used on all the sections. Traffic was applied to two 30-inch circumferential strips or zones. During the first phase of the test trucks equipped with plain solid tires were used with different truck loading while during the second phase non-skid chains were placed over the rear tires.

TABLE 1-DETAILS OF THE TRAFFIC TEST LOAD PROGRAM.