than the workability, which latter is usually of prime importance in the field. As designed by the Bureau of Standards the apparatus consists of a metalcovered table with the top so arranged that it can be raised and dropped freely through a distance of 1/2-in by one revolution of a cam (see Fig. 5). The plastic concrete is molded in a truncated cone, the mold withdrawn, and the table raised and dropped 15 times in 10 seconds. The resulting diameter of the mass, divided by its diameter as molded, and multiplied by 100, is considered an index of the

FIG. 7. Effect of Quantity of Mixing Water on the Strength of Concrete under Various Curing Conditions

consistency or flow. A slump of 6 to 7 in might be obtained from a concrete showing a flow of 210 to 230.

10. Consistency. IT HAS BEEN ESTABLISHED BEYOND REASONABLE DOUBT

THAT THE PROPORTIONAL AMOUNT OF MIXING WATER EMPLOYED HAS A VITAL AFFECT UPON THE STRENGTH OF THE RESULTING CONCRETE.

Within the range

of consistencies practicable in building construction, the driest mixture that can be properly transported, deposited, and compacted gives the strongest concrete. As the amount of water is increased there is a surprising decline in the strength of the concrete at all ages, especially during the early stages of hardening. Fig. 6 shows graphically the relation between the amount of mixing water, expressed as a ratio of the volume of water in cubic feet, and also in gallons per sack of cement, to the volume of cement (1 cu ft 94 lb) and the compressive

=

strength at 28 days for a wide variety of mixtures and consistencies. Curve A shows the strength that may be obtained under laboratory conditions, and curve B, giving values about 500 lb per sq in less, indicates the strength that may be expected where operations are not under rigid control. Although these curves only show the variation in the compressive strengths of concretes mixed to different consistencies, it should be noted that corresponding increases occur in flexural strength, bond resistance, abrasive resistance and impermeability as the water-cement ratio is reduced. It is also true that the shrinkage of plain concrete is proportional to this same ratio.

Fig. 7 shows the effect of quantity of mixing water on the strength of concrete subjected to various curing conditions, and tested at an age of 28 days.

FIG. 8. Operation of the Inundation System

It should be noted that increased amounts of water result in greatly reduced strength even when the samples suffered through too little moisture during their curing period. NORMAL CONSISTENCY* (relative consistency = 1.00) is too stiff for practical use in building construction. For average commercial aggregates, and the conditions obtaining on the job, the driest mixtures that can be properly and economically handled have relative consistencies ranging from 1.05 to 1.20, corresponding to slumps varying roughly from 3 to 7 in. The Joint Committee, 1924, recommends that for reinforced concrete the maximum slump for thin, confined, horizontal sections, be limited to 8 in; that thin vertical sections and columns have a slump not exceeding 6 in; the slump for heavy sections is limited to 3 in. Referring to Fig. 6 it is seen that a concrete having a compressive strength at 28 days of 2 000 lb per sq in when mixed with a water-cement ratio of 1.00, would probably develop less than 1 300 lb per sq * "Normal consistency" is defined as that giving a slump of 1⁄2 to 1 inch.

in when the ratio was changed to 1.25, a difference of 700 lb per sq in and a 35% reduction in strength, still well within the working limits of the average job. In fact a "GOOD WET MIX, THAT RUNS IN NICELY AROUND THE STEEL," usually has a water-cement ratio in excess of 1.25, except where workability is obtained by an exceptionally suitable aggregate or an excessive quantity of cement. A 1:24 MIXTURE, CONTAINING FAIRLY WELL-GRADED AGGREGATES, AND HAVING A WATER-CEMENT RATIO OF 1.30, MIGHT WELL HAVE A PROBABLE STRENGTH AT AN AGE OF 28 DAYS UNDER 1 000 LB PER SQ IN OR LESS THAN ONE-HALF THAT WHICH IS ORDINARILY ASSUMED.

Unfortunately, a wet mixture flows more easily into the forms and requires less labor in depositing and compacting than a comparatively dry mixture. Consequently, for the sake of saving an expense which is almost negligible, when

All percentages based on Tube No. 1-Dry Rodded Sand

Courtesy of the Portland Cement Association FIG. 9. Effect of Moisture upon Sands

all elements of the work are considered, the entire structural frame of a building is frequently built of concrete which, even under the most favorable curing conditions, may not ever attain more than about three-quarters of the designed strength. To control the strength of a concrete requires an intelligent specification and vigilant field inspection supplemented by frequent compression tests made upon samples of concrete taken at the job. As the amount of water required under job conditions to obtain the plasticity necessary for handling is always in excess of that required for the hydration of the cement, the quantity should be gauged to produce the DRIEST MIXTURE THAT CAN BE PROPERLY TRANSPORTED, DEPOSITED, AND COMPACTED. The consistency will consequently be different for various classes of work. The slumps given in the specification of the Joint Committee, 1924, offer an approximate idea of the limiting consistencies which should be permitted under AVERAGE CONDITIONS,

and their relation to the probable strength values of the concrete should be noted.

In practice it is not hard to determine, for any particular mixture, the driest consistency that can be efficiently handled. The difficulty arises in HOLDING THIS CONSISTENCY during the progress of the work. Any appreciable variation in the GRADING of the aggregates changes the VOID CONTENT, and the mixture requires more or less water to obtain the same consistency. Small variations in the moisture-content of the sand not only introduce into the mixture a far from negligible amount of water, but, through the bulking effect of the water upon the sand, completely alter the proportions when volumetric measurement is employed. Mechanical control of the consistency is consequently very difficult. It is possible, however, to compensate by inundation for the varying degrees of moisture in the sand, and the increase of volume which results. This method of measuring aggregate, Fig. 8, consists essentially of a tank-like device for measuring the volume of the sand in a submerged condition, permitting a more exact evaluation of the water-content and of the actual volume of aggregate. The BULKING of sand is very appreciable, some increasing as much as 40% in volume upon the addition of 5% of water. Fig. 9 shows the effect of varying degrees of moisture upon a medium-sized sand. Besides the moisture in the fine aggregate, there is often considerable free water carried by the particles composing the coarse aggregate. Both may also have a very appreciable degree of ABSORPTION, varying from 0% to 1.3% for sand, and from 1.0% to 9.0% for coarse aggregate, the percentages being taken by weight. The following tables give approximate quantities for various aggregates.

Approximate Quantities of Free Water Carried by Average Aggregates *

* From the publications of the Portland Cement Association.

If it is desired to hold a CONSTANT WATER-CEMENT RATIO, this can be done by mechanical means, and the problem of STRENGTH becomes one of WORKTo obtain a workable mixture, with a constant water-cement ratio,

ABILITY.

the inundator has a distinct value. If such means are not used, the moisturecontent of the sand can be periodically determined by test, and the appropriate changes made in the volume or weight of the water added to the batch. Variations from the desired consistency should, in all cases, be corrected by changing the quantity of the combined aggregate. For example, if the batches are coming out too dry THE QUANTITY OF AGGREGATE MUST BE REDUCED INSTEAD OF INCREASING THE QUANTITY OF WATER. Unfortunately, when the mixture is too dry, the tendency is to add water instead of reducing the aggregate. The result is weak concrete. For extimate purposes it may be interesting to note that, although the amount of water required for a consistency corresponding to a compressive strength at 28 days of 2 000 lb per sq in is approximately 74 gallons for each bag of cement, the job allowance should be from 60 to 75 gallons per cu yd of concrete.

The following excerpt from the Preliminary Draft of Proposed Standard Building Regulations for the Use of Reinforced Concrete (Proceedings of the American Concrete Institute, 1925, page 427) points the way toward obtaining concrete of PREDETERMINED STRENGTH AND SUITABLE WORKABILITY.

"CONCRETE QUALITY. Provisions for the design of structures embodied in these regulations are based on the presumption of concrete of certain strength. To produce concrete of the required strength, the proportion of the mixing water to the cement shall be accurately controlled. To obtain the strengths indicated in the following table, the ratio of water to cement shall be in the proportions shown. The strengths indicated represent the minimum ultimate strength in compression which may be expected at 28 days when cured and tested as specified in Section C-1.*

*The committee calls attention to the fact that these proportions of water to cement are based on a wide range of tests and experience and therefore they may be expected to apply to a majority of cases. However, there may be localities where the available materials are such that different water-cement ratios are required for the strengths indicated. It would be advisable therefore for each municipality to conduct the necessary tests to determine the suitability of these limits. For the purpose of a building code it is recommended that the strengths specified be not over 80% of the average values shown by the tests.

"Water or moisture contained in the aggregates must be included in determining the ratio of water to cement.

"All structural drawings and plans submitted for approval shall show the