Variations in the strength of Portland cement

Abstract

Summary Data on the variation in compressive strength of ordinary Portland cement have been collected from tests carried out by a number of investigators. The tests were made on mortar cubes made and tested as described in B.S. 12: 1947 or on 4 in. concrete cubes made and tested as described in B.S. 1881:1952. The principal conclusions, based on tests at 3, 7 and 28 days, are as follows. (1) When several vibrated mortar cubes are made from a blended sample representing a single batch of cement, the variation in the compressive strengths obtained will be primarily due to the experimental errors arising from unknown variations occurring in the making and testing of the specimens. The standard deviation of the compressive strengths obtained from a single batch increased from 150 lb/in2 at 3 days to about 280 lb/in2 at 28 days. The coefficient of variation was approximately 4%. (2) The variation between batches of cement from any one works corresponds to a standard deviation generally greater than 400 lb/in2 in the compressive strengths of vibrated mortar cubes. This does not increase in proportion to the increase in average strength with age, so that, for example, the coefficient of variation fell from as much as 25% at 3 days to 8% at 28 days. Tests covering deliveries of cement over periods of four months and longer showed a greater variation than tests covering a shorter period. This appears to be the result of a random variation combined with a long-term variation. Further work would be required to establish the separate causes of the long- and short-term variations. (3) There is a variation between cements produced by different works which, in one series of tests, corresponded to a standard deviation of 620 lb/in2 (13% of the average) in the strengths of vibrated mortar cubes and in another series to a standard deviation of 770 lb/in2 (16% of the average) in the strengths of concrete cubes. If the variation between batches from anyone works could be reduced, less cement would be required to produce a concrete of a specified minimum strength, especially at early ages such as 7 days. The variation in cement, however, provides less than one-half of the variation in works cube strengths and, for this reason, the saving would be less than 6%. In any case, the minimum strength at greater ages might be reduced. The variation between cements from different works will enhance the variation of the works cube strengths on those contracts where cement is drawn from more than one works, and this practice should, therefore, be discouraged. Where cement is used from several works, this source of variation might in an extreme case lead to an increase of up to 12%, in the amount of cement required to meet a minimum strength specification. It must be recognized, however, that although the compressive strength of concrete provides a general indication of its quality, other properties such as durability or impermeability may often be of importance so that compliance with a strength specification may not always produce the most satisfactory material. Further, the choice of works from which cement can be obtained will normally be limited.