FACTORS INFLUENCING THE RATIO OF MORTAR TO CONCRETE STRENGTHS G K Moir Blue Circle Industries PLC United Kingdom ABSTRACT. The factors which influence the ratio of EN 196-1 mortar to BS 4550 concrete strengths have been examined using data derived from the testing of random audit samples taken as part of the BSI Kitemark Scheme. Selected cements with a range of chemical and physical characteristics were also tested in EN 196-1 mortar, ASTM C 109 mortar and a range of concretes prepared using BS 4550 aggregates. It is concluded that no single test, whether in mortar or concrete, can be expected to reliably predict concrete strength over a range of cement contents. Nevertheless for a given cement source a significant change in mortar strength is likely to be reflected in a significant change in concrete strength. Keywords: EN 196-1 mortar, ASTM C 109 mortar, BS 4550 concrete, Alkalis, Cement content, Portland cement. Dr G K Moir is Chief Chemist at the Blue Circle Industries PLC Technical Centre at Greenhithe in Kent. His main responsibilities and interests concern cement production quality control and the factors which influence both the short and long term performance of cements in use. He is currently chairman of CEN TC51/WG10 (masonry cement) and of CEN TC51/WG12/TG1 (sulfate resistance test method). Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 50 Moir INTRODUCTION All national cement standards have minimum strength requirements which are determined utilising a standardised test procedure. Cement type may also be classified according to strength and the current British cement standards, in common with the proposed European Standard for Common Cement (prEN 197-1) has several strength classes with well defined upper, as well as lower strength limits. In the world today most national standards utilise one of 4 test procedures whose main characteristics are summarised below: Table 1 Standardised strength tests for cement used internationally DESIGNATION MIX PROPORTIONS CA:FA:C:W* SPECIMEN DIMENSIONS mm CURING TEMP °C BS 4550 concrete* EN 196-1/ISO 679 mortar ASTM C109 mortar BS 4550 mortar 3.5 : 2.5 : 1 : 0.60 0 : 3 : 1 : 0.50 0 : 2.75 : 1 : 0.485 0 : 3 : 1 : 0.40 100x 100x 100 4 0 x 4 0 x 160 5 0 x 5 0 x 50 70.7 x 70.7 x 70.7 20 20 23 20 * CA = Coarse aggregate, FA = Fine Aggregate, C = Cement, W = Water J Now withdrawn but maintained as BS 1881 : Part 131 and as the UK Cement Industry Test Procedure. In the UK the vibrated mortar cube test was introduced in 1940 but was never popular with either cement users or manufacturers. The cement industry developed its own 4 inch cube (~ 100 mm) concrete method and this was standardised in 1978 with the publication of BS 4550 : Part 3.4. This method formed the basis of quality control testing within the industry until the introduction of BS 12 : 1991 which required strength testing using the EN 196-1 mortar prism method. The adoption of the new testing procedure meant that customers were supplied with test certificate data based on mortar rather than concrete, and an ongoing debate has continued regarding the significance of these results to concrete production quality control data. Schramli  has investigated the relationship between standard mortar test results and standard concrete results with the results of ready-mixed concrete strength testing. It was found that the standardised concrete test was no better than the standard mortar test for predicting the strength of production concrete. This was attributed to the relatively low variability of the cement used in the investigation, and the poor reproducibility of the concrete test. Gaynor  has analysed ASTM C 109 mortar and field concrete data and concluded that provided reliance is not placed on the results of single tests, mortar results do give a meaningful guide to concrete strengths. The relationships improve markedly if cement variability is increased, and/or repeat testing and average results are used. De Hayes  investigated the relationship between ASTM C 109 mortar strengths and concrete cylinder strengths. Although a good correlation was found between 1 day mortar and concrete strengths, correlations at later ages were poor. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Mortar and Concrete Strengths 51 Bloomer  has investigated the factors which affect the ratio of EN 196-1 mortar to concrete strengths. It was concluded that cement properties other than strength have little effect on this ratio. This paper compares strength data obtained when cements from different sources were tested in mortar (ASTM C 109 and EN 196-1) and concrete. Factors responsible for differences in the ratio of mortar to concrete strength are examined. SOURCES OF DATA In the preparation of this paper, data have mainly been drawn from 3 sources: • • • Cement plant routine quality control data Audit sample testing at Blue Circle Industries PLC central laboratory An experimental programme utilising 17 cements from different plants in the UK and overseas. REPRODUCIBILITY OF TEST METHOD If a test method is to yield meaningful results, it is essential that the inherent variability of the method is relatively low in relation to cement variability. The reproducibility of the EN 196-1 and ASTM C 109 methods can be assessed from the annual round-robin testing exercises organised respectively by the French Cement industry association (ATILH) and the US Cement and Concrete Research Laboratory (CCRL).Typical mean values in recent years for all participating laboratories are given in Table 2 Table 2 Reproducibility of EN 196-1 and ASTM C 109 METHOD EN 196-1 ASTM C 109 COEFFICIENT OF REPRODUCIBILITY 3% 7% Within the cement industry itself, a higher level of reproducibility and repeatability is normally achieved, and within laboratory coefficients of variation of 2% and 4% should be achievable for the EN and ASTM methods respectively. Thus, the European method has a level of inherent variability, which is approximately half that of the ASTM method. This is borne out by quality control data for US plants, where despite similar levels of control of cement chemistry, fineness and mineralogy, the coefficients of variation of the mortar test results are approximately twice those of UK plants which employ the EN 196-1 mortar test procedure. These differences in apparent strength variability are illustrated in Figure 1. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 52 Moir UK UK UK UK UK UK H UK £ UK <S UK J UK EN 196-1 MORTAR NA NA NA NA NA NA NA ASTM C 109 MORTAR 10 cv% Figure 1 Coefficient of variation of 28 day strength results for 1997 Apr-94 Dec-94 May-95 Nov-95 May-96 Nov-96 May-97 Nov-97 May-98 DATE Figure 2 Works A audit samples tested at Central Laboratory 70 OH PC 42.5 • SR42.5 60 s PC 52.5 50 40 - < ^ E a c h p o i n t r e p r e s e n t s the m e a n o f 2 3 tests on c e m e n t s O sampled over a 4 year period ~7 30 ON ^ 10 2010 , 20 I 30 • - J 40 1 50 1 60 BS 4550 CONCRETE MPa Figure 3 Relationship between EN 196-1 mortar and BS 4550 concrete strengths Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Mortar and Concrete Strengths 53 The BS 4550 concrete testing procedure has been withdrawn, but continues within the UK cement industry as the Cement Industry Test Procedure and in other laboratories as BS 1881 : Part 131. The method uses 10 mm crushed granite aggregate from a single source and selected quartz sands. In a competent laboratory the coefficient of variation of 28 day strength determined using this method is as good as that for the EN 196-1 mortar method, that is - 2 % . C O R R E L A T I O N B E T W E E N E N 196-1 M O R T A R S T R E N G T H S A N D B S 4550 C O N C R E T E S T R E N G T H S Parallel testing of cement samples in mortar and concrete has demonstrated that a significant change in mortar strengths will always be reflected in concrete strengths, and vice versa. This relationship is illustrated in Figure 2 where data from the testing of audit samples from a cement source which shows higher than normal variability are plotted. Since March 1994 the 6 random despatch audit samples taken each year for 16 cements, (13 PC 42.5, 2 sulfate resisting PC 42.5 and 1 PC 52.5), as part of the requirements of the BSI Kitemark Scheme have been tested in concrete, as well as mortar, at the Blue Circle UKAS accredited central laboratory. The sand used during the period March 1994 - December 1997 was Normensand from Germany and the method of compaction used was vibration. Figure 3 illustrates the relationship between mortar and concrete 2 day and 28 day strengths. Each point represents the average result obtained for 23 samples from each cement source/type. The relationship appears satisfactory, but when the data are examined, it can be seen that there are systematic differences in mortar:concrete (m/c) ratio for different products. At 28 days the m/c ratios range from 1.19 (PC 52.5) to 1.35 (relatively coarse PC 42.5). Figure 4 illustrates that there is clearly a relationship between m/c ratio and cement surface area. The surface area of PC 42.5 cement reflects the fineness of grinding required to achieve a mean 28 day EN 196-1 mortar strength of -59 Mpa. The fineness of grinding required is influenced by • • • the cement milling equipment (eg. open or closed circuit) the clinker chemistry (silicates content, alkali level, etc.). clinker mineralogy and tends to be characteristic of a given plant. The reduction in m/c ratio with increasing fineness is related to the influence of cement fineness on the degree of compaction achieved during specimen preparation. Prism densities were determined by weighing the mortar specimens in air and water immediately following demoulding at an age of 24 hours. In Figure 5 it can be seen that prism density decreases with increasing cement surface area. This suggests that mortar compaction and the elimination of entrapped air is inhibited with finer cements, at least with the vibration method of compaction. Figure 6 confirms that m/c ratio increases with increasing mortar density. It can be deduced that mortar compaction is influenced by cement fineness to a greater extent than is the case with concrete compacted using the BS 4550 hand tamping procedure. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 5 4 Moir 2,310 1.4 PC 42.5 SRC 42.5 PC 52.5 PC 42.5 SRC 52.5 PC 52.5 2,300 1.35 B 2,290 P 1-3 % 2,280 in 1.25 Q 2,270 1.2 2,260 —I i I i I i I 1 1 1 1 1 L_ 2,250 320 340 360 380 400 420 440 460 480 _l i I i L_ 1.15 320 340 360 380 400 420 440 460 480 2 2 S S A m /kg S S A m /kg Figure 5 Relationship between prism density and cement SSA Figure 4 Relationship between m/c strength ratio and cement SSA 1.4 m / c ratio = 1.73 - 0 . 0 0 1 S S A - 0.008C3A • 1.35 • • 1.3 • • U < 1.25 - . • • 1.2 • PC 42.5 SRC 42.5 PC 52.5 • % • 2,250 2,260 2,270 2,280 2,290 2,300 2,310 DENSITY kg/m PC 42.5 SRC 42.5 PC 52.5 • @ • 1.15 1.15 i 1.2 I 1.25 . 1.3 I , 1.35 , ! 1.4 3 Figure 6 Relationship between m/c strength density and mortar density PREDICTED Figure 7 Actual v predicted m/c Strength ratio Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Mortar and Concrete Strengths 55 A secondary influence identified using multiple linear regression analysis is cement C3A content. Figure 7 illustrates that m/c ratio can be satisfactorily predicted using the expression m/c = 1.73 -0.001 SSA -0.008 C A. 3 Thus the 28 day m/c ratio is not the same for all cements, but unless significant changes are made to cement fineness or clinker chemistry the ratio remains a characteristic of that particular cement. INFLUENCE O F M O R T A R TEST P R O C E D U R E AND CONCRETE CEMENT CONTENT Whilst it has been shown that m/c ratio is not the same for all cements, the implications for the prediction of concrete strength need to be assessed taking into consideration the other important influences on concrete strength, such as: • • • Cement content w/c ratio Aggregate type In order to investigate the influence of mortar test method and concrete cement content, 19 different cements were obtained from the UK, US, Australia and South Africa. This selection enabled a much wider range of cement characteristics to be investigated, than would have been the case had only UK class 42.5 cements been included in the programme. All of the cements were tested in concretes prepared from BS 4550 aggregates, with nominal cement contents of 200, 260, 310, 360, 430 and 550 kg/m . The concretes were gauged to a constant slump of 60 mm at w/c ratios ranging from 1.10 to 0.36. In addition, 17 of the cements were tested in BS 4550 concrete (w/c 0.60) and using the EN 196-1 and ASTM C 109 mortar methods. 3 Figures 8 and 9 respectively illustrate the relationship between EN 196-1 and ASTM C 109 mortar strengths, and those given by the BS 4550 procedure. Note that these are the results of single pairs of tests and thus the scatter is considerably greater than in Figure 3. Figure 10 illustrates that if the cements are ranked according to BS 4550 28 day strength the overall agreement with both ASTM and EN 196-1 strengths is relatively poor. However, it can also be seen that the agreement of BS 4550 concrete strengths with those of concretes having different cement contents is also poor. Several factors affect the response of different cements to changes in cement content and thus w/c ratio. One of the most important is cement alkali content. Figure 11 illustrates that in the rich 550 kg/m mix the cement alkalis have the expected negative influence on 28 day strength. However, in the lean 200 kg/m mix, where the w/c was —1.1, the cement alkalis had a positive influence on 28 day strength. Thus, no single test method, whether based on mortar or concrete, can be expected to predict performance in a range of production concretes. It is likely that the closer the w/c of the concrete to that used in the standardised test, the better will be the relationship. 3 3 Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 56 Moir 70 E a c h result is a s i n g l e test 60 OH S 50 2 § 4 0 30 28 2 8 ddays a v s 7 days 2 days 20 10 10 20 30 50 40 10 20 BS 4550 < cu EN 196-1 A B C D E F G H I J K L M N O P Q H A B C D E F G H ASTM C 1 0 9 I sex z z 36 38 40 42 44 46 48 MPa 50 200 kg/m3 (60 mm slump) 40 Figure 9 Relationship between ASTM C 109 mortar and BS 4550 concrete Figure 8 Relationship between EN 196-1 mortar and BS 4550 concrete z 30 BS 4550 MPa BS 4550 MPa < A B C D E F G H I J K L M N O P Q 46 48 50 52 54 56 58 60 62 MPa 36 310 kg/m3 (60mm slump) 38 40 MPa 42 44 550 kg/m3 (60mm slump) | A B C D E F E G 1 = n — 1 i J K Z < K H L H M H N ET O EZ p inn Q D M N O PE Q 10 12 MPa 14 34 ^ , 1 1 ^ , • ,.. i ' = ^ , i - ; , i 1 \ i i 1 : 36 38 46 40 MPa 75 MPa 80 Figure 10 Influence of test procedure and cement content on 28 day strength Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. M o r t a r and Concrete Strengths 8 0 0.2 0.4 0.6 0.8 1 0 1.2 EQ Na 0 2 0.2 0.4 0.6 0.8 EQ Na 0 1 2 Figure 11 Influence of cement alkali content on concrete 28 day strength 80 BS 4550 AGGREGATES EN 196-1 MORTAR RESULTS ARE M E A N S OF TESTS USING 17 60 C E M E N T S FROM DIFFERENT S O U R C E S ASTM C109 MORTAR 40 Cu 20 0.2 0.6 BS 4550 aggregates Concretes gauged to constant slump of 60mm. 0.8 1.2 W/C RATIO Figure 12 Relationship between cement content and 28 day strength Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 57 1.2 58 Moir Apart from its lower inherent variability a further advantage of the EN 196-1 procedure, compared to the ASTM C 109 procedure, is that the strengths obtained are similar to those given by a BS 4550 type concrete mix, proportioned to give a w/c of 0.50. The ASTM test gives much lower strengths, despite the lower w/c of 0.485. This difference is illustrated in Figure 12 using average data for the 17 cements. The main reason for the much lower strengths given by the ASTM method is the much higher level of air in the mortar. The ASTM C 109 mortars have typical air contents of 10%, whilst in EN 196 mortar the level should be less than 4%, and in BS 4550 concretes less than 2%. CONCLUSIONS 1. The ratio of EN 196-1 mortar to BS 4550 concrete strength is mainly influenced by cement surface area. A secondary influence is cement C3A content. 2. The ratio of EN 196-1 mortar to BS 4550 concrete at 28 days strengths can be estimated using the expression m/c - 1.73 -0.001 SSA-0.008 C A 3 3. The decrease in m/c ratio with increasing cement surface area is associated with reduced mortar prism density. 4. For any given cement source a significant change in EN 196 mortar strength is likely to be reflected in a significant change in concrete strength. 5. It is not possible to predict concrete strengths over a range of cement contents from any single strength test either based on standard mortar or concrete. 6. The EN 196 mortar strength is more satisfactory for quality control purposes than the ASTM C 109 test procedure, on account of its lower inherent variability, and the much better agreement between strengths obtained and those expected for a concrete of similar w/c ratio. REFERENCES 1. SCHRAMLI, W, and WOLTER, H. Determining the quality of cement by means of standard mortar or standard concrete testing - an investigation. Zement Kalk Gips, Vol 41, No 10, 1988, pp 403-408. 2. GAYNOR, R D. Cement strength and concrete strength - an apparition or a dichotomy. Cement Concrete and Aggregates, Vol 15, No 2, 1993, pp 135-144. 3. De HAYES, S M. C 109 vs concrete strengths: a preliminary investigation. Proceedings of the 12 International conference on cement microscopy, Vancouver, 1990, pp 246-263. th 4. BLOOMER, S J. Effect of cement properties on the strength performance of various cements in commercial concrete. Advanced Concrete Technology Project 95/4. Institute of Concrete Technology, 1995. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved.