1. INTRODUCTION
Quality control in concrete technology continues to rely mainly on results of tests on hardened concrete, particularly on the assessment of compressive strength. The non-compliance of the properties of hardened concrete with the required specifications gives rise to difficult and costly remedial actions.
Such actions often require a removal of substantial quantities of hardened concrete. The over-reliance on the assessment of hardened concrete can be reduced significantly and the concrete construction process can be made simpler if a fast, uncomplicated, reliable and inexpensive method for direct checking of the composition of a fresh mix were available. The knowledge that the fresh concrete placed consists of correct constituents in the required proportions and that this composition is within acceptable tolerances will improve greatly the quality of concrete.
A modern batching plant can provide a continuous record of what type of material has been discharged into the mixer and its quantity. To some extent the records can be used as evidence of the composition of a fresh mix. An independent, direct check, however, which will verify the composition and assess the within-batch homogeneity of the freshly mixed and transported concrete remains very desirable.
The potential advantages of such a system have been appreciated for a long time and several methods for analysis of fresh mixes have been proposed. These methods were based mainly on the following principles:
(a) chemical / mechanical separation of one of the basic constituents of the mix, particularly cement,
(b) electrical tests for resistance and capacitance,
(c) nuclear methods, x-ray, γ-ray or neutron activation.
Practical methods of analysis currently used are almost entirely based on principles listed under (a) above. The other principles have been used for research and laboratory work rather than for a practical assessment on sites.
Most of the methods proposed were not at all comprehensive. The methods were normally based on a direct determination of the content of only one constituent alone. The proportions of the other constituents had to be obtained from additional tests carried out independently. Despite the very considerable progress which has been achieved recently in the analysis of the fresh mixes, a method which would be truly comprehensive and entirely satisfactory has not yet been developed.
Although a completely satisfactory method has not developed as yet, considerable progress has been achieved. The accuracy of the analytical methods currently available has improved and the analysis is now capable of coping with mixes other than simple cement-water-aggregate ones. The time required for the completion of the analysis has shortened to approximately 10 - 15 minutes for one complete test. However, the apparatus required for the shortest tests which also offer acceptable accuracy has remained expensive and its use has been therefore limited to large projects only.
The origins, principles, apparatus and test procedures for significant methods currently in use are described below.
2. RAPID ANALYSIS (RAM) TEST
 |
| Fig. 1: Rapid Analysis Machine type 300 (Photograph courtesy of the BCA Special Projects group). |
The Rapid Analysis Machine (RAM) has been developed in Great Britain during the late 1970s and early 1980s. It is sometimes known as the C&CA Rapid Analysis Test or as the Constant Volume method.
The method is based on a separation of cement and similar fines by their flocculation from water within a container of constant volume.
Application:
The test method can be used for analysis of fresh mixes with a maximum size of aggregate up to 40 mm, including mixes containing some types of cement substitutes (ground granulated slag) and an air-entrained concrete.
The main application of the RAM test is for an assessment of compliance with specifications of minimum cement contents of concrete mixes, such as are required in BS 8110 : 1985 (ref.l), BS 8007 : 1987 (ref.2), etc.
Description:
The Rapid Analysis machine (RAM) is a floor mounted unit as shown on Fig. 1. It consists of a frame which supports the Elutriation column and the Conditioning vessel. At the top of the Elutriation column is a Sampling head and the bottom of the column is fitted with a Dump valve. At the top of the Conditioning vessel is a Vibrating sieve and a Constant Volume vessel is attached to its lower end.
Basic control of the RAM is by using the panel on its front. The testing cycle itself is carried out automatically.
The most recent version of the RAM (type 300) has a total mass of approx. 145 kg when empty and dimensions of approx. 66 cm x 78 cm x150 cm.
The machine requires a supply of electricity and uses about 80 litres of water per test (ref.l) . The main parts are identified on Fig. 2.
 |
| Fig. 2: Main parts of the Rapid Analysis Machine |
- RAM testing machine,
- Laboratory balances capable of weighing 10 kg, accurate to 1.0 g, and 3000g, accurate to 0.10 g,
- Spare nozzles and a supply of chemicals,
- (Optional), a transformer to convert the available voltage to the 110 V used by the RAM.
Size of the sample:
The standard sample has a mass of 8 kg, equal to approx. 3.5 l of a concrete mix of normal density. It is possible to analyse samples of a smaller size, down to approx. 6 kg.
Operating Instructions:
Calibration:
A calibration chart is prepared from test results obtained from three calibration mixes. The mixes consist of approx. 7 kg of clean aggregate which is combined with 0 g, 750 g and 1500 g of cement and enough water to make the mixes workable. The mixes are tested using the standard procedure and a calibration chart such as is shown on Fig. 3 is prepared. The RAM should be re-calibrated in regular intervals, depending on the amount of usage, and after each replacement of the siphon nozzle.
 |
| Fig. 3: A calibration chart for Rapid Analysis Machine |
1. The sampling head is filled with approx. 8 kg of fresh concrete.
2. The 'start' button is pressed. The signal initiates the pumping of water which washes the sample at an accurately controlled rate. The cement particles are washed up and over the top of the Elutriation column where the 10% of the sample is removed into three sampling channels. The remainder of the sample is dumped to waste.
3. The proportion (10%) of the original sample which has been retained then passes through a 150 μm vibratory sieve and into the Conditioning vessel.
4. Chemical agents are stirred into the particle suspension in the Conditioning vessel. The chemicals cause the cement particles to agglomerate and sink to the bottom of the Constant Volume vessel.
5. After all the cement particles have been collected in the Constant Volume vessel the excess water is removed by coarse and fine syphoning.
6. The Constant Volume vessel is removed from the RAM and weighed on a balance accurate to 0.1g.
7. The cement content of the concrete sample is determined from a calibration diagram and adjusted for silt content.
8. Cement content per 1 m3 of fresh concrete mix is calculated.
Interpretation of the test results:
The RAM test determines primarily the cement content. It is therefore normally supplemented by three other tests to determine:
- the water content,
- aggregate content / grading of the mix
- the amount of ground slag in case it has been used as a cement substitute.
The test for water content is based on drying out of a 2.5 kg sample of the fresh concrete, eg. by using an industrial microwave oven and a suitable container. The process takes approximately 15 minutes (ref.3) and can be carried out simultaneously with the RAM analysis. The precision of the determination of the water content depends on the advance of the hydration of the cement in the mix tested. It has been suggested that the faster drying out in a microwave as compared with an ordinary fan-assisted oven, causes much more of the mixing water to be evaporated before it becomes tied in the hydration products.
The aggregate content is determined from the aggregate retained in the Elutriation column and which is dumped, and from the 150 μm sieve through which approx. 10% of the sample has passed. Aggregate from both the sources is dried out separately and the proportions of coarse and fine aggregate are established by appropriate factoring, sieving through 5 mm sieve. The silt content obtained from separate RAM tests is also noted.
On large projects the delivery of batches of fresh concrete and their placing is not normally held up until the results of the analysis are known. Instead, regular consecutive testing is carried out and the results shown as diagrams.
The trends indicated are constantly monitored and corrective actions taken when required. The test for the slag content is based on the determination of the sulphide content of the cementitious material, including any silt, which accumulates in the Constant Volume vessel at the end of the RAM test. A test set for detection of hydrogen sulphide is required. The material is dried and a small sample (1 g) is mixed with hydrochloric acid. The hydrogen gas produced is drawn by a vacuum pump into a disposable detection tube. The slag content is then determined by comparing the reading on the detection tube with a calibration chart obtained from testing of laboratory samples of the slag-cement mix used (ref.3).
Precision:
The test results are subject to a total error which is made up principally of the sampling error, batching error, machine error and testing error due to the silt content of the concrete. The magnitude of the silt error is approximately equal to the other errors combined. This means that the total error is greatly reduced in cases of concretes with very low silt content or when the silt content is known and an appropriate correction can be made to the result of the analysis. The total standard deviation of the cement content obtained by the RAM varies from approx. 7 to 13 kg/m3 , depending largely on the accuracy of the silt correction (ref.4).
Duplicate samples should be always tested and the results compared. If the difference between the cement contents obtained is not greater than 20 kg/m3 the mean of the two test results is recorded as the final result. If the difference between the two test results is greater, the test is abandoned and new samples have to be obtained and tested.
Delay between the mixing of the concrete and the RAM test has an effect on the cement content determined; the longer the delay, the lower is the cement content measured. A three-hour delay reduced the value of the cement content measured to 92% of the actual content. Such reductions however have been shown to be reasonably predictable (ref.5).
Air entrainment and coarse aggregate with maximum size greater than 25 mm increase the variabiity of the RAM test results. In case of air-entrainment it is recommended to remove the entrained air by mixing the fresh concrete with a suitable de-training agent such as tri-n-butyl phosphate.
An extensive comparative assessment carried out by Dhir et al. (ref.5) indicated that the precision and repeatability of the RAM test results were much better than that of the Buoyancy method which used to be standardised in Great Britain (ref.6).
Advantages:
- Speed of testing. The latest model of the RAM uses a 6-minute automatic operating cycle. The total time from loading of the RAM with a sample of concrete to reading-off of the test result from the graph is not more than 10 minutes. The test is faster than any of the other tests based on separation of concrete constituents. A fully comprehensive analysis of the composition of the fresh mix could be completed within approx. one hour, depending on the number of operators involved.
- Convenience. The RAM test is highly automated. It is therefore relatively easy to carry out, compared to other methods for analysis of fresh concrete.
Disadvantages:
- Cost. The apparatus is expensive. As an alternative to an outright purchase, the RAM can be hired for specific tests or for the duration of a particular project.
- Calibrations. Each testing machine is calibrated when manufactured and the calibration should remain essentially the same in use. However, it is recommended to re-calibrate the machine in regular intervals. The silt correction is based on results of RAM tests carried out on mixes using the same aggregate with a known cement content or by testing the aggregate directly and obtaining an average silt correction amount. The silt correction is essential and it should be checked regularly.
Standardization:
The RAM test method is not yet covered by a standard.
3. OTHER METHODS
3.1 Buoyancy method
The Buoyancy method uses the principle of displacement of water by different basic materials which are separated from the fresh concrete. The test was standardized in Britain until 1983 (ref.6). The test method is based on an assumption that solid particles having size lesser than 0.150 mm are almost entirely those of cement. The method is suitable for all types of concrete. In case of air-entrained mixes, the air has to be removed from the sample by a de-entraining, air-releasing admixture prior to the test.
 |
Fig. 4: Basic apparatus for the buoyancy test
|
The equipment for the test requires a special balance for weighing of materials in water and in air. This can be either a special balance, as shown on Fig. 4 or an electronic balance placed on a suitable frame above a water tank from which a special copper bucket can be suspended. Sieves with 5.0 mm and 0.15 mm apertures and washing out facilities are also required.
The concrete sample is weighed in air and water using the special copper sample bucket. The mix is then washed over 5 mm and 150 μm sieves until each sieve contains clean material.
The coarse aggregate retained on the 5 mm sieve and the fine aggregate retained on the 5 μm sieve are then weighed in water using the same equipment as for the fresh mix. The cement content is then calculated using the following equation:
mass of cement (g) = (cw - ca ) x 3120 / (3120 - 1000)
where:
cw = mass of concrete under water (g)
cw = mass of concrete under water (g)
ca = mass of concrete in air (g)
The precision of the method depends on the content of non-cementitious material consisting of particles with sizes < 150 μm, such as the silt or clay and on a correct assumption for the density of the cement tested.
The test also indicates the content of the fine and coarse aggregate. The water content of the mix should, however, be preferably determined by a separate test.
3.2 Pressure filter method
The pressure filter method is also sometimes known as the Sandberg method. It is based on the separation of solid particles from the fresh mix by pressure filtering.
A concrete sample of a known mass is stirred with added wash water and then wet-sieved through sieves of sizes from the maximum down to 0.150 mm. The wash water and all the fine particles are then passed through a filter paper assisted by compressed air. All the water which was filtered is collected and its mass determined. The filter paper with the fines are weighed and then dried in an oven to a constant weight.
The mass of cement is corrected for the silt content and the mass of the filter paper used. Further correction is required if the cement used showed some water solubility. The cement content of the sample is then converted into the cement content per 1 m3 of the mix.
3.3 Chemical method
Chemical methods are mostly based on the determination of calcium content of the fresh mix. Provided the amount of the calcium which can be brought into the mix as part of the aggregate and the type of cement used are known, the results in the form of the calcium content can be converted into cement content with the aid of a conversion chart established for each particular case.
Typical examples of the chemical methods are the two methods (A & B) described in the ASTM C 1078 - 87 (ref. 7) or the ' GLC' method developed in Great Britain in early 1970's (ref.8).
 |
Fig. 5: Main parts of the apparatus for the rapid filter method
|
In the 'GLC method a sample of concrete is washed through a 300 μm sieve and a sub-sample of the material which has passed the sieve is treated with nitric acid. The concentration of the calcium in the sub-sample is then determined by flame photometry. The results do not have to be adjusted for the content of fines, eg. the silt in the mix such as in the case of the RAM method. The method is, however, considered unsuitable for concretes containing fine calcareous aggregate.
The ASTM methods are also based on the separation of a mixture of fines and cement and on the treatment of a sub-sample by nitric acid. The method A determines the concentration of the calcium ion in the sub-sample by titration using a solution of di-sodium ethylenediamine tetraacetate in the presence of a colour indicator (Eriochrome Black T). The cement content is determined by converting the result of the titration using a calibration diagram. The ASTM method B is based on a fluorometric determination of the calcium ion concentration in a sub-sample also treated with the nitric acid. The titration is based on ethylene glycol-bis-tetra-acetic acid solution and a Calcein indicator is used. The process is carried out in an automatic titration apparatus. The calcium content measured is converted again into the cement content using a calibration diagram.
In all cases supplementary tests are required to determine the content of other concrete constituents, particularly water, in cases were w/c ratios have been specified.
A test procedure used by the 'GLC test, similar to the ASTM C 1079 - 87 (ref.9) standard test for water content is based on the measurement of the dilution of a chloride added to the fresh mix. Sodium chloride is normally used.
The determination of the chloride concentration is either carried out by titration or by using a suitably sensitive proprietary chloride meter.
The chemical methods offer precision better than that achieved by the separation methods such as the RAM test. The assessment of the precision of the chemical test methods has been largely confined to research rather than site laboratories, thus the practical advantage may be less than claimed. The presence of calcium in the aggregate affects strongly the results and if the calcium content is variable, regular re-calibrations have to be carried out.
Disadvantages of all the chemical methods include the use of analytical equipment and operators with skills which are above those normally required for operators carrying out other tests on fresh and hardened concrete in a site-laboratory which provides continuous quality control of manufacture of concrete. Potentially toxic materials have to be used in some of the tests.
The chemical tests are more appropriate to situations in which samples of fresh concrete which is known or strongly suspected of non-conformity have to be analysed with greater precision.
3.4 Physical separation method
The test method known sometimes as the 'Laing' test is based on the principle of separating cement from aggregate by the use of a liquid separating medium.
The 'Laing' method uses bromoform which is a liquid of approx. density of 2900 kg/m3 . This density is between the density of ordinary aggregate (approx. 2600 to 2700 kg/m3 ) and cement (approx. 3150 kg/m3 ).
The fresh mix is washed out through sieves and the suspension of cement and fines passing 212 μm sieve is retained. Bromoform is then added to a sub-sample of this suspension and the mixture is centrifuged to determine the cement content (ref.8).
4. REFERENCES
1. BS 8110: 1985, Structural use of Concrete, British Standards Institution, London, Great Britain.
2. BS 8007: 1987, Code of Practice for Design of Concrete Liquid retaining Structures , British Standards Institution, London.
3. D.M.Weatherill, Developments in the Analysis of Fresh Concrete, BCA Bulletin, British Cement Association, Slough, 1988, 5.
4. M.R.Hollington, Development of Compliance Rules for the Analysis of Fresh Concrete, RILEM Symp. Quality Control of Concrete Structures, Stockholm, June 1979.
4. M.R.Hollington, Development of Compliance Rules for the Analysis of Fresh Concrete, RILEM Symp. Quality Control of Concrete Structures, Stockholm, June 1979.
5. R.K.Dhir, J.G.L.Munday, Nyok Young Ho, Analysis of Fresh Concrete: Determination of Cement Content by the Rapid Analysis Machine, Mag. of Cone. Res.,June 1982 (34) 59 - 73 .
6. BS 1881: Part 2: 1970 (withdrawn) Methods for Testing Fresh Concrete, British Standard Institution, London, Great Britain.
7. ASTM C 1078 - 87, Standard Test Methods for determining the Cement Content of Freshly Mixed Concrete, American Society for Testing and Materials, Philadelphia, U.S.A.
8. P.M.Barber, Analysis of Fresh Concrete, Concrete, June 1983 (17) 12-13.
9. ASTM C 1079 - 87, Standard Test Methods for Determining the water Content of Freshly Mixed Concrete, American Society for Testing and Materials, Philadelphia, U.S.A.
9. ASTM C 1079 - 87, Standard Test Methods for Determining the water Content of Freshly Mixed Concrete, American Society for Testing and Materials, Philadelphia, U.S.A.