SAMPLINGTaking the clinker sample for microscopical examination has, as yet, no formally accepted procedure and several techniques are currently used, largely dependent on the purpose of the investigation. Because of time constraints during clinker analysis, the clinker sample must necessarily be small and, therefore, the conclusions must be cautiously drawn. A grab sample is preferable to composite samples for most investigations.Hofmänner (1973) recommends the following sampling technique:1. At intervals of five minutes or less take three 2-kg samples, mix, and quarter down to 500 g.2. Crush the 500-g sample to 5-mm particles.3. Quarter until a sufficient amount remains forencapsulation with resin in a 25-mm-diametercup. Two encapsulations are recommended toget a “representative average of the randomsample.”Ono (1981) recommends a grab sample every eighthour shift during clinker production; hourly samples are taken during kiln startup.Hicks and Dorn (1982) recommend the Ono test (except birefringence) once per day and every time a change is made in the burning process and, once per week, a polished-section examination of the 0.84- to 0.59-mm (No. 20- to 30-mesh) granulated clinker.For determination of the phase content of clinker, Chromy (1983) utilized polished sections made from the quartered residue from 0.5 kg of clinker ground to a particle size passing a 1.0-mm sieve. A 20-mm-diameter polished section of particles embedded in epoxy was prepared.One of the most popular methods involves crushing a random clinker sample of roughly a liter volume (1 to 2 kg) to approximately 2- to 4-mm-diameter particles, from which a few particles are randomly selected or riffled (1) for encapsulation in epoxy for reflectedlight examination and (2) for further crushing to a powder for immersion in oil on a microscope slide for examination in transmitted polarized light. Several encapsulations can be made, thereby increasing the probability of studying particles representing most of the original clinker sizes. One should be aware that different size fractions of crushed clinkers may have different phase abundances (some alite-rich, others belite-rich).J.D. Dorn (personal communication, 1985) stated that clinkers less than approximately 25 mm are virtually the same and that larger clinkers exhibit effects of different cooling rates. Dorn recommended passing a liter of clinker through a crusher, producing particles of approximately 5-mm diameter, followed by riffling to a volume of 1/4 liter and pulverizing to less than 0.59 mm. The 0.59- to 0.30-mm (No. 30- to 50-mesh) fraction is used for a polished section.Centurione (1993) recommends an initial 15-kg clinker sample, which is then quartered to 2.5 kg and sieved. The sieved fractions are crushed, sieved into 2.4-, 0.6-, and 0.3-mm fractions, and blended. A 50- gram sample is taken for microscopy, XRF, and chemical determination of free lime.One problem with the crushing of clinker prior to examination is that microcracks seen in polished-section or thin-section study are ambiguously interpreted. Microcracks that are not artifacts of sample preparation may, in some investigations, be related to strain caused by thermal stress (Hornain and Regourd, 1980), crystal reorganization, hydration, and expansion.If the clinker is extremely sandy or dusty, crushing prior to sieving may not be necessary. A random spoonful taken from a well-mixed sample will likely be adequate.Other workers prefer to sieve the clinker sample, after which representative portions of arbitrarilydefined coarse, medium, and fine fractions are selected for analysis. Whole or crushed clinkers are encapsulated in epoxy and polished sections are prepared. Powders for study in transmitted light can be made from representative portions of the same sieve fractions.Long (1982a) stated that the sampling technique must be dependent on the kind of problem under investigation. A constant cement-quality problem might be studied with a clinker grab sample. However, for analysis of process variations of several days, for example, a combination of several clinker samples to form a composite might provide an abundance of information, particularly if the clinker shows variability. Hourly samples may also be studied as kiln modifications take place. Long recommends taking a 15-kg sample, crushing it to less than 6 mm, then riffling or quartering and separating the 2- to 4-mm fraction for microscopical study. Dusty or sandy clinker should be sieved into a coarse fraction (greater than 2 mm) and a fine fraction (less than 2 mm). The coarse fraction is then crushed to supply the 2- to 4-mm-size material for microscopical study as a companion to the fine fraction. Whole clinker nodules can also be studied. These should proportionally represent the sizes of the nodules in the grab sample and typically number 10 to 12.The sampling method normally followed by the author is to restrict the microscopical investigation to clinkers from only a broadly defined modal-size class from which a number of clinker nodules (at least 30) are randomly selected and crushed to 1.0 to 2.0 mm, some fragments for encapsulation in epoxy and others further crushed and treated with KOH-sugar solution for powder-mount examination and X-ray diffraction (see Chapters 4 and 5). The broadly defined modal class is presumed to represent that part of the clinker size population that volumetrically supplies most of the cement and, therefore, has the dominant influence on the cement’s hydraulic characteristics. Thus, by neglecting the largest and smallest clinkers, one studies the most common clinker sizes that perhaps more accurately reflect the burning conditions and the nature of the raw mix. Sampling just downstream from the cooler is also recommended because the clinkers represent a relatively narrow range of kiln conditions, simplifying the interpretation.The sampling of cements appears to present no major problems. Care should be taken, however, to avoid bias from samples unduly rich in coarse or fine particles, or samples representing areas that might be affected by moisture condensation—unless incipient hydration is the object of the investigation.In conclusion, sample volumes and sampling techniques appear to be largely the arbitrary choice of the microscopist, with objectivity and relevance to the aim of the investigation as the primary considerations. A standard practice for sampling and sample preparation is needed for routine microscopy. For certain studies, clinker nodules can be halved, one half for microscopy, the other half for chemistry and X-ray diffraction (XRD). Only one kiln should be represented in a single clinker or cement sample. A composite clinker sample can be somewhat confusing due to the possible variety of burning conditions represented. Systematic microscopical analyses of the clinker with its corresponding raw mix and cement are highly recommended. It is not uncommon for the writer to place a portion of the greater than 45-µm cement and raw mix in the same cup with the clinkers for epoxy impregnation and polished thin-section examination.
Sample Storage
Preventing atmospheric hydration and carbonation of cement and clinker is a difficult but, for most microscopical studies, not an insurmountable problem. Sample contact with water, atmospheric or otherwise, should be minimized. For long-term storage, glass jars or vials with corks or screwtops that have been sealed with molten wax appear to be moderately effective. During routine examinations, the author stores cement or crushed clinker sieve fractions (after wet sieving with an isopropyl alcohol spray) in 15-mL screwtop glass vials. Only the less than 75-µm size (No. 200-mesh sieve) is retained. To help prevent hydration, the vials can be stored over DrieriteTM or similar hydrophilic material in a vacuum jar. Various types of plastic bags with sealable tops are available and may suffice for temporary storage. However, pinholes produced by abrasion are not uncommon if the samples have been subjected to jostling or other types of rough handling. Metal cans with tight-fitting lids (the type in which paint is supplied) are also relatively satisfactory for sample storage.Regardless of the type of clinker storage container, if a significant quantity of free lime is present in the clinker, disintegration of the clinker nodules will probably occur as a result of lime hydration (air slaking) forming calcium hydroxide. A dry (humiditycontrolled) storage room or cabinet is recommended.
Storage of Prepared Specimens
Polished sections and thin sections can be protected during storage by mounting the cover glass with a drop of epoxy (without hardener) on the prepared section surface. A small dropper bottle containing
epoxy resin (without hardener) is kept at approximately 40°C on the slide warmer for the purpose of mounting cover glasses. Keeping the resin at this temperature in the bottle seems to minimize the crystallization that may occur at room temperature. The cover glass can be easily removed for several months, but even the epoxy (without hardener) will eventually bond the cover glass to the section. Then the problem becomes one of trying to remove the cover slip with a razor blade or by grinding. The polished surface can be protected with an acrylic spray, which can be removed by gentle rubbing with an acetone- or xylene-soaked rag. An acrylic spray eventually cracks, however, and does not prevent hydration of free lime exposed on the section surface.Immersion of epoxy-encapsulated materials in polished sections in an anhydrous lightweight oil (preferably odorless) in a wide-mouth glass jar with a screwtop lid effectively minimizes, but does not eliminate, hydration. If the specimen is re-examined microscopically, the oil appearing on the polished surface can be removed with a sonic cleaner containing isopropyl alcohol, followed by a forceful isopropyl alcohol spray. Dorn and Adams (1983) used Freon in a sonic cleaner to remove residual oil on polished-section surfaces. In the writer's experience, oil droplets on a polished section can be removed with a brief application of acetone, followed by an alcohol spray wash, and blow drying.
