Document 3Q5ZEO2QeO0p7jmLmy304dB8D

ASBESTOS A Symposium of Articles by M. S. Badollet, B.S., M.S., Chem. Eng. Papers Presented at Annual General Meetings of THE CANADIAN INSTITUTE OF MINING AND METALLURGY 1948 to 1958 Reprinted by The Quebec Asbestos Mining Association 1958 MTC 000651 CONTENTS Research on Asbestos Fibres (1948) ............................................ 1 Filterability of Asbestos Fibres Used in Wet Processes (1949) 5 Processing Asbestos Fibres: Effect Upon Physical Properties (1950) ....................................................................................... 10 Asbestos, A Mineral of Unparalleled Properties (1951) ............ 15 Asbestos Floats (1952) ......................................... 25 Asbestos Fibers: Production and Usage (1953) ........................... 30 Heat Treatment of Chrysotile Asbestos Fibers (1955) (with W. C. Streib) ........................................... The Role of Asbestos in Plastics (1956) (with M. R. Ximenez .... 38 Identification of Minerals Associated with Asbestos by X-Ray Diffraction Patterns (1958) (with J. P. McGourty) ............ 44 33 MTC 000652 Research on Asbestos Fibres By M. S. BADOLLET* ' (Jubilee Annual Meeting, Vancouver, B.C.) (Transactions, Vol. LI, 1948) Introduction During the processes of liberat portant. As the fibre strength is ing the fibres from the wall-rock it decreased by a processing method, V_ERY little fundamental re- is necessary to use some type of jjearcli" has "been ' ddne~qn~aj- equipment that will crack the rock bestos fibre to take maximum ada nd release the fibre. The maximum so also is the modulus of rupture of an asbestos-cement board or sheet decreased. vantage of its properties an3 to strength of the fibre is preserved Typical tensile strength values "determine in what commerciaFprod- when this type of releasing is gentle of chrysotile asbestos crudes range ucts it can best be utilized. and does not produce fractures at from 40,000 to 100,000 pounds per For the research chemist who is right angles to the fibre bundles. square inch. Sometimes these values associated with asbestos develop It has been possible to submit are well over 100,000 provided the ment work, a thorough study of min- crude bundles to processing equip filaments have not been deformed eralogical information on asbestos ment such as crushers, rolls, fiberiz- by wall-rock movement during the deposits would furnish reliable in ers, etc., and to test the tensile period of crystallization. formation on the rock formations and the mineral impurities closely associated with the asbestos. Furth ermore, he must have an intimate knowledge of the uses of the fibre in various products so that he can visualize the importance of the fibre and the role it plays in industry. Some of the properties of asbes tos fibres such as strength, com strengths of the fibres before and after processing in order to deter mine which type of equipment is best suited for opening fibre bundles without destroying the fibre strength. Even the speeds at which some pieces of equipment are op erated have a definite effect upon the destruction of fibre strength and a change of several hundred Other fibres, such as amosite, will have values ranging from 15.000 to 90.000 pounds per square inch, while crocidolite will show strengths from 100.000 to 300.000 pounds per square inch. Some of these values are far greater than the tensile strength of the usual grades of re inforcing steel. position, length, harshness or soft ness, and mineral associates are of r.p.m. may transform a strong fibre into a weak fibre. Composition of Asbestos Fibres considerable importance to the man ufacturer because they are respons ible for many production difficulties. When studying the fibre from a new deposit, it is important to know the strength of the fibre in its nat Commercial asbestos fibre, when carefully separated into various fractions, is found to be made up ural form, as 'crude', as well as of a number of component parts: Fibre Strength after it has been processed. The crudy fibres, in which the slender basic information gained by this threads or filaments are not sepa It has been known for some time type of test is important when con rated, but remain in bundles; dust, that asbestos filaments have con sidering the installation of a new consisting of extremely fine par siderable tensile strength. These mill to process the fibre. If present ticles, either fibrous or granular; strengths have been referred to milling practice is to he used, it fairly small size particles of ser merely as `weak' to `very strong'. may mean that certain pieces of pentine rock or other minerals close Not until recently have numerical equipment may require by-passing ly associated with the fibre de | values been applied to the strengths in order to preserve the fibre posit; and relatively pure fibres of asbestos fibres. But today we can strength and not destroy the fibre separated into thin threads or fila take a thin fibre filament and length. ments. All of these fractions go to mount it between two holders and The strength of asbestos fibres make up a commercial fibre and apply a gradually increasing load is also important in certain commer each component part plays a defin or pull until the fibre breaks. By cial products, as for example tex ite role when tile fibre mass is used measuring the cross sectional area tiles, papers, and asbestos-cement in industry. For example, the crudy of the filament and the amount uf products. If tlic strength of the bundles improve filtration in wet load applied, we can calculate the fibre is low, the loss in the textile processes and contribute towards tensile strength of the fibre in terms plant will be high due to drops porosity, but they also impart to of pounds per square inch. This from the cards and willows. In asbestos papers a rough texture type of test is slow and rather papers, weak fibres will be ground which may be objectionable. Dust difficult, but the fundamental data down by the beater roll. Lifting of acts as a filler which decreases por obtained furnish important basic the roll may be required, or even osity, decreases filtration rates, and information that can be applied to reducing the beater cycle, in order increases the density. Rock and mechanical processing equipment to obtain a well formed paper where other granular material may drop and to certain asbestos products. the fibres retain their maximum out as a loss during processing or. *Johns-Manville Research Center Manville, New Jersey, U.S.A. strength. In asbestos-cement prod if it remains with the fibre, it will ucts, the fibre strength is im increase filtration rates, increase MTC 000653 porosity, and give a rough surface texture to papers or pressed prod ucts. If the granular material con sists in part of magnetite particles, this is considered objectionable from the standpoint of the elec trical industry. Therefore, each component of an asbestos mass plays a definite part and it is an extreme ly difficult task to separate these materials from the fibre mass com mercially and to eliminate those that may be objectionable. Knowing, then, that commercial grades of fibre contain opened fibres, crudv bundles, fines, and granular materials, it is possible to separate these by a special labora tory technique and express their values in percentages of the orig inal fibre mass. From our experience, it has been noted that certain grades of fibres require certain combinations of the fibre component parts, and that a change of 10 to 15 per cent in these fractions will make a large and noticeable difference in the manu factured product. Likewise, when we know the type of fibre and the percentages of its component parts needed for a product, we can devel op this grade of fibre for this par ticular product. Furthermore, when working with asbestos in plastics, the various component parts of a given grade of fibre are extremely important factors. In this case, we are in volved in such physical measure ments as absorption, wetability. fluidity, impact strength, power fac tor, bulking, and surface area. This means that the research chemist must know the effects of crudv bundles, fines, and granular mate rials upon the properties of absorp tion, wetability, fluidity, etc. After these facts have been established, he must understand the effects of various combinations of the fibre component parts (dust, crudy bun dles, rock, etc.) upon these physical properties. With this fundamental information as a background, it is then necessary to make up molding powders, press them in molds, and study the properties of the molded material to see if the objective has been attained. The results of all these studies place the producer of asbestos in a position to discuss problems under standing^ with the plastics manu facturer and to help him with his problems, particularly in the selec tion of the proper fibre grade. Fibre Lenoth The measurement of the true staple length of the fibres is prob ably the most difficult of the as bestos producer's jobs. Asbestos is unlike cotton, rayon, or other organ ic fibres that can be combed out and the lengths actually measured in inches or fractions of an inch. The staple lengths of fibres in any given mass vary from those ap proximating Yh in. and longer to those of microscopic size. Methods that have been successfully applied to organic fibres have failed to give satisfactory results when applied to asbestos. Wet screening methods have given fairly good classifica tions providing the fibre surfaces are presented horizontally to the screen opening during the test. If the fibres pass lengthwise through the screen opening, the test loses most of its reliability. A test that can measure the true fibre lengths and can be completed within a few minutes would most certainly be welcomed by the indus try. Further effort should be di rected by the research chemist to the solution of this problem. Harshness and Softness of a Fibre The terms `harsh' and `soft' are often used by the miner and manu facturer in connection with asbestos fibre. In many cases, the harshness of a fibre is erroneously used to mean crudiness. This point should be clearly defined as the two terms are entirely different. The term 'crudiness' means the presence of many hundreds of fibre filaments in bundles that are unopened and are wider than a thirty-second of an inch. The bundles also tend to ap pear stiff and will not bend through a 90 arc without fracturing. Of course, these crudy bundles could be soft, silky, or harsh, but, after they are properly opened, they hive lost their crudiness as bundles or aggregates and appear as opened or willowed fibres that can be bent at right angles without breaking. Harsh fibres, on the other hand, are usually weaker than soft fibres and. during a processing method whereby the thin filaments are sep arated and bent at right angles, they have a tendency to fracture easily, with resultant loss of fibre length. Some investigators have attributed the harshness of a fibre to the presence of mineral impurities which have crystallized close to the fibrils (1, 2, 3). Others believe it to be due to the water of crystallization or water of constitution of the as bestos (3). In any case, the degree of harshness of a fibre is an im portant factor in its use in industry. If a harsh fibre is micro-pulver ized and examined under a micro scope, it is noticed that the fibres show an abrupt breakage and the pattern resembles a needle structure instead of the wavy pattern shown by a soft, silky asbestos fibre. The degree of harshness of a given fibre may be estimated by a number of tests such as: (1) open ing of the fibre in a standardized piece of equipment and measuring the length loss and increase of fines produced; (2) the tensile strength; and (3) its filtration properties be fore and after opening. The results of these tests may be expressed as the percent change and the numer ical values used for comparison purposes. It is interesting to note that a harsh fibre always retains a relative ly fast filtration characteristic re gardless of the type of processing. On the other hand, a so-called soft or silky fibre decreases in porosity at a rate proportional to the de gree of opening of the fibre fila ments. It is the fast filtering characteris tic of a harsh fibre that makes it attractive to the manufacturer, pro vided the fibre strength is not too poor. This type of fibre is an ex cellent material for 'freeing-up' a stock, giving porosity and aiding filtration. However, the manufac turer must be careful in handling harsh fibres or he may destroy the maximum fast filtering properties of the fibre by over processing. Soft, silky fibres are normallv difficult to handle by the manufac turer producing wet machine prod ucts. These fibres have a greater surface area than harsh fibres, and are also slower filtering. The de gree of softness of a fibre can also be estimated by the same procedure employed for evaluating a harsh fibre and numerical values can be used for comparisons. Usually, it may be said that the softer and silk ier the fibre, the greater the dif ficulty in handling it during the stages of manufacturing. Therefore, it is necessary for the manufacturer to have some knowledge of the workability of soft fibres before he attempts to use them, and it is usual ly the research man that can deter mine these properties and supply the necessary information. Filtration The filtration characteristic of any fibre mass is an important fac tor to the manufacturer using wet --2-- MTC 000654 methods for the production of as bestos products. For example, the rate of removal of water, either by suction or by pressure, when manu facturing asbestos papers, mill boards, shingles, asbestos-cement sheets, etc., will, to a certain extent, govern the production rates. Therefore, in order to understand the problems facing the manufac turer, it has been necessary to de velop a laboratory filtration test that can evaluate fibres as fibre masses, or as fibres in combinations with other ingredients that are used in commercial products. The first step in this type of in vestigation was to determine the ef fects of all mineral impurities closely associated with commercial asbestos fibres. The next step was to study the filtration properties of asbestos fibres from all parts of the world and also, at the same time, determine the percentage of each mineral impurity. This was followed by studying the effects of electrolytes, wetting agents, soft or harsh fibres, fibre particle size, weathered fibres, blends containing various types of fibre, crudy fibre bundles, and many others upon a standardized fibre in order to show the advan tages or disadvantages as regarding the elimination of water from as bestos slurries. With this type of information available, it has been possible to express the degree of filtration change in percentage and then set up methods of correction so as to obtain the maximum improve ment in filtration. Many production problems involving the elimination of water from fibre slurries have been investigated in the laboratory and the information has been suc cessfully applied to the factory pro cesses. Value of X-Ray Studies Today, X-ray diffraction studies are becoming one of the standard tools of the research chemist. They give the chemist a chance to obtain a preliminary answer to some of his problems within a few hours, whereas a chemical analysis may require weeks of analytical work. Quite often, the chemical analysis of asbestos fibres shows the pres ence of small quantities of mag nesium, silica, or water in excess of the normal amounts found in an average sample of asbestos. When these data are calculated in terms of minerals, they may indicate small amounts of talc, brucite, chlorite, etc., occluded in, or crystallized be tween, the fibrils. In such cases, a careful X-ray investigation shows the typical ring structure of the minerals and the data can then be recalculated showing the percent ages of the mineral impurities pres ent in the fibre. X-ray studies, along with petro graphic investigations, have fur nished valuable clues to chemical reactions that may take place not only in asbestos fibres themselves, but in finished products. Knowledoe of Geology of Deposit as an Aid to the Manufacturer If the manufacturer purchases fibre at random, the material he ob tains may include asbestos from several different deposits, and the mineral formations of each locality will contribute some important fac tor to the fibre that may or may not be objectionable to his process. For example, suppose a certain asbestos-cement manufacturer for a number of years purchased his fibre supply from one mine, where the rocks were very little faulted and hence there was very little fault fibre, and where such minerals as brucite, picrolite, and magnetite were present in only small amount. This manufacturer would have all of his processes well standardized on the fibre from this mine and everything operating in a satisfac tory manner. Now suppose that he decided to change his source of fibre supply and the fibre came from a deposit in another locality where there was considerable fault ing, and considerable amounts of fault fibre, brucite, picrolite, mag netite, and other minerals. What would happen? First, if he did any re-process ing of the fibre in his factory, such as willowing, he would notice a loss in fibre due to formation of dust or fines caused by the wil lowing action on the fault fibre, brucite, and minerals present. Second, the fibre, having a total overall lower strength, would have a tendency to lose length during any processing stage. Third, the surface area of the processed fibre would be increased and it would have a tendency to re quire more binding agent. Fourth, the fibre, when wet, would be slimy and difficult to handle. Fifth, the porosity of the fibre would decrease and the removal of water by normal filtration means, either by suction or by pressure, would be slowed down, thus result ing in a decreased output of the pro ducts. Sixth, the strength of the final products would probably show a decided decrease in modulus of rupture. If these changed conditions were t suddenly thrust upon the manufac turer, without any forewarning, he would be in a difficult position. Consequently, today, the research chemist studies all of these prob lems long before the manufacturer needs to make the fibre change. He first consults the geological litera ture of the locality in which the fibre occurs and becomes familiar with the rock formations. With this information as a background, he ob tains large representative samples of the fibre and studies its physical properties in the laboratory. This is followed by pilot-plant tests, where products are made under con ditions similar to regular manufac turing procedures. When all of this work is completed he is ready to make a recommendation to the manufacturer as to the value of the fibre and as to whether it can be used successfully, and, if so, what changes in processing will be nec essary. It may be urged that these tests are time consuming and may not be worth the expense, but, if a manu facturer loses production in his factory for any length of time, his financial loss is considerably greater than the cost of the surveys and tests made by the research and de velopment chemist. Future of the Asbestos Industry The asbestos industry appears to have reached a stage where consid erable expansion of output of fibre is necessary. In making this state ment it is, of course, assumed that there will be no displacement of asbestos by other materials in the products for which it has been used in the past and also that the many new developments now in progress will materialize commercially in the near future. The demand for asbestos today far exceeds the supply, and it has been necessary for many producers of asbestos products to curtail pro duction because of the shortage of fibre. The situation has been relieved somewhat as a result of many re finements in extraction at the mills which have made it possible to in crease recoveries and throw away less fibre in the tailing product. Developments in the plastics in dustry in Canada and the United States during the past few years open up an estimated potential --8-- MTC 000655 market for short asbestos fibre that may reach 8,000 to 11,000 tons per year. The asbestos-cement products in dustry in the United States and Canada, which includes flat sheets, corrugated sheets, pipe, shingles, etc., may, it is estimated, consume 200,000 tons of asbestos per year. The asphalt tile industry, which uses short asbestos fibre, has shown considerable expansion during the past few years, and this consump tion may be estimated at 120,000 tons of asbestos per year. There is also an increasing pro duction of asbestos papers, mill boards, brake linings, clutch fac ings, pipe coverings, roofings, black line, etc., that require asbestos fibres of various grades. The total amount of fibre used in these prod ucts in Canada and the United States might be estimated at 130,000 tons per year. Therefore, it may be said that the potential volume of asbestos that could be consumed in Canada and the United States might reach a total of 450,000 to 500,000 tons per year for the next few years provided the fibre is available and the asbestos products manufactur ers can continue, uninterrupted by labour conditions or shortage of materials, to expand the present factory equipment. Conclusions , In concluding, the writer wishes to point out that, in our work on asbestos fibres in the Research Cen ter, we have set up a procedure to be followed during our fundamental and development investigations. This plan of investigation includes the following types of approach: (1) Obtain the complete history of the source of the fibre. (2) Study its physical and chemi cal properties. (3) Improve the physical or chemi cal properties of the fibre so that it will be better suited for its specific use. (4) Study the processing technique and develop the best method in the pilot plant for handling the fibre and then recommend the procedure to be followed in the factory. (5) After the preliminary investi gations are completed, the processed fibres are submitted to other pilot-plant operations in the Research Center, where regular products are made on a semi-commercial scale. If the problem involves very large tonnages of asbestos fibre, the work is transferred to the large-scale pilot plant at the mines at Asbes tos, Que., where large size produc tion milling equipment is used. In this manner, we are able to follow the fibre from the earliest stages in the crude or mill rock throughout all its investigations in the laboratory and pilot plants, and during the time when the fibre is incorporated into the final product. This procedure gives the research man full responsibility for the re search and development work until the factory accepts the material and takes over the production job. References Allen, M. A. and Butler, G. M., Univ. of Arizona Bull. No. 113, Mineral Technology Series No. 24, June 1st, 1921, pp. 1-29. Keep, F. E., Geology of Shabani Min eral Belt; Geol. Surv. Rhodesia, Bull. No. 12, 1929. Cooke, H. C., Geol. Surv. Can., Mem. No. 12, 1937. --4-- MTC 000656 t Filterability of Asbestos Fibres Used in Wet Processes * By M. S. BADOLLET (Annual General Meeting, Montreal, Que., April, 1919) (Transactions, I'olume LII, 1949, pp. 260-264) Introduction by mechanical equipment with a minimum reduction of fibre length. THE manufacture of some as The many physical properties bestos products involves the use peculiar to asbestos fibres (1) should of asbestos in combination withbe well understood by the manu water, cement, and granular fillers. facturer. They may be briefly This type of production is often called the wet or semi-wet process enumerated as follows: fibre strength, fibre composition, slimi and requires the use of either pres sure or vacuum to eliminate the ness, length, harshness, softness, and filterability or porosity. This excess water in the fibrous blend so that the product may develop last named property is important and it is the purpose of this paper the proper strength during the cur to present some basic studies on ing stage. filtration as applied to production Asbestos-cement products avail problems. able today are numerous and are a valuable contribution to the build ing industry. Their production in Filtration Studies quantity depends upon the type of equipment employed, the speed at which it can be operated, and the factors involving the blending of The term filtration is used here to designate the removal of water from a slurry or wet matte of as bestos fibre by applying pressure asbestos, cement, water, and fillers. or vacuum. Assuming that all mechanical troubles at a plant have been elim A simple laboratory procedure has been devised for accurately inated or are under control, it is still advisable to study the basic properties of the raw materials used in the asbestos-cement products. The type of cement used is an important factor, but usually the manufacturer of the cement is will ing to work with the asbestos de velopment engineer and recommend the proper type to be used for the product. The source of the water should not be overlooked as a general fac tor involved in the production cycle. Waters containing a high percentage of contamination such as silts, hu mus, or soluble salts may adverse ly affect the setting of the cement or cause a slowing down of the production rate in the factory. Asbestos in combination with ce ment is an ideal material for build ing products. It is inorganic, will not deteriorate, and can be handled measuring the filterability of as bestos in water, or of asbestos in combination with water and ce ment. The apparatus (Figure 1) consists of a vertical bomb or cylin drical iron pipe threaded at each end so that it may be capped. The bottom cap is fitted with a screen and outlet so that the filtration proceeds on this disc of a definite area. The water or filtrate which passes through the filter matte is passed up vertically through a rotometer which accurately measures the rate of flow. Pressure is applied gradually by the use of a pressure regulator so that a constant rate of flow is maintained throughout the filter cycle. Slight agitation is maintained within the bomb by al lowing the entering air to percolate (1) See Balollet, M.S., Research, on Asbestos Fibres; C.I.M., Trans., Vol. LI, 1948, pp. 131-134. `Research Centre, Johns-Manville Corporation, Manville, New Jersey. Figure 1.--Apparatus for measuring the filterability of asbestos. --5-- MTC 000657 upward from a point or several points above the filter disc. This prevents settling of the slurry and allows a gradual deposition of the fibre upon the filter disc. The data obtained by such filter tests are plotted on co-ordinate paper and the areas under the curves are measured in square inches so that the values can be compared directly with each other. Procedure is as follows: The top of the filter unit is removed and a small sample of the fibre un der study, made up as a slurry in water, is poured into the bomb and the cap replaced. A low pressure is applied and the entire amount of fibre in the bomb is allowed to deposit on the filter disc or leaf to form a thin pre-coat. A larger sample of the fibre is now made up as a slurry and placed in the bomb for the actual test. By means of a pressure regulator, the float in the rotometer is maintained at a standard rate of flow (in terms of c.c. per minute). The pressure is continually applied so that the flow rate is kept constant. Pressure read ings are recorded every 30 seconds and the pressure values are plotted against time in order to present the data graphically. Table I.--Effect of Asbestic Fines Area Designation (See Figure 2) Percentage of Fines in . Sample E................ D............... C +D............... B+C+D.............. A -f B -f C -f D................ 0% 10% 20% 30% 40% Area Under Curve at 6 Min. 10.326 sq. in. 12.148 sq. in. 15.989 sq. in. 22.655 sq. in. 27.373 sq. in. Percent Decrease in Porosity Due to Presence of Fines 17 64% 54.84% 119.40% 165.10% Calculation (example): 12.148 - 10.326 X 100 ----------------------------------------- = 17.64% 10.326 ence of asbestic fines is objection able since they adversely affect the filtration characteristics of an as bestos fibre. The manufacturer ob jects to the presence of asbestic fines because they slow down the wet process machines and also in crease his overall fibre losses by as slow filtering, poor saturation, and increased density, and, finally, will give a brittle product. Effect of Temperature of Water In many wet processes, the water temperature is an important fac tor and production rates can be in creased by taking advantage of this fact. A series of four tests were con ducted at temperatures (F.) of Effect of Asbestic Fines A quantity of commercial asbes tos fibre was first cleaned to re move all dust. rock, and asbestic fines. This clean fibre was then set aside for the filtration studies. The asbestic fines or dusts used in these studies were obtained from an asbestos-mill dust house. A micro scopic examination of the dust showed that the particles ranged in size from 0 to 40 microns and were of irregular shape with practically no fibrous structure. Blends of these fines with the clean fibre were made up, contain ing, respectively, 10. 20, 30 and 40 per cent fines. Filtration studies were made on each of these blend* and also on the clean fibre. The results of the tests are given in Table I and are shown graphically in Figure 2. The data, when plotted graph ically (Figure 2), give typical fil tration curves and from these we are able to measure the area under each curve and calculate the effect of the fines upon filtration. It will be quite evident from in spection of Figure 2 that the pres Figure 2.--Effect of fines on filterability of asbestos. Figure 3.--Effect of water temper ature on filterability of aiberto*. passing through the equipment screens into the waste water. It is known that commercial fibres contain various amounts of admixed asbestic fines and that, if these form a high percentage of the total fibre mass, they will cause serious manufacturing troubles such 50. 72.5, 86 and 122. In each case the same fibre was used, so that the only variable was the tem perature of the water. The results of the tests are pre sented in Table II and, graphically, in Figure 3. Table II.--Effect of Water Temperature Temperature f. Area Under Curve Increase in Filtra- at 6 Min. tion Rate 59 .................................................... 72.5 .................................................... 86........................................................... 122........................................................... 25.918 sq. in. 21.629 sq. in. 18.481 sq. in. 13.295 sq. in. 16.6% 28.9% 48.7% Calculation (example): 25.918 - 21.629 X 100 -------------------------------------- --- 16.6% 25.918 -- 6-- MTC 000658 These tests show that, by in creasing the water temperature, it is possible to remove the water from an asbestos fibre slurry in a shorter time and thereby give the manufac turer an opportunity to increase his production rate. In some plants, warm water is used so as to gain this added advantage. Table III.--Effect of Serpentine Rock Screen Size of Rock Calc. Specific Area Under Surface Area Curve at 6 Min. 0.................... -- 8 + 10 mesh . . - 10 + 14 mesh - 20 + 35 mesh.......... - 65 + 100 mesh.......... -150 + 200 mesh. . __ 11.8 cm2/gm. i 16.8 cm2/gm. 37.4 cm2/gm. 154.0 cm2/gm. 294.0 cm2/gm. 28.9 sq. in. 14.8 sq. in. 11.8 sq. in. 9 0 sq. in. 16 1 sq. in. 17.6 sq. in. Improvement in Filtration __ 48.8% 59.2% 68.8% 44 3% 39 1% Effect of Serpentine Rock of Different Sizes Serpentine rock relatively free from asbestos was ground in a ballmill and screened for size by siev ing on a rotap apparatus. A con stant weight of each screen fraction was then added to a constant weight of asbestos in order to determine which size is the most effective in improving filtration. The assump tion was made that the rock par ticles were perfect spheres and that the average diameter of each particle was the average of the screen open ings. Although it is realized that this involves a certain amount of error, the method does give a rough comparison of the specific surface area in terms of cm2 per gram. 28.9 - 14.8 100 Calculation (example):------------------- ------------- 48.8% 28.9 upon filtration and upon the qual ity of the manufactured products. A quantity of crudy fibre was selected and fractions of various screen sizes were separated on a rotap apparatus so that a constant weight of these bundles could be added to a constant weight of fibre. greatest effect upon speeding up the filtration properties of an as bestos fibre (see Table IV and Fig ure 5). As the surface area of the hundles increased, the tendency was for filtration to be slowed down, thus indicating that, if all the bundles were completely opened, It was found that, of the several screen sizes of serpentine rock used in the tests, the most effective in improving the filtration characteris tics of the asbestos fibre was the -- 20+35 mesh material. The larg er sizes apparently do not give as good a porous structure to the fibre, and with the smaller sizes, also, the filtration rate is lower, due to decrease in the porosity of the fibre mass (see Table III and Fig ure *). Figure 4.--Effect of rock particles on filterability of asbestos. Figure 5.--Effect of crudy fibre bundles on filterability of asbestos. Although the presence of some suitably sized serpentine rock is beneficial in improving the filterability of asbestos fibres, care must be taken not to use an excessive amount, which would adversely af fect the physical structure of the asbestos-cement products. If the percentage of grit or serpentine rock exceeds that normally present in commercial grades of asbestos, then the asbestos-cement products will have lower strengths and in creased densities. Effect of Crudy Bundles or Pencils Milled fibres as sold to custom ers frequently contain crudv bundles that are not properly fiberized. The presence of these bundles, if not further opened by re-processing, will have certain adverse effects In this case it was necessary' to make the assumption that the fibre bundles were perfect cylinders. The lengths and diameters of these hundles were measured under the microscope and then averaged for each screen fraction. The crudy bundles with the small est specific surface area had the the filtration rate of the fibre wculd approach that of a normal fibre and would be classed as slow. Screening of the crudy bundles on the 10 mesh screen indicated that the grading for size was not ac curate, as the specific surface area of these bundles should have been between 102 and 127 cm2 per gram. Table IV.--Effect of Crudy Fibre Bundles Screen Size Original.. -4+8 mesh --10 + 14 mesh -14+28 mesh -28+33 m:sh I Calc. Specific I Surface Area 0 102 0 cm2/gm. ! 168 0 cm2/gm. 127 0 cm2/gm. 164 0 cm2/gm. Area Under j Improvement in Curve at 6 Min. filtration 28 9 sq. in. 7 1 sq. in. 9 4 sq. in. 11 6 sq. in. 13 2 sq. in. 75.5% 67 5% 59 8% 54 3% Calculation (example): 28.9 - 7.1 X 100 ---------------------------- 75.5% 28.9 --7-- MTC 000659 The other specific surface area calculations seem to agree with the general trend and indicate a decrease in the fibre bundle cross-section. As shown graphically in Figure 5, the --4+8 mesh bundles in creased the filtration rate more than did the --28+35 mesh bundles. From the viewpoint of the manu facturer. the presence of the crudy fibre bundles would he considered satisfactory -- particularly if he did any reprocessing -- as he would have more effective fihre present in his product. It might even en able him to reduce the quantity of fihre used and make a superior product at a reduced fibre cost. However, the presence of crudy fibre bundles has a detrimental ef fect in some products because they impart a rough surface, and any subsequent sanding or polishing would result in the product having an uneven texture. Fibres From Different Deposits For these tests, samples of as bestos crudes from a number of sources were carefully processed so as to avoid contamination with par ticles of the wall-rock, and, after a preliminary opening by hand, were passed through a laboratory-type micro-pulverizer so as to obtain fibres of approximately the same size. The pulverized fibres were measured under the microscope and were then tested for filterability. The cross-section measurements indicated that, for each prepared sample, the widths ranged from 1 to 5 microns and the lengths from 10 to 200 microns. Figure 6 shows graphically the results of filtration tests. The slow est filtering fibre tested is the Arizona soft fibre (curve A). The prepared Canadian soft fibre (curve C) is faster filtering than the Arizona soft fibre. The Italian slip chrysotile (curve D) is faster filtering than the Canadian soft fibre. The Cyprus chrysotile (curve F), which has a semi-harsh feel to the hand, would be considered fair ly fast filtering in comparison with the other soft fibres. Curve E rep resents a typical Canadian semiharsh fibre and it would be consid ered fast filtering. Curve B was ob tained for a very harsh fibre from Arizona; this is extremely fast fil tering. The area measurements of each curve are given in Table V. In order Table V --Chrysotile from Different Sources (Fibre width, 1-5 microns; fibre length, 10-200 microns) Identification Arizona soft................. Canadian soft ... Italian slip.. Cyprus........... Canadian semi-harsh. Arizona harsh............ Curve (Figure 6) A C D F E B Area Under Curve at 4 Min, 13.80 sq. in. 8.50 sq. in. 6.40 sq. in. 3.30 sq. in 1.78 sq. in. 0.24 sq. in. Filter- ability (Can. soft - 100%) ,37 6% 00.0% 124.7% 161.2% 179.0% 197.2% to compare the curves we have taken the Canadian soft fibre as the standard. By this method we are able to express the filterability of each fibre as a percentage, with Ca nadian soft fibre taken as 100 per cent. Since the Arizona soft fibre was so slow filtering it was impos sible to measure the area of its curve at the end of six minutes. As a con sequence. the area measurements in these tests were all made at the end harshness, the fibre will filter fast er. Application of Technological In formation to Manufacturing Processes The importance of the filtration characteristics of asbestos fibres should not be overlooked by the manufacturer of asbestos products which are made by wet or semi-wet methods. In the manufacture of paper, millboard, shingles, pressed sheets, or pipe, the ease of the re moval of the water will to a cer tain extent affect the production rate. If a fibre is difficult to dewater, then the manufacturer eith er changes his process mechanical ly, changes his grade of fibre, or uses some fibre blend that will al low him to maintain the production rate. By adding a fast filtering fibre to a slow filtering fibre it is pos sible to step up production rates, but the manufacturer must be care ful not to overdo this mixing for fear of losing strength in the prod uct. Figure 6.--Filtration characteristics of chryaotile fibre from varioua localities. of four minutes instead of six min utes. On this basis of comparison, the Arizona soft fibre (curve A) is 62.4 per cent slower filtering, and the Italian slip fibre (curve D) is 24.7 per rent faster filtering, than the Canadian soft fibre (curve C). For the Cyprus, Canadian semi-harsh, and Arizona harsh fibres, the filtra tion rate is faster than for the soft chrysotiles. A physical examination of each of these fibres would show that the softer and silkier a fibre, the more difficult it is to filter. As the soft ness of texture decreases and the fibre assumes a certain degree of It is usually considered that fibres ranging from semi-harsh to harsh are fast filtering and like wise that the fibre strength decreases with its degree of harshness. There fore the manufacturer must exer cise care in selecting his fibre, or the ratio of soft to harsh fibre in his fibre blend must be well plan ned, in order to obtain the best fibre for his particular process. At the present time, United States manufacturers of asbestos products who employ the wet or semi-wet process consume approximately 340,000 tons of asbestos per year. This includes approximately 11,000 tons of fibres from countries other than Canada and the United States. This immediately raises the question as to why these foreign fibres are used. Would they be used if Can ada could supply the total tonnage required? The answer to this ques --8-- MTC 000660 tion is not so much concerned with the availability of the fibre as with the particular physical properties of the foreign fibres that give the manufacturer the effects desired in his operations. These foreign fibres have the required free filtering properties and at the same time have strength. They also give the manu facturer sufficient latitude for blending, so that he can sweeten his slow filtering fibres with strong fibres that are fast filtering. Conclusions (1) The presence of asbestic fines or shorts decreases the filterability of a fibre mass in proportion to the quantity of fines or shorts present in the mixture. (2) By increasing the water tem perature it is possible to increase the filterability of a fibre mass and thus improve the operation of thej process. (3) The presence of serpentine rock will increase the filterability of a fibre mass provided the rock is not too small in particle size. In the tests made, --20-|-35 mesh ma terial gave the maximum improve ment in filtration; with smaller sizes, the filtration rate began to decrease, due to decreasing poros ity of the mixture. (4) Crudy fibre bundles will im prove filtration. As the cross-sec tion of the bundles is decreased, the filtration rate is slowed down. The presence of large quantities of crudy bundles may also affect the surface texture of the product, a factor which must be taken into con sideration. (5) Fibres from different depos its show considerable diversity in filtration characteristics. Soft, silky fibres are difficult to filter; seiniliarsli to harsh fibres filter easily. (6) By knowing the filtration characteristics of different fibres it is possible to select those that fil ter rapidly or to blend fibres to obtain the desired filtration results to fit the manufacturing process concerned. --9-- MTC 000661 Processing Asbestos Fibres: Effects Upon Physical Properties By M. S. BADOLLET* (Annual General Meeting, Toronto, On/., April, 1950) (Transaction>, Volume LIII, 1950) Introduction mand for a particular fibre grade is great enough to warrant special ASBESTOS as received by the milling techniques. In such a case, manufacturer of asbestos prod the mill places this specially pre ucts is in the form of either crudepsared fibre in its regular produc or milled fibres processed to differ tion line. Each time it decides to ent degrees of texture. Frequently, produce a special fibre, several im the manufacturer finds that the fi portant points must be considered, bres as he buys them are not adapt such as: (1) Will this special fibre able to his plant processes, and, require new processing equipment therefore, it is necessary for him or can existing equipment be used? either to contact the mine supply (2) Will the over-all capacity of ing the material and request a spe the mill be reduced? (3) Does this cific texture for the fibre or to re special fibre have specific physical process the fibre in his own plant. properties that cannot be matched When the mine or mill manager is bv present fibres offered to the requested to change his milling public? and (4) Can the extra pro equipment to produce a special duction cost be recovered? grade of fibre with a specific tex ture, he would like to comply. How Fibre milling is a technique that ever, if he began to make special differs somewhat with each plant. fibre grades for each customer, he All producers are continually mak would be in the position of con ing mechanical changes in order to stantly rearranging his milling tech improve their products while main nique, adding more equipment, or taining their production schedules. building a new mill. In the end, This problem is no easy one, and it his mill would be producing hun requires full co-operation between dreds of different fibre grades of the Sales staff and the Mill Depart varying degrees of texture, which ment. is a difficult accomplishment if he desires to maintain any quantity of Viewpoint of the Customer production. The customer has an established Viewpoint op the Miner plant which produces a certain line of asbestos products and probably It is well known that asbestos fi bres as produced at the Canadian mines fall into definite groups, such as crudes and milled fibre. Each of these groups has a number of sub-divisions, and many of the fi bres have different textures such as crudy, semi-crudy, open, and well opened. No doubt these several de grees of texture have been estab lished after many years of experi ence based upon the general demand of the customer. Therefore, the mill tries to maintain a set of standard fibre textures, with gaps between the different grades for future ex pansion if the demand is sufficient. In some instances, the customer de at one time he set up his plant with the minimum of equipment ne cessary for utilization of the regu lar grades of fibre as produced at the mines. After a number of years of competition and new develop ments, his products possibly did not meet the newer specifications, and the first question that came up for discussion was asbestos fibre. Was he using the proper fibre, or should he install equipment to re-process the fibre in order to change its characteristics? At this point, the Research and Development Engin eer should step in and review the entire problem with the production department and decide whether the fibre must be changed to fit exist *Research Centre, Johns-Manville Corporation, Manville, N.J. ing equipment, or whether the equipment should be modernized to handle the fibre. Since many man ufacturers of asbestos products pur chase their asbestos from more than one source -- Canada, United States, Africa, Russia, and Aus tralia -- it is important for them to know how to handle the several fibres properly. This is particular ly true when fibres of the same grade are purchased from two or more different sources and, conse quently, may have entirely different physical characteristics. To illustrate some of the prob lems facing the manufacturer of as bestos products, typical shipments of asbestos have been obtained from different sources and the effects of willowing upon the physical prop erties of these fibres have been in vestigated. African Blue (Crocidolite) This variety of asbestos would be difficult to use in its raw condition, as received in the United States, be cause of its crudy nature, quantities of unopened bundles of fibres, and the presence of pieces of rock of varying size. Therefore, the cus tomer must either have some tvpe of processing equipment to convert this fibre to the proper condition for use in his products or he must try to convince the miner in Africa to process his fibre to the customer specifications. Years of experience have shown that the easiest answer to this problem is for the customer to re-process the fibre himself, be cause he knows exactly what he wants. African Blue fibre is strong but harsh, and by successive willowing action the fibre becomes bulky and loses length. A normal Quebec-screen test, af ter one willowing action, shows that approximately 13 ounces of the total 16 ounces remain on the top screen. Successive willowing stages do not change this result sufficiently to show what is happening to the phys ical properties of this fibre. -- 10 -- MTC 000662 By a careful water eleutriation method, a definite indication can be obtained of reduction in fibre length after each willowing action. Figure 1 shows that, for five willowing ac tions, the +14 mesh fibre dropped from 52.6 per cent to 30.6 per cent, a length reduction of 40 per cent. Coincident with decrease in fibre length, there is also an increase of fines (-- 200 mesh) from 30.4 per cent to 45.4 per cent, or an increase of 49 per cent. Therefore, to wil low this fibre successfully, a mini mum amount of willowing action should be used in order to avoid a loss in length of fibre and the for mation of fines which would be lost in later stages of processing or product formulation. In the production of asbestos products by wet methods, the dens ity of the opened fibre as well as its ability to remain in suspension are important factors. The buoy ancy of a fibre can be determined from a slurry of fibre in water, with a definite weight of fibre in a constant and definite volume of wa ter. After thorough mixing, the slurry is allowed to settle, and the rate of settling is read in cubic cen timeters for a given time. In most cases, the settling Teaches equili brium after one hour. The density can be determined by filling a stan dard measure with a capacity of, say, a cubic foot, and obtaining the weight of its contents. The results are then expressed as pounds per cubic foot. Figure 2 shows that the buoyancy of the Blue fibre increases rapidly for the first willow pass, and that after the second pass the slope of the curve decreases. Conversely, the curve for the density values shows that the density decreases rapidly for the first pass and after the sec ond pass changes only slightly. An interpretation of the buoyancy and density curves indicates that this particular Blue fibre reached a high degree of opening at two pass es through the willow and that further willowing, at added expense, would effect relatively little im provement in the quality of the fibre. During the process of willowing a fibre there is another important change taking place in its surface area. This change, also, is indi rectly reflected in the buoyancy and the density measurements. Samples of the Blue fibre were measured for surface area by the air permeability method, using the Bowen apparatus and following the Lea and Nurse technique (1). A number of methods have been advo cated for determining the surface area of solids and each has advan tages and disadvantages. However, after considerable experimentation, l it is thought that the method used in this investigation, while not giving absolute values, shows the general trend and indicates physical changes to the fibre after each processing action. Since it is known that wil lowing action on a fibre will open bundles into thinner cross-sections, reduce length, and create fines, there should result a cumulative measure of these effects by obtain ing the surface area measurements. Figure 3 shows the surface area measurements of the Blue fibre in its original state and after each willowing action. The results are expressed in square centimeters per gram of fibre, and the data, ex pressed graphically, yield a curve which shows that: (1) the surface area is increased after each willow action; and (2) this measurement shows a greater change per willow pass than the tests for buoyancy and density, because of the change in fibre length and the formation of fines. The conclusions that can be drawn from these test data on open ing African Blue are: (1) the fibre length decreases after each willow ing pass, with resulting increase of fines; (2) the buoyancy and density values of the fibre show only small changes after two passes through the willow; (3) surface area meas urements show that, after each wil low pass, the area is increased; (4) a total of two passes through this particular willow would be the maxi mum treatment to produce a fibre for use in a plant; (5) one willow pass would be acceptable, since it produced a fairly open fibre with the minimum of length loss and creation of fines. Rhodesian Fibre (Chrysotile) This particular grade of fibre as received in the United States would also be difficult to use in most wet processes for the production of as bestos products, because of the pres ence of crudv fibre bundles. The same procedure has been ap plied to this fibre as was used with the African Blue. The same equip ment and rates of feed were adopt ed for all tests, so that the mechan ical processing was constant in all cases. (1) For leferences, see end of pa per. This particular fibre was given a total of eight passes through the willow in order to determine the maximum opened condition of the fibre. The Quebec-screen tests increased slowly to the maximum value of 12.6 ounces on the second screen at four passes, and then decreased to 11.7 ounces on this screen at the eighth pass. The water eleutriation procedure for length indicated a decrease in fibre length ( + 14 mesh) after each pass, and a small increase in fines (--200 mesh). Figure 4 shows the two curves, which are entirely dif ferent from those for the African Blue fibre and which do not inter sect at any point. The slope of the curve for the + 14 mesh fraction begins to level off after six passes while the curve for the --200 mesh fraction shows only a small increase in fines after six passes. In other words, the + 14 mesh fibre decreased in length 61 per cent, but increased only 24 per cent in the objectionable --200 mesh fines. The buoyancy of the fibre con tinues to increase rapidly after each pass. Figure 5 indicates that the slope of the buoyancy curve would begin to level off at some point after eight passes. Density measurements show that the density of the Rhodesian fibre decreases rapidly for the first two passes and thereafter at a lesser rate, with indications that it would probably Teach its minimum density after eight passes. Surface area measurements made ru this fibre show that it has less surface area than the African Blue after each pass through the willow. The curve (Figure 6) has a lesser slope than that for the Blue fibre, which would indicate that this fibre does not break down into shorter fibre lengths as rapidly as the lat ter. In other words, it resisted de structive action better than the Afri can Blue tinder identical test con ditions. Canadian Chrysotile -- Semi-Harsh A quantity of Canadian semiharsh fibre was obtained and sub jected to numerous willowing actions to determine its resistance to fibre length destruction. This particular fibre contains considerable quanti ties of crudv fibre bundles that should be opened before it is usable in an asbestos product; therefore, it was subjected to a total of ten wil low treatments. The Quebec-screen test indicated -- 11 --- MTC Figures X, 2, ind 3.--African Blue. Figures 4, 3, and 6.--Rhodesian. Figures 7, 8, and 9.--Canadian Semi-Harsh. 1 i 4< I --'.---- ^ N. * > i-------r *----- r---- Figures 10, 11, and 12.--Canadian Soft. V T 1. m i* t 1 vj * 7 *< * At J Figures 13, 14, and 15.--Canadian Harsh. -- 12 -- r l-- * 4 i> MTC 000664 a large increase on the second screen for the first pass, and only small increases for each succeeding pass up to six. From that point up to ten passes, the test indicated minor changes only. The water eleutriation test indi cated an 18i/j per cent decrease in length for the + 14 mesh fibre for the first two passes. Examination of the curve (Figure 7) reveals that the remaining willowing actions from two passes to ten had only minor effects upon the destruction of the fibre length. The fines (--200 mesh) show a 14 per cent increase for the first two passes and only a slight increase for the re maining treatments. From these data it can be concluded that this fibre resists destructive action and remains in good conditoin even after ten passes through the willow. The buoyancy value of this fibre, as shown by Figure 8, increases rapidly for the first four passes, and then the slope of the curve begins to decrease. The density values (Figure 8) show a large change for the first two passes and only minor changes for the remaining tests. Surface area measurements are not increased as rapidly as in the case of either the Rhodesian or the Blue fibre. Figure 9 illustrates the increase in surface area and, as will be noted, the slope of the curve is less than for the two African fibres. From the test data obtained on this Canadian fibre, it can be stated that willowing action is not detri mental to its physical properties and that it resists destructive action bet ter than the African Blue or the Rhodesian fibre. Canadian Sort Chrysotile Fibre A typical Canadian soft, silky fi bre was willowed a total of sixteen times in order to determine its re sistance to the action of a willow. - The Quebec-screen test indicated a change on the second screen after the first pass. This same screen began to indicate a decrease after nine passes and, at sixteen passes, had decreased to a value slightly above that at the starting point. The water eleutriation curves (Figure 10) show that this fibre, although it lost length and some pul verizing took place, did not disin tegrate; therefore, it can be con sidered highly resistant to process ing equipment. It would probably be satisfactory for use in its orig inal state unless an increased fibre buoyancy was desirable for some specific usage. The +14 mesh fibre decreased 20 per cent after sixteen passes, and the --200 mesh fines increased 14!/i per cent for the same number qf passes, a further proof that this fibre resisted the destructive action of the mechanical equipment. The buoyancy of this fibre, as il lustrated by Figure 11, shows a rapid increase for the first two to three willowings and only a gradual increase by subsequent treatments. However, the buoyancy values are all lower than those for the three fibres previously discussed, which shows that this fibre is relatively free from crudy fibre bundles and is fairly well opened. The density values decrease for the first two to three passes and then practically level off, with only minor changes brought about by further processing. A total of two willow passes would probably be sufficient to place this fibre in good open condi tion. From the curve obtained (Figure 10), it can be concluded that this number of willowing ac tions did little harm as regards de stroying fibre length or pulverizing the fibre. The curve also indicates that the quantity of unopened bun dles was small and that the fibre was fairly well opened at the asbes tos mill prior to shipment to the customer. Therefore, this fibre can be used in its original condition, as received from the mine, without fur ther processing on the part of the purchaser. Surface measurements on the Canadian soft fibre are given in Figure 12. The fibre in its condition as received presents a greater sur face than any of the other fibres discussed in this investigation. This would indicate that the combination of fibres of various lengths and fines created a large surface area. The surface area of this fibre at any given pass is greater than that of any of the other fibres so far dealt with. Therefore, if a manu factured product requires a fibre of great surface area, this one should fit the specification. However, since this physical property is only one of many to be recognized, we must balance it against the other factors before reaching a final conclusion. Canadian Chrysotile -- Harsh Fibre This harsh fibre of the chrysotile variety has some interesting char acteristics from the viewpoint of its physical properties. It was subjected to a total of six passes through a willow, and, in the Quebec-screen test, each pass indicated an increase on the second screen. The crudy Fibre bundles opened easily and became fairly bulky, so that most of the fibre re mained on the second screen. Water eleutriation tests indicated a definite length loss in the +14 mesh fibre after each pass through the willow (see Figure 13). After six passes, this decrease in length amounted to 81 per cent and the quantity of fines (-- 200 mesh) showed an increase after each wil low action. However, the amount of this increase is only 22 per cent, which would not be considered alarming. These data indicate that the long fibre will not resist flexing ation by willowing, and that shorter lengths were produced without pul verizing to dust. Therefore, any mechanical re-processing of this fi bre after shipment to a customer should be extremely mild, or the fibre should be used without addi tional processing. The bucvancy value of this fibre (see Figure 14), while showing an improvement after each pass, does not equal that of the other fibres discussed up to this point. The density shows a rapid im provement for the first two willow passes and only minor changes after subsequent passes. Therefore, for re-processing of this fibre, from the viewpoint of buoyancy or density, a minimum of two passes, or even one pass, would be considered sufficient. However, there is an important loss of length after one or two passes, and it might be advisable to avoid any further processing of this fibre. Surface area measurements on this harsh fibre, shown graphically in Figure 15. indicate a smaller in crease in area than for any of the other fibres tested in this investi gation. In other words, this fibre maintains low surface area and low buoyancy, although it loses length, in processing. For certain products these properties are important and should not be overlooked; for other products, great care must be exer cised in handling the fibre before and during its use. The surface areas of all fibres tested are charted in Figure 16, which enables a direct comparison to be made between the several fi bres for the same number of passes through the willow. The two ex treme fibres are the soft, silky chrysotile of large surface area, and -- 13 -- MTC 000665 Figure 16.--Effect of willowing on surface area, various fibres. the harsh chrysotile, having small surface area. Effects of Willowino Upon FlI.TERABII.ITY OF FlBRE When a fibre is opened by mech anical means, a fibre of greater surface area results. Figure 17 shows, for each of the fibres inves tigated, the effects of four willow passes upon its filterability. The soft, silky chrysotile fibre rapidly becomes more difficult to de-water after each opening process, with the result that plant produc tion is retarded. The Rhodesian fibre remains fast filtering up to four passes through the willow, at which point it begins to show a trend toward more diffi cult filtration. However, after this number of willow passes, the fibre would still be considered satisfac tory for any wet process. Of all the fibres tested in this investigation, the filtration charact eristics of the two identified as 'semi-harsh' and `harsh' are least af fected by the willowing process. Both would be considered ideal for sweetening poor filtering fibres, even if they had been highly opened by mechanical processes. The filterability of African Blue fibre is only slightly affected by the willowing process, and at four passes it is still considered a fast filtering fibre. Figr-e 17--Effect of willowing on filterability of verioui fibre*. Correlation of Test Data In general, there is good correla tion between buoyancy and surface area values at a given number of willow passes. When the surface area data and buoyancy values are expressed graphically, most of the points fall upon a straight line, in dicating that either test can be used as a guide for determining the de gree of opening of a given grade of fibre. When the surface area measure ments and the decrease in --j-14 mesh fibre expressed in percent are shown graphically for each willow pass for each kind of fibre, there ap pears to be a fairly good correlation, since most of these points also fall upon a straight line. This is an indication that the surface area of processed fibres is greatly influ enced by the destructive action of willow hammers. Application of Test Data The manufacturers of some as bestos products require that asbes tos fibres meet certain specifications as to rate of filtration, fibre strength, length, buoyancy, density, and surface area. In a previous pa per on filtration, presented at the 1949 Annual General Meeting of this Institute (2), it was pointed out that, by opening a fibre, the rate of filtration is decreased. In all of the investigations discussed in the present paper, each fibre shows a trend to become more difficult to filter after each willowing action. Therefore, some decision must be made on the basis of the general over-all effect a willowing action has upon physical properties of the fibre. Fibre strength (3) is important, and soft, silky fibres retain their strengths better than harsh fibres when processed by mechanical equipment. If filtration is the most impor tant factor governing the production rate of an asbestos product, then, given a fibre of satisfactory strength, a minimum amount of wil lowing should be used. If a low density or a high buoy ancy value is required and filtra tion is not important, then the fibre should be well opened to obtain its maximum fluffing condition. If a fibre is required to have strength and high surface area with a minimum loss of fibre length and minimum formation of fines by a willowing action, then it is neces sary to select a fibre or a blend of fibres that will meet these require ments. The Research or Development En gineer should make a thorough study of the available fibres. Then knowing the fibre specifications re quired for a given asbestos prod uct, he should make definite recom mendations to the manufacturer as to the kind of fibre and type of re processing necessary to obtain the maximum value of the fibre for the specific utilization. Conclusions (1) Asbestos fibres as received by the manufacturer of asbestos prod ucts usually require some type of re-processing before they can be used to best advantage in a product. (2) Any fibre re-processing meth od adopted by a manufacturer will produce changes in physical prop erties, such as a lowering of fibre strength, loss in length, formation of fines, different degrees of buoy ancy, changes in density, increase in surface area, and a tendency to be come more difficult to filter. (3) The soft chrysotile fibres re sist the destructive action of a wil low fairly well, but have the great est surface area and are the most difficult to filter at any given num ber of willow passes. (4) African Blue fibre loses length, increases rapidly in surface area, becomes relatively buoyant after each willow pass, and filters rapidly. (5) Rhodesian fibre loses length, increases in buoyancy, has less sur face area than either the Canadian soft fibre or the African Blue fibre, and its filtration characteristics re main good even after four willow passes. (6) The semi-harsh fibre resists the destructive action of the wil low fairly well, but is less resistant to willowing action than the Cana dian soft fibre, increases rapidly in buoyancy, is rapid filtering, and has less surface area than the Blue, Rhodesian, or Canadian soft fibre. (7) The harsh fibre loses length rapidly, increases in buoyancy slow ly. lias the lowest surface area of all fibres studied in this investiga tion, and remains fast filtering after each willow stage. References (1) Lea, F. M., and Nurse, R. W,, Specific Surface of Fine Powd ers; Soc. Chem. Ind., Trans., Sept., 1939, p. 277-283. (2) Badollet, M. S., Filterability of Asbestos Fibres; C.I.M., Trans., Vol. LII, 1949, pp. 260-264. (3) Badollet, M. S., Research on Asbestos Fibres; Can. Min. Jour., April 1948, pp. 213-216. -- 14 -- MTC 000666 Asbestos, A Mineral of Unparalleled Properties *1 By M. S. BADOLLET* (Annual General Meeting, Quebec City, Que.. April, 1951) (Transactions, Colume LIf', 1951. pp. 151-160) Inrtoduction This information was obtained over a period of years by search of the THE DEMAND for general literature and by experimental in knowledge on asbestos fibres vestigations. has increased considerably in recent years. A few publications have printed data showing some of the Properties of Asbestos Fibres physical and chemical properties Table I (1. 2) presents, in con of asbestos, but in many cases this venient form for easy comparison, information is difficult to find and data on the principal properties of is seldom available when needed. the several varieties of commercial During the past ten years, Johns- `asbestos' -- actinolite, amosite, an- Manville have received many inquir thophyllite, chrysotile, crocidolite, ies on the physical and chemical pro and tremolite. Under each variety perties of asbestos fibres. Fortun- of fibre is a brief statement describ atelv, we had some of the informa ing the properties in terms of struc tion in our files, and we willingly ture, mineral association, origin, supplied the data. etc. Some of the more important To help relieve this need, we have data in Table I are expanded and tried to present briefly in this paper presented in Table II (3). some of the interesting and out standing properties of asbestos. Solubility of Asbestos "Research Center, Johns-Manville Corporation, Manville, N.J. (1) For references see end of pa per. A report (4, 5) discussing the effects of acids and caustic on as bestos fibres was published in Ger many in 1927. Re-published several times, it appears to be the only in formation available on the subject. When, several years ago, it be came necessary to obtain technical data, not available in the literature, on the solubility of commercial grades of asbestos, tests were ar ranged following the plan adopted by the author of the original article (4). Samples were obtained of actino lite from Canada, amosite from Af rica, anthophvllite from Georgia, chrysotile from Canada, crocidolite from Africa, and tremolite from California. These samples repres ented fibres obtainable in commer cial quantities. They contained some mineral impurities and thus their degree of solubility was not identical with that of hand-picked, highgrade crudes from these localities. The acids used in these tests were hydrochloric, acetic, phosphoric, and sulphuric. All were diluted to a 25 per cent acid solution, by weight. Table II.--Physical Properties of Asbestos Chrysotile Specific heat B.t.u./lb./F........... Tensile strength, lb./sq. in.......... Temp, at max. ignition loss, F.................. ...................... Filtration properties..................... Electric charge............................... Fusion point, F............................ Spinnability.................................... Resistance to acids & alkalies Magnetite content........................ Mineral impurities present.......... Flexibility....................................... Resistance to heat......................... Ionizable salts, micro-mhos........ (Relative dec. conductance) Colour............................................. 0.266 80,000 100,000 1,800 Slow Pos. 2,770 Very good Poor 0-5.2 Iron, chrome, nickel, lime High Good. Brittle at high temp. 1.82 Green, grey to white Amosite Anthophyllite 0 193 16.000 90.000 0 210 4.000 & less 1,600 to 1,800 Fast Neg. 2.550 Fair Good 0 Iron 1,800 Medium Neg. 2.675 Poor Very good 0 Iron Good Good. Brittle at high temp. 1.34 Poor Very good 0.58 Yellowishbrown Yellowishbrown, Some times almost white Crocidolite Tremolite 0 201 100.000 300.000 0 212 1,000 8,000 1.200 Fast Neg. 2.180 Fair Good 3 0-5.9 Iron 1.800 Medium Neg. 2,400 Poor Good 0 Lime Good Poor, fuses Poor Fair to good 0.84 -- Blue White Actinolite 0 217 1,000 & less -- Medium Neg. 2,540 Poor Fair -- Lime, w 1 "L Greenish -- 15 -- MTC 000667 PROPERTIES OF ASBESTOS FIBRES 'flocont studios claim that ( A t cryiti/ slructur* is monoclinic. 4HU M 4-- -*x 4X 40 X O 4o X* <W :s5, 0 0 M u Ul U 4 a x w x - # w a u X 3 0W ?5 xa 9 5 a * 0M w X0 X X a - 4 1333 4 t0 c &M .3 >x xu J *. x Jf XX x3 o- ?5i' C-- a -1 X0 wX -4X4 XX 0 M4 w XX 1i Xa w j u X0 No 4 Nu X V) u a 44 0 m -JX u p4 -- 0 a --X 0 xa o 0w M xo X X o ?: 1 N X xx o 2 X X ax Zi4 u So Irt U0 3 8 i u 1 1 \ * > at Ul -il > ol ff] 51 MIX 1-- (BL MX S5 w o U j X 5 m lb a X x w e M 0 X^ a c - X o X lb -- 4X 1_S _J 4 0 WO X4 6*^ OX x 3a Jhh WWW 4*X 3^0 a - --4 -o X o -x 4 iw U0 X W * X4w 4 J--w O XXo 9 C 0VX V* a y w x 0 X a t, -HO 3 3: o W : a x -- X 4 y H l S NX S S 3 mO w x w o - X -i W X W 4 " w y s 4 * J Jnw P WX m- 4 1 Jlii < o- X - 3 X a X -- a o 3 a O o a w X M W -- Ill Jw J fc> 4W ww xa o --4 4 4 UX x-- oo ss wx - OX au WJ w XX M NO 33 a -- - ww 0 u -X XX 4 o ** 4 4w N X X o --w X X z0 g* w 0 X o" a A S u 40 A X ^o Zl" O a 0 U 4 0 0-- ow a4 za o s" N w N I X K eg s 5 aM w 1 40 g r4 0 ? f o fj X X "T O X &i oi X 00 UW4 wax 0XX A ? a X 2 `fc' a o 2 M 3 4 a 4 0 e a u. a H C ; 4 (J 4a 0X X- o 4 W a w 2 4a 0X X-- , a0 o X X --0 o- 4 X wX o 4 4I X -|J j j SI O0 4U ua X A 0*w x -- M i/> a u s: X2 -- X 0V 4 0O lx -- x a 4u rs 34 5 a lb X JX uuu 4M ww X ? *- 0 3 XX 4 io ft Z * (J a U ^0 2 g i i *si XX - a o3 1O a 0 j 2 X c X X u X 1 X X u i 2 X* 0 XX w0 w-- x a a-i X w 0X --0 X- . g > 4 w-i >X W O 0 X w irt , X w 3 3 o w X 0X _l " w a * 0 0 mJ t/i w X 0 XX 2 Jt --w 3> X - . a X wX --w XX aa 0o 0 XX (H 0 X w 0 a w 0 > X o 0 X w a 4 ax x& <3 4 a a 0 a ww 04 > X 20 ww X (A O N ix i 7 (A a 9I N A O 10 9N 11 ** A <x 1 <a * ** oN A 11 M A X1 a w X JW " 0 0 o o 0X w w 3* X w 3i +1 a. a W0 o 4U X 40 --X aw a W ww a8 0A (3 < z4 0 4 w * a M a ib wa 0-- 0X --w xw ax 0M iS 4-- X0 wo a s w a 0 0 X o X A X a0 xu 4a xa 0a gw 3! X0 ww s 0 X 4 0 a aa 4w sX I* o4 XX j1 0 wa xa 4b XX a xa A W4 ax a4 a X -i 0X w X H 3 -J 0 o --x - 40 W 2 X g w X r* - A 0 J a aX w o 0 XX 4X a0 w xa 0 4 Ir 4 wa 0 X a 0 a> X a a -i 0 4 Ib 0 4 x WW 4U w J u 04 lb 0 4 wM bO J0 --2 40 aw a i 8 - * 4 W a a lb ie A W aX 0 a aX w X XX | a4 w 0 a-- a X w-- a 0f 0 ao w 0 X o b. M w Wb > A A< A4 0 w WJ 0 -X a2 -X X a X w X aw 40 X O2 w Xo aX --w X o Na 4 n W0 uw X 0w w aa A X X c -x a- 0X >0 X o ft. 40 aa w w Xo a ui X --w 4U lb 4 _4 3 w a0 X X w X u 04 a .. > H mb N 4 aoa w wX s 0xx w-- A - -1 0 ;; - 0 0 o Jt ? X -- WX Ou -- a 0, 40 a<i N aw u w w-- 5 --a o X u o wJ 0 4-- w J5 X 4 4 U a MX -- 2 x C - 4 W a a ib w a X w w a 3 S X a0 30 Wa w 0 A X 4 1 M 0 a ww X4 O -1 4 0 aw w- a X0 4X 0 0 4 a m X 0 XO 4a * Ib a w wa X wa *0 X Xw > 5 ft. w 5 3 Ui %S ^x s kS u Z IX 0 o -J ae 0 M UJ h UJ xc O > _ 0- Ul zz Ul A at >* Ui < o00 a. P- -- a --J 9 -J ID p" 0 (9 0- -- UJ Ul Ul A g oc -- -- z i h- oe or s :0 o A CD -- s * toc 0 z-- ZO 0 tu > >> 3 v> at 3 IX M Ul > w Uu o M uo at lx 3 IX Xz 0* K A dk A -- 16 -- MTC 000668 For the caustic solubility tests, solid pellets of sodium hydroxide were dissolved in distilled water to obtain a 25 per cent solution, by weight. 'Each fibre sample, consisting of 10 grams, was accurately weighed and placed in a flask containing 300 c.c. of the 25 per cent acid or caus tic solution. Two sets of tests were conducted, one at room temperature (26C.) for indefinite periods, the other at boiling temperature in a reflux con denser for two hours. At the end of each test, the fibres were removed from the solutions, washed free from acids or caustic, dried, and weigh ed. Table III shows the solubilities of these particular asbestos fibres in the four acids used and in caus tic soda. The action of boiling acids on asbestos fibres is severe, with ehrysotile the most soluble. Acetic acid was not so effective as the mineral acids in dissolving chrysotile, and caustic had even less action on this mineral. After chrysotile fibres had lost 55 per cent of their weight, exam ination showed the fibre structure to be almost all silica, brittle, and very fragile. The high solubility values for actinolite are, it is believed, due to the presence of soluble minerals as impurities in the commercial product used by industry. Anthophyllite seemed to resist acid action better than any other variety, with crocidolite second best, tremolite third, and amosite fourth. All these solubilities would vary depending upon the source of the fibre and whether the samples rep resented pieces of `crudes', or milled fibres as purchased on the market. The action of acids and caustic Table III.-- Solubility of Asbestos Per Cent Loss in Weight, Re-fluxing Two Hours 25% Acid or Caustic HC1 Actinolite..................... Amosite......................... Anthophyllite.............. Chrysotile..................... Crocidolite................... Tremolite..................... 20 31 12 84 2.66 55.69 4.38 4.77 CHaCOOH t 12.28 2 63 0 60 23.42 0.91 1.99 HjPO, h,so4 20.19 11 67 3 16 55.18 4.37 4.99 20 38 11.35 2.73 55.75 3.69 4.58 NaOH 9.25 6.97 1.22 0 99 1.35 1.80 Per Cent Loss in Weight, Room Temperature 26C. for 528 Hours 25% Acid or Caustic HC1 Actinolite..................... Anthophylite............... Chrysotile..................... Crocidolite................... Tremolite..................... 22 55 12.00 2 13 56.00 3 14 4.22 CHaCOOH 12 14 3 08 1.04 24 04 1.02 1.41 HaPO, h,,so4 20 10 11 83 3 29 56 45 3.91 4.89 20 60 11.71 2.90 56.00 3.48 4.74 NaOH 9 43 6 82 1 77 1 03 1.20 1.65 at room temperature is slower than at boiling temperatures, but if the fibres are kept immersed in the sol vent for a sufficient length of time, the solubilities finally reach the values obtained at boiling tempera tures. The tests at room temperature were conducted for periods of 24 hours, 192 hours, 360 hours, and 528 hours. After each time period, the fibres were removed, washed, and weighed, and the solubility cal culated. The fibres were then re placed in fresh solutions of acid or caustic to begin the next cycle. For convenience, only the solubil ity data for 528 hours are given in Table III since at this point most of the fibres apparently reached their maximum solubility, and this was approximately equal to the sol ubility after a two-hour treatment in boiling acids or caustic. Such information is invaluable to a manufacturer desiring to produce an asbestos product which will be exposed to conditions similar to those in the tests. The most resist ant fibre, if available in commercial quantities, naturally will be select ed for the product. Effect of Heat on Asbestos Since asbestos fibres frequently are exposed to elevated tempera tures, it is necessary to determine which fibre most satisfactorily meets the demands. With this in mind, the same fibres as used in the solubility tests were subjected to two hours' exposure at temperatures varying from 400 F. to 1,800F., and the weight loss measured. The samples, after dry ing to eliminate surface moisture, were weighed, placed in a muffle Table IV.-- Effect of Temperature on Loss in Weight of Asbestos Fibres Temp. F. 400 600 700 800 900 1,000 1,100 l)200 1,400 1,500 1,600 1)700 1,800 Time 2 hr. " " Amosite % 0 23 0 57 0.80 0 98 1 07 1 16 1 36 1 39 1.43 -- 1.52 -- 1.53 Loss in Weight Anthophyllite % Chrysotile % 0 05 0 24 0 30 0 38 0 41 0 44 0 52 0 54 0 54 0 64 1.12 1.73 2 39 0 30 0 85 1 78 2 17 2 83 3 99 10 38 12 75 13 43 13.62 -- 13.77 Crocidolite % 0 08 0 25 0 49 0 73 0 83 0 86 1 00 1 04 1 03 -- 0 93* -- 0.77* Tremolite % 0 04 0 08 0 13 0 22 0 26 0 29 0.37 0 37 0 47 0.56 0.67 1 40 2 18 Iron changing in weight due to oxidation. -- 17 MTC 000669 where temperature was maintained automatically, removed after a twohour exposure, cooled to room tem perature in a desiccator, and re weighed. The percentages of loss in weight are presented in Table IV. Generally speaking, the amphibole fibres such as anthophyllite and tremolite resisted the heat fair ly well at temperatures below 700 F. The chrysotile fibre, on the other hand, began to show a sudden ly increased weight loss at 700F., and a rapid increase above 1,000F. At higher temperatures, all fibres began to change in physical charac teristics, but the amphiboles did not disintegrate so rapidly as the chry sotile fibres. The crocidolite fibre at 1,600F. and higher showed a de finite oxidation of iron and became brittle. Chemical analysis of chrysotile fi bres showed that the ignition loss values varied with the mineral im purity present. Therefore, it was decided to determine the nature of these impurities and their effect on ignition loss values. A series of chrysotile asbestos fi bres from Arizona, Australia, Rho desia, Russia, and Canada were se lected. These fibres were individual ly tested by placing a sample in an ultimate analysis train, with heat ing units electrically controlled. The air taken into the combustion chamber was dried by passing it through sulphuric acid and absorb ing the water and carbon dioxide with magnesium perchlorate, Mg(C104)2, and ascarite. Tlie temperature of the combus tion unit was accurately determined by a chromelalumel thermocouple connected on one end to a potentio meter, and on the other to the side of the asbestos sample in the com bustion tube. Any gases, such as water and car bon dioxide, driven off the asbestos sample were weighed afteir being re covered in a second absorbing train containing Mg(C104)..: and ascarite in two different U tubes. Prior to the test, the combustion train was purged for half an hour at 220F. with dry air. Then the temperature was raised by 1,000F. and the absorbed gases weighed. Later, the test was repeated by rais ing the temperature to 1,800F. and weighing the absorbed gases. The data shown in Table V give the percentages of water and car bon dioxide driven off each fibre at the two temperatures. The Arizona fibre had a high ig nition loss at 1,800F. This loss consisted of 12.89 per cent water and 1.91 per cent carbon dioxide. At 1,000F., 0.54 per cent carbon dioxide was expelled from the sam ple. Calculated as MgC03, this in dicates the presence of 1.03 per cent MgCO,. As the temperature was raised to 1,800F., additional quantities of carbon dioxide were driven off, which, assumed as hav ing been combined with CaO, would represent 3.11 per cent CaCOs, in dicating that the Arizona fibre test ed contained a high percentage of calcium carbonate. These calculations of CO? as rep resenting MgC03 and CaCOa, resspectively, are based upon the disso ciation temperatures of these com pounds. Magnesium carbonate dis sociates at temperatures below 1,000F., and calcium carbonate at a temperature of approximately 1,630F.; a temperature of 1,800F. for one hour, therefore, should be sufficient to drive off all the CO. The Australian fibre contained 12.48 per cent H20 and 0.96 per cent CO_.; loss in weight on igni tion was 13.33 per cent. Recalcula tion of the carbon dioxide as car bonates indicated that this fibre con tained 1.26 per cent MgCOn and 0.68 per cent CaC03 as impurities. The Canadian fibre identified as "A" shows a water content of 14.28 per cent, which is higher than the theoretical amount of water in the asbestos molecule. X-ray diffraction patterns of this fibre show the pres ence of brucite, Mg(OH)s, which would break down and produce wa ter during the heating period, ac counting for the high water content of the fibre. The magnesium car bonate is calculated at 1.62 per cent and calcium carbonate at 0.27 per cent. The Rhodesian fibre contained a much higher percentage of magne sium carbonate than any of the other fibres tested -- 4.85 per cent. The calculated amount of calcium car bonate is 0.79 per cent. The Canadian fibre identified as "B" resembles the "A" sample in having a water content higher than that of pure serpentine, and here again the X-ray diffraction pattern showed the presence of admixed brucite. The magnesium carbonate content was calculated to be 0.86 per cent, and calcium carbonate, 0.07 per cent. The Russian fibre is a harsh chrysotile having a low water con tent (11.74 per cent at 1.800F.). The presence of 1.47 per cent mag- Table V.-- Combustion Analysis of Chrysotile Asbestos Source of Fibre Temperature F. Arizona............................. Australian ....................... Canadian (A).................... Rhodesian.......................... Canadian (B).................. Russian.............................. H Canadian (C).................... 1,000 1,800 1,000 1,800 1,000 1,800 1,000 1,800 1,000 1,800 1,000 1,800 1,000 1,800 Per Cent H20 Per Cent C02 Total Per Cent H2O-PCO2 3.37 12 89 1 50 12.48 5.47 14.28 1.82 12.00 3.75 13 31 1.98 11 74 2.25 11 93 0 54 1.91 0 66 0 96 0 85 0.97 2.54 2.89 0.45 0.48 0.77 1 06 0.47 0.51 3.91 14.80 2.16 13 44 6.32 15.25 4.36 14.89 4.20 13.79 2.75 12.80 2 72 12 44 Per Cent Ignition Loss OF Fibre 3 87 14.80 2 06 13.33 6.23 15.17 4.28 14.84 4.14 13.75 2.64 12.72 2.66 12 40 Carbon Dioxide Calculati:d in Terms of PerceNTAGE OF MgC03 CaC03 1.03 1.03 3 11 1 26 1 26 0.68 1.62 1.62 0.27 4.85 4.85 0.79 0.86 0.86 0.07 1.47 1.47 0.66 0 90 0.90 0.09 -- 18 -- MTC 000670 nesium carbonate and 0.66 per cent calcium carbonate was indicated. Canadian fibre identified as "C" is a semi-harsh fibre with a low wa ter content (11.93 per cent). The total amount of CO2 driven off is small and is calculated as represent ing 0.90 per cent MgCOa and 0.09 per cent CaC03. From these data it can be seen how readily ordinary ignition-loss figures can be misunderstood if the presence of mineral impurity is not taken into account. For example, most ignition-loss tests on chryso tile asbestos are made to determine molecular water content. If the fibre contains impurities such as brucite, magnesium carbonate, or calcium carbonate, an error is introduced in to the calculations and the fibre is shown to have an ignition loss great er than the theoretical quantity of molecular water. This could be a serious error in determining the as bestos content of an asbestos tex tile, from which the organic fibre is burned purposely and a correction factor is applied for the molecular water content of the asbestos. Effect of Heat on Tensile Strength A piece of Canadian crude was set aside and a number of small fibre bundles, approximating 20 to 30 microns in cross-section, were se lected and heat-treated at different temperatures in an automatic con trolled muffle. The original fibre bundles had a tensile strength of 131,000 lb./sq. in. Four sets of fibre bundles were heated for 3-minute periods at tem peratures of 600F. , 800F., 1,000F., and 1,200F. The effects on the tensile strengths of the fi bres are shown in Table VI. At temperatures higher than 1,200F., the fibre bundles were too brittle to handle and were con sidered rather weak. The same fibre, when heated for one hour at the temperature stated, had a tensile strength as follows: 400F., 129,000 lb./sq. in.; 600F., 100,000 lb./sq. in.; 1,200F., less than 2,000 lb./sq. in. These data show that tempera tures and times of exposure affect the tensile strength of chrysotile fibres, especially at 1,000F. and higher, where the strength begins to decrease rapidly. Tensile Strengths of Various Materials Reference has been made on sev eral occasions to the fact that as Table VI.-- Effect of Heat on Tensile Strength of Canadian Chrysotile Crude Original crude--No heat.............. Heated 3 min. at 600F.............. .................... 800F............... " .............. 1.000F........... .................... 1,200F........... Tensile Strength (lb./sq. in.) 131,000 120,000 96,000 78,000 42,000 Per Cent of Original Tensile Strength _ 91.6 73 3 59.5 32.0 bestos fibres are strong. As an in teresting comparison, data have been selected showing the approximate tensile strengths of such materials as iron, steel, cotton, rock wool, glass, and several varieties of as bestos. Table VII. -- Comparison of Tensile Strengths of Various Materials Type of Material Tensile Strength (Lb./Sq. In.) Ingot iron................... 45,000 Wrought iron............. 48,000 Carbon steel............... 155.000 Ni-Cr steel................. 243,000 Piano steel wire......... 300 000 Cotton fibre............... 73,000 to 89,000 dividual fibrils not visible to the eye. This would indicate that asbestos should have a great surface area when fully opened. Table VIII.--Comparison of the Surface Area of Various Fibres Type of Fibre Surface Area by N2 Adsorption (Sq. Cm./Gram) Nylon.......................... 3,100 Acetate rayon............ 3,800 Cotton......................... 7,200 Silk............................... 7,600 Wool............................ 9,600 Viscose rayon............. 9,800 Asbestos (Chrysotile) 130,000 to 220.000 Rock wool.................. 60,000 Glass fibre.................. 100,000 to 200,000 Chrysotile asbestos .. 80,000 to 100,000 Crocidoliteasbestos.. 100,000 to 300,000 Amosite asbestos.... 16,000 to 90,000 Tremolite asbestos... 1,000 to 8,000 The data in Table VIII show the surface areas of nylon, rayon, cot ton, silk, wool, and chrysotile asbes tos. The values for asbestos are so large in comparison with the other fibres tested that it seems hardly possible that they could have such great surface area. The values given in Table VII are approximate and are cited sole ly for the purpose of illustrating the comparative strengths of some materials used in industry. It will be noted that chrysotile and crocidolite asbestos fibres are exception ally strong and are comparable in that respect to glass and to souie grades of steel. The tensile strength values for the asbestos fibres are based upon breaking loads applied to fibre bun dles measuring approximately 20 microns in cross-section. Comparison of Surface Area of Fibres It is a recognized fact that as bestos fibres appear as bundles con sisting of many thousands of in The large surface area value of asbestos is a very important factor. It gives wide coverage and at the same time furnishes strength to as bestos products. Comparsion of Fibre Diameters The approximate diameter of fi bres of various types has been measured and recalculated in terms of the number of fibres or fibrils required to equal one linear inch. The results are interesting and thev show that a very large number of as bestos fibrils is required to make a linear inch. Table IX gives approxi mate data for fibres of hair, ramie, wool, cotton, rayon, nylon, glass, rock wool, and asbestos. The ex tremely small cross-section of asbes tos fibres as compared with others will be noted. -- 19 -- MTC 000671 Table IX.-- Comparison of Approximate Fibre Diameters Type of Fibre Fibre Diameter in Inches Human hair................................ 0 00158 Ramie........................................... 0.000985 Wool............................................. 0 0008 to 0 0011 Cotton.......................................... 0.0004 Rayons....................................... 0.0003 Nylon........................................... 0.0003 Glass............................................. 0.00026 Rock wool................................... 0.000142 to 0.000284 Asbestos (Chrysotile)................ 0 000000706 to 0 00000118 Fibrils in One Linear Inch I, 630 1,015 900 to 1,250 2.500 3,300 3,300 3,840 3,520 to 7,040 850,000 to 1,400,000 method of investigation is of con siderable interest. On a number of occasions we have had the opportunity of examining asbestos fibre under the electron microscope through the courtesy and co-operation of Dr. James Hillier*. The several varieties of fibres stud ied and discussed in this paper were submitted to the RCA Laboratory to be photographed at two different magnifications to show the struc ture of the fibrils. Chrysotile The soft Arizona chrysotile at a 4,000 magnification shows fibre bundles which appear wavy with frayed ends (see Figure 2)t. The same fibre at a magnification of 35,000 shows bundles in which Harsh type Semi-harsh type Figure 1.--Photomicrograph* of chryaotile asbeltoi. x 100. Soft type Photomicrographs of Chrysotile Three samples of chrysotile were selected, representing three differ ent degrees of texture. Eacli of these three fibres, mounted on a cover glass, was placed under a microscope and photographed at 100 magnifica tions. They are shown in Figure 1. The harsh chrysotile fibre at this magnification resembles the struc ture of splinters and has a needle like structure. The fibres are springy, bulky, fast-filtering, and usually not so high in tensile strength as soft, silky fibres. This type of fibre will break quickly un der flexing action. The semi-harsh fibre appears par tially as a harsh, and partially as a soft, fibre. The fibres have lost most of the needle-like structure and ap pear as thick, wavy bundles. It is difficult to show in a picture that they are different from soft fibres. Chemically, they are the same as the latter, but the measurements show quite different physical properties. Such physical tests as filtration, willowing. surface area, bulk, and strength will bring out the true properties of this fibre and show how it differs from the soft, silky chrysotile. The soft fibres appear as thin threads which are very soft, not bulky, slow-filtering, and extremely strong. As noted above, they have the same chemical composition as the semi-harsh fibres, but quite dif ferent physical characters. The soft, silky chrysotile fibres make up the bulk of asbestos used in industry. They are available in large quantity and usually are cheaper than the semi-harsh or harsh fibres. Electron Micrographs of Asbestos The electron microscope is a val uable tool for examining materials which are difficult to see by ordin ary microscopic means. Since asbes tos exists in thin cross-sections, this the fibres appear as hollow tubes (Figure 3), This feature was first reported by Dr. Hillier in April, 1949 (6), using a magnification of 94,000, and it has since been re ported by Bates. Sand, and Mink (7), of Pennsylvania State College. Our measurement of the thin fibrils indicated a cross-section in the range of 214 and 284 At units, while Dr. Hillier reported fibril diameters of 180 A units. This would account for the - fact that soft chrysotile fibres have large surface areas and show a preference to open length wise if properly fiberized. Amosite The amosite from South Africa, at a magnification of 4.000, shows RCA Laboratories, Princeton, N.J. Jin this and ail succeeding Fig ures showing electron micrographs, the magnification has been reduced from that stated in the text. The cap tion below each Figure gives the magnification as thus reduced. fAngstrom. -- 20 -- MTC 000672 straight bundles criss-crossing the field (Figure 4). The same fibre at 35,000 magni fication shows rather thick bundles that still remain straight and indi cate a plane of cleavage that is clean-cut and abrupt (Figure 5). Whenever a small fibril is split from a small bundle, its cross-sec tion appears to be approximately 300 A units. However, only a few small bundles appear in this pic ture, while chrvsotile fibre, on the other hand, apparently breaks down lengthwise easily and shows numer ous small fibrils within the range of 300 A units and less. The fact that amosite fibre tends to remain as large, thick bundles may well account for its freeness in filtration, small surface area, and bulky appearance. Antho phyllit r At a 4.000 magnification, anthoplivllite fibres appear as a mixture of thick and thin bundles which show clean, even breaks at the ends (Figure 6). At a 35.000 magnification, some fibrils appear thin and flexible while others are stubby, thick bundles (Figure 7). The cross-section of the thin flexible fibrils is equal to, and possibly smaller than, that of chrvsotile fibrils, placing them at ap proximately 200 A units. It is to be noted that no hollow-tube structure appears within these thin fibrils. This particular sample of anthophvllite is soft, not strong, and breaks easily during processing. The mineral impurities in the fibre are visible as plates in the electron micrographs. This fibre is found in Georgia as a mass fibre, and considerable non-fibrous material is associated with the recovered fibre. Crocidolite The crocidilite used in this in vestigation was the typical blue fi bre from South Africa. Figure 4.--Electron micrograph, amosite asbestos, x 1,733. ; 4 i Figure 5.--Electron micrograph, amosite asbestos, x 15,200. -- 21 -- MTC 000673 Figure 6.--Electron micrograph, anthophyllite asbestos, x 1,733. Figure 7.--Electron micrograph, anthophyllite asbestos, x 15,200. At a 4,000 magnification, the fi bre appears as straight, stiff, brit tle bundles, and in most cases the ends show clean breaks (Figure 8). At a 35,000 magnification, the bundles appear fairly thick, but there is evidence that slender fibrils of approximately 300 A units exist (Figure 9), though they are few and far apart. The brittle nature of the fibre probably prevents individ ual fibrils from splitting off length wise, and consequently all that is seen is the large bundles. Bolivian Blue Bolivian blue asbestos is a type of crocidolite fibre which differs somewhat from normal African cro cidolite in chemical composition. It is a soft, silky fibre, not very strong and easily pulverized. At a 4,000 magnification the fi bre bundles are clean, tightly bound, and show evidence of brit tleness (Figure 10). At a 35,000 magnification, the bundles still appear clean and tight ly held together (Figure 11). An examination of the few bundles in the field indicates that, if the bun dles are properly opened, they will show fibrils of approximately 300 A units. Tremolite The tremolite under investigation is from California and is considered of good quality. At a 4,000 magnification, the cross-section of the bundles varies from thick to thin (Figure 12). The bundles do not appear to be held tightly together and they show a tendency to fray out. Examination of the bundles at 35,000 magnification shows evidence of individual fibrils existing at ap proximately 300 A units (Figure 13). Figure 8.--Electron micrograph, crocidolite asbestoi. x 1,733. Figure 9.--Electron micrograph, crocidolite aebettoe. x 15,200. -- 22 -- MTC 000674 Figure 10.--Electron micrograph, Bolivian blue asbeetoa. * 1,733. Figure 11.--Electron micrograph, Bolivian blue aibestoa. x 13,200. Actinolile The sample of actinolite was ob tained from Canada. Previous tests showed high solubility, indicating the presence of impurities. The electron micrograph at 4,000 magnification verifies this fact as it shows only a few fibre bundles, with a high percentage of plate-like minerals present (Figure 14). Sampling this grade of actinolite for electron microscope studies is a considerable problem and many ex aminations would be required to ob tain a satisfactory field. However, it was recognized that this sample was a milled commercial product and the pictures do show the true nature of the material. At a 33,000 magnification, only cne large, thick bundle appears in the picture. There is a suggestion of breaks along the end, but there is no indication as to the size of the individual fibril (Figure 15). Conclusion The various properties of the six varieties of asbestos investigated are given in convenient table form to serve as a quick reference. Solubility data of the six varie ties of asbestos show that chrvsntile asbestos is readily attacked by acids and that the amphibole fibres are fairly acid-resistant. The effect of heat on the differ ent varieties of asbestos varies. Chrysotile shows small ignition losses at temperatures below 700 F. and a rapid increase in loss at 1,000F. and above. The amphiboles show small ignition losses since they contain only small percentages of water of crystallization. Ignition losses also indicate the presence of water and carbon di oxide. Some chrysotile fibres show the presence of brucite. magnesium F'gurei. 12.--Electron micrograph, tremolite aebeatoa. X 1,733. Figure 13.--Elecrton micrograph, tremolite aibeitoi. x 15,200. -- 23 -- MTC 000675 f. I Figure 14.--Electron micrograph, actinolite aibcstos. * 1,733. Figure 15.--Electron micrograph, actinolite aebcatoe. x 15,200. carbonate, and calcium carbonate; therefore, great care should be taken in interpreting ignition-loss data. High temperatures and lengths of exposure decrease the tensile strength of chrysotile fibres. At temperatures above 1,000F., the strength decreases rapidly. Charts are presented to show that asbestos fibres have great tensile strength and large surface areas. Photomicrographs at 100 magnifi cation show that the soft, silky, chrysotile fibre is wavy, whereas the harsh fibre has a straight or needle-like structure. The electron micrographs of the six varieties of asbestos at 4,000 and 33,000 magnifications show the fibre bundles, individual fibrils, and their approximate cross-sec tions, and may help to explain some of the physical properties of asbes tos fibres used in commercial prod ucts. Acknowledgments The author wishes to thank Dr. James Hillier and the RCA Labor atory for their help in producing the electron micrographs of the six varieties of asbestos described in this paper. Reference; 1. Cummins, A. B., unpublished pa per. r., ana KERR, P. F Optical Mineralogy, 1942, pp. 277' 282, 289, and 363. 3. Casey, K. S., Materials of Con struction; Indust, and Eng. Chem., Vol. 40, Oct., 1948, p. 1793, also pp. 1837-1860. 4. No Author, Acid Resistance of Asbestos; Gummi Zeitung, May, 1927, p. 1861. 5. No AUTHOR, Acid Resistance of Spinnable Types of Asbestos; As bestos, 1931, pp. 22-24. 6. Turkevich, J., and Hillier, J., Electron Microscopy of Colloidal Systems; Anal. Chem., Vol. 21, Ap ril, 1949; pp. 480-481. 7. Bates, T. F., Sand, L. B., and Mink, J. F., Tubular Crystals of Chrysotile Asbestos; Science, May 12, 1950, p. 512. 1. -- 24 -- MTC 000676 Asbestos Floats By M. S. BADOLLET* i (Ottaxca General Meetiny, January, 19.52) (Transactions, Volume LV, 1952, pp. 185-189) Introduction sbestos floats may be A defined briefly as blends of airborne particles of fibrous asbes tos and dusts produced during the milling stages and collected by Cot trell precipitators, dust sheds, or baghouses. The various grades of asbestos floats on the market contain fi brous asbestos particles ranging in length from microscopic to approx imately 14-inch. The fines or dusts present in the floats are from 40 microns to 2 microns or less in crosssection. Airborne asbestos particles are of diverse length because they are collected at numerous places in the mill. There are: (1) dusts removed from dry rock storage bins, (2) dusts removed from rock and fiber screens, (3) particles from the tops of collectors, and (4) fines removed by suction from the processing equipment. Sometimes the airborne particles are kept segregated, so that there is a partial separation by gravity. This is accomplished by a series of dust chambers enclosed in a long, rectangular building. The longest and heaviest particles fall nearest to the air flow entrance, the small- Research Center, Johns-Manville Corporation, Manville, New Jersey, U.S.A. est at a point farthest away and nearest the exit. Sometimes the air from the dust chamber building is passed through a series of baghouses or electric precipitators to collect the extremely fine particles which normally escape to the at mosphere. In other instances, the airborne particles by-pass the dust chambers and are passed through rows of bag collectors. Regardless of the method of col lecting the particles, the milling de partment must select proper fiber sizes for blending, producing uni form floats to meet the requirements of the trade. Sometimes one grade of float can be used successfully in many com mercial products; in other cases, great care must be taken to pro duce a float for use in a specific product. The present total production of all grades of Canadian floats ap proximates 40,000 tons per year and it is possible that this quantity may be greatly increased by the proposed expansion programmes now going into effect in Canada. Since floats are valued at $40 per ton at the Canadian mines, a tonnage of 40,000 has a potential sales value of $1,600,000. This di vision of the asbestos industry is a good business, and it is well worth the concentrated effort of research and development engineers to find new uses, and so expand the mark et. for all available floats. Physical Properties ok Floats According to the Quebec Screen classification, there are two general classes of floats, known as 7RF and 7TF. Samples of various commercial floats were obtained on the market and a series of tests were devised to determine the differences in their physical properties. Some of these tests were similar to those made by many manufacturers who use floats in their products. The results of some of these tests are given in Table I. They do not, however, in clude the official Quebec Screen test. Manufacturers using floats usually set up their own specifica tions, based upon requirements for their particular products. Loose Density The loose density of floats can be measured by slowly sifting the pow der into a can or container of defin ite dimensions so that the weight per cubic foot can be calculated. Although in some products the loose density of the floats is not par ticularly important, in others it is critical, so each float must be considered as an individual case. In the present investigation eleven 7RF floats were tested and their loose densities were found to range Table I Identification Loose Density Ib./cu. ft. (1) 7RF.... (2) 7RF.. .. (3) 7RF.... (4) 7RF.... (5) 7RF.... (6) 7RF.... (7) 7RF___ (8) 7RF___ (9) 7RF___ (10) 7RF___ (H) 7RF.... (12) 7TF. ... (13) 7TF.... (14) 7TF. .. 19 20 21 22 22 24 24 20 25 23 14 23 23 26 Material + 100 M % 3 4 4 4 3 5 5 6 7 7 1 2 2 3 Fines -200 M % 96 93 93 93 95 90 89 91 84 88 96 96 95 91 Grit +48 M % 0 04 0 14 0 24 0 16 0 12 Tr 0.06 0 12 0 24 0.46 0.04 Tr 0.06 0.00 Whiskers + 48 M % 1.90 2 50 2 26 1.30 0 80 1.16 1.28 4.12 0 70 2 38 0.50 0.30 0.60 0 10 Grit -48 M +80 M % 0 02 0 14 0 12 0 20 0 10 0.04 0.10 0.56 0.46 0.54 0.02 Tr 0.06 Tr Whiskers -48 M +80 M % Oil Adsorption c.c./g. fiber 0 64 0 66 1 18 0 94 0 46 0.96 1.12 1.10 1.30 0.92 0 84 0 66 0 74 0 75 0 68 0 65 0 68 0 71 0 66 0.65 0.65 1.05 0.50 0.70 0 16 0 75 0 65 0 68 -- 25 -- MTC 000677 from 14 to 25 pounds per cu. ft. The heaviest is No. 9 and the light est No. 1 1. Nos. 1 to 5. inclusive, and No. 8, all with loose densities between 19 and 22 pounds per cubic foot, may be considered as falling within the same rang?, as also may Nos. 6. 7, 9. and 10. which are slightly heavier (28 to 25). Float No. II is in a class bv itself and is light in weight as well as bulky (loose density It). If density were the only physical property that had to be considered, this record would enable the manufacturer to select the proper one for his purpose. The three TF floats included in flic Table differ from the RF floats in degree of particle fineness. Nos. 12 and 18 have identical loose den sity (25) while No. It is heavier (26). Classification bv Wet Washing Fibre and Grit + 100 Mesh A classification of the particles can be obtained by a wet-washing method which determines the amounts of fihre and grit plus-100mesh and of fines minus-200-mrsh. This information is particularly use ful if the floats are to be used in molding, extrusion, or spraying. The RF float No. 11 contains enlv I per cent of material larger than 100 inesh. In the other RF floats, the content ranges from 3 to 7 per cent, indicating the presence of material that is sometimes re ferred to as 'whiskers'. In the plas tics industry, particles plus 100 mesh may be helpful in producing impact strength in the product. The three TF floats contain al most identical amounts of plus-100mesh material. In two of them the percentage is lower than in the first ten RF floats. Fines --200 Mesh The fines minus-200-nu'sh in the RF floats range from 81 to 90 p-r cent, and in the TF floats from 91 to 9G per cent. The material between plus 100 mesh (liber and grit) and minus 200 mesh (fines) would he collected oil the 200 mesh screen. Its amount can be obtained by adding the plus 100 mesh and minus 200 mesh lalues shown in the Table and sub tracting the sum from 100. This test, although useful, does not distinguish between, or give the relative amounts of, `whiskers' (as bestos fibers) and particles of grit that might be present in the floats. Grit +48 Mesh In many cases, the grit particles in the floats are of very diverse sine. Usually, they are small pieces oi serpentine rock and magnetite, or small bundles of asbestos 1 fiber. Complete classification, therefore, requires a further breakdown to de termine the quantities of these ob jectionable particles present in the floats. This test is made by water elutriation and by careful separation ot the grit from 'whiskers' or fiber bundles. The amount of grit plus-48-mesh in the eleven 7RF floats ranges irom a 'trace' to 0.46 per cent. The percentage is lowest in samples Nos. 1, 0, and 11 and highest in No. 10. Although these quantities of grit seem small, they could cause considerable trouble in a manufac turing process by plugging extrusion apparatus or scoring the walls of expensive molds. The three TF floats are almost free from grit plus-48-mesh. `li'hiskers' +48 Mesh 'Whiskers' are short asbestos fi bers that may be valuable or ob jectionable, depending upon the use for which the product is required. The quantity of whiskers in the eleven RF floats ranges from 0.50 per cent (in No. 11) to 4.12 per cent (in No. 8). In the TF floats, the range is from 0.10 per cent (No. 14) to 0.60 per cent (No. 13). Grit --48 Mesh +80 Mesh In some special products, the presence of grit smaller than 48 mesh, and even smaller than plus 80 mesh, is considered objection able. Therefore, a further separation is made to determine the quantities of this grit that may be present in these floats. As shown in Table I. the range lor the eleven RF floats tested is from 0.02 per cent (Nos. 1 and 11) to 0.36 per cent (No. 8). Two of the TF tloats are free from this size of grit and the third (No. 18) co.i.ai.u only 0.06 per tent. `Il'hishers' --48 Mesh +80 Mesh The presence of whiskers in the sizes of minus 18 mesh and plus 80 mesh is also objectionable because they affect the surface of certain products. The eleven RF floats tested con tain whiskers of this size ranging from 0.46 per cent (No. 5) to 1.30 . ,-- -/ '-'"responding figures for the TF floats are a low of 0.16 per cent (No. 14) and a high of 0.70 per cent (No. 13). In the manufacture of plastics, the presence of whiskers results in non-uniform pourability of the molding powders. This necessitates a re-screening operation so that uniormlv coated powders will prop erly fill the melds. In many cases, the whiskers and crudv fiber bun dles are not thoroughly covered by the organic plastic material and this increases the water absorption properties of the plastic material, which may result in the formation of blisters. Oil-Adsorptive Capacity The capacity of a float to adsorb oil is important because it gives an indication of ability to hold an or ganic liquid when manufacturing a plastic or caulking compound. This test, made by manufactur ers, consists of titrating a given weight of float with oil until its adsorptive capacity is reached. In practice, the adsorptive capacity is considered to have been reached when oil can be released from a ball of the thoroughly mixed float and cil by applying a small amount of pressure against it. The results of this test are expressed in terms of cubic centimeters of oil per gram of float. For the eleven RF floats tested, the oil adsorptive capacity ranges from 0.65 to 1.05 cubic centimeters per gram. The value for No. 11 (1.05 c.c./gm.) is much higher than that for any of the others, which all fall between 0.65 and 0.68 c.c./gm. except Nos. 2, 3, and 7 (0.71 to 0.75 c.c./gm.). Two of the TF floats. Nos. 13 and 14, have closely similar adsorp tive capacity, 0.65 and 0.68 c.c./gm.. respectively. No. 12 gave a higher value (0.75 c.c./gm). Miscellaneous Physical Tests Numerous other tests, of import ance to the manufacturer of certain asbestos products, are made on floats. Data for these are given in Table II and are summarized in the following paragraphs. Surface Area Study of the surface area of each float was made to obtain data on its covering power as a filler. In this test, the surface area is measured by the air permeability method by means of the Bowen apparatus, fol- -26- MTC 000678 Table It Iden* TIPICA- Surface TION Area Density cm* /rm. ',000 p.s.i. :.000 p.s.i. (1) 7RF 14.800 89.9 94.7 (2) 7RF 11.300 93.6 97.0 (3) 7RF 13.000 93.5 99.2 (4) 7RF 11.800 95.5 99.1 (5) 7RF 13.000 95.0 98.2 (6) 7RF (7) 7RF (8) 7RF (9) 7RF flO> 7FF ill) 7RF 14,000 13,900 12.500 7.500 8,600 20.800 100.0 100.8 102.7 99.6 101.2 93.6 105.4 105.2 107.7 106.8 105.8 99.2 (12) 7TF 15.400 95.0 100.0 (i3> TTr 11.300 94.8 99.4 (14) 7TF 15.600 96.0 99.5 Compressibility* % Thickness at 200 Compress. 3,000 p.s.i 29 32 28 32 28 32 29 32 28 31 28 32 28 32 28 32 27 31 28 31 35 T9 27 31 27 ?n 27 9 Springback. % Thickness at Max. Compress. 7.000 p.s.i. 2,000 p.s.i. 2.5 .8 3.1 3.7 :.2 3.3 3.3 3.4 2.6 3.4 3.8 4.0 3 4 4.0 3 6 4.2 3.4 39 3 0 3.4 5.8 6.7 3.3 3.7 ?.5 3.3 3.3 ?3 5uFr. Coi^d. Cf PREFSiD Specimen 5:m'l whisMcrs & bundles Grit & few whis. & bund. Sm'l ami. of < rit & * bs. Coed, srrcolh Grit rrostly Good surface Good surface F ?r{e j nt Good surface fONIZAFLE Salts. Micro* rrho*/ cm / 1U tolls 219 218 222 2C6 212 379 3^3 '21 190 243 ?eo 221 176 3S6 Dry --W a--ter-Fo--lihle Bulk. Chlo- Mag- VOL. OP Solids ioo g. % FINE NETJC as Cl Rating CC. % 350 0.16 0.017 4.7 335 340 0.17 0.024 0.18 0.029 6.3 6.1 3C5 0.16 0.025 6.0 300 0.17 0.022 5.7 320 0.36 0.15 4.1 295 T60 038 0.F6 0.26 O 15 5.1 5.1 2F5 0.12 0 0049 f O 285 0.14 0 0038 96 475 0.15 0 0016 3 4 0 15 no?2 5n ?ro 0.14 V 034 j.n 75 o.;6 .16 T9 *Mea8urements start at 200 p.s.i. as zero point. lowing the Lea and Nurse tech nique (1). The eleven RF floats tested show a wide range in surface area, from 7,500 cm2/gm. (No. 9) to 20,800 cm-/gnt- (No. 11). In other words, float No. 9 has the lowest, and float No. 11 has the greatest, covering power as a filler. In plastics, the surface area of the float governs the amount of filler that can be used successfully prior to pressing in a given mold. Two of the three TF floats (Nos. 12 and 14)' have the same surface area, approximately 15,500 cm2/gm. For No. 13 the value is 11,300 cm2/gm. The high value of 15,500 for Nos. 12 and 14 indicates the presence in these floats of a large percentage of fines. Density at 2,000 and 3,000 p.s.i. Ten-gram samples of each float were subjected to pressures of (1) 2,000 p.s.i. and (2) 3,000 p.s.i. and their densities calculated in pounds per cubic foot. These pressures were selected because they correspond to the range of pressures used in the plastic-molding industry, and the data obtained are helpful in select ing the proper asbestos filler for molding powders. At 2,000 p.s.i., the densities range from 89.9 lb./cu. ft. (No. 1) to 102.7 lb./cu. ft. (No. 8). Several of the samples have substantially identical density, but for the group as a whole the differences in density are such as would influence the selection of the filler to be used for certain products. At 3,000 p.s.i., the densities range from 94.7 lb./cu. ft. (No. 1) to 107.7 lb./cu. ft. (No. 8). The in crease in density for the additional (1) For references, see end of pa per. 1.000 p.s.i. pressure ranges from 3.2 to 6.2 lb./cu. ft. Compressibility of Floats at 2,000 and 3,000 p.s.i. In this test, a 10-gram sample of the float is compressed at 200 p.s.i. and its thickness measured. It is then compressed at (1) 2,000 p.s.i. and (2) 3,000 p.s.i. and its thickness again measured. The re duction in thickness, expressed as a percentage, gives a measure of the compressibility for each float. Thirteen of the fourteen floats test ed show about the same percentage of compressibility (27 to 29). For No. 11, the value (35) is appreci ably higher. This is an important property and must be taken into consideration when pressing a molding powder to a definite size after the molding cavity has been properly filled. Uniform compressibility is impor tant as a bulky fiber will also ex hibit poor dimensions after press ing. Springback After Release of Pressures The amount of 'springback' of a molding powder is another im portant property considered by plastics manufacturers. If the springback is too great, it will be difficult to obtain the proper di mensions for a molded product. Too mucli emphasis cannot he placed upon this property. All fourteen floats were tested for springback at pressures of botli 2.000 and 3,000 p.s.i. As will be noted from the data in Table II, the springback factor is uniformly low for all of them except No. 11 which, also, is the sample with the highest percentage of compressibil ity. To use float No. II, with a springback of 5.8 and 6.7 per cent, allowance must be made for its com pressibility or the plastics manu facturer cannot control the di mensions of his product. Surface Condition of Pressed Blocks The surface characteristics of pressed blocks of the floats are not ed in Table II. The surfaces usual ly show long 'whiskers', short stub by fiber bundles, and grit, which generally are considered poor prop erties from the viewpoints of ap pearance and water adsorption. A smooth surface is desirable, provided the molded product meets all other specifications of the manu facturer. Five of the floats give good smooth surfaces and prohably will be satisfactory for use in mold ing powders, provided other proper ties are acceptable. Ionization Salts For the manufacture of some spe cial types of molding powders, it is essential that the ashestos fibre have low electrical conductivity. In this case, some plastics companies make a test on water-soluble material leached from floats. Apparently, all floats contain a certain quantity of water-soluble material. Its conduc tivity is determined, and the re sult of the test is expressed in terms of micromhos per centimeter per 10 volts. For the eleven RF floats test ed. this factor ranges from 190 (No. 9) to 421 (No. 8). For the TF floats. Nos. 12, 13, and 14, the values are, respectively, 221, 176, and 396. It is not known at present which values are objectionable, but it ap pears that a conductance of 300 mi cromhos or higher per centimeter -- 27 -- MTC 000679 per 10 volts may cause trouble, with the molded plastic product having a low dielectric strength. Dry Bulk The bulking values of floats are sometimes considered important fac tors in plastics as well as in drywall joint fillers. By some manufacturers, the bulk ing value is expressed as the vol ume. in cubic centimeters, of a given weight, say 100 grams, of the float. Some fillers should have high drybulk value while others should not; this depends entirely upon the in tended application. The eleven RF floats have drybulk values ranging from a low of 255 (No. 9) to a high of 475 (No. 11). For the three TF floats, Nos. 12, 13, and 14, the values are, re spectively, 295, 290, and 275--dif ferences that may lie within experi mental error. Water-Soluble Material All floats were analyzed for wat er-soluble materials by allowing a sample of each to stand in distilled water for 63 hours and filtering off the clear water. Aliquots of the fil trates were dried and weighed to determine total solids, and these were then analyzed for chlorides. The amount of water-soluble ma terial in the eleven RF floats ranges from 0.12 per cent (No. 9) to 0.38 per cent (No. 7). In the TF floats the range is about the same, with a maximum of 0.36 per cent for No. 14. No good correlation exists be tween the ionizable salt values and the quantity of water-soluble mate rial. The percentage of chlorine in the residues also was determined. For the eleven RF floats it ranges from 0.0015 to 0.16 per cent, and for the three TF floats from 0.0034 to 0.16 per cent. Magnetic Rating Since many floats are used as fil lers in products employed by the electrical industry, it may be impor tant to learn something about their magnetic ratings. The test, an adap tation of the Mapes (2) test, meas ures the amount of magnetite pres ent. Ratings of the eleven RF floats range from 3.4 (No. 11) to 9.6 (No. 10). The other nine RF floats have ratings between 4.1 to 5.9 with an average of 5.4. The three TF floats range from 3.9 to 5.0, averaging 4.3. The maximum magnetic rating value for a successful float is not known precisely, but it is believed that the manufacturer would prefer the lowest rating available. It is known that the smallest trace of magnetite can be easily detected in a magnetic field and it may result in the rejection of the plastic ma terial. Use of Asbestos Floats Asbestos floats have a definite use in such industrial products as automobile body coatings, adhesives, caulking compounds, dry-wall joint filler, lubricating greases, asphalt and cold-water paints, plastics, molding powders, welding rods, in secticides, radiator scaling com pound, acoustical plasters, cements, and many others. Asbestos floats offer to manufac turers the advantages of being a low-cost inorganic filler, which will improve impact strength and pro vide good workability, good binding qualities, heat resistance, fair acid and alkali resistance, large surface area, and availability in commercial quantities. A recommendation for the best grade of asbestos floats for a par ticular product should be made care fully, since each product has its own requirements and specifications. Automobile Body Undercoating A float for automobile body coat ing to be applied by spray gun may require the absence of 'whiskers' and large particles of grit. If the coating is to be applied by brush, however, the presence of whiskers and grit would not be objectionable. Therefore, the selection of a float for this job depends upon the meth od of application and such other factors as bulk and fluidity when incorporated in a binder. Adhesives The use of floats in adhesives depends upon the type and appli cation of the adhesive product. The proper recommendation can be made only after knowing the desired prop erties of the adhesive and the meth od of application. The amount of floats in adhesives (3) varies be tween 9 and 40 per cent of the mix, depending upon the product. Caulking Compounds Caulking compounds may be sen sitive to changes in grade of floats. For example, a gun grade of caulk ing compound may require a float gru, wnue. tor a caulking compound applied by knife, a float containing whiskers and some grit could be used. In general, the use of floats in caulking compounds (3) may range from 4 to 15 per cent by weight of compound. Jf all Joints The manufacturer of drv-wall joint filler is very careful in his selection of floats. His specifica tions fcr asbestos fiber are rigid and he insists upon good bulking value, freedom from grit, good cov erage. and no whiskers. Very few natural floats on the market today fulfill these requirements and. therefore, a special float must be made for this job. The amount of floats (3) required for this use uiav vary between 5 and 20 per cent of the dry mix ready for the market. Various types of floats have been used in some grades of lubricating greases, but in most greases grit is objectionable. Whiskers may be permitted unless the customer's spe cification indicates otherwise. The amount of float used in greases var ies between 10 and 25 per cent of the product (3). Paints Paints, with either asphalt or cold-water base, may require a gritfree and whisker-free float if smooth surface is demanded. Other wise, some grit and whiskers are allowable. Paints may contain from 5 to 11 per cent of floats. Plastics and Molding Powders Floats, with specifications carv ing from customer to customer, are used extensively for plastics and melding powders. The general opinion of the plas tics manufacturer is that fibrous float structure increases impart strength, allows a wide range for adjusting quantity of floats as filler, furnishes good binder retention and workability, imparts hardness and toughness to the moldings, increases heat and fire resistance, reduces natural shrinkage of resins and binders, reduces warpage and de formation, makes cold-molding pos sible by controlling the flow under pressure in the molds, allows blend ing of mineral and organic fillers, imparts good finish, improves the --- 28 -- MTC 000680 electrical properties, and reduces the cost of moldings. Molded plastics are carefully checked for specific gravity, flex ural strength, tensile strength, com pressive strength, impact strength, water absorption, dielectric strength, power factor, shrinkage, heat re sistance, visual observations of molding qualities, and ejection from the mold. It is important, therefore, that the plastics manufacturer select asbestos floats with care so that molding powder and molded prod uct have the desired properties. The asbestos floats used as a filler in these materials generally form from 60 to 80 per cent of the mix (3). Welding Rods The welding rod manufacturer desires a float which will produce good coating as a filler, and which lacks whiskers and grit that might clog extrusion equipment. Floats also should be free from any sul phide mineral, or at least should contain less than the 0.01 per cent of sulphur that would produce ob noxious gas during welding. The amount of floats used for coating rods varies between 8 and 10 per cent*. Insecticides Floats sometimes are used as fill er for insecticides to be dusted or sprayed onto flowers, weeds, and other plants. In this case, the float should be a fine dust so that it will impart large coverage. Floats for 'British Patent 535,355; Canadian Patent 365,792. this purpose must compete with such other powders as talc and diatomaceous earth. The amount of float in a ready-to-use insecticide is approximately 4 to 5 per cent. i Radiator Sealing Compound Various combinations of mate rials, including fine asbestos, are used in sealing compounds for au tomobile radiators. The asbestos forms from 1 to about 6 per cent by weight of the solids content of the mixture!. Acoustical Plasters Based upon the weight of the dry compound, some acoustical plasters contain from 6 to 15 per cent of asbestos floats. In this ap plication, the floats function as a filler to provide bulk and fire re sistance. Whiskers and grit are not objectionable unless a spray gun is used. Cements Cements and adhesives are con sidered similar by some authorities, and the amount of floats needed for adhesives would apply also to cements. Depending upon the type of cement and the application, the asbestos floats content ranges from 9 to 40 per cent (3). In many cases, whiskers and grit are permitted. Conclusions (1) Fourteen asbestos floats pro duced in Canada have been exam ined in tests similar to those made fU. S. Patents 2,391,737, 2,315,321. by customers in their own labora tories. (2) Many of these floats have almost identical characteristics; however, one float, No. 11, differs considerably from all the others and has many unique properties. (3) The type of float required for a particular product depends upon the use for which the product is manufactured, and each manu facturer has his own specifications and methods for testing the float. (4) Float characteristics such as grit, whiskers, compressibility, springback, and magnetite rating are usually carefully checked by the plastics manufacturer, since these qualities affect his products. (5) Bulk, surface area, and den sity values of floats are considered important factors in such products as wall joint fillers, caulking com pounds, some pressed products, and acoustical plasters. (6) The amount of ionizable salts, water-soluble solids, and chlorides in the floats is a matter of importance and frequently is checked by plastics manufacturers interested in improving the elec trical properties of their products. References 1. Lea, F. M., and Nurse, R. W,, Specific Surface of Fine Powders; Soc. Chem. Ind., Trans., Sept., 1939, pp. 227-283. 2. A.S.T.M., Standards for Textile Materials, D118-50T, Oct., 1950, pp. 197-202. 3. Bennett, H., The Chemical Form ulary, Vol. IX, 1951. Henley, Normann, Henley's Twen tieth Century Book of Formulas, Processes, and Trade Secrets, 1947. -- 29 -- MTC 000681 Asbestos Fibers: Production and Usage By M. S. BADOLLET* (Annual General Meeting, Edmonton, April, 1933) (Transactions, Volume LVl, 1953, pp 237-289) DURING tlie past five years, grade, and quantities of rock, dust production of asbestos fibers and such foreign materials as wood has increased in all countries thaatnd wire. Comparisons are made list asbestos as one of their mineral with standard Canadian grades, or, resources. Despite the many small if the fiber is of some other variety, mines that have been opened in comparison is made with shipments Canada, United States, Mexico, of that grade received during the South America, Italy, South Africa, past five years. If the sample is Australia. Yugoslavia, and Rho only a small quantity, such as a few desia, the supply still does not meet grams, it is placed between two the demand. It is estimated that glass plates and reflected against total 1952 world production of as a white wall, where the lengths may bestos fibers of all varieties ap be measured and the impurities proximated 1,600,000 tons, with counted. When large samples are Canada suplying 966,382 tons (I), available they are given special or 60 per cent of all fiber mined. tests to evaluate the fibers for use Asbestos fiber brokers continu in regular products. These special ally are offering odd-lot samples of tests include air separation and chrvsotile, amosite, crocidolite, screening in a pilot plant to deter tremolite, and anthophyllite fibers mine impurities ranging from large to manufacturers, quoting fantastic -|-4 mesh fragments of rock to --35 prices for tonnages varying from mesh granular material. Then the just a few to 50 or more iii a clean fibers as well as the original month. In many cases, the samples fibers are combined separately with are not properly graded and con the necessary ingredients to produce tain quantities of rock impurities a commercial product. Physical tests and finely ground dust. on the product furnish further fac When vendors submit samples of tors for evaluating the fiber. asbestos fibers, the manufacturer Should a sample of fiber pass should consider them in terms of laboratory tests, it is customary to price, quality, quantity, and as pos request three to ten tons for a plant sible substitutes for fibers current test before final approval is given. ly used in his products. When the large shipment of fiber Grading of these odd-lot fibers arrives it is thoroughly sampled in most cases has been poorly done, and re-tested, and the data com with resultant non-uniform fiber pared with those of the original lengths. In some instances, a num sample. If both sets of data are ber of miscellaneous lots of fibers comparable and the plant tests are have been blended for sale. This satisfactory, fiber specifications are blending produces a fiber that is drawn up for future purchases of difficult for the manufacturer to ac this grade and variety of asbestos. cept as a substitute for the wellstandardized Canadian fibers. All fibers offered to the pur chasing departments of most manu facturers usually are forwarded to their research units, where they are In some cases, the quality of the fiber is so poor that anyone accus tomed to fiber examinations will realize that it is worthless before tests are made. carefully examined for length, The three varieties of asbestos used in large quantities industrially ''Research Center, Johns-Manville Corporation, Manville, New Jersey, U.S.A. (1) Preliminary estimate by In dustry and Merchandising Division, Dominion Bureau of Statistics, Ot tawa. are chrysotile, amosite, and crocido lite. Although actinolite occasional ly appears on the market, no pub lished records show how much is used, and we may consider it of minor importance. Chrysotile Chrysotile asbestos is the most popular variety for industrial pur poses and probably accounts for 90 per cent of all asbestos fibers used in industrial products. The quantity of chrysotile im ported by United States industry in 1952 was approximately 734,500 tons (2), 668,800 tons of which came from Canada. When available, some Rhodesian and Arizona chry sotile fibers are used in textiles, special papers, and asbestos-cement products. Last year the textile in dustry in the United States im ported 26,445 tons (2) of chryso tile fibers, including No. 1 and No. 2 crudes and Group 3 fibers. This amount includes 24,631 tons of spinning fiber from Canada, and 1,824 tons of C.&G. No. 1 and No. 2 from Rhodesia. These fibers usually are re-processed by the consumer and then curded and form ed into blankets, rovings, threads, cords, ropes, and yarns of various classifications depending on demand. Asbestos cloths, tapes, wicks, and braided tubing all are made to meet particular specifications. Fibers prepared in the textile plant (3) are used in such products as awnings, brake linings, blankets, conveyors, curtains, clothing, gloves, bag filters, draperies, coverings for electrical and power plant equip ment, flexible metal tubings, pack ings, chemical filters, thermal in sulating materials, friction mate rials, clutch facings, gaskets, elec tric blankets, ironing board covers, joint fillers, turbine jackets and blankets, and plastics. The shortage of spinning fiber has resulted in off-grade so-called2 3 (2) Bowles, Oliver, U.S. Dept of Interior, Washington, D.C., private communication. (3) A Primer on Asbestos Textiles, prepared by the Asbestos Textile In stitute, 1946. -- 30 -- MTC 000682 `textile' fibers being offered to customers. Most of these fibers come from sources other than Canada and often are poorly graded. They sometimes are brittle and contain high percentages of non-fibrous ma terials. On several occasions, small sam ples of textile fibers from foreign countries have been submitted and found to be satisfactory in length, strength, and in quantity of such impurities as magnetite, limestone, rock, and dust. An order of a ton of the fiber, however, proved en tirely different and resulted in large card drops because of brittle ness and non-uniformity of fiber length. Brokers submitting samples of textile fibers should get a guaran tee from the miners that shipments of fibers will be equal to samples submitted to prospective customers. For papers and millboards, most manufacturers use fibers of Groups 3 to 6, depending upon the kind of products being made. The longer fibers are used in special papers or diaphrams requiring sufficient length to give reinforcement. The types of paper products manufac tured by a paper mill are numerous and may be mentioned briefly as electrolytic cell or diaphram, wick, pipe covering, millboards, roofing papers, laminated roll board, roof ing felts, pipe line felts, welding paper, and similar materials. Fibers for asbestos-cement pro ducts vary from Groups 4 to 7, in clusive, depending upon the methods of manufacture. This group of pro ducts includes shingles, pipes, jackets, flat sheets, corrugated sheets, wall boards, and electrical panels. Asbestos asphalt roof coatings usually contain fibers of Groups 6 and 7, and floats, depending upon the particular coating being manu factured. The use of floats was dis cussed in Asbestos Floats (4) pre sented at the Ottawa General Meet ing of the Institute in January, 1952. In this presentation, many of the physical properties and uses of floats were discussed. The popularity of floor tiles made from organic resins and as bestos fibres has resulted in ex ceedingly large sales. Group 7 fibers used in floor tiles by all lead ing United States manufacturers amounted to approximately 170,000 tons in 1952. (4) Badollet, M.S., C.I.M., Trans., Voy. LV, 1952, pp. 185-189. Asbestos-cements contain fibers of Groups 6 and 7, as well as floats, the fiber used being governed by the cement. The quantity of fibers used in these products in the United States in 1952 was approximately 30,000 tons. i Amosite United States industry imported 18,319 tons of amosite fibers (5) from South Africa last year. This was approximately 30 per cent of the total African production. Samples of amosites are being offered by many vendors and brokers. In some cases, the same samples appear under different identification numbers. These en velope samples actually are too small for evaluation and many times appear to be of good quality, al though larger samples of the same fiber contain high percentages of rock and are rejected. In many cases, the fiber length is so short that it falls into the filler classifica tion. When properly processed, this variety of asbestos is bulky and light in density, and frequently is suitable for insulations. Amosite imported into the United States gradually is becoming short er in length and dirtier, and con tains considerable rock. It is neces sary, therefore, for the consumer to reprocess the fiber to remove the non-fibrous material which should have been removed in Africa. This additional processing cost plus loss of the non-fibrous material lias raised the cost of this fiber to a point where substitutes will be given consideration. It is not known what kind of control tests are maintained on the production of amosite fibers in Africa, but it is suggested that, whatever they are, they should be reviewed in terms of consumer needs. Amosite is not as strong as chrysotile or crocidolite. nor does it have the good flexing properties of these fibers for use in textiles. Occasional samples of amosite, however, possess sufficiently high strength and good flexing properties to be used in textiles, if they were avail able in quantity. Products requiring light weight and not needing high strength can be greatly improved by the use of this variety of asbes tos. Amosite fibers, with or without other varieties of fibers, will exhibit (5) Bowles, Oliver, U.S. Dept, of Interior, Washington, D.C., private communication. low densities, fairly good strengths, and good insulation properties, and. when properly processed, can be formed into magnesia blocks and pipe coverings of low density and good insulating properties. Amosite, combined with other fibers and fillers, is used to produce lightweight asbestos-cement wallhoards that are attractive, easy to install, and possess good insulation properties. A most recent installa tion of this construction is found in the fireproof walls of the new American ship, United States. Crocidolite Crocidolite asbestos is found in commercial quantities in South Africa and also in Austria. Small er mines in other countries do not produce much fiber of this variety. Africa produced 44,734 tons of crocidolite last year. United States industry imported about 7.100 tons, including 274 tons of Australian Blue fiber (6). That the quality of crocidolite fiber has decreased during the past five years is evidenced by a de crease in length and an increase in rock and extreme fines in the form of finely ground ironstone. Manv vendors and brokers are offering samples of crocidolite as odd lots. Some fibers are so short that they fall into the filler class. If the quality of this variety of fiber con tinues to decrease, it will be nat ural for customers to investigate substitutes. Because of its physical proper ties, crocidolite lends itself to some special applications that are not practical for chrvsotile. For exam ple, this type of fiber has very high tensile strength, has good acidresisting properties, will give good porosity when formed into a filler mat, and will filter out small par ticles of dusts, smokes, and sus pended particles in gases. Crocidolite can be converted into textiles with some difficulty but it does not lend itself to carding and weaving as easily as does chrvsotile. Probably its greatest application is in asbestos-cement products where its properties of strength and increased porosity improve manu facturing processes and boost pro duction rates. Crocidolite is ex tremely effective in acid-resisting gaskets and the so-called service (6) Bowles, Oliver, U.S. Dept, of Interior, Washington, D.C., private communication. -- 31 -- MTC 000683 sheets. Filter mats or filter pads made of crocidolite, or of crocidolite in combination with chrysotile or with such organic fibers as cotton wool, or cellulose, have appeared on the market. These filter pads are found in gas masks, industrial masks, and more recently as the small plug in the top of some cigar ettes. Applications for crocidolite are broadening. Anthophyllite Anthophyllite asbestos is found in several parts of the world in de posits of varying size. Its colour varies from white to brown and some of its physical properties dis courage its use in commercial prod ucts. No estimate has been pub lished of the quantity of antho phyllite fiber used by United States industry in 1952, but the amount is very small. Because the strength and flexing properties of this fiber are poor, it would not be considered a good reinforcing fiber for asbestos-cement products. In most samples of an thophyllite submitted to industry, the fibers are short and would be equivalent to Groups 6 and 7 and floats. On rare occasions, however, the fibres have lengths equal to Group 3 and are bulky. Anthophyllite of good length can be processed to produce a bulky mass with fairly good insulating properties. There is a tendency, however, for the fibres to disinte grate into short fibres during most processing methods. Special mech anical processing equipment, there fore, should be used in milling antho phyllite. Most manufacturers consider this type of asbestos as a filler to be used in products where strength is not an important objective. As a filler it might be considered as a partial replacement for the stronger chrysotile fiber, but only at a sacrifice of some product strength. In asbestos-cement products, this fiber will decrease strength and in crease bulk. For some time, antho phyllite has been used for welding- rod coatings, and has been consid ered successful as a cheap filler. In plastics, a good grade of antho phyllite is bulky and does not give the impact strength exhibited by chrvsotile. It also is difficult to coat completely and `bundles' ap pear on the surface of the molded product. Recently, this fiber has been tried as a filler in floor tile in competi tion with chrysotile. Here, colour and fiber strength are important factors. In tiles of light colours, the brown anthophyllite requires more white pigments to offset the dark colour. White anthophyllite does not present this problem. As for strength, anthophyllite lacks the re sistance of chrysotile to the process ing action of the mixers and rolls and probably would be reduced in length. Finally, decrease in floor tile strength will depend upon the quantity of anthophyllite in the mix. There is no doubt that the use of this fiber as a filler will be in vestigated in many products. Its success will depend considerably on freedom from impurities, fiber quality, uniformity of fiber length, and the conditions of the fibers as sold to the consumer. Quality con trol to insure standard production will be a critical factor. Tremolite Tremolite asbestos is found in many parts of the world in varying degrees of purity. Some deposits exhibit fibers of good strength and flexibility while others yield a prod uct that is rather weak and reduces to powder during processing. Some deposits of tremolite have long, strong fibers that may be converted into textiles and are satisfactory for use in asbestos-cement products, while fibers from other deposits can be used only as a cheap filler. Many samples of tremolite being submitted by vendors and miners possess fair to poor quality, indicat ing contaminaiton. The demand for this variety of fiber is small and only a few customers are interested. The quantity of tremolite consumed ----- --out it can be assumed to be small. Often tremolite is purchased and re-purified by an acid treatment and sold for special uses in the fil tration of chemicals, pharmaceuti cals, blood plasma, and in Gooch crucibles in analytical chemistry la boratories. These specialty items bring high prices. ACTINOLITE Actinolite asbestos occurs in ser pentine areas in various parts of the world and sometimes is closely associated with tremolite. The quantity of actinolite used by in dustry is not known, but it may be presumed to be small. Because of unattractive physical properties, including brittleness and low strength, it is difficult to use actinolite interchangeable with other asbestos fibers. Very little use is made of actinolite except as a cheap filler, and even for this pur pose the quantity incorporated in a product must be closely controlled. Conclusions The demands for all varieties of asbestos fibers have increased greatly during the past five years and in most cases shortages still exist. The quality of Canadian chryso tile fibers has been kept at a high level, but the quality of fibers from other countries has, in general, de clined. This is particularly true of amosites and crocidolites from Africa. Producers of amosite, crocidolite, and anthophyllite fibers should re view their specifications for con trolling fiber quality. This should be accomplished by consulting with the manufacturer to obtain his view point on fiber requirements. Manufacture of asbestos products in the United States, Canada, and other countries has expanded great ly during the past few years, and this expansion, in turn, has in creased exploration and mining ac tivities for asbestos deposits all over the world. -- 82 -- MTC 000684 Heat Treatment of Chrysotile Asbestos Fibers By M. S. BADOLLET* and W. C. STREIB* y (Annual General Meeting, Toronto, 19.55) (Transactions, Volume LVIII, 19 55, pp. 33-37) ABSTRACT A commercial method has been de veloped for controlling the texture rf chrysot le asbestos fibers so that they may possess properties compar able with natural semi-harsh fibers for use in wet processes. The physical properties of the soft, silky and slimy fibers are partially altered by a flash-heating process which removes a portion of the mole cular water. The heat-treated fibers show an improvement in f lterability, an in crease in bulk, absorption, and sur face area, and better electrical prop erties, such as vjlume resistivity. harsh tvpes they are as low as llM2/g." The reason fcr these differences in fiber texture has been a matter of conjecture for some time, and the literature (2) contains several suggested explanations, such as the MgO : SiO- ratio, lime content, presence of calcite, pressure of formation, weathering, transition be tween asbestos and talc, and mole cular water content of the fiber. After examining fibers of widely differing textures, it was found that the best correlation seems to exist between the water of constitution and the degree of harshness. The paramount differences between Introduction harsh- and soft-textured fibers, how ever. are in their respective HRYSOTILE ASBESTOS fi strengths and filtration characteris C bers are found in nature having tics. Furthermore, it is the wide dif varying degrees of texture, referredference in these properties which to as soft, semi-harsh, and harsh. directly influences their industrial Fibers of each type of texture have usages. For example, a soft, silky inherent characteristics which di fiber usually has high tensile rectly affect their usefulness to in strength and is extremely flexible. dustry. Visual differences are ap However, it is also usually slimy parent under a microscope where it and difficult to de-water when used is seen that the soft-textured fibers commercially to produce products have a silky, serpentine or wavy ap involving wet methods. On the other pearance, whereas the harsh-tex- hand, harsh and semi-harsh fibers tured fibers are coarser and spicu- de-water rapidly, but are generally lart. Semi-harsh fibers, of course, low in tensile strength, and. in many have properties intermediate be cases, brittle to the point that the tween these extremes. flexing characteristics are vastly in ferior to those of the soft chryso Fankuchen and Schneider (1) tiles. have shown by the use of low-angle Obviously, it would be desirable X-ray scattering that the diameter to produce a fiber incorporating the of the harsh type of chrysotile strength properties of the soft-tex liber is somewhat greater than that tured fibers with the rapid filtering of the soft variety. Surface area properties of the harsher types. The measurements on various chrysotiles commercial or industrial importance of different textures tend to con- of such a fiber would be an increase iirm their findings. By nitrogen in the production rate of wet-ma adsorption methods, it was found chine products and possibly the util that the surface areas of soft types ization of lower grades of poorer of chrysotile asbestos are approxi filtering fibers where high product mately 23M'/g whereas for the strength might not be of the utmost `'Research Center, Johns-Mansville Products Corporation, Manville, N.J., U.S.A. fPhotomicrographs showing these differences were publ'shed in C.I.M. Transactions, Vol. LIV, 1961, p. 156. (1) For references see end of pa per. importance. Since the water content of chrysotile appears to be corre lated with its texture, it was de cided to examine heat-treatment as a means of modifying the texture of the soft fibers and improving their physical properties. Effect of Heat on Texture of Fiber Samples of chrysotile crudes, carefully cleaned of wall-rock and fiberized, were prepared and classed as soft, silky, and slimy when wet by water. Careful heating of these samples in the laboratory under controlled conditions at various tem peratures and for various durations of exposure definitely indicated that a soft fiber can be converted to one exhibiting the characteristics of a harsh or semi-harsh fiber. Dehydration studies of chrysotile asbestos have been made by numer ous investigators (3) and the ef fects of temperature are well known. However, the purpose in the pres ent application was to dehydrate the chrysotile molecule only partially and to bring the water content in the range normally encountered in the harsh-textured fibers. The "loss on ignition" shown in analyses of semi-harsh to harsh chrysotile fibers usually ranges from less than 12 per cent to 13 per cent, depending upon the degree of harsh ness. The theoretical percentage of molecular water or water of consti tution in pure asbestos is approxi mately 13 per cent. This is repre sented hv the hydroxyl radical in the molecular structure as expressed by the formula Mg,,(OH ),,Si,0,,,. where the heat first breaks down the components to 3Mg;,Si01. SiO:. and 4H.O. Therefore, any reduc tion in the number of hydroxyl rad icals present, as would be effected by heat, will decrease the total amount of water of constitution, and this affects the physical properties of the fiber. For the applications under con sideration, the water of constitution was to be reduced about 30 per cent, or, in other words, the fiber sample was to be heated until its loss in weight amounted to about 4 per cent. In most of the tests, however, this loss actually ranged from 0.3 per cent to 3 per cent. Excessive heating has a deleterious effect upon -- 33 -- MTC 000685 fiber strength (4). and if carried too far severely reduces the useful ness of the fiber as a reinforcing filler. Hence, a carefully controlled heat-treatment process is required to attain the desired product. Laboratory tests were conducted also on Canadian Group 6 fibers from regular production in which they, too, were exposed to heat treatment at various temperatures for varying time intervals. Again, the fibers produced possessed tex ture characteristics similar to those of semi-harsh and harsh fibers. In the fibers from regular production, there are impurities such as brucite, carbonates, and others. Heating will result in ignition losses higher than the theoretical 13 per cent of water of constitution of pure asbestos. Physical Properties ok Heat- nTreateo 6 Fiber Tabic I lists some of the physical properties of the 6I) fihers before and after the heat treatment. The data show that as the heat-treat ment temperatures and times are increased, filtration is improved, water of constitution is reduced, magnetite is decreased, and the strength of an asbestos-cement block containing the heat-treated fiber is decreased. Figure 1. in which percentage in crease in filterahility is plotted against temperature, shows the ef fect of the heating at time inter vals of fifteen minutes and three minutes. Figure 2, showing the percentage increase in fiber filterahility at dif ferent heating-time intervals at tem peratures of 700F. and 1.200 I'., demonstrates that heating at a high temperature for a short period will yield a fiber having the same de Table I--Physical Properties of 6D Fiber ; Before and After He\t Treatment Temperature F. -- 600........................ 600..................... 900....................... 900....................... 1.200.................... 1,200.................... 1,200 .................. Time Min. ___ 15 30 7 15 1 2 3 Increase Residual in Filter- HjOin ability Sample % ___ 7.0 25.0 25.0 79 0 25 0 67.0 71.0 % 100 0 98.4 98.7 97.1 95 3 99.4 94.9 95.3 Residual Magnet ite % Asbestos Pan ___Modulus of Rup ture, p.s.i. Residual STRENGTH % 100.0 90.2 86.9 90.2 82.0 -- 85.3 78.7 3,310 ___ ___ 3.280 2.490 -- -- 2.760 100.0~~~ 98 2 74 5 82.6 gree of filterahility as that obtained by heating at a lower temperature for a longer period. Another interesting comparison is that of the percentage of residual water with the percentage of in crease in filterahility. As shown in Figure 3, there is a definite cor relation. and the amount of water in the molecule governs the degree of filterahility of the fiber. The effect of heat upon the mag netite present in chrysotile from regular production is to convert, partially, the F304 to Fe203, which is non-inagnetic. In Table I. the re duction in the amount of magnetite remaining in the fiber after heating at 600. 900. and 1,200F. shows that the magnetic character of the chrysotile is considerably reduced bv heat treatment. Tabic I also shows the effect of heat treatment upon asbestos-cement panels made of a Canadian 6/) fiber. In the Table, the strengths of these panels are expressed as modulus of rupture in pounds per square inch. Panels made with fibers heated at the temperature of 900 F. for 15 minutes still retained 74.5 per cent of their normal strength. Panels made with fibers heated at l^OO' F. for 3 minutes still retained 82.6 per cent of their normal strength. Another advantage of heat-treat ment of chrysotile is its effect upon the volume resistivity of the fiber. Table II illustrates the vast in crease in the resistivity of a heattreated Canadian 6D fiber which is used in the manufacture of asbestos paper. Vhe.sc data show that, as the molecular water was reduced, the volume resistivity expressed as megohm inches was greatly in creased. Figure 4 illustrates this graphically by a plot of the percent age of residual water against the logarithm of the volume resistivitv. The area between these curves rep resents the decrease in electrical resistivity due to the adsorption of surface moisture upon the fiber sur face. However, even when the fi bers were tested at 7tF. and 14 per cent relative humidity, the im provement in electrical properties was highly significant. A log-log plot of residual water and resistivity has much the same contour as that in Figure 4. The combined effect of improved filterahility and electrical o 200 4oo 600 aoo iooo mo woo *oo TEMPERATURE, "f. Figure 1.--Percent increase in fdurability n. temperature (F.) oraeamauita TIME OF HEATINft. MINUTES Figure 2.--Percent increase in filterability n. time of heating (min.) -- 34 -- MTC 000686 OOSSSOWaOTSTOfcSM % RIMPIML Ht0 or CRYSTALLIZATION Figure 3.--Percent residual H2O n. percent increase in filterability. tO 65 TO 75 80 85 80 95 100 * RESIDUAL WATER OF CRYSTALLIZATION Figure 4.--Effect of heat treatment upon volume reiietivitv of 6D chryeotile asbestos. properties is of direct interest and importance in the manufacture of electrical papers. X-ray diffraction studies were made on a Canadian chrysotile fiber heated for one hour at temperatures of 400, 500, 600, 700. 800, and 900F. They showed that the structure remained typical of chrysotile. At 1,000F. for one hour, the surface was amorphous, but, after carefully scraping off the surface, examination of the interior revealed the X-ray diffraction pat terns were mostly those of chryso tile. At 1,400 F. for one hour, the entire mass had the structure of olivine. This study indicated that chrysotile, after losing small amounts of its molecular water, still retains much of its general struc ture. Anderson and Clark (3) heat ed chrysotile to constant weight at 900C. and showed that the result ing fiber structure was not so defi nite as in natural fiber. Application of Heat Various types of heating units were investigated, such as a muffle, infrared, dielectric, different types of commercial furnaces, and flash calcining. The main problem was to obtain a uniformly heated asbestos particle throughout any given mass, whether suspended in a gas atmos phere or as a thin or thick layer on a conveyor belt. Another problem was to devise a temperature control system that could be regulated man ually or made automatic. A final problem was cooling of the treated fiber before it was bagged. Several years were required to in vestigate all of the possible methods of applying heat to produce heat- Table II.--Effect of Heat Treatment on Resistivity of Fiber >-03 Residual Water % 100 9/3 97.0 94.5 93.3 76.7 89.4 68 2 Residual Water, Log 2 000 1 988 1.987 1.975 1 970 1 885 1.951 1.834 ESISTIVITY 44RH Megohm inches Log resistivity 0.016 0 023 0 026 0 075 0.110 13.30 0.25 45.30 -1.796 -1 638 -1 585 -1.1249 -0.9586 + 1.1239 -0 6021 1 656 Volume Resistivity Bone Dry Megohm inches Log resistivity 0 08 0 29 0 29 4 10 4.80 52.000 41 0 300.000 -1.097 -0 538 -0 538 +0 613 0.681 4.716 1 613 5.477 treated fiber. Finally, the flash-cal cining system was selected as the most appropriate. Heat-Treatment Pilot Plant A flash-calcining unit was pur chased from the Nichols Engineer ing and Research Corporation, and a pilot plant with a processing ca pacity of two tons of fiber per hour was erected at Asbestos, Que. A diagram of the unit is present ed in Figure 5, which shows the path of asbestos particles and gases throughout the system. Thermocou ples were installed at strategic points, and the temperatures were recorded on charts for a permanent record. After some experience, it was decided which thermocouple provided the most critical control, and from them on, the desired tem peratures were regulated by this particular thermocouple reading. The furnace, which was oil-fired, left no deposit of soot or other im purity in the fiber product since complete combustion took place in the chamber outside of the calcining tower. A gas-fired furnace would be ideal because the combustion cham ber could be reduced in size or en tirely eliminated. The pilot plant building was a three-story structure located on a railway siding near the asbestos mill at Asbestos. In the basement was placed the mechanical driving mechanism for operating the air locks and a fan. The first floor con tained the combustion chamber, pyrometer recorders, oil burner, and fiber-feeding conveyor. The second floor contained the bagging spouts where the fiber was sampled, bagged, sewed, and chuted to the boxcar for shipment. The flash calciner unit was J(i ft. high by 8(-. ft. in diameter, and was lined with fire brick. The bot tom of the calciner was conical to enable the heated fiber to drop au tomatically into an air lock and be removed from the heating unit as rapidly as it collected. The natural fiber was fed to a fan unit that gently willowed the fiber, freed it of lumps, and suspended it in an air stream so that it would fall by gravity down through the hot gas-heating zone. Peepholes -- 36 -- MTC 000687 along the side of the calciner per mitted a clear view of the fiber falling down through the hot gases. A complete check was made of the mechanical operation of the furnace, and gas analyses as well as air balances were conducted on the entire system. The time of fiber exposure to the heat was approxi mately 5 seconds. Any impurities, such as wood or organic materials, were consumed by the hot gases and disappeared com pletely before reaching the bagging spouts. Therefore, no complaints were ever made on' the presence of wood in the fiber product. In 1947, the calciner was oper ated at a capacity of two tons per hour, and it required 3 to 3 Im perial gallons of oil per ton of fiber, depending upon the tempera ture of operation. The total operat ing cost, taking into account losses and operational expenses, varied from $3 to $10 per ton. Three men could run the pilot plant after ad justment of temperature. One man fed the conveyor, one man bagged, and one man sewed the bags. The entire unit was closely supervised by additional men, who gathered data on the operation. During the years of 1947 and 1948, the pilot plant was run on all grades of fibers varying from Group 3 to Group 7 inclusive, and the resultant fibers were investi gated in the Company's plants and converted into products. Based upon their filtration properties, the fibers produced by heat-treatment belong in the texture class of semi-harsh to harsh, depending upon the tempera ture of the treatment. These fibers showed improvements in filterabilitv and absorption, which can be trans lated into increased production in wet processes, and in better satura tion of asbestos felts. The major fields in which heat-treated fibers can be utilized are asbestos-cement products, millboards and asbestos papers, and felts. 'Floats' produced as a by-product of the heat-treating system were promising; they were bulky and had excellent electrical properties. The magnetite content of the floats de creased as the temperature was in creased; however, the total iron by chemical analysis remained approxi mately unchanged. The surface area by gas-absorption techniques showed a definite increase with temperature increase. Bulk, water absorption, and oil absorption also increased as the temperature was increased. All these factors are favourable in heattreated floats to be marketed as fill ers for such products as plastics. Figure 3*---Diagrammatic .ketch of pilot plant. The method of producing partial ly dehydrated chrvsotile fibers is covered by U.S. Patent 2,616,801 and Canadian Patent 504,310 (6) and applications have been filed in other countries. Packing fiber in jute or paper bags by mechanical or hydraulic packers is now becoming popular in the asbestos industry. Normal chrysotile fibers which are soft or silky are unaffected by this type of pack ing. Heat-treated fibers, when com pressed to a pressure of 500 p.s.i., showed only slight changes in phys ical properties. There was no de crease in length but there was a slight increase in density and a small decrease in absorption and filterabilitv. These small changes in physical properties were considered unimportant, and, therefore, pres sure-packing of heat-treated fiber is satisfactory. In the heat treatment process, the colour of the fibre changes from greenish-grey to cream or light brown, depending upon the temper ature of the treatment. The bulky nature of the heattreated fibers is very apparent in such tests as the Quebec screen, Rotap, density, and absorption. Although heat-treated fibers showed no evidence of fusion, micro scopic examination did reveal that fiber bundles exhibited fraying ef fects that increased fiber surface area and coverage. Advantages of Heat-Treated Fibers Since heat-treated fibres are somewhat similar to natural semiharsh to harsh fibers, they can com pete with these commercial fibers in industrial usage. Deposits of semi-harsh to harsh chrysotile fibers exist in various parts of the world but very seldom does any one deposit contain a complete range of different fiber grades. Generally, there is little variation in the degree of harshness of the fiber in any one particular orebodv. Industrial Use of Heat-Treated Chrvsotile Asbestos Fibers The process of heat-treating as bestos fibers allows the producer to make products of any degree of harshness he desires for any partic ular usage. Operators of mines in which the ore contains a slimy chry sotile fiber which is difficult to dewater may now have a means of converting their raw material into a product with the filtering character istics that are requisite in wet pro cesses. Furthermore, the texture of the product can be controlled, and at the same time its physical prop erties can be improved at a nominal cost per ton. The manufacturer who uses heattreated fiber will obtain increased production rates on wet-machine op erations, improved saturation capac ity of his paper felts, and better electrical properties. Also, these treated fibers will enable him to use lower grades that are difficult to dewater. Conclusions (1) A heat-treatment process has been developed for converting soft, silky, slimy chrysotile asbestos fiber -- 36 -- MTC 000688 into fibers possessing properties competitive with those of natural semi-harsh to harsh fibers. The pro cess, a flash-heating system, was operated as a pilot plant at the rate of 2 tons per hour for several years. (2) The process allows the mine and mill operator to control his product and to produce any desired texture for fibers grading from Group 3 to Group 7 and 'floats,' at nominal additional cost. (3) By heat treatment, the phys ical properties of the fibers have been partially altered by the re moval of small quantities of mole cular water and a partial conversion of magnetic iron oxide (magnetite) to the non-magnetic oxide, Fe203. (4) The improvements in filterability, increased bulk, increased ab sorption, increased surface area, and better electrical properties, such as volume resistivity, are all a direct function of the temperature and time of heating. (5) The tensile strength of heattreated fibers, as well as the strengths of asbestos-cement panlels containing them, will decrease as the temperature and time of treatment is increased. However, this disad vantage is minor and the products are equal to those obtained by using natural semi-harsh to harsh fibers. References (1) Fankuchen, I., and SchneidER, M., Low-Angle X-ray Scat tering from Chrysotilea; Am. C'hem. Soc., Jour., Vol. 66, 1944, p. 500. (2) Soboleff, N. D., and Latarinoff, M. W., The Cause of Brit tleness in Chrysotile Asbestos; Econ. Geol., Vol. 28, January-- July, 1933, p. 171. (3) Nutting, P. G., Some Standard Thermal Dehydration Curves of Minerals; Professional Papers 196-198, P.T.B., 1941--1943. Gill, a. F., The Thermal De hydration on Composition of Certain Minerals and Salts; Can. Jour. Research, Vol. 10, 1934, pp. 703--712. Wolochow, D., and White, W. H., Thermal Studies on Asbes tos; Can. Jour. Research, Vol. 19, Feb., 1941, pp. 49--55. (4) Badollet, M. S., Asbestos; A Mineral of Unparalleled Prop erties; C.I.M., Trans., Vol. LIV, 1951, pp. 151--160. (5) Anderson, H. V., and CIlark, G. L., Application of X-rays in the Classification of Fibrous Silicate Minerals Commonly Termed Asbestos; Ind. and Eng. Chem., Vol. 21, No. 10, 1929, pp. 924--933. (6) Badollet, M. S., and Streib, W. C., U.S. Patent 2.616,801, Partially Dehydrated Chrysotile Fibers and Method of Making; also Canadian Patent 504,310. -- 87 -- MTC 000689 The Role of Asbestos in Plastics By M. S. BADOLLET* and tit. R. XIMENEZf (Annual General Meeting, Quebec City, April, 1956) (Transactions, J'olume LIX, 1956, pp. 283-288) Introduction (6) Increase heat and fire re sistance URING the past twenty years Dthe use of asbestos filler in (7) Allow ejection of the cold molding compounds has increasedmoldings from the mold, and im considerably, and hence the plas prove handling and movement of the tics field has served as an outlet for uncured moldings without danger large quantities of asbestos fibers of deformation or warpage from Canada. It is estimated that Unfilled moldings would be im approximately 14,000 tons of as practical in the cold-molding field bestos fibers varying from Group 4 of bitumens and phenolics. Because to floats are used by the plastics tbe filler must produce sufficient industry. handling strength, asbestos, with its The utility of fillers in molding fibrous character, is desirable. compositions has long been known. Binders normally used are asphalts, Their original purpose was to help pitches, gilsonite, polymerisable make molding practical and to re oils, resins, and similar compounds. duce the product cost. In the early Cold-molded products of the bitu days, these fillers consisted of saw minous type are sometimes loaded dust, cotton waste, powdered asbes with as much as 85 per cent of as tos, silica, diatomaeeous earth, sand, bestos. Such a product possesses clay, powdered glass, cork, marble good heat resistance and fairly good dust, slate flour, or other materials. electrical insulating properties, and Each filler had certain good char is low in cost. Its strength is not acteristics as well as certain objec equal to that of hot-molded prod tionable ones. ucts; however, it is sufficient for In this presentation it is proposed the type of moldings made with to discuss only asbestos as a filler cold-molding powders. in molding plastics. Cold-Molded Phenolics Types of Plastics In general, molding plastics can be divided into the following class es: (1) cold-molded, (2) thermo plastics, and (3) thermosetting. Cold-Molded Plastics Asbestos fillers have the follow ing advantages in the cold-molding field: (1) Lower molding cost (2) Make cold-molding possible by controlling the flow of material under pressure while it is in the cold mold (3) Improve mechanical proper ties (4) Increase hardness (5) Decrease molded shrinkage ,Johns-Manvil]e Research Center, Manville, New Jersey. fFormerly of Johns-Manville Re search Center; now retired. Viscous solutions of a suitable phenolic resin are used as binders in cold-molded phenolics. The ad vantages of this higher-priced type of cold-molding compound are: bet ter surface finish, lighter colours, more dimensional uniformity, and better mechanical and electrical properties. Manufacturing Method Cold-molded compounds are in general made bv loading the fillers into a heated internal mixer, after which the binders and other in gredients are added in liquid form or in solution. This mass becomes somewhat tacky and, after dis charge, it is allowed to 'season'. Later on it is screened to separate lumps, whiskers, and other extra neous material to provide a fairly uniform granular compound for molding. The grades of asbestos used in the cold-molding field are normal ly described as floats and shorts. The quality and physical charac teristics of the asbestos are import ant and should be controlled. The most important and troublesome variables within this field are: (1) Texture and quality of fiber should be as constant as possible so that the molder can set up his working formula and not be forced to make repeated changes. (2) Bulking property of fiber should be kept constant. For exam ple, if the fiber is increased in bulk ing properties, more binder is re quired and this increases product cost. On the other hand, if the fiber is too heavy or less bulky, the molding material becomes tacky because it does not absorb the prop er quantity of binder. This condi tion also causes screening and mold ing difficulties, sticking to the molds, increased baking time, and softer moldings. Often, the mechan ical strength of the product is low ered. (3) Excess moisture in the as bestos also can cause trouble, such as sticking and blistering. Normal ly, asbestos as received by the mold er contains from one to two per cent of moisture. Under some stor age conditions in customers' plants, the fiber may pick up additional moisture; this is why some plant operators pre-dry their fiber as a protection against any moisture in crease due to storage conditions. (4) Non-uniform fibers should be at an absolute minimum. They may be described as those that have `whiskers' or long fibrils in excess of the average fiber length for that particular grade. These 'whiskers' twist and tangle and interfere with the efficiency of the screening pro cess. They also cause sticking to the molds, produce surface blisters, and in most cases affect the finish and the appearance of molded products. -- 38 -- MTC 000690 After the molding powders are properly sized by screening, they are ready to be placed in the mold. At this point, the bulk factor is extremely important because pre forming and final molding are one and the same operation. Moreover, positive molds are used, and most molded parts are each fitted with from one to several metal inserts. The molds are loaded by means of a small hand rake or mechanical rake. Good pourability and proper bulk of the molding compounds are indispensable at this step for good and efficient cold-molding. If the molding compound charge is insuf ficient, then the molded part is offsize; and if the charge is excessive, the product is heavy and in some extreme cases may even damage the mold. The cold-molding operator has his own responsibilities. He must de velop a well-balanced compound in which the ratio of binder to filler is adjusted in relation to the bulk value of the filler. If he does this, sticking and off-size products, as well as other troubles, will be mini mized or eliminated. In most cases, the compounds are dusted with a powdered lubri cant before they are loaded into the mold, to prevent any sticking tendency. If the formulation is made correctly, the cold-molded com pound will flow properly and pro duce good material. The main fac tors the operator must constantly watch are: (1) bulk of the asbestos filler; (2) ratio of binder or filler; (3) proper size of particles of the molding compound; and (4) mold ing pressure as well as baking tem peratures and cycles. In general, it may be stated that cold-molded products are not suit able for power factor applications of any importance; but they are used extensively as heater plug parts and as housings for the shield ing of electrical units. High load ing of asbestos as a filler imparts the heat and flame resistance re quired. Thermoplastict In the thermoplastics field the fillers generally used are zinc oxide, titanium oxide, blanc fixe (barium sulphate), talc, diatomaceous earth, and asbestos. The common names of the most familiar thermoplastics are: cellu lose nitrate, cellulose acetate, cellu lose acetate butyrate, ethyl cellu lose, vinyls, polystyrenes, vinylidene chlorides, and esters of acrylic acid. The disadvantages of fillers in thermoplastics may be summarized as follows: (1) Mineral fillers increase the specific gravity and molding weight of both thermoplastics and thermo setting molding compounds. (2) Fillers preclude production of transparent or translucent plas tics (molded, cast, or fabricated). Translucency can be obtained only to a limited extent when the re fractive index of the plastic and the filler are approximately the same; in these cases, however, the mold ings lack quality. (3) Fillers decrease water and moisture resistance. (4) Fillers also reduce electrical and mechanical properties. The ex ceptions are resistance to arcing and compressive strength. (5) Fillers in most eases damage the delicate molding and fabricating equipment. (6) Fillers frequently introduce new and sometimes complicated problems in the difficult art of col ouring, mottling, and configuration of thermoplastics. (7) Fillers in thermoplastics im pose the need for special types of processing, molds, and methods of molding. (8) Fillers increase the quantity of waste and re-work. (9) Fillers make injection and extrusion molding methods troublesrme and wasteful. Consequently, work along these lines is chiefly experimental at present. Asbestos improves the heat re sistance and compressive strength of the thermoplastic moldings. It also reduces the cost and the tend ency to cold-flow, with resultant deformation and warpage. Although hardness is improved, the increase is only moderate. Resistance to flammability and arcing is also im proved by asbestos; however, this is true of all mineral fillers. Incorporation of fillers in ther moplastics is usually handled by one of three different methods, briefly described as follows: (1) Preliminary mixing of the plastic binder in powdered or flake form with the plasticizers, fillers, and other ingredients in ball mills or in internal mixers. (2) Blending of the ingredients, including binders, plasticizers, and fillers, with or without solvents, in heated kneaders. Hot-rolling is fol lowed by calendering, stamping, or by molding. (3) Incorporation of the fillers and plasticizers on the hot rolls after the plastic binder has been consolidated into a plastic sheet. This procedure is used for blanking and/or fabricating and/or molding. To obtain a good product requires considerable gocd judgment on the part of the operator. This is in fluenced by: (1) method of incor poration; (2) time of mixing; (3) temperature of mix; (4) percentage of filler; (5) type of filler; (6) mixing technique; and (7) type and condition of the plastic binder. Because of the conditions under which hot-rolling is practical in the thermoplastics industry, and be cause the binders are plasticized, the operation is a safe and efficient one for the operator. Binders re main fusible indefinitely, and this permits perfect blending by pull ing the filled compound in fairly thin sheets over and over as re quired. Filler loadings are usually on the low side. In most cases, a well-formulated compound can be processed and molded without sticking to hot steel; however, in some special cases lu bricants are used to minimize pos sible difficulty. Most of the thermoplastics used for regular molding can be produced in a large variety of molding flows, because the plastic base and the plasticizer content, or their ratios, can be changed within wide limits. In thermoplastics, the molder must give close attention to: (1) nature, grade and flow of the base; (2) nature of the plasticizer, wheth er it is a solid or liquid; (3) per centage of plasticizer; (4) nature and percentage of fiber; (5) pre heating; (6) molding temperature; and (7) molding pressure. Special techniques have been de veloped by the trade for the drill ing and machining of each par ticular type of thermoplastic, filled or unfilled. Asbestos, especially light-coloured fibers, can be used in moderate percentages in the manufacture of thermoplastic molding or calender ing materials based on ethyl cellu lose. vinvlchloride acetate, and vinylchloride (polyethylene has been tried experimentally). The as bestos filler improves the hardness, heat resistance, burning rate, com pressive strength, and arcing resist ance of the product but it affects -- 39 -- MTC 000691 adversely such characteristics as specific gravity, translucency, ap pearance, depth of colour, other col our possibilities, flexibility, shock and flexural strength, water resist ance, and electrical properties. In the thermoplastics field, the use of fillers is very limited. Vinyl chloride acetate and vinylchloridebase floor tiles, however, constitute a very important and growing field. Work is also being done on floor coverings. Asbestos is used in some types of shellac-base compounds also, but not to the extent it was in the past. One manufacturer in the United States has been producing injection moldings in very large quantities with a compound made with miner al fillers, asphalts of high purity, pitches, plasticizers, and hardening or strengthening agents such as as bestos or other fibers of short or medium length. We understand that the cost of these moldings ranges from $20 to $200 per ton. Thermosetting Plastics The thermosetting field of plas tics is large, and use of fillers has many advantages. Filled moldings are cheaper, and the fillers control the flow of the molding compound. In addition, they improve the me chanical strength and dimensional uniformity, they make the moldings more dependable and durable, and they decrease the shrinkage and warping tendencies of moldings after their ejection from the hot mold. Filters shorten the curing time and increase hardness; they im prove heat, fire, and arcing resist ance. Some of the disadvantages in the use of commercial fillers in these plastics are: increase in specific gravity and molded weight; de crease in water and moisture resist ance; lowering of electrical proper ties; increase in wear and tear on equipment and molds. Moreover, as bestos introduces grit, rock, and magnetite. These impurities increase deterioration of equipment; they make drilling and machining opera tions difficult. In the production of thermoset ting plastics it is generally accepted that, if the compound has a tend ency to stick to the hot-mixing rolls, it will also stick to the molds and cause rejections, expensive delays, and cleaning jobs. Moreover, the mirror polish of the cavities may be damaged, and internal lubrication of the molding compounds is indis pensable. If the compound flows poorly, production is decreased. Also, mold ings are mechanically weak, untrue, and defective dimensionally, and electrical and other properties are erratic. On the other hand, if the com pound flows immoderately and too rapidly, it causes excessive case hardening, excessive and wasteful flash, lack of dimensional accuracy, decreased mechanical and electrical strength, and upsetting of pins, re movable parts, inserts, and the like. It also interferes with the efficiency of the ejecting devices. Chrysatile asbestos as a filler has a number of advantages, some of which are common with other fill ers. Advantages of chrysotile are: (1) it is mineral in nature; (2) it is available in large quantities; (3) it is low or moderate in cost; (4) it is fibrous in structure; (5) it is pos sible to adjust percentages used in relation to fiber length; (6) it is easy to process and mold in com pounded form; (7) it retains the binders; (8) it is inert to reactions involving rolling, molding, and cur ing of the compounds; (9) it per mits the manufacturer to use high percentages of filler, e.g., 63 to 70 per cent; (10) it allows the molder to vary the percentage and fiber length to achieve desired mechanical strengths; (11) it imparts tough ness; (12) it provides better surface finish than soft wood flours, barytes, or mica; (13) it retards the burning rate; (14) it supplies heat and fire resistance; (15) it in creases hardness; (16) it reduces the natural shrinkage of the resins and plastic binders; (17). it allows blending with mineral and organic fillers; and (18) it increases the moisture resistance of moldings partly filled with cellulose fillers. Compared with other mineral fillers, chrysotile has the following advantages: (1) it is fibrous and available in lengths desired; (2) it has excellent binder retention; (3) it is easy to process and mold; (4) it may be loaded at higher levels than other mineral fillers; and (5) it produces the hardest and tough est moldings at loadings impossible with other fillers. Asbestos - filled thermosetting moldings excel all others in heat re sistance, up to temperatures of 450F., because asbestos loadings can be made at higher percentages than is possible with other mineral fillers. The impact strength can be varied within wide limits by vary ing the quantity and length of as bestos. Chrysotile asbestos has some dis advantages however. Chief of these are: (1) Combined Water Combined water is present as a part of the chrysotile structure. If temperatures of molded products reach 450F. or higher and some of the combined water in the asbestos is liberated, the molding will blister and crack. This difficulty is par ticularly significant when heater plugs and other electrical and heat ing appliances are involved. (2) Fiber Length Fiber lengths of asbestos used in plastics vary, depending upon the grade. The molder must understand the type of fiber and something about its properties, especially the specific gravity, bulk, molded weight, mechanical and electrical properties, water and moisture re sistance. In addition, he must under stand mixing techniques, colouring, surface finish, molded shrinkage, and percentage of fiber available in the asbestos grade used. (3) Dust or Fines The so-called `dust' or `fines' in asbestos may be granular, fibrous, or talcy. Variations in the ratio of dust to fiber, and its degree of fine ness, will affect all those properties mentioned under the heading of fiber length. For example, surface quality of the molding will improve as the percentage of dust increases; however, all other conditions being equal, the surface quality of the molding will be lowered as the fibrous structure of the dust, or as the particle size of the dust, in creases. In other words, a fine gran ular dust will impart the best sur face finish. (4) Talcy Dust Talcy dust adheres to the sur faces of the fibers, producing what is called the `talcose' effect. It re tards, and in extreme cases pre vents, adhesion of the resin binder to the asbestos filler and adhesion of the blending charge to the hotter of the mixing rolls. This often re sults in increased rolling time, stick ing, fouling, irregularities, poor de tail, inferior surface finish, and in creased rejections or waste. (5) Rock The so-called 'rock fraction' in asbestos, which is mostly ground serpentine, is usually present in dif -- 40 -- MTC 000692 1 fering percentages. This fraction has a tendency to segregate when the bestos. Its removal is difficult. The effect of magnetite on electrical ders of the desired fineness; (2) preliminary dry-mixing in ball mills ingredients forming the molding properties has been the subject of or in internal mixers followed by compound are given the preliminary much discussion, particularly when hot-friction blending in Banbury dry- and cold-mixing. Accumulation it is coated by a resin. Its black mixers, and grinding to powders of of rock on the bottom of the mixer colour is objectionable in moldipgs the desired fineness. By this meth or ball mill unbalances the disposi of light colour or shade. Electrical od of treatment, fiber length is tion of the ingredients. Presence of properties of the molding may or sometimes reduced, and consequent this rock in a molding compound may not be affected by the pres ly the toughness and mechanical will tend to scratch the delicate mir ence of magnetite. strengths of the moldings are de ror polish of the mold, which is creased. However, good blends are costly to repolish. (11) Alkalinity made and the curing cycles are re duced. (6) Pencils of Fiber Chrysotile asbestos has a high al kalinity because of the nature of The purpose of the hot-rolling or hot-blending in Banburys is two 'Pencils' are defined as fiber the molecular grouping of its ele fold, (1) to ensure thorough blend bundles not well opened, and thin ments, which, in the presence of ing of the resin, fillers, dyes, and in cross-section. These pencils cause water, combine to form a small per lubricants, and (2) to facilitate the the charge to stick to the hot rolls centage of Mg(OH)s. The pH is advance of the resin in order to de and molds as a result of the pencils usually around 9.3 to 9.7. In many crease molding time. Moreover, re being crushed between the friction molding powders this alkalinity is duction in the amount of condensa rolls or under the molding pressure. not objectionable; in others, such tion produced in the mold favours When this happens, the newly ex as urea formaldehyde, alkalinity is production of perfect, blisterless posed fiber surfaces remain partly troublesome. However, asbestos is moldings and tends to reduce mold coated and do not ride with the not used to fill resins of this type. ing flash. Thermoset flash is prac charge on the rolls or within the mold. When the pencils remain uncrushed they are enclosed by the binder, and, when opened by the grinding operation, they appear as poorly bound units, or as uncoated fibers or specks that show on the exposed surface of the moldings and (12) Capacity for Moisture Absorption Asbestos, unfortunately, has a high capacity for absorbing mois ture. This effect can be counter acted, however, by oven-drying be fore the molding powders are made. tically useless. On the other hand, the thermoplastics type can be re worked and mixed with virgin com pound. with or without the addi tion of plasticizers to compensate for volatilization in molding. Preforming.--Phenolic molding compounds are usually preformed throughout their mass. These pen in suitable equipment. Preforms cils may also be blamed to a cer (13) Surface Finish Imparted by produced with the original powders tain extent for blistering of mate Asbestos are made in various forms and sizes rial within the mold or in service. (7) Lumps of Fiber If the fiber is lumpy before the molding powder is made it will cause about the same trouble as pen cils. Lumps must be removed to en sure good disposition. The fiber product as produced from the mine should be void of lumps; if not, it should be screened or lightly fluffed before use. Floats of good grade impart good finish as well as the desired tough ness and mechanical strength. As the result of considerable experi mentation, floats of excellent qual ity are now available. (14.) Phenolic Molding Compounds Phenolic resins include the con densation products resulting from the controlled reaction of phenol with formaldehyde or with furfural- or pellets, or in shapes approximat ing the final molding. This method accelerates production, assures uni formity of the molding charge, and reduces waste and flash to a mini mum. Preheating.--Molding preforms are usually preheated in special ovens to bring them to the proper degree of plasticity. Preheating re duces molding time. When long or very long asbestos fibers are used in phenolic molding (8) Whiskers Small percentages of long fibers in an extremely short fiber product are called `whiskers'. Although these whiskers may be only a frac tion of an inch longer than the main makeup of the fiber grade, they cause sticking and affect sur face quality and appearance. (9) Grit Grit is usually smaller in particle size than the rock discussed earlier; it is harmful, however, and can damage the molds. (10) Magnetite dehyde. The ratio of the reactants can be varied. The variations as well as the type of condensing agents, modifiers and/or additives used determine the characteristics of the resultant resin. One- and twostep resins and variations are made; they are used within the field dis cussed, the type selected being based on the properties desired in the end product. Phenolic molding com pounds are made by blending shortfibered asbestos and or floats and subjecting the mixed ingredients (including fiber) to some type of mechanical or heat treatment while the mass is in motion. Chief among these treatments are: (1) prelimin ary dry-mixing in ball mills or in compounds they are mixed with the finely powdered resins and other ingredients in internal mixers, or by spraying liquid resins, or in solu tions. The resultant mixture, al though not very uniform, is used for molding. Mechanical mixing does not affect fiber length, but it does create a problem of loading the mold because of the high-bulk factor. Toughness and mechanical strengths are improved, hut flow within the mold is retarded. Conse quently the molding cycles are long er, more internal lubricant is usual ly needed, surface quality is poor, and the molded mass is not so uni form as that resulting from short asbestos grades. Magnetite, an iron oxide having magnetic properties, is usually pres ent in most grades of chrysotile as internal mixers followed by hotblending in friction rolls, and fin ally grinding of the mass to pow The furan resins include the products resulting from the reaction of furfuryl alcohol and furfural; -- 11 -- MTC 000693 ASBESTOS-FILLED PLASTIC MOLDING COMPOSITIONS (1) Thermosetting Types Phenolics Molded I Cast Specific gravity.................. ....................................... 1 52-2.00 Specific volume, cu. in. /lb.............................................. 18.2-13.8 Mold shrinkage, in. per m............................................... 0.0005-0.006 Tensile strength, p.s.i......................................................... 4.000-6,500 Compressive strength, p.s.i............ ...................... 15.000- 35,000 Flexural strength.............................. ... 7.000-15,000 Impact strength, ft. lb. iin on notch-lzod test......... 0.27-3 5 Hardness, Rockwell......................... M95-M11S Thermal conductivity. !0-4 cal sec /sq. cm per lC 'em 8-16 Resistance to heat (continuous). F............................... 350-400 Heat distortion temp.. F.. ............. 290- 350 Dielectric strength, H in. thickness, short time. volts per mil..................................................................... 100-350 Dielectric strength, H in. thickness; step by step.-- volts per mil..................................................................... Dielectric constant, 10* cycles.......................................... 75-325 5-7 Dissipation (power) factor, 10 cycles.......................... 0 10-0.50 Arc resistance, seconds . ................................. Traces Water absorption, 24 hr., H in. thick, r/c...................... 0 10-0.50 Burning rate................................................ im Effect of weak acids........................................................ I None to slight Effect of strong acids. . . ............... ............... I Decomp, by oxidizing acids; others si. /none Effect of weak alkalies Effect of strong alkalies Depends on al kalinity; si. or marked Effect to organic solvents. None on bleedproof materials Machining qualities........... Molded qualities................ Cure....................................... Poor Fair to good In mold 1.70 16.3 3.000-6,000 10,500-12.500 5.000-8.000 Rl 10 8.4 300 __________ Almost nil Same as molded In mold Furans Molded 1.75 15.80 3.000-4,500 10.000-13,000 6.000-9.000 Rl 10 265-330 0.01-0.2 Slow Same as phe nolics None None to siight Completely resistant Fair to good Good In mold or oven Melamine Formaldehyde Molded 170-2.00 16.3-13.8 0.005-0.007 5,500-7.000 30.000 9.000-11.000 0.28-0.40 Ml 10 13-17 250-400 265 350-400 320 6.1-6.7 0 041-0.050 120-140 0.06-0.14 Ml None to slight Decomposes Very liRht att'k Slight attack None on bleedproof olours Fair Good In mold Cold Molded 0.010-0.017 1.400-3,000 6.000-15.000 3.700-10,000 040 M80-M90 500 >400 85-115 50-75 6.0 0.07 75-200 5.5-2.U Nil Slight Decomposes Inorganic 1^00^2 2 0.000-0.010 2.200 Z.OOO-'tkx) 0 40 M75-M9S 900-1300 >400 50-80 100-500 0.5-15 Nil _ flight Decompoaes None Decomposes Attacked by some Poor to fair Fair to good In oven None None None Poor Fair to good I n oven (1) Data taken from Modern Plastics Encyclopedia. 1955 edition. fiirfurvl alcohol with formaldehyde; furfuryl alcohol and ketones, and finally the product resulting from the polymerization of furfuryl al cohol. Resins made by reacting phenol and furfural also belong to the furan group, hut their properties place them with the phenolics. (15) Melamine-Formaldehyde Molding Compounds Calcium cyanamide. a fertilizer prepared by heating calcium carbide with nitrogen under pressure, is the starting point in the synthetic prep aration of melamine. This watersoluble material reacts with formal dehyde adjusted to a plf of 7.2 to 9.0 hv means of caustic soda. The primary advantages of melamine formaldehyde resins compared with cheaper urea formaldehyde types are increased stability to heat and hot water, and greater resistance to alkalies and fruit juices. Moreover, the melamine formaldehyde resins permit use of alkaline fillers, such as asbestos. Such fillers cannot he used with the urea resins because they would inhibit setting. Molding powders are prepared by mixing the liquid resin with the desired percentage of filler. Double arm open mixers provided with ducts to remove the formaldehyde fumes are used. The damp mass is dumped onto a screener provided with a device to break up the damp resin-filler mixture. From this screener. the compound is conveyed to the oven drier. The galvanized iron shell of the driers must he acidproof. The continuous convey ing belt should be chromium plated. The compound, spread in thin lay ers. is dried under very closely con trolled temperature and humidity conditions. During the drying cycle the water content drops from 50 to 1 per cent and chemical condensa tion of the resin proceeds. Cutting and grinding are the next steps. Grinding is done in ball mills where pigments, lubricants, and catalysts are added. The catalysts develop acidity when the compounds are heated to molding temperatures. These curing agents reduce the cur ing time and make the molding op eration faster and more economical. The compound taken from the hall mills is ground to a fine powder to provide a more uniform molding powder. If a molding compound of higher density is desired, the material is heated from 120 to 220F\ in Ban bury mixers. The partial fusion thus obtained reduces the volume of the molding compound. Powders must he kept in cold storage under controlled conditions. (16) Color of Thermosetting Moldings For most phenolic molding com pounds used for industrial and many other applications, the colour is not important. Colours other than black and brown are available todav, how ever, and the filler has to be selected accordingly. (17) Textures Asbestos of the chrvsotile variety, whether in the form of shorts or floats, should be of soft texture. Semi-harsh and harsh chrvsotile fibers have a tendency to pulverize during the processing cycles and are more difficult to coat with re sins than are the soft chrysotile fibers. Likewise, the molder will have less trouble in formulation if he knows that the physical proper ties of his fibers are not altered during his processing. In the paper. Heat Treatment of Chrysotile Asbestos Fibers (1). it was stated that this kind of fiber in the form of floats is very good for use in plastics because of its im proved physical properties. Considerable research is being done in the fields of polyester and epoxy resins with asbestos in vari ous forms. These fields may be very important in the not-too-distant fu ture. (18) Physical Properties of FiberContaining Moldings. The accompanying Table, taken from the Modern Plastics Encyclo- (1) Badollet, M. S., and Streib, W. C., Heat Treatment of Chrysotile Fibers; C.I.M., Trans., Vol. LVIII, 1955 pp. 33-37. MTC 000694 pedia for 1955, shows the various properties of the asbestos-filled molded compositions of the thermo setting type. Conclusions The role of asbestos in plastics has been reviewed from the view point of its usage in the cold-molding, thermoplastic and thermosetting fields. The advantages and disad vantages of chrysotile asbestos in these fields have been discussed, as well as general methods of com pounding the fillers with the resins. The use of asbestos in plastics is important to both the plastics man ufacturer and the asbestos indus try, and it will continue so for a long time. Today, in the competitive market, it is necessary for the asbestos fi ber salesman to understand the us age of asbestos in plastics and to be able to help the manufacturer wher ever possible. His contacts furnish the leads for future development by the mines of the proper grades of asbestos for the plastics industry. The closest possible co-operation between the mines and the plastics manufacturer is essential to assure the success of the product by keep ing variations in quality within the narrowest limits. -- 43 -- MTC 000695 Identification of Minerals Associated with Asbestos by X-Ray Diffraction Patterns By M. S. BADOLLET* and J. P. McGOURTYf [Annual General Meeting, Vancouver, April, 1958) (Transactions, Volume LXI, 1958, pp. 169-174) InTIIODUCTION als which the chemical analyses in dicate to be present. T HAS LONG hern customary J for chemists to analyze samples Dr. Cooke was very co-operative of asbestos and their associated in furnishing us with samples of rocks bv routine chemical proce material identical to those he used dures. Results of these investiga in his studies. We accepted his tions are usually reported in the chemical analyses and his re-cast conventional way. i.r., the percent minerals data and have applied the ages of the various constituents X-ray technique for checking and existing in the sample under investi gation. Often an examination of the identifying the minerals actually present in the samples. The dual data shows that re-grouping of the chemical analyses and re-calculation purpose of this investigation was. first, to explore the limits of the of the data in terms of minerals will X-rav diffraction method for iden shed more light on the true com tifying the minerals; and second, position of the sample. to develop a simplified technique for applying this information in The late Dr. H. C. Cooke in his sample identification. Such a tech G.S.C. Memoir 211 (1) presented nique would be of great value to an exhaustive study of analyses of company geologists in searching asbestos and rocks taken from the for new asbestos deposits, as well Thetford, Disraeli, and eastern as in the examination of drill cores. half of Warwick map-areas in the Province of Quebec. Also he dis cussed the composition of asbestos X-Ray Technique Employed and other fibers of the Thetford district in a paper he presented to the Royal Society of Canada in 1935 (2). Both papers include a series of analyses of fibers from the chemical point of view. The an The X-ray pictures were taken hy the Norelco X-ray diffraction unit with a Debye-Scherrer camera for a four-hour exposure, using cop per radiation with a nickel filter. alyses are accompanied by a re cast of the data in terms of miner als in which Dr. Cooke sought to All samples were carefully powd ered in an agate mortar before they' were examined in the unit. deduce the mineral composition of each sample. This presentation of data was well done. It has been stimulating to the X-ray specialist The X-ray pictures accompany ing this paper carry the following identifications: who is interested in carrying the investigation further to confirm the predictions of the chemist, miner alogist, and geologist, and to make definite identification of the miner- `Section Chief, fResearch Engi neer, Johns-Manville Research-Cen ter, Manville, New Jersey, U.S.A. S--Serpentine..................... C--Chrvsotile......... A--Antigorite. .. B--Brucite......................... F--Iron oxide. . . G--Grammatite (trerrolite) E--Epidote........................... Q--Quartz........................... A Line 2.49 2.44 2.53 2.35 1.61 8.4 2.9 3.34 (1) Cooke, H. C., Mem. 211, Geol. Survey, Dept, of Mines and Re sources, Ottawa, 1937. (2) Cooke, H. C., Roy Soc. Can., Trans., Sec. IV, 3rd Series, Vol. XXIX, 1935. When doubt existed in identifica tions, it was necessary to adopt spectrographic and petrographic tech niques. Di cussio.v of Data Sample 1* ------ _ .. r. a normal red-weathering serpentine from a chromite pit in the northwest corn er of lot 19 N.W.. Range X. Coler aine township". The chemical an alysis (Table I) revealed that it contained SiO-, MgO, and HjO, with small amounts of other metals. Re-casting into minerals bv Dr. Cooke indicated that the sample was mainly serpentine rock. The reddish colour probably was due to weathering of the small amount of iron present. During recent years the X-ray diffraction patterns of serpentine have been fairly well indexed and identifications have been made easier. The pattern of Sample I showed clearly the typical ring structure of serpentine with a con siderable amount of crystalline brucite. These two diffraction lines are identified in Figure 1. The chemical analysis showed an excess of MgO over the amount required to combine with SiCK to form serpentine, HMg.,Si-Os; this excess of MgO crystallized in the form of brucite, Mg(OH).. which is intimately associated mechan ically with the serpentine. The absence of evidence in the dia gram of the presence of other minerals is due to the fact that any compounds formed by the minor constituents shown by the chemical analysis are not in suffi cient quantity to yield definite Xrav diffraction patterns. The brucite is so closely associated with the serpentine that it could not be re moved by mechanical means but it could be' dissolved out chemically. *The descriptions and chemical analyses of samples 1 to 21 are re produced from G.S.C. Memoir 211, hy Dr. H. C. Cooke. Numbering of the samples is the same as in that report. -- 14 -- jjlTC 000696 I Sample 2 Sample 2 represented the same material as in Sample 1, but it ex hibited "material blackened over a zone a foot wide and flanking a dyke about a foot wide". Chemical analysis (Table 1) showed the presence of A1=03, Fe203, and FeO in quantities that indicated other minerals than serpentine must be present as impurities. The X-ray diffraction patterns did not reveal the presence of brucite, thus confirming the chemical analysis, which showed sufficient SiOs to combine with all the MgO to form serpentine; hence, no ex cess MgO remained to form brucite or any other magnesia mineral. Al though the chemical analyses showed the presence of aluminum and iron compounds, they were not identified in the X-ray pattern be cause they are present only in very small amount. Sample 3 Sample 3, which represented "chrysotile asbestos taken from the King mine", is typical of the fiber from that area. The chemical an alysis (Table 1) disclosed a com position equivalent to that of chrysotile asbestos, with small traces of impurities. Dr. Cooke, in his re-calculation of the analysis, deduced the possible presence of minerals such as chlorite, talc, and periclase as impurities associated with the chrysotile asbestos. X-ray diffraction lines did not indicate the presence of brucite, but they did show a trace of quartz. Al though the chemical analysis, when re-cast in terms of minerals, indi cated the presence of chlorite, talc, and periclase, lines for these miner als were not visible in the X-ray pattern. This sample was also analyzed quantitatively by emission spectro scopy* with known oxide mixes as standards; the data obtained, on an ignited basis, showed the presence of AljOs, 0.10%; Fe2Os, 1.3%; and CaO, 0.11%. If the iron in the orig inal sample was in the form of Fe304, it would have been oxidized to Fe203 by the treatment pre ceding to the spectroscopic examin ation. The total amounts of A1303, Fe303, and CaO by spectroscopic * Emission spectroscopy is not a means of identifying compounds but can be used as a supplement to or dinary wet chemical analysis. Table t.--Analyses, Samples 1 to 4 3102 Al20g F20s -- F0 .. CaO MgO H20 + -<H2O-----T102 MnO .... C02 * CrjQg... HiO----- Total* JL 34.40 0.50 i 0.82 3.79 0.02 42.36 15.17 0.47 Mil 0.94 0.38 0.48 0.35 2. 40.42 2.24 2.11 4.26 0.29 39.44 11.11 0.10 3 42.05 Mil 0.96 0.39 0.05 43.30 12.52 0.75 99.68 99.97 100.02 4* 43.48 1.41 0.60 2.36 0.12 39.50 12.69 0.45 100.61 Figure 1. S--serpentine line; B--brucite line. Figure 2. S--serpentine line. Figure 3. C--chrysotile line; Q--quartz line. Figure 4. C--chrysotile line; F--iron line. 45 -- MTC 000697 analysis did not quite agree with the chemical analysis; this points to the difficulty of interpreting the chemical data. Sample 4 Sample 4 represented a chrysotile fiber from the Johnson mine which was "darkened and made brittle apparently by the blackened zone close to the acid dyke". The chemical analysis (Table 1) indi cated chrysotile asbestos, with Al,Oa and FeO and an excess of Si02 present. The data, when re cast into minerals, showed the pos sible presence of chlorite, brucite, and silica. The X-ray diagram did not show brucite, silica as quartz, or chlorite. The chemical analysis, however, in dicated an excess of Si02 over the amount required by the MgO to form asbestos. This Si02 may have been in the form of amorphous Si02, which would not show in the X-ray diagram; or it may have been in combination with the minor con stituents shown by the chemical analysis, and the compounds thus formed are not in sufficient quan tity to produce an X-ray diagram. The X-ray film did indicate iron as Fe304 or Fe203. Emission spectroscopy of this sample showed the presence of A1203, 0.20%; Fe,03, 4.2%; and CaO, 0.03%. These spectroscopic data did not agree with the chem ical analysis, again pointing to the difficulty of identifying the true minerals present. Sample 7 Sample 7 represented a so-called "high-grade chrysotile asbestos taken from the Beaver mine". The chemical analysis (Table II) showed minor amounts of iron and alu minum, possibly indicative of the presence of other minerals. The re-cast data indicated chlorite, bru cite, and periclase. The X-ray diagram did not show lines for brucite, chlorite, or peri clase; although the chemical analy sis indicated a slight excess of MgO it was not identified in the X-rav diagram. Iron as Fe203 did appear in the X-ray diagram. Emission spectroscopy showed the presence of A1203, 0.12%; Fe203, 1.70%;; and CaO, 0.03%. Sample 8 Sample 8 represented a "highgrade chrysotile asbestos taken Figure 5. C chrysotile line; F--iron line. Figure 6. C--chrysotile line. Figure 7. C--chrysotile line; F--iron line. Figure 8. C--chrysotile line; F--iron line. Figure 9. C--chrysotile line; F--iron line. Figure 10. C--chrysotile line; F -- iron line. -- 46 -- MTC 000698 from Deloro township, Province of Ontario". The chemical analysis (Table II) showed the presence of A1203, Fe203, FeO, and CaO as impurities. The re-cast data indi cated the possible presence of chlorite, talc, and periclase. The X-ray diagram did not show chlorite, talc, brucite, or periclase. If these minerals were present, the quantities were insufficient to pro duce an X-ray diffraction pattern. Emission spectroscopy demonstrated the presence of A1203, 0.07%; Fe203, 1.0%; and CaO, 0.02%; amounts which differ from the chemical analysis. Sample 9 Sample 9 is a chrysotile fiber from the "British Canadian mine at Black Lake". The fiber is further identified as a semi-harsh type, i.e., not actually brittle, but not soft and silky. The chemical analysis (Table II) showed small amounts of A1203, Fe203, and CaO, and almost 2 per cent FeO. Calculated in terms of minerals, the presence of chlor ite, talc, and periclase was indi cated. The X-ray diagram did not re veal the presence of brucite, chlorite, or periclase, but it did show small amounts of Fe304. Emission spectroscopy showed the presence of A1203, 0.16%; Fe_.03, 4.4%; and CaO, 0.02%. If Fe304 were present originally it was probably oxidized to Fe203 by the ignition process in preparing the fiber for this test. Sample 10 Sample 10 is an extraordinary chrysotile asbestos fiber of harsh texture bordering upon a brittle nature. It is from the "Vimy Ridge mine in Coleraine township". The chemical analysis (Table II) indi cated a low MgO-Si02 ratio, a low combined water content, as well as a high percentage of Fe203. Re cast in terms of minerals, the pres ence of chlorite, talc, and periclase was indicated. The X-ray pattern did not show lines for brucite, talc, periclase, or chlorite. If these minerals were present they must have been in very small quantities, not detectable by the X-ray technique. Emission spectroscopy showed A1203, 0.52%; Fe203, 3.0%; and CaO, 0.03%; amounts which do not agree with the chemical data. Table II.--Analyses, Samples 7 to 10 S102 Al20g FeaO,.,. FeO . CeO ... MgO . .. H20f H2O------ 7 41.13 0.80 i 1.01 0.58 0.03 43.24 12.74 0.45 9 42.40 0.14 0.97 0.19 0.14 43.09 12.60 0.45 Totals. 99.98 99.98 9 41.80 0.64 0.80 1.89 0.14 42.18 12.07 0.20 99.72 /O 42.95 Mil 2.08 0.84 0.19 41.60 10.26 1.61 99.53 Figure 11. C--chrysotile line. Figure 12. B--brucite line; C--chrysotile line (barely visible). Figure 13. B--brucite line. Figure 14. C--chrysotile line; A--antigorite linp. -- 47 -- Figure 15. C--chrysotile line. MTC 000699 Sample 15 Sample 15 represented a type of "slip serpentine taken from a broad fault-zone, drift 505X, King mine, Thetford". The analysis (Table III) showed some impurities, such as Fe203 and FeO. Re-cast in terms of minerals, chlorite, brucite, and periclase were indicated. The X-ray diagram showed no trace of brucite, periclase, or chlorite; however, the diffraction lines for chrysotile and iron are clearly defined. Sample 16 Sample 16 is another specimen of "slip serpentine taken from a fault-zone in drift 504X of the King mine, Thetford". The analysis (Table III) showed that the Si02 was slightly low, and the MgO rath er high, for serpentine. Minor con stituents were A1203, Fe203, FeO. and C02. The data, re-cast as min erals, suggest the presence of chlorite, brucite. and hydromagnes ite. The X-ray diagram showed no trace of brucite, chlorite, or hydro magnesite. It did show, however, chrysotile, and iron as Fe30,. Sample 17 Sample 17 represented a "tough, stringy fiber from a fault, drift C502E, near D505X, King mine, Thetford". Chemical analysis (Table III) showed substantial amounts of Fe203 and FeO. Re-cast of the analysis in terms of minerals indi cated the presence of chlorite, bru cite, and periclase. The X-ray diagram showed no trace of chlorite, brucite, or peri clase, but the L-hrysotile line is well defined. Sample 18 Sample 18 was a "soft, chrysotile asbestos fiber taken from drift 304, King mine, Thetford". It was further identified by Dr. Cooke as a fault fiber. The chemical analysis (Table III) showed relatively large amounts of A1203, Fe203, and FeO, indicating considerable contamina tion. Si02 was extremely low, H20 very high, and MgO unusually high, indicating a peculiar fiber composi tion. Re-calculated data indicated that the minerals present might be brucite, chlorite, and periclase, with brucite comprising 57.61 per cent of the sample. This is an interest ing sample for further investigation. The X-ray diagram showed a very high percentage of brucite, Table III.--Analyses, Samples 15 to 21 5109 u.28,... ro- CaO----HgO K1*22O6.4'..`.1 XnO N10 ncoo82.... /X 4?.45 MU 0.56 1.13 0.09 42.78 12.98 0.39 16 39.2* 0.03 1.13 0.55 p.07 42.93 14.20 0.36 >7 60.61 0.49 1.49 1.07 traea 42.31 13.27 1.36 1.69 /$ 14.97 1.53 1.61 2.61 0.07 56.69 22.46 0.76 0.59 0.57 0.93 6.09 0.19 62.01 28.76 0.14 0.32 Z.O 62.16 0.41 1.99 1.72 0.09 40.X 12.X 0.37 0.47 a1 60.63 2.27 1.78 2.53 0.32 39.88 12.47 0.47 Total- 100.38 100.X 100.60 IX,75 99.60 99.29 100.35 Figure 16. C--chrysotile line; B--brucite line Figure 17. G--grammatite line; E--epidote line. with some chrysotile present. No evidence of chlorite or periclase was indicated by the X-ray diagram. Sample 19 Sample 19 was a "fault fiber tak en from the Beaver mine at Thet ford". Chemical analysis (Table III) indicated only small amounts of Si02, A1203, Fe203, and CaO, but a relatively high percentage of FeO. The MgO content (62 per cent), and the combined H20 (28.76 per cent), indicated the presence of some magnesium mineral other than asbestos or serpentine. The re-cal culated data by Cooke showed about 95 per cent brucite together with small amounts of chlorite, periclase, and magnetite. The X-ray diagram showed that the specimen was practically all brucite. The diagrams displayed no evidence of chlorite, periclase, or magnetite.. . Sample-iO " Sample 20 was identified as a "green, fibrous picrolite taken from a fault plane at the 300-foot level in the King mine in Thetford". Chemical analysis showed (Table III) Si02 and MgO, the chief con stituents, were in about the right proportion to form serpentine, but the presence of relatively large amounts of Fe203 and FeO suggest ed an admixture of other minerals. Re-calculated data indicated the presence of chlorite, talc, periclase, and hydromagnesite. The X-ray diagram showed the material to be chrysotile and antigorite. No evidence of chlorite, talc, periclase, or hydromagnesite was observed. Sample 21 Sample 21 was described as a "semi-fibrous picrolite filling vein at the 300-foot level in the King mine at Thetford". Chemical an alysis showed (Table III) that it contained appreciable amounts of A1202, Fe203, and FeO denoting the presence of other minerals than picrolite. Re-cast of the analysis in dicated that there might be chlorite (19.12%), talc (2.23%), and peri clase (0.92%). The X-ray diagram did not show lines for brucite, chlorite, talc, or periclase, but it did show the dif fraction line for chrysotile. -- 48 -- MTC 000700 Sample 50* Sample 50 was described as a "tough stringy fiber forming crossfiber veins at drift 505X in the King mine, Thetford". It was more specifically identified as a softchrysotile fiber. Chemical analysis (Table IV) showed the presence of Al203,'Fe203, FeO, and CaO, as well as a small amount of Na20 and K:0. Re-cast by Dr. Cooke in terms of mineral composition the analysis indicates that chlorite 8.01%, and brucite 9.19%, may be associated with the chrvsotile in this sample. The X-ray diagram disclosed the presence of chrvsotile and brucite, but no chlorite. Sample 63* Sample 63, described as "pseudo asbestos", represented fibrous veins in the footwall of the Federal pit near Robertsonville, Quebec. The analysis is given in Table IV. This specimen was an interesting sam ple for X-ray analysis since it showed a high percentage of A1203, Fe203, FeO, and CaO, with MgO exceedingly low. The X-ray diagram showed lines corresponding to grammatite (tremolite) and epidote. Conclusion Chemical analyses of chrysotile, picrolite, and other varieties of ser pentine invariably, or almost invari ably, show that, in addition to the MgO, Si02, and H20 necessary to form chrysotile of theoretical com position H4Mg3Si209, they contain `The descriptions and chemical analyses of samples 50 and 63 are reproduced from the paper by Dr. H. C. Cooke in Volume XXIX, 1935, of the Transactions of the Royal So ciety of Canada. Numbering of the samples is the same as in that paper. Table IV.--Analyses. Samples 50 and 63' Sample 50 Si02............ AI2O3.......... F62O3.......... FeO............. CaO............ MgO........... h2o+......... HjO--......... TiOs............ MnO........... NiO............. NasO........... KjO............ 37.44 0.58 1.23 0.54 0.20 43.49 14.60 1.71 -- -- -- 0.28 0.05 100.12 Sample 63 46.361 11.39 9.30 3.18 18.47 8.94 1.87 0.12 0.05 0.61 0.15 -- -- 100.44 also a small, or in some specimens a very appreciable, percentage of Al203,Fe203,Fe0, etc. By re-casting the analyses in terms of minerals these may be assigned to various minerals such as brucite, chlorite, periclase, and hydromagnesite, which may or may not actually be present admixed with the chrysotile. Except when in very small amount, definite verification of the presence of one or more of these minerals can be made by application of Xrav diffraction, emission spectro scopy, or petrographical methods of examination. These methods, and in particular the X-ray diffraction method, have been used in the pres ent investigation of specimens of serpentine from a number of asbes tos mines in the Eastern Townships of Quebec, and of one from Deloro township, Ontario. A critical study of chemical, and re-cast mineral, analyses of the specimens had been made earlier by Dr. H. C. Cooke, who very kindly supplied the writ ers with the material used in their investigation. For some specimens, the X-ray diffraction patterns confirmed the mineral composition as derived by Dr. Cooke in his mineral re-cast of the chemical analysis; for some others. in which the extraneous mineral or minerals were in very small amount, diffraction lines for these did not appear. Clearly, with a sample con taining two or more minerals, only those present in sufficient quantity to give diffraction lines can be defi nitely identified as being present. Samples of the latter type were analyzed by emission spectroscopy and in a general way, though not quantitatively, the results were in agreement with those derived by Dr. Cooke from his study of the chemical analyses. Petrographic ex aminations under the microscope were also helpful in cases where doubt existed. This investigation has indicated that no one method is adequate in itself for complete and positive identification of all the minerals present in a sample. For this it is necessary to resort to a combination of several methods. For complete identification of all the minerals present in samples obtained in pros pecting or in drill cores of asbes tos bodies, not only chemical an alyses but also the results of X-ray diffraction, spectrograhic, and pe trographic studies are necessary. Acknowledgments Chemical data used in this paper were taken from G.S.C. Memoir 211, by the late Dr. H. C. Cooke, and from a paper by him in the Transactions of the Royal Society of Canada, Volume XXIX, 1935. This investigation was initiated six years ago after a discussion with Dr. Cooke, in Ottawa, on the use of X-ray technique for identification of minerals. Dr. Cooke supplied all samples used in the investigation. The assistance of Mr. Charles Whinfrey, formerly of the JohnsManville Research Center, who made the preliminary X-ray studies, is appreciated. -- 49 -- MTC 000701