Click for next page ( 30


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 29
Appendix B LAYER IGNITION TEMPERATURE BACKGROUND The committee' s Panel on Dust Test Equipment considered five methods for determining the layer ignition temperature of combustible dusts: 1. The hot plate method of the U . S . Department of the Interior, U . S . Bureau of Mine s ~ private communication, M . Jacobson, U. S . Bureau of Mines, 1977) . 2 . The modif fed Godbert-Greenwald furnace method of the U . S . Department of the Interior, Bureau of Mines ~ Dorsett et al. 1960 ~ . 3. The standard method of test for ignition properties of plastics of the American Society for Testing and Materials CAST designation D 1929-68~. 4 . The hot plate method deve loped by the British Fire Research Station (private communication f rom the International Electrotechnical Commission SC 31H/WG2, Fe bruary 1975 ~ . 5. The hot plate method proposed as the International Electrotechnical Commission ~ IEC ~ test method by IEC Subcommittee 31H, Working Group 2 ~ IEC 1975) . It was the consensus of the panel that the test method to be recommended should be a hot plate method because of the simplicity and availabi lity of the equipment involved. Available test data indicate that similar test results are obtained when using different hot plate test methods, when different laboratories use the same method, and when repeated tests are perf armed in the same laboratory . It also was noted that the difference in results between the U. S. Bureau of Mines' hot plate method and furnace method is about +20C. The proposed IT test method was chosen by the panel as the foundation of its recommended method because there i s a relatively wide data base for this method that indicates that it produced results as reproducible as any method studied. The method recommended by the panel for determining the layer ignition temperature of combustible dusts is described below, and the differences between it and the IEC test method are discussed. Mo st of the data readily available in the United States on the ignition temperature of combustible dusts layers is the result of tests by the U. S . Bureau of Mines using dust passing through a 0. 0 74-mm ~ 2 00-mesh) sieve. The 29

OCR for page 29
30 test method recommended by the IEC requires that the dust pass through a 0.2mmm (70-mesh) sieve, with an exception if it is necessary to test coarser dust. There are no reported data correlating the layer ignition temperatures using dust samples identical to each other except for the particle size. Although most experts believe that the difference in layer ignition tempera~ ture between dusts passing through 0.2mmm and 0.074mmm sieves is minimal, the panel decided to recommend use of the 0.074-mm sieve because of the lack of reported correlation data and the mass of information already available on layer ignition temperatures using dust passing through a 0.074-mm sieve. This decision errs, if at all, on the side of safety. Other relatively minor changes from the IEC recommended test method (e.g., thickness of dust layer) also have been made by the panel to reduce the likelihood of differences between layer ignition temperatures obtained using the test method recommended and layer ignition temperatures reported by the O.S. Bureau of Mines. The panel's recommendation that the melting temperature be considered the ignition temperature if the material melts before an ignition is Observed is based on the behavior of molten material, which is different from the behavior of material in the form of dust. Electrical equipment evaluated and found acceptable for use in the presence of dust may not be acceptable when exposed to molten material. RECOMMENDED TEST METHOD FOR IGNITION TEMPERATURES OF DUST LAYERS Scope This test method is intended to determine the minimum temperature of a hot surface that will result in the ignition of a layer of particulate solid, or dust, of specified thickness deposited on it. The test is not suitable for use with substances having explosive properties. Definitions For the purpose of this recommendation the following definitions apply: 1. Ignition--The initiation of combustion in the material under test. Ignition should be considered to have taken place at the minimum hot plate temperature at which: (a) there is visible evidence of combustion such as a red glow or flame, (b) the slope of the temperature-time curve for a thermocouple in the center of the dust layer continues to increase, (c) a 50C temperature rise above hot plate temperature in the dust occurs, or (d) the dust melts.

OCR for page 29
31 2. Ignition Temperature of a Dust Layer--The lowest temperature of a hot surface, rounded to the nearest integral multiple of 10C, at which ignition occurs in a dust layer of given thickness on the hot surface when the procedure in this recommendation is followed. Preparation of Dust Sample The dust should be able to pass a 0.074-mm (200-mesh) sieve. If necessary, any dust passing a O.84mmm (20-mesh) but not a 0.074-mm (200-mesh) sieve should be ground to reduce the particle size until all of the dust passes a 0.074-mm (200-mesh) sieve. The ground and unground (fine dust) samples then should be mixed. The sample must be representative of the dust received and the dust used in the test should be well mixed. Any changes caused in the properties of the as-received dust samples (e.g., by sieving or grinding) should be reported. General Description of Apparatus The apparatus is shown schematically in Figure B-1. Essential details and performance requirements are described below. Heated Surface. The heated surface should consist of a circular stainless steel plate 200 mm in diameter and not less than 20 mm thick. The plate should be heated by an electrical heating element and its temperature should be controlled by a device for which the sensing element is a thermocouple mounted in the plate at the center and with its junction in contact with the plate and within 1 +0.5 mm of the upper surface. The same thermocouple should be connected to a temperature recorder for recording the temperature of the plate during a test. The heated plate and its controller should satisfy the following performance requirements: 1. The plate should be capable of attaining a maximum temperature of 400C without a dust layer in position. 2. The temperature of the plate must be constant to within +5C throughout the period of the test. 3. When the temperature of the plate has reached a constant value, the temperature across the plate should be uniform to within +5C when measured across two diameters at right angles using a procedure like that illustrated in Figure B-~ (this requirement shall be satisfied at plate temperatures of 200 +5C and 3SO +5C measured at the center of the plate). 4. The temperature control should be such that the recorded plate temperature will not change by more than +SC during the placing of the dust layer and will be restored to within 2C of the previous value within S min of placing the dust layer.

OCR for page 29
old , ~ H I - ;~ ~ ) ~ 11 1 1 art E 1. ~J K g It (t ~ 11 1 \ w-~7 ,7-~ \ ~ 1 Age/ l ~ ., ~ liar) _L,L~_ A Heated Plato B Ski rt C Heater D Heater Base E Heater Connection to Power Supply and Controller F Ring for Dust Layer G Plate Thermocoupic to Controller H Plate Thermocouple to Recorder J - I Dust Layer Thermocouple to Recorder Screw Adjustment for Thermocouple Height K Coil Spring FIGURE B-1 General arrangement of hot plate (not to scale).

OCR for page 29
33 5. Temperature control and measurement devices should be calibrated and should be correct to within +3C. Test Thermocouple. A fine thermocouple (O.20 to O.25 mm in diameter) should be located so that the junction is at the geometric center of the dust sample. The thermocouple should be stretched across the heated plate parallel to the surface. This thermocouple should be connected to a temperature recorder for observing the behavior of a dust layer during tests. Temperature Measurement. Temperature measurements with the thermocouple should be made either relative to a fixed reference junction temperature or with automatic cold junction compensation. In either case, calibration should satisfy the above +3C accuracy requirement. Ambient Temperature. The ambient temperature should be measured by a thermometer placed in a convenient position within 1 m of the hot plate but shielded from heat convection and radiation from the plate. The ambient temperature should be within the range of 10 to 30C. Dust Layers. Dust layers should be prepared by filling the cavity . formed when a stainless steel ring is placed on the surface of the hot plate and by leveling the layer to the top of the ring (Figure B-2~. The ring should be 100 mm in diameter and should have slots at opposite ends of the diameter to clear the test thermocouple (Figure B-3~. The ring should be left in place during a test. A given dust should be tested in a layer that is 12.5 mm deep. A ring of the appropriate depth will be required. ~ , ~ A; - . =_ FIGURE B-2 Scoops recommended for forming dust layers. The scoop with the concave edge supports the ring and collects excess dust swept toward it in leveling the layer formed inside the ring.

OCR for page 29
34 6.25+0.1 mm 1 ~ il 1 ~1: L 2+ 0.2 mm Slots 12.510.1 mm 6.25 + 0.1 mm ~1 1 +0.2 mm 1 ~ 100+2.0mml/D FIGURE B- 3 Ring f or f arming dust layer s . Density of Dust Layer. The dust layer should not be compressed unduly ( i . e , =e muse ~ beheld oe put into the ring with a spatula and distributed mainly with sideways movement of the spatula until the ring is slightly over-filled; the layer then should be leveled by drawing a straight edge acro ss the top of the ring and the excess dust should be swept away ~ . To minimize spillage, it is convenient to form a pan around one half of the ring arid then to draw the straight edge towards the pan.

OCR for page 29
35 A layer of each dust should be formed in the above manner on a tared sheet of paper and weighed. The apparent density should be calculated from the weight of the dust and the filled volume of the ring and should be reported. These data are only to provide a reference should data on a similar material yield significantly different results. Procedure General Basis. Ignition in particulate or porous solids exposed to elevated temperatures generally is preceded by a more or less protracted period of self-heating (usually due to atmospheric oxidation). Depending on the temperature of exposure, self-heating may result in no more than a transient, although sometimes substantial, rise in temperature within the solid that does not lead to the propagation of combustion. Further, the "induction period" for ignition at temperatures near the minimum required for ignition is usually many times greater than for ignition in dust clouds or in gases and vapors. For these reasons, recognition of the minimum ignition temperature for layers is less straightforward than for dust clouds or for gases and vapors, and it is necessary to be certain that failure to ignite at a given temperature is not merely a result of premature termination of a test. The occurrence of ignition in a layer of dust on a surface at a given temperature depends critically on the balance between the rate of heat generation (self-heating) in the layer and the rate of heat lost to the surroundings. The temperature at which ignition of a given material occurs therefore depends on the thickness of the layer. Values determined for two or more thicknesses of a given dust may be used for predictive purposes (see the section below on application of results). If the dust is seen to flame or glow, this is sufficient evidence of ignition. If flaming or glowing is not seen, the behavior of the dust layer is to be observed by means of a fine wire thermocouple supported within it and connected to a temperature recorder. It usually will be found that, provided the temperature of the plate is high enough, the temperature in the layer will increase slowly to a maximum value that may be in excess of the temperature of the hot surface and then fall slowly to a steady value below the temperature of the hot surface. This behavior is evidence of self-heating in the dust layer and may often be accompanied by a discoloration of the dust but not by active and visible combustion of the layer. Discoloration shall not be considered to be ignition. If the temperature of the hot surface is slightly higher, it will be found that the temperature measured in the dust layer will continue to rise (instead of passing through a maximum) and lead to ignition. Simple temperature-time curves illustrating this behavior are shown in Figure B-4. If there is no ignition within 30 minutes, the test can be considered concluded. With organic dusts, combustion usually will take the form of charring followed by

OCR for page 29
36 / / Dust Layer Hot Surface TIME A. Self-heating without ignition LO LL CL up 1 Dust Layer J Hot Surface / / / / 1 TIME B. Ignition FIGURE B-4 Typical temperature-time curves for ignition of dust layer on hot surf ace .

OCR for page 29
37 smoldering and glowing that will progress through the layer and leave a residue of ashe Sugars, starches, and some other dusts turn dark, melt, expand, foam, and sometimes char with or without ignition. For these dusts, visible observations and notes should be included with the temperature-time curve. With dust layers composed of certain divided metals, ignition may be characterized by the relatively sudden appearance of highly incandescent smoldering combustion progressing rapidly through the layer. In determining the ignition temperature for a layer of given thickness, repeated trials should be carried out using a fresh layer of dust each time and adjusting the temperature of the hot plate until a temperature is found that is high enough to cause ignition in the layer but that is no more than 10C higher than a temperature which fails to cause ignition. Tempera- tures at which ignition fails to occur must be confined by continuing a test long enough to establish that any self-heating is definitely decreasing in rate (i.e., the temperature at the point of measurement in the layer is decreasing to a steady value lower than the temperature of the plate). Method. The apparatus should be set up in a position free from drafts and under a hood capable of exhausting smoke and fumes. An angled mirror some distance above the test sample or equivalent means for visual observa- tion of the dust during the test should be provided. The temperature of the hot plate should be adjusted to a desired value and should be allowed to become steady within the prescribed limits. A ring of the required height should be placed centrally on the surface of the plate and should be filled with the dusts to be tested and leveled off within a period of 2 min. The test thermocouple recorder then should be started. The surface of the heated plate and the ring should be cleaned after each test. Results. Tests should be repeated with fresh layers of dust until an ignition temperature has been determined. This should be the lowest hot plate temperature, rounded to the nearest integral multiple of 10C, at which ignition occurs in a layer of a given thickness. At least two observa- tions of ignition at temperatures differing by no more than 20C should be recorded. The temperature at which ignition does not occur also should be recorded. This temperature should not be more than 10C lower than the temperature at which ignition does occur and should be confirmed by at least two more tests. The test should be discontinued if ignition of a dust layer does not occur below a hot surface temperature of 400C. This fact and the maximum duration of the test should be reported. Time to ignition, or time to the maximum temperature in the case of no ignition, should be measured to the nearest 2 minutes from the time of placing the dust layer on the hot surface and should be reported. If melting occurs, this fact and the melting temperature should be recorded and

OCR for page 29
38 the test should be discontinued. The melting temperature should be considered to be the ignition temperature.* If flames appear above the surface of the dust but the dust itself does not ignite, the temperature at which flames appear above the dust should be considered to be the ignition temperature.** If foaming of the dust layer occurs, this fact should be recorded and the testing should be continued until ignition, flaming, or melting is observed.*** Validity of Test Results Repeatability. Duplicate results obtained by the same operator on different days should be considered unsatisfactory if they differ by more than 20C. The tests should be repeated. Reproducibility. Results obtained in different laboratories should be considered to be unsatisfactory if they differ by more than 20C. The tests should be repeated. Difficult Materials. Repeatability and reproducibility sometimes may be very poor for reasons associated with the physical nature of the dusts and the behavior of layers during the testing. When this occurs, it should be reported and all results should be accepted as equally valid. The test report should include a brief description of the nature of the combustion following ignition, noting especially behavior such as unusually rapid combustion or violent decomposition. Factors likely to affect the significance of the results also should be reported; these include difficulties in the preparation of layers, distortion of layers during heating, decrepitation, and melting. Reporting of Results The test report should include the name, source, and description (if not implicit in the name) of the material tested; the date and serial number of the test; the room temperature; and the apparent density of the material as tested. The report should state that the determination of ignition temperature has been carried out in accordance with this recommended method. The ignition tests should be reported in the manner shown in Table B-1 (showing results in descending order of temperature rather than in the order in which tests were performed). All test data should be reported (e.g., the * Some materials, such as sulfur dusts, melt prior to ignition. ** This phenomenon may occur with some hydrides, for example. *** Some dusts, such as starch dusts, may foam when heated.

OCR for page 29
39 ignition temperature for the 12.5 mm layer described in Table B-1 would be recorded as 170C). Trials in which the hot surface differed by more than +20 C ~ ~ ~ ~ ~ from the recorded ignition temperature need not be reported unless unusual observations were noted at temperatures higher than the recorded ignition temperature. TABLE B-1. Typical Table of Test Results Depth of Set Surface Time to Ignition or Layer, mm Temperature, C Result of Trial Maximum Temperature' min 12.5 180 Ignition 16 170 Ignition 26 160 - No Ignition 30 160 No Ignition 30 160 No Ignition 30 150 No Ignition 30 Application of Results The values of minimum ignition temperature determined in accordance with this recommended test method apply to layers having the thickness used in the tests. It is possible to estimate minimum hot surface temperatures for the ignition of layers of a given dust of lesser or greater thickness by linear interpolation or extrapolation of the experimental results plotted as the logarithm of the thickness versus the reciprocal of the ignition temperature in degrees Kelvin. This is the simplest predictive procedure that has some theoretical justification. More elaborate treatment based on thermal ignition theory will permit estimates of the ignition temperatures of layers in other configurations (e.g., layers on curved surfaces); however, if accurate predictions of ignition temperature under widely different conditions of exposure (in particular, exposure to a symmetrical high-temperature environment rather than to an unsymmetrical environment like that on a hot plate) are desired, it is preferable to use results obtained for an experimental procedure matching the different environment more closely (e.g., ignition in an oven). When extensive prediction is intended, it is recommended that ignition temperatures be detained for at least three layer thicknesses and that thicker layers be emphasized. Construction of Heater Surface Provided the requirements presented above in the section describing the heated surface are satisfied, the detailed construction of the heated surface is not critical. It should consist of a circular plate of stainless

OCR for page 29
40 steel provided with a "skirt" (Figure B-1) and may be mounted on any suitable electrically heated hot plate commercially available. Aluminum and ordinary steel are not recommended for the heated surface because of the potential for corrosion problems and the possibility that an aluminum surface could be destroyed when metal powders are being tested. There are two ways of achieving a sufficiently uniform temperature distribution across the heated plate, the choice of which depends primarily on the heating device available. If the heater consists, for example, of exposed coiled filaments intended to run at red heat, there should be an air gap of about 10 mm between the heater and the plate so that heat transfer occurs by radiation and convection. If, however, the heater is designed for direct contact and heat transfer occurs mainly by conduction, the plate needs to be much thicker if hot spots are to be avoided. A thickness of not less than 20 mm is recommended. The general arrangement shown in Figure B-1 is self-explanatory. It is preferable to insert indicating and controlling thermocouples in holes drilled radially from the edge of the plate and parallel to the surface at a depth of 1 mm from the surface. The base of the hot plate should be provided with feet in order to clear the support for a thermocouple stretched horizontally across the surface. This thermocouple is to be mounted between spring-loaded carriers on threaded vertical rods and height adjustment should be provided, e.g., by means of wing nuts. Measurement of Temperature Distribution on Hot Plate A suitable piece of apparatus for measuring the temperature distribution across the hot plate is illustrated in Figure B-5. The measuring element should consist of a fine thermocouple with the junction flattened and brazed to a disc of copper or brass foil (5 mm) in diameter. This should be placed at a measuring point and covered with a piece of insulating material (asbestos millboard) 5 mm in thickness and 10 to 15 mm in diameter, held by a vertical glass ron Gnat moves Freely In a tubular guide and to which a fixed load is applied. Temperature measurements should be made along two diameters at right angles and at points 20 mm apart and recorded as in Figure B-6. The thermocouple must be allowed to reach a steady temperature at each point. The measured surface temperature usually will be lower than the set point temperature of the plate depending on the detailed construction of the Theo couple. This is immaterial and can be ignored. The essential requirement is an accurate measurement of temperature differences rather than of absolute values.

OCR for page 29
41 / _ i' 1 ~5 o cr: ~ ' _ ~ _ o ~ ~ 4 - ~ ~ ~ An, - J In ~ _ _ ~ O SO a' C) U) o a) F: SO U' Up 1

OCR for page 29
i 346 42 Plate Temperature: 350 Scale 1:1 343 - 345 - 345 - \ 1 1 1 1 1 1 1 1 1 1 344 346 342 346 - 346 347 346 346 346 C 347 - 347 - 346 346 - - . ~ FIGURE B-6 Variation of surf ace temperature over entire plate--SOC (maximum difference from desired temperature--8OC). REFERENCES Dorsett, H. G., Jr., Jacobson, M., Ilagy, J., and Williams, R. P., Laboratory Equipment and Te st Procedure s f or Evaluating Explosibility of Dusts, Report of Investigations 562 4, U. S . Bureau of Mine s, Pittsburgh, Pe nnsylania, 1960 . International E1 ectrotechnical Cocci s sign, An Apparatus and Method f or Determining the Set f-lgnition Temperature of a Dust Layer, Report IEC/31 H/WG2, Geneva, Switzerland, 1973.