US3142180A - Method and apparatus for making grain shape analyses - Google Patents

Description

July 28, 1964 w. H. GLEZEN ETAL 3,142,180

METHOD AND APPARATUS FOR MAKING GRAIN SHAPE ANALYSES Filed July 10, 1961 0.0m INCH ASTM SIEVE 0.0I65 INCH ASTM SIEVE I5 I L (DISCARD) SINGLE PARTICLE DROPPING MECHANISM 2| 36 38 39 20 27 XL ELECTRONIC DIGITAL 40 W COUNTER RECORDER V26 23 22 RECORD Q A 31 TAPE zgfi FIG. I

INVENTORS WILLIAM H. GLEZEN BYJOHN C. LUDWICK, Jr.

ATTORNEY United States Patent 3,142,180 METHOD AND APPARATUS FOR MAKING GRAIN SHAPE ANALYSES William H. Glezen, Hampton Township, Allegheny County, and John C. Ludwick, Jr., Pittsburgh, Pa., assignors to Gulf Research & Development Company,

Pittsburgh, Pa., a corporation of Delaware Fiied July 10, 1961, Ser. No. 122,706 7 Claims. (Cl. 73-432) tribution is related to the abrasion history of the sand.

Another example occurs in the grinding art, because the shape distribution of the particles of an abrasive material affects the cutting efiiciency of the material, and therefore it is advantageous to specify the grain-shape distribution as well as the grain size itself when identifying an abrasive material.

Heretofore the only reliable way to classify grains as to shape has been by actual observation and measurement of one or more cross-sections of each individual particle. Attempts to segregate as to shape the particles of a particulate material have been made, as shown for example in US. Patents 1,659,153 and 2,658,616, but these and similar devices leave much to be desired as to precision and particularly as to the reproducibility of results.

Accordingly, it is an object of this invention to provide a method and apparatus for measuring a shape-sensitive parameter of a particle whereby by making measurements on a large number of particles of a particulate material one can obtain the grain-shape distribution function for the material. The grain-shape distribution function is a curve whose ordinate is related to the number of particles and whose abscissa is related to particle shape.

This and other useful objects are attained by this invention in the manner disclosed in this specification of which the drawing forms a part and in which FIGURE 1 is a diagrammatical representation of a practical embodiment of a grain shape classifier that may be employed in this invention, and FIGURE 2 is a section through the plane 11-11 of FIGURE 1.

Prior to the application of the method of this invention the particulate material is sieved in order to obtain a sample whose particles have the same uniform nominal size. The particles of the sample will normally pass through a standard sieve but will be retained on the sieve having the next smaller size openings. The uniformly sized sample is also given a specific gravity separation treatment in order that the sample shall contain particles of the same uniform specific gravity. In the method of this invention each of the particles in the thus se- 3,142,180 Patented July 28, 1964 lected sample is then bounced down an inclined, roughened, enclosed chute of substantial length hereinafter termed a channel, and the time of descent of each particle is measured. It has been found that spherical particles traverse the channel in a minimum time and attain a higher mean terminal velocity as compared to non-spherical particles, and that the traverse time is increasingly longer and the mean terminal velocity is lower for particles whose shape departs farther from a spherical shape. Each particle during its traverse executes a plurality of bounces against the bottom and side walls of the channel. It has been found that a particle quickly attains a mean terminal velocity that both theory and observation show to be determined by particle param eters directly related to the shape of the particle. Theoretical analysis as well as observation have shown that spherical particles execute a minimum number of bounces per unit of channel length, whereas less spherical particles execute more bounces per unit of channel length.

A particle whose shape departs farther from a spherical shape will consequently execute a greater number of bounces in descending the channel, and the bounces will also be more erratic in direction. As the number of bounces per unit of channel length increases, the mean terminal velocity attained by the particle decreases, and therefore the time required for the particle to traverse the'channel increases. The traverse time of each particle is measured and from this the mean terminal velocity of the particle can be deduced. From the traverse-time measurements a frequency distribution curve of mean terminal velocity is obtained for the particulate material being studied. It has been found that such a frequency distribution curve is a measure of the grain-shape dis tribution function of the particles of a particulate material. While the effect of the erratic bouncing trajectory of the particle results in variations in traverse time among reruns of the same particle, this effect can be reduced by increasing the length of the channel traversed by the particle.

An apparatus for carrying out the method of this invention is illustrated diagrammatically in the accompanying drawing. In the drawing the numeral 10 indicates a closed chute or channel which the particle traverses from top to bottom. Channel 10 is inclined to the horizontal by an angle 11 which is greater than the angle of repose of the particles of the particulate material being studied. The angle 11 should be as small as will permit running of the particles. A smaller angle 11 will result in a greater difference in traverse time between particles of dilference shapes. In the event that the particulate material being studied contains a substantial fraction of thin tabular particles, it may be necessary to remove this fraction prior to study of the material in order to prevent these particles from fouling the channel. Accordingly, a particle entering the top of the channel will transverse the channel executing a series of bounces against the walls and floor and eventually be deposited in a tray 9 located at the lower end of the channel.

It has been found that the cross-section of channel 10 should be rectangular and of relatively small inside horizontal dimension 8 (i.e. at right angles to the plane of FIGURE 1) and of considerably larger inside transverse dimension 12. The inside horizontal dimension 8 of the channel should be larger than the maximum dimension of any particle to traverse the channel so as to permit free passage of all particles, but should be sufiiciently narrow to constrain the bounces to be in a substantially vertical plane. It is preferred that the dimension 8 be two to seven times the nominal diameter of the particles. The inside transverse dimension 12 of the channel should be sufiiciently large so that contact between particle and ceiling is avoided, i.e. the particles should not strike the ceiling in traversing the channel. It has been found that a rectangular cross-sectional configuration illustrated in FIGURE 2 having its larger dimension 12 in the vertical plane provides optimum resolution between particles of various shapes. By way of example, a channel having inside dimensions 8 of 0.125 inch (perpendicular to the plane of FIGURE 1), and an inside dimension 12 of one inch, and an effective length of about 112 inches is satisfactory and has adequate resolution when used with particles whose nominal size is about 0.020 inch.

In order to obtain significant results it has been fotuid necessary to employ particles of substantially the same uniform nominal size and specific gravity. Sizing is accomplished for example by initially screening the particles by means of a standard sieving technique. Particulate material to be tested is placed in sieve 13 which may for example be a standard 0.0197 inch ASTM sieve. The material that passes through the sieve 13 is placed in sieve 14 having the next smaller size openings, for example a standard 0.0165 inch ASTM sieve. The material which passes through sieve 14 is discarded as indicated by the line 15. Specific gravity separation is accomplished by the well-known sink-float method as for example using bromoform or some other liquid of appropriate specific gravity. The sample thus prepared is employed in making the channel-traverse time measurements. The sieving process and the specific gravity separation produces particles of substantially the same uniform nominal size and density.

The sample on which shape analysis is to be made is transferred to a dropping mechanism 17 as indicated by line 16,. The purpose of the dropping mechanism 17 is to isolate and deliver single grains of the sample to the channel 10. The device 17 may be any type of device which removes at random a single particle, and delivers it at its output opening. Alternatively an operator may use a pair of tweezers to isolate individual grains of the material. The operator or the dropping mechanism 17 at regular intervals delivers single particles and drops the particle into a funnel 13 which delivers the particle to the channel 10. The channel may have an entrance fitting 2t) closed by a perforated plug 21 through which the funnel tube 19 is inserted. The inside diameter of the funnel tube 19 should be no larger than necessary to reliably discharge any particle without chance of catching.

In the slant portion of channel 10 near the upper end thereof the top and bottom walls of the channel each have a transparent (e.g. glass or transparent plastic) window indicated by 22 and 23. The windows 22 and 23 have a relatively small dimension parallel to the length of channel 1% and form transverse light slits. The window material is inserted in the wall of channel 10 so that the inner surfaces of windows 22 and 23 are flush with the top and bottom inside surfaces respectively of the channel Til in order that there be no mechanical obstruction inside the channel It). A light source 24 with lens 25 provides a light beam through the Windowed slits 22 and 23. The light beam passes through the channel 10 at right angles to the longitudinal dimension thereof and emerges on the other side, where the emerging light beam is focused by a second lens 26 onto a phototransistor 27. The light source 24 may be powered by means of battery 28 or by any other conventional means. It is apparent that when a particle drops into the channel 10 and traverses the upper light beam, an impulse will be generated in the circuit of upper phototransistor 27.

Near the lower end of channel 10 the top and bottom surfaces of the channel are provided with transparent windows 2? and 3&9 similar to 22 and 23 and forming transverse light slits. Windows 29 and 30 are flush with the inside surface of channel 10 so as to form no interior mechanical obstruction. A light source 31 energized by battery 32 supplies light which passes through a lens 33 and through the transparent windows 29 and 30, through the channel ill at right angles to the longitudinal dimension thereof, and upon emerging is again focused by lens 34 to fall on phototransistor 35. Accordingly, an impulse will be generated in the lower phototransistor 35 whenever a particle traverses the lower light beam.

For purposes of illustration the optical systems are shown in FIGURE 1 as operating with transmitted light as would be employed when the particles are opaque. However, when the particles are transparent or translucent, it is more effective to employ incident and emerging beams that are at right angles, because transparent and translucent particles are more reliably observed against a dark background by their reflected light than by their effect on a transmitted beam.

In order to measure the transit time of a particle between the upper light beam and the lower light beam the phototransistors 27 and 35 are connected to amplifiers 36 and 37 respectively, and the amplifiers are connected to the input circuits of an electronic counter 38. The phototransistors are conventional and may for example be type 800 made by Texas Instruments, Inc. The phototransistors are biased with D.-C. in conventional manner and are capacity coupled to the grid of the first tube of the amplifier as is customary. The amplifiers 36 and 37 are conventional and serve only to amplify the impulses from the phototransistors to a level sufficient to reliably operate the counter 38. The electronic counter 38 is conventional in all respects and may for example be a Model No. 522B Electronic Counter made by Hewlett-Packard Company of Palo Alto, California. The electronic counter 38 comprises an oscillator whose oscillations are counted during the interval between the impulse received from the phototransistor 27 and the impulse received from the phototransistor 35 in well-known manner. The read-out of electronic counter 38 may be observed visually if desired, and recorded manually, but it is preferred to connect counter 36 to a digital recorder 39. Such digital recorders are conventional and serve to print on a record tape 40 the count indicated by electronic counter 38. A satisfactory digital recorder 39 may for example be a Model No. 560A made by Hewlett-Packard Company of Palo Alto, California. The tape 40 will therefore show a series of numbers which are the respective transmit times of individual particles released into the channel 10.

The preliminary preparation of the sample indicated in the figure by elements 13, 14, 15, and 16 is performed prior to operation of the shape-classifying apparatus and is indicated only schematically since it comprises. wellknown techniques. It is apparent that the shape classification obtained as a result of the operation of the apparatus of this invention will depend on the care with which the preliminary size and specific gravity selections are performed, since these control the degree of uniformity (except as for shape) of the particles placed in the dropping mechanism 1'7. The rate at which the dropping mechanism 17 delivers particles to the channel 10 is adjusted so that a particle may traverse channel 10 before a succeeding particle is introduced at the top of channel 10.

It has been found that the degree of discrimination against particles of various shape will be increased if the channel is made longer. It has been found however that a channel of about 112 inches is sufficiently long to give effective discrimination without being unwieldly or requiring excessive operating time. We have found that sand particles whose nominal size is about 0.020 inch will traverse a channel of the dimensions stated (i.e. /8" x 1 x 112") when inclined at an angle of 26 degrees to the horizontal in the neighborhood of 10 seconds, the exact traverse time depending on the shape of the particle. The inside surfaces of the track 10 are made rough for example by sandblasting, so as to attain a roughness equivalent to that of 600 grit emery paper. It has been found that the nature of the inside surface of channel 10 materially affects the transit time of the particle, and thereby has a substantial influence on the quality of the classification effected by the apparatus.

The angle of inclination 11 of the channel 10 also materially affects the transit time of the particles, and also materially affects the quality of the resulting classification. We have found that for sand particles whose nominal size is about 0.020 inch an inclination angle 11 in the neighborhood of 26 degrees results in obtaining a satisfactory shape classification.

It has been found that the shape resolution obtained with the apparatus of this invention increases with those parameters that increase the time of descent of the particles, i.e. decreasing the angle of inclination or increasing the length of the channel.

Certain precautions are necessary in the operation of the apparatus in order to eliminate spurious effects. It has been found that electrical charges develop on particles of quartz sand and the resulting electrical forces tend to produce spurious effects. In order to eliminate these forces, it is desirable to subject the particles to radioactive emanations (not shown) which serve to produce ionization which discharges the particles in well-known manner.

In order to prevent further development of electrical charges on the particle as it traverses channel 10, it is preferred that the channel 10 be made of an electrical conductor such as aluminum or brass. We have found that the channel 10 may satisfactorily be made of aluminum alloy, as for example type 2024T4. Channel 10 is conveniently made by assembling four fiat sides which can be individually roughened by sandblasting and subsequently assembled. It has been found that the walls and bottom of channel 10 must be scrupulously clean before assembly, and finger prints and all traces of grease must be avoided.

After the transit time of a large number of grains has been measured and recorded on the record tape 38, these data may be classified into convenient intervals and their frequency distribution plotted. The resulting curve will indicate the shape distribution of particles in the particulate material being studied. Alternatively the electronic counter 38 may be connected to a conventional encoder (not shown) which in turn is connected to a conventional di ital magnetic recorder that records its output on magnetic tape. The latter may then be processed by an electronic computer to automatically tabulate the shape distribution curve of the particulate material being studied.

It is also contemplated that the lower end of tube 10 may be provided with an automatic flapper gate (not shown) controlled by the electronic counter 36 so that each particle discharged from the channel 10 is switched to the appropriate one of a series of receptacles according to its respective transit time. In this manner the apparatus may be employed to automatically sort the individual particles as to their shape.

The described embodiment of the apparatus of this invention is adapted to the determination of the grainshape distribution function of relatively small sand particles, but this is by way of example only and the invention is applicable to larger particles by appropriately increasing the cross-sectional dimenisons and length of the channel 10. The invention is also applicable to smaller particles, but it is believed that a lower limit of usefulness occurs where surface forces alfecting the particles become large compared to body forces. Experience has shown that for silica sand this occurs when the particles are smaller than about 0.35 millimeter.

In comparing various samples of particulate material on the basis of grain-shape distribution, the samples are first sieved and separated by specific gravity to obtain fractions each of which is composed of particles of substantially th same uniform nominal size and specific gravity. The invention is then applied to the size and specific gravity fraction that is common to all the samples, and the particle shape frequency distribution functions of that fraction of the various samples compared.

What we claim as our invention is:

1. Apparatus for classifying a particle with respect to its shape comprising a channel adapted to be traversed by the particle, the interior boundaries of said channel being roughened surfaces, means connected to said channel adapted to insert the particle into one end of said channel, means propelling the particle through said channel by executing a plurality of irregular bounces off the boundaries of said channel, and timing means connected to said channel adapted to measure the time required for the particle to traverse said channel.

2. Apparatus for classifying a particle with respect to its shape comprising a channel adapted to be traversed by the particle, said channel having roughened interior boundaries and a length sufficient to permit the particle to attain substantially a mean terminal velocity during its traverse thereof, means connected to said channel adapted to insert the particle into one end of said channel, means propelling the particle through said channel by executing a plurality of irregular bounces off the boundaries of said channel, and timing means connected to said channel adapted to measure the time required for the particle to traverse said channel.

3. Apparatus for classifying a particle with respect to its shape comprising an inclined roughened channel, means connected to said channel adapted to insert the particle into the top of said channel, whereby the particle traverses said channel under the influence of gravity by executing a plurality of irregular bounces off the boundaries of said channel, and timing means connected to said channel adapted to measure the time required for the particle to traverse said channel.

4. Apparatus for classifying a particle with respect to its shape comprising an inclined channel, said channel having a rectangular cross-section two sides of which are in substantially vertical planes and at least the bottom and sides of said channel having roughened surfaces, means connected to said channel adapted to insert the particle into the top of said channel, whereby the particle traverses said channel under the influence of gravity by executing a plurality of irregular bounces off the boundaries of said channel, and means connected to said channel adapted to measure the time required for the particle to traverse said channel.

5. A method of making a grain shape analysis of particulate material which comprises propelling individual particles of the material through a roughened channel in a path that includes a plurality of irregular bounces, and

measuring the time required for the respective particles to traverse said channel. 6. A method of making a grain shape analysis of particulate material which comprises dropping individual particles of the material into the upper end of an inclined roughened channel whose angle of inclination is such that the path of the particle includes a plurality of irregular bounces, and

measuring the time required for the respective particles to traverse said channel.

7. Apparatus for classifying a particle with respect to its shape comprising an inclined channel whose angle of inclination is greater than the angle of repose of the particle in the channel,

said channel having a rectangular cross-section two means connected to said channel adapted to measure sides of which are in substantially vertical planes the time required for the particle to traverse said spaced 2. distance in the range two to seven times channel. the nominal diameter of the particle and at least the bottom and sides of said channel having roughened 5 References Cited in the file of this patent Surfaces UNITED STATES PATENTS means connected to sa1d channel adapted to insert the particle into the top of said channel, 2781127 Bgcker 1937 whereby the particle traverses said channel under the FOREIGN PATENTS influence of gravity by executing a plurality of ir- 10 regular bounces off the boundaries of said channel, 471,449 Canada Feb. 6, 1951 and

Claims (1)

1. APPARATUS FOR CLASSIFYING A PARTICLE WITH RESPECT TO ITS SHAPE COMPRISING A CHANNEL ADAPTED TO BE TRAVERSED BY THE PARTICLE, THE INTERIOR BOUNDARIES OF SAID CHANNEL BEING ROUGHENED SURFACES, MEANS CONNECTED TO SAID CHANNEL ADAPTED TO INSERT THE PARTICLE INTO ONE END OF SAID CHANNEL, MEANS PROPELLING THE PARTICLE THROUGH SAID CHANNEL BY EXECUTING A PLURALITY OF IRREGULAR BOUNCES OFF THE BOUNDARIES OF SAID CHANNEL, AND TIMING MEANS CONNECTED TO SAID CHANNEL ADAPTED TO MEASURE THE TIME REQUIRED FOR THE PARTICLE TO TRAVERSE SAID CHANNEL.

Images