\ ul ‘lelMllH‘llh § ,1 M 1: .l ‘ .Wl I l 1 I w ! IN I l #4 ON (JG—x ":‘E’: 333363 Q? :AK S"?§2§9 ““533 233:5 Tags-Es f0; 1'33 339m; 31‘ 3.3. Mai”. GAB: 3 EAT? 3615.265. 2} 9“( $0393 2'55 E..- a?” sea 355‘- 7" 1 I v. . Ir‘v" THE EFFECTS OF LAKE SUPERIOR SHORE CURRENTS ON RECENT SE DIMENT S By JOSEPH L. ggTRICK A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Geology 1955 ri‘yifiis W—Io—Sé JOSEPH L. PATRICK ABSTRACT Sedimentologists have devoted considerable effort to the study of sediments in the laboratory and in field investigation attempting to determine the effects that tranSportation by water has upon grains of sand. This study is devoted to a laboratory analysis of an artificial sediment which has been subjected to abrasion by natural forces. During the mining of copper ore in the Keweenaw Peninsula of Mich- igan the finely ground waste rock from the milling operation was dumpediinto Lake Superior. The current and wave action has .formed this new sand into extensive beaches. Samples of sand were collected at regular intervals from one beach and analyzed in the lab- oratory. The data from the analysis show a definite increase in round- ness and a corresponding decrease in sphericity with distance from the o riginal source. ii ACKNOWLEDGMENT S The writer is indebted to many individuals who aided with the investigation and text. Particularly is he indebted to Dr. B. T. Sandefur for suggesting the problem and who gave freely of his time and knowledge. His patient direction in the laboratory and helpful suggestions with the organization of the text are greatly appreciated. He is indebted to Dr. S. G. Bergquist for his constant en- couragement and for editing the text. Special thanks go to Dr. W. A. Kelly for his assistance in planning the field work and to Dr. Justin Zinn and Dr. J. W. Trow for their friendly interest and many helpful sugge stions. iii INTRODUCTION . . . Purpose of the Study Location ...... FIELD SAMPLING . Location of Sample Sampling Method TABLE OF CONTENTS ooooooooooooooooooooooooo S .......................... LABORATORY PROCEDURE ...................... Quartile Measure 5 ooooooooooooooooooooooooooo Comparison of Quartile Measures ................. Quartile Skewness SPHERICITY AND R0 General Statement Preparation for R0 Sphericity Analysis and Kurtosis ................. UNDNESS .................... ooooooooooooooooooooooooooo undness and IIIIIIIIIIIIIIIIIIIIIIIIII 22 25 28 28 29 Shape and Roundness with Lateral Movement ......... CONCLUSION . . ............................... Page 30 36 38 39 TABLE 1. LIST OF TABLES Tyler Standard Screen Scale Sieves .......... Comparison of Sorting ................... Comparison of Kurtosis and Skewness ........ vi LIST OF ILLUST RAT IONS Figure Page 1 Index Map Showing Sample Distribution ........ viii 2-12 Cumulative Curves ...................... 10-20 13 Comparison of Sphericity and Roundness with Lateral Movement ................... 31 14 Photomicrograph of Control Sample .......... 32 15 Photomicrograph of Sample at Five Miles ...... 33 16 Photomicrograph of Sample at Ten Miles ...... 34 Figure 1 . .' -. o ' :E:E::JL'A \i ' F: \ “.['. to M . "t on." 9 ‘ 6° I 3 C3 “" ‘ _ _ .. A“ ' t f I Q /‘\ T I I \I . c) I ”"8": Erin any I I . | t I — ----- 0 3 3 - . I ,' I P :52»; I I0 ' . \v 9 ‘ 8 7 6 5 « Edgm::3 on“ mag. F'.J:g~’6/§:}¢ion Hill 0mm“ 'Chompio’n Mill ‘ Hou man I | 2 some [0050" ’3 TrimouMoIn '3 ’t Champion _ P ‘R Glob. . \ 4 . 470L— Porrago p“ fairy VA 3 Q, .7 1‘ t f. I. - I Q -. U.\ f) LOCATION OF SAMPLE DISTRIBUTION KEWEENAW PENINSULA 0° CDTIES ‘F‘I—i—i—i— RAILROADS 5 O 5 .0 Location when sump!" won comer“ II ,-- I :— 4 Nun-bu mm Ioculon mucous "on" nunbov . 2 NINES —'-— CG. LINE MILES. - __ _-- _ - l J INTRODUCTION The Keweenaw district of the Upper Peninsula, Michigan, con- tains the world's outstanding deposits of native copper. Exploitation of the great ore reserve started in 1844, and mining Operating con— tinued to expand until World War I. Since then there has been a constant decrease in the number of active properties. Entire com- munities have been abandoned and in many places only the founda— tions of the mills and houses remain. Some of the striking features, remaining as monuments to man's industry, are the beaches of black sand which mark the location of the mill sites. It is estimated that a total of 600 million tons of ore has been processed since the beginning of mining in the area. The mills crushed and processed the ore for the native c0pper which it con— tained. The tailings, or waste rock, from the process was dumped into the nearest convenient body of water, there to be moved and rearranged by current and wave action into extensive beaches. The writer was intrigued by the possibility of making a study to deter- mine the effects of transportation by water upon the particles of waste rock. Purpose of the Study Krumbein and Griffith (1938) describe a sediment which would be ideal for study as: One in which the environment is relatively self-contained; that is in situations in which the complete history of the sedi- ments may be traced from source to final deposit. If, in addi- tion a simple lithological setup is involved so that variables in terms of particle density and the like may be avoided, it seems likely that a more complete picture may be obtained, free of complexities which cannot be evaluated directly. The sediment described in this report is of simple composi- tion. It began its tranSportation cycle from a known point as a well sorted, angular sand. The wave and current action of Lake Superior has been active in translocéating and abrading the material for almost a century. The purpose of this study was to determine what effect the wave and current action had on the size and shape of the indi- vidual grains involved in the process. Location The beach chosen for this study lies on the west side of the Keweenaw Peninsula. It extends for a distance of ten miles from the Champion Mill in the Village of Freda to the breakwater of the Portage Lake canal. Long-shore currents produced by prevailing westerly winds have carried the new material northeastwardly until it has completely isolated the old beach with its towering wave-cut cliffs of Freda (Cambrian) sandstone. Indentations of the old shore line have been filled, and in these local areas the beach is often a mile wide. Its average width, however, is approximately 100 yards. FIELD SAMPLING Location of Samples The points at which the samples were collected along the beach are shown in Figure l. The first sample, taken immediately below the tailing flume of the Champion Mill, had not been subjected to any natural abrasive wave action and was used as the "control" for the experiment. The other ten samples were collected at one- mile intervals along the beach, the distance between samples being measured by pacing. The boldness of the cliffs along the shore was such that the writer was able to mark the position of each sample on aerial photographs. The pacing was checked by measuring from easily identifiable natural and man-made landmarks. Sampling Method Each of the eleven serial samples collected for study weighed approximately four pounds. With the exception of the first, or "control," the samples were taken within three feet of the water's edge, At each site a small trench was dug to a depth of 18 inches and one wall of the trench was very carefully channeled from top to 4 bottom. The material was collected in a paper bag placed at the bottom of the trench, then carefully transferred to regular sample bags which were labeled in consecutive order. LABORATORY PROCEDURE Preparation for Analysis Laboratory preparation for the analysis was limited to drying and then reducing the samples to 200 grams. The wet samples were placed in an electric oven and dried for approximately two hours. A Jones sample splitter was used after the drying oper- ation to reduce the sample. Sieving The sieving analysis was made with the aid of a Ro-Tap shaking machine using the Tyler standard screen scale sieves shown in Table 1. Each ZOO-gram sample was subjected to sieving in the Ro-Tap machine for 15 minutes. At the end of the sieving period the screens were removed and cleaned; each sieve fraction was weighed and placed in a separate container for later analysis. Mineralogy Under microscopic examination the mineralogy of the grains proved to be simple and homogeneous. The bulk of the material is 6 TABLE 1 TYLER STANDARD SCREEN SCALE SIEVES Size of Ope ning Screen Number Openings per Square Inch Inche s Millimete r 6 0.131 3.36 6 8 0.093 2.38 8 12 0.065 1.68 10 16 0.046 1.19 14 20 0.0328 0.84 20 30 0.0232 0.59 28 40 0.0164 0.42 35 50 0.0116 0.297 48 70 0.0082 0.210 65 100 0.0058 0.149 100 140 0.0041 0.105 150 200 0.0029 0.074 200 very fine-grained basalt consisting of pyroxenes, epidote, and olivine. Grains of secondary calcite are common and usually are quite spherical and well rounded. STATISTICAL ANALYSIS The curnulative curve is a graphic statistical device commonly used in the study of sediments. The cumulative frequency curve is a curve based on the original histogram data and is made by plotting ordinates which represent the total amount of material larger or smaller than a given diameter. [Krumbein and Pettijohn, 1938.] Cumulative curves based upon the weight and size of the indi- vidual fractions were constructed for each of the eleven samples. These curves, Figures 2 to 12, have the independent variable sizes plotted along the "X" axis, and the dependent variable, frequency, plotted along the verticle "Y". aXis. Quartile Measure 5 Quartile measures, introduced by Trask (1930), have been widely used in comparing sediments. In his paper Trask describes a quartile: As that fraction of the sediment which is composed of particles larger 'in diameter than the dimension given for that percentile. Thus, if the three quartiles were 15, 4 and 11‘ respectively, it would mean that 25 per cent. by weight of the sample was composed of particles larger in diameter than 15 K, 50 per cent. of the constituents greater than 4}! , and 75 per cent. larger than 1% . This, with the percentage by weight 9 10 Figure 2 $323.: 2. ”Em _.o no _ N sou. map—ED: t... Olin—Om 8a. mmuzsuxm \ \ MJQSZM. Jomkzoo ...._o w>mDo m>_._.<43_2:o om om ow 00. Percent Weight 11 Figure 3 mmwkmszqdi z. wN_m _.0 m6 _ N 8.. «.85.... 3.. 62.5.3 coo. mmuzsuxm m.:2 _ ._.< unis—4m m0 m>m50 m>:.<.5230 ON 0% 0m 00 00_ Percent Weight 12 Figure 4 mmm._.w2_4..__s_ z. wN_m _.o . 0.0 _ N n v ON om :n. 935:; 2... 62.53 «no. «323....» . \ on N 00. mm.:s_ N ._.< mums—4m “.0 m>m50 m>_._.m30 m>_._.m30 m>_.r<.._3230 ON oe 0m 0m 00_ Pa rcent Weight 15 Figure 7 mmmhm—zjnzz Z. uN_m _.o 0.0 _ N ONO. mmuzBuxm \ mun. wacky—3x \ an; 022.com \ mun—.2 m ._.< wiEsz mo m>m30 w>_._.<.5550 ON 04V 00 Ca 00. Percent Weight 16 Figure 8 mmwkmijdz z. wN_m _.o no _ N who. map—.53. \ to; m¥2h¢Om n50. mmwzefixm mwnzz m ._.< uni—24m ..._O w>mao m>.._.m8 m>_._.<.s§o II\/III|II il‘I 0 ON .m O¢ m .e W t n e c u .00 P Om 00. 21 of the sand, silt, clay and colloid gives an adequate picture of the sediment. Percentiles are especially advantageous for comparing deposits with each other, as they give exact numeri- cal criteria for classifying the size-distribution. " Quartile deviation when used with the median is a measure of the average Spread of the sizes and may be shown as arithmetic,‘ geometric, or logarithmic quartile deviation. The arithmetic quartile deviation is the average between the difference of the two quartiles, and it serves to illustrate the size factor. In the formula shown immediately below, QDa represents the arithmetic mean, Q3 the larger quartile, and Q1 the smaller quartile. The geometric deviation or sorting coefficient (Trask, 1930) is the square root of the ratio of the quartiles. So zl/Q3/ 01 The logarithmic quartile is simply the log of sorting to the base 10. 22 Comparison of Quartile Measures The cumulative curve of the control sample, Figure 2, shows that the material, as it comes from the mill, is distributed rather evenly throughout the various sizes. The succeeding curves, Figures 3 to 12, become progressively steeper and show that a preferential sizing occurs very soon in the transportation cycle. The quartile data, Table 2, taken from each of the cumulative curves, illustrate that coarse particles continue to be separated from the fine as the material is transported along the beach. The first and third quartiles show a decrease in the size distribution about the median and an increase in the general size of the material as it is carried away from its source. The mechanical sizing and screening during the milling opera- tion precludes the possibility of poor sorting in the original sediment. The control sample, as may be suSpected, falls within the 2.5 well- sorted classification of Trask. Of interest, however, is the fact that the sorting coefficient of already well-sorted sediment continues to increase with distance from the source. A comparison of the log quartile deviation, the last column in Table 2, further confirms the sorting effect of transportation. TABLE 2 COMPA RISON O F SORTING 23 Median Q1 Q3 Sample (mm) (mm) (mm) QDa So Log 1050 Control 1. 65 0. 90 2.60 0.85 .77 0.248 Mile 1 2.04 1.67 2.54 0.43 .05 0.021 Mile 2 1.63 0.90 2.80 0.95 .76 0.245 Mile 3 2.32 1.30 3.21 0.95 .57 0.196 Mile 4 2.00 1.25 2.84 0.70 .51 0.179 Mile 5 2.65 1. 79 3. 22 0.71 .33 0. 124 Mile 6 2. 54 1. 84 3. 39 0.72 .34 0.127 Mile 7 1.55 1.01 2.76 0.87 .66 0.220 Mile 8 1.75 1.60 2.18 0.29 .17 0.068 Mile 9 2.62 2.09 3.10 0.50 .03 0.013 Mile 10 2. 80 2. 30 3. 30 0. 50 .43 0. 155 24 The material under consideration undergoes progressively in— creased sorting during transportation. Small sizes are reduced to very fine particles which are carried away in suspension, possibly to be deposited in the deeper portions of the lake. Larger particles, which are moved either by traction or saltation, continue to be moved along the shore until reduced and carried away in suSpension. In this study of a beach sand, Krumbein (1938) shows that a variation of the phi standard deviation exists along a beach. Vari- ation in size and sorting of the sand grains may be either an ex- pression of changing conditions of deposition, or the result of changes in direction of movement of the sand or the SIOpe of the beaches. In certain local areas any change in the character of the shore line may vary the intensity of wave action with a resulting change in the size of the material being deposited. An unknown factor which in- fluences the grain size and Spread is the "travel distance" perpen— dicular to the axis of the beach. Quite obviously, the horizontal distance along the beach is not an exact measure of travel distance for the material as it does not take into consideration the forward and backward movement caused by wave action, and undertow, which at times may be perpendicular or nearly perpendicular to the direc- tion of horizontal movement. 25 Quartile Skewness and Kurtosis Quartile skewness is a measure of the "degree of symmetry" of the size distribution. If the median coincides with a point half- way between the quartiles, the curve is symmetrical. Skewness is the measure of the departure from the median. In its simplest form the arithmetic skewness emphasizes the size factor, and a symmet- rical curve has a value of one. If the value is greater than unity the sorting of the specimen lies on the coarse Side, or if the Spread is greater on the small side the value will be less than unity. Coefficient of the geometric quartile skewness is derived by use of the following formula (Twenhofel and Tyler, 1941). Q1 QB SK = ‘T M Q1 2 first quartile 3 = third quartile 2 . M = square of the median Quartile kurtosis is a measure of "peakedness" of a fre- quency curve. Pettijohn (1938) defines kurtosis as: A comparison of the Spread of the central position of the curve to the spread of the curve as a whole, the values decrease with increasing peakedness. 26 The formula for kurtosis according to Kelley (1924) is: K = Q3 - Q1 2(P10 ‘ P90) K = kurtosis Q3 = third quartile Q1 = first quartile P90 2 ninetieth percentile P10 == tenth percentile Values for skewness and kurtosis are shown in Table 3. The kurtosis of the various samples used in this study does not show any great degree of departure from that of the control sample, and the inference is that the degree of spread of the central position to the Spread of the curve remains constant. The skewness curves of all samples approach unity and may lie very slightly to either the coarse or fine side with a preference shown for the fine. This preference may be a function of the orig- inal sediment. TABLE 3 COMPARISON OF KURTOSIS AND SKEWNESS Kurtosis Skewness Sample _....._ K Q3 - Q1 01 Q3 - z(1’10 - P90) - M2 0 (control) 0.257 0. 855 1 0.150 1. 060 Z ,0. 311 0. 932 3 0.345 0.772 4 0.289 0. 888 5 0. 326 0. 820 6 0. 575 O. 975 7 0. 296 1.160 8 0. 172 1. 122 9 0.296 0. 953 10 0.313 0.858 SPHERICITY AND ROUNDNESS General Statement Two important fundamental properties of sediments which have received considerable attention in field studies and laboratory experiments are .sphericity and roundness. Wadell (1932), who was probably the first to consider Sphericity and roundness as inde— pendent variables, expressed Sphericity as: A ratio of the surface area of a Sphere of the same volume as the particle to the actual surface area of the particle . He expressed the degree of Sphericity by the formula: 3 = degree of true Sphericity Where "5" is equal to the surface area of a sfiiere of the same volume as the grain and "S" is the surface area of the particle. This method of determining the Sphericity of small grains is not only time-consuming, but also difficult. The short method, as used by Pettijohn (1938), is expressed as a ratio of the diame- ters of the inscribed circle to that of the circumscribed circle. Measurements‘are made in the projected plane of the grains. 28 29 The concept of roundness differs from Sphericity in that the former is a measure of the angularity of the reSpective corners of a fragment. Roundness of a particle is the summation of the roundness of the individual corners divided by the number of corners measured in the plane of projection. Wadell (1934) expressed the formula for roundness as: 3:" Roundness z _ R- N- Preparation for Roundness and Sphericity Analysis Analysis for Sphericity and roundness measurements was completed on a portion of each of the eleven samples. The fraction between 0.42 and 0. 59 millimeter diameter was. carefully split to approximately 100 grains. This portion was mounted on slides in a chemical compound with an index of 1. 66. The slide was placed in a microscope for magnification and grains were projected with the aid of a camera lucida. The diameter of the individual corners, as well as the diameters of the largest inscribed and smallest circumscribed circles of fifty grains from each slide, were measured. A celluloid circle scale such as described by Wadell (1935) in his study on the "Volume, Shape and Roundness of Quartz Particles" 30 was used. A scale of this type is made by drawing a series of concentric circles, with an increasing diameter of 2 mm per circle, on a small sheet of celluloid. Shape and Roundness with Lateral Movement A graphic comparison of shape and roundness with lateral movement is shown in Figure 13. Roundness of the grains in the control sample, as. may be suSpected, is very low. Examination of the grains, Figure 14, shows them to be angular with many small corners. The data show that rounding progresses at a rapid rate for the first mile of movement. This accelerated rounding is prob- ably accomplished by fracturing and removal of corners which may have been partially broken during the milling process. After the initial mile the rate of rounding continues at a reduced pace, but in general, it increases with distance from the source. Results of this roundness study closely parallel the experi- mental findings in the tumbling barrel studies by Krurnbein (1941). In his study, values for roundness show a rapid increase from 0.16 at zero miles to 0.5 at 1 mile. In this study the rounding is not as rapid; the rounding progresses from 0.18 at zero miles to 0.25 at 1 mile. Unlike the tumbling barrel, the distance in this case is not a 31 Figure 13 o. m o e. o o v n N . o .o. o_. . on us . . . a. .h N ... a nu . 2.. unuzozaoc l. I i .. . C6293...“ .. i .Nua nus O s .3... and 2. 86-3.6.5! no men n. 4 «NW ch mm s . N 3 ... m mum nhm a. . 'Na. Ohm o n 8 IA s. III Q ”Nos ”h \o 11/1 ‘9’: I \\\\\G \/ x /o. lJ..\ . an. 2.. s s o o x o hm. OH 9/. O s \ j x, .. 3 on. \\\ II \ \\\\\ 11/ \s a“. - 0. PI \\\0\ I!” \s d a on «a. Hzm2m>02 44mm._.<4 I._._>> mmmzozaom 024 >._._o_mmIn_m mo zom_m