THE UN I V E R S IT Y OF MICHIGAN COLLEGE OF ENGINEERING Department of Naval Architecture and Marine Engineering Final Report INVESTIGATION OF EFFECT OF SHIP MOTION ON THE TRANSPORTATION OF THE NEW ORE CONCENTRATE Henry C. Adams Supervisor.Bertram Herzog Advisor Max Coon Technician UMRI Project 2810 under contract with: PICKANDS MATHER AND COMPANY DULUTH, MINNESOTA administered by: THE UNIVERSITY OF MICHIGAN RESEARCH INSTITUTE ANN ARBOR January 1959

STATEMENT OF PROBLEM The problem discussed in this report has to do with the effect of ship motion on the transportation of the new ore concentrate. Because this ore concentrate is a new material, no information was available on its shipping properties. Therefore it seemed desirable to compare the properties of this ore to some similar materials which are shipped on the lakes. Two materials were used: one for which there are no particular shipping problems, and one for which some precautions must be taken. Wheat and flax meet these requirements respectively, and thus the properties of wheat, flax, and the new ore concentrate were compared through the tests described in the following pageso

OUTLINE OF TESTING PROCEDURES1 The tests used to determine the static and dynamic properties of wheat, flax, and ore were as follows. TEST I This test consisted of determining the angle of repose and the density of the materials as a function of moisture and loading procedures. TEST II In this test the effects of vibration and moisture on the angle of repose were investigated. TEST III This group of experiments was designed to simulate the motion of a ship by using a model as a pendulum. 1A detailed description of these tests and the data obtained from them will appear latero 2

RESULTS From the data obtained, it can be concluded that the new ore concentrate is no more dangerous to ship than wheato To our knowledge, no precautions are taken when wheat is shipped on the lakes. It was also found that the "flatter" the ore is originally loaded, the more stable the cargo will be. However, if for some reason a slight cargo shift does occur, the ore is sufficiently mobile to allow the cargo to redistribute itself properly. NOTE During the study of the physical properties of the ore, it was found that a crust could be formed on a pile of ore by spraying the surface with water. This crust will remain for several days and stop any surface shift of the ore. 5

TEST I The density of wheat, flax, and ore was determined for both the packed and unpacked conditions. Ore of three different moisture contents was usedo The results of this test are given in Table Io To determine the angle of repose2 of these materials, a drop test was set upo To simulate different loading conditions, two drop tests of different types were used on the oreo The wheat and flax were subjected only to the low test because the loading properties of these materials were not of interest. It was found that a high drop test would give a sharper pointed cone. It was also found that the dry ore would give a flatter cone. The apparatus used in this drop test is shown in Fig. 1. The data from this test are shown in Table IIo Actually the included angle of the cone was recorded in place of the angle of repose because it seemed more pertinent. However, the angle of repose in degrees is (l80-0)/2, where 0 is the included angle of the cone in degrees.

0 Ho4vr, 11i.i o fet0 2'_ 4."'.'. f 1' ^ T 4 "'" -*..(o) (b) (c) Fig. 1. Drop test. (a) shows the high drop test; (b). shows the low drop test; (c) shows a typical truncated cone of ore, wheat, or flax. It was possible to obtain a full cone by pouring the last of the material slowly. However, this did not affect 0. 5

TABLE I PHYSICAL PROPERTIES OF CARGOES TESTED (Density in lb/ft3) Ore Wetteda "As Is"b Dryc Wheat Flax Packedd Unpacked Packed Unpacked Packed Unpacked Packedd Unpacked Packedd Unpacked 180 119 145.5 137 166 15305 51 49, 6 45.4 42 178 3 148 139 131 173 155.5 49.5 47.6 44 8 41.4 178.3 144 137 113.5 168 153.5 50.4 48 5 44.8 41.4 i80 151 134.5 111 166 155.5 183 152 139 114 168 153.5 ao 183 146.5 138 116 168 148.5 183 150 137 113 173 151.5 178 138 135 108 178 151.5 183 142 134 112 178 153.5 183 142 132.5 111 Average 181 146 137 116.7 171 152.8 50.3 48~6 45 41.6 a. The term "wetted" means that the ore was saturated with water. b. The term "as is" means that the ore which was received was used directly. c. The term "dry" means that all the moisture was removed from the ore. d. The packed condition was obtained by shaking a container until no more packing could be observedo

TABLE II INCLUDED ANGLE OF CONE IN DEGREES Ore Dry.As. Is.. t._Wheat Flax High Drop Low Drop High Drop Low Drop Low Drop Low Drop 109 115 77.5 80 133 134.5 105 116 80 80 127 139.5 106o5 118 80.5 80.5 131.5 135 114 118 91 108 77 5 74 75 71 80 81.5 Average 108.5 11607 80 82 130.5 136.3 Note: When the data were not consistent, /more tests were performed. 7

TEST II In this test a small container was connected to a vibrating table and samples of ore, wheat, and flax were tested, It was found that the angle of repose of the materials was affected by acceleration more than by velocity. It was also found that dry ore is the most unstable. In the "as is" condition, the ore was mobile, but as more water was added the ore became more stable.

TEST III This series of experiments was the most pertinent to the problem. DESCRIPTION OF APPARATUS USED IN DYNAMIC TESTS A 1/24-size plywood model of 72 ft of the hold of an ore carrier was constructed. The model was supported by 3/4-in. pipe flanges and nipples placed at the ends of the model. These nipples were placed on smooth metal supports which would permit the model to oscillate freely and simulate the rolling of a ship. The points of support were so placed that the surface acceleration of the ore in model and prototype was constant.3 A transducer and oscillograph were used to obtain the dynamic data. The angular displacement of a model could be read directly from the oscillograph. The model is shown in Fig, 2. DESCRIPTION OF DYNAMIC TESTS4 General Remarks.-For a 1/24-size model of three hatch openings, 520 lb of ore was needed. In the tests on wheat and flax, an attempt was made to keep the geometry of the loads the same. This required 172 lb of wheat and 166 lb of flax. These weights of wheat and flax would not give the correct weight loads for the model, but in these tests the geometry was the important factor. For all three materials, the loading was accomplished by slowly pouring the material into the center of the hatch openings. This made the load consist of three cones with intersecting bases. This loading procedure gave the most unstable load, which was desirable for testing purposes. All the ore used in these tests was dried because it was found that this was the most unstable ore condition. Ore, wheat, and flax were used in all the following tests. 3Previous tests on a vibrating table showed that the angle of repose is affected by acceleration, rather than by velocity, DaSa J-:ro?- these tests will appear later in tabular form. 9

. _____ _All dimensions shown are Inside. 6I,,~ /, / Total model weight is 95 ib. Model is swung from 4 "water pipe Interior is finished smoothly. Fig. 2. Model of three hatch openings (1/24 size).

Test -II-lo -This test was designed tD simulate the rolling motion of a shipo The model was loaded and forced to oscillate about its natural equilibrium position. The data recorded on the test were the angular deflection which first caused the particles to move alon-g the surface of the several materials, the maximum angular deflection, and the final angle of heelo Test II'-2o — This test was designed to simulate the case of a ship with an initial heel (which might be caused by a broadside wind). First the model was loaded and then a weight was added to one side of the model. This gave the model an initial heel and a new equilibrium position, about which the model was then forced to oscillateo When the model came to rest, the' weight was removed and the model was forced to oscillate about its natural equilibrium positiono This second oscillation was used to equalize any cargo shift caused by the first oscillationo The recorded data were: the angular deflection caused by the added weight, the maximum deflection of both oscillations, and the angle of heel after both oscillationso Test iTI-o3.-This test was designed to simulate the case of a wave which is out-of-phase striking a ship during its natural rolling motiono In this test, the model was loaded and forced to oscillate about its natural equilibrium position. However, a spring was placed a short distance from the model which interfered with the model's natural oscillationo The model assumed its natural angular deflection on one side of the equilibrium position but not on the other side Two different positions of the springs were used in this test. The recorded data were: the distance from the spring to the model, the maximum angular deflection on both sides of the equilibrium position, and the final angle of listo Test TIT-4.-This test was designed to simulate a ship slowly rolling over to give angular deflection and then beginning to oscillateo This might be encountered when a ship turns into a large broadside wave. The model was loaded and then slowly forced into an initial angular deflection. Then it was permitted to oscillate about its natural equilibrium position. This test was made twice, using two different initial angles. The recorded data were: the initial angle and the final listo 11

SYMBOLS USED IN FOLLOWING TABLES Symbols Quantity Dimensions h Height of cone-of-cargo Inches 40 Included angle of cone Degrees 0da List of model Degrees P 1;~List of model caused by added weight Degrees Q Angle of oscillation Degrees Subscript i's are used to indicate that the quantity is an initial quantity (measured before the test)o Subscript f s are used to indicate that the quantity is a final quantity (measured after the test)O A subscript 1, 2, or 3 is used to indicate that quantity will be used againo In some tests Ol. = Of9and (O = ai for another test. This means that the final angle of list from one test is used as the beginning angle of list for the next test. 12

Test III-1 Simulation of Rolling Motion Wheat Flax Ore r- r hi 9-1/4 10-3/4 10-1/2 I i 140 130 117 1hii |h hf 7-3/4 8-1/2 9-3/4 1 1 1Of 164 158 1 H 09'i 0o o -B0 (ra* ~A rocking motion of increasing amplitude gave an angle of first shift, Qs 6 6 3-1/2 Sand l1+Imax = 1-Imax 16 15 15 f< y \hf Cf 0 3/4 1/2

Test III-2 Simulation of a Heel Caused by an External Force Wheat Flax Ore hi 9-1/2 10-5/4 10-1/2 di^.~~~~~~~~~ f | l~4o 150 119 hi hf 7-1/2 8-1/4 9-1/4 L I.~~~~~~.j1~ 0f 167 157 155 -^r' Y I —-\ A weight L caused an angle 6 6 5 A rocking motion caused ~Imax = IYllmax 14 14 i4 When oscillation had stopped a+ f 8 9 6 When L was removed there was an angle of heel af= 5-1/2 5-1/2 2 + 8 r =^ ai 5 3-1/2 5-1/2 2 A rocking motion gave 1l=1ax v \y \hf Iimax 14 15 14 Of f52 1

Test 11I-3 Simulation of Out-of-Phase Wave &1 = 5-1/2" b2 = 4-1/4" Wheat Flax Ore Wheat Flax Ore hi 9-1/4 10-3/4 10-1/4 9-1/4 10-3/4 10-1/4 |,q^it Xi 140 128 115 139 129 119, ~ |hf 8-1/4 9-1/4 9-3/4 7-3/4 8-5/4 9-1/4 ( i>_\ / Of 159 150 151 159 152 155 Spring ea.ai 0 0 0 0O O A rocking mo- I@Q+max 14 13 13 14 14 12 tion of in- 1- lmax 11-1/2 9 11 9 7 7 f - \ creasing amplitude caused \ I\ / a 1/2 0 0 1/2 1/2 0

Test III-4 Simulation of Slow Roll Wheat Flax Ore * ^fhi 9-1/4 10-1/2 10-1/4 >hi|~.^~0i 142 124 116 J~~~~~i'~~~~ ~hf 7-1/2 9 8-1/2 \f / 10 165 155 lio 01\& ci ^0 0 0 The model was slowly inclined to an angle +'max 8 7-1/2 11 \^ \N ^The model was then allowed to oscillate until it stopped, leaving an angle 1i = 1 1/2 0 8 e^^a c1=c 1 1 1/2 o The model was slowly inclined to ^h an angle -l+Im 17 14 19 hf \The model was then allowed to oscillate until it stopped, leaving an angle f 2 2 5

gA\\y5 02493 0o70