i 1 5 £53: lllllllllHHlH: WARD IAN. ANCHORAGI INVESTIGATION Mkthlt WflAflW 4huumfibafifl 1940 THESIS If, ’l‘ .t‘lgaxa.‘ ‘1..’ I.; .I» I Guard Rail Anchorage Investigation A Thesis Submitted to The Faculty of MICHIGAN STATE COLLEGE of AGRICULTURE AND APPLIED SCIENCE by James F. Bovill Candidate for the Degree of Bachelor of Science July 1948 THESIS PREFACE This investigation is a continuation of H. I. Bacon's "Experimental Investigation of Michigan State Highway Guard Rail Anchorages.“ Mr. Bacon established a method of experimentation and constructed the neces- sary equipment for such experiments. The procedure of experiments is similar to Mr. Bacon's, but attempts are made to verify certain fur 4- mental relationships in guard rail anchorage not yet established. 207240 GUARD RAIL AHCHDRACE INVESTIGATION The study of guard rail anchorage is a current project in progress at the Michigan State Highway Department Research Laboratory, and this is a phase of the more general subject which will be beneficial for actual use in our present day highway construction. The idea of contributing to a project which is designed to eliminate aguess work" in engineering prac- tice directed my selection of thesis subject to guard rail anchorage. PURPOSE The guard rail anchorage project is past its primary stage. Mr. Bacon has set up scale model apparatus nd determined a method of experi- mentation. The secondary stage, which is the present phase of the pro- ject, cons'sts of experiments with the scale model apparatus to determine certain graphical relationships existing under controlled conditions. The third and final stage will be the full scale pull outs conducted by the Michigan State Highway Department. HISTORX A history of guard rail anchorage experimentation can be simply described as insufficient. D. C. Hubbard, working for the A. B. Chance Co. on patented metal anchors, gives a general picture of full scale anchor pull outs. His tests were made in various soils and were accomplished by means of a pole, block and tackle, a strong spring scale and a caterpillar tractor. No control was made of variable and only depth of burial was changed. A curved line variation was stated in regard to depth and 15; variation in results was recognized due to different soil conditions. R. T. Haggerstrom wrote in his thesis, “An Experimental Study of Special Anchors in Sand and Clay," that a 45° pull out angle and a greater depth of burial provided larger pull out forces. “An Experimental Investigation of Michigan State Highway Cuard Rail Anchorages,’I by H. I. Bacon is very valuable because of his methods and apparatus. He has set up a reliable method of experimentation which has operatus is quite consisoant as to 4. been changed only slightly, and his a results. The specific purposes of this investigation are to find: the rela- tibnship between anchor area and pull out value, the relationship between anchor depth and pull out value and the relationship between soil moisture content and pull out value. These factors can be estimated from practical eXperience or Luessed at from other works on similar problems, but the object of this study is to control the variables in such a manner that guess work can be eliminated. APPARATUS Figure 1 shows the scale model apparatus built by H. I. Bacon. A 5 foot by 4 foot by 28 inch box was built to the scale of 5 inches = 1 foot for the purpose of holding the sand. It has a glass observation win- dow opposite the angle iron bracket. The bracket is arranged to accommo- date the applied load through a pulley system. he vertical adjustment of the pulley m kes possible the control of the pull out angle. Brackets are attached to the inside of the box to hold the arm of a portable vibrating machine. .2- Figure In. General 1710! of apparatus showing observation window. Figure lb. General view of apparatus showing load-Application system. Figure 1c. General view of apparatus showing shoulder section. View of typical anchor pull out. Figure 1d. MATERIALS As this is a controlled experiment, the type of anchor is constant. The anchors are made of hard wood and are all of the oval type as shown in Figure 2. The oval type anchor was used because it proved to have the greatest pull out value according to both Mr. haggorstrom and Mr. Bacon. The hard wood was treated against moisture absorption by boiling it in linseed oil for one-half hour. The anchor specifications are as follows: Type Dimensions '_§§§ Beight a oval-wood 1/2 in. x 7.8 in. x 3.9 in. dia. 24 sq. in. 1 lb. b I I 1/2 in. x 7.2 in. x 5.6 in. dia. so I I 1-1/4 in. c " ' 1/2 in. x 6.6 in. x 5.5 in. dia. 56 ' ' 1-1/2 in. d - I 1/2 in. x 5.88 in. x 2.94 in. dia. 42 n a 1-5/4 in. The sand was graded sand and air dried prior to beginning the experiment. PROCEDURE A five-gallon can with the bottom out out was forced down into the sand to such a depth that the center of gravity of the anchor lay at the specified depth below the sand surface. The sand contained in the can was removed and the anchor placed at a 45° angle. The sand was replaced and vibrated for 15 minutes. An initial load was always applied to take up any slack in the cable. Any disturbances in the shoulder section were repaired after each experiment. The glass observation window was not used as very little movement occurred in the sand near it. Shoulder damages and pull out cracks were similar to those found b" Bacon. . J Figure 2. View showing oval type anchors. The pull out force was the only measurement needed in this experi- ment and any unusual cases of pull out values or pull out actions were discarded. Purpose DATA To determine the relationship between the anchor area and the pull out value. Apparatus and Conditions .1.— a. Oval type anchors as shown in Figure 5 with areas of 24, 30, 58 and 42 sq. in. These areas were determined by the following method: P1 . k; =-§.= 2 . . x = 2y y y(x) + .785(y)2 = A 2y2 + .785(y)2 = A 2 = A y 2.785 x = 2y 0 2: 42 ' ' yl 2.785 2 - 5O ys " 2.785 b. Air dried sand. c. Twelve—inch burial depth. 2 = 58 2.785 2: 24 2.785 y4 = 5.88 X4 = 2.94 Y4 , I'"I",""'I"'l“'0'5'I"’I'”l"'l'$'""'_'§'m . . _ Isl . ‘ (“I . 1r~ .13 . r '4. L”. 1)firiflqufib‘lfifihfinr.fining}: nt-a.' ...... , ....... :5...'.. I 7 Y i1.1.!.'.r'.l1...I...‘..~..'-‘.B'v.v.r‘v. ‘ . 56 sq. in. are: Figure 5. View of four anchors used in Area vs. Pull Out teat. Results Burial Depth Dead Weight Pull Out Force 39; Area Moisture Content 1 24 sq. in. Air Dried 2 I II I II I! 5 1 50 sq. in. Air Dried 2 I I I I I 5 I! II II I I 1 58 sq. in. Air Dried 2 I I I n a 5 a n I u I 1 42 sq. in. Air Dried 2 R I II II II 5 I I I I I Remarks 12 in. I! ll 12 in. I! II I I 12 in. II I! I I 12 in. I I I I mean mean mean Figure 4 shows the results of this test graphically. 141 lb. 150 " z" u l... qr C210 ‘.5 lb. k 167 lb. 171 “ 172 ' 170 lb. 233 1b. 191 " 238 ' 212 1b. NRJNR‘I H 880" o O) (1"! CD (1) *5 :03 CR :15 H (3" O The S 1.1 gilt upward trend seems to indicate a greater pull out value per square inch of surface area with a larger area as this is nearly a straight line graph, the tendency would be to interpret it as such, and to account for any curvilinear variations by such reasons as: variation in soil com- paction, frictional resistance of the sand, slight differences in the rate of dead load application and the slight difference in the weights of the anchors. -5- FORCE IN POUNDS PULLOUT 280 260 f 240 -__ —-—— / 7 / “—7“ 220 A ’ 2.. / .ao .— ~_I——.— / I60 / I40 / I20 IOO I8 24 30 36 42 AREA OF ANCHOR IN SQUARE INCHES EFFECT of ANCHOR AREA m PU LLOUT VALU E DATA (continued) Purpose To determine the relationship between the depth of burial of the anchor and the pull out value. Apparatus and Conditions a. Thirty-six sq. in. oval type anchor. b. Air dried sand. c. Six in., 8 in., 10 in. and 12 in. burial depths. Results Dead Weight No. Area Moisture Content Burial Depth Pull Out Force 1 58 sq. in. Air Dried 6 in. 56.75 1b. 2 " " " ” ' fl ' 54.00 1b. 5 ' “ ' ' ' ' ' 53.00 lb. mean 53.50 1b. 1 56 sq. in. Air Dried 3 in. 135.25 1b. 2 ' ” ' ' ' ' ' 104.25 1b. 5 ” ' ' ' ' ' ' 102.25 1b. mean 1.04 .25 lb. 1 56 sq. in. Air Dried 10 in. 158.00 1b. 2 ' " ' ' ' ' ' 155.00 1b. 5 " ' ' ' ' ' ' 151.75 lb. mean 154.25 lb. 1 56 sq. in. Air Dried 12 in. 208.00 1b. 2 ' ' ' ' fl ' ' 201.00 lb. 5 ' ' ' ' ' ' ” 211.00 lb. mean 206.00 1b. Remarks Figure 5 shows the graphical results of this test. It is a straight line variation which would be expected. his is also an indication that the Area vs. Pull Out graph (Figure 4) is fundamentally a straight line. 220 ZOO / I80 ' r" " / Ieo / I40 A / I oo —-- l .0 // _________ .0 -/ A 6 IN POUNDS FORCE PULLOUT i \ 4O 8 IO I2 DEPTH OF ANCHOR IN INCHES EFFECT of DEPTH of ANCHOR we PULLOUT VALUE DATA (continued) Purpose To determine the relationship between the moisture content of the sand and the pull out value. Apparatus and Conditions a. Thirty-six sq. in. oval type anchor. b. Moisture content of sand varied between 0 and 20 per cent. 0. Twelve inch depth of burial. Results Dead Weight No. Area moisture Content Burial Depth Pull Out Force 1 58 sq. in. 0% 12 in. 209.00 1b. 2 ' ' ' ' " " 131.00 1b. 5 I I I I I I 255.00 1b. mean 212.00 lb. 1 56 sq. in. 63 12 in. 221.00 1b. 2 ' ' ' ' ' ' 234.00 1b. 5 I . I n I I 22735.75 lb. mean 250.25 1b. 1 56 sq. m. ’ 9% 12 5.11. 240.00 lb. 2 ' ' ' ' ' ' 241.00 1b. 5 I II I! n I ” 2310.00 lb. mean 257.00 1b. 1 55 sq. in 18% 12 in. 255.00 1b. 2 I I I I I I 250.00 lb. 5 I I I I l ' 255.00 1b. mean 245.00 1b. 1 56 sq. in. 20% 12 in. 242.00 1b. 2 I I I I l 3 248.00 lb. 5 ' " ' ' ' “ 255.00 lb. mean 222.00 lb. Remarks Figure 6 shows the results of this experiment graphically. The apparent rise and decline of the pull out value at increasing moisture 260 —T 250 -— /, AA‘ 240 / 230 5 FORCE IN POUNDS \ 220 / PU LLOUT 2|O 20° 0 5 IO I5 20 PERCENT MOISTURE IN SAND EFFECT (7)5 MOISTURE CONTENT m2 PULLOUT VALUE contents can be explained this way: The optimum moisture content of this sand is 18.65, and by checking the graphical results one can see that the maximum pull out force came at about the optimum moisture content. It seems reasonable that at the point of maximum density, the greatest pull out value would be obtained. A rough check of saturated sand was made and the pull out force was found to be considerably less than at 0% moisture. -9- Three tests were run: Anchor Area vs. Pull Out, Anchor Depth vs. Pull Out and Moisture Content vs. Pull Out. Anchor Area vs. Pull Out is a curvilinear variation almost approxi~ mating straight line variation. For all practical purposes it can be interpreted as a linear graph. Thus, the crossectional area of the anchor is directly proportional to the force required to cause failure of the soil to hold the anchor. Anchor Depth vs. Pull Out is a linear variation making the depth of burial directly proportional to the force required to cause failure of the soil to hold the anchor. Moisture Content vs. Pull Out is a curvilinear graph showing that at the optimum moisture content of the soil, the greatest pull out force can be expected. Further investigation on full scale models can be governed by these results. As these results are based on a small scale, accurate predic- tions are not warranted but general trends and estimates of values may be made successfully from any of the information submitted. ~10~ r01 USE“ mm w," II I 1le 31293 i d' .1. m 0319 If I ‘l 6 Ml'I’W‘ 'SMN STAIE UNIVERSITY I ll I” 'l 298 L38 xii la! 2 RAF'Ei I H