SHEAR INTERACTION OF TWO I-BEAMS FASTENED TOGETHER, ONE ABOVE THE OTHER, BY VA R YIN G LENGTHS OF WELD

SHEAR INTERACTION OF TWO I-BEAMS FASTENED TOGETHER, ONE ABOVE THE OTHER, BY VA R YIN G LENGTHS OF WELD BY FABIAN ROBERT PETERSON B.S.C.E., University of Denver, 1951 TH ESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE IN THEORETICAL AND APPLIED MECHANICS IN THE GRADUATE COLLEGE OF THE UNIVERSITY OF ILLINOIS, 1955 U R B A N A, IL L IN O IS

U N IVERSITY OF ILLINOIS THE GRADUATE COLLEGE 5 July 1955 I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY FABIAN ROBERT PETERSON. E N T I T L E D SHEAR INTERACTION OF TWO I-BEAMS FASTENED TOGETHER, ONE ABOVE THE OTHER, BY VARYING LENGTHS OF WELD BE ACCEPTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR Committee on Final Examination Required for doctor s degree but not for master s. 5M 4-54 54947

iii TABLE OF CONTENTS Title Page I. INTRODUCTION... 1 A. History of Problem... 1 B. Notation U s e d... 2 II. THEORETICAL DISCUSSION... 4 A. Formulas U s e d... 4 B. Position of Neutral A x i s... 6 C. Assumptions... 6 III. PROCEDURE FOR T E S T... 8 A. General Description... 8 B. Number of Test Series... 8 C. Properties and Equipment... 8 IV. CORRECTION FOR DEFLECTION R E A D I N G S... 2 V. R E S U L T S... 3 VI. CONCLUSION... 32 BIBLIOGRAPHY... 36 APPENDIX I... 37 APPENDIX I I... 38 APPENDIX I I I... 39

iv LIST OF FIGURES Figure No. Title Page 1. SHEAR DISTRIBUTION D I A G R A M... 7 2. LOCATION OF STRAIN R E A D I N G S... 1 3. GAGE ARRANGEMENT FOR DEFLECTION CORRECTION... 22 4. GRa PH OF LOAD VS. DEFLECTION OF SUPPORTS... 23 5. GRAPH OF PER CENT RESTRAINT VS. DEFLECTION WITH LARGER BEAM ON B O T T O M... 27 6. GRa PH OF PER CENT RESTRAINT VS. DEFLECTION WITH SMALLER BEAM ON B O T T O M... 28 7. PLOT OF STRAIN R E A D I N G S... 29 8. VIEW OF GENERAL S E T U P... 34 9. CLOSE-UP OF BEAM WITH MIDPOINT L O A D I N G... 34 1. CLOSE-UP OF BEAM WITH THIRD POINT LOADING... 34 11. CROSS SECTION OF TEST B E A M... 35 12. CONTROLS OF OLSEN TESTING M A C H I N E... 35 13. WELDING P A T T E R N... 35

V l i s t o f t a b l e s Table No. Title Page I. t a b u l a t e d d e f l e c t i o n r e a d i n g s, UNCORRECTED... 11 II. t a b u l a t e d s t r a i n m e a s u r e m e n t s... 17 III. CORRECTION D A T A... 21 IV. t a b u l a t e d v a l u e s o f r e s u l t s... 24

v i ACKNOWLEDGEMENT The author is indebted to Professor James O. Smith for his guidance and helpful suggestions on both the experimental and analytical phases of this investigation. The author also wishes to thank Professor Clyde Kesler for his efforts in arranging for the use of the testing equipment and to Mr. Ernest Ryner for his shop work and welding that was needed during the tests. The material used in the tests was provided by the Department of Theoretical and Applied Mechanics.

1 I. INTRODUCTION The investigation of shear interaction between two I-beams joined together, one above the other, by varying amounts of weld is the subject of this paper. was limited to two I-sections. The beams used in this experiment This shape of section and the wide flange section are the most common used in structural design practice which require, on occasion, to be reinforced. The subject of shear interaction is important when reinforcing an I-beam or wide flange beam with a similar section under it. This type of reinforcing is sometimes used in practice. Old industrial buildings which need stronger floor beams or girders, due to changing load requirements, create this condition. The question of how much weld, rivets, or bolts should be used to tie the two members together so they act as an equivalent solid section should be answered by this investigation. Some engineers believe that if enough restraint is given to take care of the horizontal shear, the beam will have the stiffness of a solid beam of the same cross section. The deflection should then be about the same as for an equivalent solid section and the carrying capacity of the beam could be determined by the Mc/l relationship. Commenting on the strength of wood beams made up of two separate sections and bolted together, S. Timoshenko states in his book on Strength of Materials1* that beams of solid section * Denotes the reference number in the bibliography.

of the sane dimensions are stronger. 2 Experiments made have shown built up wood beams to have about 75 per cent of the strength of the solid beam. This may or may not be true of steel beams welded together. This writer found no written material which dealt directly with this subject. The closest approach to the problem was an investigation of the deflection of laminated beams consisting of two layers bolted together at one end and in the middle. 2 The author L. G. Clark concluded that two layer laminations joined at an infinite number of points reduced the problem to that of a simple beam. A beam made up of two beams connected only at a small per cent of their common plane will have actually two types of cross- sections. One will be joined together to form a section closely resembling a solid section. See Figure 11. The other will still remain two independent unconnected sections, one above the other. In the connected section, shear will flow through the welds and the interaction will take place. In the unconnected section, shear cannot be transmitted across the common plane except the small amount that friction might contribute. Just how much influence this lack of interaction might have on the total deflection of the beam, is one of the answers sought in the experimental tests. Notations used in this thesis. a = Area of cross section b = Width of cross section d = Depth of cross section E = Modulus of elasticity.

3 G L I M P t Us V = Shearing modulus of elasticity = Total length of span = Moment of inertia = Bending moment = Total load = thickness = Total strain energy due to shear = Total vertical shear = Distance from neutral axis to center of area above the plane of shear computed = Total deflection = Deflection due to flexure = Deflection due to shear = Deflection due discontinuous shear interaction unknown causes = Unit stress = Unit shear dx dy = Differential length = Differential width

4 I I. THEORETICAL DISCUSSION The usual formula derived for deflection of an initially straight beam is based on the bending moments only. Experiments at the University of Washington3 show that for concentrated loads on short span beams shear can contribute over fifty per cent of the total measured deflection. In this problem, the depth to span ratio will always be large, which means a short span beam. The original unreinforced beam would, on the average, have about a one to twenty depth to span ratio. The A.I.S.C. Code4 limits the span to one to twenty-four of its depth. Thus, if the original beam is reinforced by another beam under it of approximately the same depth, the depth to span ratio will be increased twice its original amount. Beams with depth to span ratios of of one-twelfth or more, can be classified as short span members. The theoretical determination of the total deflection of the double I-beam is the sum of the deflections due to bending moment and the deflection due to shear plus deflection due to the discontinuous shearing action between the unconnected beam section. Total deflection The first two terms can be determined by proven equations.5 Deflection due to flexure may be found from the following equation: For concentrated load at the middle this becomes: deflection

For concentrated load at the third points this becomes: deflection = 5 The deflection of the beam due to the vertical shearing forces can be determined using Castigliano's theorem for expressing energy.6 The strain energy due to shear is: The cross section shape of two I-beams welded together produces an irregular shaped section. The unit shear, T, was determined at a number of points on cross section and plotted on a graph. See Figure 1. The average value of T for each part of the cross section was estimated from the graph and the value substituted into the above strain energy equation. See Appendix II. dx dx dy b Substituting the numerical values into the above equations and solving (see calculations, Appendix III) the following equations for shear are obtained. shear for midpoint concentrated load. shear for 1/3 point concentrated loads.

The deflection is the unknown term and its value is 6 determined from the tests made on the beams loaded under two types of loading. Data was obtained using a concentrated load at the midpoint and also with concentrated loads at the third points. The will be the difference between the total measured load and the sum of the computed deflections due to flexure and shear. When the laminated beam is fully restrained and approaches the stiffness of a solid section, its neutral axis will fall on the theoretical neutral axis of the section. If the beams are not restrained but rest one of the other, each beam will have its own neutral axis and the stresses will be distributed accordingly. As restraint in the form of welding causes shear interaction between the two beams, the neutral axis of each beam should approach the neutral axis of the fully restrained beam or solid section. The following assumptions are made in this investigation: (1) Planer distribution of stress. (2) The distance between the two beams is very small, so the second moment of area (moment of inertia) is and remains constant. (3) Friction is neglected, though it does play a part. (4) The amount of weld in the first series of welding will not fail due to horizontal shear.

FIGURE I. SHEAR DISTRIBUTION DIAGRAM 7

8 III. PROCEDURE FOR TEST The experimental teste were carried out in Talbot Laboratory at the University of Illinois. The beams used were as described in the introduction. They were cut to a length of 8 ' 6" and set up to span 8 1 ". This left a 3" overhang for bearing. The supports consisted of a fixed round block at one end and a 2 " diameter free roller at the other. See Figure 8. The tests were divided into thirteen different series. Each series varied as to span and amount of restraint (length of weld) between the beams. Three different spans were tested. The first six series tested the beams spanning 8 ' ". Three series were made on the beams spanning 6 ' * and four series on beams spanning 4' 6". Material, properties, loading conditions, and equipment: Material Structural steel - rolled section Modulus of elasticity: 3,,psi Shearing modulus of elasticity: 12,,psi Types of loading (a) Concentrated load at center of span. See Figure 9. (b) Concentrated load at third points of span. See Figure 1. Speed of loading:.3" per second Speed of unloading: same Gages used: Ames dial gages (least reading.1") for deflections.

Berry Strain Gage (4" length) for strains. Position of beams (a) Small beam. See Figure 2a. (b) Large beam. See Figure 2b.

1 (a) Large Beam On Bottom (b) Small Beam On. Bottom FIGURE 2. LOCATION OF STRAIN READINGS

TABLE I. TABULATED DEFLECTION READINGS, UNCORRECTED 11 Total Load Position Arrangement Amount of Weld Deflection in Pounds of Load of Beams Inches: Per Cent in Inches (Uncorrected) Series la Span 96" third point Larger None. 25.81 5.133 75.186 Series lb Span 96" midpoint Larger None. 25.98 5.158 75 no reading Series lc Span 96" third point Smaller None. 25.51 5.14 75 no reading Series Id 25 4 75 Span 96" midpoint Smaller None..61.1 no reading Series 2a 25 5 75 1 Span 96" third point Smaller (1 2 " ) 25%..3.62.93.123 Series 2b 25 5 75 1 Span 96" midpoint Smaller (12 " ) 25%..36.73.19 no reading Series 2c 25 5 75 1 Span 96" midpoint Larger (1 2 " ) 25*..37.74.115 no reading

TABLE I (continued) 12 Total Load in Pounds Position of Load Arrangement of Beams Amount Inches: of Weld Per Cent Deflection in Inches (Uncorrected) Series 2d 25 5 75 1 Span 96" third point Larger (12") 25%..33.53.94.126 Series 3a 25 5 75 1 Span 96" third point Smaller (24") 5%.3.59.88.117 Series 3b 25 5 75 Span 96" midpoint Smaller (24") 5%.355.69.13 Series 3c 25 5 75 Span 96" midpoint Larger (24") 5%.35.7.14 Series 3d 25 5 75 1 Span 96" third point Larger (24") 5%.3.595.88.117 Series 4a 25 5 7 1 Span 96" third point Smaller (36") 75%.35.59.82.115 Series 4b 25 5 75 Span 96* midpoint Smaller (36") 75%.355.685.1

13 TABLE I (con tin u ed ) Total Load in Pounds Position of Load Arrangement of Beams Amount Inches: of Weld Per Cent Deflection in Inches (Uncorrected) Series 4c 25 5 75 Span 96" midpoint Larger (36") 75%.35.69.12 Series 4d 25 5 75 1 Span 96" third point Larger (36") 75%.3.59.885.1165 Series 5a 25C 5 75 1 Span 96" third point Smaller (48" ) 1%..31.56.87.115 Series 5b 25 5 75 Span 96" midpoint Smaller (48") 1%.36.68.12 Series 5c 25 5 75 Span 96" midpoint Larger (48" ) 1%.38.715.155 Series 5d 25 5 75 1 Span 96" third point Larger (48") 1%.31.6.89.117 Series 6a 25 5 75 1 Span 96" third point Smaller Continuous solid weld.31.6.89.1175

TABLE I (continued) 14 Total Load in Pounds Position Arrangement Amount of Weld Deflection of Load of Beams Inches: Per Cent in Inches (Uncorrected) Series 6b 25 5 75 Span 96" midpoint Smaller Continuous solid weld..375.7.13 Series 7a 25 5 75 Span 72" midpoint Smaller Continuous solid weld Series 7b 25 5 75 1 Span 72" third point Smaller Continuous solid weld Series 7c 25 5 75 1 Span 72" third point Larger Continuous solid weld Series 7d 25 5 75 Span 72" midpoint Larger Continuous solid weld.19.37.55 Series 8 and 9 were abandoned because as weld was removed the beam bowed, giving the deflection readings too great an error. Series 1a 25 5 75 Series 1b 25 5 75 1 Span 54" midpoint Larger Span 54" third point Smaller 6(") 22.2'.11.12.35 (6") 2 2.2#.17.32.46.61.1.19.28.38

TABLE I (continued) 15 Total Load in Pounds Position of Load Arrangement of Beams Amount Inches: of Weld Per Cent Deflection in Inches (Uncorrected) Series 11a 25 5 75 1 Span 54" midpoint Smaller (1 2 ") 44.4% Series llb 25 5 75 1 Span 54" third point Smaller (1 2 " ) 44.4%..95.18.27.355 Series 11c 25 5 75 1 Span 54" midpoint Larger (1 2 ") 44.4*. 11. 2 1.31.4 Series lld 25 5 75 1 Span 54" third point Larger (1 2 ") 44.4* Series 12a 25 5 75 1 Span 54" midpoint Smaller (18") 67* Series 12b 25 5 75 1 Span 54- third point Smaller (18") 67* Series 12c 25 5 75 1 Span 54- midpoint Larger (18") 67*. 12.215.31.4

16 TABLE I (continued) Total Load Position Arrangement Amount of Weld Deflection in Pounds of Load of Beams Inches : Per Cent in Inches (Uncorrected) Series 12d Span 54" third point Larger (18") 67* 25.11 5.2 75.29 1.37 Series 13a Span 54" midpoint Smaller (24" ) 89% 25.9 5.18 75.27 1.36 Series 13b Span 54" third point Smaller (24") 89% 25.7 5.15 75.24 1.31

TABLE II. TABULATED STRAIN MEASUREMENTS 1 7 Strain readings were taken at twelve points on the beam. Location of these points is shown in Figure 2. Total Load in Strain Measurements; inches in a 4-inch gauge length. Pounds a b c d e f g h j k l m Series la. Third point load. No weld or restraint. Large beam..3.8.9.14.18.7.21.4.48.84.86.62 75.1. *.3.18.18. *.27.1.4.87.91.56.3. *.9.13.17. *.21..47.84.91.62 Series lb. Midpoint load. No weld or restraint. Large beam..2. *.9.12.16.7.21.99.47.84.93.63 5.7. *.3.1.16. *.27.98.4.88.92.6.3. *.9.12.15. *.2.99.47.92.92.63 Series 1c. Third point load. No weld or restraint. Small beam..4.53. *.38. *.41.15.38.98.54.35.82 5.5.54. *.43. *.39.17.29.94.58.35.78.45.55. *.39. *. *.56.29.99.54.36.83 Series ld.midpoint load. No weld or restraint. Small beam.. 1 1. 1 6. *. 2 *.7.8.92.63.2..47 4. 1 6. 1 6. *.5. *.5.84.92.6.26.2.42. 1 1. 1 6. *. 2. *.7.82.92.63.21..47 Series 2a.Third point load.length of weld 12".Small beam..53.59.9.43.81.51.2.34.93.6.43.89 1.62.65.89.47.78.46.27.39.88.63.4.82.53.6.91.46.82.51.2.35.85.6.42.89 Series 2b. Midpoint loading.length of weld 12". Small beam..55.6.91.44.75.53.21.34.35.6.83.89 75.6.63.88.45.75.45.24.38.89.62.86.8.53.59.91.45.74.51.21.35.9.6.86.88 Series 2c. Midpoint load. Length of weld 12". Large beam..28.55.23.73.38.37.66.62.37.62.13.93 75.37.67.29.74.36.32.73.6.33.6.8.88.29.63.3.75.41.39.66.62.36.64.12.97 Series 2d. Third point load. Length of weld 12". Large beam..29.61.24.75.42.4.63.57.33.7.1.9 1.41.6.25.73.39.36.73.6.34.71.9.89.33.6.24.78.42.38.65.59.35.65.12.97 Series 3a. Third point load. Length of weld 24". Small beam..52.56.9.4.71.48.2.21.86.58.76.84 1.6.6.9.4.69.4.23.36.89.6.77.76.5.54.92.39.69.48.19.31.81.57.81.84 * = No reading taken.

TABLE II (continued) 18 Total Load in Pounds Strain Measurements; inches in a 4-inch gauge length. a b c d e f g h j k l m Series 3b. Third point load. Length of weld 24". Small beam on too..53.57.93.41.72.5.19.33.92.6.82.86 75.6.61.91.42.7.43.24.39.91.63.86.78.52.56.93.39.7.49.18.33.89.59.8.85 Series 3c. Midpoint load. Length of weld 24". Large beam..47.73.38.94.59.51.82.7.54.86.31.14 75.52.73.39.94.56.47.89.73.52.8.26.6.53.74.39.95.6.58.82.73.54.85.3.13 Series 3d. Third point load. Length of weld 24". Large beam on too..45.73.4.95.58.54.82.8.56.82.31.13 1.55.75.37.95.53.49.9.78.54.85.24.5.46.73.4.91.59.54.81.77.54.84.28.12 Series 4a. Third point load. Length of weld 36". Small beam..5.55.86.31.65.45.13.28.86.53.64.79 1.54.48.34.31.64.36.17.32.84.52.72.72.49.54.86.36.65.46.12.29.86.54.74.81 Series 4b. Midpoint load. Length of weld 36". Small beam..51.55.9.35.66.47.12.32.88.57.78.84 75.56.59.9.33.65.4.2.34.89.54.73.74.5.56.88.35.66.46.16.31.86.56.78.83 Series 4c. Midpoint load. Length of weld 36". Large beam..49.73.38.94.6.57.8.77.53.86.33.1 75.54.7.34.87.55.48.85.77.5.84.23.4.52.73.39.89.59.55.79.77.52.86.27.11 Series 4d. Third point load. Length of weld 36". Large beam..48.71.35.9.62.54.79.77.52.85.28.1 1.59.76.35.86.54.47.84.76.49.81.23.3.46.73.35.89.56.52.77.75.5.81.26.9 Series 5a.Third point load. Length of weld 48". Small beam..53.58.92.38.67.49.11.32.89.58.78.82.57.61.93.39.66.4.21.36.9.67.76.72.51.55.91.37.65.44.14.32.9.58.76.79 Series 5b.Midpoint. L e n g t h o f weld 48" Small beam..49.55.92.37.67.46.14.32. *.57.7.79 75.55.59.93.33.66.4.21.35.91.6.77.75.5.56.94.37.67.46.13.3.9.55.76.79 Series 5c. Midpoint load. Length of weld 48". L a rg e beam on.46.71.4.95.6.53.76.77.51.85.27. 1 75.54.73.37.94.56.49.84.76.51.85.27.6.47.73.39.96.58.53.76.76.53.88.26. 1

TABLE II (continued) 19 Total Load in Strain Measurements; inches in a 4-inch gauge length. Pounds a b c d e f g h j k l m Series 5d. Third point load. Length of weld 48", Large beam..44.69.37.95.57.5.77.75.52.87.27.1 1.52.71.36.93.52.44.83.76.52.83.25.4.45.69.37.92.56.52.75.75.52.85.27.9

2 I V. CORRECTION FOR DEFLECT ION READINGS During the tests, the deflection measurements were read on an Ames Dial Gage. This gage was fastened to a stand and set on the base of the testing machine. In order to eliminate the deflection due to the end supports and the bed of the testing machine, a set of readings were taken to determine the deflection which should be deducted from the deflection taken during the testing of the beam. This correction will give a more accurate and true deflection value for the beam. Figure 3 shows the method used to obtain these corrections. A pin was driven through the web of the beam near the theoretical neutral axis. Dial gages were mounted on each side and measured the deflection on the two sides relative to two straight edges which rested of the beam supports. A third dial gage was set to measure the deflection at the same place and under the same conditions as during the tests. Table IV gives the tabulated data for the deflection correction. Readings were taken for a number of different support conditions. The first set of data was obtained with a support built up with a roller. The second set of data was a built up support without a roller, but using a flat plate. The third set of readings was taken using a solid cast steel support which had a rounded supporting edge.

t a b l e III. CORRECTION DATA Load Dial No. 1 Dial No. 2 Dial No. 3 Deflection in in in in correction pounds inches inches inches in inches Data for built-up support with rollers. 1.5.5..3 22.5.6. 1.3 26.7.13.7.6 41. 1.19. 1 1.8 55.14.25.15. 1 68. 18.29.19.1 78.19.34.23.13 9.23.37.25.13 98.24.41.28.15 15.27.43.29.15 111.27.46.32.16 126.32.51.36.17 Data for built-up supports without rollers. 11.4.6.6.1 2.5. 1.1.3 49.17.26.23.6 56.15.25. 2 1.7 72.28.4.36.8 87.24.37.3. 1 15.29.44.35. 12 145.4.6.47.16 Data for solid cast steel support. 2.5.9.7.3 355. 1.16. 1 1.6 62.17.23.2.4 63.17.27.19.9 93.26.39.27.13 99.29.37.28.9 11.32.46.32.14 121.34.45.34. 1 1

22 FIGURE 3. GAGE ARRANGEMENT FOR DEFLECTION CORRECTION

23 Deflection in Supports; I/IOOO Inches 2 4 6 8 1 Total Load in 1 Pounds FIGURE 4. LOAD vs DEFLECTION IN SUPPORTS

TABLE IV. TABULATED VALUES OF RESULTS 24 Tabulated results of the load at the third points with the larger beam (6 inch I-beam) on the bottom. These tabulated results are also shown on the graph in Figures 5 and 6. The total deflection is the corrected deflection using the correction indicated for each load on graph in Figure 4. Amount of Load Span Measured Calculated Discontinued Weld in in in Deflection Deflection Interaction Per Cent Lbs. Inches in Inches Bending Shear Deflection 25 5 96.56.44.3.9 5.53.6 75.53.6 1.5.3 Solid.54.7 25 75 96.83.67.5.11 5.78.6 75 7 96.74.65.5.3 1.77.67.5.5 Solid.79 25 1 96.11.88.7.15 5.14.9 75.12.7 1.12.7 Solid.14.9 22 5 54.14.8. 2.4 44.13.3 67.13.3 89. 1. 22 75 54.19. 1 2.3.4 44.18.3 67.18.3 89.15. 22 1 54.25.16.4.5 44.24.4 67.23.2 89.2. Solid weld indicates that the intermittant welds were filled producing a continuous solid weld on each side of the beam.

25 TABLE IV Tabulated values of the (continued) load at the midpoint of the span with larger beam on the bottom. Amount of Load Span Measured Calculated Discontinued Weld in in in Deflection Deflection Interaction Per Cent Lbs. Inches in Inches Bending Shear Deflection 25 5 96.67.52.5. 1 5.63.6 75.62.5 1.62.5 Solid.64.7 25 75 96.99.78.8.13 5.93.7 75.9.4 1.92.6 Solid.93.7 22 5 54.15.9.3.3 44.15.3 67.13.1 89. 12. 22 75 54. 2.14.3.3 44. 2 1.4 67.19.2 89.17. 44 1 54.28.18.6.4 67.24. 89.24. Tabulated values of the load at third point of span with the smaller beam (4 inch I-beam) on the. bottom. 25 5 96.58.44.3. 1 1 5.54.7 75.54.7 1.55.8 25 75 96.85.67.5.13 5.79.7 75.79.7 1.8.8

26 TABLE IV (Continued) Amount of Weld in Per Cent Load in Lbs. Span in Inches Measured Deflection in Inches Calculated Deflection Bending Shear Discontinued Interaction Deflection 25 5 75 1 22 44 67 89 22 44 67 89 22 44 67 89 1 96.12.1 1 1.1 1.1 1 1 5 54 no reading.15.14.14 75 54 no reading.23.23.2 1 54 no reading.26.25.25.88.7.25.16.15.16.9.3.3.2.2.14.4.5.5.2.18.6.2. 1.1 Tabulated values of the load at midpoint of span with the smaller beam (4 inch I-beam) on the bottom. 25 5 75 1 25 5 75 1 5 96.68.64.63.65 75 96.15.94.92.95.52.5. 11.7.6.8.78.8.19.8.6.9

27 Deflection in Inches Ab + A s + A uk FIGURE 5. PER CENT RESTRAINT VS DEFLECTION

Total Deflection in inches A b + A S + A uk 28

FIGURE 7. PLOT OF STRAIN READINGS 29

3 V. RESULTS From the experimentatl data obtained, it appears that the efficiency of the reinforced beam spanning 96 inches reaches a maximum when restrained approximately 5 to 6 per cent. Shear interaction is complete enough to give approximately 9 per cent the stiffness of an equivalent solid section. The data shows that greater percentage of restraint is required by the shorter beam which has a span of 54 inches. The maximum efficiency of the shorter span occurs when the two beams are about 3 to 9 per cent restrained. When the small beam is on the bottom, however, the curves show not much gain after 67 per cent restraint. The increased restraint needed for the shorter beams indicates that shear becomes more critical as the depth to span ratio increases. The deflection, A Uk, due to discontinuous interaction varies a great deal. In the short span it disappears when 89 per cent restraint is given the laminated beam. This would indicate 1 per cent efficiency compared with a similar solid section. Due to the inaccuracy in measurement and other causes, such as, residual stresses and non-uniformity of the beams, the A uk disappeared by choice. The correction factor read from the graph in Figure 3 could have been chosen, because of spread of plotted points, to reduce the corrected total deflection so that A uk would become a negative value. This would not be unreasonable, because warping of either beam during welding can and does pre-stress the members and stiffens the laminated beam.

31 The data also indicates that the most efficient arrangement calls for the larger beam to be placed on the bottom. This arrangement caused less deflection. In general, the tabulated values and the charts show that the depth to length of span relationship and the intensity of load influence the percentage of weld needed to approach a value of the solid section. The curves show a steep drop from the no weld condition. This indicates a good gain in stiffness when the beams are caused to act together. The strain readings, though they were crudely measured, indicate that the strains act as stated in the theoretical discussion. The plot of strain readings shown in Figure 7 is a fair indication that the neutral axis of the solid section is the neutral axis of the combined beams section when full interaction is obtained. It shows that partial interaction splits the neutral axis and each moves toward its own independent beam axis. The presence of residual stresses tends to interfere with the stresses due to load and scatters the readings. This makes it difficult to get a good stress distribution diagram.

32 VI. CONCLUSION The results of the experiment are adequate to justify the following conclusions. To reinforce an I-beam or wide flange beam it is necessary to weld the flanges together 5 to 6 per cent of their span distance. For very short beams this will increase to approximately 8 per cent. This amount will give enough interaction to obtain good efficiency from the beam. It also indicates that for static loading, intermittant welding gives good results. However, in actual practice, it will be better to weld continuously along both sides of the beam. Continuous welding will shut out moisture and stop corrosion between the flanges. The fatigue strength of intermittant and continuous welds are about the same for to tension loads at low number of cycles but continuous welds are better for a large number of 7 cycles. It can be pointed out that in this type of beam reinforcing the position of the weld is near the neutral axis where the stresses in the elastic range are not very great. The conclusions made from the results of the test must be limited to the range of dimensions which the beams covered. The loading conditions consisted of concentrated static loads. Concentrated loads cause more shear deflection than the loading usually existing in a structure. Building floor loads usually are uniform loads. However, the concentrated loads put the results on the side of safety. Within the working range the equation r = Mc/I will give a satisfactory answer in determining

the section of beam needed to satisfy a greater load requirement. The amount of error in the readings taken, no doubt, 33 influenced the results to some extent. Since no effort was made to control the dimensions of the beam or its properties the existance of practical conditions was not removed. In any future investigation it would be better to provide a means of obtaining deflection readings of the beam, which would not include the deflection due to supports or deformation of the web over the supports. In terms of further investigation of shear interaction, it would be interesting to learn the influence of fixed end conditions on the action of a reinforced beam as described in the introduction.

34 FIGURE 8. GENERAL SET UP OF TEST FIGURE 9. CLOSE UP OF BEAM WITH MIDPOINT LOADING FIGURE 1. CLOSE UP OF BEAM WITH THIRD POINT LOADING

35 FIGURE II. CROSS SECTION OF LAMINATED BEAM FIGURE 12. CONTROLS OF OLSEN TESTING MACHINE FIGURE 13. WELDING PATTERN

36 BIBLIOGRAPHY 1. Timoshenko, S., Strength of Materials, p. 127, New York, D. Van Nostrand Company, Inc. 2. Sergev, Sergius, "The Effect of Shearing Forces on the Deflection and Strength of a Beam," University of Washington Experimental Station Bulletin No. 114, Seattle, 1947. 3. Clark, L. G., "Deflection of Laminated Beams," Separate No. 331 Proceeding of A.S.C.E., November, 1953. 4. American Institute of Steel Construction, "Manual," p. 283, New York, 1951. 5. Seely, Fred B., Resistance of Materials, p. 137ff, New York, John Wiley and Sons, 1947. 6. Seely, Fred B. and Smith, James O., Advanced Strength of Materials, p. 442f, New York, 1952. 7. Wilson, Wilbur M., and Munse, William H., "Fatigue Strength of Various Details Used for Repair of Bridge Members," p. 38, Bulletin 382 Engineering Experiment Station, University of Illinois, 1949.

37 APPE N D I X I Calculations to determine the position of the neutral axis of the laminated beam and the moment of inertia of the combined section Moment of Inertia of 6" I-Beam Area = Area of two flanges plus area of the web Moment of Inertia, "I "equals For the 4 " I-Beam Area = 2.21sq. in. (A.I.S.C.) Moment of Inertia = 6. in.4 (A.I.S.C.) Location of neutral axis: Area Arm Moment Distance from base to neutral axis = 4.85 inches Moment of Inertia of combined section; by transfer method Moment of Inertia, Total - 59. 2 in.4

APPENDIX II 38 Calculation of shear,, in the cross section of the- beam. The shear interaction is provided by the weld. Shear flow is assumed to divide at the center flanges and one-half of its value flow each way. The values are plotted in Figure 1. Upper flange: V(.656 x.312 x 4.89 ) = 3.2V I x (.312) I Upper web and upper flange: V(2.62 x.312 x 4.89) = 21.5V I x (.18 7) = I Upper web and lower flange: V (3.37 X.1 87 X 3. 8 ) + 4.2V = 3 1. 8V I x (.1 8 7 ) I Lower flange of upper bean: V(.312 x 1.3 1 x 1.27) + 5.97V = 11.2V I x (.312) 2 I Top flange of bottom beam : V(.75 x.34 x.91) + 6.49V = 1.3V I x (O.314) 2 I Upper web of lower beam: V(2. 62 x.31 x 1.27+3 x.3k x.91) + (5.97V) = 26.8v I x (.2 9 5 ) I Lower web of lower beam: V(3 x.34 x 4.68) = 16.2V I x.295 I Lower flange of lower beam: V (.7 5 x. 34 x4.6 8 ) = 3.5 V I x (. 34 ) I

39 APPENDIX III Calculation of strain energy duo to shear; and the deflec tion resulting from the shearing action. The unit shear,, used is the average reading chosen from the shear diagram in Figure 1. The equation for strain energy due to shear: Top flange = Web (Upper) = Center flanges Web (Lower) Lower flange Total strain energy due to shear, Total Shear V Midpoint and third point loading for midpoint load for third point loads