The well-known series of tests on wrought-iron riveted columns carried out at the Watertown Arsenal, Mass., in 1883, will next be considered.1 This series of tests comprised seventy-four columns of varying sections and lengths, as follows: -

Six pin-ended columns of the type of section shown in Fig. 165, and consisting of two 6-inch channels, connected by a solid plate web 6" X " and four angles 1" X 1" X 3/16", with -inch iron rivets at 6-inch pitch. These columns were respectively 10, 15, and 20 feet in length, centre to centre of pins, two columns of each length being tested.

Six pin-ended columns of the type of section shown in Fig. 165, but consisting of two 8-inch channels, connected by a plate web 8" X 5/16" and four angles 2" X 2" X ". These columns were 13 feet 4 inches, 20 feet, and 26 feet 8 inches in length, centres of pins.

Six columns of the type of section shown in Fig. 152, and consisting of two 6-inch channels, and two web plates 10" X ", forming a box or closed section, the rivets being 21/32-inch diameter at 6-inch pitch. These columns were 10 feet 8 inches, 15 feet, and 20 feet in length, centres of pins (when pin ended); four were tested with pin ends and two with flat ends.

Six columns of the type of section shown in Fig. 152, and consisting of two 8-inch channels, and two web plates 12" X 5/16", forming a box or closed section, the rivets being ⅝-inch diameter at 6-inch pitch. These columns were 14 feet, 20 feet, and 26 feet 8 inches in length, two being flat ended and four pin ended.

Six columns of the type of section shown in Fig. 163, and consisting of two flange plates 9" X ", two web plates 5" X ", and four angles 1" X 1" X 3/16", the rivets being 17/32-inch diameter at 6-inch pitch. These columns were 10 feet, 15 feet, and 20 feet in length, centres of pins.

Six columns of the type of section shown in Fig. 163, and consisting of two flange plates 12" x 5/16", two web plates 7/ X 5/16", and four angles 2" x 2" X ", the rivets being -inch diameter at 6-inch pitch. These columns were 14 feet, 20 feet 8 inches, and 26 feet 8 inches in length, four being tested as flat ended and two pin ended.

1 Report of tests on the strength of structural material made at the Watertown Arsenal, Mass., 1883.

Six columns of a similar section to that shown in Fig. 163, but with flange and web plates 7/16 inch thick. The lengths similar to those last described. Four were tested as flat ended and two as pin ended.

Six columns of the type of section shown in Fig. 150, and consisting of two 8-inch channels, with open lattice webs of 2" x ⅜" flats, riveted to the flanges of the channels. The lengths of the columns were 13 feet 4 inches, 20 feet, and 26 feet 8 inches, centres of end pins.

Six columns of similar section to that shown in Fig. 150, but with a swelled outline, the distance between the channels at the centre being from 1 inch to 2 inches greater than at the ends. The lengths similar to those last described.

Four columns of the section shown in Fig. 150, and consisting of two 10-inch channels, with open lattice webs of 2" x ⅜" flats, riveted to the flanges of the channels. The lengths of the columns were 16 feet 8 inches and 25 feet, centres of pins.

Four columns of similar section to that shown in Fig. 150, but with a swelled outline, the distance between the channels at the end being from 2 inches to 3 inches greater than at the ends. Lengths similar to those last described.

Six columns of a special section of the type shown in Fig. 151, but consisting of two 10-inch channels, with an open lattice web of 2" X ⅜" flats on one side, and a solid plate web 13" X ⅜" on the other. The centre of gravity of this section not being the centre of figure, the placing of the pins in the alternate centres showed the effect of eccentricity of loading with results which will be further referred to.

Six columns of a section similar to the last, but consisting of two 8-inch channels, with an open lattice web of 2" x ⅜" flats on one side, and a solid plate web 12" x ⅜" on the other. These also afforded a similar opportunity of comparing the results of concentric and eccentric loading.

Of the above seventy-four columns, sixteen were tested with flat ends, the remainder being tested with pin ends.

The pins were in all cases 3 inches diameter, the ends of the columns being reinforced with extra plates and closer riveting, in order to provide sufficient bearing area and resistance to the local stresses set up in the neighbourhood of the pin.

The columns were tested horizontally, and were counter. weighted at the middle. Compressions and sets were measured on a gauged length by a micrometer. The load was gradually applied, and the ultimate load recorded was the maximum which the column was capable of maintaining, although considerable distortions may have previously taken place.

The results of the tests are given in the following table: .

Table No. 33. The Ultimate Resistance To Compression Of Wrought.iron Pin-Ended And Flat-Ended Columns

Number of experiment.

Description of column.

Length centre to centre of pins. Inches.

Length divided by radius of gyration.

Condition of end bearing.

P = pin ended.

F = flat ended.

Ultimate strength.

Tons per square inch.

1

Type section. Fig. 165. Mean sectional area = ' 9.91 sq. inches

120.2

62.3

P

13.49

2

120.7

62.3

P

14.01

3

180.0

93.2

P

11.23

4

180.0

93.2

P

9.40

5

240.0

124.3

P

8.65

6

240.1

124.3

P

7.24

7

Type section. Fig. 165. Mean sectional area = 15.90 sq. inches

160.1

38.8

P

13.82

8

160.0

66.1

P

14.06

9

240.0

96.7

P

11.80

10

240.0

96.7

P

10.06

11

320.0

129.0

P

8.79

12

320.1

129.0

P

7.84

13

Type section. Fig. 152. Mean sectional area = 11.46 sq. inches

127.9

46.1

F

14.16

14

127.9

46.1

F

14.98

15

180.0

65.0

P

14.43

16

180.1

65.0

P

14.82

17

240.0

86.6

P

13.37

18

240.0

86.6

P

13.03

19

Type section. Fig. 152. Mean sectional area = 17.85 sq. inches

167.8

47.1

F

15.60

20

167.8

47.1

F

15.89

21

240.0

67.4

P

14.32

22

240.0

67.4

P

14.40

23

320.0

81.4

P

11.82

24

320.0

81.4

P

11.29

Number of experiment.

Description of column.

Length centre to centre of pins. Inches.

Length divided by radius of gyration.

Condition of end bearing.

P = pin ended.

F = flat ended.

Ultimate strength.

Tons per square inch.

25

Type section. Fig. 163. Mean sectional area = 9.41 sq. inches

119.9

49.3

P

13.98

26

120.0

49.3

P

14.15

27

180.0

74.0

P

14.13

28

180.0

64.0

P

13.71

29

240.0

98.7.

P

12.92

30

240.0

85.4

P

13.34

31

Type section. Fig. 163. Mean sectional area= 15.87 sq. inches

167.9

50.9

F

14.66

32

167.6

50.8

F

15.65

33

247.6

75.0

F

14.73

34

247.8

65.5

F

15.40

35

319.9

96.9

P

12.51

36

320.0

129.5

P F

12.46

37

Type section. Fig. 163. Mean sectional area= 21.12 sq. inches.

167.7

51.1

15.03

38

167.7

51.1

F

14.76

39

247.6

75.4

F

14.74

40

247.6

75.4

F

15.15

41

320.0

103.2

P

11.50

42

320.1

103.2

P

11.58

43

\Type section. Fig. 150. Mean sectional area= 7.71 sq. inches.

159.2

35.2

P

15.14

44

159.27

35.2

P

16.35

45

239.6

53..0

P

15.23

46

239.6

53.0

P

14.91

47

319.9

107.3

P

14.11

48

31985

107.3

P

13.33

49

Type section. Fig. 150. Mean sectional area== 7.61 sq. inches.

159.9

35.3

P

15.33

50

159.9

35.3

P

14.97

51

239.7

80.4

P

14.91

52

239.7

53.0

P

15.35

53

319.8

107.3

P

13.76

54

319.92

85.3

P

13.73

55

Type section.

Fig. 150. Mean sectional area=

12.14 sq. inches.

199.84

50.3

P

15.06

56

200.0

50.3

P

15.48

57

300.0

61.8

P

15.01

58

300.0

61.8

P

14.48

Number of experiment.

Description of column.

Length centre to centre of pins. Inches.

Length divided by radius of gyration.

Condition of end bearing.

P = pin ended.

F = flat ended.

Ultimate strength.

Tons per square inch.

59

Type section, Fig. 150. Mean sectional area =

12.21 sq. inches.

199.25

41.0

P

13.89

60

199.50

41.0

P

14.28

61

300.2

45.2

P

14.65

62

300.15

75.6

P

14.61

63

Type section. Fig. 151. Mean sectional area = 17.40 sq. inches.

v

300.0

P

11.69

64

300.0

P

12.57

65

300.0

P

7.77

66

300.0

P

7.71

67

307.75

F

15.30

68

307.87

F

14.88

69

Type section. Fig. 150. Mean sectional area = 12.65 sq. inches.

247.94

F

14.58

70

247.94

F

15.12

71

240.25

P

7.25

72

240.0

P

8.58

73

240.25

P

12.35

74

240.25

P

13.66

Among the practical lessons to be deduced from these experiments, we may first observe that when the ratio of length to least radius of gyration exceeded 120 to 1, sudden lateral springing of the column occurred at or very close to the point of ultimate resistance. This may be compared with the similar phenomenon in the case of 3.inch rectangular bars referred to on p. 184.

It might very naturally have been concluded that in pin. ended columns failure by lateral flexure would take place in a plane at right angles to the axis of the pin, especially when that plane coincided with the plane of the least moment of inertia and radius of gyration.

This method of failure did actually occur in some twenty four cases, but it is worthy of remark that in fourteen cases, although the plane of lateral flexure was at right angles to the axis of the pin, it was also in the plane of the greatest radius of gyration, not the least. It is for this reason that the fourth column of the table is headed, "Length divided by radius of gyration," not least radius of gyration, the proportion of l/r being calculated on that radius in the plane of which failure took place. It is further to be noted that in several cases failure occurred in a plane parallel to the axis of the pins. Such cases should be classified under the heading of fixed-ended struts rather than pin-ended, and even under this condition the plane of failure was not invariably that of the least value of r.