Dimensions of the specimens and main factors

The geometry of the specimens is described in Figures 3-3 and 3-4, in the latter of which the layouts of the splices are shown. In a specimen, two prefabricated parts were bolted together, such that they formed a flanged splice for a tube. In the prefabricated part, one end of the rectangular cold-formed tube CFRHS 250x150x10 was welded to the end plate of the splice, and another end was welded to the support plate. The span between the support plates was 1 m in every specimen, such that it was measured from the centre planes. The lower edges of the nominally 20 mm thick support plates were chamfered, as shown by detail A of Figure 3-3. The specimens were assembled (bolted) by the testers in the laboratory. The manufacturer of the prefabricated parts was MR-Steel Oy from Peräseinäjoki/Finland. The dimensions of the loading shelf can be found in Appendix A of the test report (Perttola & Heinisuo 2012). The material data of the steel parts are given in Section 3.2.2. In Table 3-1, the considered six splice tests are grouped by the main factors varied in the tests. Four tests (TE1, TE2, TE3 and TE11) were carried out, such that the tube and its splice were subjected to biaxial bending. In the biaxial tests, the angle between the direction of the load (vertical) and the strong axis of the tube section was 35° (Figure 3-3). In two tests (TE7 and TE8), weak axis bending was arranged to apply on the tube and splice. The end plates were varied between the thinnest (TE1, TE7, TE11) and thickest ones (TE3, TE8), with 10 mm and 20 mm plates, respectively. Also, a splice with intermediate thickness made of 15 mm plate was tested (TE2). The measured thicknesses of the end plates are given in Table 3-3. The bolts were placed at the corners of the end plates (corner bolts) in all splices except for in test TE11, which was arranged with the “mid-side bolts”, as shown in Figure 3-4.

The bolts M20-70 (DIN 933, hexagon head full thread bolt) grade 10.9 were used in all splices. The grade of the nuts (DIN 934/10 M20 PLAIN) and the washers (DIN 125 ZN HV300 M20) was also 10.9, which is in accordance with good practise. The heights of the screw head and the nut were 12.5 mm and 16 mm, respectively. The inner and outer radii of the 3 mm thick washers were 21 mm and 36 mm, respectively. Those bolts supposed to act as tension bolts in the splice under bending were pre-tensioned with the tightening moment of 400 Nm, whilst the bolts assumed to locate in the compression zone were tightened only to 70 Nm. Therefore, the three lowest bolts were pre-tensioned using the larger moment in biaxial tests TE2 and TE3. Moreover, the two lowest bolts were the more tightened bolts in the weak axis tests TE7 and TE8. An exception was test TE1 in which the tightening moment was 70 Nm for all bolts. In biaxial test TE11 with mid-side bolts, only the bolt in the middle of the longer higher side of the end plate was assumed not to be the tension bolt (compare with the FE analysis results shown in Figure 5-56). The gaps are related to the insufficient flatness of the end plates caused, mainly, by the welding of the tubes to the end plates.

The bolts were pre-tightened in order to minimise the gaps between the end plates, especially on the

35

tension side where the prying effect may occur. This is, presumably, more difficult in the case of thicker end plates. In practise, it is not possible to eliminate the gaps associated with the initial imperfection completely by pretension. On the other words, an ideal initial interface without gaps between the end plates cannot be achieved by the used means.

The edge of the tube wall was chamfered for the butt weld connecting the tube to the end plate. The leg lengths of the “extra fillet weld” in addition to the butt weld were measured to be 2-4 mm in the direction of the end plate and 8-12 mm in the direction of the tube wall. The welds were made by hand and perhaps more carefully than usual by the fabricator. It is noted that no fracture in the welds was observed in the tests due to poor quality. The tubes were welded to supports plates by 8 mm fillet welds.

The arranged tests stand for the splices analysed as the driving examples of the thesis as presented in Table 3-2. Tests TE1 and TE7 with 10 mm end plates differ only with the orientation of the tube and the specimen of these tests represent the same splice marked as S1 but in different loading conditions (i.e., under weak and biaxial bending). Correspondingly, tests TE3 and TE8 with the 20 mm end plate represent splice S3 under weak and biaxial bending. Splice S2 with the 15 mm end plate is represented, in turn, by test TE2 only. Splices S1, S2 and S3 with the corner bolts form a series with varying end plate thickness (stiffness) when the bolt size (M20) is fixed. The

“comparative test” TE11 without the corner bolts is marked here as S4.

Table 3-1. Tests grouped by main factors.

Loading condition End plates Bolt placing Test

Biaxial bending

10 mm plate corner bolts TE1

15 mm plate corner bolts TE2

20 mm plate corner bolts TE3

10 mm plate mid-side bolts TE11

Weak axis bending 10 mm plate corner bolts TE7

20mm plate corner bolts TE8

Table 3-2. Correspondence between the studied splices S1 to S4 and the arranged tests.

Splice

Nominal end plate thickness

[mm] Bolt placing Tests

S1 10 corner bolts TE1 (biaxial), TE7 (weak axis)

S2 15 corner bolts TE2 (biaxial)

S3 20 corner bolts TE3 (biaxial), TE8 (weak axis)

S4 10 mid-side bolts TE11 (biaxial)

36

Figure 3-3. Specimen in the biaxial and weak axis bending test. Layouts of the end plates are given in Figure 3-4.

37

Figure 3-4. End plates x 290 x 390. The diameter of the holes is 22 mm for the M20 bolts.