Failure of a long I-beam. - Fractography

Failure of a long I-beam. - Fractography - Rail defects

Figure 1: 12 m long I-beam split along the web. With kind permission of Reinhold Publishing Company.

Defect name: No data.
Record No.: 2948
Type of defect (Internal/Surface): Surface
Defect classification: Fractography, rail defects
Steel name: Steel
Steel composition in weight %: No data.
Note: When components are manufactured from metals and alloys by the various operations, an important parameter that is overlooked is the residual stress retained during fabrication. Unless adequate precautions are taken to reduce or eliminate them, residual stresses can lead to unexpected failures. This is illustrated by an incident in Belgium in 1934. A 12 m long I-beam of high-strength steel fractured suddenly while lying flat on the shop floor, with an explosive noise. Figure 1 shows the beam, which had split along the entire length of the web. Both halves had curved with an outside concavity of the flanges, with deflections up to 8 cm. On the day before, skew cuts had been made at its two ends with a torch. The spontaneous fracture was attributed to the state of stresses introduced during the hot rolling of the beam. The web cooled faster than the flanges. The contraction of the flanges by cooling was prevented by the web, which induced high stresses in the beam. This state of precarious internal equilibrium was disturbed by the flame cutting. It is believed that under the heavy state of residual stresses, flame cutting followed by cooling produces a kind of artificial strain aging and hardening locally and generates cracks. These cracks must have grown slowly, until reaching the critical length.
A somewhat similar beam failure was reported at the Indira Gandhi Centre for Atomic Research, Kalpakkam, the crack in this case following a path along which there was excessive sulfur segregation.
Reference: Not shown in this demonstration version.