Case Studies

Repair Welded Steel Rail

Background

A steel rail in a rail system servicing Central Queensland mine sites failed in service. The rail system in which the failure occurred comprised primarily 60 kg/m rail with approximately 1000 km of total track length. The failure manifested as a transverse fracture, necessitating removal and replacement of the failed section.

Initial Hypothesis

Fracture occurred due to progressive crack growth from an internal discontinuity (transverse fissure) or surface/near-surface discontinuity (detail fracture).

Data Collection

Visual examination showed that the rail had a transverse defect covering around 50% of the rail head cross-section area (Fig. 1). The transverse defect was biased toward the gauge side of the head.

Figure 1: Fracture face showing transverse defect (approximately bounded by broken lines). Arrows mark proposed direction of crack progression from internal pore.

Macroetching showed that the rail had been subject to running surface weld repair. The weld layer was considerably thicker on the gauge side versus the field side by a ratio of around 5 to 1 (Fig. 2).

Figure 2: Macroetched cross section showing differences in weld repair thickness from field side to gauge side (black bars).

A gas pore around 2 mm in diameter was seen within the weld repair region on the gauge side, near the centre of the transverse defect (Fig. 1). The maximum allowable gas pore size in an arc weld repair as per AS 1085.20 is 0.3 mm diameter.

Hardness testing showed that the hardness in the repair weld fell below the minimum requirement on the gauge side and exceeded the maximum allowable level on the field side.

Micrographs from the weld repair showed shrinkage voids and gas pores (Fig. 3). Cracking was seen to extend into the weld repair core from a surface-breaking pore (Fig. 4). The microstructure in the weld mostly comprised bainite and ferrite with minor pearlite and martensite. A soft, cracked region (240 HV) adjacent to the large pore consisted of ferrite and pearlite.

Figure 3: Unetched micrograph showing shrinkage voids with a gas pore (arrow) in the weld repair.
Figure 4: Cross-section showing pore and crack (arrow) in repair weld. Upper weld run comprises ferrite (light) and pearlite (dark) with minor bainite (circled). Lower weld run comprises majority bainite (blue, B) with minor ferrite (light).

Analysis

The repair weld did not comply with AS 1085.20 requirements in two ways: its hardness was outside the specified limits, and it contained a pore considerably larger than the maximum allowable size.

The low hardness near the large pore, combined with the stress concentrating effect of the large discontinuity, resulted in the onset of fatigue cracking in the rail head. Fatigue cracking propagated under normal operating stresses until the transverse defect covered around 50% of the rail head cross-section, at which point fast fracture occurred and the rail section separated.

Conclusion

The hypothesis is accepted — fracture occurred due to progressive crack growth from an internal discontinuity. The internal discontinuity was the 2 mm pore.

The root cause of the failure was improper design and/or execution of the weld repair procedure, resulting in the generation of numerous voids (gas pores and shrinkage) and regions with non-compliant hardness.

Recommendation

It is recommended that repair weld procedures and welding personnel be qualified by an appropriate laboratory to reduce or eliminate the risk of the production of non-compliant welds.

Visit Us
3/627 Boundary Road

Archerfield, QLD 4108

Call Us
07 3034 1110 | 0499 294 926
Send a Message
info@faceconsulting.com.au