The flashing begins at small contacts

Flash-butt welding places two rail ends opposite one another and passes current through their contact points. Those small contacts have high current density, so material heats, melts and is expelled in the characteristic shower of particles. The ends progressively shorten as this flashing continues. TWI describes the process as resistance heating supplied largely by the workpieces themselves. The visible sparks are therefore part of a controlled way to heat the joint faces and remove contaminated material, rather than a separate filler being deposited between the rails.

Sources: weld-twi

A forging force consolidates the interface

After the planned burn-off, the machine pushes the heated ends together with a substantial forging force. This upsetting stage consolidates the interface. TWI distinguishes the resulting forge butt weld from a joint that retains a pool of melted metal between the parts. Heating and pressure have complementary jobs: the first prepares a softened zone and clears the faces; the second brings the material together. Simply making two ends hot would not describe the whole joining mechanism. The force and the progression of the cycle matter too.

Sources: weld-twi

Displaced metal forms a removable collar

The upsetting stroke moves material outward around the rail section. Plasser & Theurer describes trimming this upset metal immediately after the stroke in its mobile welding process. This explains why making the bond and restoring the rail’s outline are separate stages. The machine has produced excess material at the joint while pressing the ends together; the finished running profile cannot retain that protruding collar. The manufacturer’s account also identifies the process as using the rail material itself, without adding a separate filler metal to form the connection.

Sources: weld-plasser

Cooling changes the surrounding steel

A weld has a heat-affected zone even when no filler is added. RDSO’s rail-welding manual explains that head-hardened rail can lose hardness around a weld because cooling differs from the original head-hardening treatment. The resulting hardness variation can lead to uneven plastic deformation under wheel contact and local cupping of the running surface. Other rail grades have different treatment requirements. The metallurgical question therefore extends beyond whether the faces joined: the heated steel must recover an appropriate structure and hardness distribution for that rail grade.

Sources: weld-rdso

A sound bond still needs the right shape

A rail is also a precision running surface. Voestalpine’s account of its welding plant notes that heat can introduce small deformations, followed by straightening. Joining rails into long lengths in a factory reduces the amount of construction work required at the site, but the joint still has to fit the intended rail geometry. This is why a weld’s outward appearance and alignment are meaningful engineering properties. A connection can be metallurgically joined while still needing further work to produce the required shape along the rail.

Sources: weld-voestalpine

The joint has a manufacturing history

Weld assessment combines physical checks with a record of how the joint was made. RDSO’s manual includes dimensional, hardness, transverse-load and metallographic examinations in its quality framework. Voestalpine describes registration numbers that make individual welds traceable. Together, these approaches connect a particular joint to its manufacturing history and examination results. The useful output of automated welding is consequently more than a repeatable movement of the machine: it is an identifiable joint whose process and resulting properties can be evaluated.

Sources: weld-rdso, weld-voestalpine