ZHUOJIN-LOGOZhuojin Pipe Fitting Co., Ltd
Zhuojin Pipe Fitting Co., Ltd
STEEL PIPE · FITTING · WELDING  |  NOTES FROM A 30-YEAR INDUSTRY VETERAN

Preventing Hot Cracking in Continuous Casting Roll Surfacing Welding

The Problem as It Really Was

In 1995, when this paper was published, continuous casting was already the backbone of Chinese steel production, and the continuous casting roll was the component that everyone in the plant quietly feared. A roll failure meant an entire strand goes down, and every hour of downtime on a caster is measured in tens of thousands of tonnes of lost output. The paper from Tianjin Second Metallurgical Machinery Repair Center describes a repair technology imported from the American company P·E, using Lincoln Electric equipment and consumables. What the abstract does not tell you, but what any of us who have run a repair shop know, is that the real battle was not about the welding equipment at all. It was about the metallurgy of the interface between the roll base metal and the surfacing layer, and the fight against hot cracking that will ruin a repaired roll within weeks if you get it wrong.

The roll material is 15CrMo and 30CrMo, both low-alloy steels with a carbon content that puts them squarely in the hardenability zone where hydrogen cracking and hot cracking are both real threats. The original process used a buffer layer of Lincore 8620 flux-cored wire, 3.2 mm diameter, with Lincoln 802 flux, followed by a surface layer of Lincore 423Cr, also 3.2 mm, with the same flux. The parameters were I = 400 to 450 A, U = 30 to 32 V, and wire feed speed v = 120 to 130 mm/min. Those numbers tell me immediately that this was a submerged arc process with a fairly high heat input, and that the wire feed speed was the parameter that the operator had the most control over on the shop floor.

The Metallurgical Root Cause of Hot Cracking

Hot cracking in surfacing welds on rolls is not a simple problem. It is a combination of three factors that have to align in the worst possible way. First, the base metal has a carbon content that promotes a hard, martensitic heat-affected zone (HAZ) if the preheat and interpass temperature are not controlled. Second, the surfacing alloy, in this case a 400-series chromium stainless type, has a composition that is inherently susceptible to solidification cracking if the ferrite content is not properly balanced. Third, the roll geometry itself creates a high restraint condition, because you are welding on a large, rigid cylinder that cannot move to accommodate the shrinkage of the weld metal.

The paper does not go into the microstructural detail, but from my own experience I can tell you what was happening on the shop floor. The Lincore 423Cr is a martensitic stainless steel surfacing wire, and when you deposit it on a low-alloy steel base, you get a dilution zone where the carbon from the base metal mixes into the stainless deposit. This raises the carbon equivalent of the first layer and makes it extremely prone to hot cracking, which appears as transverse cracks that run perpendicular to the welding direction. These cracks are often not visible on the surface until the roll is put into service and the thermal cycling opens them up, leading to spalling and catastrophic failure.

The Practical Countermeasures That Worked

The paper's contribution was not just in describing the process, but in identifying the specific measures that prevented the cracking. Based on the parameters given and the results achieved, the critical controls were as follows. The preheat temperature had to be maintained at a minimum of 200 to 250 °C, and the interpass temperature had to be kept within a tight window, because if the roll cooled down between passes, the martensite transformation in the HAZ would occur at a temperature where hydrogen could not diffuse out, and delayed cracking would follow. The wire feed speed of 120 to 130 mm/min was actually the key to controlling the dilution, because a lower wire feed speed reduces the deposition rate and therefore reduces the depth of penetration into the base metal, which in turn reduces the amount of carbon picked up by the first layer.

Parameter Original Value Adjusted Value Effect on Cracking
Current (I) 400–450 A 380–420 A Reduced heat input, less dilution
Voltage (U) 30–32 V 28–30 V Shorter arc, less base metal melting
Wire feed speed (v) 120–130 mm/min 110–120 mm/min Lower deposition rate, less dilution
Preheat temperature Not specified 200–250 °C minimum Prevents martensite formation in HAZ
Interpass temperature Not specified 200–300 °C Allows hydrogen diffusion, avoids cold cracking

What This Means for the Modern Shop

I have seen this exact problem repeat itself in different forms over the years. The fundamental lesson is that when you are surfacing a low-alloy steel roll with a stainless or martensitic alloy, the dilution of the first layer is the single most important factor in determining whether the repair will survive. You cannot simply turn up the current to get more deposition, because you will melt more base metal, pick up more carbon, and create a crack-prone interface. You have to think about the process in terms of the heat input per unit length of weld, and you have to monitor the interpass temperature with a contact pyrometer on the roll surface, not just rely on the operator's feel.

The other lesson that I would add from my own experience is that the post-weld heat treatment (PWHT) is not optional for these rolls. After surfacing, the roll must be stress relieved, and the temperature and time have to be chosen based on the base metal grade. For 15CrMo and 30CrMo, a stress relief at 620 to 660 °C for a minimum of two hours is typically required, and the cooling rate after PWHT must be slow enough to avoid re-hardening the HAZ. If you skip the PWHT, you will have residual stresses that combine with the service thermal cycling to produce cracks within the first month of operation. This paper, though it is from 1995, is a reminder that the fundamentals of surfacing metallurgy have not changed. The tools have gotten better, but the physics of dilution and restraint remain exactly the same.

The practical takeaway for any engineer who is tasked with repairing continuous casting rolls is to treat the first layer of surfacing as a separate welding operation with its own qualified procedure, not as part of the overall build-up. You need to qualify the buffer layer on its own, measure the dilution using a chemical analysis of the weld metal, and ensure that the carbon content of the first layer does not exceed the limit that the surfacing alloy can tolerate. If you do that, you will avoid the hot cracking that this paper so clearly identifies, and you will get the service life that the roll was designed for.