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

Post-Weld Heat Treatment of CLAM Steel TIG Welds — Tempering Parameter Selection

Why CLAM Steel Welds Need PWHT

The paper by Qiao Jiansheng, Huang Yina and Wan Farong, published in Nuclear Science and Engineering (2009, Vol. 29, No. 3, pp. 239–246), addresses a problem that I have wrestled with many times in my career: what to do with a martensitic steel weld that has come out of the arc too hard and too brittle. The material in question is CLAM steel (China Low Activation Martensitic steel), which is being developed for fusion reactor first-wall and blanket structures. It is a 9Cr–1Mo–V–W type reduced-activation steel, and it has the same fundamental welding metallurgy as the P91 and P92 steels that we use in high-temperature power plant piping.

The problem is straightforward. When you weld a martensitic steel with a TIG process, the weld metal and the heat-affected zone cool rapidly through the martensite transformation temperature range. The result is a fully martensitic microstructure that is hard, strong, and low in toughness. In the as-welded condition, a P91 weld can have a hardness of 400 HV or more, and the impact toughness can be well below the required minimum. This is not acceptable for a pressure-retaining component, and it is certainly not acceptable for a fusion reactor blanket where the material must withstand neutron irradiation and thermal cycling.

The authors set out to find the optimal post-weld heat treatment (PWHT) parameters for CLAM steel TIG welds. They investigated the effects of tempering temperature, tempering time, and repeated tempering on the hardness of the weld zone. Their goal was to reduce the hardness to an acceptable level while maintaining the strength and toughness balance required for service.

Tempering Study and Recommended Cycles

The authors conducted a systematic study of tempering parameters. They varied the tempering temperature over a range that spans the typical tempering window for 9Cr martensitic steels, and they varied the tempering time from short to long durations. They also examined the effect of repeated tempering cycles, which is a practical consideration when a weld has been tempered once but the hardness is still too high.

Tempering Temperature Tempering Time Hardness Result Assessment
700 °C 2 h Still high Insufficient
730 °C 2 h Moderately reduced Marginal
760 °C 2 h Acceptable Recommended
760 °C 1 h, repeated twice Acceptable Alternative recommended
790 °C 2 h Reduced, but risk of over-tempering Not recommended

The two optimal heat treatment cycles identified by the authors are:

The first cycle is straightforward and is the standard approach for most martensitic steel welds. The second cycle is interesting because it recognises that in some cases, a single tempering cycle is not sufficient to achieve uniform tempering through the full thickness of the weld. This is particularly relevant for thick-section welds, where the centre of the weld cools more slowly than the surface and may retain a higher hardness.

From a metallurgical standpoint, the tempering of 9Cr martensitic steel involves the precipitation of fine carbides and the recovery of the martensite lath structure. At 760 °C, the steel is in the lower part of the tempering range, which gives a good balance between hardness reduction and strength retention. At higher temperatures, the carbides coarsen rapidly and the strength drops, which is why 790 °C is not recommended.

Comparison with Conventional Martensitic Steels (P91/P92)

The CLAM steel PWHT parameters recommended in this paper are remarkably similar to those used for P91 and P92 pipe welds in the power generation industry. For P91, the typical PWHT is 730–760 °C for a minimum of 1 hour per 25 mm of thickness, with a minimum of 1 hour. The authors' recommendation of 760 °C for 2 hours is at the upper end of this range, which is consistent with the fact that CLAM steel has a slightly different composition and may require a slightly higher temperature to achieve the same degree of tempering.

Steel Grade Typical PWHT Temperature Minimum Time Hardness Target
P91 730–760 °C 1 h per 25 mm ≤ 250 HV
P92 730–760 °C 1 h per 25 mm ≤ 250 HV
CLAM (this paper) 760 °C 2 h Not specified, but reduced

The key practical point is that the PWHT window for these steels is narrow. If the temperature is too low, the hardness remains high and the toughness is poor. If the temperature is too high, the strength drops below the required minimum and the creep resistance is compromised. The difference between a good PWHT and a bad one can be as little as 20–30 °C, which means that temperature control during PWHT is critical.

Practical Implications

For the pipe and fitting industry, this paper is a reminder that PWHT is not a "set and forget" operation. It requires careful control of temperature, time, and heating/cooling rates. I have seen P91 welds that were PWHT at too low a temperature because the thermocouples were not properly attached to the pipe surface, resulting in a weld that passed hardness testing at the surface but was still hard in the centre. I have also seen welds that were over-tempered because the furnace was left at temperature for too long, resulting in a loss of creep strength that only showed up years later in service.

The repeated tempering cycle recommended by the authors is a valuable practical insight. In thick-section welds, or in welds with complex geometries, a single tempering cycle may not be sufficient. The first cycle tempers the outer regions, but the centre of the weld may still be hard because the heat from the tempering furnace takes time to penetrate. A second cycle at the same temperature, but with a shorter hold time, can complete the tempering of the centre without over-tempering the outer regions.

For CLAM steel, which is intended for fusion reactor applications, the additional consideration is that the material must be low-activation. This means that the alloying elements are restricted to those that do not produce long-lived radioactive isotopes under neutron irradiation. The PWHT parameters must be chosen to maintain the required mechanical properties without introducing any microstructural features that would degrade the material's performance under irradiation. The authors' work provides a good foundation for this, but it is worth noting that the irradiation behaviour of the tempered microstructure is not fully addressed in this paper.

In summary, this paper provides practical PWHT recommendations for CLAM steel TIG welds. The recommended cycles of 760 °C for 2 hours, or 760 °C for 1 hour repeated twice, are consistent with the practices used for conventional 9Cr martensitic steels and should be adopted as the baseline for CLAM steel fabrication. The key lesson for production is that PWHT is a precision operation, and the difference between success and failure is measured in tens of degrees and minutes, not in broad strokes.