How to Weld Heat-Resistant Alloys
Choosing and processing heat-resistant alloys always raises metallurgical questions. Let’s provide some answers.
As a metallurgist, I am frequently asked for assistance on metallurgical issues relating to alloy products and selection. Here are the most frequently asked questions and a guide to answering these questions.
What is the best alloy to use in my new application?
The best alloy choice depends on many factors, including maximum metal temperature, atmosphere, severity of thermal cycling, intended processes and the expected frequency of use.
In heat-treating facilities, 1600°F (871°C) would be the first critical temperature. Operations such as annealing and neutral hardening of carbon steels are done primarily at temperatures lower than 1600°F. Atmospheres for these processes would include air, inert gas or low-potential endothermic gas that matches the carbon content in the alloy being heat treated.
For air and inert-gas environments, stainless steels including 309, 310 and 253MA would be preferred because of the relative low cost. In the case of neutral hardening in a low-carbon-potential endothermic environment, an alloy with at least 35% nickel would be preferred in order to resist carburization since the carbon content of the alloy would be appreciably lower than the steel being heat treated. Stainless steels will not resist carbon diffusion and will quickly embrittle and crack.
Stainless steels are subject to sigma-phase embrittlement over time between 1100° and 1600°F. At room temperatures, an embrittled part can shatter from a forklift impact. Liquid quenching has the potential to cause cracking once this sigma phase forms. For this reason, an alloy with higher nickel content should be used for fixturing that will go into a liquid quench. In the event that flatness is critical, a higher creep-strength material or casting should be considered.
Nitriding is done well below 1600°F. Despite this, stainless steels are not suitable for any nitriding operation because they will not effectively resist nitriding. An alloy with at least 35% nickel would be needed here. The most common choices would be RA330® or 600. Alloys 601 and X could also be used.
At temperatures of 1600-1800°F (871-982°C), creep strength and resistance to atmospheres become the key factors. 309, 310 and RA330 all exhibit usable creep strength to about 1800°F. 253MA® has usable creep strength well above 1800°F. Nickel alloys 600, 601, RA333®, 602CA®, HR120®, Haynes 214® and heat-resistant castings all have usable creep strength up to 1800°F and in some cases beyond. All of these alloys can be used in air or inert environments to 1800°F.
Carburization is typically performed at 1600-1750°F (871-927°C). A minimum of 35% nickel is required for carburization resistance. RA330 and 600 are the most commonly used wrought alloys. Alloys 601 and X are sometimes used in carburizing applications. High-nickel castings may be used for fixturing when straightness is critical, but exposed to high carbon they are brittle and susceptible to cracking, especially when subject to water quenching.
What about temperatures above 1800°F?
Stainless steels and nickel alloys are annealed, and stainless steels are hardened above 1800°F. Fixturing and furnaces for these operations will require alloys with usable creep strength above 1800°F. Such alloys include 601, RA333, 602CA and high-nickel castings (Fig. 1).
On occasion, alloys such as HR120 and Haynes 214 may also be used. All of these wrought-nickel grades will hold up to repeated liquid quenching, though they can be expected to show some warpage. All of these alloys should exhibit carburization resistance as well.
One stainless steel, 253MA, has oxidation resistance up to 2000°F (1093°C) and usable creep strength to 2100°F (1150°C). Unlike the nickel alloys, 253MA does not exhibit carburization resistance, and its oxidation limit will decrease in the presence of water vapor. It is suitable in air-only environments to 2000°F and in inert atmospheres to 2100°F. In order to be used in inert environments between 2000° and 2100°F, all metal surfaces must be in inert atmospheres, not just the inside surfaces.
Are these expensive nickel alloys cost-efficient?
Relative costs of the various wrought plate and sheet materials are shown in Table 1 (as of 5/1/19). Stainless steel 316L plate is used as the baseline. All other alloys will have a ratio reflecting its price compared to 316L plate. Material with a ratio of 3 is three times the cost of 316L plate. The table is for comparison and estimation only.
Cost versus benefits must be considered on a case-by-case basis. A good example is deciding between cast and wrought radiant tubes. Experience shows that a three-legged radiant tube from 3/8-inch wall cast tube will last four years on average. For approximately a 40% cost premium, the same configuration tube can be fabricated from 602CA sheet with only 1/8-inch-thick wall and last almost nine years.
This additional life more than offsets the additional upfront cost of the wrought material. Furthermore, the one week of lost time for tube replacement could be worth much more in additional revenue. The overall savings over the life of the tube are quite significant if an organization can wait the 8-9 years to realize the entirety of the savings. A similar case can be made for replacing continuous muffles with 602CA alloy.
Another significant application is wrought fixtures (Fig. 2) instead of cast fixtures. RA330 grids will warp some after continued prolonged usage. Cast grids will not warp. They will just start cracking at some point. On a component cost basis, there may not be justification to switch to wrought. If a lighter wrought grid can be used and the lighter grid can support an extra large part or more than one large part and still be within the furnace capacity limit, however, significant productivity improvements can be made because one is now using more of the furnace BTUs to heat more parts instead of fixturing. Such savings can be significant.
How do I weld these alloys to each other and to other alloys?
Information on welding can be found on suppliers’, producers’ and specialty-metal distributors’ websites. There are a few unique welding combinations where guidance is not easily found. Two such cases are welding heat-resistant alloys to carbon steels and welding heat-resistant alloys to cast heat-resistant parts.
Welding of carbon steels to heat-resistant stainless and nickel alloys can be tricky for many welders because the techniques used to produce good welds in carbon steels are exactly what should NOT be used to produce good welds in heat-resistant alloys. In carbon steels, welding heats the metal to liquid and the solidification and subsequent cool-down is analogous to quenching and tempering. Quick cooling forms a brittle phase known as martensite, which must be tempered to prevent self-cracking.
Preheating and post-heating slows down the cooling enough to prevent martensite formation or sufficiently temper any martensite formed. In heat-resistant nickel-alloy welding, material is always a single phase: austenite. There is no phase transformation and therefore no hardening. Instead, these materials need to be rapidly cooled through the temperature range over which the alloy solidifies.
That is the key to preventing these welds from cracking. Alloy selection is critical to produce a crack-free weld. For stainless steels to carbon steels, 309 wire is the alloy of choice. For nickel alloys to carbon steel, 82 wire is the alloy used most frequently. Other options for nickel alloys to carbon steel include RA330-04 and RA333.
When welding new castings to heat-resistant nickel alloy, 309 and 82 are the most common alloys of choice. RA330-04, RA333 and RA 602 CA have also been used, depending on the chemistry of the casting. Repair welding of used castings is much more complicated. Castings become embrittled with continued service. Simply checking how magnetic the cast product is can help determine if weld repair is feasible. Highly magnetic alloys are brittle and prone to cracking. When there is only light magnetism or just a slight pull, the chances of successful weld repair improves. The shielded metal-arc process is the preferred method for repair welding.
GMAW or gas-metal arc welding (often referred to as MIG) and GTAW or gas tungsten arc welding (often referred to as TIG) are the preferred methods for welding heat-resistant alloys. Both of these arc processes employ bare-metal consumables and require inert-gas shielding.
Bare-wire welding is preferred for fabrication of heat-resistant alloys. However, shielded metal arc or stick welding is preferred for repair welding where protection is provided from the flux and gases in the flux coating. Flux-core arc welding is another welding process in which both fluxes and shielding gases provide protection.
Flux-core arc welding is a productive process capable of running at high speeds. These fast speeds are sufficient to entrap slag particles in the weld (Fig. 3) before they have a chance to rise to the surface. The slag becomes a defect in such instances. While not as prevalent, slag entrapment can occur in shielded metal-arc processes too. For these reasons, any slag-creating weld method is less preferred, except for repair welding, where the stick process can produce the best results and the least cracking.
When should I use wrought material, and when should I use a casting?
There are compelling reasons to justify the use of both wrought materials and castings in heat-resistant applications. The primary advantages of castings include: higher creep strength due to higher carbon; cheaper unit costs when they can be mass produced; the ability to design and manufacture complex shapes; and the ability to cast more complex chemistries with higher percentages of beneficial alloying elements such as chromium (Cr) and aluminum (Al).
The advantages of wrought materials include: lower cost for smaller quantities, better inventories of wrought materials, ease of fabrication, better weldability (particularly for repair work), thinner individual size availability and typically shorter lead times for smaller-quantity orders.
Another consideration is when lighter is better. A cast radiant tube, with a higher creep strength than a wrought tube, would appear to be the best option. A commercial heat treater used a wrought radiant tube of 1/8-inch-wall 602CA, however, which lasted slightly more than twice as long as a 3/8-inch-wall cast tube in the same furnace. Since the wrought tube weighed 1/3 as much as the cast tube and only had to support its own weight, the properties, such as high ductility at temperature, more then compensated for the higher creep strength of the cast product.
There are several considerations for cast versus wrought as it relates to grids. It is the experience of many heat treaters that cast grids will stay straight while wrought grids will deform and warp. On the other hand, the cast grids often break at joints over time. There is no real downtime associated with replacing a broken grid, unless the grid breaks during heat treatment and parts are lost, in which case there can be significant downtime. Cast grids can also weigh significantly more than the wrought grid. This means that more parts can go on a wrought grid before furnace capacity is reached. A lighter wrought grid also enables more BTUs to be used to heat parts instead of fixtures, reducing unit costs of heat treating.
Why do some fixtures crack or warp, even when operating below perceived temperature limits?
One factor that is often overlooked by designers of high-temperature components is thermal expansion and thermal gradients. Coefficient of thermal expansion is an intrinsic alloy-specific property that increases with temperature. The use of different materials with different coefficients of thermal expansion leads to one component growing more than another at elevated temperatures.
Large components can be subject to thermal gradients when some parts are in the hottest area of the furnace and others are shielded. The hot areas want to expand, while the colder, shielded areas prevent such expansion. When the component cannot move, grow or contract freely, this creates stresses in the component, and the restrained area can bend, buckle or crack (Fig. 4).
Methods to minimize the impact of thermal gradients include a planned gap, expansion joints and using thinner materials in critical areas. Designers need note that stiffeners, which are great tools to employ at room temperature, have the opposite effects at high temperatures. They prevent other components from growing, which forces deformation.
Is this alloy machinable?
Just as there are websites to help in the selection of welding wire, there are some websites with very specific information on machining and relative machinability numbers.[3,4] Most alloys have a relative machinability number that is the percentage of the machining speed of the alloy in question compared to the machining speed of free machining B1112.[3,4] Relative machinability of the common heat-resistant alloys is shown in Table 2.
Consider a company that regularly uses 316L that now has to machine a component from RA330. The table shows that 316L has a relative machinability of 44, while RA330 has a relative machinability of 24. Using this data, the starting speed for RA330 should be (24/44), or 55% of the speed used for 316. From there, minor adjustments might need to be made.
Note that these numbers are based on high-speed steel tools. Today, these alloys are often machined with carbide tools instead of high-speed tool steels. The carbide tools enable faster speeds, but for the most part the relative machinability does not change.
- RA330 and RA333 are registered trademarks of Rolled Alloys.
- 253MA is a registered trademark of Outokumpu.
- 602CA is a registered trademark of VDM.
- HR120 and Hayes 214 are registered trademarks of Haynes International.
For more information: Contact Marc Glasser at Rolled Alloys, 125 West Sterns Road, Temperance, MI; tel: 800-521-0332, e-mail: firstname.lastname@example.org; web: www.rolledalloys.com. See Rolled Alloys’ website for a weld-wire selection guide and comprehensive fabrication/welding bulletins for many different alloys. Weld wire selection guide may be found at: https://www.rolledalloys.com/technical-resources/fabrication-information/welding/
- Kelly J, Heat Resistant Alloys, Art Bookbindery,. Winnepeg, Manitoba, Canada. 2013
- Machinery’s Handbook, Twenty First Edition. Industrial Press Inc. New York, NY. 1980