Q&a Of Welding Stainless Steel

Stainless Layout:Layout 1

2/10/09

2:43 PM

STAINLESS Q&A

Page 20

BY DAMIAN J. KOTECKI

Q: We are making 1-in.-thick butt joints between 304L stainless steel and ASTM A516-70 carbon steel, using submerged arc welding with ER309L electrode. We use a single U-groove with a 1⁄4-in. root radius, 1⁄8-in. land, no root opening, and 20-deg included angle. Then we backgouge and finish the joint with FCAW or SMAW. We have a lot of trouble with centerline cracking in the SAW root pass, unless we substitute SMAW with 309L filler metal for the root pass. What can we do to make SAW root passes successfully?

A: The root cause of your root cracking problem is dilution from the carbon steel side causing the root pass to solidify as primary austenite, instead of the desired primary ferrite. The root pass cracks during solidification and it ends up with virtually zero ferrite content. Shielded metal arc welding gets around this problem by reduced dilution as compared to SAW, but its productivity is low. One way to combat this is by using DCEN polarity for SAW at low wire feed speed (low current) to reduce dilution to levels comparable to those obtained with SMAW. But that is a relatively low productivity approach also, and the DCEN root pass bead shape is often not desirable. A more robust (forgiving) approach is to design the joint to get away from symmetry. The problem with high dilution in a symmetrical joint, such as a single Ugroove, is that the dilution comes approximately equally from the two base metals. It is dilution from the carbon steel side that is responsible for loss of ferrite in the weld metal. Dilution from the 304L side of the joint has a more-or-less negligible effect on root pass ferrite content (or solidification mode). That is because common stainless steels like 304L, 316L, 317L, 321, and 347 are generally designed by the steelmaker to solidify as primary ferrite, just like common stainless steel weld metals like 308L or 309L. This helps them to get a higher yield from hot working of the steel. You can easily establish this for yourself by simply making a GTAW beadon-plate weld without filler metal on one of these base metals — you will almost certainly find a magnetic response in the weld indicating presence of ferrite, even if the base metal is completely nonmagnetic before welding. The approach I suggest is to make the joint preparation asymmetrical so that more of the dilution comes from the 304L side of the joint than from the carbon steel side of the joint. This can be done simply by, for example, changing from a single U-

20

MARCH 2009

Fig. 1 — Alternate joint preparations for carbon steel to 304L with ER309L filler metal.

groove joint preparation to a compound joint preparation consisting of a single J preparation on the 304L and a single bevel preparation on the A516-70, so that the SAW arc will impinge mostly on the 304L. A second possibility is to insert a 304L stainless steel backing bar into the joint, so that, again, the arc will impinge mostly on the 304L, although in this case it is the backing bar. These two alternatives, as compared to your symmetrical joint preparation, are sketched in Fig. 1. Figure 1A is your symmetrical joint preparation, while Fig. 1B and C are asymmetrical joint preparations as outlined above. In welding either the joint preparation of Fig. 1B or that of Fig. 1C, you should still align the electrode along the joint preparation centerline, not along the 304L/A516-70 interface. Then the weld bead will wash up onto the carbon steel without a lot of penetration into it. Most of the arc energy will be spent melting 304L base metal (possibly including the backing bar should you choose to use that) and ER309L electrode. Because the joint preparation is asymmetrical, you should find that you will have to backgouge a bit deeper in order to reach sound metal, but that is preferable to cutting out a root pass with a centerline crack as you have been experiencing with the symmetrical single U-groove joint preparation. In fact, this will be evidence that you have achieved lower dilution from the carbon steel. After a successful root pass, the arc impinges mainly on previously deposited weld metal that contains ferrite, and the low dilution effect from the carbon steel side continues until joint completion. Once you get the idea of the way this

asymmetrical joint preparation works, I am sure that you can come up with a number of variants on the approach to suit other situations. For example, the backing bar need not be inserted into the root for a single-side joint preparation — a 304L backing bar wider than the root opening could be used underneath the joint preparation with the same effect. In that case, backgouging to remove the backing bar would only have to remove the backing bar and very little of the joint, so that only a very shallow backside weld, if any, would be needed to complete the joint to full penetration. This approach is not restricted to SAW. It can be used to improve solidification cracking resistance in any situation in which a stainless steel expected to provide a bit of ferrite (316L, for example) is to be joined to carbon steel or even to a stainless steel in which no ferrite would be expected (310 or 320, for example).◆

DAMIAN J. KOTECKI is president, Damian Kotecki Welding Consultants, Inc. He is a past president of the American Welding Society, currently treasurer and a past vice president of the International Institute of Welding, and a member of the AWS A5D Subcommittee on Stainless Steel Filler Metals, and the AWS D1K Subcommittee on Stainless Steel Structural Welding. He is a member and past chair of the Welding Research Council Subcommittee on Welding Stainless Steels and Nickel-Base Alloys. Email your questions to Dr. Kotecki at [email protected], or send to Damian Kotecki, c/o Welding Journal, 550 NW LeJeune Rd., Miami, FL 33126.