Wrapping welds and undercut on fillet welds

By catdaddy68 Date 06-18-2007 18:08

I am looking for input about undercut on fillet welds.

Some welders in our shop have trouble with burnout or blowout or whatever you call it when they come to end of a fillet weld on a knife plate. The end result is undercut exceeding 1/32" for about 1/4" long or less.

Table 6.1 of the D1.1(2006) says "For material less than 1 in. [25 mm] thick, undercut shall not exceed 1/32 in. [1 mm], with the following exception: undercut shall not exceed 1/16 in. [2 mm] for any accumulated length up to 2 in. [50 mm] in any 12 in. [300 mm]."

Can I interpret this to mean that the 1/4" of 1/32" or more undercut is acceptable? Granted, it looks like garbage, but is it rejectable under D1.1? Does the undercut at the end of the knife plate have more chance of failure than undercut in the middle of the weld?

My next question is wrapping welds at the end of knife plates. A special inspector visiting our shop commented that by wrapping welds around the end of the plate introduced perpendicular stresses to the fillet welds on the sides and increased the chance of failure. He said that the D1.1 had a section on this in Chapter 2 (Joint Design) but was unable to find it for me. What are some of your opinions on wrapping or not wrapping welds? Also, if you don't wrap your welds, do you stop short of the end of a plate? By how much? If you do stop short, then the first problem I mentioned tends to fix itself.

Thanks in advance.

By swnorris (****)

Date 06-18-2007 20:17

Your special inspector is correct. Fillet weld terminations should be located approximately one weld size from the edge of the connection to minimize notches in the base metal. We allow one exception, which is when the material is to be hot dip galvanized. If this is the case, we weld all around, while paying close attention to make sure we have no undercut.

By catdaddy68 Date 06-18-2007 21:09

Hot dog. Can you tell me the section in D1.1 so I may reference it in the future?

By swnorris (****)

Date 06-18-2007 21:51

What I posted came directly from the AISC's Steel Construction Manual, 13th Ed., page 16.1-97. Also, take a look at 2.8.3.1. It states in part that fillet welds may be stopped short........

By thcqci

Date 06-18-2007 22:37

That was one of the harder things I had to overcome when I came here was to get them to quit wrapping and welding off the edge of plates. The majority of the time, there would be undercutting at the corners and edges. They said "it looks better" or "it doesn't look done when we don't wrap". "Better" being a relative term because I certainly didn't think the undercut looked better!

The "weld-all-around" symbol used by the detailer is often not correct to just plain lazy on the detailer's part. We weld 2 sides of shear tabs, knife plates etc. Likewise, we do not wrap toes of columns on base plates even when the weld symbol is weld-all-around. Along with the above reference, look thru the tables and figures in Chapter 2 of D1.1.

By swnorris (****)

Date 06-19-2007 10:12

Doug,

I agree with your comments 100%. I hear and see the same things.

By catdaddy68 Date 06-19-2007 14:02

Gentlemen,

Thanks very much for your input. This will allow me to settle disputes between production and quality control.

By CHGuilford (****)

Date 06-19-2007 16:46

Fillet weld termination is addressed in D1.1-2.8.3. What I get out of that is that it is ok to have an end return (wrap), or to stop at the ends of the joints, or to stop the welds short if not specified otherwise. The rest of 2.8.3 lists the exceptions to be considered.

Good workmanship practices and exceptions aside for a moment, my personal feeling is: don't put welds where they are not specified to go. If a weld wraps the ends a bit, it is probably no big deal- if it were, then there would be a detail to show "weld hold back".
Also, weld all of the weld that is supposed to be there. If the weld is a little short, also no big deal but be careful because hold back dimensions are not given.
If every weld is right at the end, nothing short, nothing extra, then the welder is pretty good.

The easiest way to weld knife plates with no/minimal undercut at the ends: start welds on the ends and stop in the middle. Then start at the other end and tie in at the middle, filling the crater. This is not hard to do unless the weld joint is very short.

Special inspectors and others need to be careful that their comments are either in agreement with the specifications, or that they clearlly indicate they are expressing an opinion.
(I ain't special, but that's my opinion.)

By drifter57 (**)

Date 06-19-2007 17:49

I am just a little lost on this. If a Crater is filled to the full cross section of the weld how can there be undercut?

By waynekoe (**)

Date 06-19-2007 17:56

I have to agree with Chuck. Particularly in respect to knife plates in brace connections. In my research from my last post on this same subject, of immediate concern for weld termination and allowable undercut depth is the designation of the part in question. A simple shear clip that knifes thru a column could/would still be concidered as a primary member and subject to the demands of 7(B) in table 6.1., D1.1. Knife plates for brace tubes and any main piece of that structure, i.e back to base plate connection welds and yield point stiffeners are only allowed the minimum of undercut, .01",and wrapping the ends, while optional, is desireable. The tube to knife plate connections are designed with a specific weld length. Generally, the length of tube that swallows the knife plate, say, 12", and the minimum weld length specified is 1/2"x12", then you have to run the full length, you can't be short. On the other hand, Weld washers, filler plates ect., , can have the max. allowable for undercut.

By hogan

Date 06-19-2007 18:07

i would think that the contractor would understand that the inspector (special or not) has no authority to direct work. also, that they are responsible for the fabricated product. i don't think that inspectors need to watch what they say. i think contractors need to know what they are doing. not relying on the inspector as a fabrication aid.
inspectors are the eyes and ears of the engineer, not the mouth.