Isn't there a provision in the inspection clause which specifies that the ends past the effective length of intermittent fillets are not subject to inspection?  Pardon me in advance if I'm speaking out the wrong end.

Not sure if you're just wanting to use that weld picture as an example for a question, or if you are wondering why no one is calling out any issues with it.  The weld in the picture you attached that I made was only made to try and simulate the cathode cleaning action pitting that occurs while welding over abraded surfaces.  I was trying to find out what is causing the fluorescent dye to hang up on the toes of the weld.  The crater pit and the overlap at the start and finish are rejectable defects and good teaching aids when teaching about PT. I couldn't be doing R & D at work when I'm supposed to be teaching the students without actually teaching the students. ;-) ha ha  Sometimes, on say a half inch stitch weld, the welder welds 1" centered.  So you have a 1/4" of weld on either side of the good 1/2" weld that isn't necessarily needed.  I've seen some inspectors allow those 1/4" areas for a cold start like you see in the picture and lapped finish as long as there is a good 1/2" weld centered where it's supposed to be.  I want to say I remember reading in an AWS standard about stitch welds that went something along the lines that says as long as there are no defects in the length of weld called out it was acceptable.  Your allowed to have more weld, reducing the space in between your increment welds, but not increase the length of space in between increment welds.   Maybe someone on here can speak more to that and sight bible and verse.  I haven't had my nose in a AWS standard for a minute.

Overlap is located at the toe of he weld bead where the weld "spills" over on to the adjacent base metal without fusing to the base metal. My definition, not per AWS 3.0. 

Here's the definition found in AWS A3.0:2001.

Best regards - Al

Martin,

Greetings.  I have seen so many variations to the application of code and definition of overlap that it makes my head spin.  I have seen so many welds marked as having overlap that in my mind were no where close to being overlap as defined in the attachment Al gave. 

But how do you apply the definition?  It is stated 'beyond' the toe... thus, when a fillet weld with a convex profile has a bit extra of a drop at the toe area many will mark it as 'overlap'.  But it doesn't take much of an eye to see that that weld reinforcement does not protrude beyond the toe.  And, the toe itself is clearly visible to see that there is no lack of fusion at the toe.

IMHO, we are really looking for two things: 1) Lack of fusion which could lead to failure especially under certain stresses, and 2) Stress raisers caused by unacceptable weld profiles.

So, if you can see the toe and confirm that there is indeed fusion, and, the weld does not protrude PAST THE TOE, it is NOT overlap nor lack of fusion.  That does not mean that it could not be improved upon and in some cases be detrimental to the integrity of the component as if not a totally perfect profile it will indeed fail sooner than is desirable. 

Take a former job I mention often, Log Stackers in the NW.  If the gussets on many locations of that machine are not perfectly smooth they fail so much sooner per frequency of use than those that are very smooth and even somewhat concave instead of convex. 

The ultimate answer is in the application for the job at hand, the IOR and EOR's desirable finish, and your conscience. 

Have a Great Day,  Brent

In reading through this discussion again I think I noticed one more point that needs to be made, at least for my jobs.

In some regards it may be true that the weld develops the proper amount of strength per the engineer's plans as long as the weld has the proper length of sound weld.  BUT, the areas outside that length, while not important as to strength calculations, are still critical for member integrity. 

Do you really mean to tell me that any good, competent, reliable, inspector would let 1/8" undercut go without correction/repair just because it is outside the length required per plans for weld.  No discontinuity is acceptable in weld profiles just because they are beyond the length specified.  I may be able to leave an undersized area of weld as undersized because it does have the correct length of properly sized weld, but undercut, overlap, porosity, or a crack?  No way.  Not on any job I am on. 

Yes, there are aspects that can be passed up because the weld is longer than it needs to be, but if it is an unacceptable discontinuity then it is unacceptable anywhere.  That would be like saying an arc strike can be left alone because it is so far away from the edge of a beam or any other welded component.  Not a Chance.

Have a Great Day,  Brent