JimA:
You noted "...the metal is actually torn out." This is typical for a fillet weld. If you can see any part of the corner of the vertical member, then incomplete fusion is present.

If this is a structural steel weldment, then refer to the D1.1:2002, 4.30.4.1 for the acceptance criteria for fillet weld break tests.

Thanks CWIC

Thanks CWIC, We're overseas at a remote location and just trying to get our local welders some qualification test practice before our CWI arrives.  I have a couple more questions on the break test.  So far we've had a pretty good variety of fractures, and most are easy to evasluate.  A few I'm not sure about.  For example, a couple have areas where the weld has pulled away from the vertical leg, and has taken a paper thin amount of metal with it. You can't see any of the original surface of the vertical leg, just freshly exposed metal. So there was penetration, but just barely.  Would that be a pass or fail?  Also, we had one that failed at the toe of the weld all the way across.  The weld bead is intact, just the leg tore away from the base plate.  There was penetration the whole way.  Is there any significance to that type of fracture versus the weld fracturing?  Thanks for your help.

JimA:
If you are indeed removing a "...paper thin amount of metal...", make sure it is base metal and not mill scale which is a common failure mechanism for fillet welds.

The depth of penetration is not an issue unles it is specified that your fillet weld's actual throat complies with the criteria for your project. I failed a fillet weld test once (aerospace application) where the criteria had specified a minimum effective throat for the weld. The profile and fusion were satisfactory, but the weld (throat) needed to be increased. This was based on a % of the material thickness.

As far as the one that failed "...all the way across...", I would think (without seeing the test plate) the welder simply needs more practice. It sounds like he/she is getting incomplete fusion which is again commmon with fillet welds.

I have seen very experienced welders fail simple fillet welds tests as they spend time trying to make the weld look pretty as opposed to getting the proper penetration and profile.

Thanks once again CWIC.  You really helped us out a lot and we appreciate it!
Jim A.

JimA:
Good luck with your qualifications and please don't forget to let the forum know how the testing came out.

CWIC,
OK - I sure will. It will be in about two weeks or so.  Some of the gus are ready, and most of the others are close and should be OK with a little more practice.  One of the main problems is they don't do a very good job of getting the plates cleaned up before welding, but they're learning.  Thanks agian for your help.
Jim A.