Stress Concentration of Undercut Weld

By jmaus3 Date 09-09-2005 18:10

I am trying to determine the stress concentration factor (K factor) of the undercut section of a seam weld. I am certain that there must be a paper in the AWS archive that will help with this subject, but I was unable to find anything. I am experiencing some failures with an undercut weld and I would like to be able to calculate the impact that the undercutting is having on the structure. If anyone has any information on this subject please help. Thanks.

By RBeldyk (**)

Date 09-09-2005 19:07

See table 37 in "Roark's Formula for Stress and Strain" Warren C. Young, 6th edition.

By D.X.J Date 09-15-2005 07:49

I have a same question with you,

By jwright650 (*****)

Date 09-15-2005 12:41

One obvious thing that comes to mind about undercut...If you have a 1/4" thick plate and it has 1/16" undercut, you have reduced the cross section of the plate from 1/4" to 3/16".

By - Date 09-15-2005 15:36

John,
My understanding is that the notch effect is much more injurious than the reduction in thickness, especially for joints transverse to the primary direction of tensile stress. My experience on fatigue sensitive offshore structures has been that in some cases, the Engineer (via the spec and drawings) requires the toes of welds to be ground such that the base material thickness is intentionally reduced (ref D1.1 C2.20.6.6 for a discussion of this).
This is normally accomplished with a pencil grinder using a bit with a radius based on the material thickness, and the grinding is normally to a depth of 2-3mm.
Mankenberg

By jwright650 (*****)

Date 09-15-2005 16:19

Kipmank,
You are so correct, I totally over looked that fact. Smooth transitions transfer stresses without concentrating them such as in notches from undercut or poorly executed flame cuts(ie. copes). Other examples are transitions in thicknesses and widths are required to have a transition of 2.5 to 1 slope between the parts(ref. D1.1:2004 Figure 2.2 and Figure 2.3) also see Table 2.4 for more pictorials and explanations.
John Wright

By henri

Date 10-06-2005 08:12

Perhaps you may have to determine KIc with the method(s) outlined in "Fracture Toughness of Metals (K-value)" - ASTM E 399