Ultrasonic inspection may provide some measure that is useful. Marking the length of the crack, assuming it is not all around the joint, then cyclic loading and inspection to assess any crack growth. 

Such cracks can sometimes be avoided with preheating before welding. Corrective strategies when the cracked part must be reused include grinding out all of the weld and any cracked metal adjacent to the weld.  Then preheat, weld and immediately post weld stress relieve. This might require an inserted piece if needed.

Hi Luis Hernandez,

it's true Robert, Ultrasonic testing will best detect those more critical planar discontinuities such as cracking and incomplete fusion. UT is most sensitive to discontinuities that lie perpendicular to the sound beam. Because a variety of beam angles can be used UT can detect laminations, incomplete fusion and cracks that are oriented such that detection with radiographic testing would not be possible, but The major disadvantage of ultrasonic testing is the advanced skill required to properly interpret the results. Weld design, location of the defect, internal structure, and complexity of the weldment affect the interpretation of the ultrasonic signal.

Regards,

Thanks guys. I've asked the ultrasonic testing but the result didn't arrived yet. We have permission to do destructive mechanical testing, so if you have any sugestion, tell me.

If so, you can cut the samples of material directly from joint and perform the destructive mechanical tests, particularly, in tension conditions.

This give you valuable data concerning strength characteristics as well as limit elongation and other parameters. The determination of impact toughness of samles is seemed useful too. Best regards

Hi  Luis Hernandez,

The most common types of destructive mechanical testing are known as free bend, guided bend, nick-break, impact, fillet welded joint, etching, and tensile testing. The primary disadvantage of destructive testing is that an actual section of a weldment must be destroyed to evaluate the weld.

Regards,

I agree with Mr. Serge and Luis. perform UTS, BEND, IMPACT etc

It seems to be clear transverse cracks and attributed (most probably) to the carbon equivalent. Measure the chemical composition and estimate the carbon equivalent (should not more than 0.35). Weld decay is also the cause for transverse bead cracks. Measure the hydrogen content in weld by extracting piece from weld region only.