weld inspection

If it is a critical joint use X or gamma rays depending on the walls thickness

Hi,

One thing to have into account is the weld thickness (assuming there are full penetration welds). For heavy thicknesses I would suggest ultrasonic testing which will allow you to check all the fusion zones for lack of fusion/penetration. You have techniques (tip diffraction, for example) which can measure the real dimensions of the possible defects.

For using eddy current I see advantages if the inspection can be done from the root side of the weld, due to the limits on depth penetration but if the welds are partial penetration ones, this will give always an indication which can be more difficult to evaluate.

Radiographic examination for this type of weld would need two different expositions minimum, because the maximum sensibility of the method is for defects parallel to the beam direction and you have two surfaces in which lack of penetration can occur (tube and plate) and they are at 90º. If the pipe diameter is not too large, several more expositions would be needed because no panoramic exposure is possible.

Also, besides weld penetration, you have to check above and below the weld for de-lamination of the plate material. We do that with a simple dye penetrant test.

Have to agree wholeheartedly with Post 2, in that your selection of Method will depend heavily on the thicknesses (and geometries) involved. Also, are these regular carbon steel, stainless steel, or (?). Is the plate cut-open to the bore of the tube, or left solid? How large a piece are we talking about (could it be lowered-into a small immersion tank for ultrasonic testing)? Is it "in-situ" high up in a difficult-to-access location? Many criteria will determine which Method & Technique are likely to give you the most profound results.

Also, what type of welding procedure was utilized, and what will be your critical defect size?

Eddy current will work. X-ray is better. If you plan on making several hundred welds of this type, look up Southwest Research Institute. I have a buddy there using a magnetostrictive technology thats out of this world literally.

Hi folks,

I'll try to answer in only one post to several questions (mainly related to posts 2 and 5):

1. The original poster said: "Can anyone tell me which would be better for inspecting how well the weld has bonded/penetrated the surfaces of the tube and the plate" Bold is mine.
2. Obviously there are many type of defects that can occur in a weld (porosity, slag inclusions, hot and cold cracks, lamellar tearing, excessive undercuts, etc) depending upon the welding procedure used.
3. I think that freddie refers to lamellar tearing which may happen in the plate due to stresses normal to the thickness when segregations are present in the plate. The dye penetrant inspection proposed is O.K. if you have few parts to inspect or they are non ferromagnetic. Otherwise, magnetic particle is much more faster and sensitive than dye penetrant.
4. Selection of any NDT procedure needs some knowledge of the type of defects that can produce the manufacturing process used and the specific material properties and behaviour. Just read the requirements of ASNT or EN-473 for qualification of Level III personnel.
5. Eddy current I think (with the few information given) is not the most suited method because of inherent characteristics of the method itself: You can have different signals from base metal, heat affected zone and weld deposit which may impair the inspection. Furthermore, you can detect just any variations produced in the close vicinity of the coil and to inspect a weld of that type you would need 4 or 5 different scannings by very skilled people.
6. Finally I completely agree, obviously, isn't it? with post 4.

Thank you all very much, this has been very helpful so far. I will post again when I have some more specifics, but for now, the tube is 2" OD, 1/16" wall thickness welded at a 90° angle to a 3/32"x7"x7" plate with a hole cut through the middle. So, yes, I think it would easily fit into a submersion tank.

I was told by a sales rep that in the world of NDT, no one has yet developed a way to test the penetration of the weld into the work pieces, so it's good to hear that ultrasonic testing may be able to do that.

For the eddy current probe method, the weld could be analyzed from the root side, since it's a hollow tube.

Greetings to all ~ ..."still more data is needed". You didn't address 2 points in Post 4:

"what type of welding procedure was utilized, and what will be your critical defect size?"

You state that there's a hole cut in the plate's middle; how large, how close to the inside diameter of the tube?

Such thin material (still of unknown steel-grade). ... How critical does the "bond" need to be? Is it for shear or torsional strength, or vacuum/pressure app?

"PS" ~ magnetostrictive technology *DOES* have enormous potential to cover vast areas of large diameter pipelines, seeking areas of pitting, etc. But its "bulk-wave" principle is not applicable to such a small component as this.

It does now!!!!!!!!!!!!!