UT-welds early-stage training-plate idea

Hello all

Help wanted on idea for early-stage training on UT (Ultrasonic Testing) of welds...

There is this idea you can see - tried making one - at

formatting link

This shows the concept of a "training plate" for people at the early stages of learning UT of welds.

Perception that these people need to be familiar with all the external-shape echoes - many which are strong reflectors but not defects - before trying to identify defects in training plates with known deliberately-induced defects.

Concept is ensuring people can walk before you try to get them to run...

Several of us recently gone through UT training were trying to work out how to have a stage between finishing theory and trying to find defects in training-plates with deliberately-introduced "defects".

Why?

Because the "with defects" plates are expensive and unique - hence - the class gets broken up into every individual doing their own thing, unable to support each other and the instructor being overworked and unable to give more than one individual sufficient attention.

If you could have as many cheap identical early-stage external-shape-effects training plates, the instructor could guide the class together through the common surface effect echoes, concentrating on the majority which are not defects... The instructor could slave their UT-set to a projector, so that all could see the reflection-type and its echo-dynamics which is being studied, and the class could support each other.

Things I reckon people need to know

  • "noisy root" - the root makes a big many-dB's echo but is not a defect

  • mode conversions - the multiple peaks close and connected in the echo dynamics when there is a mode-conversion longitudinal wave reverberating between root and cap (given using shear-wave probe)

  • reflections off the cap (OK - a cap would have to be defective in shape to produce echoes and that would be obvious if you were UT'ing on the cap-side, but anyway...)

You can see what the intention is...

Anyone have additional things to look for, better alternative ideas for how to do the same training, etc, etc.???

Richard S

Reply to
Richard Smith
Loading thread data ...

I like the idea. I can see having a plate made with some narrow laser cuts that you cap weld top and bottom to create "lack of fusion" areas

You could mill a narrow groove on the backside near the root to create a "non-relevant indication".

Reply to
Ernie Leimkuhler

I think slits done with a 1mm thick slitting disk in a 4inch or

4~1/2inch angle-grinder would do just fine??? Cheap and readily done. Given idea is that anyone can make one of these "training plates" so must use equipment you'd find in any workshop. The UT beam won't know there is a 1mm slit behind the supposedly crack-like "lack of fusion" defect. Would make no difference to the reflected signal, echo dynamics, etc. And it would make no difference looking from the other side because - no beam skipping into the weld is going to see a l-o-f at this high angle to that beam. The only non-real aspect is going to be a reflection off the bottom of the slit groove when skipping into the weld from the opposite side. And that is not going to be much of an indication - only 1mm wide - and highly off-angle to the beam coming up from the skip.

I'm not following you here. Again one could use a 1mm wide slitting disk to make two parallel slits about say 3mm apart overall edge-te-edge and grind out middle with 3mm thick disk to make square-edged "pseudo" lack-of-root-fusion where there is unmelted weld-prep at the root.

That isn't what you mean though, is it?

In all cases, the idea is that this block only have external geometric effect reflectors. So making a manageable step of only considering non-defect reflectors like "noisy root" and "mode conversion peaks". So yes you could make samples with "cap-toes cracks" by slitting the top surface at an angle and location representing the imaginary weld preparation - then deposit weld metal over at low current, so as to preserve the "defects". So that might be a leap too far. Maybe a second class-size set of samples with such cheap reproducable "pseudo" internal defects???

Richard S

formatting link

Reply to
Richard Smith

hi again Ernie

Lasers to create UT-weld train "defects"

Thought about this myself for making pseudo internal defects.

So again, idea is to make "class set" of identical samples for each defect to be studied, so the instructor can take the entire class through, say - the echo appearance and echo dynamics of a sidewall l-o-f (lack-of-fusion) ..., etc

You only need about 20mm to 25mm of width of "sample" down which to point your beam and reflect off the surface (given beam width). OTOH in early-stage training you need to be able to explore the "echo-dynamics" - one movement of which is to sweep a sector with the beam always pointing down radii to the same point on the defect. Which requires a wider plate to move the probe around on.

So cut plate with inspected area with pseudo-defect as a protrusion?

[plan view]

|--------| | | |________| /______ laser'ed "defect" | ^ | \ |------------ ^ ------------| | ^ | | ^ | | ^ | | ^ | | ^ | | ^ | | Probe | | | | | ----------------------------------

You could do in-depth slits

- down the 30deg angle to represent sidewall l-o-f

- horizontally to represent inter-run l-o-f

- if you ran the beam through-thickness but jiggled it around, would that reasonably simulate a weld centre crack?

The whole idea falls apart if lasers available do not have the property that they can easily and cheaply punch a narrow slit through about 20mm of steel...

???

Richard Smith

Ernie Leimkuhler writes:

formatting link

Reply to
Richard Smith

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.