Checking Welds

I've been following the learning to weld thread and didn't want to pollute it with a semi off topic question

Also mentioned was the ability to weld copper if you get good enough and have the gear, all very useful to any ME chappies

My query is on the checking and testing of welds, a water/hydraulic pressure test is the obvious one, or at greater risk to the user, an air pressure test

What about the welds themselves though?

I believe they use ultrasound/x rays for welds in critical industries like nuclear submarines

Does anyone know if they are producing any cheap weld checking technology out there yet, especially since smartphone technology etc is advancing so fast

edit from yahoo

X-raying or radiograph will show lack of side wall penetration and impurities (slag or gas holes) within the weld.

Ultrasonic testing will show the same and is cheaper than xray.

Magnetic particle testing will show cracking if any

-------------------

wiki

An interesting new technology called SIP is now being used by the car industry, using electrical signals

It looks like it's only a matter of time before this kind of thing hits the market

Edited By Ady1 on 23/10/2015 11:19:07

Look back at BBC programme (Building Cars live BBC2 19.30hrs Tuesday this week) on Cowley works Mini production. Good portion on spot weld testing using an U/S probe nd also tearing them apart in the classical way.

My intent was to introduce a bit of light relief, instead I seem to have convinced people I want a welding simulator at home and another round of reflex China-bashing...

On your topic, x-ray, ultrasoinic and dye penetration come to mind but this website offers you five options:

**LINK**

My approach is 'pull it and if it doesn't fall off it's OK'.

Neil

In a previous life I used to do a lot of NDT (no destructive testing) of welded joints, so the methods available are as follows: dye pen, mag particle, ultrasonic and x-ray.

What you can do yourself is limited to dye penetration testing and magnetic particle testing. The Dye Pen method is to clean the test area with a special cleaner, spray the test area with a dye, usually red, and let it dry, then clean it off with a cleaner 'soaked' (lightly coated) rag, then spray with a very light coat of a usually white topping. The crack will show up as a red line. Magnetic particle is much the same, clean, spray with a white coat, let dry, spray with a fluid containing iron filings and then apply a magnet. The filings line up along the crack. Although both methods are a bit on the DIY side, they do need the right instruction and application to enable the method to be done properly, so some training is required. Like most things the tests are useless if you don't know what you are doing.  A finer more sensitive method for both is a UV version of both - and for both these training is definitely required!

The problems with both these methods, good as they are, is that it only shows up surface defects on the tested side, not any defects within the weld. To do that you have to examine the weld using either ultrasonic testing or radiography (X-Ray), or both (boiler tubes on coal-fired power stations would employ 100% ultrasonic PLUS 10% X-ray). For both you need very well trained and competent personnel to carry out the test and interpert the results, an x-ray is not as easy to view and determine what it shows as one might think. X-ray testing is not foolproof either, the angle at which the image was taken can mean a weld with a crack within it can seem to show a perfectly good weld on the x-ray print. An example of this at the power station was one weld where the ultrasonic guy said there was a crack in the root of the weld but the x-ray was OK - I know as I also viewed it. The ultrasonic guy was adamant the crack was there, the weld was retested with a shot from a different angle and there was the crack clear as clear can be.

With ultrasonic testing, one basically used a probe which sent out an ultrasonic sound wave out and through the metal and an electronic screen which displayed what waves came back. The probes are produced to send the ultrasonic waves out at a predetermined angles from 0 deg. (straight down) to close to 90 deg. The equipment is calibrated before use by using known targets on the probe in use and so setting the screen to display the results to a certain magnitude. A 0 deg probe finds the thickness of the plate or tube being tested and also any laminations within the plate, the other probes send waves out to search the weld. If all is good the waves just disappear down the metal but any defects within the weld send a signal back (like radar) which is then plotted on a drawing, noting where the probe was when the signal was received, so a bit of maths was involved, to show what and where the defect is and likely to be, like lack of root penetration, root crack, slag inclusion and so on and so forth plus its size and position.

There are times when, having made an ultrasonic examination and you have put yourself on the line by saying for example that there is a crack in the boiler furnace tube from here to here and to within 3mm of the surface and therefore the boiler needs an expensive repair before being returned to service, you are rewarded by the boiler repair guys tell you after that the crack was exactly where you said it was, from there to there and to within ⅛" of the surface - that would do for me! There would also sadness when an ultrasonic test of a boilers riveted seams shows a massive amount of cracks at just about every rivet hole meaning the boiler was a scrapper. Especially when this boiler was at a wood 'factory' or yard and all the scrap wood and sawdust was burnt in this boiler to produce the steam the yard required for its processes, a very waste-not want-not system, and a sad day telling the manager his boiler was a scrapper, he'd have to get a new one.

Chris

Edited By ChrisH on 23/10/2015 14:48:43

Edited By ChrisH on 23/10/2015 14:50:58