Ultrasonic testing V/s Rediography

radiography shines x-rays through the metal. Areas with voids will let more xrays through, so voids in welds can be seen. They can show on a screen or with film to be developed. The film method is a mature technology, but you have the delay to develop the film and the cost of a film labs on the spot or the long wait to send them out. The solid state detector matrixes are lower in resolution, cost more to buy, but have lower operating costs and give fast results. You can also print with a laser printer on the clear film so it gives a final product that resembles the film. Both use a high intensity radioactive source that is licensed and quite dangerous and needs to be operated with care to expose only the film/matrix and not you.

A few years ago a couple of guys were killed when they scrapped a source for scrap metals and drove around with the scrap n the back of their truck. They died a slow lingering death from radiation.

Ultra sound send high frequency sound in a path that includes the weld. If their is a crack or void, the ultrasound will reflect from that discontinuity and be picked up by the test sounder(which both emits and listens for reflections). It is similar to baby ultrasound in principal. It has a screen and you can print it out.

read here

http://www.google.ca/#hl=en&source=hp&q=weld+%2Btesting+%2Bmethods&btnG=Google+Search&meta=&aq=f&oq=weld+%2Btesting+%2Bmethods&fp=ada65163c2197ad6

Aurizon has provided a good summary. In practice, the best choice depends on the materials and the type of flaw you are trying to detect. For example, if you have two metal plates laying on each other and you want to know if they are one piece or two by examining normal to the plane of the discontinuity, you will not be able to tell with an X-ray, but an ultrasound will easily make this distinction. If you inspect the joint from a 90 degree axis, the X-ray may provide clearer results.

And it also depends on the sensitivity requirement. If properly selected, UT not gives a far higher sensitivity, but infact some times it is too sensitive to give false signals.

additionally signal attenuations are a bit more of an issue in these. I have had quite a few joints passed in RT but failed in UT - re-tested again passed RT- but the UT failure forced us to repair.