The best NDT method is competent visual inspection, both at fit-up and post weld.
UT and RT are for verifying internal soundness of welds.
The best choice between those two is dependant on circumstances such as joint geometry, access, availabilty and cost.
Each method has abilities to identify discontinuities the other has missed, so they are often used in tandem on highly critical joints.
Ultimately, the final definitions of which type and extent of NDT is to be performed is made by the project engineer and agreed to by all parties in contract documents.
Tim
Greetings Simon,
WELCOME TO THE AWS WELDING FORUM!!
To add to Tim's response, there is NO BEST NDT METHOD.
They all have their advantages and disadvantages. They all have limitations. They all have practical application.
As Tim mentioned, the best QC/QA system starts with trustworthy, knowledgeable, thorough Visual Inspection. Even most codes specify that prior to other methods of NDT, VT will be performed first. There are many reasons for this but I will refrain. But what needs mentioning is that VT is also an NDT method, nothing is destroyed and it is the most complete method to be sure all code requirements are met. UT cannot tell you if the preheat was right. Can't tell you if the electrode was stored properly. Can't tell you if the welder was properly qualified or that the welding was done within compliant parameters.
Now, all other things being completed properly, from a cost, safety, accessibility to members, time, aspect you are normally going to be best off for structural applications with UT. That does not mean RT does not have it's place and both have weaknesses and strengths.
Personally, in this order for my work, VT, MT, and UT. VT of the whole process. MT of root passes on CJP's and of a percentage of surface passes and fillet welds. UT of CJP's and when applicable Lam scans.
But, it is the EOR's job to choose how this is to be handled on their job to best meet the needs of the project.
Oh, and I am wondering why you are asking about structural RT vs UT in the ASME section?
He Is In Control, Have a Great Day, Brent
Apples and oranges. One is better suited for planar flaws (cracks, lack of fusion etc), the other for voluminous (porosity, slag, etc).
Neither are a magic bullet nor are they replacement for weld program control.
For either to find anything, there must first be a flaw to find. While even the best can make mistakes, too many times people focus on the NDE rather than maintaining their welding program properly.
Beyond that, your engineers need to define what type of flaw is most agregious to your product form and proceed accordingly.