From ISO 5817 quality level to weld radiography: technique class and acceptance level

"Quality level C according to ISO 5817." In fabrication, pressure equipment and piping specifications this line shows up again and again, and it tells the people doing the radiographic testing more than it seems. It sets the technique class used to make the image, per ISO 17636, and the acceptance level used to judge the indications, per ISO 10675. This tool rebuilds that chain for steel, nickel and titanium (ISO 5817) and for aluminum (ISO 10042), with film, digital detectors or radioscopy, and adds the notes that help when reading the images.

The result updates with every choice, together with a specification line you can paste into the inspection plan. The other tab applies the same logic to ultrasonic, penetrant and magnetic particle testing, limited to the cases that can be checked against public sources. When a case is outside the standards cited, the tool says so instead of guessing.

Where the correspondences come from, and where they stop

For radiography the source is ISO 10675-1 (2021 edition), whose Tables 2 and 3 link quality level, technique class and acceptance level, the first for film (RT-F) and the second for radioscopy and digital detectors (RT-S and RT-D). Part 2 of the same standard, also from 2021, repeats the structure for aluminum with the quality levels of ISO 10042. Level B leads to class B and acceptance level 1, C to class B and level 2, D to class A and level 3. A footnote adds that for level C on circumferential welds, the minimum number of exposures may follow class A.

The limit is set by the 2025 edition of ISO 17635, the general standard for nondestructive testing of welds. Its scope warns that acceptance levels are not a direct translation of the ISO 5817 and ISO 10042 quality levels, because they are linked to the overall quality of the batch of welds. A joint accepted to acceptance level 2 in radiography can therefore still miss quality level C because of a surface imperfection, such as undercut, that is measured by visual testing. That is why ISO 10675-1 asks for accessible areas to be visually tested before radiography.

Two decisions belong to other documents and stay outside the tool. How many joints to test is set by the application standard or the specification, and in steel structures it comes from EN 1090-2 through the execution class, as explained in the article on who decides weld radiography acceptance levels. The written procedure and personnel qualification to ISO 9712 remain with whoever performs the test, and the practical choices are covered in the guide on how to organize weld radiography.

The costliest step is the last one, checking that the system on your shop floor can actually deliver the required class. Class B calls for stricter IQI values, tighter conditions on source and distances and, with digital detectors, a higher signal-to-noise ratio, which a tube or detector chosen for class A does not always reach, especially when the part geometry forces short distances. The possible setups, from portable tubes to systems tailored to the application, are collected on the industrial radiography page.

The full correspondence table

The table collects every case covered by the tool, in the form an engineering office needs. The radiography rows come from Tables 2 and 3 of both parts of ISO 10675, published in 2021, and the ultrasonic, penetrant and magnetic particle rows from the 2018 edition of ISO 11666 and from Tables 5 and 6 of IACS Unified Requirement W33, listed in the sources.

Radiography: quality level, technique class and acceptance level

* On circumferential welds the minimum number of exposures may follow class A. Suffix X can be added to acceptance levels 2 and 3, making any indication longer than 25 mm unacceptable. Other metals, such as copper, are not covered by the ISO 10675 series except by agreement between the parties.

Ultrasonic, penetrant and magnetic particle testing on ISO 5817 joints

In the 2018 edition of ISO 11666 the table refers to the 2014 edition of ISO 5817. The PT and MT levels (2X and 3X) come from Tables 5 and 6 of IACS Unified Requirement W33 (Rev.2, July 2026), which refers to ISO 17635. Check them against Annex A of the edition named in the contract. For austenitic stainless steel, nickel and titanium the PT row uses values taken from a requirement written for hull steels (IACS UR W33), so confirmation against Annex A of ISO 17635 is essential. Outside the scope here are ultrasonic testing of materials other than ferritic steel, magnetic particle testing of non-ferromagnetic materials and penetrant testing of aluminum (ISO 10042).

What to send us for an assessment

If the result raises a real issue, for example a class B requirement your current system has to meet, a few details are enough for us to assess it:

• quality level and standard (ISO 5817 or ISO 10042) stated in the specification, with any suffix X;
• material, thicknesses and joint diameters;
• required technique class and the technique you use today, film, CR or digital detector;
• available source and distances imposed by the part;
• volumes, in joints or exposures per month;
• the defect that worries you most, volumetric or crack.

You can send them by email to info@dipitech.com or on WhatsApp, with an extract of the specification, or use the contact form. To choose between methods, see also the comparison of radiography and ultrasonic testing and the page on PAUT and TOFD ultrasonic instruments.

Cite or link to this tool

If you work in an engineering office, teach at an NDT school or answer questions in industry forums, you can link to the tool from your pages or course notes. The code below is ready to paste.