NDT of drum

[fazel1]

Based on ASME SecI PW11 it is needed for one of our HP Drums to be full RT , but our licensor have told us to do 100% RT and 100% UT on all circumferrential and longitudinal butt welds of the drum. (The drum tickness is 90mm and the pressure is about 150bar). what kind of defect can we find by UT which we can not find by radiography?

 

[metengr]

fazel1;
Are you sure about having to perform both types of volumetric examinations for the drum welds? Normally, either RT or UT is performed, unless you are using the electroslag welding process. For this process Table PW-11 requires UT and RT.

Regarding advantages and disadvantages, the sensitivity of RT can be limited because of base metal thickness and geometry of the component. Ultrasonic examination is more tedious and requires operator experience. UT is conducive to finding lack of fusion defects, soldification cracks and hydrogen embrittlement (delayed) cracks.

 

[JR97]

fazel1

I concur with metengr's synopsis on UT. It is not customary to impose both inspections to insure weld integrity. In my experience when the customer imposed both RT and UT, he intended to implement an in-service program using UT - only need access to OD, no radiological restricted area, etc. The pre-service inspection was to establish a baseline for the in-service inspections. That may be the reason why the double inspection.

 

[ijzer]

RT of thick wall vessel will not detect defects parallel to the surface , like laminations in the plate and lack of interpass bonding. These defects can better be detected by US.

 

[VeryPicky]

RT and UT were historically a powerful combination for a critical equipment. RT is great to find and evaluate voluminous defects and planar defects that happen to be in the plane of X-ray. Manual UT is excellent for all the defects that RT does not find. RT was the only volumetric NDE method accepted by ASME BPV Code and UT was not due to a sensitivity to the operator technique and lack of permanent record. Code case 2235 allows use of automated UT (TOFD or Phase Array) in lieu of RT. Those new UT methods satisfy both ASME concerns and could replace RT.
Having said that, I understand that you are applying the licensor requirements. My advise: talk to them and find out whether these are not outdated. If they insist on both NDE methods, well, you don't have much choice, do you? That is: if you want to have their warrantees...
I hope this helps.

Putting Human Factor Back in Engineering

 

[AIAIAANI]

Since ASME Sect.I requires RT (although UT RECORDED, is permitted by Code Case 2235) Engineering Companies usually require both because for heavy thickness UT is more reliable. Mauro

 

[ndeguy]

Which welding process will you be using and what bevel angles? What is the drum's OD? Automatic welding or manual?

RT is insensitive to lack of sidewall defects unless the beam axis is aligned parallel to the bevel face. And if using automatic welding without control of stop-start locations then stacked LOF defects could give you a short defect of great through-thickness height. Most automated UT acceptance criteria will call for their repair based on Engineering Critical Assessment (ECA). Read ASME Code Case 2235-9 to get an idea of how acceptable indication lengths diminish with increasing indication height.


RT may have a part to play in the inspection, especially looking for volumetric indications such as porosity or slag (if using a flux welding process) but at 90mm wall thickness you have to choose your radiographic method very carefully in order to obtain meaningful results. This will be at the extreme limit of the penetrating power of Iridium-192 - perhaps beyond its capabilities which are normally put at about 75mm. Exposure times can certainly be very long, and keeping the source and film stable can be a problem if there is a possibility of vibration or shocks in the area. However, both image sharpeness and contrast are notably better than with Co-60. If you are able to use solid-state radiography instead of film that will shorten the exposure time a lot.

If the drum geometry precludes single wall techniques so you have to use a double wall technique then RT will be limited to Cobalt 60 or a specialist set-up such as a Linac.

You could RT after filling about 25mm in order to ensure a sound root before full welding. Combine this with UT of the root because at this thickness you may be constrained to use UT as the final acceptance so its better to know what root indications the UT will show before fully welding.

Once fully-filled then pre-PWHT TOFD of the completed weld and manual pulse-echo UT of the root and cap as some defect indications may be masked by the backwall and lateral wave signals on the TOFD display. Finally repeat the TOFD and manual UT after PWHT.

Good luck!


Nigel Armstrong
Karachaganak Petroleum
Kazakhstan