NDT for socket welds 2

[shonuff]

Previously had socket weld(fillet)crack, analysis revealed the failure was fatigue due to harmonics. Changes are being made to correct the problem at other welds. Would like to know other than dye/pen what is another ndt method to inspect fillet socket welds to check for cracks??

 

[pennpiper]

First, the correct nomenclature (terminology) for what you are asking is "NDE"/
N = Non-
D = Destructive
E = Examination

Examination by definition is "Non-Destructive"

If you submit your weld sample to a lab for "Testing" they will destroy it in a "Test" machine therefore it is NOT non-destructive.

Now, what other NDE options do you have for Socket Weld joints?
VT - Visual - Good for all weld types, but not perfect
MT - Magnetic Particle - Not good for Socket-Weld joints
PT - Dye Penetrate - Good for surface examination
RT - Radiological (X-Ray) - Good for Butt-Welds, Not good for Socket Weld joints

As you have already found out there are few options.
In view of your problem with Vibration I recommend you take out all the Socket-Weld piping and replace it with Butt-Weld piping.

prognosis: Lead or Lag

 

[gtaw]

I disagree with the comment regarding the use of MT to examine socket joints. As long as the base metal is ferromagnetic, MT is a valid test method. As a matter of fact, I prefer fluorescent wet magnetic particles if the test area is free of ambient white light.

NDT, NDI, NDT, the only difference is the length of the stick stuck up the person's .....

Best regards - Al

 

[DekDee]

pennpiper,
Please explain where your interpretation of "correct terminology" comes from ?
NDT is Non Destructive Testing and has been since Jesus was playing fullback for Jerusalem !!!
Do a search of various codes - AWS references NDT only.
ASME references both.
ASNT stands for American Society for Non Destructive Testing - have you seen ASNE referenced anywhere ?

In answer to shonuffs question - if it is C/S then MT is your best option for surface defects.
On a previous project the Quality Manager misinterpreted the requirements of B31.3 regarding the 1/16" gap before welding and wanted to RT the sockets after welding to confirm the 1/16" gap after welding.
Showed a lot of defects in the fillet welds and cost the project huge amounts of unneccessary time and money but if you want to be 100% sure you have no cracks - RT and MT on your socket welds.
Regards,
DD

 

[racookpe1978]

If you have continued vibration, socket weld joints are NOT recommended against vibration.

'Tis expensive, but consider replacing them with butt-weld joints wherever vibration is high or medium.

 

[SectionIX]

Agreed with GTAW re: MT of ferromagnetic socket welds. Far superior to PT. For the record, RT is an acronym for Radiographic Testing.

 

[Duwe6]

NDE vs. NDT: The 'T' for testing implies that VT - Visual Inspection is excluded. 'E' for examination includes VT.

Nuke industry has been having SW cracking caused by harmonics. Their solution, developed thru EPRI, is a 2:1 weld. If the fitting requires a 1/4" socketweld, the weld leg on the pipe is to be at least 1/2". Seems to have cured the cracking issue.

 

[shonuff]

Duwe6, very interested in info for this solution. Thanks

 

[11echo]

Are you sure the cracking was caused by harmonics? ...And NOT by the welders bottoming out the pipe in the socket when they do their weld? They ARE suppose to first bottom the pipe in the socket, then scribe a line to indicate the depth of the socket, then back out the pipe so there is a 1/16" gap between pipe end and socket depth. This is for any growth between the weld and the socket end and the accompanying stress on the weld ...semi common problem these days.

 

[gtaw]

The problem comes about when the weld and adjoining base metal cools. As the weld is being deposited, the base metal is heating and expanding. Once the weld is completed, the weld and the adjoining base metal cools. In doing so, the weld is contracting, thus pulling the pipe further into the joint. As the pipe is pulled into the socket, the gap is closed somewhat. Were it not for the presence of the gap between the end of the pipe and the shoulder of the socket, the weld root would be placed into tension resulting in a root crack unless the weld is very ductile (as is the case with austenitic stainless steel).

Welds that are high strength with low ductility tend to crack when the welder forget to leave a gap between the end of the pipe that is inserted into the socket fitting and the shoulder of the fitting. As mentioned, the codes require a minimum of a 1/16 inch at the time the pipe and socket are fitted together. When these same joints are radiographed, it is common to see the gap is less than 1/16 inch. Again, the reduction of the 1/16 inch dimension is the result of the weld contracting upon cooling and drawing the pipe further into the socket. The larger the fillet weld, i.e., the more weld passes deposited, the smaller the gap becomes.

The suggestion that the weld against the pipe be made larger is something practiced by the U.S. Navy for many years. The nominal fillet size for Navy work is T by 1.75T where T is the wall thickness of the pipe. The long fillet leg is placed against the pipe. It has been shown to increase the life expectancy when fatigue is a consideration.

A good friend did his Master's Thesis in this very subject.

Best regards - Al

 

[shonuff]

We use gap-o-let to ensure 1/16 gap, and we are using RT now for info only, which shows adequate gap. Harmonics ratio calculations have been completed and that is where the information was generated. It was a "miss" on the engineering side, we are looking at replacing the socket welds with butt welds in the future, have redesigned dia. and leng of thermo-welds (probes) to alleviate some of the issues. Thanks for everyone's input. Another question, lol how effective is RT through insulation? Just curious

 

[DekDee]

shonuff,
Will not work well at all.
There is no problem with the source being a distance away from what is being radiographed but the film must be in contact.
If the film is not in contact the images portrayed on the film are distorted making interpretation to a code impossible.
Regards,
DD

 

[Duwe6]

RT thru insulation; it will allow you to see the corroded profile of pipes - ID & OD, up to about 18"NPS, and it will allow you to see the gap on SW's. Not very good for flaw detection of butt welds.

 

[SectionIX]

Hi Shonuff,

RT effectiviness is dependant upon (among others):

+ type/thickness of insulation
+ source (x-ray/gamma)
+ source to object/film distance
+ exposure geometry
+ film type
+ joint geometry
+ type/size of imperfection/discontinuity you are looking to find
+ Operator training/experience
+ material thickness
+ etc

 

[DekDee]

Section IX,
Isn't the most important aspect in this example - object to film distance if you are looking for a shot that is actually going to tell you something ?
Cheers,
DD

 

[SectionIX]

DeeKee,

Yea, the intent was to illustrate RT effectiveness is dependant on a plethora of factors. In this case, RT would be less than ideal for the application.