penetrant test on piping

[smart143]

We had done 3" branch weld in A182 F304 material on header of 30" (A358 304)per B31.3. the inside weld penetration was flush grind with Branch ID per customer requirement.

As per customer drawing 100% LPI was specified for branch weld.

We have performed LPI from outside weld & it passed without any Indication.

The subjected branch weld was LPI tested by customer from inside at their end & they have send rejection note to us saying that branch weld fails in LPI. The indications were rounded one but report mentions that those are crack like indications. Now we are in discussion with customer on below two issues

1. LPI from inside for 3" branch weld is not required, as in principal LPI for branch is done from outside(Is it in line with ASME B31.3??)
2. even though customer has done LPI from inside , the indications are rounded but they report it as crack like indication. per my understanding only cracks are not acceptable in LPI per B31.3 , the size of indication doesn't matter.


Can any body put insight on this issue.

 

[racookpe1978]

Have to see the pictures of the dye pen test.

A crack "can" be indicated by a row of small "dots" ... Or not. A single "dot" could be a hole, or could be meaningless.

 

[DekDee]

smart143,
What was the thickness of the branch pipe ?
If it is Sch 5 or Sch 10 then surface porosity is not acceptable.
If it is Sch 40 or Sch 80 then it is N/A (Table 341.3.2 Note 5)

There are two ways of interpreting the 100% LPI detailed.
If is is a general statement it means all branch connections require LPI.
If it notes 100% LPI and identifies a specific weld (eg.in the tail of a weld symbol) then that means both surfaces of the weld, external and internal.

Some contracts state that the client has the right to perform additional inspections - if the additional inspections show no defects the costs are borne by the client, if defects are identified then the costs are borne by the contractor

Regards,
DD

 

[smart143]

Its difficult to attach picture but what I can see a single rounded indiaction of 3 mm Diameter at 4 locations spread over 4 nozzles, means each nozzle has onnly one indication of max dia 3 mm.


The thickness of branch is sch 80(7.62 mm) so per note 5 of Table 341.3.2 surface porosity is not a cuase for rejection. the indications looks like a surface porosity to me.

Again the 100% LPI is written specific against the branch weld in tail of weld symbol.

So does this mean that customer has right to perfrom LPI from inside & outside. I often see the same symbol for all drawings they supply. if that is the case then it is not practicalbe to perform inside LPI when header is of 4" or smaller with branch of 1.5".

 

[smart143]

Its difficult to include pictures but all 4 indications are of rounded nature with Max dia of 3 mm. they apart from each other by atleast 30 mm.

branch is of sch 80(7.62 mm) so per note 5 of Table 341.3.2 surface porosity is not a cause for rejection.

The 100% LPI is spefied against branch weld with weld symbol tail showing 100% LPI.

So you interprete it as inside & outside requirement.but do we have any chance to defend it.

I see quite a few drawing having same weld symbol with same 100% LPI requirment for 4" header with 1.5" branch connection. In that case inside LPI is not at all possible.

 

[DekDee]

smart143,
Most of my work is for clients rather than contractors so I will give you my opinion of what I think may have happened.
When your pipe and branch have arrived at the customers premises they have visually inspected the branch and have seen something they were not happy about and have decided to LPI the inside of the branch to confirm.
Firstly, I would question the report you have been issued - there is no defect in any code I know of that can be noted as a "crack like indication" - it is either a crack or it isn't.
How can they reject something that looks similar to a crack but they are not sure ?
If it is a "crack like indication" it can be reported as such but it cannot be sentenced - it is purely noted so further investigation can be conducted.(further excavation and rechecking with LPI)
Secondly, just remember - if you as the contractor could access the branch pipe due to size of run pipe to get in and grind it smooth then obviously the customer can as well.
If you had not ground the weld it is possible there would have been no indications noted.
You may have exposed internal porosity by grinding off the weld metal so you should have performed LPI on the internals just to be safe.
Easy to say in hindsight but maybe a (hopefully not too expensive) lesson for the future,
Cheers,
DD

 

[tbuelna]

Interesting topic. Unfortunately I am not familiar with the acceptance criteria for penetrant inspection noted for piping system welds. But I am familiar with similar specifications used for aerospace welds. And in these specs a single rounded surface defect with a diameter of .50T or less is not cause for rejection for any class of weld with a wall thickness of .18 inch or more.

Also, as another poster pointed out, based on the shape of the indication it is likely a gas porosity defect. And it was only exposed by the penetrant inspection as a result of the overlaying weld surface being ground away. If the overlaying surface of the weld had not been removed, the penetrant inspection would likely not have exposed this subsurface defect.

In the end, most weld specs allow this type of defect to be ground away and re-welded. So it would seem the problem can be resolved fairly easily. In the future, Smart143 could make things easier by agreeing to a contract that allows the penetrant inspection to be performed prior to grinding the inside surface of the weld joints.

 

[DekDee]

tbuelna,
If you perform dye penetrant on the as welded surface you then have a dye penetrant report that is useless after the reinforcement is ground away.
The purpose of a surface method inspection is exactly that - an inspection of the exposed surface.
Hypothetically you could have a 100% pass after dye penetrant testing on the as-welded surface, walk away and after the reinforcement is ground away it is now 100% fail but no-one will know and it will be put into service "as-is".
Regards,
DD

 

[tbuelna]

DekDee,

Thanks for the response.

When you say "grind away the reinforcement", what does that mean? I assumed the post-weld grinding of the inside (backside) face of the single-sided pipe butt welds was primarily to remove weld drop or blend mismatched edges, to improve flow conditions. Is there actually some type of temporary reinforcement or locating feature used on the ID of pipes that gets ground away after butt welding?

I've designed some butt welded tube/duct assemblies for critical aerospace applications, and sometimes it is allowed to remove a limited amount of material to blend the flow surfaces after welding, or to locally remove very small amounts of surface material from a weld to blend a small surface defect. With regards to penetrant inspection of welds, if there were any grinding/material removal allowed on the weld, a penetrant inspection was usually required before and after any significant material removal from the weld area. In fact, sometimes a chemical etch of the weld area was required prior to the penetrant inspection to remove any smeared material from grinding or machining or to help expose cold laps in the weld edges.

Obviously, penetrant inspection can only detect surface defects. But if subsurface weld defects are a concern then wouldn't some other type of NDI process (like ultrasound or eddy current) be a better choice for the application?

I am interested to learn more about how other industries do things. So thanks again for taking the time to respond.

Regards,
Terry