Welding and ASME Standards

Neither. 100% of the given lot size are only required to be redone on after the third repair/failure. You need to specify how many welds failed in your #3. (By the way the min. requirements of B31.3 would not disqualify the welder, only the welds - to completely disqualify the welder would be up to the owner/design spec)

Upon failure of subsequent testing in your #3 lot the answer would depend:

1. What is the actual lot size specified? (100, 300, ...?)
2. What is the total number of welds in the entire system?
3. What is specified by the engineering design?

The purpose of progressive sampling is to provide a means to detect problems and rectify them before the entire piping system is completed. The 100% rework rule is to allow this before the system is done and has to be completely reworked. (eg. a matter of redoing 100 welds versus 500)

does anybody else experience a bad link for the image linked above?

right-click and "Save Target As" results in an 88 byte "non-image" as well...

I don't have a problem from my work computer (was uploaded from home) - if more people post saying there are issues I can resubmit it when I get home again.

Thank you sir.

I have seen a flow chart. I have some doubt that in flow chart is following,

1)Lot means whether it should be same kind of piping material classes or it may be different classes

2)Whether i can consider this flow sheet progressive sampling applicable for piping material class basis and/or welder basis.

Waiting for your reply with thanks

The code is pretty vague regarding this except for "each welder's and welding operator's work shall be represented".

It is left up to the owner/engineering design to prescribe a lot size and extent.

Take a hypothetical situation (and in my opinion, what I would do):

*******************

Installation of a new heat exchanger and associated piping (shell in/out, tubes in/out)

Shell side is cooling water, class 150 flanges, STD wall carbon steel; tube side is process gas, class 600 flanges, XS wall stainless steel.

Both are specified as normal fluid service, but with a 10% random RT on circ welds (which is above and beyond code min of 5% random RT on circ welds)

Both the shellside in/out & tubeside in/out each have (coincidentally) 100 welds each.

One would specify separate lots for each - which makes sense - you have different materials, different welding procedures, and different classes.

Shellside:

• 2 welders - each do 50% of welds, ie. 50 welds each
• 10% RT will require 10 welds in total, it would be prudent to split equally so that 5 welds of welder A and 5 welds of welder B are examined.

Tubeside:

• 11 welders - each do approx. 9 welds.
• HOWEVER, as mentioned above you need to include (represent) each welder.
• Therefore, you would need to, AT A MINIMUM, examine 11 welds (1 from each welder), which works out to 11% RT, but that is the way it has to be.
  

Mr Rajan, before anything you have to designate the lot to be (NDT) Tested. I had obtain easy results using LOTS no more than 20 welds. (For same welder, same process .......), dayly production could be also considerate a lot, but is no a good control. Please review ASME B31.3 point 344.1.3 foot note 2.

The rule I use is 3 strikes you are out. No matter how many welds a welder has made on the third repair he loses his job. If I pay for 100% welds I expect 100 % because a weld repair is not a chance thing that can't be avoided as some would think. If a welder is worth a darn he knows what will pass and what won't and he knows how to fix a bad spot as he makes the weld.