Acceptable Branch Connection B31.3

[BCbrad]

I am looking for an acceptable branch connection, where the branch is 1-1/2" OD and the run is 4-1/2" OD. Fig 328.5.4D illistrates 5 connections none of which suit my application. The latter part 328.5.4(a) states that 'the illistrations are typical and are not intended to exclude acceptable types of construction not shown'. So I am continuing to look.
The branch deos not require reinforcement, the branch connection goes through the wall of the run (4" Sch40). I would almost like to describe is as a socket weld as per fig 328.5.2C but I dont know if that adquately describes the connection.

Any thoughts or recommendations? Thanks.

 

[ColonelSanders83]

A connection of that nature is not shown due to its poor mechanical strenth.

If you look at the branch connections allowed by code you will notice that they all involve foll penetration attachment welds into the pressure boundary at the branch.

A fillet weld joining a pipe as you have shown is not acceptable. I would not allow such a joint in any of my code systems. The fillet weld you have shown on your sketch is not even large enough to develop full strength of the pipe. You should use one of the detailes shown, I would recomend fig 328.5D detail 1 for your application. Of course this assumes that you dont need additional reinforcment per the calculation you have done in accordance with para. 304.3.3

A question properly stated is a problem half solved.

Always remember, free advice is worth exactly what you pay for it!

http://www.ap-dynamics.ab.ca/

 

[BCbrad]

Thank you for the reply. I am well above the area required for reinforcement. I suppose it was the Fig. 304.3.3 that threw me off. The welds illistrated are simple fillets showing no penetration.

Is there a direction that I can look to find out the strngth of a simple fillet weld? I agree that the details 1 would be the most effective. However, there are 1440 of the branches to make and the fabrication time needs to be considered. So if I can argue that the fillet is inaquate then I have some ammo. If the fillet is adequate then I can sleep at night. The max pressure is 100psig, if that helps.

 

[ColonelSanders83]

What temperatures will the system be operating at and how many cycles will it undergo?

Branch details are governed more by cyclic design (usualy thermal growths but not always)than by straight pressure retention.

What are the stresses like in the branch? Is it acceptable using the SIF's for an unreinforced branch connection, or do you need a fitting tee for it to pass?

Even though there may be 1440 branches the weld you propose is not a code complient joint. End of story. Without doing a detailed engineering analysis in accordance with section VIII of the boiler and pressure vessle code you cannot qualify the joint for use.

Also I don't know if you are the owner or not, but if you are doing the work for someone else, the owner can throw a spanner in the works and force you to meet code at any time.

A question properly stated is a problem half solved.

Always remember, free advice is worth exactly what you pay for it!

http://www.ap-dynamics.ab.ca/

 

[ColonelSanders83]

In Reading your second post again I should also mention that Fig. 304.3.3 is for Area replacment calculations only (and is ilistrated accordingly, physical construction must be in accordance with Fig. 328.4.4D in order to meet acceptability. You might notice that all acceptable branch details shown in code figures have varying types of full pen welds.

A question properly stated is a problem half solved.

Always remember, free advice is worth exactly what you pay for it!

http://www.ap-dynamics.ab.ca/

 

[rneill]

I have never seen anyone attempt to make this sort of piping connection with a simple fillet weld. I have only ever seen branches done using one of the details shown in Fig 328.5.4 with a full penetration weld.

B31.3 does not permit partial penetration welds for branch connections and specifies that the weld dimensions can not be less than shown in Fig 328.5.4D. Refer to para 328.5.4 (b) and (d) in the 2008 edition of B31.3.

If the system is required to be Code compliant, you must use a full penetration weld regardless of any analysis you may perform. If the system is not required to be Code compliant then it is up to the end user to determine whether the detail you propose and your supporting analysis is acceptable. If I were the end user, I would not accept such a joint even in a low pressure system due to the poor fatigue performance you would have and the potential for corrosion related problems (depending on the service environment).

 

[BCbrad]

The 1.5' OD tube has a wall thickness of 0.105". There is some concern of burning throught the wall using a full penetration weld. We will have to do some practice welds using full penetration to get an idea of the feasability of this project.
An alternative is sockolets, but this also causes some challenges because the span between the 4" headers is 15'. Each header has (13) 1.5"OD tubes at 5" spaces. So I am concerned about mis-alignment. Regardsless, staying within the code requirements will provide an intresting challenge.

 

[rneill]

The 0.105" is incredibly thin for such a large diameter tube. At that size (1.5') is is easy to deal with internal pressure but not so easy to deal with thermal (expansion / contraction) and longitudinal (self weight and external loads) stresses.

I recommend that you make sure you have checked all these other load cases quite thoroughly for such thin wall tubing.

It would be possible for loads in the branches to cause movement (flexing) of the header leading to fatigue failures so you might want to look at that as well.

You might also want to look at PFI ES-7 "Minimum Length and Spacing for Welded Nozzles".