Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

O-Let Welding

Status
Not open for further replies.

Mechomatic

Mechanical
Apr 23, 2013
50
I know this topic has been beaten to death, as I've gone through pages of search results on the topic. I wanted to throw an application/proposed solution out to see if there might be something else I should consider.

1" Sch.40 SA106B pipe requires a 1/2" NPT connection on the side. GTAW, GMAW, and SMAW welding methods available.

-I've considered using Sch.80 pipe and backwelding a threaded Tee into place, but would prefer to not do so because the customer has been finnicky and would probably squash any backwelding plans.

-Using a half coupling was considered, but rejected for a small number of reasons.

-The (presently) prefered method is to weld a copy-cat Thread-O-Let to the pipe. I cannot claim to be a welding expert, but based on the information I can dig up, my proposed soultion is:

1) Clamp pipe near both ends into a length of unistrut and clamp the unitstrut in the middle (beneath where the o-let will be welded) into a bench vise, both to hold the member during welding and keep the pipe from bowing downward and spreading the sides of the unistrut.
2) Set welding machine to high end of amp range. I would leave the exact setting up to the welder since I'd rather they be able to make a good weld than sacrifice weld quality by specifying an amp setting they are not able to comfortably use.
2) Tack the o-let in place at the 12,3,6, and 9 o'clock positions (pipe axis along 12 to 6 line).
3) 1st/root pass in quadrants- From 1:30-4:30, 10:30-7:30, 1:30-10:30, and 4:30-7:30.
4) 2nd/cap pass in quadrants- From 12-3, 9-6, 12-9, 3-6.
5) Grind fillet weld to a flat finish.


One thing I'm not sure of is wait/cooling time between weld stitches and between passes. Any quick suggestions, or just the obligatory "hit the books" to figure it out? Are there other options I should consider to reduce weld distortion, or any changes to my proposed procedure that are obvious to those more experienced than I?
 
Replies continue below

Recommended for you

If you require a 1/2" NPT branch on a 1" run, Id suggest you take a 1"x3/4" reducing tee, and weld a 1/2" B16.11 half coupling on the tee's branch. For a cl. 3000 half coupling, the OD of the tee's branch almost exactly matches the OD of the half coupling. Taper the half coupling on the inside at the welding end to make sure to get a smooth transition in the root.
 
RE: O'let -- Unistrut will just bend. You would need to use 3"C-channel(+), and 'bridge clamps' or 3/4" long(+) tack welds for attachment. Let the weld bead in each quadrant of the O'let cool to below 250°F, unless you immediatly weld the opposite quadrant. Then, when 'welded out', perform a 'hillbilly PWHT" by torch-heating the weld area on the pipe to 1000°F or so*, and allow to cool slowly inder insulation. Should be [almost] straight when the tacks are released.

*Make d@*m certain that you do not exceede 1250°F in any spot while PWHT'ing.
 
Thank you for your replies. Not sure why I didn't think about a 1x1x3/4" reducing tee earlier. Probably would have gone that route if I'd thought of it sooner. Unfortunately, I believe the spool in question is needed to be finished tomorrow to be assembled and sent to painting. I think, in that case, I'll get with our friendliest welder about trying Duwe's suggestion. Fortunately, he- and not the usual guy- is the one with the spool drawing. The usual guy is a good welder, but grumpy as all get out and not likely to follow such a detailed weld procedure from anybody, let alone the newest, youngest engineer on staff.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor