Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

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

Standard face to face tolerances on welded components. 1

Status
Not open for further replies.

bryansonnier

Mechanical
Dec 20, 2011
80
0
0
US
I'm looking for the end to end acceptable tolerance for welded components. We manufacture valves here and we are getting the flanges (2" 600 sch80 in this specific case) welded to the end connections for the valve. The best answer I can find is in ASME B16.10 section 5.1 (straightway valves) where they list a tolerance of "+/- 0.06in shall be allowed on face-to-face and end-to-end dimensions of valves NPS10 and smaller". My issue the valves it seems to show in this spec look like fully casted valves so i was wondering if there is a different spec for welded end connection valves or components.

We are trying a new welding shop and are getting push back on the tolerance i've listed on the weld drawing of +/- 0.03" for the face to face of the end connection (2 ends at 0.03 each gives 0.06). They are telling me that they weld pipe all the time and are used to 1/8" tolerances. Any help would be greatly appreciated.
 
Replies continue below

Recommended for you

Desertfox,

Thank you for the help. I wonder if there is a specific code governing that dimension. It looks like it may be ASME B31.3 so i'm looking now but haven't found it yet.

It looks like 1/8" for the section, which would be 2 welds giving a 0.0625" tolerance per weld. That would be a little easier for the welder to hit than 0.03 which i have right now.

Still a specific code and location would make me feel a lot better. I can't imagine i'm the first person to ask this question so i'm sure there is a spec that governs this directly.
 
Ok, so the spec from the Pipe Fabrication Institute PFI ES-3 is what the link was based on (referenced at bottom of page) but it seems a little loose for what i'm looking for. It basically says for a straight run of pipe it has a tolerance of 1/8" unless there are a bunch of connections then it can be more. That is a little too unspecific for my taste.
 
Hi Bryansonnier

Well given the number of configurations and flanged joints on pipe work I can’t see how they could tie it down, there are an infinite number of configurations one can have.
The pipe people you are dealing with have mentioned 1/8” so why not stick with that as a maximum and then if you have a complicated configuration where they want more leeway for welding make sure they contact you before welding said configuration, that way you are still in control.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein
 
My problem is we are fabricating a valve with flanged ends which it seems should conform to ASME B16.10 which gives a +/- 0.06 (2mm) for valves of the size we are making. However, all the images that reference valve face to face for flanged valves all seem to reference a fully casted valve where tolerances would be easier to hold as the ends would be machined. This is from API 594 to ISO 5752 to IS 130 (india standard) and they all seem to mirror or flat out reference the ASME B16.10 spec leading me to believe that one is king.

My problem with all that is maintaining a +/- 0.03 per weld is asking a lot of welders.

I think i will continue to ask them to try for 0.03 but so long as its 0.06 or less i'll have to be happy with that.
 
Do you need tolerances that tight? If you do, is machining AFTER welding a possibility? You have to remember that things move during welding, and then often shrink after cooling.

The devil is in the details; she also wears prada.
 
Hi Bryansonnier

Why not buy blank flanges and finish machine them post welding? Apart from the increased cost I can’t see a problem doing that.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein
 
Personally i'm not sure what tolerance i need as i'm not the one using these valves. And short of doing a massive customer survey with customers who are usually extremely resistant to giving any kind of information i will have to rely on some sort of industry standard. These valves aren't specifically an "industry standard" type valve but i want to avoid causing the customer unnecessary hardship trying to get my product to fit in their system.

Yes machining AFTER would be ideal accept it would add a LOT of extra cost and we are already slightly more expensive than our competitors. We make up for it with service and features but that will only go so far. So i'm stuck trying to ask tolerances of welders they are not used to and probably can't reliably hit. But if you shoot for the moon and only make it half way you've still traveled a long distance. At least i think that is the approach i will have to take on this unless i can be pointed to a specific spec that deals with welded valves. That way i can point to it when the customer complains to my boss that valve face to face dimensions aren't as advertised (even though the issue was probably with his pipe and not our valve).

The PFI ES-3 spec is probably the one i need to hit but i'm not 100% sure. This would still require +/- 0.06" per weld which is better but still less than the +/- 0.125" they tell me is the best they can do.
 
Desertfox,

That would be THE problem however. Our product is not a 1 off specialized product where that sort of cost can be easily passed down to the customer. Ours is more of a configurable commodity item but our sales are not nearly large enough to justify a individual casting for every connection type. So i'm kind of stuck welding standard weld neck flanges to our valve ends.
 
PFI ES-3 is the industry standard used for tolerances for fabricated pipe parts including valves, pipe fittings, and flanges. Your customers should be aware of this.

The devil is in the details; she also wears prada.
 
DVWE,

Thank you. I wasn't exactly sure the full range that spec encompassed it will help a little i guess as 0.06 is twice a easy to hit as 0.03, lol. Thank you for your help.
 
bryansonnier,

Your fabricator is the expert on what tolerances they can achieve. You can always try other fabricators. Failing that, you need to make your design work with [±]1/8"[ ]tolerances.

--
JHG
 
There is a certain expectation that the flanges on both ends of a valve will be somewhat parallel to each other. And that the flange holes be "clocked" to each other, or some feature on the valve.

Length is not the only thing to consider.
 
Yes i am aware, thank you. Thankfully these haven't given us any trouble from the weld shops. It looks like i got the tolerance on these correct through a bit of blind luck.
 
bryansonnier

I don't have that specification available could you take a screen shot of the specific paragraph.
I have not worked with standard pipe specifications, however I have worked exotic shuttle, centaur fuel lines.
and I had to over come extremely close interface of the flanges and it had to be square parallel's to each each other
and the distance of the flange face had to be precise.
I set up weld fixtures to hold the pipe. starting with the inner most welds and worked out towards the flanges., I left stock on the end of the pipe for
the last weld. because each weld will have weld shrink. I ran test on set up parts as how much contraction, then allowed excess at the last end of the weld.
the flange faces were bolted in place. as not to move for the last weld. since there will be weld shrink on that last weld , there was two removable fixtures at each end, one
for welding and one for final size. from memory each weld with that material and that thickness and that weld schedule had approximately
.060 inch shrink at each weld. but this was not cheap,
in your case have them weld record the weld shrink, than allow the excess on the last welds, have them make a weld fixture to maintain face interface.
I can say it was closer than .030.

 
Status
Not open for further replies.
Back
Top