Mating RFWN Flange Joints of Unequal Thickness 1

[auzie5]

Is it a code requirement (ASME B31.3 or CSA Z662) that flanges have matching inner diameters?

For example:

Can an NPS 8 x STD (ANSI Class 300) RFWN flange be joined to an NPS 8 x XS (ANSI Class 300) RFWN flange without any consideration given to the step introduced from the different flange IDs?

In my mind, the 6.3mm step between flange bores would encourage turbulent flow into the gasket and increase the potential for gasket failure. However, I’ve searched through ASME B31.3 and CSA Z662 and cannot find any restrictions on mating flanges of unequal wall thickness.

In cases where pipe of unequal wall thickness (i.e. XS vs. STD) mate at a flange connection, I’ve always understood it to be prudent to use identical flanges (i.e. matching IDs) and make the following butt joint tapered (as per code allowed detail) to suit the mating pipe wall thickness.


Thanks in advance for any replies.

 

[tothepoint]

Suggest you use a reducing flange: NPS 8- NPS 6 for instance. A class 150 flange bolt pattern will not match up with a class 300 bolt pattern.

 

[gr2vessels]

There is no code restriction on joining flanges with different ID, provided they have same NPS and same rating and the same standard. Remember, the vessel nozzle neck thickness is designed / selected per pressure, temperature, CA, tolerances and external loads, as opposite to the interconnecting piping selected only on the basis of pressure and temperature. It is quite frequent the nozzle neck ID to differ from the piping ID, mainly because there is no link between the piping designers and the pressure vessel contractor.
While it's not unheard of same rated piping using different wall thickness (for other than code reasons), it is prudent to taper the joint as you mentioned.
Cheers,
gr2vessels

 

[racookpe1978]

I think, tothepoint, that the OP is confusing things with his choice of words.

I think - but I can't tell for sure - that he has two Class 300 8 inch dia flanges, and wants to weld a Std (Sch 40) pipe to one end of the resulting assembly, and a different Sch 80 thicker-walled pipe to the other end of the 2nd flange.

 

[auzie5]

Thanks gr2vessels for pointing to the vessel/piping boundary limits as an example where one typically encounters this issue. I was also looking at flanged valve (ASME B16.34) connections as another example where the bore of the valves' flange will not match the bore of the mating pipes' flange (in most instances).

And thanks racookpe1978 for pointing out that I could have been more clear in describing the issue at hand.

To clarify:

Assembly 1:
NPS 8 x STD RFWN flange (ANSI Class 300) welded to NPS 8 x STD pipe

Assembly 2:
NPS 8 x XS RFWN flange (ANSI Class 300) welded to NPS 8 x XS pipe

Bolt Assembly 1 to Assembly 2 resulting in a 6.3mm step between flange faces at the bolted connection (due to difference in ID).

An example where this may be encountered is when tying into existing piping systems where the engineer has decided that the new spool need not match the existing spool wall thickness. Rather than introduce a transition piece, the engineer has the choice to introduce a set of flanges to step-down/up the new spools' WT.

I still believe it to be good engineering practice to avoid this step where possible using identical flanges (i.e. matching IDs) and taper the following butt joint to suit the mating pipe wall thickness but clearly in certain cases (like flanged vessel or valve connections) this may not always occur.

I'm good.

Thanks.

 

[rmw]

First, any answer given here below assumes that the service requires no more than the Std wall flange and pipe and that the XS flange and pipe are overkill. If XS is required by the service, then run don't walk to get the std wall flange and pipe out of there and replaced with XS stuff.

That aside, you are absolutely right, there will be a step due to this mismatch introducing some small amount of flow disturbance, either a sharp edge flow area reduction or increase at the step depending on the direction of flow.

But let's talk real world scenarios. As already mentioned, sometimes flanged valve to flanged pipe connections also contain steps and if a pipe fitter does a crappy job of aligning flanges when they are bolted up and has one high to the extent of the bolt hole slop and the other one low as it can go in the bolt holes, there can also be a step in one half of the pipe that introduces a flow disturbance.

Often times there is a similar mismatch where, for example, std wall elbows and fittings - flanges, for example - are used with lighter schedule pipe with thinner walls. At least in those cases, if the code is followed, the thicker piece will be beveled to the thickness of the thinner piece so the step is not so pronounced, but there is nonetheless a step there. As lighter schedule fittings are often difficult to source to go with the lighter wall pipe, this can occur many times in a pipe run.

If your piping design is so tight that the flow loss introduced by this small step in the piping is going to cause problems, your process is in trouble already.

So I think you are right. But just being right might not justify the cost of making it right. Adding a spool piece to transition between the two will still involve a (beveled) step somewhere in the transition run, so the ony out you have is to replace the entire XS run. Can you afford to do that just to be right? Only you can make that determination.

rmw

 

[BigInch]

The pressure issue is real. The welding mismatch in wall thickness is real. The flow disturbance issue is BS.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[jtseng123]

I will provide my two cents and realistic answer to this thread:
1. How many 8" STD ID flanges bolted to 8"-XS ID flanges: if not millions, it will be thousands around the world.
2. For a typical tower with 50 nozzles, how many this kind of mismatch will occur: more than 50% of all nozzles.
3. Why this happened all the time: because ASME SEC VIII Div. 1 has minimum nozzle thickness requirement, that drives the thickness one schedule higher than piping designed per B31.3.
4. Why Div 1 is doing that: I 'guess": it is easier to replace piping than repair a vessel nozzle, so make the vessel nozzle stronger won't go wrong, and cost minimal.
5. Does anyone care about ID mismatched: hardly heard.
6. Does anyone try to taper the small ID so to make a smooth transition: hardly heard. (this taper means 3:1 or 4:1 taper, just at the edge of the flange)
7. Is it recommended to taper the ID ? see point 5, and if it is directly connected to pump, compressor, or the velocity is extremely high, just do it. It cost nothing.
8. Is there any additional spool needed for the mismatch issue: never heard of, and you don't need it.
9. Will the flow turbulance be taken care by the taper: yes, and the behavior has no difference with the same ID of flanges, only velocity minimally increased. If you don't believe, run CFD to see it.
10. So what is the conclusion: if you worry it, do whatever you want or see point 7, otherwise, leave it alone.

 

[curtis2004]

If velocities are low - leave it alone.
If velocities are high and expected to affect piping system connected then there are the following things can be done in order to mitigate possible problems:
1. Transition pipe from one thickness to another;
2. Taper inside piping;
3. Machined radius at the place of step.

Regards,
Curtis

 

[BigInch]

I believe he is talking about joining flanges face to face, so wall thickness mismatch when welding neck to neck comments would not be applicable.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[rneill]

Hmmm, we don't specify the flange bore on the valves we buy and install and we use the same ASME Class 600 flanged valve regardless of whether the attached pipe has a Sch 40 bore or a Sch 160 bore ... we've never worried about it for valves so I'm not sure that this is a real problem ...

Of course, for a given pressure class, you aren't going to end up with too much of a range of possible flange bores. If you have a very large difference in mating flange bore, it may become an issue?

 

[BigInch]

Yes it is a problem due to stress concentrations and axial pressure stress, due to Poisson's ratio, x area of steel on one side is not equal when multiplied by area of steel on the other side, which produces unequal axial loads. As those axial forces are redistributed, higher stresses at the joint can be created. The ASME codes have maximum allowed mismatches for thickness, which must be followed. Mismatches can also be a problem for a good weld too.

If it ain't broke, don't fix it. If it's not safe ... make it that way.

 

[phamdongduong]

Hi auzie5

I have the same issue regarding what is max. difference (step) between Equipment Nozzle ID and Pipe ID which not require any taper or trasision peace?
Since I just newly reading this post so do you have already conclusion?

Thanks for your response

 

[zdas04]

Typically 90-95% of the bulk flow is in the middle 80% of the pipe. Velocity outside of the middle 80% of the pipe flow area rapidly falls to zero at the wall (remember "no flow boundary" from college?). Or in BigInch's words "velocity concerns are BS".

When I look at stress concentrations due to welding I see a couple of orders of magnitude more localized stresses than you would ever see from Poisson's ratio considerations. Gotta call BS on that one as well.

8-inch pigs are pretty resilient, so you are unlikely to stick a pig in the difference between a standard bore flange and XS pipe. In smaller sizes this transition might impact a pig choice, but not in pipe that big.

My welding inspector friends are concerned about tapering the joint to get equal penetration of the welds. The explanation that I got for this lost me in the details, but the bottom line seemed to be a concern about creating laminations in the steel of the thicker pipe that would be weaker than a full penetration weld. He said it all with a straight face just after rejecting a 20-inch weld for misalignment so I think he believed it.

David Simpson, PE
MuleShoe Engineering

"Belief" is the acceptance of an hypotheses in the absence of data.
"Prejudice" is having an opinion not supported by the preponderance of the data.
"Knowledge" is only found through the accumulation and analysis of data.

 

[phamdongduong]

Hi zdas04

Thanks for your information, my issue here is flange connection between Equipment Nozzle and Piping Flange which has difference ID. So the question here is Does Flange need to be taperred to match Nozzle ID? and if need what is max. much difference (step) which we need to consider about tapering to match nozzle ID.

Thanks alot in advance

 

[moltenmetal]

Choose the right gasket, and you can mate any RF flange to any other RF flange of the same size and class. If it were not so, you'd have a problem with pretty much every single valve, pump or other flanged item you bolted up to, none of which are guaranteed to have the exact same ID as your piping.

Choose the WRONG gasket, and you will have problems mating RFSO flanges to anything else.

 

[jnich]

You will always want to match the w.t. of your RFWN flange with your pipe (ie STD x STD or XS x XS). This has to due with weld penetration ability, as zdas04 mentions, and passing your weld x-ray tests per DOT (if you're DOT regulated, like all the work I do). I don't believe what BigInch states about axial stress differences is the primary concern.

For ASME B16.5 RFWN flanges, unless you specify otherwise it's assumed to be a STD w.t. As I understand, there are no issues with different pipe w.t. connecting together via the same Class 300 ASME B16.5 RFWN flanges.

I'm not sure about B31.3, but B31.4 addresses unequal w.t. throughout the spec. See section 434.8.6 in the 2009 edition for more details.