INTEGRALLY REINFORCED OUTLET FITTINGS - WELD SIZE 4

The problem with not filling the entire weld bevel on a weld-o-let fitting is the creation of a "notch" effect or stress riser. This notch will result in the intensification of thermal/mechanical stresses leading to possible crack initiation in service.

Are you filling to the OD of the fitting? There is a weld bevel that does need to be filled. However to go beyond the required weld size could cause Lamellar Tearing. It is also a waste of time and money.

The transition of the weld to the base metal is very important. See ANSI/AWS B1.11 or the CWI course book for some info.

Go to this website and download their Branch Connection Catalogue and look on page 35 there is instructions on installing a Weld-O-Let.

tr6,

I have heard of this Bonney Forge welding problem many times over the years. The problem is really one of design perceptions, welder expectations and the odd geometry of these larger fittings.

When welding small diameter "o-lets" to much larger diameter piping, "filling the weld to the bevel", is obvious. The contour of the finished acceptable weld looks similar to a standard weld with a fillet reinforcement.

However, when larger diameter "o-lets" are welded to piping of nearly the same size, the bevel on the "o-let" gets extreme and it is easy to assume that the weld must be larger than just the beveled area. ( You kind of have to see this to understand it..)

Bonney Forge has had the question asked a zillion times.... In, my opinion, the welding instructions included in the catalog could be made a little clearer when larger "o-lets" are installed.

my opinion only


MJC

Part UW 16 deals with the type of fitting,and the ones with outside welds require a tremendous amount of weld, I use them sometimes only when strictly necessary,

(I do not know about the B31.requirements but can be similar)
GB

tr6,
As metengr says the problem is not with the pressure retainment of the weld but with the stress levels during operation when the "notch" is subjected to bending loads which fluctuate. The original "Designer" will have assumed a certain SIF for the branch (if it were stress analysed) as indicated in the codes. These SIF's are based on the geometry of the fabricated assembly. If the actual assembly is somewhat different(due to insufficient weld being input or a poor geometry) then the SIF could be significantly higher. I have come across this time and time again whereby the fabricator assumes that the thickness of the weld to be applied only needs to be the same thickness as the branch pipe to which the O'let is welded. It is even more serious on stainless steel lines for example when you have a 40"nb Sch10s line which has had a 20"nb weldolet specified and due to distortion problems the fabricator has applied minimal weld to the branch weld whereas the Stress Analyst had assumed the Code SIF's when he analysed the piping to minus 196 Deg C. Scary or what???

These are unlisted components so the required weld size is that which is specified by the manufacturer of the component. Some fitting manufacturers have a clear weld fill point. Others, not.

The thickness of the weld actually has very little to do with the SIF. It is primarily driven by the contour of the weld, whether the contour creates a notch, or not. It is the weld on the sides of the run that is the primary issue.

bvi

Parts UW and PW deal with the size of the weld and the type of fitting demands the size. In this case the mfr only follows the ASME/ANSI B16 commands to fabricate a fitting
to comply with the Code so they are "listed" fittings and
thanks to the fitting people as BForge we have a guide to weld them.
GB

bvi,
Suggest you read up on SIF's. If the weld is small then you have an inherent "notch" formed and hence a stress raiser. In order to use the Code SIF's the weld profile should be as indicated by the Manufacturer and be "smooth". If the thickness of the weld is so small as to not "fill" the correct profile then the SIF is higher than the Code. So the thickness of the weld does have bearing on the SIF.

GenB -

The subject is piping so Parts UW and PW are not applicable. For B31.3 piping, the requirement is that the branch connection weld be full penetration. However, for these unlisted components, the size of a full penetration weld must be specified by the manufacuturer. It should correspond to the weld size they used in their proof testing of the fittings.

There is no B16 standard for branch connection fittings, saying they meet the intent of some B16 Standard does not make them a listed fitting, nor does it make them comply with the Code. The is a listed MSS standard, MSS SP-97 for integrally reinforced carbon steel branch outlet fittings. Branch connection fittings need to comply with either the MSS standard, or need to be qualified as unlisted components for use in B31.3 piping. A typical means of qualifying the component is proof testing.

DSB123-

I am well read on SIFs and your information is incorrect. Read, for example, "Effects of Weld Metal Profile on the Fatigue Life of Inegrally Reinforced Weld-on Fittings," Woods and Rodabaugh, First International Symposium on Process Inustry Piping, Dec 14-17 1993, MTI. There is no correlation between weld size and SIF. In fact, some of the larger welds that were fatigue tested had higher SIF's. The profile is critical, but is not in fact driven by weld size; the critical factor is how the weld blends into the pipe wall. Of course, the minimum weld size required by the manufacturer must be used, that is a minimum requirement.

bvi,
1993 paper seems to be well up to date!!! If you were to read my original post and the subsequent one I am basically saying that if the well is not large enough on the weldolet then the PROFILE will cause a notch. Is that plain enough??

The weld should be sufficient to meet the requirements of B16.25 par 127.3 and 127.4.2, ie, a thickness transition not exceeding a 30 degree slope in the vicinity of the weld, over a lenght at least equal to 1.5 times the min wall thickness. As Meteng said, there is a "notch "effect casued by the thickness transition if the 2 materials are of siialr properties, and there will be a further metallurgical notch effect if the materials are dissimilar.

DSB123

If you are interested, I suggest that you get the paper I referenced. From looking at the weld profiles shown in the paper and the resulting SIFs, you will see that the SIF really does not depend on the weld size. The notch is not the gross notch that you may think. It is how the weld toe of the fillet transitions into the run pipe. Too large a weld in fact can create a higher SIF because the weld has to come back to the run pipe (out-plane side) creating a notch. This is the only paper I am aware of that shows cross sections of weld profiles for Branch Connection Fittings and the resulting SIFs. The fittings were actually cut in half after testing, and the weld profiles are clearly shown in the figures in the paper. Excellent job by Glynn Woods. If anyone is aware of any other publication that shows this type of information, I would certainly appreciate the reference.

Thank you bvi and davefitz. Star for both of you for much needed info.

Hmm. I just bought a large number of olet fittings from WFI international, and I requested that they provide me with the minimum weld size that was required to meet the requirements of code. I believe there have been several interpretations in the past that indicate that the weld design needs to be considered in the qualification of these fittings for there respective pressure/temperature ratings. Our intermediate supply house had a fit when I told them I wanted that information, but I told them to go to WFI and request it, and they had no problem doing it for us. (I did pay a little bit for the information.) I believe I also got a statement saying that SIF's were less than or equal to "code recommended" SIF's.

I can't pass the info around, because I beleive there is some proprietary information involved, but if you ask a reputable vendor, they should be able to provide all this information.