Effects of axial stresses on fillet welds 1

[brimmer]

Hello all,

I am seeking some information regarding what efefct axial stresses/loading/tension might have on fillet welds, and if it is a poor idea to use fillet welds where axial stresses from environment exist. The scenario would involve welding a cylindrical sleeve on piping, using fillet welds at the ends of the sleeve. If the pipe is in an area of higher axial load from the environment (buried underground pipeline), what effect would this have on the fillet welds? Would these areas become stress concentrators and develop cracks at the toe? I will ask the same question for pipelines in areas of bending stresses. Thanks for any info or literature you can provide.

 

[dhengr]

Brimmer:
How else would you make that lapped pl. weld without some stress raisers? The fillet weld is at least pretty clean and simple; and a bevel groove weld, or some such, would introduce many of the same problems, and maybe some new ones. I assume you are talking about girth welds btwn. the sleeve and the main pipe, where the main pipe has significant longitudinal tension. Why is the sleeve needed, and how is it loaded, what are its stresses? The trick is to minimize the affects of those stress raisers, and deal with them so you are comfortable with your design. I don’t know off the top of my head exactly what the various codes say about this girth weld condition. It’s not appreciably different than a beveled girth butt weld where you can’t get back in to back gouge and reweld at the root. Imagine the flow of those tensile stresses up into the sleeve (or out of the sleeve), they want to follow the general shape of the fillet weld, and that’s good. But, a couple thoughts: an undercut at the toe of the weld on the main pipe is a stress raiser, across the stress flow path, to be avoided; the optimal shape for that fillet might actually have a longer leg on the main pipe, since this causes a nice gradual transition; ripples btwn. weld passes in that girth weld will all be stress raisers; the outer corner of the sleeve, beyond the vert. toe of the fillet is literally unstressed by this stress flow. You want good fit-up btwn. the sleeve and the main pipe so you get good penetration and fusion at the root of that girth weld, because the flow of the tensile stresses is perpendicular to that root and tends to put it in tension, not a particularly good condition. You don’t want a crappy root pass or any root cracking. But, any other girth weld will likely end up with a similar root condition. Depending upon the predominant stresses in the main pipe, and what the sleeve is intended to do, there might be an argument for a longitudinally split sleeve, with longitudinal welds and no girth weld. Don’t forget that there will be radial and tangential stresses and bending normal stresses in the main pipe too, this might dictate where you make welds or where the max. stresses are.

 

[desertfox]

Hi brimmer

Firstly let me say all fillet welds are generally considered to fail be shearing outwardly from where the two components touch each other to the main weld surface at an angle of 45 degrees.

http://m.weldingdesign.com/archive/weld-orientation-has-consequences

Now if the sleeved pipe is buried would this not put a compressive stress onto the sleeve as opposed to a tensile stress?

Here is another site where there are examples of fillet weld stress calculations

http://www.roymech.co.uk/Useful_Tables/Form/Weld_strength.html#Design_St

L
I don't know whether the sleeve your using is going to be pressurised internally from the main pipe or not but if I assume it is pressured internally then the weld will be subject to hoop stress and longitudinal stree but the former is the critical one normally, if the sleeve is not pressurised then I think the welds would only have very small stresses, unless there is a significantly large axial stress due to temperature change for example, however in this case the hoop stress would be small and the weld would need to be analysed for axial stress only.

In my experience welds on pressurisd pipes and cylinders usually crack from the inner surface to the outer surface, this is because in the case of a pipe or vessel, the tensile hoop stress is greatest on the inside face and the stress drops off to a lower value as it reaches the outside, if the pipe is subject to pressure fluctuations the stress on the inner face is cycling and when a fatigue crack forms it starts in this region and eventually makes it way to the outer surface.

It's hard to give more information as we know little of the conditions your sleeved pipe is being subjected too, perhaps you can provide more information.

 

[brimmer]

Thanks for your responses dhengr and desertfox. I'll explain a litle more. So this is a B-Sleeve (pressure containing sleeve) welded onto a 12" oil pipline. The sleeve will be used to cover a defect in the pipeline, as per code the sleeve must be pressure containing (hence the fillet welds on the circumferential ends of the sleeve). So it will be a split sleeve with a longitidunal weld (on the sleeve only), and then the ends sleeve of the sleeve fillet welded to the pipe. There is no temperature fluctuation, temps are low, the fillet welds themselves will be subject to hoop stress (which is lower pipeline operates at approx 30% SMYS), etc. We have used this repair many times, have well developed weld procedures, and welders to make the repair, so assume very good weld quality with no intitial defects. The caveat in this case is the pipeline itself is located in an area where high axial stresses will be present from the environment, from the soil moving along the pipeline putting the enitre pipe in the area in tension. It is really the outside stresses applied to the pipe I am worried about, not any stresses coming from operation of the line. My concern is that the fillet welds on the pipe will become the weakest point, and develop cracking. Any additional insights much appreciated, I'm not sure this area has been really studied, or even what effect bending stresses may have on fillet welds?

 

[weldstan]

Unless the split sleeve is in bending at its location, the fillet weld will not really experience axial tension loads. Remember that the pipe is still intact and the axial loads that you indicated are experienced by the pipe itself. You have simply added some reinforcement.

 

[dhengr]

Brimmer:
Take a look at this thread, thread378-383519, in the pipeline forum, and see if there is anything in that thread that pertains to your problem.

 

[desertfox]

Hi brimmer

Well if your sleeve is made to a code does it not have allowances for the sleeve to experience tensile stress?
All I can suggest is that you size and stress the weld for whatever tensile stress you believe the sleeve to be subjected too, in line with the calculations on one of the links I posted, although you might find this link useful.

http://www.paulin.com/library/fatigue_info/hinnant_asme_plant_engineering_presentation.pdf

 

[brimmer]

Thanks guys.

@dhengr - yes, so this is the scenario that will play out, only not girth welds, but at the circumferential fillet welds on the sleeve. So would unstable soil on a hill putting axial stresses on the entire pipeline in this area cause an increased risk for these welds to develop crakcs, whether cracks appear in the fillet welds, or in the HAZ at the toe of th weld.

@desertfox, yes, the code has limits, but nothing pertaining to outside forces, but in any event, once the soil starts moving and putting stresses on the pipe, I really have no way to determine how what those might be.

 

[dhengr]

Brimmer:
Well, they are both girth welds, and as relates to the orientation of some of the stresses they see they can have some of the same problems. In some important respects, the girth welds on your pipe sleeve and the girth butt welds btwn. the main pipe sections have significant similarities. And, in these respects they both have the potential of causing problems on a highly stressed section of pipe. Obviously, the API codes and the like, will pay much attention to the girth butt welds because they are so common in a long pipeline. From a stress analysis, stress raiser, crack starter standpoint the roots of these welds are problematic. Ideally you would like to be able to get at the root of the welds, and back gouge it and then do a good job of rewelding it, but many times you just can’t do that. This leaves a less than perfect root area however you prepare it or weld it. Then the normal stress analysis accounts for all of the normal loads, but may not account for unusual conditions, new tensions due to the lay of the land, localized bending due to soil settlement or shifting, corrosion, erosion, inferior metallurgical conditions, etc. As mentioned above, your sleeve may not see much in the way of preexisting stresses, but it will see any newly added stresses, stresses in (and from) the sleeve itself, or some of the main pipe stresses if the line is shut down to make this repair. If, as you suggest, the sleeve might be under pressure, you have a significant change in cross section at the sleeve’s end girth welds, and flexing due to pressure under the sleeve can cause prying tensile stresses across the root of that weld. That weld is trying to act like a fixed end on a beam. The biggest problem that I see is the root if these welds and the way stresses flow past and around it, much like the butt girth weld btwn. the main pipe sections. While they are different welds, it’s the quality of the weld at the roots and the way the stresses flow in this area which makes them much the same as potential problems. Then the longitudinal welds on your split sleeve will intersect with the end girth welds and this causes a nasty triaxial stress and residual stress condition in a localized area.

 

[weldstan]

The sleeve could experience a direct, applied shear load from the soil bearing upon it because of its raised geometry from the pipe itself and the fillet would be the link.