Welding A Stainless Steel Manifold

[valmeidan]

I am designing a stainless steel manifold like the one attached. I am however a little confused as how they made this. It appears they used a stainless steel pipe then welded some couplings onto it. In order to do this would you simply bore the holes in the pipe first to match the OD of the coupling then insert the coupling and weld? Or am I missing something?

 

[metengr]

In order to do this would you simply bore the holes in the pipe first to match the OD of the coupling then insert the coupling and weld?

Yes, most likely. These couplings look to be stubbed into the manifold or pipe run and GTAW welded. The external fillet welds were blend ground smooth, contoured and polished or plated (can't tell).

 

[valmeidan]

So you have a coupling sitting inside the pipe? So if you are to look down the pipe you would see the bottoms of all the couplings? What if you want smooth flow through this instead of hitting these couplings?

 

[metengr]

The coupling can be stubbed in and flush with the ID pipe surface.

 

[valmeidan]

Sorry but, what does stubbed in mean.

 

[SJones]

The picture is at a difficult angle to tell exactly how the branches have been made, but apart from weld on, there are at least two other methods that could be used: hot isostatic pressing and extruded branches. The shaping of the crotch area and the lack of a joint further up the branch in the picture does suggest weld on. The weld on approach will require creation of the hole, the other two methods create the hole during manufacture.

Steve Jones
Materials & Corrosion Engineer

http://www.linkedin.com/pub/8/83b/b04

 

[Duwe6]

A better term for 'stubbed in' is 'set in', which is where the hole in the manifold/header is large enough for the branch pipe/coupling to be set into the header. The weld takes place between the ID and the OD of the header pipe.

The other method is 'set on', where the hole in the header matches the ID of the branch pipe/coupling and the weld is all on the OD of the header.

 

[valmeidan]

I would imagine the set in would be easier however if flow is of concern through the header you would get a lot more turbulence running down the header because of hitting the ends of the coupling

 

[paddingtongreen]

Could the taps be welded into place first and then a bore run through them to open the connection?

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[Duwe6]

Boring after welding is not "normal", but it is fully acceptable.

 

[dvd]

If turbulence is really a concern, the manifold should be sized large enough so that the manifold velocity is very low.

I would also consider coping the nipples and beveling coped ends similar to a weldolet connection.

 

[gtaw]

There are several design alternatives to consider, but determining which is best requires some information such as:

What code is any is required or is applicable?
What is the design/operating pressure?
And what is the design/operating temperature?

The US Navy uses MIL-STD-22D as their standard for joint details. Many of the joint details for branch fittings involve welding the fitting to the run pipe and drilling the hole after the welding is completed.

If no code requirements need to be met, you could simply weld half couplings to the run pipe. Even if ASME B31.X is applicable, the diameter of the couplings and the run pipe must be considered. B31.3 for Normal and Category D for instance limit the coupling to 1/4 the run pipe diameter.

Whether the fitting is a CJP or a simple fillet weld is also dependent on code requirements if one applies.


Best regards - Al

 

[MintJulep]

Or use one of the -olet fittings from Bonney.

http://www.bonneyforge.com/products.php?pg=branch

 

[BrianGar]

Is it just the one you have to make?

Id be looking at casting it in stainless(polish it), or brass, and plating it, if it doesn't have to be stainless that is...

Its going to be tough to blend in the welds, and having it all look nice.

Depends on numbers though...

Brian,

 

[MikeHalloran]

The pictured manifold comprises an apparently thick wall run pipe with a female thread counterbored into at least one end, and an array of half-nipples welded in and blended to a high polish finish.

<People, please respect the difference between nipples and couplings, and use the proper root word in discussion.>

The nipples do not appear to be extraordinarily thick, though it's hard to tell from the photo. It's also possible that the female run thread is actually a thin half-coupling pressed rather deep into a thin wall run tube, welded around the coupling's outer circumference, with severely blended welds. The female run thread's recessed position limits the sort of fittings that can properly mate; a typical hex nipple probably could not engage enough to seal properly before fouling the hex points.

Manifolds are much more commonly made with half-couplings instead of half-nipples, because the nipples' external threads are easily damaged in handling, and not so easily repaired.

I would conjecture that appearance and compactness of the next higher assembly is of paramount importance, that this part does not have a formal pressure rating, does not meet any code, accepts only a limited number of styles of mating parts, and may not even have a drawing, certainly not a proper one.


Mike Halloran
Pembroke Pines, FL, USA

 

[BillPSU]

Many hydraulic cylinders use this type of port. The port is machined with a radius matching the barrel OD. The barrel has a hole predrilled for the port. This hole is used to locate the barrel in the welding fixture which is a circular spud welder with a cam following vertical slide. As the welder rotates the cam causes the weld gun to rise and fall matching the barrel OD.

The interior of the barrel will be deformed by the welding which may require resizing after welding the port on. Normally this is performed using a hone. Some manufacturers use pre-honed material and live with the weld distortion on the ID of the barrel.

You didn't say what type of material is flowing through this manifold but unless there is a very substantial flow rate the size of the ID versus the port sizes there shouldn't be much turbulence even with the fittings sticking into the flow stream.

I assume your biggest problem is the cost of the shown manifold. There is probably more cost associated with the finishing then the manufacture of the tube. The polishing will require substantial hand work and then an environmental unfriendly plating process. This part would not be a part I would recommend to in-source unless your company has experience with this type of part.