Question on welding Fish-Mouth pipe connections

If both pieces have the same diameter, you can't drill a hole in the base that's the same size as a vertical. OTOH, It CAN be done by heating, punching, and drifting.

I use fishmouth joints to butt weld similar and different sized pipe pretty regular. To my knowledge, none of the joints have ever failed - although some of the gates are showing evidence of some serious whiskey bumps.

Seems the same to me, although if the welds break, the drilled one will hold in shear. Compression should be worse in the drilled one because the lack of material will allow bending in the long pipe.

Personally I think I'd go with fishmouth since it's easier than drilling that bigass hole... but then if I had a Bridgeport, everything would look like a job for an endmill...

Tim

-- "That's for the courts to decide." - Homer Simpson Website @

When I was building fishing boats we fitted 4" to 6" stack pipes all the time. We *always* used welded fishmouth joints.

By the way, you can't drill a 2.25" hole in a 2.25" pipe, you'd wind up just making two fishmouth joints!

Pipe welding is harder than it looks. I used to see these guys who could weld up pipe joints way out of position, often using a mirror while lying in a bilge under some machinery, and the welds came out looking machine made. I always thought these guys should get Olympic gold medals or something, to me this bordered on magical.

Grant

JR North wrote:

For whatever it is worth; steel tube airplane fuselages are built with the truss members fitted to the joint contours. The longerons are nevcer drilled.

Well, drilled "rosette" welds are acceptable practice in longeron repair, but only in addition to angle splices or fish-mouth splices. I think they're still part of the FAA repair manual, which is available as a PDF file somewhere online.

A book that tube welders would find useful, which describes this technique as well as a lot about making structural joints, fish-mouthing, etc., is the EAA's "Aircraft Welding" book. It's actually a collection of old articles from various sources, including the USAF repair manual, some of them quite old. But it's quite a bag of tips and tricks. Anyone who's welding structural joints in tubing would benefit from it.

Ed Huntress

Several people have pointed out that, due to the wall thickness, you can't drill a hole into a pipe large enough to admit the end of a same-size pipe. It has also been pointed out that drilling such a hole is more difficult than forming a fish-mouth. So, for practical purposes the question has been answered.

However, just for the sake of discussion, lets suppost that the branch pipe is slightly smaller than the main pipe, so a hole can be drilled. It would still be weaker than a fish-mouth joint, because the metal across the side opening is gone, so it can't contribute any strength.

The only justification for drilling the hole would be if the T-connection has to carry flow, in which case you would drill a hole AND form a fish mouth.

If you try to drill a clearance hole for a 2 inch pipe in another piece of 2 inch pipe, you'll sever the pipe. You'll wind up with

2 fishmouthed pieces made from the 1 piece of pipe you tried to drill. So now you have 2 fishmouth joints to weld instead of one. It won't be stronger than a single fishmouth weld to an intact pipe, it may be weaker. Gary

Various replies:

Why can't you? Of course you can - you're only drilling 1/2 way through to make a T joint.

Bob

Depends on a few things.

First, I presume you're talking about using the pipe strictly for mechanical purposes and not for carrying a liquid or gas.

Drilling and fishmouthing will give you exactly the same shape on the surface of the pipes. On the crosspiece of the T, you would of course drill only halfway through the pipe. On the leg of the T, you'd drill completely through. Drill both pieces, and it's just like fishmouthing both and would give you a joint that would allow flow.

As I said, the shapes would be identical on the surface. But the pipe walls have thickness. Drill just one piece - either one - and the full wall thickness of one pipe will be in full contact with the outside of the other. Drill both pieces, and they will be in contact only at the surface - the wall thicknesses will look FROM THE INSIDE as though you had vee'd them at 45 degrees each for weld penetration. Sort of like vee'ing out two butting plates full thickness, but then welding them from the back side. If you can weld them thay way and get full penetration and bridging of the vee, the joint would be fine.

Otherwise, and if it doesn't need to pass a flow, I'd drill just one piece - either one.

John Martin

A only-slightly-better-than-ASCII sketch (577k BMP - sorry, Paint was the only thing that I could find to sketch with):

I don't think there's a question. You start with a sound through-member. Don't screw it up. Fishmouth the T-leg. If the wall is considerable grind the weld prep on the fishmouth. Instantly learn how to weld in all positions.

Hth, Fred Klingener

--

I think that the strength comes from the weld. And the two ways of fitting yield the identical joint, so they have the same amount of weld. I don't think there is *any* difference in strength, nor in amount of material. I'm wondering - after you drilled the 7/8" hole with the hole saw, did the waste lift out? I guess I was thinking pipe with a wall thickness of at least 3/16", not tube - I don't have any experience working with tube.

Grant

JR North wrote:

You COULD drill halfway through the main run pipe with a flat bottom drilling device, but I still think the fishmouth is easier and stronger overall.

RJ

Another question arises. Which method of fitting of the two pipes creates the most curvature of the main run during welding? In case others are un-aware, there is a tool to prevent this sort of curving during welding. It looks like a large wheel puller with hooked 'feet' to hook the main run, and a jackbolt to push on the main run at the outside of the intersection to prestress the pipe before welding.

RJ

By drilling out the main (long) tube, you've destroyed its tensile and bending properties. You don't come even close to getting them back by transferring longitudinal stress in the long member to transverse stress in the walls of the Tee-leg.

Another poster talked about controlling weld distortion. If you leave the long tube intact, its structural integrity means that weld distortion will be much more difficult to control/correct (especially in stainless.) On the other hand, the predrilled joint will be much less sensitive to distortion and easier to straighten out what does occur _because_ of its inherent weakness in that direction.

Don't take my word for it. Take a couple of pieces of tubing each a couple of feet long. Weld a drilled tee joint in the center of one, a fishmouth joint in the other. Line them up afterwards to compare distortion. Set them up spanning a pair of tables. Load them with something at the weld joint. Measure natural frequency. Measure deflection as a function of load. Load them to failure. Report back here.

It really boils down to your structural requirements on the long piece. If it's a primary load-bearing member, don't prep it and deal with the distortion some other way. If it's cosmetic, then it doesn't matter.

After you make a couple hundred fishmouth joints, you get the idea why people like square tubes. They're far easier to jig, weld beads are far easier to run, and it's easy to design machined preps to meet special load conditions without destroying basic section properties.

Hth. Fred Klingener

If you look at the forces resolved through the pipe with a hole in it, you'll see that any tensile or compressive forces transmitted along the side of the pipe with the hole have a break, where the forces now have to resolve up through the sides of the perpendicular pipe and around the edges of the hole. Where the hole is, there is no steel to resist tension or compression transmitted along that pipe.

It's much weaker and it's less stiff. If someone has the time to run it through an FEA program, it could be calculated exactly.

Ed Huntress

That's what I was wondering about- the transfer of stress

*through* the vertical tube wall.

Actually, this was just a thought excersize. I'm not building a Space Shuttle :)

Thanks

JR Dweller in te cellar

-------------------------------------------------------------- Home Page:

If you're not the lead dog, the view never changes Doubt yourself, and the real world will eat you alive The world doesn't revolve around you, it revolves around me No skeletons in the closet; just decomposing corpses

-------------------------------------------------------------- Dependence is Vulnerability:

"Open the Pod Bay Doors please, Hal" "I'm sorry, Dave, I'm afraid I can't do that.."