Swing joints. How do they work?

ChrisJ

I'm sorry if this is a dumb question but I'm trying to understand swing joints better.

Does the piping actually turn inside the fittings or is it to align piping in a way that it can flex and the actual threading inside the fitting doesn't move?



I'm having a hard time understanding if the piping actually turns inside the fitting how doesn't this create leaks?

icesailor

Swing Joints:

The alignment of any fitting depends on the other fitting being in the ALMOST exact same plane. Some number that is divided by 90 degrees. With one ell, you can only operate through 90 degrees. With two, one plane will be within 90 degrees, the other plane can be anything. If you add a third, you can swing throughout any plane. With four ells, there is no plane that you can not swing through. You can make a swing joint with two ells when connected to a stationary point. Like an oil tank where two ells will give you any offset but you need a third to get back plumb.

It comes into play when you are rising from the vertical and you have to go to the horizontal and need to travel with pitch at say 1/4" per foot. Unless you use a crooked thread on the start of the horizontal, you will need two ells to go from the horizontal. If you do not use two ells on the end of the pitched horizontal run, it will not be 90 degrees plumb. You need another ell. Unless you have a crooked thread.

If that makes any sense to you.

ChrisJ

Swinging

So if I'm understanding correctly "swing joint" is a standard piping term in regards to when you use an elbow  so the pipe can "swing" to a different angle?



My assumption was it was a term only used in steam piping.

jonny88

hot water risers

we also use a six elbow swing on hot water risers on high risers.My friend an incredible plumber showed me how but due to my computer inadaquacies or spelling I cant draw it out for you.It is beautiful though

nicholas bonham-carter

Piping geometry

A happy, and informative time can be had at the hardware store, trying out different combinations of 90's, and 45's to get an idea of how stiff piping can be made to go between 2 points without using a single straight line.

Ice, why not go the hardware store, and put some joints together, and take a picture of them so we can see what your friend showed you.--NBC

Mark Eatherton

Not just a steam term, but….

Swing joints are critical for the longevity of a piping system with extremely long runs. If left to its own action/reaction, a piping system can tear itself apart at the joints due to uncontrolled expansion AND contraction. I have seen long runs of PEX used for snowmelt system that literally tore itself out of the plastic supports due to contraction, a condition that no one anticipates on the start up of an extremely cold slab.



A properly installed expansion compensation system will have the ability for the pipes to move without causing undue wear on the pipes (slides and or guides) and is anchored such that the growth/contraction is directed towards the expansion compensator, and is virtually silent in operation.



Expansion joints (other than near boiler piping on steam boilers) can be comprised of 4 elbows arranged in a large U shape (there are actually required leg lengths based on anticipated growth) or in the case of copper or plastic, a large (again, calculated) circular loop, or a telescopic device sealed with O rings.



Again, even though it may be installed, if improperly blocked and supported, they may be useless if the expansion and contraction are not being directed towards the compensation device. In the case of contraction, the initial setting would be half way between compressed and relaxed to compensate for contraction. If used for heating only, they would be set in a contracted mode to allow for expansion.



This required anchoring is a detail that is typically overlooked by the design engineer, the inspector and the contractor and the expansion joints can be a waste of time if improperly applied and controlled.



Residentially, long runs of PEX, if improperly applied and controlled, can equate to a very noisy piping distribution system. In my opinion, "Comfort", does NOT include any noises associated with the operation of the system. Control is as simple as blocking and locking the tubing on the far ends, having an expansion joint in the middle of the run, and supporting the tubing with glides that will allow for a smooth movement of tubing without creating ticking and clicking.



This is one major reason I am not a fan of suspended tube, a.k.a. staple up tubing for RFH applications. When the tubing cools to room temperature,and then is heated up due to call for heat, it sounds like a herd of crickets running through the floor joist…



Got questions ?

icesailor

Expansion/Contraction:

Expansion is one thing. It grows and just compresses upon itself.

Contraction is a different animal, If in the act of expansion, it resolves its position in life and takes a new set, and then contracts, the same rules that apply to expansion, apply to contraction. But it can rip itself out of a fitting if it can't stretch anymore.

With copper tube, buried under a floor in screened sand, it will expand and push sand away, to fill in behind. Then, when it contracts, it will rip copper tube out of a copper fitting and leak.

Never underestimate the power of contraction.

icesailor

The Google:

Do The Google and put in the term "piping  swing joint". you will end up knowing more than you ever want to know.

ChrisJ

Google

I tried that, I couldn't really find anything that talked about whether or not threaded pipes actually twist and move inside the threaded fittings.

mroberts5

I know this is an old thread, but figured it was better to add to than start new.

Can the run-out at the top of a riser be parallel to the main below it, or does it have to run off at an angle?

GroundUp

Yes you can absolutely run them parallel like that

JUGHNE

All of your horizontal steam pipes cannot be level. They have to have a slope for condensate drain back.
Looking at your "main" it should have slope in it..therefore it is not level. Then your riser you would want to be plumb up and down. So 1 90 degree elbow will not let you make that connection without bending the riser. There you would add a short nipple and another 90 ell (swing 90's) to correct the plumb of the riser. This would apply to all fittings above that point until you get to the radiator....2 more sets of swing 90's...you want the riser at the radiator to be plumb in all planes.

mroberts5

Makes sense. So yes, I can run parallel, but also need to "relevel" the final outlet to hit my radiator.

Thanks

acwagner

You also don't want the condensate the cross the path of the steam flow. In your sketch, when the condensate flows out of the radiator and down the short vertical section of pipe, when it hits that 90 it will drip and cross the path of the incoming steam that coming along the top of the pipe. If you use two 90's, you can avoid that.

mroberts5

And as my main is counter pitched, doubling back would make the run-out pitched the wrong way.

Who would have though 3 ft of piping could be so complicated

For as thorough as Lost Art is, more detail on piping could be helpful

mroberts5

Would this be better?

acwagner

Yup. The double 90 allows you to pitch the pipe as well.

mroberts5

Last question: I only need the two double 90s, correct? I can do a single 90 or 90 valve at the top?

acwagner

Yes, just the 90 valve at the top. Either make the vertical pipe slightly off vertical so the rad has some pitch or get the pitch from the union connection.

Jamie Hall

To go back to one of the original questions -- if the swing joint fittings are threaded, the pipes twist on the threads. Not much, but they do -- and the result is that the compression/expansion is taken up with no linear or bending stress on the pipes. Or, for that matter, on whatever is trying to hold them in place.

Expansion loops are quite different, since all the piping will be in one plane (look at an air photo of an oil tank farm for instance) and there the expansion or contraction is taken up by bending in the pipes, rather than twisting in the fittings -- hence, as was mentioned, very specific dimensions for the things so that the bending is within the elastic limit of the metal.

To me the place where this gets really fascinating is in railroad practice, where the steel rails may be welded together in lengths up to several miles long. There the expansion and contraction is resisted along the entire length of the rail, by anchors on the ties -- but the forces involved can be very large indeed (several hundred thousands of pounds to a million pounds). If the weather is very hot or very cold, the railroad will slow traffic or even stop it altogether to avoid breaking or kinking the rail.

ChrisJ

@Jamie Hall

What I find odd is when I tried to pull apart very old threaded joints near the old boiler I sure couldn't get them to move. Even with two 48" wrenches.

I also don't see how NPT connections can move back and forth thousands of times and not create leaks?

Jamie Hall

The twist is very small, @ChrisJ , although the forces are huge (a lot more than you or I with our pipe wrenches!). Suppose you have a 100 foot steel main going from room temperature to boiling -- it will expand about a tenth of an inch. And suppose then it is hooked through a swing joint to a foot long bit of pipe fixed at the other end so it can pivot but not move. Such as, it might be, a typical drop header ( ). That works out to half a degree of rotation in the two threaded bits together.

The point is -- it can give. Copper fittings arranged as a swing joint can't give, so unless that foot long length is able to bend, all the twisting force gets taken up in the solder which, no surprise, eventually fatigues and breaks. And leaks...

ChrisJ

> @Jamie Hall said:
> The twist is very small, @ChrisJ , although the forces are huge (a lot more than you or I with our pipe wrenches!). Suppose you have a 100 foot steel main going from room temperature to boiling -- it will expand about a tenth of an inch. And suppose then it is hooked through a swing joint to a foot long bit of pipe fixed at the other end so it can pivot but not move. Such as, it might be, a typical drop header ( ). That works out to half a degree of rotation in the two threaded bits together.
>
> The point is -- it can give. Copper fittings arranged as a swing joint can't give, so unless that foot long length is able to bend, all the twisting force gets taken up in the solder which, no surprise, eventually fatigues and breaks. And leaks...

Do brazed or welded joints fail in those locations?

JUGHNE

Brazed and welded joints have a stronger bonding material than solder, pretty well goes with the temp involved to make the joint, IMO.
If the risers on a boiler are fairly tall with or without swing joints of any material, it seems to me that the deflection force on the boiler joint would be less. An exaggeration would be with 10' risers and a 5' header, that header would expand lengthwise and the deflection angle/force at the boiler opening would be minimal. Perhaps this makes no sense but........the threaded fittings will rotate ever so slightly without leaking because of the rust left there by the Dead Men.

Jamie Hall

@ChrisJ -- welded joints don't usually fail -- if they are done right, that is. But the force has to go somewhere... and it's not always desirable. And @JUGHNE 's comment is spot on.

Solid_Fuel_Man

> @Jamie Hall said:

> The point is -- it can give. Copper fittings arranged as a swing joint can't give, so unless that foot long length is able to bend, all the twisting force gets taken up in the solder which, no surprise, eventually fatigues and breaks. And leaks...

Isn't this the only reason copper isnt good for steam, and why threaded pipe is best?

Intplm.

@ChrisJ The way it was first explained to me was when I was a apprentice and was taught that a swing joint needed to be installed on the fill and vent lines serving a buried underground oil tank.
The fill and vent extended horizontally underground from the buried tank to the side of the house where the homeowner wanted the fill and vent to come out of the ground. The horizontal fill and vent had two elbows installed on the horizontal with a six inch nipple in between the ells.
This piping arrangement, or swing joint, would ever so slightly move up and down with the under ground temperature conditions. Frost heaves etc. This would stop the pipes from permanently becoming bent.
So a threaded joint is arguably considered one of the best swing joints installed. And they serve very many different applications
as you have read above.
And as you can see.....not a "dumb question."

JUGHNE

For our UG NG system we typically come to the house with 1 1/4" coated iron pipe. The riser is made by bending the pipe up into a 90 sweep. We use a pipe bender for that sweep.
If we go straight into the house off the top of the riser there is a risk that the ground settlement will pull the line sweep down.
We have an insulating plastic/pvc bushing in the riser line above ground. (cathodic protected system)
There is a risk of the bushing being pulled out because of the ground settling or frost heaving.
So enter the house about 6" left or right of the riser, use a 90 and nipple to 90 into the house. This gives us a swing joint for movement and protection of the bushing.

If the riser is not part of a sweep bend then UG 90's are allowed.
There must be 2 90 with nipple between. Fittings are to be heavy duty. The theory is that at the alley meter if the riser is hit by a car then the riser stub can swing/flex in both planes somewhat without cracking the single 90 underground.