I was having a debate with a group of guys who are insisting that a U-bolt when tightened is an anchor. I am of a different opinion. Please is that the case? Also is it possible to determine the amount of force or stress exerted on a pipe support (i.e a U-bolt) as a result of thermal expansion?

Grateful stress guy

I think a U-bolt ist not a "rotation restraint".
Axial restraint is it only with friction.

Hello Ama_1,

Mr. Ohliger's observation is astute and concise as usual. So, that "conciseness" allows me a "window" in which to offer "a few" additional words. Most of us have the same opinion as Mr. Ohliger, and I will attempt to expand on it.

The question really begins with semantics. How do YOU define "anchor"?

For most of my colleagues, an anchor is a theoretical "fixture" that results in six degrees of fixity (3 translational and 3 rotational) at the location on the pipe chosen by the designer. Very simple that, but as the section inertial values of the pipe increase (i.e., as larger diameter pipes are considered), the harder it will be to actually build a fixture that can provide an "anchor" as defined. The old Kellogg book showed some ideas of "an anchor" and they were well conceived (albeit large). But even an "anchor" of the Kellogg design would allow some moments to "leak through" beyond the "anchor". In the real world, with real world economics being a large influence, it is virtually impossible to build the "anchor" that is defined above.

A "U" bolt is a "structure" that will, as manufactured, provide two mutually perpendicular restraint vectors - on one plane (albeit, the plane is not necessarily one of the three described by your model's coordinate system). If the "U" bolt is placed such that it restrains a pipe that has its longitudinal axis in the horizontal direction, the resulting pair of restraint vectors are one vertical and one horizontal and both mutually perpendicular to the pipe centerline. It is argued by some that if the nuts are very tight, the resulting friction will also provide a restraint vector parallel to the pipe centerline but any amount of power plant experience would convince them that there are uncertainties in this assumption. You can add another (parallel to the pipe centerline) restraint vector by welding lugs to the pipe that would preclude the pipe "sliding through" the "U" bolt. This is similar to the technique used with vertical pipe at clamp locations. What the "U" bolt will NOT provide (in this example) is restraint against rotation about the pipe's vertical axis. Nor will it provide restraint against rotation about the pipe's horizontal mutually perpendicular axes. So, if the "traditional" definition of "anchor" is assumed, the "U" bolt will not fit the definition.

Also, the flexibility of any restraint will always be a consideration (remember, the "anchor definition" above implies there is complete stiffness). If you want an accurate calculation of the thermal expansion loadings on your restraint (e.g., "U" bolt) you will have to accurately model its flexibility (and if it is "tightened" enough to provide some resistance to axial movement you will need to model the resulting friction). Some companies use a "rule of thumb" that advises that any restraint that has a stiffness (inverse of flexibility, in terms of deflection per unit load) of more than 10 times the stiffness of the pipe can be considered "rigid" (in that direction). While this is probably reasonable for small bore piping (however that is defined), the "rule" is subject to discussion as the section modulus of the pipe increases. The "U" bolt can be modeled in an FE analysis to develop approximate stiffnesses for including in our Caesar II model and that would also tell us something about the stresses in the "U" bolt (if we are SURE about the its material of construction). When we are not sure of the "U" bolt's pedigree, the question of the tensile and shear strength of the "U" bolt under operating loadings is yet another issue.

An associated issue for discussion is: "is it really necessary for your single restraint (in and of itself) to resist moments". Most of the time, "supports" and "restraints" can be located such that they will "break-up" the moment into its resultant force couples. It is much easier to design structures (and restraints) to accommodate forces than it is to design them to accommodate moments. So, if you were to locate TWO adjacent "U" bolts along the piping (as described above) the vertical moment would be transformed into a horizontal force couple that loads the adjacent "U" bolts as opposing horizontal forces. The moment that would be about an axis perpendicular to the pipe's centerline would be accommodated by the pair of "U" bolts in a similar fashion. NOW the question becomes: "how far apart must these "U" bolts be placed". "Right next to each other", is certainly NOT the correct answer.

Now consider the item that we call the "U" bolt. How was it manufactured? Was it cold formed around a mandrel after threading? That would imply that the material of construction was already forced to yield and it has some degree of "work hardening". How reliable would this item be if subjected to alternating or cycling loads? Were the threads "cut" into the metal or were they "rolled" into the metal? Each thread root could be an incipient crack (and with cycling loadings....)! And what happens to this item when it is subjected to seasonal weather? Have you considered crevice corrosion and fretting due to oxidation? AND, how was the item procured? Was it purchased from Anvil-Grinnell as "figure 137, with double bolting" (a load rated item meeting MSS Standard SP-58) or did your purchasing department order it from the McMaster-Carr catalog as "hardware" of unknown pedigree?

For every complex question there is a short and easy answer, and (except for Mr. Ohliger's observations) it is usually wrong. I can make no claims about the "rightness" of long winded answers. Ladies and gentlemen, have a pleasant and productive day.


Best regards, John.

Amachree,

I have seen u-bolts used as anchors in many systems (in one absurd case, the engineer used two u-bolts on a 16” line with 7000 lb of axial load); I have seen several situations in which the “anchor” failed; and I have seen various formulae for calculating u-bolt restraint forces, all of which I think are dubious. In addition, I’ve worked with many clients who do not permit u-bolts at all, in part because of their doubtful restraint capabilities and various other reasons (nuts fall off; nuts not replaced after a turnaround; u-bolts intended to be double-nutted with a gap are tightened down hard; etc.). Personally, I prefer not to use them. In practice, however, they should at least be limited to small lines, used only where low forces are expected, and never used as the primary piping support.

Regards.

To add my 2 cents...I am not fond of U-bolts either, but they do have their place, as mentioned.

Howdy John Breen. Very thorough answer, as always...but I think you exceded Ohliger's window...

Its time i must improve my english, that my wondows will be greater . Smile.
Sure i can give many more detail answer give her, but my thought and complet knowledge to translate in english is to expenditure for me. I hope still
welcome her.

Short and sweet... an old saying here in the U.S.A. (I believe its a quote from someone) "brevity is the soul of clarity"...

Your statement said it all as far as I was concerned....

"I think a U-bolt ist not a "rotation restraint".
Axial restraint is it only with friction."

Between you and Mr. Breen the subject ist kaput!... Your english is far better than my deutsch as my relatives in Osterreich constantly tell me!

Mr, Ohliger,

Your contributions to this forum are always welcome and highly regarded. We (the community) thank you for sharing your thoughts with us.

Some people (not to be named :rolleyes: ) simply cannot seem to use an economy of words. As my Großvater, John Luf, says, "short and sweet....".

Regards, John.

The Pittsburgh effect.... my relatives in Butler PA have the same proclivities... must be something in the water.... If you ever do move to Texas I reckon your ways will change...

Hi all,

a very timely case for me where I need to use a U-Bolt as longitudinal restraint for a small bore pipeline, 1". The temparature is high and also high pressure (B31.1 category). Assuming that support point I'll be using a U-Bolt as longitudinal restraint is not subjected to any horizontal and vertical moment but purely longitudinal, what friction value should I use to restraint the horizontal force? I remembered mechanics as F x fric. factor = N where F is the horizontal load (longitudinal) and N is the normal force to F (or could be the tightening load to the U-Bolt?) Looks like I need to set any friction factor as a varying factor to enable Caesar to determine the horizontal load, then only I could use the horizontal load result from Caesar to multiply with the friction value I used to determine the required tightening force on U-Bolt. Please advice. Thank you so much.

Gimini55

1.Tightened gripping u-bolts apply crushing stresses to the pipe wall. not good if high pressure.
2.U-bolts have a tendancy to dig in with large
displacements
3.U-bolts can loosen when subjected to cyclic movements.
4.Mr fitter usually fits the wrong bolt the wrong way.

U-bolts are fine if used for their proper purpose, ie not stress critical piping.
Play safe and if the support is critical, make it a proper shoe. a no risk option!

Good day Gimini55,

The "U" bolt will be in contact with the OD of your hot pipe. When the pipe is new and cold and you tighten the "U" bolt enough (cold) to develop friction how long will it take for the tension in the bolt to creep out at your design temperature? Once the tension has been reduced by creep and the expansion loading axially displaces the pipe, you will have to again tighten the "U" bolt (after repositioning the pipe) to regain the tension (friction). Of course, when the piping system cools to ambient temperature, the permanantly deformed "U" bolt will not return to its original cold (and pipe constraining) configuration. So on the next heat-up it will take a while (perhaps a matter of minutes) for the pipe to "grow" to its hot circumference and in that time the pipe will NOT be axially restrained. So, the pipe will slip through the "U" bolt a little each temperature cycle. Maybe someone will notice that and again tighten the nuts on the "U" bolt. How many years will that scenario be replayed before the "U" bolt "necks down" enough to break?

That is the difference between design and analysis. You can assume all is well at your "U" bolt when you develop your analysis model but in the real world it won't be so for long. Design for the real world and do not count on the friction that is required to "hold" the pipe in position. Find a vertical piece of pipe (upstream or downstream of the horizontal) for the "U" bolt that will "control" the horizontal movement. And as Ricardo says above, ...."never use (it) as the primary piping (weight) support".

Regards, John.

Thanks Superpiper and Mr. Breen. I'll consider your advices very seriously. As always I do appreciate much your time.

Regards,

Gimini55

U-Bolt = Anchor = Bad Idea