How to find the thickness of pipe after bending 6

[GOBLINTECHNIC]

Dear sir,

I want to know how to find the thickness of pipe after bending

Thank & Regards

 

[Waramanga]

Whats wrong with using a thickness approach. Straight pipe = 1, if radius is longer then the same material must still be there but thinner no?

 

[SNORGY]

An approximation I recently tried successfully for a boiler tube bend repair simply applied conservation of mass and density before and after bending relative to the outermost band of fibers. I rationalized that the tube diameter(s) would be more or less equal before and after bending, so that the radial (circumferential) strain would be negligible. The change in thickness was then approximated as a function of the change in average fiber length between the bend radius to the outside ID and the bend radius to the outside OD of the pipe.

The Regulator accepted the calculation on the premise that the assumption of constant density - while probably not entirely accurate - produced a more conservative result relative to the wall thickness reduction.

Regards,

SNORGY.

 

[BigInch]

I'm sure it did pretty well.

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[SNORGY]

Yes, but sadly...

I didn't get a "star" for it.



Regards,

SNORGY.

 

[FeX32]

Post the calcs. Maybe you will get a star. Or not, since I don't agree with the const. density assumption.
jk



Fe

 

[SNORGY]

Not sure why. Even my wife agrees with the "constant density" assumption. At least, that's what I conclude when she's constantly telling me:

"You're always just so dense."

Regards,

SNORGY.

 

[FeX32]

I was kidding. Although it wasn't me that gave the star. At first I thought that dislocation motion (caused by plastic deformation) would change the density enough to matter. However, I don't believe it does matter.

I get the dense thing too. But, its more of a " your soo god dam hard headed"



Fe

 

[GOBLINTECHNIC]

Thank for all comment

 

[AlyB]

If you want to do a bit more math and get what some of my profs* believed to be a more accurate answer, try checking in a mechanics of materials textbook. A good one is Intermediate mechanics of Materials by Vable. There are definitely ways to approach the problem that will get you closer than Hooke's law, though of course you'll never get a real world situation that looks just like the formulas.


*I am a recent grad with an engineering job, please don't kick me off the site for mentioning profs.

 

[BigInch]

* OK, but you're getting close

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[wimple]

why not perform FEA ?
That should provide some good guess or rough figures

Wimple

 

[BigInch]

Is your FEA's mechanics valid in the plastic range?

**********************
"Pumping accounts for 20% of the world’s energy used by electric motors and 25-50% of the total electrical energy usage in certain industrial facilities."-DOE statistic (Note: Make that 99% for pipeline companies)

http://virtualpipeline.spaces.live.com/

 

[wimple]

Material models shall be entered as inelastic (plastic in this case) and Non Linear solver is to be activated in FEA software.

 

[racookpe1978]

Constant density? Sure. The metal is the same.
Constant weight (total weight, that is.)? Certainly!
Constant outside radius? Probably not - The outside is deforming inward as the metal deforms plastically.
Constant inside radius? Close to constant, but small waves and compression zones/heights are likely.
Constant thickness? Well, we don't know.
Constant strain? Nah.
Constant stress? Nah.
Consistent position of the neutral axis at pipe CL? Probably not.

So, though my wife agrees with my constant density of mind, I'll say "Bend it, drill it, mic it." Can't see how FEA could be assumed accurate enough for all of the approximations. (Whether it is worth paying for the FEA modelling time and computer time is another question all again....)

 

[FeX32]

"Bend it, drill it, mic it."
Forget about dislocations. Material science brothers...

My ass is so 'strain hardened' my gf can't break it.



Fe

 

[FeX32]

For those who have a punt for FEA.
I read (a long while ago) a document that dealt with some implicit and explicit combination's for plastic behavior. See below.


Fe

 

[wimple]

racookpe1978

I guess that your point, is to clearly avoid use of calculation/simulations to guess the thickness? you are considering that measurement (e.g. NDT) is almost the only proper way. Do you confirm?

Frankly speaking, I am not specialist of such problems. I have suggested FEM, as it is a tool that engineers have in hand, so why not consider it?

In my optinion, we know the initial geometry before bending (probably straight pipe) and the final one (loaded). So Starting from this, one should try to simulate with FEM the behavior on an iteratively manner. That can be a possibility.

I guess the problem stands in the non linear nature of equations and especially the material deformation occuring in the plastic range.

Plasticity being "non nonservative", the final state is dependant from the path of the load vs. time. Thats add a complexity. I understand however that FEM codes are now quite confortable for solving such non-linear and non-conservative problems.

My point is that in any case, we need to know what is level of accuracy required, what is the type of application involved (e.g. lethal service?) and what are the applicable Codes & Standard? Afterward one could evaluate wether FEM simulation could produce some orientation to Engineer or is clearly jumping in the dark...

Wimple

 

[racookpe1978]

Well, there is seldom any problem with guessing (more accurately, making well-founded assumptions based on experience and based on reasonable extrapolations from previously tested models) ... But.

A FEA model is best used within its rannge of accuracy: nothing can come closer to showing the high stress areas when a good, close-fitting model is subjected to outside strains. Then what?

Well, the usual FEA model exaggerates the movement (distortion) from an original shape to a distorted shape based on outside forces. The exaggerated movement (strain, thus stress) is usually shown in colors - but those forces are applied up to the elastic limit, within the 2% limit for permanent plastic deformation in all the results I've seen. (Which is NOT a large sample of more than a few 100's, mind you!) Pipe bending will be at least to a 90 degree permanent angle, permanently stretching the steel wall well past the small movement where the "rules" of the FEA equations from cell to cell apply. For example, what happens to the FEA "cubes" when the material moves through the first FEA box all the way through the next 24 into a 25th "cube"? Is the original equation of strain in one cell that affects adjacent cells still valid when there is no material left "in" the first, second, or third cell?

Will FEA be accurate over that large a movement? Before we assume it will, we should (I would, at least) actually test the validity of the model over the range in question.

But, returning to the original focus of the question: Why is this thickness being questioned? What are the budget/time/resources available to his project team and WHY are these resources beng spent? How many times will this problem be faced, and in how many different configurations (what combinations of pipe wall, pipe material, pipe diameter, bend radius, and bend angle) will this problem be faced?) If there is only one bend, test that configuration properly and be sure. An FEA model will only tell you how accurate the program "thinks" its answer is after x number of iterations with a cube size of y and material properies of wxyz, but that answer has nothing to do with realty.

The tone of the original question doesn't appear to warrant assuming that questioner is facing a situation where determining the presence (or absence!) of a 1/100 of a millimeter (a few 1/1000 of an inch) in a pipe wall bent into a 90 (?? - how far is he concerned?) will affect the operable pressure in his system.

 

[saplanti]

I have found an article about the subject on the following link:

http://www.ndt.net/article/wcndt2004/pdf/petrochemical_industry/130_iimura2.pdf

I believe you will find it useful for the discussion.

Kind regards,

Ibrahim Demir

 

[FeX32]

Good one Ibrahim


Fe