Hello Fellow Stressers,

There is a formula for one component of axial stress in ISO 14692 2017. It is given as (Do/2C)/E.

where:

Do = Pipe OD
C = Curve Radius
E = Elastic Modulus

I have two questions:

1) Shall the formula be written as Do/2C/E since Do/2C will be unitless and E will be in Mpa?

2) Is this a compressive axial stress for a curved pipe? How the formula was derived?

Your response is highly appreciated. Please correct any wrong statement I've made.

Thanks & Warm Regards,

1. Obviously, you are right. Note that for example formula (19) of ISO 14692-3:2017 is correct.

2. The root is unknown for me. I am in doubt the subject is an uniform axial stress following an uniform strain r/R=OD/2C over the section. Maybe is a bending stress with no strain in axis and r/R elastic strain(or OD/2C) in the extremity of cross section. Sometime such formula is used in elastic cold bends calculation. I cannot argue too much being not able to understand the reason behind.

Hi Mariog,

Thanks for your response. I am working in Pipeline and currently have this elastic cold bends.

Actually, I don't have a copy of ISO 14692 2017 (just seen the formula in help file).

As long as the formula is clearly written in the Code, I have no more further question.

Just being curious [it's just me, always wanting to satisfy my curious mind :)]

Cheers!!!

Hi Borzki,
The formula is Do*Ea/2C. It's used as both a tensile and compressive in the sum of the axial stresses.

The formula is derived in the same way as for curved pipes subject to bending moments. This gives the axial strain as the distance from the neutral surface over the radius of curvature. Applying Hooke's law then gives the stress.

Thanks Johannes for that information. Nice to know that it's from curved pipes subject to bending moments.

Now the picture is clear to me by the following:

1) If I have a straight pipe of of original length "Lo". The center line of pipe and outside surface of pipe are all same length "Lo".

2) If I bend it by 90 deg. with a bend radius "C". The original length "Lo = 1.5708 C" at the pipe centerline. However the extrados of the bend will get longer "Lextrados = 1.5708 (C+r), where "r" is the pipe outside radius. And the intrados of the bend will get shorter "Lintrados = 1.5708 (C-r).

3) Therefore the axial tensile strain in the bend extrados would be {(Lextrados-Lo)/Lo} which will lead to "r/C or Do/2C" and the axial compressive strain in the bend intrados would be {(Lintrados-Lo)/Lo} which will lead to "-r/C or -Do/2C".

I think, this would be negligible for 1.5D elbows under ASME B31.3 and only applicable to large radius bends for Pipelines.

Please correct any wrong statement I've made especially on the last sentence.

Cheers!!!

Is this formula also applicable to ASME B31.4 and ASME B31.8?

Thanks & Warm Regards,

Johannes.

What Borzki describes is an elastic bend, applying a bending moment to a straight pipe as to obtain a 90 deg bend.

When you say "curved pipes subject to bending moments" does this assume that you start from a straight pipe? Or better said does ISO 14692 assume such thing? Otherwise, it is hard to understand such approach when you start the calculation with a fabricated curve.

Hi Johannes & Mariog,

I think for curved pipes subject to bending moments, the axial stress will be from M/Z with SIF's and considering flex factors of the bend. This will be carried out in beam analysis software considering the stiffness matrix of a bend (inverse of the flexibility matrix).

Just correct any wrong statement I've made.

Thanks,

Hi all,

I've had a look in ISO 14692-3/2017, Sect. 7.8 ("Allowable Stresses" - not quite proper title, since it presents actual hoop and axial stresses induced by applicable external loads).

The axial stress-component under discussion, "sigma"_ac = OD/(2*C)*Ea, is defined as "roping curvature" stress. Other details are not provided.
Since this stress does not depend of any external applied loading/strain, I guess Borzki opinion regarding cold forming-bending origin of this stress makes sense and is reasonable. Thus, this stress should be considered for FRP/GRP/GRE bends/elbows only.
In addition, please note that all the other axial and hoop pressure-induced stresses, and axial bending-induced stresses (e.g. SIF x M/Z general formula based stresses) are considered separately from "sigma"_ac "roping curvature" stress.

My opinion is that ISO 14692 based stresses for FRP/GRP/GRE pipework cannot be extended to ASME B31.4 and B31.8 metallic pipeline components. FRP/GRP/GRE materials are un-istropic, with different "load-strain-stress" inter-dependence behavior than that of metallic materials.

I hope these thoughts might help...

Best regards,

In other section of ISO is written "Installed curve radius (roping)". In my opinion this stress must be counted only for such cases.

ASME B31.4 & B31.8 do not consider this effect, however there are limitations for radius of cold bends. In addition, a common practice in the industry is to limit outer fibre elongation to 2% - 3% which reduces importance of this stress category.

Thanks Mariog and Dorin for sharing the contents of the ISO 14692 Code. Now the intent of the code is clear to me. And also the limit for cold bending as industry practice for Pipeline.

I think, RTR pipe is more susceptible to this failure since the properties is not isotropic and have many uncertainties. So the Code includes all possible stress which may affect its integrity unlike the high strength (e.g X60) pipes in pipeline industry.

Any other opinion is highly appreciated. Just correct any wrong statement I've made.

Cheers!!