Elastic Bend Radius Checks in Pipeline Stress Tools (CAESAR II / Simflex)

[svi]

I wanted to open up a discussion around a gap I’ve noticed in conventional pipeline stress analysis tools like **CAESAR II** and **Simflex**, particularly when dealing with **elastic bends (EBs)** in flowline or long pipeline routes.

In practical pipeline design—especially in surface-laid flowlines and cross-country pipelines—**elastic bends are a common and intentional feature**. However, tools like CAESAR II:

Do not explicitly identify or model **elastic bends** unless modeled as a sequence of straight segments.
Do not offer any warning if the **minimum bend radius** has been breached.
Ignore **strain levels** and **residual stress** resulting from cold bends or natural pipe curvature.
Focus entirely on beam-element stress compliance (code stress, expansion, etc.) while **geometric strain control is left to the user**.

Simflex does better in terms of a **chainage/bearing-based alignment interface**, but still lacks automatic radius compliance checks.

The input data is already there:

3 consecutive nodes give us enough info to calculate **local radius of curvature**.
Using this, software could:

Warn if the radius is below a user-defined (e.g., 15D or 20D),
Compute **local bending strain** as t/2R,
Flag segments exceeding an allowable strain threshold (say 0.25%),
Optionally visualize curvature hotspots in the 3D model view.

Minimum radius violations are **real-world concerns** that can cause excessive ovality, local buckling, or strain concentration—especially during installation or operational settlement.

CAESAR II gives detailed stress breakdowns but leaves users blind to **geometric non-compliance**, which is ironic given how easy this would be to implement.

Have others encountered issues with **undetected tight elastic bends** during construction or post-installation inspection?
Wouldn't it be practical for CAESAR II / Simflex to include this as a **standard warning or check**?
Any workarounds or custom tools that others use to check local curvature and strain from model geometry?

I’m curious what the broader community thinks. Is this something worth pushing for with Hexagon or software developers? Or have we just become accustomed to handling this outside the software?

Looking forward to your thoughts.

 

[LittleInch]

Weel to start with I would take issue with your statement "**elastic bends are a common and intentional feature**"

I accept thar especially for smaller bore flowlines, especially surface laid ones, you do see them, but common and intentional is, IMO, not correct for larger buried lines.

An acceptable elastic bend radius depends on many things such as internal pressure, material strength and thermal expansion. You can easily do equivalent stress checks using bending moment formulae which sometimes limits you > 1000D.

15 or 20D is a purpose made cold bend, not elastic.

So I would think that this sort of thing is sorted in simple design and routing and not somehting you need to do in a stress analysis program.

Any sort of "snaky lay" to alleviate axial thermal stress is quite complex and often in practice ends up more Z shaped than S shaped and then breaks the pipe....

 

[svi]

Thanks for your reply. Accept that it may not be common on larger size pipelines. My judgement is based on the numerous 4" surface laid flow lines in the middle east that I am dealing with now. One needs to a stress run invariably for all of the lines. I had this silly doubt. if there is a vertical EB and that it has a residual stress from the EB. When modelling in CII this residual stress is not accounted. If the flow line displaces from the loading and displacements in the stress model. resulting a bend radius less than the beginning. The actual stress at the bend may exceed the plastic limit. The stress calc will not show this as the residual stress was not accounted. CII will only calculate for the additional displacement imposed on the EB and will be lower. Is this a realistic scenario. Please clarify. Yes I understand that you had said that one needs to check this out outside of the stress run. One has to consciously remember to see if there is a displacement and the resulting bend radius is not reduced lesser than what it began with. will it not be easier to check on the geometric model than doing it outside.

 

[LittleInch]

There is a practical issue that the program doesn't know that the pipe started as a straight line and was then forced into an elastic bend. It has to assume that at the point and temperature of installation the only stress is self weight.

Running stress for a 4" flowline, especially if surface laid is a waste of time.

I've never seen a vertical elastic bend on an onshore development.

If you start with straight runs you don't have this problem. Using elastic bending to route lines is pure laziness IMHO. Put in formed bends, either 40D or 25D or 10D cold bends or 5 D induction bends.

If you're surface laying 4" flowlines they will snake naturally to accommodate thermal stress / expansion.