I am trying to model a 6" API 10000 injection line connected to an oil well. It has some unconventional components.

1) A Target elbow,

http://www.wfi-intl.com/index.php?page=seamless-target-tee

In comparison to a standard B16.9 elbow, this one must be a lot stiffer. Should I model this as a rigid. What SIFs will be applicable to them.

2) The piping is cradled in a rubber inserted Stauff clamp against FIV.

http://www.stauff.com/stauffone/en/#18

The clamp holds the pipe tight with a rubber insert. The clamp has a limit load of 40 KN for the heavy series 10 S that I am considering. Beyond this load the clamp will allow the pipe to slide against the rubber. The friction factor between steel and rubber is 0.6. The clamp can be modelled as a 4 way restraint with stiffness included for the rubber insert. Is there a way I can model this limit load for a restraint.

I want to use the model to do a Fatigue screening needed by Chapter IX of B31.3.

Target/Cushion Tees are typically custom. As such, I don't consider them to be stiffer than normal tees simply for having a blinded end. They must also be significantly thicker.

Calcs:
Option 1: Model them all at the thicker wall, and have CAESAR calculate all SIFs and Flexibilities.

Option 2: Model an example one and pull the SIFs and flexibilities out and apply them to normally modeled tees. (This will be conservative, as the tee will have less material to it.)

Option 3: Model the component in Nozzle Pro and calculate stiffnesses and flexibilities. Apply in CAESAR.

Modelling:
Option 1: Modelling the components at real thickness will result in good numbers.

Option 2: Model as normal pipe with the same OD/WT as the rest of the pipe, but use offsets and calculated SIFs. Modeling them as a rigid components will ignore stresses.

Thank you Michael for your reply,

The target tees are specified at the same thickness as the pipe, but since they are machined off forging stock, they end up thicker than the pipe, the additional stiffness is also attributed to the rectangular cross section of the stock.

I have modelled them as a regular but welded tee. From your reply, Calcs Option 2. How do I model an example. Is it modelling this in Caesar? Does this not happen on it's own in Caesar, the applying of the SIFs and Flexibility of the regular tee.

The wall thickness for the piping is calculated using the Chpt IX equation. Caesar does not have a provision to include this code calculation. I am trying to do this run and it fails in sustained as the thickness using normal service equation will be higher.

Model the tee with its higher thicknesses... for example, nodes 10-20-30 are the run and nodes 20-40 for the branch.

Under "Environment" click "Review SIFs at Intersection Nodes." Input 20.

With regards to your second query, I suggest you look into the X2, Y2, and Z2 restraints.

I see the Y2 restraints are the bi-linear supports which can have an ultimate load. Which appears close enough in concept to what the limit load for the Stauff clamp is. However the bi-linear support model is included in the Caesar UG modeller and not as a regular restraint.

You can use a Y2 restraint in any model, usage of bi-linear restraints is not restricted to the Buried Pipe Modeler.

The target elbows which physically appears as a tee with only the run-in and the branch being connected to the connected piping. The run-out is a solid lump. I modelled them as tees and used the resulting built in SIFs corresponding to the tee. Is this correct?

Can you a point to a link for the modelling of a Y2 restraint, which is not for soil modelling.

It would have been good if this forum allowed a subject wise categorization and search to look up previously answered queries.

I'm not aware of any convenient tutorial to accomplish your specific request, but this is how I would proceed:

1. Your pipe will heat up and your clamp will experience infinite stiffness, but what holds your clamp in place does not. It will deflect some distance X, and the pipe will exert some force Fy (40 kN) at which point the clamp will slide. K1 will be this Fy/X.

2. Fy is Fy.

3. For K2, you're going to need a separate model with +Y and friction value of 0.6. Run and calculate displacements and forces. This force divided by displacement will be K2.