Stress analysis - Strange thing 1

[Resegone]

Good evening or Good mornign everyone.

I kindly ask your support for a strange case in stress analysis.

Let me describe the problem

I have a condenser with very very low allowable nozzle load.

The external engineer that provide us the stress anlysis in Caesar provide us a strange results.

Due to very very low allowable nozzle load, i kindly ask to them to put an Anchor (Fixed point) next to the nozzle in order to avoid problem.

But the results is different from what i thought, instead to have low value i have a very high value, but for me is strange

Please find here below the configuration, i have the flange attached to the nozzle, an elbow installed to the flange and then a piece of pipe and next to the elbow i will install the pipe shoe welded to the structure.

Please consider that Design pressure is 24 bar(g), and design temperature is 80°C so even very low and the pipe is 3'' with Sch.40 in SA 106 Gr.B
on the flange i have a load of 10'000N on Y direction and even higher on other.

But this value for me is strange and quite difficult to understand

 

[RVAmeche]

I assume your nozzle anchor is modeled as perfectly rigid? An anchor next to a nozzle modeled this way will show extremely high forces despite small thermal growth associated with short pipe runs because it has no where to go; the pipe wants to grow 1/16" and its stuck between two rigids and can't, so high forces result.

There are some people who say this problem doesn't exist in real life and is a phantom of the math/imperfect nature of stress analysis.

If you can capture the actual stiffness of the condenser nozzle and/or apply a stiffness (allow deflection) at your pipe anchor, you should see forces go down dramatically. No pipe anchor is perfect so if you assume 1E7 stiffness, it will deflect slightly, and forces will drop.

 

[NovaStark]

Do you have the details on how the external engineer modelled the anchor on the condenser ?

Are the loads on the rest of the piping comparable or do you have case of lift-off occurring and thus causing more loads to be transferred towards the anchor ?

Are the allowable loads on the condenser based on guidance from the OEM or are they from a project standard that are generalized ?

 

[Snickster]

As RVAmeche indicated with the cooler nozzle modeled as a perfectly anchor and the anchor modeled as purely rigid, the pipe between the anchor and the nozzle is completely restrained. In this case a maximum restraint force will be developed. It does not matter how long the pipe is that is restrained between two anchors as you will get the same force even if the pipe was 10 feet long. This is because stress (and therefore force) is proportional to the strain or change in length over original length of pipe not the pipe length.

However the actual Thermal expansion deflection of the pipe between the anchors in your case is only about 0.01 inches based on about 16" inches of restrained length in each direction of carbon steel pipe at about 106 F delta temperature. This small deflection I do not believe will cause a overstress of the cooler connection.

That being said I would try to avoid putting an anchor at that location. For 3" piping at 130 F delta temperature, I believe you can make the connecting piping flexible enough not to need to use an anchor so close to the cooler. You could just put some guides or limit stops in the piping to direct the thermal expansion away from the nozzle, if needed.

Another thing is that if you really need to put an anchor as last resort, the anchor should be removable (bolted) so you can undo the flange connection if required for maintenance. With the detail you have you would have to cut the weld off from the structural support if you ever have to break the connection flange.

 

[Resegone]

@RVAmeche: you have right, this explain everything. The one that made the stress put 0 on the anchor instead of 0.0001. So this explain everything.


@Novastark: The problem is the manufacturer of the condeser that declare an allowable load on the nozzle of 100N, 10kg.
This value is very low and without a fixed point in any case i could not respect their request.

@Snickster: The problem is that i have showed you only one connection of the cooler but for inlet such in my case i have a common Header with two branch that go to the two nozzle.
With fixed point the Tee of the branch goes in overstress.

Due to the very low value of allowable load: 100N the anchor is the only i can imagine to guarantee that value

 

[KevinNZ]

100N load is silly. How are flange studs done up?

 

[RVAmeche]

It is but its not unheard of. Hell, most vendors on my projects just say zero.