Pipe Stress and nozzle load based on thermal

[NewENG102]

Hi, I am working on calculating what the end load / nozzle load would be and I coming up with a really high number.
Based on a 25' pipe and change in temp of 50F this would calculate to about 1/8" thermal growth.
Below is how I came up with the end load. Would be great if someone can chime in.

Pipe length = 25'
Pipe OD do= 36"
Pipe ID di= 35.25"
Pipe Cross section area A ((pi*do^2-di^2)/2) = 41.97 in^2
Modulus of Elasticity E = 29500000 psi
Change in temp dT = 50 F
Coefficient of thermal expansion e = .00000633 in/in-F
Stress S = E*dT*e = 9336.75 psi
Force on Nozzle / End Load S*A = 391,860 lb

The nozzle load seems very high compared to the 1/8" of thermal growth the pipe will see.

 

[XL83NL]

A routing sketch would help. Can you share one?

 

[TGS4]

Nope - you got the calc right. The forces relating to restrained free thermal expansion are very high.

 

[TGS4]

Where's the other end of the pipe?

 

[NewENG102]

@TGS4 lets assume its fix at other end- does this help?

 

[LittleInch]

Trying to stop thermal expansion of a large steel pipe is not easy. In reality the other end of the pipe is not as fixed as you might think.

It's also why pipe bellows exist.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[TGS4]

Then my initial response stands.

 

[NewENG102]

@TGS4 - assuming the same scenario but removing thermal out of the equation, during installation lets assume the pipe was bolted with 1/8" misalignment in the axial direction which causes stress on the pipe. Now, I dont think the same approach (cals) will work in determining the load at the ends - can you point me in the right direction?

 

[RVAmeche]

Pipe grows/contracts due to temperature. This always happens and you need to route your pipe or otherwise accommodate this effect.

You have 25ft of straight piping going to a pump suction nozzle. If the other end of the pipe is a pipe anchor, you need a bellows expansion joint or a different routing that has more flexibility (changes in direction) to allow the pipe to move. Otherwise, as you've already shown, a small amount of growth (with nowhere to go) results in an extremely large force.

 

[TGS4]

Your misalignment problem is a "comparability" problem. Displace your 25' long beam to back-calculate the force.

 

[LittleInch]

New eng. It is the same calc.

In reality the two end points move a very small amount so the forces reduce.

Same thing with the expansion.

What usually happens is the nozzle moves as it can't resist the force. Now whether that movement is acceptable is up to you.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[NewENG102]

@littleInch - I was also thinking shouldn't the two ends see half the movement / force. Would you say the ends will see half the movement / force? Thanks.

 

[RVAmeche]

That's why it was asked what kind of support is on the "other end" of the 25 ft section, which you said assume fixed (doesn't move).

 

[NewENG102]

@RVAmeche - Got it. I should have said anchor. So in this case both ends will see the half the force or move half the distance? (assume same physical properties for both ends).

 

[JJPellin]

Modeling software would normally be used to analyze this situation. The model can be manipulated to change the pipe routing or the attachment conditions. If you use the software, beware. A rigid support is never perfectly rigid. A guide or shoe is never perfectly free (no friction, no rust). I have seen these programs show acceptable pipe strain on a pump when the reality was strain so great the 3 inch thick pump feet cracked.

Attempting to restrain thermal growth in a straight piece of steel is not easy to do. Look up ribbon rail for railroad track and the phenomenon called sun kink.

Johnny Pellin

 

[racookpe1978]

NO!

Absolutely not the same.

In the first, you have "told us" that both ends of the 25 foot pipe are restrained, then the pipe (tries to) expand (in length) 0.125 inch due to thermal conditions. The entire pipe length MUST then be compressed by the length it cannot expand. You calculated the force above required to "compress" that pipe by 0.125 inch.

Second case. One end fixed, one end free. Same temperature rise. The fixed end stays in one place = no movement. The free end moves 1/8 inch. ZERO additional stress in the pipe due to the UNRESTRAINED thermal growth.

Third case. Both ends free to move, therefore we assume the center is fixed in place. Center does not move, left end moves 1/16 inch, right end moves by 1/16 inch, ZERO additional stress in the pipe.

Fourth case. One end fixed, pipe is pushed axially (sideways) 1/8 inch. If the pipe is supported horizontally, then you only need hand pressure of 15 or so pounds force. Near-zero. If the pump is on a pallet or the floor, friction becomes a concern.

 

[LittleInch]

New eng 102

You keep changing the end conditions of the pipe.

If indeed one end is fixed the it is fixed - i.e. doesn't move. Now in reality this is very difficult to achieve, but the piping stress programs just enter it as an infinitely rigid thing.

The other end is either fixed exactly in position onto your pump, but then has thermal stress due to temperature increase which exerts a large force on the flange.

The other case appears to be the force required to move it 1/8". This will also be very large and essentially the same calculation. The key point is that one end is rigidly fixed in terms of the analysis.

the force won't be halved unless the end point instead of being rigid allow movement of 1/16".

you need to draw this a free body diagram.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[NewENG102]

@Racookpe7978 & LittleInch thank you both for the detail explanation.

I know thermal growth is taking in to consideration when a pipe line is designed - a expansion loop or a bellow joint may be used to absorb the movement and not cause any stress on the pump/nozzle. In my scenario lets assume its a 25' pipe bolt to a pump flange and the other end of pipe has an anchor (usually never the case), no bellow no pipe loop between the pump and anchor- I am trying come up with the force that will be on the pump flange/nozzle based on an 1/8" thermal growth. Based on some of the response the force I came up with (391,860 lb) will not be halved, this force will be on the pump flange / nozzle?

 

[RVAmeche]

Are you looking at 25' of thermal growth = 1/8" or a 1/8" misalignment? It's very confusing when you keep proposing different scenarios.

Regardless, in your configuration with 25' of straight pipe, a pump flange on one end, and an assumed anchor on the other: yes, the amount of force at the pump flange (and anchor on the other end) will be very very large. I cannot confirm if 391 kips is right, but it's not unbelievable. This can lead to damage of the supports, the pump, and the piping system.