Is there friciton between steam and steel pipe?

[muruep00]

I calculate the ducts for air cooled condenser conducts which come out of turbines and end up in condensers, in EfW and CCGT plants.

I have always assumed a static fem model with pressure, temperature... etc, but in reality, the system is dynamical because the steam is constantly flowing through the ducts at a certain speed.

I assume that some friction must take place between the interior duct steel walls and the flowing of steam, which would create stresses on duct wall in the direction of flow.

I want to address how strong these stresses are in an approximated way to find if they are negligible, and for that, I need to know the speed of the steam, humidity, temperature, grade of turbulent/laminar flow, and friction coefficient between steel and steam.

I can ask the supplier for all this data except for the friction coefficient, and I am struggling to find it on the net.

Any ideas of how to get this friction coefficient?

 

[LittleInch]

I'm not sure that there is any appreciable force that isn't orders of magnitude less than pressure forces and thermal forces.

For turbulent flow, the diagrams and theories mainly point to essentially zero velocity right at the fluid / solid boundary with all the flowing friction taking place in the boundary layer and fluid turbulence.

On a practical sense take the surface of your car. You can sweep off dust / snow etc very easily with your hand, but drive at 70 mph and it's still there.

Ditto fine water drops.

Tells me there is negligible force present right at the actual fluid to solid surface.

Happy to be proved wrong, but I don't think a friction factor exists for fluid v solid interfaces, or at least not in the same context as solid to solid friction factors.

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

 

[muruep00]

I'm not sure that there is any appreciable force that isn't orders of magnitude less than pressure forces and thermal forces.

For turbulent flow, the diagrams and theories mainly point to essentially zero velocity right at the fluid / solid boundary with all the flowing friction taking place in the boundary layer and fluid turbulence.

On a practical sense take the surface of your car. You can sweep off dust / snow etc very easily with your hand, but drive at 70 mph and it's still there.

Ditto fine water drops.

Tells me there is negligible force present right at the actual fluid to solid surface.

Happy to be proved wrong, but I don't think a friction factor exists for fluid v solid interfaces, or at least not in the same context as solid to solid friction factors.

I was also thinking that maybe the flow is concentrated at the central axis of the tube, and near the walls, the speed is small/zero.

Be aware that there is also around 1.5bara of pressure exerted on the ducts from inside.

 

[hacksaw]

you can always measure you pressure drop

 

[muruep00]

you can always measure you pressure drop

I introduce pressure loads in a 3D fem model, which I believe takes into account pressure drops due to section reduction. But my model is static, Im talking about the difference with a possible dynamic one.

If you are asking to measure the actual duct, beware I design before construction and for international project (thus, I cannot go and take measurements).

 

[LittleInch]

So I think what you're thinking about is is there or what size is an axial force on dead straight duct connected to say a fan outlet flange, assuming no friction on the duct supports?

This is not the momentum force or end cap effect at a bend or a tee, which I think is likely to be much higher than any friction force effect on the duct, if it actually exists.

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

 

[muruep00]

So I think what you're thinking about is is there or what size is an axial force on dead straight duct connected to say a fan outlet flange, assuming no friction on the duct supports?

This is not the momentum force or end cap effect at a bend or a tee, which I think is likely to be much higher than any friction force effect on the duct, if it actually exists.

You are right but at the end it's a sum of forces and my duct support are sliding in axis direction, so all these loads are absorbed by the turbine exhaust to which the duct is connected, and this is very sensitive

 

[LittleInch]

Any residual "drag" on the duct is going to be negligible compared to any end cap or thermal loading.

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

 

[muruep00]

Any residual "drag" on the duct is going to be negligible compared to any end cap or thermal loading.

Thus, in term of forces and stresses, it doesn't matter if the steam is flowing or not. Mmmm...

 

[hacksaw]

The only reference on hand: Marks Handbook that refers to ASRAE.

 

[EdStainless]

You have low velocity and low mass flow (compared to the mass of the system) so any such forces would be negligible.
Thermal expansion and vacuum forces will be much greater concerns.
And air-cooled condensers have enough other issues (vacuum leaks, low heat transfer, steam distribution, keeping external fins clean, ...) that I wouldn't worry about this one.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Snickster]

Say you have a section of pipe (or duct) between two elbows. There is a static tensile force in the pipe due to the internal pressure acting on the upstream and downstream elbows which produces a tensile stress in the crossectional area of the pipe. On the upstream elbow there is a pressure force of P1 x A pushing on the elbow in the upstream direction. On the downstream elbow you have a internal pressure force of P2 x A pushing on the downstream elbow where P2 is P1 minus the frictional pressure loss in the piping between. This frictional loss produces a shear force between the fluid and the pipe wall of (P1-P2) x A and shear stress between the fluid and the pipe wall equal to (P1-P2) x A divided by the surface area of the wall.

Therefore the pressure force on the upstream elbow is balanced by pressure force on the downsream elbow plus the shear force of the fluid on the pipe wall due to friction and static equillibrium exists. So if you just model the system with the longitudinal stress in the pipe wall due to internal pressure value at the upstream elbow you will get a correct model of the stress.

 

[1503-44]

I agree. The end cap forces in a pipeline with flow would not be equal, as it is P1 x A on one end and P2 x A on the other end. The difference must be the fluid's frictional shear x internal circumference integrated along the length of pipe. It is not necessary for the infitesimal layer of fluid immediately adjacent to the pipe to move relative to the pipe in order for it to transfer shear stress. Shear stress in any medium is between molecules. The movement of the fluid molecules immediately adjacent to the static fluid film on the wall will transfer their frictional shear stress to the molecules of the film. Since the film is restrained by its contact with the roughness of the pipe, the shear is then transferred to molecules of the pipe.

The longitudinal stress of the pipe is the pressure hoop stress x poisson ratio. Thus longitudinal stress in a flowing pipe is different at upstream and downstream cross sections. That difference / poisson ratio is the pressure drop, which balances the frictional shear, so it happens if you have actual end caps, or even no end caps at all.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[IFRs]

Are there any turns in the duct?