Forces on Pipe Bends and Nozzles
A pipe bend must be held in place against the water it turns.
Before this Lesson
Dig up a buried water main where it turns a corner, and you'll usually find a mass of concrete poured between the outside of the bend and undisturbed soil. It's called a thrust block. Leave it out, and the bend can shove its way through the soil until a joint pulls apart.
How hard does the bend shove? A 12 in. main at 100 psi, turning 90° while it carries 3 ft³/s, pushes on its bend with about 16,000 lb. Yet the water inside creeps along at 3.8 ft/s, slower than you walk. A jet that slow, carrying the same flow square onto a wall, would push it with about 22 lb. Where does the rest come from?
Two pushes on the water
Treat the bend the way Forces from Jets on Plates and Vanes treated a vane. Draw a control volume around the water inside it, cutting square across the pipe at section 1, where the water enters, and at section 2, where it leaves, and list every outside force on that water.
For a free jet, the list held only the vane, since open air at zero gauge pressure surrounds the jet. Inside a full pipe, each cut has more water pressing on the other side, with a force equal to its pressure times the pipe's area. Pressure only pushes, never pulls, so both forces point into the control volume: downstream at the inlet, and back against the outgoing flow at the outlet. Add the bend's own push on the water, . The water's weight acts straight down, square to a horizontal bend, so it drops out.
Write once per direction, with x along the incoming flow and y along the outgoing flow. Along x, shoves the water forward, yet its x-momentum drops to zero, so the bend has to cancel the shove and stop the water too. Along y, shoves the water back, yet it gains y-momentum, so the bend has to beat the shove and speed the water into the turn:
So the water pushes on a bend in two ways, by its pressure and by having its momentum changed, and here the two add in each direction. The bend has to answer both, and the anchor holds the bend while it does.
In the Hook's main, each pressure force is lb, while lb. Each component is 11,332 lb, and the resultant is lb. Nearly all of it is pressure, so even with a valve shut downstream and the water stopped, the bend would need almost the same block. Mind the units: psi times in² gives pounds, but psi times ft² comes out 144 times too small.
The Handbook's is the same: the "force exerted by the bend on the fluid," with the fluid's force on the bend "equal in magnitude and opposite in sign." So find and reverse it. The water pushes this bend along the incoming flow and back against the outgoing flow, outward from the turn, which is why the thrust block goes on the outside. The block, holding the bend still, pushes back on it with itself, so the anchoring force has the same size as the water's push.
Use gauge pressures. The atmosphere presses all the way around the outside of the pipe too, and a pressure that's the same on every side pushes nowhere. The Handbook doesn't say, calling just the "internal pressure in the pipeline."
The vane's steps carry over, with one addition:
- Draw the control volume, cutting square across the pipe at each end.
- Point x along the incoming flow, with y toward the turn.
- Add a gauge pressure force at each cut, pointing in. A cut through a free jet gets none.
- Find the bend's push on the water from in each direction, then reverse it for the force on the pipe.
Bends at any angle
Most bends turn less than 90°. Let be the angle the flow turns through, measured from its incoming direction, as for a vane. The inlet terms still lie along x. Both outlet terms lie along the outgoing pipe, pointing back in and pointing onward, so both resolve with along x and along y:
Solve each for the bend's push, reverse it, and combine the components into the resultant, . A bend that also changes diameter works the same way, with each section's own and . The Handbook prints this pair under Pipe Bends, Enlargements, and Contractions:
Its sketch draws pointing upstream, as in its Fixed Blade case, hence the minus sign. Its bend turns upward in a vertical plane, so the water's weight enters the y equation; for a horizontal bend, drop it. Its , measured from its horizontal inlet, is the same turning angle.
Reducers: the pressures push against each other
Straighten the bend out to and you have a reducer. Both pressure forces now lie along the pipe, pushing into the control volume from opposite ends, and the x equation is all there is:
Watch the two pressure forces fight. The upstream pressure acts on the bigger area and wins easily, while the water's speed-up asks for only a modest net push downstream. So comes out negative: the reducer holds the water back, and the water pushes the reducer downstream just as hard. A sudden contraction gets the same control volume and the same equation.
Problems usually give only the upstream pressure. Find the downstream one first, by Bernoulli, or by the energy equation when a loss is stated, and then write the momentum equation.
From The Energy Equation, for a level line, with the velocities from continuity, :
With no loss stated, and it's Bernoulli: where the water speeds up, its pressure drops to pay for it.
Nozzles: the outlet is open air
A nozzle is a reducer whose outlet is a free jet, and that changes one thing. At the tip the jet is surrounded by air, so its gauge pressure is zero, and the outlet's pressure force disappears, as in the right panel above:
A problem often gives the flow and not the inlet pressure, and it doesn't need to. Write Bernoulli from the inlet to the jet, with and no change in height:
So a nozzle takes three equations in a row: continuity for the two velocities, Bernoulli for the inlet pressure, and momentum for the force. Given the pressure instead, run the same Bernoulli backward, with , for the velocities. The answer always points the same way. The inlet pressure pushes the water downstream harder than the speed-up needs, so the nozzle holds the water back, and the water pushes the nozzle downstream, as if to tear it off the hose. The coupling holds it on, in tension.
Checking the setup
The arithmetic is only as good as the arrows. Before you calculate, check where each pressure force acts, which way it points, and whose force you're about to report.