
Fluid Mechanics
Pour, push and spin fluids to see how they really move
Ten chapters, 15 interactives, few words. Every chapter is mapped to the Anna University, VTU, SPPU, Mumbai University and KTU units it covers, from fluid properties to turbines, pumps and compressible flow.
- AUCE3391Fluid Mechanics and Machinery
- VTUBME403Fluid Mechanics
- SPPU202049Fluid Mechanics
- MUMEC402Fluid Mechanics
- KTUPCMET303Fluid Mechanics and Machinery
By Skill-Lync · Updated 16 Nov 2026 · 10 chapters · 15 interactives · free, no sign-up
Fluid properties and viscosity
Which syllabus covers thisAU CE3391 Unit IVTU BME403 Module 1SPPU 202049: fluid propertiesMU MEC402 Module 1.1KTU PCMET303 Module 1
What makes a fluid different from a solid?
A solid resists a shear force by deforming a fixed amount and stopping. A fluid keeps deforming for as long as the shear acts, however small it is. That is why we describe fluids by rates (velocity, rate of shear) rather than displacements, and treat them as a continuum: a smooth material whose density and velocity are defined at every point, which works as long as the flow is much larger than the spacing of its molecules.
What are density, specific weight, specific volume and specific gravity?
Density is mass per unit volume (water about 998 kg/m³ at 20 °C). Specific weight is weight per unit volume (water about 9.79 kN/m³). Specific volume is . Specific gravity compares a liquid's density with water's: SAE 30 oil is about 0.89, mercury 13.6.
What is viscosity, and what does Newton's law of viscosity say?
Viscosity is a fluid's resistance to shearing. Slide a plate over a thin film and the fluid touching each surface moves with it (the no-slip condition), so the velocity changes across the gap. Newton's law says the shear stress is proportional to that rate of change, with the dynamic viscosity as the constant (below). Dynamic viscosity is in Pa·s; kinematic viscosity is in m²/s (1 centistoke = m²/s). Liquids get thinner when heated; gases get more viscous.
What are Newtonian and non-Newtonian fluids?
For water, air and oils is a straight line through the origin when plotted against : they are Newtonian. Paints, blood and ketchup are shear-thinning (they flow more easily the faster you shear them); cornflour in water is shear-thickening; toothpaste is a Bingham plastic that needs a yield stress before it flows. A simple model is the power law .
What are surface tension and capillarity?
Molecules at a liquid surface are pulled inwards, so the surface behaves like a stretched skin with tension (N/m; water 0.073). In a narrow tube this lifts water that wets the glass and depresses mercury that does not: . In a 2 mm glass tube water climbs about 15 mm; that is why manometer tubes should be wide.
What are bulk modulus and vapour pressure, and why do they matter?
Bulk modulus measures how hard a fluid is to compress: about 2.2 GPa for water, so liquids are treated as incompressible except in water hammer. Vapour pressure is the pressure at which a liquid boils at a given temperature (water 2.34 kPa at 20 °C). If the local pressure in a pump or propeller falls below it, vapour bubbles form and collapse: cavitation.
Halve the thickness of the oil film under a sliding plate and keep its speed the same. What happens to the drag force?
With a linear profile . Halving doubles the velocity gradient, so the shear stress and the force on the plate both double.
Worked examples
Plate sliding on an oil film · Plate area m², speed = 1.2 m/s, film = 1.5 mm, SAE 30 oil = 0.29 Pa·s
- = 0.29 × 1.2/0.0015 = 232 Pa
- = 232 × 0.4 = 92.8 N
- Power to keep it moving = 92.8 × 1.2 = 111 W
Answer: = 232 Pa, = 92.8 N, about 111 W
Capillary rise and depression · Glass tube = 2 mm; water = 0.0728 N/m, ; mercury = 0.485 N/m, , = 13 600 kg/m³
- Water: = 4 × 0.0728/(998 × 9.81 × 0.002) = 0.0149 m
- Mercury: = −0.0047 m
Answer: Water rises 14.9 mm; mercury is pushed down 4.7 mm
Fluid statics, manometers and buoyancy
Which syllabus covers thisAU CE3391 Unit IVTU BME403 Module 1SPPU 202049: fluid statics, buoyancyMU MEC402 Module 1.2KTU PCMET303 Module 1
How does pressure change with depth, and what is Pascal's law?
In a fluid at rest the pressure at a point is the same in every direction (Pascal's law), and it rises linearly with depth (below). Pressure measured above atmospheric is gauge pressure; below it, vacuum. Absolute pressure = atmospheric + gauge. At 10 m under water the gauge pressure is about 98 kPa, roughly one extra atmosphere.
How do piezometers and manometers measure pressure?
A piezometer is a vertical tube: the liquid rises to . For higher pressures, or gases, a U-tube with a heavy liquid (mercury) is used. Write pressures level by level from one end to the other: going down adds , going up subtracts it. For a differential U-tube between two pipes at the same level carrying the same liquid, . Bourdon gauges are the mechanical alternative.
What is the total pressure and centre of pressure on a submerged plane surface?
The resultant force on a plane surface equals the pressure at its centroid times its area, . It does not act at the centroid: pressure grows with depth, so the resultant acts lower, at the centre of pressure. For a vertical surface ; for a horizontal one the pressure is uniform and the two points coincide.
How do you find the force on a curved surface?
Split it into components. The horizontal component equals the force on the vertical projection of the surface. The vertical component equals the weight of liquid directly above the surface, up to the free surface (real or imaginary). Combine them as ; the resultant passes through the centre of a circular gate.
What is buoyancy, and how does the metacentre decide whether a floating body is stable?
Archimedes: a body in a fluid is pushed up by the weight of fluid it displaces, acting through the centroid of the displaced volume, the centre of buoyancy B. Tilt a floating body and B shifts to the low side. The vertical through the new B cuts the centreline at the metacentre M. If M is above the centre of gravity G (positive metacentric height GM) the couple rights the body; if M is below G it capsizes. The metacentric radius BM is the second moment of the waterplane area divided by the displaced volume.
A ship sails from a river into the sea. What happens to how deep it floats?
Sea water is about 2.5 % denser than fresh water. The ship's weight is unchanged, so it needs about 2.5 % less displaced volume and rises. Ships carry load lines (the Plimsoll mark) for exactly this.
Worked examples
Differential mercury manometer · Two water pipes at the same level, mercury deflection = 150 mm
- = (13 600 − 998) × 9.81 × 0.150
- = 18.5 kPa
- As a head of water: = 1.89 m
Answer: = 18.5 kPa (1.89 m of water)
Inclined gate in a tank · Rectangular gate 2 m wide, 3 m long along the slope, top edge 1.5 m below the water surface, inclined at
- , = 2.799 m
- = 998 × 9.81 × 6 × 2.799 = 164.4 kN
- = 3.000 m
Answer: 164.4 kN acting 3.00 m below the surface (0.20 m below the centroid)
Is the pontoon stable? · Rectangular pontoon 12 m long, 6 m wide, 2 m high, density half that of water, centre of gravity 1.2 m above the keel
- Draft = 0.5 × 2 = 1 m, so = 0.5 m
- = (12 × 6³/12)/(12 × 6 × 1) = 3 m
- Weight 12 × 6 × 1 × 998 × 9.81 = 705 kN
Answer: GM = 2.3 m: stable
Fluid kinematics
Which syllabus covers thisAU CE3391 Unit IVTU BME403 Module 2SPPU 202049: fluid kinematicsMU MEC402 Module 2.1KTU PCMET303 Module 2
What are the types of fluid flow?
Steady flow does not change with time at a point; unsteady flow does. Uniform flow has the same velocity along a streamline at an instant. Laminar flow moves in smooth layers; turbulent flow mixes in eddies. Flows can be one-, two- or three-dimensional, compressible or incompressible, rotational or irrotational. The Lagrangian view follows each particle; the Eulerian view, used almost everywhere, watches fixed points in space.
What is the difference between streamlines, pathlines and streaklines?
A streamline is everywhere tangent to the velocity at one instant, so no flow crosses it. A pathline is the track of one particle over time. A streakline joins every particle that has passed through one point, like a line of dye or smoke from a nozzle. In steady flow all three coincide; in unsteady flow they differ.
Why can a fluid particle accelerate in a steady flow?
Acceleration has two parts: local (the velocity at a point changing with time) and convective (the particle moving to a place where the velocity is different). In a steady nozzle the local part is zero, but particles speed up as the passage narrows, so the convective part is not. In 2D the x-component is written out below.
What is the continuity equation?
Mass is conserved. For steady flow in a pipe what goes in comes out, so for a liquid the discharge is the same at every section: halve the area and the speed doubles. At a point, for incompressible 2D flow, the velocity field must satisfy the differential form below: any stretching in one direction is matched by squeezing in the other.
What are the stream function and the velocity potential?
The stream function is constant along a streamline, and the difference in between two streamlines is the flow rate between them; any automatically satisfies continuity. The velocity potential exists only for irrotational flow. Lines of constant cut streamlines at right angles, and together they form a flow net. Simple flows (uniform stream, source, vortex, doublet) add up: uniform + doublet gives the flow round a cylinder.
What are rotation, vorticity and circulation?
A fluid element can spin as it moves. Vorticity is twice its rate of rotation; flow with zero vorticity is irrotational. Circulation is the line integral of velocity round a closed curve, equal to the total vorticity inside it. A spinning cylinder in a stream has circulation, which produces lift per metre (Kutta-Joukowski): the Magnus effect behind a curving cricket ball.
Water flows steadily through a converging nozzle. Is the acceleration of a water particle inside it zero?
Steady means nothing changes with time at a fixed point (zero local acceleration). Each particle still speeds up as it moves into the narrower section: that is convective acceleration.
Worked examples
Checking a velocity field · 2D field , (m/s, x and y in m), point
- Continuity: = 2 − 2 = 0, so it is a possible incompressible flow (subscripts are partial derivatives)
- = 2 × 2 + (−2) × 0 = 4 m/s²
- = 0 + (−2) × (−2) = 4 m/s², so the acceleration is 5.66 m/s²
- Vorticity = 0 (irrotational); stream function , potential
Answer: Valid, irrotational flow; = 5.66 m/s² at (1, 1)
Rankine half-body · Uniform stream = 2 m/s plus a source of strength = 4 m²/s
- Stagnation point upstream of the source: = 0.318 m
- Far downstream the body's half-width is = 1 m
Answer: Nose 0.318 m ahead of the source; the body grows to 2 m wide
Lift on a spinning cylinder · Air = 1.204 kg/m³, = 10 m/s, circulation = 6 m²/s
- = 1.204 × 10 × 6
- 72.2 N per metre of span
Answer: 72.2 N per metre, at right angles to the stream
Bernoulli, flow measurement and momentum
Which syllabus covers thisAU CE3391 Unit IVTU BME403 Module 3SPPU 202049: fluid dynamicsMU MEC402 Module 3KTU PCMET303 Module 2
How is Bernoulli's equation derived, and when does it apply?
Apply Newton's second law to a small element moving along a streamline, with pressure and gravity forces only: that is Euler's equation, = 0. Integrate it for an incompressible fluid and the sum of pressure head, velocity head and elevation is constant along the streamline. Its conditions: steady, incompressible, frictionless, along one streamline, no pump or turbine in between. Real pipes add a head-loss term.
How do a venturimeter and an orifice meter measure flow?
Both narrow the flow so it speeds up and its pressure drops. Measure that drop and combine Bernoulli with continuity to get the discharge. A venturi's gentle diverging cone recovers most of the pressure (coefficient of discharge about 0.98); an orifice plate is cheaper and shorter but loses far more energy ( about 0.6). With a mercury U-tube under water, a reading is a head of water.
How do a Pitot tube, an orifice and a notch work?
A Pitot tube faces the flow and brings it to rest, so the extra head it reads over the static pressure is the velocity head: . Flow out of an orifice in a tank follows Torricelli's result with the head in place of , reduced by contraction and friction through a discharge coefficient. Notches and weirs measure open-channel flow from the head over the crest: discharge grows as for a rectangular notch (formula below) and as for a V-notch.
What does the momentum equation tell you, for example about a pipe bend?
Newton's second law for a control volume: the net force on the fluid equals the momentum leaving per second minus the momentum entering, , as vectors. For a bend, the forces are the pressure forces on the two ends and the push from the bend wall. Turning 0.1 m³/s of water through 90° in a 200 mm pipe at 150 kPa needs about 7.1 kN of anchoring, which is why bends get thrust blocks.
What force does a jet exert on plates and vanes?
A jet hitting a fixed flat plate head on loses all its forward momentum: . If the plate runs away at speed , only the relative speed counts. A wheel of many vanes catches the whole jet, , and its efficiency peaks at 50 % when . Curved buckets that turn the jet almost back on itself, as in a Pelton wheel, raise the ideal peak close to 100 %.
Water flows through a horizontal venturimeter. Where is the pressure lowest?
The throat is the narrowest section, so continuity makes the velocity highest there and Bernoulli makes the pressure lowest. The diverging cone recovers most of it.
Worked examples
Pipe reducer · Horizontal pipe 300 mm to 150 mm, = 0.1 m³/s of water, = 200 kPa, no losses
- = 0.1/0.07069 = 1.415 m/s, = 0.1/0.01767 = 5.659 m/s
- Bernoulli, same level:
- = 200 000 − 499 × (32.02 − 2.00) = 185 020 Pa
Answer: = 185.0 kPa
Venturimeter with a mercury manometer · Inlet 300 mm, throat 150 mm, water, mercury reading = 200 mm, = 0.98
- = 0.2 × (13 600/998 − 1) = 2.525 m of water
- = 0.07069 m², = 0.01767 m²
- Theoretical discharge from the formula above: 0.1285 m³/s
- Actual discharge = 0.98 × 0.1285 = 0.1259 m³/s
Answer: = 0.126 m³/s (126 L/s)
Jet on Pelton-type buckets · Jet = 100 mm, = 40 m/s, buckets at = 20 m/s, outlet angle , friction factor = 0.9
- Mass flow = 998 × 0.007854 × 40 = 313.5 kg/s
- = 11 722 N
- Power = 234.4 kW; jet KE = 250.8 kW
- = 234.4/250.8 = 93.5 %
Answer: 11.7 kN, 234 kW, hydraulic efficiency 93.5 %
Force on a 90° pipe bend · Horizontal 90° bend, = 200 mm, = 0.1 m³/s of water, = 150 kPa at both ends (losses ignored)
- , = 3.183 m/s
- In the inlet direction: = 4712 + 318 = 5030 N
- In the outlet direction, the same size: = 5030 N
- = 7114 N at
Answer: About 7.1 kN on the bend, pushing outwards at 45°
Laminar viscous flow
Which syllabus covers thisAU CE3391 Unit IIVTU BME403 Module 2SPPU 202049: laminar flowMU MEC402 Module 4KTU PCMET303 Module 3
What did Reynolds' experiment show?
Osborne Reynolds (1883) injected a thin dye streak into water flowing through a glass pipe. At low speed the streak stayed a straight thread (laminar); at higher speed it wavered and then broke up and filled the pipe (turbulent). What decides it is not speed alone but the Reynolds number, the ratio of inertia to viscous forces. In pipes, flow is usually laminar below about 2000 and turbulent above about 4000.
What is the Hagen-Poiseuille equation for laminar pipe flow?
In fully developed laminar flow the pressure gradient balances the viscous shear. The velocity profile is a parabola, maximum on the axis at twice the mean velocity, zero at the wall. The pressure drop is linear in velocity and very sensitive to diameter; in Darcy's form it is a friction factor of 64/Re (below).
How does shear stress vary across a pipe?
The shear stress varies linearly from zero on the axis to a maximum at the wall, , in laminar and turbulent flow alike (it follows from a force balance alone). In laminar flow the wall value is .
What are Couette and plane Poiseuille flow between parallel plates?
Between two plates a distance apart, a pressure gradient alone gives a parabolic profile with the mean velocity two-thirds of the maximum (plane Poiseuille). One plate sliding with no pressure gradient gives a straight-line profile (Couette), the model behind a journal bearing: the power lost is the shear force times the surface speed.
Oil flows in laminar flow through a pipe. Keep the flow rate the same but double the diameter. The pressure drop becomes:
At the same , , and . Doubling divides the pressure drop by 16.
Worked examples
Oil pipeline · SAE 30 oil ( = 891 kg/m³, = 0.29 Pa·s), = 50 mm, = 0.5 m/s, 100 m of pipe
- = 891 × 0.5 × 0.05/0.29 = 76.8: laminar
- = 32 × 0.29 × 0.5 × 100/0.05² = 185.6 kPa
- = 64/76.8 = 0.833; = 1.0 m/s; = 8 × 0.29 × 0.5/0.05 = 23.2 Pa
Answer: 185.6 kPa lost per 100 m
Power lost in a journal bearing · Shaft = 100 mm at 1200 rpm, bearing length 150 mm, radial clearance 0.2 mm, oil = 0.1 Pa·s
- Surface speed = 6.283 m/s
- = 0.1 × 6.283/0.0002 = 3142 Pa
- = 148.0 N; torque = 7.40 N·m
- = 930 W
Answer: About 930 W turned into heat in the oil film
Flow through pipes and losses
Which syllabus covers thisAU CE3391 Unit IIVTU BME403 Module 3SPPU 202049: friction and minor lossesMU MEC402 Module 5KTU PCMET303 Module 3
What is the Darcy-Weisbach equation?
It gives the head lost to wall friction in a pipe of length and diameter . The friction factor is in laminar flow. In turbulent flow it depends on and the relative roughness (roughness height over diameter); commercial steel has a roughness of about 0.045 mm. Some Indian textbooks write the same law with in place of , where : check which friction factor a question means. Chezy's formula is the older open-channel form of the same idea.
How do you read the Moody diagram?
The Moody chart plots against for different . On the left, laminar flow follows for any roughness. Turbulent curves fall with and then flatten: at high the flow is fully rough and depends only on . The curves come from the Colebrook equation; Swamee-Jain gives directly within about 1 %.
What are minor losses?
Fittings, valves, bends, entrances and exits cost head too, written as . Typical values: sharp entrance 0.5, exit into a tank 1.0, sudden expansion , sudden contraction about , open globe valve about 10. In a short pipe with many fittings, the 'minor' losses can be larger than the friction loss.
What are the hydraulic and energy gradient lines?
Plot the total head along a pipe and you get the energy (total energy) line, which only ever falls in the flow direction: steadily with friction, in steps at fittings. The hydraulic gradient line is , one velocity head below it; it shows how high water would rise in a piezometer. Where the HGL drops below the pipe, the pressure is below atmospheric.
How do pipes in series and parallel behave?
In series the same flow passes through every pipe and the head losses add. In parallel every branch has the same head loss and the flows add, so a wider branch takes far more than its share: at equal and length, flow scales as . An equivalent pipe is a single pipe that carries the same flow with the same total loss.
Water flows through a long, rough pipe in fully rough turbulent flow. Double the flow rate. The friction loss becomes:
In the fully rough zone is constant, so . Doubling the flow doubles and quadruples the loss (and the pumping power goes up 8 times).
Worked examples
Friction loss in a water main · Commercial steel pipe, = 150 mm, = 500 m, = 0.045 mm, = 30 L/s of water
- = 1.698 m/s
- = 998 × 1.698 × 0.15/0.001 = 2.54 × 10⁵; = 0.0003
- Colebrook: = 0.0173
- = 0.0173 × (500/0.15) × 1.698²/(2 × 9.81) = 8.45 m
Answer: = 8.45 m of water
Sudden expansion · Pipe enlarges from 100 mm to 200 mm, = 40 L/s
- = 5.093 m/s, = 1.273 m/s
- = (5.093 − 1.273)²/(2 × 9.81) = 0.744 m
Answer: 0.744 m lost at the step
Two pipes in parallel · Two 400 m pipes, = 200 mm and = 300 mm, both = 0.02, total = 0.15 m³/s
- Equal head loss means : = (0.2/0.3)^{2.5} = 0.363
- ,
- = 1.271 m/s, = 0.02 × (400/0.2) × 1.271²/(2 × 9.81) = 3.30 m
- Check with pipe 2: = 1.557 m/s gives the same 3.30 m
Answer: The 300 mm pipe carries 73 % of the flow; head loss 3.30 m
Tank draining through a pipe and valve · Water level 10 m above a free outlet; 50 m of 50 mm commercial steel pipe; sharp entrance ( = 0.5), valve = 5
- Energy: , solved with Colebrook
- = 2.66 m/s, = 1.33 × 10⁵, = 0.0213
- Velocity head m: friction m, valve m, entrance m, jet m
- = 5.22 L/s
Answer: 5.22 L/s; friction takes 77 % of the 10 m
Dimensional analysis and model studies
Which syllabus covers thisAU CE3391 Unit IIIVTU BME403 Module 4SPPU Not in 202049 course objectives (check your syllabus copy)MU MEC402 Module 2.2KTU PCMET303 Module 3
What are fundamental dimensions and dimensional homogeneity?
Mechanics needs three fundamental dimensions: mass M, length L and time T. Everything else is built from them: velocity LT⁻¹, force MLT⁻², pressure ML⁻¹T⁻², dynamic viscosity ML⁻¹T⁻¹. A correct physical equation is dimensionally homogeneous: every term has the same dimensions. That alone catches many algebra slips and is the basis of dimensional analysis.
How do Rayleigh's method and the Buckingham pi theorem work?
Rayleigh's method writes the quantity as a product of powers of its variables and matches the exponents of M, L and T. Buckingham's theorem is the general version: variables with fundamental dimensions form independent dimensionless groups. Drag on a sphere depends on : five variables, three dimensions, two groups, so . One curve replaces endless experiments.
Which dimensionless numbers matter most?
Each compares inertia with another force. Reynolds number (viscosity) for pipes, aircraft and submarines; Froude number (gravity) for ships, spillways and anything with a free surface; Euler number (pressure) for flow through valves; Weber number (surface tension) for droplets and sprays; Mach number (compressibility) for high-speed gas flow.
What are similitude and model laws?
A model predicts its prototype only if it is geometrically similar (same shape), kinematically similar (velocity ratios the same) and dynamically similar (the governing dimensionless number the same). Ship and spillway models keep Froude number equal, so the model runs at of the full-size speed; pipe and submarine models keep Reynolds number equal. Matching both at once with the same fluid is impossible, which is why ship resistance is split into friction and wave parts.
What is a distorted model?
River and harbour models are often given a larger vertical scale than horizontal (say 1:100 vertically, 1:1000 horizontally). An undistorted model at 1:1000 would have water a few millimetres deep, where surface tension and laminar flow spoil the result. The price is that slopes and velocities must be corrected with separate scale ratios.
A 1:25 Froude model of a ship is towed in a test tank. Compared with the ship's speed, the model runs at:
Equal Froude numbers mean is the same, so = 1/5.
Worked examples
Drag on a sphere by Buckingham pi · Drag depends on , , ,
- Variables = 5, dimensions = 3 (M, L, T): = 2 groups
- Repeating variables ; gives = −2
- gives , the inverse of
Answer:
Ship model test · Ship towed at 10 m/s; 1:25 Froude model in water; model resistance 40 N (same fluid)
- = 10/5 = 2 m/s
- Forces scale as : = 40 × 25³ = 625 000 N
Answer: Model runs at 2 m/s; the ship's resistance is about 625 kN (before the friction correction used in practice)
Why pipe models are hard · Prototype pipe 1 m diameter carries water at 2 m/s; a 1:10 model also uses water
- Equal :
- = 2 × 10 = 20 m/s
Answer: The model needs 20 m/s, ten times the real speed: Reynolds models often switch fluid (air, or pressurised tunnels)
Boundary layer, drag and lift
Which syllabus covers thisAU CE3391 Unit IIVTU BME403 Module 4SPPU 202049: boundary layer, external flowMU MEC402 Module 6KTU PCMET303 Module 3
What is a boundary layer?
Prandtl (1904) saw that at high Reynolds number viscosity matters only in a thin layer next to a surface, where the velocity climbs from zero (no slip) to the free-stream value. Outside it the flow behaves as if frictionless. On a flat plate the layer starts laminar and thickens downstream; around = 5 × 10⁵ it turns turbulent, grows faster and rubs harder.
What are displacement and momentum thickness?
The 99 % thickness is where the velocity reaches 99 % of the stream. Displacement thickness is how far the wall would have to move out to carry the same flow in a frictionless stream ( roughly in a laminar layer). Momentum thickness measures the momentum lost to friction and gives the drag directly. Their ratio is about 2.6 in a laminar layer and 1.3 in a turbulent one.
How do you find skin-friction drag on a flat plate?
Drag on each wetted side is a coefficient times the dynamic pressure times the area. For a laminar layer the coefficient is the Blasius result below. For a layer that turns turbulent at = 5 × 10⁵ it becomes , valid up to about = 10⁷. The turbulent part costs much more per square metre.
What is boundary-layer separation, and how is it controlled?
Behind the widest point of a body the pressure rises again. The slow fluid near the wall cannot push against it, stops, reverses, and the layer lifts off, leaving a wide low-pressure wake. That wake, not friction, is most of the drag on a bluff body. A turbulent layer has more momentum near the wall and separates later: dimples on a golf ball cut its drag sharply. Control methods include streamlining, suction, blowing, vortex generators and slots.
What are drag and lift, and why do streamlined and bluff bodies differ?
Drag acts along the flow, lift across it, both written with a coefficient times . Drag has two parts: skin friction and pressure (form) drag. A streamlined body (an aerofoil, a fish) keeps the flow attached, so friction dominates and can be under 0.05. A bluff body (a flat plate facing the flow, a truck) separates early: around 1 to 2. A car with = 0.3 needs about 8.5 kW just to push the air aside at 100 km/h. An aerofoil turns the flow downwards and gains lift until, at too high an angle, it stalls.
Why do the dimples on a golf ball reduce its drag?
A turbulent boundary layer carries more momentum close to the surface, so it stays attached further round the back of the ball. The wake narrows and pressure drag drops, more than paying for the extra skin friction.
Worked examples
Laminar layer on a plate in air · Air ( = 1.204 kg/m³, = 1.51 × 10⁻⁵ m²/s) at 5 m/s over a 1 m long plate, per metre width, one side
- = 5 × 1/1.51 × 10⁻⁵ = 3.31 × 10⁵: laminar all the way
- = 8.5 mm
- = 0.00231
- = 0.00231 × 0.5 × 1.204 × 5² × 1 = 0.035 N
Answer: Boundary layer 8.5 mm thick at the end; drag 0.035 N
Plate in water, laminar then turbulent · Water ( = 1.0 × 10⁻⁶ m²/s, = 998 kg/m³) at 2 m/s over a 3 m plate, 1 m wide, one side
- = 2 × 3/1.0 × 10⁻⁶ = 5.99 × 10⁶
- Transition at = 0.25 m
- = 0.00297
- = 0.00297 × 0.5 × 998 × 2² × 3 = 17.8 N; end thickness = 49 mm
Answer: Turbulent beyond 0.25 m; 17.8 N of drag
Air drag on a car · = 0.3, frontal area 2.2 m², 100 km/h, air = 1.204 kg/m³
- = 100/3.6 = 27.8 m/s
- = 0.3 × 0.5 × 1.204 × 27.8² × 2.2 = 307 N
- Power = 8.5 kW
Answer: 307 N, 8.5 kW of engine power just for air drag (it grows with speed cubed)
Compressible flow
Which syllabus covers thisAU not in syllabusVTU BME403 Module 5SPPU not in syllabusMU not in syllabusKTU not in syllabus
What is the speed of sound, and what does the Mach number tell you?
Sound is a small pressure wave; in a perfect gas its speed depends only on temperature (below): about 343 m/s in air at 20 °C and 295 m/s at airliner cruise altitude. The Mach number sorts flows: below about 0.3 density changes are negligible; subsonic below 1, supersonic above it. A supersonic object cannot send pressure signals ahead, so the air meets it as a shock wave.
What are stagnation properties and isentropic flow?
Bring a gas to rest without friction or heat transfer and its temperature and pressure rise to stagnation values and , fixed for the whole isentropic (reversible, adiabatic) flow. Static values follow from the local Mach number.
Why does a supersonic nozzle diverge?
Combine continuity with isentropic flow: . Below Mach 1 a narrowing passage speeds the gas up, as in a garden hose. Above Mach 1 the sign flips: the gas expands so fast that it must have a widening passage to keep accelerating. So a rocket nozzle converges to a throat, where = 1 when it is choked, then diverges. Lowering the back pressure further cannot push more mass through a choked throat.
What happens across normal and oblique shocks?
A normal shock is a jump a fraction of a micrometre thick in which supersonic flow becomes subsonic: pressure, temperature and density rise abruptly and stagnation pressure is lost (entropy rises). At = 2 in air the pressure jumps 4.5 times and 28 % of the stagnation pressure is lost. An oblique shock turns the flow at an angle, as at a wedge or an intake ramp, and is weaker.
What is computational fluid dynamics (CFD), and what are its limits?
CFD solves the governing equations (continuity, Navier-Stokes, energy) numerically on a mesh of small cells, giving velocity, pressure and temperature everywhere. It replaces many wind-tunnel and rig tests in car, aircraft, pump and HVAC design. Its limits: turbulence is usually modelled rather than resolved, results depend on mesh and boundary conditions, and every simulation still needs checking against theory or experiment. The worked examples in this book are the hand checks a CFD engineer uses.
A converging nozzle is already choked (Mach 1 at the exit). You lower the back pressure further. The mass flow rate:
Once the throat reaches Mach 1, pressure changes downstream cannot travel upstream through it. The mass flow is fixed by the stagnation conditions and the throat area.
Worked examples
Normal shock at Mach 2 · Air, = 1.4, = 2
- from the shock relation = 0.577
- = 1 + (2.8/2.4)(4 − 1) = 4.5
- Density ratio = 2.4 × 4/(0.4 × 4 + 2) = 2.667, so = 4.5/2.667 = 1.6875
- Stagnation pressure ratio = 0.721
Answer: Subsonic = 0.577; 27.9 % of the stagnation pressure lost
Supersonic nozzle exit · Converging-diverging nozzle, exit-to-throat area ratio 2, air, isentropic supersonic design
- Solve = 2 on the supersonic branch: = 2.197
- = (1 + 0.2 × 2.197²)^{-3.5} = 0.0939
Answer: = 2.20, exit pressure 9.4 % of the stagnation pressure
Turbines and pumps
Which syllabus covers thisAU CE3391 Units IV and VVTU not in syllabusSPPU not in syllabusMU not in syllabusKTU PCMET303 Module 4
How are hydraulic turbines and pumps classified?
Turbines take energy from water; pumps give energy to it. Impulse turbines (Pelton) convert all the head into a free jet at atmospheric pressure; reaction turbines (Francis, Kaplan) run full of water, with pressure falling through the runner. By flow direction: tangential (Pelton), radial-inward or mixed (Francis), axial (Kaplan, propeller). Pumps are rotodynamic (centrifugal, axial) or positive displacement (reciprocating, gear, vane).
What do velocity triangles and Euler's equation tell you?
At the inlet and outlet of a blade the absolute velocity equals the blade speed plus the velocity relative to the blade, drawn as a triangle. The work done per unit weight is the change in , where is the whirl (tangential) component. A pump with radial entry has no inlet whirl, so its head is ; a turbine designed for radial exit has .
How do Pelton, Francis and Kaplan turbines differ, and how is one chosen?
Pelton wheels suit high heads and small flows: a nozzle makes a jet that splits on double-cup buckets, running best near . Francis turbines cover medium heads; water spirals in through guide vanes, turns in the runner and leaves through a draft tube that recovers pressure. Kaplan turbines suit low heads and large flows, with adjustable propeller blades. The specific speed , the speed of a similar turbine making 1 kW from 1 m head, sorts them: typical textbook ranges are Pelton 8.5-30, Francis 51-255, Kaplan 255-860. Governors adjust the jet needle or guide vanes to hold speed as load changes.
How does a centrifugal pump work, and what are its performance curves?
The spinning impeller flings water outwards, raising its speed and pressure; the volute casing turns that speed into more pressure. Head falls as flow rises, and efficiency peaks at the best efficiency point. The pump runs where its head curve crosses the system curve (static lift plus pipe losses). Changing speed moves the curve by the affinity laws: , , power . A pump must be primed (full of liquid) to start, and the inlet pressure must stay above vapour pressure with margin (NPSH) or it cavitates.
How does a reciprocating pump work, and what are slip and air vessels?
A piston draws liquid in through a suction valve and pushes it out through a delivery valve. The theoretical discharge of a single-acting pump is below. Slip is the shortfall of actual discharge through leaking valves; negative slip happens when a long delivery pipe's momentum keeps the delivery valve open and the actual flow exceeds the theoretical. The indicator diagram (pressure against stroke) shows the work done. Air vessels on the suction and delivery sides smooth the flow, cutting the acceleration head and saving work (about 84.8 % of friction work for a single-acting pump).
A run-of-river site has 15 m of head and a large flow. Which turbine fits best?
Low head and high flow give a high specific speed (the worked example below gets about 413), which is Kaplan territory. A Pelton wheel would need an enormous, slow wheel.
Worked examples
High-head site · = 400 m, = 1.5 m³/s, = 500 rpm, overall efficiency 0.9
- = 998 × 9.81 × 1.5 × 400 × 0.9 = 5287 kW
- = 20.3
Answer: = 20.3: a Pelton wheel
Low-head site · = 15 m, = 50 m³/s, = 150 rpm, overall efficiency 0.9
- = 998 × 9.81 × 50 × 15 × 0.9 = 6609 kW
- = 413
Answer: = 413: a Kaplan turbine
Centrifugal pump head from the velocity triangle · Impeller = 300 mm, outlet width 20 mm, 1450 rpm, outlet blade angle , = 0.05 m³/s, radial entry, manometric efficiency 0.75
- Blade speed = 22.78 m/s
- Flow velocity = 2.653 m/s
- Whirl = 22.78 − 4.59 = 18.18 m/s
- Euler head = 42.2 m; manometric head = 0.75 × 42.2 = 31.7 m
Answer: Euler head 42.2 m, manometric head 31.7 m
Reciprocating pump slip · Single acting, bore 150 mm, stroke 300 mm, 60 rpm, actual discharge 4.8 L/s
- = 5.30 L/s
- Slip (5.30 − 4.80)/5.30 = 9.5 %; = 0.905
Answer: Slip 9.5 %
Keep going
- GATE Mechanical: Fluid Mechanics, 26 solved previous-year questions
- All Skill-Lync CFD courses compared
- Reynolds number calculator
- Pump head calculator
- CFD engineer salary in India
Recommended reading
- Hydraulics and Fluid Mechanics, P. N. Modi and S. M. Seth
- A Textbook of Fluid Mechanics and Hydraulic Machines, R. K. Bansal
- Fluid Mechanics, Y. A. Cengel and J. M. Cimbala
- Introduction to Fluid Mechanics, Fox, McDonald and Pritchard
- Fluid Mechanics, F. M. White
- Introduction to Fluid Mechanics and Fluid Machines, S. K. Som, G. Biswas and S. Chakraborty
How this book was made
Chapters follow the official syllabi listed above. Every interactive's formulas and every worked-example number were checked by script. Explanations, animations and examples are original to Skill-Lync. Topics informed by MIT OpenCourseWare 2.25 (Advanced Fluid Mechanics), CC BY-NC-SA. All explanations, animations and examples are original to Skill-Lync; MIT does not endorse this page.