Fluid Flow & Aerodynamics
why things fly, and why the usual explanation is wrong
Concept

Dynamical system — where it appears

A set of equations giving the rate of change of a state in terms of the state itself. Fluid flows are of this kind with infinitely many degrees of freedom, and studying small truncations of them was how the modern account of transition began.

Named by 12 essays across 5 fields — each of them below, with the objects they name alongside it.

Three equations, and the set they never leave. The Lorenz trajectory at r = 28, projected on x and z, after the transient has been discarded. It never repeats, never leaves, and never crosses itself in three dimensions. The Lyapunov exponent printed beside it is measured on this system by separating a nearby pair, so the claim of sensitive dependence is a computation.

Three numbers left of a fluid

Saltzman truncated convection to three Fourier modes and Lorenz studied what was left. The result changed science, and it stopped being a description of a fluid at about a fifth of the way to the parameter everybody quotes it at.

turbulence · Convection A millionth, doubling every three-quarters of a second. The separation of two trajectories started a millionth apart, on log axes against time. It grows as a straight line until it saturates at the size of the attractor, and the slope of that line is the largest Lyapunov exponent — measured here on the system by renormalising a nearby pair, not quoted. Sensitive dependence is what the straightness of the line means.

A millionth is enough

Two trajectories a millionth apart separate by a factor of e every three-quarters of a second, so a millionfold improvement in the measurement buys about ten seconds of extra prediction. That exchange rate, and not the size of the error, is what limits forecasting.

turbulence · Convection One streamline, sectioned, in two steady flows. Every time a single streamline crosses the plane z ≡ 0 going upwards, a point is plotted. On the left the flow is integrable and the points lie on a curve, however long the trajectory is run. On the right one coefficient of the same exact solution has been changed and the same single streamline scatters over a sixth of the plane. Both flows are steady, both are incompressible to machine precision, and both are exact solutions of the Euler equations.

Steady, three-dimensional, and mixing anyway

A steady flow that solves the Euler equations exactly, with its vorticity equal to its velocity to six parts in ten thousand million — and one of its streamlines wanders through a sixth of the box while another, started nearby, lies on a curve for ever.

kinematics · Advection Four vortices, and the end of prediction. The same three vortices as before with a fourth, weaker one added near the middle. Three point vortices have three independent invariants for three degrees of freedom and cannot be chaotic; four have the same three invariants and one more degree of freedom, and generically are. The energy and the impulses are conserved here to fourteen digits over the whole run, which is what makes the tangle a property of the system rather than of the arithmetic.

Three is the most that can be predicted

Point vortices are the simplest dynamical system fluid mechanics has — no cores, no viscosity, no approximations, four exactly conserved quantities. Three of them are integrable and cannot be chaotic. Add a fourth and the same equations, conserving the same quantities to fourteen digits, stop being predictable at all.

inviscid · Vortex dynamics The mean flux, flat across the inertial shells and equal to the dissipation. The time-averaged transfer out of the first n shells, computed as the rate at which the nonlinearity changes their energy rather than from a remembered formula. It is constant to a tenth across the middle of the ladder and equal to the dissipation, which is the cascade — and it is an average.

A flux that runs both ways

The cascade is a statement about a mean. Kolmogorov's four-fifths law fixes an average and the constant flux through the inertial range is an average, and neither says anything about what the transfer is doing at any instant — which turns out to be running backwards a substantial part of the time.

turbulence · Cascade Surge is a loop round the characteristic's peak. The two runs in the plane of flow coefficient and plenum pressure rise, over the characteristic continued to reversed flow. With B = 0.5 the state slides off the peak and settles on the stalled characteristic. With B = 2 it traces a large loop: the flow collapses at nearly constant pressure, reverses while the plenum empties, recovers to the right-hand branch at low pressure, and climbs back up it while the plenum fills, round and round.

The plenum decides whether a compressor surges

Throttle a compressor past the peak of its characteristic and it does one of two things. It settles into rotating stall, a steady state with a cell of dead flow running round the annulus, or it surges, the whole flow through the machine collapsing, reversing and recovering over and over. Which one is not decided by the blades. It is decided by the volume the compressor discharges into, against the inertia of the air in its duct — one number, Greitzer's B, which grows with the blade speed, so the same machine stalls at part speed and surges at full speed.

applied · Turbomachine Under drag, every released pair comes round and meets. One cylinder's centre relative to the other's, the stream from left to right and folded into one quadrant, for pairs released from rest six radii apart under a drag coefficient of one. Released at 10° and 20° from tandem they first drift apart, as the ideal pair would, but the drag takes their speed; the torque keeps turning them, and once past 45° they close. The farthest any gets is 9.55 radii. The two faint paths are ideal pairs released at 10° and 35°, which part for good.

Drag makes every free pair meet

Two cylinders set free in an ideal stream swing towards side by side like a pendulum, and whether they collide or part is fixed, far apart, by forty-five degrees. Give each a drag on its motion relative to the stream and the pendulum stops swinging — but it does not stop the pair. A drag coefficient of order one turns the swing into a creep, lets no pair get far from where it started, and brings every one of them round to meet near side by side.

inviscid · Dalembert The band of protections that cycle narrows and closes. The derating strengths for which the bearing cycles, against the bush's share of the housing excess. With fixed walls the band runs from 0.162 to 0.3; at β = 0.2 from 0.182 to 0.279; at 0.3 from 0.205 to 0.254. A little past that it closes: no derating of any strength makes the bearing cycle, because the warm wall has stiffened the drive past its cusp and there is no S left to cycle round.

A warm bush stops the bearing hunting

A bearing whose motor is derated on its housing's temperature can hunt, jumping between a cool film and a hot one on the housing's clock. Let the bush warm with the housing, and a second feedback joins the first, of the kind that usually makes things run away. It does the opposite: a warm wall is a stiffer drive, the stiffening closes the S the cycle runs round, and past a bush coupling of about a third the bearing settles, warm and derated, however the protection is set.

viscous · Viscous heating With a small plenum, stall feedback closes the hysteresis loop. The stall amplitude against the throttle as it is closed slowly past the peak and opened again, at B = 0.2, with no stall feedback and with feedback at twice the critical gain; closing solid, opening dashed. Without feedback the cell appears suddenly near γ₀ = 0.6094 and vanishes only at 0.6428, and the loop between the two legs has an area of 0.078. With feedback the cell grows gradually as the throttle closes and shrinks along the same curve as it opens: the area is 0.0014.

A throttle can hold the peak only in a small plenum

A compressor makes its most pressure at the peak of its characteristic, and at the peak it either surges or stalls. A throttle moved by feedback can fight both. Driven by the plenum's pressure it damps surge, with a gain that grows as the square of Greitzer's B until it meets a ceiling B does not move; driven by the flow, it cannot. Driven by the stall cell's amplitude it turns the sudden jump into deep stall into a gradual one and closes the hysteresis loop — but only while the plenum is small. Above B of about a third the same feedback turns the jump into a cycle.

applied · Turbomachine A slow bush leaves the bearing hunting, a sixth faster. The film's mean temperature excess over two cycles of a bearing hunting under a derating of 0.2 with a fifth of the housing's excess reaching the bush, for a bush that follows the housing instantly and one ten housing times slow. Both jump between a cool film and a hot one; the instant bush's period is 0.347 housing times, the slow bush's 0.291. The slow bush holds the walls at the cycle's mean, and the film's jumps are the same size.

A slow bush quickens the hunt it cannot start

A bearing's bush warms with its housing and makes the drive stiffer, and past a coupling of about a third it stops a derated bearing hunting. Give the bush a time of its own and the obvious fear is that the lag reopens the hunt. It cannot: whether a bearing hunts is decided by where its equilibrium sits, which no lag can move, and a lag on a positive link cannot make a settled state oscillate. What a slow bush does change is the hunt itself, which runs a sixth faster.

viscous · Viscous heating A cut deeper than the film's window flips the film twice a cycle; a shallower one leaves it where it is. A protection that trips at a housing excess of 3.0 and resets at 2.6, against the depth of its cut: left, the film's jumps between its branches per protection cycle; right, the cycle's period in housing times. The film has both branches between 0.6496 and 0.7279 of full drive, so a cut past 0.35 leaves it only the cool one. cut 0.3: never resets; cut 0.34: 0 flips, period 1.85; cut 0.36: 2 flips, period 0.767; cut 0.6: 2 flips, period 0.756.

A trip deeper than the film's window flips the bearing twice a cycle

A smooth derating acts on a bearing's housing temperature as it stands. Most real protections trip instead: they cut the motor at one temperature and restore it at a lower one, a hysteresis of their own laid on top of the film's. Whether the oil then jumps between its cool and hot states is decided by one comparison — the cut's depth against the width of the film's window — and past it the protection's own band sets the rhythm. A trip never lowers the film's peak at all; it only protects the housing, and a deep one buys that with a plunge to the cool branch twice in every cycle.

viscous · Viscous heating Kicked at ninety per cent of its peak flow, the compressor is held only when both modes have an actuator. The flow through the compressor, as a fraction of the peak flow, after a stall cell of amplitude 0.05 is started at 90 per cent of the peak flow, at B = 2, under four controllers: a tip injector of gain 0.7 and capacity 2 per cent of the peak flow with the throttle fixed; the injector with a pressure-driven throttle; the pressure-and-flow throttle with no injector; and both. injector, throttle fixed: surges or runs away; injector, pressure throttle: surges or runs away; throttle pair, no injector: stalls; injector and throttle pair: recovers.

A compressor needs one actuator for each way it can fail

A throttle alone cannot hold a compressor left of its peak in a large plenum: it has one degree of freedom and the machine has two ways to fail, and the throttle's own surge gain runs into a ceiling that the plenum's size does not move. Give the stall cell its own actuator — air injected at the rotor tips against the cell — and give the throttle a second signal, the flow, and a machine at B = 2 is held at ninety per cent of its peak flow, recovering from a kicked cell nearly three times the injector's capacity. Let the injection lag the cell by a radian and the line still holds, but almost no disturbance is survived.

applied · Turbomachine

Named alongside it

The objects these essays reach for when they reach for this one.

Model limitBifurcationStabilityHysteresisLubrication filmPredictabilityPump characteristicRelaxation timeStallThermal runawayTruncationTurbomachine

All concepts