Unstirring Coffee

Unstirring coffee

TDC

Kevin Cameron

EVERYONE SMOKED WHEN I WAS LITTLE, giving me many opportunities to watch smoke rise from a cigarette. Initially, the hot combustion gas and particulates rose straight up from the burning end of the cigarette in a narrow column. Then the column began to oscillate from side to side, and finally broke up into swirling eddies.

I was seeing the transition from orderly laminar flow (a flow consisting of parallel layers) to disordered turbulent flow.

Years later I would see this simulated by the pumping of viscous glycerine-water solution through a glass tube. To visualize the flow, a thin stream of red dye was injected into the center of the pipe. At low pumping speeds, the red dye drew a straight red line down the center. As the pump was speeded up, the flow would remain laminar for a distance downstream of the pipe’s inlet, but then the red line would begin to oscillate just as the cigarette smoke had done-and then it would break up into turbulent eddies. Increasing the pumping speed even more would move this point of laminar-to-turbulent transition farther upstream. Slowing the pump would move the transition downstream. The same process occurs in the boundary layer on the skin of an airplane or ship.

Cream poured into iced coffee presents a fascinating show of turbulent flow as the white cream curls around the sharp edges of the ice cubes, finally mixing fully with the black coffee at the bottom of the glass. A turbulent flow of this kind is irreversible-it’s impossible to “unstir” the cream from the coffee. But in a laminar flow, the motion takes the fonn of ordered layers sliding over each other with very little mixing. Curiously, this makes “unstirring” possible. I watched a cylindrical container of viscous liquid, marked with dye lines, as it was stirred by slow rotation past a stationary vane. The result was layer upon layer of plain fluid, separated by gracefully curved dye lines. When the rotation was reversed, the plain and dyed fluids “unstirred” and, after the right number of backward rotations, returned to a pretty accurate approximation of their original arrangement.

Laminar and turbulent flows are present in many processes important to the operation of motorcycles. Engine intake flow must be turbulent, for turbulence is as essential to fuel-air mixing as it is to the mixing of coffee and cream. A laminar flow is likened to the motion of a stack of cards that is pushed sideways-each layer slides past its neighbors, undisturbed by them, and with zero mixing between layers. In the “unstirring” experiment described above, layers of fluid were behaving just like the cards in the deck; and, like them, could be put back where they started with no loss of order.

In turbulent flow, the layers break up into random motion, and material from layers once far apart is rapidly brought together and mixed. In engine intakes, this turbulence rapidly mixes fuel vapor into the tumbling, swirling air and quickly replaces the fuel-vapor-rich “atmosphere” around each evaporating fuel droplet with fresh air, into which evaporation is accelerated.

How does the intake flow become turbulent enough to accomplish all this? The pumped glycerine experiment shows how it happens. The faster and farther the fluid moves down the pipe, the more likely is the flow to break up into turbulence. Engineers use what is called the Reynolds Number of the flow to roughly predict when laminar flow will decay into turbulence in this way. This dimensionless number is computed from fluid density, viscosity and velocity, plus a length characteristic of the flow situation. Above a certain range of values, the flow becomes turbulent. It is the high velocity in engine intakes that keeps the flow within highly turbulent, fostering rapid fuel-air mixing.

As the intake flow enters the cylinder, the energy of its high-speed motion is not lost, but fills the cylinder with turbulence as it breaks up against the surfaces within. This turbulence survives through compression and assures that when combustion begins, the flame kemel from the sparkplug will itself be whirled away, shredded and mixed into all parts of the combustion chamber. Without turbulence, flame speed would be so sluggish that piston internal-combustion engines would be impractical.

Cooling-by air or water-also requires turbulence, if airflow between cooling fins remains laminar, little heat is exchanged from layer to layer because there is so little mixing. Turbulence boosts the rate of heat exchange from hot metal to air because ¡i fresh, cooler air is continually being whirled from within the flow into contact with the metal. A recent NASA research program revealed that at the laminar-to-turbulent transition in surface flow over the Orbiter, the Orbiter’s skin temperature rises by 540 degrees F. This is because laminar flow is relatively insulating, while turbulent flow increases heat transfer. In this case, supersonic flow is heating the Orbiter, while in the cooling-fin situation, the metal fins are heating the air passing between them.

In early liquid-cooled motorcycle engines the water jackets were large, slowing water flow over metal surfaces. The resulting laminar (or only moderately turbulent) water flow produced sluggish heat transfer. Recent designs employ powerful water pumps and tight water jackets that generate fast and therefore highly turbulent flow across the surfaces to be cooled. This speeds heat removal.

Laminar flow can reduce heat transfer between oil or liquid coolant and the metal tubes through which it flows in radiators. Sometimes the flow must be forced into turbulence by building kinks into the tubes. As fluid flows around the kinks, it separates into eddies, bringing fluid from the middle of the flow into contact with the metal walls, restoring rapid heat transfer.

Although turbulence is invisible as it applies to motorcycle intake, combustion and cooling flows, we can always enjoy a simulation by ordering iced coffee and pouring in cream. □