Less resistance through bubbles only feasible at a smooth ship's hull
There is no point in trying to decrease the resistance for ships - and consequently, the fuel costs - by means of air bubbles if the ship's hull is not clean and smooth. Every rough spot, caused by, say, algal growth or corrosion, will completely reverse the effect. Researchers at the FOM Foundation and University of Twente have established this under laboratory conditions. They have their findings published in the scientific journal Physical Review Letters on 23 February 2007.
It has been known for quite a long time that a small amount of bubbles in a turbulent flow will decrease the resistance faced by a wall in a flowing liquid, up to fifty percent. This is important to the shipping trade. A ship that has a rough hull requires twenty times as much energy than a ship that has a clean, smooth hull. And if you have air bubbles flowing along such a smooth hull in order to decrease the resistance, then an additional fifty percent of energy saving may be obtained. Up to now this had been tested under 'ideal' laboratory conditions, so, with smooth walls. As walls are always somewhat rough in real conditions, the researchers were curious to know whether the decreasing resistance effect of bubbles would also occur at a rough wall.
Taylor-Couette system
Dr.ir. Ramon van den Berg, PhD student Dennis van Gils and professor Detlef Lohse (Physics of Fluids, University of Twente) have been experimenting with bubbles in a turbulent flow in a so-called Taylor-Couette system, which had been provided by professor Dan Lathrop (Institute for Research in Electronics and Applied Physics, University of Maryland, USA).
The Taylor-Couette system consists of two cylinders: inside a stationary outer cylinder there is a rotating cylinder. The space in between contains a liquid. The outer cylinder is 45 centimetres in diameter and almost 70 centimetres high. The closed parts on top and at the bottom of the cylinders have a built-in cold store in order to remain a constant liquid temperature during measuring, which is of vital importance to the properties of the liquid that is used (such as its density and viscosity). By doing so it is possible for the temperature to remain stationary up to one tenth degree Celsius and the properties of the liquid will be nearly unchanged during the measuring. The inner cylinder, 32 centimetres in diameter, consists of three parts; the middle part being the measurement section. A top and bottom part are needed in order to decrease the influence of the top and bottom of the closed construction to the measurement section. Actually, the large benefit of this construction compared to the more usual constructions for research on turbulence, is its closed construction.
The extent of turbulence
The inner cylinder rotates at a speed that can be set. By setting its velocity, the Reynolds number can be defined. This dimensionless number holds the velocity of the flow in the numerator and the kinematic viscosity in the denominator. By a dominating viscosity, the number will be small and the flow laminar. By an increasing velocity up to a particular value, then the viscosity is not able to repress the disruptions caused by this and the flow will become turbulent. Then, the result is a large Reynolds number. So, this number is an indicator for the degree of turbulence of the flow.
Roughness
First, the researchers have measured the resistance by defining the torsion on the measurement section of the inner cylinder. The torsion is the force that liquid exerts on the cylinder at a particular rotation velocity, thus friction. The higher the velocity, the more force the liquid exerts on the cylinder, and the more the force of the liquid 'frustrates' the cylinder. The torsion may be directly related to an indicator for the resistance that the liquid exerts on the inner cylinder, because the researchers at the University of Twente use a closed system.
Then, bubbles were injected at the bottom of the outer cylinder by using eight needles. The bubbles that were only one millimetre across, due to the strong flow in the liquid, will be influencing the torsion that the liquid exerts on the inner rotating cylinder. A roughness was brought in by putting perspex strips, twelve to the inner cylinder and twelve to the outer cylinder.
As a result, four different possibilities of testing arose: smooth walls without bubbles, smooth walls with bubbles, rough walls without bubbles and rough walls with bubbles.
No smoothness, no effect
The researchers proved that roughness is of vital importance to resistance. In a non-bubbly situation the necessary energy in gaining the same Reynolds number at rough walls appeared to be twenty times as much as at smooth walls. This large effect cannot be reversed by adding bubbles, as no single bubbly effect appeared to be found in a rough-wall situation.
When a decrease in resistance through injection of bubbles will be applied in shipping, then the problem of sticked walls must be solved first. This could be done, say, by putting on the right coatings.
The article is entitled: Bubbly turbulent drag reduction in a boundary layer effect, by Thomas H. van den berg, Dennis P.M. van Gils, Daniel P. Lathrop and Detlef Lohse. It will be published in 'Physical Review Letters' on 21 February 2007 on line, and on 23 February 2007 in print.
For more information, please contact Dr.ir. Ramon van den Berg (phone +31 (0)53 489 10 85) or Prof.dr. Detlef Lohse (phone +31 (0)53 489 80 76).