Swing

An overview of cricket ball swing

Rabindra D. Mehta

Sports Aerodynamics Consultant, Mountain View, California, USA

Abstract

The aerodynamic properties of a cricket ball have intrigued cricket players and spectators foryears, arguably since the advent of the game itself. The main interest is in the fact that the ballcan follow a curved flight path that may not always be under the control of the bowler. The basicaerodynamic principles responsible for the nonlinear flight or swing of a cricket ball were iden-tified decades ago and many papers have been published on the subject. Over the last 25 years orso, several empirical investigations have also been conducted on cricket ball swing, whichrevealed the amount of attainable swing and identified the parameters that affect it. Thosefindings are reviewed here with emphasis on phenomena such as late swing and the effects ofhumidity on swing. The relatively new concept of reverse swing, how it can be achieved inpractice, and the role in it of ball tampering, are also discussed in detail. In particular, the abilityof some bowlers to effectively swing an old ball in the conventional, reverse and the newlytermed contrast swing mode is addressed. A discussion of the white cricket ball used in the1999 and 2003 World Cup tournaments, which supposedly possesses different swing propertiescompared to a conventional red ball, is also included. This is a current overview of cricket ballswing rather than a detailed review of all research work performed on the topic. The emphasis ison presenting scientific explanations for the various aerodynamic phenomena that affect cricketball swing on a cricket ground.

Keywords: cricket ball, swing, reverse swing, aerodynamics, boundary layer

Introduction

Cricket is arguably one of the oldest games known tomankind. Although the origins of cricket are obscureand a source of much speculation, there is some

2005 isea Sports Engineering (2005) 8, 181192 181

evidence that it was played in England in the 1300s. Itwas certainly well established by the time of the Tudormonarchs (14851603). The first reference to cricketwas contained in a document dated December 1478and it referred to croquet near St. Olmer, in what isnow north-eastern France. The first recorded cricketmatch took place at Coxheath in Kent, England in1646 and the first Test Match took place betweenEngland and Australia in Melbourne, Australia in1877. The famous Ashes match was played inLondon in 1882, when an English newspaper printeda mock obituary notice after Australia had defeatedEngland.

People who do not play cricket know it as a game ofchildish simplicity which seemingly takes forever to

Correspondence address:Rabindra D. MehtaSports Aerodynamics Consultant209 Orchard Glen CourtMountain ViewCalifornia 94043USATel: (650) 960-0587Fax: (650) 903-0746E-mail: [email protected]

8.4.2 Sports D90 Mehta 7/6/06 4:57 pm Page 181

conclude. A pleasant game for the beach it seems, buthardly enough to hold the attention of entire nationsfor days at a stretch. Aficionados know cricket as agame of infinite subtlety, not only in strategy andtactics, but also in its most basic mechanics. On eachdelivery, the ball can have a different trajectory, variedby changing the pace, length, line or, most subtly ofall, by swinging the ball through the air so that it driftssideways. Such movement has always fascinatedcricket fans, but seldom do they understand theunderlying mechanisms that cause the cricket ball toswing. In fact, more often than not, they have to relyon myth and folklore readily and eagerly spread by themedia, rather than the basic principles of science.

It was the curved flight of tennis balls that firstinspired scientists to comment on the subject(Newton, 1672; Rayleigh, 1877). This type of spin-swing also occurs in cricket where it is usuallyemployed by the spin bowlers to generate the Magnuseffect (Mehta & Wood, 1980). However, there isanother type of swing that is unique to cricket and onethat is perhaps more intriguing. The actual construc-tion of a cricket ball and the principle by which thefast bowlers swing the ball is unique to cricket. Acricket ball has six rows of prominent stitching alongits equator, with typically 6080 stitches in each row,which makes up the primary seam. The better qualitycricket balls used in competitive cricket are, in fact,made out of four pieces of leather, so that each hemi-sphere has a line of internal stitching forming thequarter or secondary seam. The two quarter seamsare traditionally set at right angles to each other.These primary and quarter seams play a critical role inthe aerodynamics of a swinging cricket ball. It is saidthat this latter type of swing originated around theturn of the century, but there is evidence that it was inexistence well before that time. Dr. W.G. Grace, oftenacknowledged as the father of modern day cricket,and who played in the late 19th century was appar-ently an exponent of swing bowling. Bowlers fromthat era had realised that a perfectly new ball favouredthe peculiar flight, so there is not much doubt that itwas the traditional cricket ball swing and not spin-swing that the bowlers were referring to.

The first published scientific account of cricket ballswing was by Cooke (1955), who gave an explanationof why it was possible for fast bowlers to make a new

cricket ball swerve and why it became more difficultto do this when the shine had worn off the ball. Sincethen, several articles have been published on thetheories of cricket ball swing (Lyttleton, 1957;Horlock, 1973; Mehta & Wood, 1980). More recently,Barton (1982), Bentley et al. (1982) and Mehta et al.(1983) described detailed experimental investigationswhere the magnitude of the side force that producesswing and the factors that affect it were determined;see Mehta (1985) for a detailed review of the earlierwork. The relatively new concept of reverse swing,which first became popular in the late 1980s and 1990s,was first explained and discussed by Bown & Mehta(1993). A preliminary analysis of cricket ball swingusing computational fluid dynamics was described byPenrose et al. (1996). More recently, the flow fieldaround a cricket ball was measured and described byGrant et al. (1998); and Sayers & Hill (1999) publishedsome measurements of the aerodynamic forces on aspinning cricket ball. Some of the myths and miscon-ceptions surrounding cricket ball aerodynamics werepresented by Mehta (2000). All the measurementsshown in this article are taken from the authors ownresearch described by Bentley et al. (1982).

Aerodynamics of conventional swing

Fast bowlers in cricket make the ball swing by ajudicious use of the primary seam. The ball is releasedwith the seam at an angle to the initial line of flight.Over a certain Reynolds number (Re) range, the seamtrips the laminar boundary layer into turbulence onone side of the ball, whereas that on the other (non-seam) side remains laminar (Fig. 1). The Reynoldsnumber is defined as Re = Ud/, where U is the ballvelocity, d is the ball diameter and is the airkinematic viscosity. By virtue of its increased energy,the turbulent boundary layer separates later (furtherback along the ball surface) compared with the laminarlayer and so a pressure differential, which results in aside force, is generated on the ball, as shown in Fig. 1.In order to show that such an asymmetric boundarylayer separation can indeed occur on a cricket ball, aball was mounted in a wind tunnel and smoke wasinjected into the separated region (wake) behind theball, where it was entrained right up to the separationpoints (Fig. 2). The seam has tripped the boundary

182 Sports Engineering (2005) 8, 181192 2005 isea

An overview of cricket ball swing R.D. Mehta


layer on the lower surface into turbulence, evidencedby the chaotic nature of the smoke edge just down-stream of the separation point. On the upper surface, asmooth, clean edge confirms that the separatingboundary layer was in a laminar state. Note how thelaminar boundary layer on the upper surface hasseparated relatively early compared with the turbulentlayer on the lower surface. The asymmetric separationof the boundary layers is further confirmed by theupward deflected wake, which implies that adownward force is acting on the ball.

To confirm that an asymmetric boundary layer sep-aration on a cricket ball leads to a pressure differentialacross it, 24 pressure taps were installed on a ball

along its equator, in a plane perpendicular to that ofthe seam (Fig. 3). This pressure model was producedby cutting open a normally manufactured cricket ball,removing its core and gluing the two halves togetherafter the pressure taps and support tube had beeninstalled. Figure 4 shows the measured pressures onthe ball mounted in a wind tunnel with the seamangled at 20 to the oncoming flow. The data shownon the right-hand side in Fig. 4 represent the meas-urements on the seam side of the ball. At low values ofRe or U, the pressure distributions on the two hemi-spheres are equal and symmetric, so there would be noside force. At U = 25 m s1, the pressure dip on theright-hand (seam-side) face of the ball is clearly lowerthan that on the left-hand (non-seam-side) face, whichwould result in the ball swinging towards the seamside. The maximum pressure difference between thetwo sides occurs at U = 29 m s1 (65 mile/h), when the


R.D. Mehta An overview of cricket ball swing

Figure 1 Schematic of flow over acricket ball for conventional swing.

Figure 2 Smoke flow visualisation of flow over a cricket ball. Flowis from right to left. Seam angle = 40, U = 17 m/s.

Figure 3 Cricket ball with the core removed and 24 pressure taps(1 mm diameter) installed along the equator.

S

S

S: Separation point

Spin

Ball's line of flight

Turbulent boundary

layer

Side force

Laminar boundary

layer


boundary layer on the seam side is fully turbulent,while that on the non-seam side is still laminar. Evenat the highest velocity achieved in this test(U = 37 m s1), the asymmetry in pressure distributionsis still clearly exhibited, although the pressure differ-ence is reduced. The actual (critical) velocities orReynolds numbers at which the asymmetry appears ordisappears were found to be a function of the seamangle, surface roughness, and free-stream turbulence;in practice it also depends on the spin rate of the ball,as shown and discussed below.

When a cricket ball is bowled, with a round armaction as the laws insist, there will always be somebackspin imparted to it. In simple terms, the ball rollsoff the fingers as it is released. In scientific terms, thespin is necessarily imparted to conserve angularmomentum. The ball is usually held along the seam sothat the backspin is also imparted along the seam (theball spins about an axis perpendicular to the seamplane). At least this is what should be attempted, sincea wobbling seam will not be very efficient atproducing the necessary asymmetric orientation, andhence asymmetric boundary layer separation. Thistype of release is obviously not very easy to master,which is the main reason why not every bowler canswing a cricket ball effectively, even a brand new onewhich is more conducive to swing.

In order to measure the forces on spinning cricketballs, balls were rolled along their primary seamsdown a ramp and projected into a wind tunnel testsection through a small opening in the ceiling (Bentley

et al., 1982). The spin rate was varied by changing thestarting point along the ramp, and the seam angle wasvaried by adjusting the alignment of the ramp with theairflow. Once the conditions at the entry to the windtunnel and the deflection from the datum are known,the forces due to the airflow can be easily evaluated.The spin rate and velocity of the ball at the end of theramp were measured using strobe photography.Figure 5 shows the measured side force (F), nor-malised by the weight of the ball (mg), and plottedagainst the balls velocity; the side force is averagedover five cricket balls that were tested extensively. Atnominally zero seam angle (seam straight up, facingthe batsman) there is no significant side force, exceptat high velocities when local roughness, such as anembossment mark, starts to have an effect by inducingtransition on one side of the ball. However, when theseam is set at an incidence to the oncoming flow, theside force starts to increase at about U = 15 m s1

(34 mile/h). The normalised side force increases withball velocity, reaching a maximum of about 0.3 beforedeclining rapidly. The critical velocity at which theside force starts to decrease is about 30 m s-1. This isthe velocity at which the laminar boundary layer onthe non-seam side also undergoes transition andbecomes turbulent. As a result, the asymmetry



Figure 5 Variation of normalised side force with flow speed(averaged over five balls).

Figure 4 Pressure distributions on a cricket ball held at a seamangle of 20.


between the two sides (difference in the locations ofthe boundary layer separation points) is reduced andthe side force starts to decrease.

The maximum side force is obtained at a bowlingspeed of about 30 m s1 (67 mile/h) with the seamangled at 20 and the ball spinning backwards at a rateof 11.4 rev s-1. At a seam angle of 20, the Re based onseam height is about optimal for effective tripping ofthe laminar boundary layer. At lower speeds, a bowlershould select a larger seam angle so that, by the timethe flow accelerates around to the seam location, thecritical speed for efficient tripping has been reached.Of course, releasing a ball spinning along the seam(without much wobble) becomes more difficult as theseam angle is increased. Spin on the ball helps tostabilise the seam orientation. Basically, for stability,the angular momentum associated with the spinshould be greater than that caused by the torque aboutthe vertical axis due to the flow asymmetry. Too muchspin is also detrimental, since the effect of the ballssurface roughness is increased and the critical Re isachieved sooner on the non-seam side. To maximisethe amount of conventional swing, the ball surface onthe non-seam side should be kept as smooth as possibleso that a laminar boundary layer can be maintained.

The actual trajectory of a cricket ball can becomputed using the measured forces. Figure 6 shows

the computed trajectories at five bowling speeds forthe ball exhibiting the best swing properties(F/mg = 0.4 at U = 32 m s1, seam angle = 20, spinrate = 14 rev s1). The results illustrate that the flightpath is almost independent of speed in the range24 < U < 32 m s1 (54 < U < 72 mile/h). The trajecto-ries were computed using a simple relationship, whichassumes that the side force is constant and acts per-pendicular to the initial trajectory. This gives a lateraldeflection that is proportional to time squared andhence a parabolic flight path. In some photographicstudies of a swing bowler (Gary Gilmour, who playedfor Australia in the 1970s), it was confirmed that thetrajectories were indeed parabolic (Imbrosciano,1981). Those studies also confirmed that the finaldeflections of over 0.8 m predicted here are not unrea-sonable. One of the photographed sequences wasanalysed and the actual flight path is also plotted inFig. 6. The agreement is rather remarkable consider-ing the simplicity of the image processing andanalytical techniques. The data in Fig. 6 also have abearing on the phenomenon of late swing.

One of the popular theories for late swing suggeststhat a ball released at a speed just above the critical(with the boundary layers on both sides or hemi-spheres turbulent) may slow down enough duringflight so that the boundary layer on the non-seam sidereverts to a laminar state, thus creating a latemovement of the ball. However, it turns out that a ballreleased at postcritical Re slows down by less than 5%in flight, and, from the shapes of the curves in Fig. 5, itdoes not seem likely that this effect would occur inpractice. Another theory relies on a change in the ballorientation (through the gyroscopic precession effect),but test results indicate that this is not a significanteffect (Bentley et al., 1982). The suggestion thatsudden changes in wind direction can lead to lateswing (through a change in the seam angle) is also notvery likely to occur in practice (Wilkins, 1991). In fact,the data in Fig. 6 offer the best explanation for thephenomenon of late swing. Since the flight paths areparabolic, late swing is in fact built-in, whereby 75%of the lateral deflection occurs over the second half ofthe flight from the bowler to the batsman. So lateswing is most likely to be a natural, built-in part ofcricket ball swing, rather than an artifact of some new,unknown phenomenon.



Figure 6 Comparison of computed flight paths using measuredforces for the cricket ball with the best swing properties. Seamangle = 20, spin rate = 14 revs/sec.


Aerodynamics of reverse swing

Since the mid-1980s, there has been a lot of talk in thecricketing world of a supposedly new bowling conceptemployed by swing bowlers. The new concept or phe-nomenon is popularly known as reverse swing sincethe ball swings in a direction opposite (or reversed) tothat expected based on conventional cricketingwisdom and previously accepted aerodynamic princi-ples. This new form of swing bowling was firstdemonstrated (with astonishing success) by thePakistani bowlers, in particular Imran Khan andSarfraz Nawaz in the early years, followed by WasimAkram and Waqar Younis. They produced reverseswing very effectively, and generally using oldercricket balls, which obviously added to the intrigue.

Ironically, I first heard about the phenomenon ofreverse swing in the summer of 1980 from an oldschool mate of mine, none other than Imran Khanhimself, who is often considered to be the firstexponent of reverse swing. In talking about some ofthe issues regarding cricket ball aerodynamics, Imrantold me about a curious effect he had observed whenbowling. He was predominantly an inswing bowler,but he remarked that with the same grip and bowlingaction, the ball would swing away (outswinger) on theodd occasion. At the time, I honestly did not believethat such a phenomenon could occur since I could notexplain it using cricket ball aerodynamics as they wereunderstood at the time. However, in the followingyear when we started conducting experiments oncricket ball swing, the whole mystery was revealed

(Bentley et al., 1982). As discussed above, for conven-tional swing it is essential to have a smooth polishedsurface on the non-seam side facing the batsman sothat a laminar boundary layer is maintained. At thecritical Re, the laminar boundary layer on the non-seam side undergoes transition and the flowasymmetry, and hence side force, starts to decrease. Afurther increase in Re results in the transition andseparation points moving upstream, towards the frontof the ball. A zero side force is obtained when the flowfields (boundary layer separation locations) on thetwo sides of the ball become completely symmetric.In terms of reverse swing, the really interesting flowevents start to occur when the Reynolds number isincreased beyond that for zero side force. Asmentioned above, the transition point will continueto move upstream (on both sides now), setting up theflow field shown in Fig. 7. The transition points onthe two sides are symmetrically located, but theturbulent boundary layer on the seam side still has toencounter the seam. In this case, the seam has adetrimental effect, whereby the boundary layer isthickened and weakened, making it more susceptibleto separation compared to the thinner turbulentboundary layer on the non-seam side. The turbulentboundary layer on the seam side separates relativelyearly and an asymmetric flow is set up once again,only now the orientation of the asymmetry is reversedsuch that the side force, and hence swing, occurstowards the non-seam side, as shown in Fig. 7; this isreverse swing.



S

S

S: Separation point

Spin

Ball's line of flight

Side force

Figure 7 Schematic of flow over acricket ball for reverse swing.


Amongst other factors, transition is stronglydependent on the condition (roughness) of the ball'ssurface. This is demonstrated in the side force resultsfor three cricket balls with contrasting surface condi-tions (Fig. 8). The new two-piece ball (without thequarter seams) exhibits a higher maximum (positive)side force than the other two balls and the side forcedoes not start to decline until U = 36 m s1. This ballwill produce reverse swing only for velocities above45 m s1 (100 mile/h), which is not very useful inpractice, although it is worth noting that two-piececricket balls are generally not used in competitivecricket matches. However, the side force measure-ments for a new four-piece ball (with quarter seams)show that it achieves significant negative side force orreverse swing at velocities above about 36 m s1

(80 mile/h). Note how the magnitude of the negativeside force at 40 m s1 is not much less than thepositive force at 30 m s1. So it seems that reverseswing can be obtained at realistic, albeit relativelyhigh, bowling velocities. In particular, reverse swingcan clearly be obtained even on a new ball, without anytampering of the surface.

Some of the fastest bowlers, such as Jeff Thomson(Australia), Michael Holding (West Indies) and ImranKhan (Pakistan) from prior years, and Shoaib Akhtar(Pakistan) and Brett Lee (Australia) from presenttimes, have been measured bowling in the 40+ m s1

(90+ mile/h) range and so in theory reverse swingwould certainly be achievable by them. Alas, not everybowler can bowl at 40 m s1, so what about the meremortals who would still like to employ this new art?Well, there is hope, as shown in Fig. 8. The old ball,

with an estimated use of about 100 overs, gives lesspositive side force compared to new balls, but it alsoproduces reverse swing at a lower velocity of about30 m s 1 (67 mile/h). The contrasting results for thethree balls are directly attributable to the effects ofsurface roughness on the critical Reynolds number.With no quarter seams, a new two-piece ball has asmoother surface compared with a new four-piece balland the critical Reynolds number at which transitionoccurs on the non-seam side is therefore higher.Conversely, the critical Reynolds number on the usedball is lower because of the rougher surface. The keyto reverse swing is early transition of the boundarylayers on the ball's surface and the exact velocitybeyond which reverse swing is obtained in practicewill decrease with increasing roughness.

Contrary to popular belief, based largely oncomments initially made by Imran Khan and SarfrazNawaz, the cricket ball does not have to be wet onone side to produce reverse swing. The notion thatthis makes the ball heavier on this side and it wouldtherefore swing in that direction has no aerodynamicbasis to it whatsoever. However, there are somepossible advantages to wetting the ball (Wilkins,1991). For one thing, it makes it easier to gouge thesofter leather with the fingernails. Also, it is possiblethat the quarter seam may produce more roughnessby absorbing water underneath the exposed stitches,thus making the ridges more pronounced. Asdiscussed above, additional roughness reduces thebowling speed at which reverse swing can beobtained. The disadvantage of a wet ball is that it isvery likely to become heavier since the (bare) leathertends to absorb water and it will therefore swing less.

Effects of ball condition and ball tampering

For conventional swing, a prominent primary seamobviously helps the transition process, whereas asmooth polished surface on the non-seam side helps tomaintain a laminar boundary layer. Historically,bowlers have always paid a lot of attention to thesetwo features of the ball, although the scientific reasonsfor doing so may not be totally obvious to them. As alltrue gentlemen cricketers know, only natural sub-stances such as sweat or saliva can be legally used as apolishing agent, although the occasional use of



New ball without quarter seamsNew ball with quarter seamsOld ball with quarter seams

00.3

0.0

0.3

0.6

8 16 24U (m/s)

F / m

g

32 40

Figure 8 Normalised side force versus ball speed showingreverse swing.


Vaseline, Brylcreem or sunscreen lotion was often atthe centre of a ball tampering controversy, especiallyin the 1970s. Picking of the primary seam (lifting thestitching so that the primary seam height is retained)on aging balls is also technically illegal, but bowlerscan be often seen running their fingernails over thestitching.

Regardless of the chosen procedure for polishing,in order to continue obtaining conventional swingfrom a new ball, it is wise to polish the new ball rightfrom the start, but not on both sides. At the outset, theopening bowler should pick the side on the ball withthe smaller or lighter (less rough) embossment andcontinue to polish only that side during the course ofthe innings. The other (seam) side of the ball shouldbe allowed to roughen during the course of play to aidthe production of reverse swing. As shown above, theexact velocity at which reverse swing occurs, and howmuch negative side force is generated at a given speedabove the critical, is a strong function of the ballssurface roughness. Once the seam side has roughenedenough, reverse swing is simply obtained by turningthe ball over so that the rough side faces the batsman.In general, the production of conventional and reverseswing will not be affected significantly by having acontrasting surface condition on the side facing awayfrom the batsman. So a bowler bowling outswingerswill still have the seam pointed towards the slips, butwith the rough side facing the batsman, instead of thesmooth for conventional swing, and the ball will nowbehave like an inswinger and swing into the batsman.The whole beauty (and success) of this phenomenon isthat a bowler who could bowl only outswingers at theonset (with the new ball) can now bowl inswingerswithout any change in grip or bowling action. Similarly, apredominantly inswing bowler can now bowloutswingers. Of course, if the contrast in surfaceroughness on the two sides of a ball is successfullycreated and maintained, the bowler becomes evenmore lethal since he can now bowl outswingers andinswingers at will by simply changing the ball orienta-tion. Needless to say, this would make for a highlysuccessful ability since there are not many bowlerswho can make the new ball swing both ways usingconventional bowling techniques. Moreover, the fewthat can bowl inswingers and outswingers aregenerally not equally effective with both types of

swing and, of course, cannot do it with the same gripand bowling action. So the key to conventional swingbowling is keeping the non-seam side as smooth aspossible, whereas for reverse swing the non-seam sideneeds to be as rough as possible.

One of the reasons why reverse swing has gainedsuch notoriety is its constant link to accusations of balltampering. The fact that bowlers started to roughenthe ball surface illegally from the early 1980s is nowwell documented. Oslear & Bannister (1996) quoteand show several examples, and I have also personallyexamined several balls that were confiscated byumpires due to suspicions of ball tampering. The mostpopular forms of tampering consisted of gouging thesurface and attempting to open up the quarter seam byusing either fingernails or foreign objects such asbottle tops. It is rather ironic that a law prohibiting therubbing of the ball on the ground was introduced inthe same year that I first heard about reverse swing(1980). Also, in the following year (1981), the Test &County Cricket Board (TCCB) standardised the ballsso that they now had smaller seams. Of course, theimplications of these changes as they relate to reverseswing had yet to be realised.

There is another distinct advantage in maintaininga sharp contrast in surface roughness on the two sidesor hemispheres of the ball. The primary seam plays acrucial role in both types of swing. It trips the laminarboundary layer into a turbulent state for conventionalswing and thickens and weakens the turbulentboundary layer for reverse swing. During the course ofplay, the primary seam becomes worn and less pro-nounced and not much can be done about it unlessillegal procedures are invoked to restore it, asdiscussed above. However, a ball with a worn seam canstill be swung, so long as a sharp contrast in surfaceroughness exists between the two sides. In this case,the difference in roughness, rather than the seam, canbe used to produce the asymmetric flow. The seam isoriented facing the batsman (straight down the pitch)at zero degrees incidence. The critical Re is lower forthe rough side and so, in a certain Re range, theboundary layer on the rough side will becometurbulent, while that on the smooth side remainslaminar. The laminar boundary layer separates earlycompared to the turbulent boundary layer, in the sameway as for conventional swing, and an asymmetric




flow, and hence side force, is produced. The ball inthis case will swing towards the rough side. At veryhigh bowling speeds, the boundary layers on bothsurfaces will be turbulent and the ball will swingtowards the smooth side, much as with reverse swing.

This type of swing, which tends to occur when theball is older and a contrast in surface roughness hasbeen established, is often erroneously referred to asreverse swing. In order to avoid this confusion and dis-tinguish this type of swing from conventional andreverse swing, I have recently given it the name,contrast swing.

The most exciting feature about contrast swing isthat just about any bowler can implement it inpractice. As most cricketers are aware, it is much easierto release the ball (spinning backwards along theseam) with the seam straight up, rather than angledtowards the slips or fine leg. Thus, even mere mortalsshould be able to swing such a ball, and in eitherdirection, since the bowling action is the same forboth types of swing, the only difference being the ori-entation of the ball with regards to the rough andsmooth sides. In fact, medium pace seam or stockbowlers usually bowl with the seam in this orientationin an attempt to make the ball bounce on its seam sothat it may gain sideways movement off the ground.With a contrast in surface roughness, these bowlerscould suddenly turn into effective swing bowlers,without any additional effort, thus confusing not onlythe batsman, but perhaps themselves as well.

During the last two World Cups, there was a lot ofdiscussion about the swing properties of the white ballused in the tournaments. The white ball was intro-duced since it was apparently easier to see, both for theplayers on the field and for television viewers. Themain contention was that the white ball swung signifi-cantly more than the conventional red ball. Someplayers also contend that the white ball swings moreafter some use. According to the manufacturer of thewhite ball used in the 1999 tournament (BritishCricket Balls Ltd.), the only difference between thetwo balls is in the coating. With the conventional redball, the leather is dyed red, greased and polished witha shellac topcoat. This final polish disappears veryquickly during play and it is the grease in the leatherthat produces the shine when polished by the bowler.The finish applied to the white ball is slightly different.

The leather is sprayed with a polyurethane whitepaint-like fluid and then heat-treated so that it bondsto the leather like a hard skin. As a final treatment, onecoat of clear polyurethane-based topcoat is applied tofurther protect the white surface so that it does not getdirty easily. With the Dukes white ball used in the1999 World Cup and the Kookaburra ball used in2003, it is quite apparent on inspection that thesurfaces over the quarter seams are much smoothercompared with those on a conventional red ball, wherethe ridges created by the internal stitching can beclearly seen and felt. As a consequence, a new whiteball should swing more, especially at higher bowlingspeeds since a laminar boundary layer is more readilymaintained on the smoother surface. Another conse-quence of the smoother surface is that reverse swingwill occur at higher bowling speeds with a new whiteball and later in the innings at more reasonable bowlingspeeds. It was apparent during both World Cup tourna-ments that the ball became rough and dirty during thelater stages of an innings and reverse swing was readilyobserved on several occasions.

Effects of humidity

The effect of humidity on swing is by far the mostdiscussed and most controversial topic in cricket, bothon and off the field. It is quite fascinating that thistopic was discussed in the very first scientific paper oncricket ball swing (Cooke, 1955). The one bit of advicethat cricket gurus have consistently passed down overthe years is that a humid or damp day is conducive toswing bowling. However, the correlation betweenhumid conditions and swing has not always beenobvious and most of the scientific explanations putforward have also been somewhat far-fetched. Ofcourse, on a day when the ground is soft with greenwet grass, the new ball will retain its shine for a longertime, thus helping to maintain a laminar boundarylayer on the non-seam side. However, the realquestion is whether a given ball will swing more on ahumid or damp day.

As shown in the previous sections, the flow regimeover a cricket ball depends only on the properties ofthe air and the ball itself. The only properties of theair that may conceivably be influenced by a change inmeteorological conditions are the dynamic viscosity




and density. The dynamic viscosity and density bothappear in the definition of Re, but small changes inRe are unlikely to affect the side force coefficient sig-nificantly [the side force coefficient (CF) is defined as,CF = F/(

12U 2A), where is the air density and A isthe cross-sectional area of the ball]. However,changes in air density can affect the side forcedirectly, since for a given side-force coefficient, theside force is proportional to the density. However, itis rather ironic that humid or damp air is oftenreferred to as constituting a heavy atmosphere bycricket commentators, when in fact, humid air is lessdense than dry air.

A popular theory that has circulated for years, espe-cially amongst the scientific community, is that theprimary seam swells by absorbing moisture, thusmaking it a more efficient boundary layer trip. Bentleyet al. (1982) investigated this possibility in detail.Profiles were measured across the primary seam on anew ball before and after a few minutes soaking inwater. Even in this extreme example, there was no signof any change in the seam dimensions (Fig. 9). Asimilar test on a used ball (where the varnish on theseam had worn off) also showed no swelling of theseam. Rather than soaking the ball in water, a morecontrolled test was also conducted whereby a ball wasleft in a humidity chamber (relative humidity of 75%)for 48 hours. The projection test was performed onthese balls with the surface dry, humid and wet and noincrease in side force was noted for the humid or wetballs, as shown in Fig. 10.

Several investigators (Horlock, 1973; Barton, 1982;Sherwin & Sproston, 1982; Wilkins, 1992) have

confirmed that no change was observed in thepressures or forces when the relative humidity of theair changed by up to 40%. It has been suggested thathumid days are perhaps associated with generalcalmness in the air and thus less atmospheric turbu-lence (Sherwin & Sproston, 1982; Wilkins, 1992). Onthe other hand, Lyttleton (1957) and Horlock (1973)conjectured that humid conditions might result inincreased atmospheric turbulence. However, there isno real evidence or basis for either of these scenarios,and even if it were the case, the turbulence scales (sizeof the eddies) would generally be too large to have anysignificant effect on the flow regime over the ball.Binnie (1976) suggested that the observed increase inswing under conditions of high humidity is caused bycondensation shock which helps to cause transition.However, his calculations showed that this effect couldonly occur when the relative humidity was nearly100%. Also, as shown by Bentley et al. (1982), theprimary seam on almost all new cricket balls is alreadyadequate in tripping the boundary layer in theReynolds number range of interest.

There is only one published paper which claimsthat the positive effects of increased humidity onswing were observed in a wind tunnel test (Bowen,1995). Only two data points were presented, whichshowed that the side force coefficient was higher andthe drag coefficient lower for a relative humidity of54% compared with those at 36%. However, theshapes of the curves and the proposed explanation,



Figure 9 Surface contour plots of the primary seam on a cricketball to investigate the effects of humidity.

Figure 10 Effect of humidity on the measured side forces on aspinning cricket ball. Seam angle = 20, spin rate = 5 revs/sec.


that humidity increases the surface roughness on theball are both hard to believe. One would need to see alot more evidence and better explanations before suchan important result can be accepted for the first time.So there seems to be no (positive) scientific evidencewhich supports the view that humid conditions aremore conducive to swing. The only viable explanation,which was first proposed by Bentley et al. (1982), isthat humidity must affect the initial flight conditionsof the ball. There is a possibility that the amount ofspin imparted to the ball may be affected. The varnishpainted on all new balls reacts with moisture toproduce a somewhat tacky surface. The tacky surfacewould ensure a better grip and thus result in more spinas the ball rolls off the fingers, and, as shown above inFig. 5, an increase in spin rate (at least up to about11 rev s1) certainly increases the side force. So,perhaps actually without realising it, the bowler mayjust be imparting more spin on a humid or damp day.

Conclusions

The basic flow physics of conventional swing and theparameters that affect it are now well established andunderstood. However, some confusion still remains overwhat reverse swing is, and how it can be achieved on acricket field. A popular misconception is that when anold ball swings, it must be reverse swing. It is onlyreverse swing if the ball swings in a direction that isopposed to the one the seam is pointing in so that, forexample, a ball released with the seam pointed towardsthe slip fielders swings into the batsman. While it isgenerally believed (with some justification) thattampering with the balls surface helps in achievingreverse swing, the exact form of the advantage is still notgenerally understood. It is shown here that the criticalbowling speed at which reverse swing can be achieved islowered as the balls surface roughness increases. One ofthe more important points to note is that ball tamperingis not essential in order to achieve reverse swing.Reverse swing can be obtained with a brand new (red)four-piece ball, but only at bowling speeds of more than36 m s1 (80 mile/h). The whole beauty of reverse swingis that, by simply changing the ball orientation, andnothing else, the ball will swing the wrong way. With asharp contrast in surface roughness between the twosides of a cricket ball, contrast swing can be obtained

with the seam oriented vertically and pointed straightdown the pitch.

It is shown here how late swing is actually built intothe flight path of a swinging cricket ball and it is this,rather than some special phenomenon, that is oftenobserved on the cricket field. The question of theeffect of humidity on cricket ball swing is still nottotally resolved. While the effect is often observed onthe cricket ground, there is not enough laboratoryevidence to explain how the amount of swing may beincreased in humid conditions. The introduction ofthe new white ball with its unique outer cover finishhas started a new controversy on how its swing prop-erties may differ from those of a conventional red ball.So, while most of the mysteries surrounding cricketball aerodynamics have now been resolved, there arestill a few intriguing effects that will keep researchersbusy and motivated for some time ahead.

ReferencesBarton, N.G. (1982) On the swing of a cricket ball in flight.

Proc. Roy. Soc. Lond. A, 379, 109131.Bentley, K., Varty, P., Proudlove, M. & Mehta, R.D. (1982)

An experimental study of cricket ball swing. Imperial CollegeAero Technical Note, 82106.

Binnie, A.M. (1976) The effect of humidity on the swing ofcricket balls. Int. J. Mech. Sci., 18, 4979.

Bown, W. & Mehta, R.D. (1993) The seamy side of swingbowling. New Scientist, 139 (1887), 2124.

Bowen, L.O. (1997) Torque and force measurements on acricket ball and the influence of atmospheric conditions.Trans. Mech. Eng. IE Australian, ME20 (1), 1520.

Cooke, J.C. (1955) The boundary layer and seam bowling.The Mathematical Gazette, 39, 196199.

Grant, C., Anderson, A. & Anderson, J.M. (1998) Cricketball swing the Cooke-Lyttleton theory revisited. In: S.J.Haake (Ed.) The Engineering of Sport Design andDevelopment, Blackwell Publishing, Oxford, pp. 371378.

Horlock, J.H. (1973) The swing of a cricket ball. ASMESymposium on the Mechanics of Sport.

Imbrosciano, A. (1981) The swing of a cricket ball. ProjectReport. Newcastle College of Advanced Education,Newcastle, Australia.

Lyttleton, R.A. (1957) The swing of a cricket ball. Discovery,18, 186191.

Mehta, R.D. and Wood, D.H. (1980) Aerodynamics of thecricket ball. New Scientist, 87 (1213), 442447.

Mehta, R.D., Bentley, K., Proudlove, M. & Varty, P. (1983)Factors affecting cricket ball swing. Nature, 303, 78788.

Mehta, R.D. (1985) Aerodynamics of sports balls. AnnualReview of Fluid Mechanics, 17, 151189.




Mehta, R.D. (2000) Cricket Ball Aerodynamics: MythVersus Science. In: A.J. Subic and S.J. Haake (Eds.) TheEngineering of Sport Research, Development andInnovation. Blackwell Science, London, pp. 153167.

Newton, I. (1672) New theory of light and colours. Phil.Trans. R. Soc., 1, 678688.

Oslear, D. & Bannister, J. (1996) Tampering with Cricket.Collins Willow (Harper Collins) Publishers, London.

Penrose, J.M.T., Hose, D.R. & Trowbridge, E.A. (1996)Cricket ball swing: a preliminary analysis usingcomputational fluid dynamics. In: S.J. Haake (Ed.) TheEngineering of Sport. A.A. Balkema, Rotterdam, pp. 1119.

Rayleigh, Lord (1877) On the irregular flight of a tennisball. Messenger of Mathematics, 7, 1416.

Sayers, A.T. & Hill, A. (1999) Aerodynamics of a cricketball. J Wind Engineering & Industrial Aerodynamics, 79,169182.

Sherwin, K. and Sproston, J.L. (1982) Aerodynamics of acricket ball. Int. J. Mech. Educ., 10, 7179.

Wilkins, B. (1991) The Bowlers Art. A & C Black Publishers,London.


