Regal Sulkies: Track Design for Harness Racing

Poorly designed racetracks have long been identified as a major cause of economic loss to racing clubs, and avoidable injury to horses. While this article is based chiefly on Australasian practice, similar comments apply to a great many tracks of 1000 metres (5/8 mile) and less throughout the world, with the exception of certain Scandanavian tracks. The record shows that many clubs in Australia and elsewhere have been poorly served by track designers, with some tracks having to be ripped up only months after completion to correct avoidable errors in the original work. The industry pays a high price for these tracks in the form of loss of animals through avoidable injury, lower export prices, and meetings needlessly lost through wet weather. In this article the role of track banking is reviewed, and a track rating system is proposed.

'The selective breeding of racehorses for speed over several generations has resulted in the production of horses of high performance potential in what is now almost an industrial enterprise, involving considerable financial investment. Many horses are not capable of withstanding the severe stress to which they are necessarily subjected during training and racing, and hence develop locomotory lesions of varying severity at a relatively young age. The increased speed and frequency of modern racing demand high standards of training and shoeing and optimum design and maintenance of racetracks. Recent investigations have shown a relationship between defects in track design and locomotory disturbances predisposing to lameness'
— Fredricson & Alm. 1972

Since those words were written in the early 1970s, there has been a further significant increase in the speed of the Standardbred, so much so that the rate of improvement has led some commentators to forshadow a time when harness racers equal or surpass the speed of thoroughbreds. Be that as it may, the fact remains that the inadequately banked tracks ubiquitous in Australia and New Zealand (and, for that matter, in most parts of the world) extract an increasingly unacceptable price.

Senator Calvert, speaking in the Australian Senate to the reports produced by the Senate Select Committee on Animal Welfare in September, 1991, said:

'We had plenty of evidence to show that [poor] track design is one of the major contributors to animal injuries and deaths in racing . . .'

The trainer of Lucky Camilla, one of the leading contenders for this year's [1997's] Inter Dominion (Australasia's premier annual harness race), 'declared he would never run another horse at Harold Park.' - after the horse nearly fell in its first run on Sydney's top track (slightly under a half mile in circumference and considerably underbanked).

The combination of tight turns, inadequate banking and hard surfaces common in Australian tracks has drawn international criticism from equine lameness specialists, such as Dr Albert A. Gabel, of Ohio, U.S.A.:

'When I was in New Zealand and Australia, I noticed that the tracks over there are terribly hard. They're not springy; they're just hard. . . . It's no surprise that [ the horses of] Down Under trainers have far more fractured feet (coffin bones), a very serious injury, per 100 horses trained than we do. And it's because the tracks are just so hard. Fractured coffin bones are relatively rare here.' (Hoffman, 1985)

It is now well established that the consequences of inadequate track design are:

With the present critical shortage of available starters in many areas, it is folly to continue to permit injury-promoting and poorly drained tracks to be constructed.

In their 1975 paper , Fredricson et al, set forth a number of basic principles of track design derived largely from prior work in railroad track and highway construction, already known - if seldom employed - in harness racing in Australia. The paper was part of a research program for racetrack improvement sponsored by Harness Horsemen International, U.S.A., Harness Tracks of America, U.S.A., United States Trotting Association, and the Canadian Trotting Association.

The authors summarised their paper thus:

'To enable racehorses to maintain their gait, the design of racetracks must comply more closely with the principles of ergonomics and highway engineering, especially in respect of curve geometry. In future tracks of adequate length and straight/curve ratio, the curves have to be suitably banked to provide more suitable conditions of motion. Also, transition curves between the semicircular curves and the straights will eliminate present-day disturbances in gait symmetry at the entrance and especially the exit of the curves.

'Existing tracks can easily be improved by increased banking of the curves, incorporation of transition curves, and elimination of slopes in the straights. These measures may be expected to reduce the incidence of lameness and improve racing performance.'

Essentially, Fredricson & Alm (1975) argued that the flat or 'underbanked' turns common in race tracks caused significant and avoidable injuries to horses. This was not a new discovery. Rooney (1969) in the USA, and Ireland in Australia (who placed special emphasis on the adverse effect of poor banking on speed) had both published earlier articles with similar conclusions. In time, the new tracks with much steeper bank angles became common in Europe, and when introduced in the USA became known as 'Euro Tracks.'

Several articles have supported the need for optimal banking of tracks. Clayton (1987) expressed the view that:
'The majority of tracks will continue to be oval in shape but it is highly recommended that short, narrow turns be abandoned in favour of wider, more sweeping turns, with the incorporation of transition curves.

'Short tracks, although they are convenient for spectators, tend to be harder on the horses. For example, a half-mile track with sharp turns requires such heavy banking of the curves that it becomes difficult to maintain since the top cushion tends to slide down towards the inside rail. Consequently short tracks are usually underbanked.'

When tracks are underbanked they tend to injure horses. Fredricson and Alm (1972) stated:
'With optimum track banking, the strain on the horse caused by the curves will be relatively small [limited to the increase in sensible weight resulting from centripetal force], provided that the radius of the curve is not too short, even at speeds within a range around the design speed, covering most race requirements.

'Underbanked curves have been shown to cause gait asymmetry, leading to abnormal stresses in fast moving horses (Rooney, 1969; Fredricson, et al., 1975; Dalin, et al., 1973). There is no longer any justification for the heavy underbanking that exists today.'

The 1980 paper by the West Australian engineer K.J. Kelsall included, inter alia:

'8.7 On some American tracks, these basic principles [those of Fredricson, Alm and highway engineering generally] have been followed and, in 1979, a 5 furlong track in Ontario in Canada was completed with a super elevation [banking] on the curves of 18% [= bank angle of 10.2 degrees]. To a much lesser extent they have also been followed on the new layout at the Showgrounds in Hobart [Tasmania], where a 10% super elevation [5.71 degrees bank angle] has been provided.

In the article referred to earlier, 'Track design and what it means to horses', Dr. Hilary Clayton supported the earlier work of Fredricson & Alm, and wrote:

'the curves of a racetrack should be designed in a similar manner [to good highway engineering practice] so that the horse has optimum conditions of motion at a predetermined speed ("The design speed"). Under these circumstances the limb axes will be perpendicular to the track, so that hooves land flat and avoid the Mediolateral stresses that occur in underbanked turns.'

We simply cannot have it both ways. Either the track design speed closely approximates the average racing speed of a horse on that track, or the horse is denied 'optimum conditions of motion'. All Australasian tracks have 'underbanked turns', none permit the limb axes to be perpendicular to the track on those turns at racing speed, so all give rise to ' the mediolateral stresses that occur in underbanked turns.'

Dr Clayton (1987) also supports the use of transitions, 'it is highly recommended that short, narrow turns be abandoned in favour of wider, more sweeping turns, with the incorporation of transition curves.'

Kemp's Engineering Handbook (1975 Edition) has an excellent explanation of the purpose of transition turns, which may be paraphrased thus:

'The purpose of a transition curve is to achieve a gradual change of direction from the straight to the minimum radius of the curve in order to eliminate the sudden shock caused by the introduction of centripetal force.

'For any constant speed, the value of the centripetal force is proportional to the radial acceleration. A basic requirement of transition curves is that the rate of change of radial acceleration shall be constant throughout the curve and of such value that there is no discomfort to passengers travelling on the curve.

'If the rate of change of of radial acceleration is to be constant, the radius of curvature at any point on the curve will be inversely proportional to the distance of that point from the start of the curve.'

Transition turns have long been used on Australian tracks: Fairfield in 1954 (Timms - according to B.W. Ireland [private communication, October 1999], Timms' may have been the first in the world to use transitions on a harness racing track), Macau in 1978 (Ireland), Hobart in 1980 (Taylor), Harold Park in 1981 (Ireland), Albion Park in 1984 (Ireland), Southport in 1985 (Ireland), Gloucester Park 1985 (Ireland), Townsville in 1986 (Ireland), Newcastle in 1990 (Kelly & Associates), and Moonee Valley 1992 (Ireland). They have been used on the Red Mile track in the USA for over a century.

Anecdotal evidence suggests that when the angle of lean relative to the track surface exceeds 10 degrees, the incidence of injury to horses can be expected to increase sharply. The harder the track surface the sooner the onset of injury.

Notwithstanding the provision of optimum bank angles, the tighter the turn the greater will be the load on the horse's legs for any given speed. The load is simply the horse's weight plus the centripetal force accelerating the horse towards the centre of the turn. At a 2:00.0 mile rate, for example, a horse of 500 kg weight would experience an extra load of 12.47 kgs on a turn of radius 75 metres. Given a radius of 120 metres (similar to the Red Mile track in Lexington, Kentucky) and the same speed, the same horse would experience a load increase of only 7.79 kg, i.e. 4.68 kg less. At the current world record mile rate of 1:48.0 (53.65 Kmh), the difference between the same two tracks would be 5.77 kg in favour of the Red Mile. Aside from the increased load during the turns, more turns must be negotiated per mile on the smaller tracks, thus exacerbating the disadvantage. The driver and sulky also experience greater load-induced drag on small tracks for the same reasons.

This is why - all other things being equal - big tracks will always be faster than small tracks.

1400 Metre tracks (approx. 7/8 mile - such as the recently upgraded [from 1000 metres] Sportsman's Park track in Chicago USA) represent what many authorities believe to be an optimal compromise between speed, safety and maintenance costs on the one hand, and the need to have the field pass the spectators twice during a one mile race, on the other.

Fredricson's recommended curve/straight ratio of 550/400 yields the following statistics, which are supplemented with more recent work (Ireland, 1995) and provided as a guide:

Table 1.
Track circumference (metres)	Length of Straights (metres)	Minimum turn radius for a 1:50.0 mile rate (metres)	Minimum Safe Bank Angle (degrees)	Optimum Angle (degrees)
804.672	169.40	74.11	7.32	17.32
1005.84	211.76	92.64	3.26	13.26
1408.18	296.46	129.70	1.72*	9.55
1541.00	324.42	141.93	1.72*	8.74
1609.344	338.81	148.23	1.72*	8.38
1900.00	400.00	175.00	1.72*	7.11

Additional to the concepts of turn radii and bank angles, is that of the track surface finishing technique, generally described as "cushion". A correctly engineered cushion softens the surface and may alleviate some of the injuries caused by underbanked turns, while improving results from even the best track geometry. Dr Gabel had this to say (Hoffman, 1985;

'The real problem is with tracks being too hard. The track must be springy; it has to have resilience. The track has to be tuned to the frequency of the foot and then the horse will bounce off the track and that gives you speed. That's what they have done at Delaware, Ohio. About three or four weeks before the Little Brown Jug, they go out there with road equipment, dig up the track eight inches [20 cms]deep. That works air into it. They level it precisely and get it smooth. As they get it packed down, there's air in it, and if you train a horse over the track you can hear the difference. The track sounds springy, and it is springy, and that's why they get speed [but not the injuries].'

In January 1974, Prof. James Rooney wrote (Hoffman. 1985):

'Everybody blames track surfaces for their troubles and, in many cases, rightly so. What is the ideal surface on which to race or work a horse? It is obviously, the surface he has in the wild: plains country with a soil laced with with grass roots holding moisture - in other words an elastic, somewhat spongy surface that yields to the hoof and springs back again.'

However, irrespective of the excellence or otherwise of the cushion, it is the writer's belief that no tracks should be built with underbanking in excess of 10 degrees less than that called for by a design speed of 52.67 Kmh. The current Australian speed record for a horse running to a conventional long sulky is 51.27 Kmh; the present (initial) 140 metres speed record is 59.98 Kmh (equates to a 1:36.6 mile rate), and the long term trend is obviously to still greater speeds. Plainly, the future speed potential of harness horses must be planned for now in track design guidelines.

Anecdotal evidence suggests that there are major differences in speed and injury-proneness between different tracks. As a general rule, the faster the track record, the safer the track. The reason for that association is that poorly-designed tracks prevent horses reaching their full speed potential with pain.

One reason that there are no "optimum" tracks in Australia is because the difficulty of track construction and the on-going cost of track maintenance vary inversely with the size of an optimally banked track. Overwhelmingly, Australia has tracks smaller than 1000 metres in circumference. A 1000 metre track optimum for a mile rate of 1:55.0 would have curves of 92 metres radius and bank angles of 12 degrees. The maximum bank angle achieved in Australia is 5.71 degrees; in North America 10.2 degrees. Few, if any, Australian 1000 metre tracks have been built with curve radii greater than 90 metres.

Where the site is constrained by the topography the curve radii may, of necessity, be too small to allow optimum bank angles to be used with conventional surface technology. If a design speed of 52.67 Kmh (= mile rate of 1:50.0), the 70 metre radius turns would require a bank angle of 17.32 degrees. It is unlikely that such a steep bank angle could be economically maintained with conventional track construction techniques.

It appears however, that the development of new track surface technologies is well advanced - such as the shredded rubber 'Pro-Trak' type - and that these may permit the economical construction and economical maintenance of much steeper bank angles on turns than has hitherto been achievable.

A Universal Track rating System?

The single most important statistic for a track, from the safety frame of reference, is the 'design speed' of the track. The design speed being defined as that speed at which a horse can negotiate the tightest bend on the track without leaning in relation to the track surface. In other words, the speed which will allow the horse's vertical axis to remain at 90 degrees to the track surface on the tightest bend. The design speed will be found from the formula:

Design Speed (in metres/sec) = Square root of (tan a x R x g)
Where a = the bank angle in degrees
R = radius of turn in metres
g = Acceleration due gravity - 9.807 m/s/s

To convert to a mile rate, divide the result into 1609.34; to bring it to kilometres per hour, divide the result into 1000.

It is desirable that the Design Speed, as defined above, become the standard track rating system throughout the world. It is based on easily determined and fixed characteristics of tracks, it has the advantage that it is not susceptible to the hyperbole with which tracks are often described by enthusiastic, but often ill-informed operators, and it gives participants a pretty good 'handle' on what is really happening. To use the development over recent years of New South Wales' main metropolitan track (Harold Park) again as an example:

Year	Angle	Radius	Design Speed in K/h	Mile Rate
1992	2.86	61	25.41	3:48.02
1993	5.71	61	27.84	3:28.08
1996*	7.92	70	31.46	2:44.16

Plainly, there is a very considerable improvement from a 3:48.02 mile rating to 2:44.16 rating. Since the mile record (TT) on the 1992 track was 1:55.0, we might reasonably expect a considerable reduction in that rate after the latest rebuild is complete. Especially since Delaware USA (the world's fastest half mile track) has a design speed of 31.89 Kmh.

The other useful aspect of the design speed is that it is that unique speed at which there is no tendency for a sulky, mobile barrier, or any other wheeled vehicle to slide across the track. Thus the driver of the mobile will know that maintaining that speed through the turns in slippery conditions will minimise the risk of sliding sideways. He will also know that the greater the difference between the track design speed and his actual velocity through the turn, the greater will be the tendency for his vehicle to slide across the track.

As for the mobile, so for the sulky. In wet and slippery conditions, such knowledge may become a vital safety consideration. In order that participants know the track design speed, it should be displayed prominently in terms of both mile rate in minutes and seconds, and velocity in kilometres per hour.

At all velocities other than the track design speed, the tendency of a vehicle to slide laterally will be resisted by the friction between the tyres and the track surface. The combination of wet weather and track surfaces with high clay content may dramatically reduce that friction by comparison to that applying to a hard (especially unconditioned) surface under dry conditions.

Equally, at velocities above the design speed the sulky will tend to slide up the bank and away from the inside of the track, while at velocities below the design speed the sulky will tend to slide down the bank and towards the inside rail.

'Some ingenious scientific studies were carried out at the Royal Veterinary College in Sweden, to measure the strain to the fetlock joint, which is caused when trotters travel at different speeds around curves with various amounts of super elevation. The results were first published in 1973. As would have been expected, the studies showed that the strain was progressively decreased as the super elevation on the track was increased, and it was eliminated when the amount of super elevation was sufficient to fully counteract the centrifugal force [i.e. when the actual speed matched the track design speed].' (Kelsall, 1986)

From the point of view of safety, the above demonstrates again the very considerable value of knowledge of the track design speed to trainers and drivers. The trainer will know that at the design speed and no other, the strain on his horse's fetlock joint will be eliminated. If the track has properly transitioned turns, then keeping the horse to the design speed will also eliminate those 'locomotory disturbances predisposing to lameness.' Surely there must be few more essential pieces of information for a horse trainer to know?

It should be appreciated that the greater the demand made on the tyres to provide the centripetal force in the turns, the greater will be the drag or rolling resistance experienced by the wheels and the greater must be the horse's energetic cost of locomotion in overcoming that drag. In simple terms, the greater the difference between the horse's actual speed and the design speed of the curve, the greater the force exerted against the sulky tending to resist the horse's efforts to pull that sulky around that curve. This also is useful knowledge for training, racing and time trials.

With the greater use of track conditioners to prepare the track surface, comes a greater drag penalty as actual speed diverges from the track design speed. Thus the track designer is faced with the difficult - and for many, impossible - task of designing curves with radii less than 90 metres but with sufficient banking so that the design speed approximates the actual speeds of modern pacers, while at the same time incorporating a free-draining surface that will resist transport during heavy rain. Perhaps it is precisely this combination of difficulties with small tracks that has led to the continued popularity of 1400 - 1609 metre tracks (most notably in the USA) up to the present day# .

'for horses and public alike, the optimum length of a racetrack would be close to one mile. On such a course the relatively moderate slope of the turns . . . would considerably facilitate the design, construction and maintenance of the track.' (Clayton, 1987)

A 1400 metre track, built to Fredricson's bend to straight ratios, would have straights of 296.46 metres and turns of radius 129.7 metres. With a bank angle of 10.2 degrees its design speed would be 54.45 Kmh - quite satisfactory at current velocities.

Since it seems to be the case that the maximum bank angle that can be economically constructed and maintained with conventional technology is about 10.2 degrees (18%), we should aim to encourage tracks to be up-graded to that bank angle. Even tracks with 61 metre radius turns would be within 10 degrees of a 52.67 Kmh design speed (1:50.0 mile rate) at such an angle (actually, 9.5 degrees), and thus would be relatively safe for horses. The latter would have a design speed of 37.35 Kmh, compared to 19.68 Kmh for a 2.86 degree (5%) bank angle.

Ideally, of course, we should discourage or prevent the construction of tracks with such tight turns and adopt 70 metres as the smallest turn radius permissible. A track with 70 metre turns would have a design speed of 40.21 Kmh with a 10.2 degree bank angle, and be 7.1 degrees off the optimum bank angle for a design speed of 52.67 Kmh.

A 1000 metre track with 90 metre radii and 10.2 degrees of banking would have a design speed of 45.37 Kmh, but only be 3.6 degrees away from a design speed of 52.67 Kmh. This may suggest a quite simple aim; just standardise all tracks to a 10.2 degree bank angle. It will be below the average racing velocity for any track under 1400 metres, but is the maximum that can be economically maintained with conventional technology, and it will represent a considerable speed and safety improvement over present track practice.