Updated November 2017

The complex interactions in caravan and tow vehicle dynamics described here by Collyn Rivers, is a semi-technical précis of his still in progress version. It will be progressively updated until the fully referenced version is completed.

Vehicle-drawn caravans were common by the mid-1920s. From their very beginning, however, they had swaying (yawing) and handling problems. Now, reports of rigs jack-knifing and overturning appear to be increasing. Most relate to long end-heavy caravans, moreover, those towed by single or dual cab tray back vehicles.

At that same time, the commercial transport industry in the early 1900s realised that trailers with centralised axles towed via overhung hitches were unstable. Fruehauf in the USA realised its cause was lateral forces inherently imposed on the trailer by the overhung hitch. Worse, that as the trailer yawed to (say) the right, it caused the tow vehicle to yaw to the left. The longer that overhang (Moment Arm) the greater the effect.

The inevitable and unstable result is the 1800 (mechanical) phase change shown in (Figure 1).spacer height=”20px”]

Caravan sway out of phase web

This is the inherent problem with a conventional caravan. If one part of the rig yaws it causes the other to yaw in the opposite plane. Pic: copyright Caravan & Motorhome Books, Church Point, NSW, Australiacaravanandmotorhomebooks.com

Realising the forces are eliminated if the hitch is directly over the tow vehicle’s rear axle/s, led to the semi-trailer concept. Within a few years, the transport industry adopted it en masse. And to the B-double and B-triple rigs of today. The first known fifth wheeler caravan was as early as 1917.

Caravan sway fifth wheel new

The fifth wheel concept. Yawing of one or other part of the rig barely affects the other. Pic: caravanandmotorhomebooks.com

Conventional caravans back then were typically short and light, and rarely towed faster than about 80 km/h. A few rollovers occurred but it was not until the 1970s that their increasing length and weight led to serious study of caravan and tow vehicle dynamics.   

These studies, and associated practical trials, identified the dominant causes. Furthermore, that many interact.

Those that mainly affect stability are trailer yaw inertia, nose mass (mass distribution) and trailer axle position. Tyre pressure and side wall stiffness too affects stability, particularly of the tow vehicle. Speed is always a factor.

This ongoing updated article discusses the many interacting issues. Each is first described individually. Later sections attempt to explain the interactions and consequent effects. Also included are suggestions of how to minimise such effects.

In essence, any trailer pulled via an overhung hitch is fundamentally unstable. It is, however, tameable within practical  limits by minimising contributory causes.


Caravan and tow vehicle dynamics – terms used

Caravan and tow vehicle dynamics article is primarily intended for those interested but not necessarily deeply technical. Some terms used have have precise meanings that are often not understood. These are explained here in an (intentionally) non-academic manner.

Mass and weight: are different concepts. Confusing one with the other may not matter with static objects (on Earth). It does, however, for objects that move, both on Earth and in space.

Mass: this relates to the amount of matter within a body. It is quantified in terms of the resistance of a body to being accelerated.

Weight: (for non-technical purposes) can be regarded as what most people think it is. It is however the concept of mass attracting another body that has mass. That which we call ‘gravity’ is the attractive downward force of the Earth’s mass. Because there are no gravitational forces acting upon it, mass has zero weight in space. But were a mass of (say) 5 kg to be thrown in space its impact will be much as on Earth.

Laws of Motion: much of caravan and tow vehicle behaviour can be explained by the basic laws of motion defined, in the late 1600s by Sir Isaac Newton.

The first is that, unless influenced otherwise by an external force, mass remains at rest, or continues to move at constant speed in a straight line unless acted upon by some external force.

The second is that time rate of change of momentum (see below) proportional to the applied force and takes place in the direction that the force is acting.

The third is that to every action, there is an equal and opposite reaction.

Force: is that influence on a body that causes it to accelerate. The greater the force applied, the greater the rate of change of acceleration. That rate of change is directly proportional to the force acting upon it. It is inversely proportional to the mass of that body. Force has magnitude and direction and describing it meaningfully requires both terms.

In SI units, force is expressed in Newtons (N). A Newton is the amount of force required to accelerate 1.0 kg  1 m.s-2. That is: 1 N = 1 kg.m.s-2.

Moment arm: A moment arm is, in effect, leverage. It determines the effect of where weight is located along the length of a caravan. It also relates to the effect of tow hitch overhang.

Torque: is the effect of a force causing something to roll or rotate. It relates to the force itself, and to the distance that force is applied from the axis of rotation.

Torque and moment arm have similar meanings. Moment arm is more commonly used where the effect is simply leverage (as with hitch overhang).

If two people of different weight sit at unequal distances from the pivot of a see-saw, balance can only be achieved by the heavier person sitting closer the pivot, or the lighter sitting further away. A similar effect happens when a couple of spare wheels and a loaded tool box are located at the rear of a long caravan. Its undesirable effect is many times that of having it close the caravan’s axle/s.

The sketch below shows how moment arms apply to a caravan – here the effect of distancing mass from the axle/s.

Moments along a beam website

Here, +B and -B are 1.0 metre either side of the axle, +C and -C are 2.0 metres, and +D is 2.5 metres. Each position has a load of 200 kg. The dynamic effect of such loading however is 200 kg at +B and – B, 400 kg at +C and -C, and 500 kg at +D. In other words, it is not so much weight, but where that weight is located, that matters. Pic: Copyright Caravan and Motorhome Books, Church Point, NSW 2105, Australia. caravanandmotorhomebooks.com

Torque is more commonly used where there’s some form of rotation, as for example when a caravan is pitching or swaying – or you are tightening a wheel nut with a wrench.

A good way to experience this is by closing a heavy door. Pushing close to the hinge requires more force but less movement. Pushing it further from the hinge requires less force but more movement. The work done and energy required is the same regardless of where it’s pushed.

Inertia: is a major factor in caravan and tow vehicle dynamics behaviour. That mass resists inherently resists changes to its state of motion or rest is known as inertia. It is usually expressed by its mass in grams.

Momentum: is the quantity of motion. It is directly proportional to the product of the mass and velocity of a moving object, i.e. momentum = mv). By virtue of its velocity, a fired 350 gram bullet thus has higher momentum than that resulting from the slower reaction of whoever fired the gun.

Acceleration: relates to a change in a mass’s rate of velocity. It may positive or negative (as when braking). It is measured by dividing velocity (metres per second) by seconds. That’s the same as dividing distance by time squared.

The unit often shown as ‘G’ (but correctly as ‘g’) is an acceleration of approximately 9.81 m/s2.

Moment of inertia: is a measure of an object’s resistance to changes in rotation (also capacity of a cross-section to resist bending). The axis of rotation must always be specified. It is usually quantified in kg/m2.

For caravan purposes that axis of rotation is the caravan’s so-called’ radius of gyration (see below). It can be calculated mathematically, using integral calculus, by ‘cutting the caravan into thin slices’ that each has a mathematically describable shape. It’s much simpler done practically (and probably more accurately) on a friction-free turntable that is rotated, by about 30 degrees, against the force of two or more springs. It is then released. The time taken to return to centre is then used to calculate the moment of inertia.

Radius of Gyration: this, in mechanics is where the centre of mass would be, were the entire mass to be concentrated in one place. It is the square root of the ratio of the moment of inertia of a body about a given axis to its mass. (Mathematicians have a different definition – that matters here only to them.)

Work: has a very specific meaning in this and similar contexts. It can be seen as a transfer of energy or the application of force over a distance. Lifting a heavy object and putting it on a high shelf is one example. The unit of work is the Joule. One Joule is 1 Newton times 1 metre.

Energy: can be seen as the stored-up ability to perform work (or the capacity or doing work). It can be expressed as force times displacement in unit time. That involved in connection with caravans and tow vehicles is mainly potential energy and kinetic energy.

Potential Energy: is the capacity of a body to do work by virtue of its position or configuration. That held within a vehicle spring when compressed is potential energy (also aptly called ‘elastic energy’). So is that held chemically within a charged battery.

Kinetic Energy: is that associated with motion. Any moving object is able to do work as a result of it moving. It is often regarded as being the same as momentum – but it’s not. Kinetic energy is proportional to the square of the mass’s velocity. It is defined as Ek1/2mv2.

A tow vehicle and caravan travelling at 100 km/h thus has four times the kinetic energy of that same mass travelling at 50 km/h – not twice. The reason why a jack-knifing is so violent is that the kinetic energy of that rig (as the result of a force that accelerated it to 100 km/h over 100 or more seconds) is mostly released over a final second or two.

Power: is the amount of work done in a unit of time. For example the work done when a tow vehicle pulls a caravan up a hill always remains the same, but doing that at 100  km/h will need a lot more power (for a shorter time) than doing so at 50 km/h. A tow vehicle and caravan travelling at 100 km/h thus has four times the kinetic energy of that same mass at 50 km/h – not twice.

The reason why a jack-knifing is so violent is that the kinetic energy of that rig (as the result of a force that accelerated it to 100 km/h over 100 or more seconds) is mostly released over a final second or two.

Yaw: refers to a deflection in a rotational or oscillatory movement of a body such as a caravan about its vertical axis. Many people refer to yaw as sway. This can confuse. When a caravan sways (rolls) its centre of gravity moves laterally.

Yaw Force: is the influence of (say) a side force that causes the front of a caravan to be accelerated sideways. The greater that force, the greater the rate of change of acceleration. As a moment arm (leverage) is introduced by the tow hitch being located to the rear of the tow vehicle’s rear axle/s that force is increased by the length of the moment arm of that hitch.

Yaw Inertia: can be seen (here) as the resistance of mass to that change of motion when that mass is subject to a side force (‘yaw’).


Caravan and tow vehicle dynamics – tyre behaviour

Horse-drawn carriages had pivoting ‘fifth wheel’ front axles. Their iron or (later) solid rubber tyres wheels were thus in line with the horses’ pulling force. But if the carriage’s inertia overwhelmed the horses’ inertia and their hooves’ grip, the horses would lose control. The carriage would then plough more or less straight ahead.

Tyres back then had mainly to revolve, not sink nor fall apart under load. Their often marginal surface grip helped restrain sliding sideways. Rudimentary braking (mostly whilst parked) was by levering against a tyre.

The forces required for traction, steering and control downhill were thus exerted by external animal power. (The suspension, however, was surprisingly advanced – using leather straps for damping).

The advent of motive power showed dramatically that ‘horseless carriages’ were subject to Newton’s findings. There was however a huge difference. The forces required for propulsion, braking and steering now had to be applied and reacted solely through the vehicles’ own tyres. 

stagecoach accident-1856-granger

A caravan’s tow vehicle acts physically much as those horses.The caravan ultimately depends on it in much the same manner as did horse-drawn carriages.This is only too often overlooked. Pic: courtesy of fineartamerica.com/featured/stagecoach-accident-1856-granger/

That early engines developed little power so traction was rarely an issue. Nor was speed, so little braking was needed – or provided. The solid rubber tyres rolled where pointed, so steering was not a problem unless they lost all grip.

Pneumatic tyres used on early cars were like oversized-bicycle tyres. They behaved like a softer solid tyre in that they rolled more or less where pointed. They relied on friction to turn the vehicle as required, and assist react unwanted forces that attempted to deflect the vehicle. When, much as with solid tyres, forces exceeded the tyres’ grip, early tyres slid in a relatively progressive, predictable manner.

Then cars became heavier and faster. Tyres became balloon-like. Wheel rim diameters reduced accordingly. Owners (now owner-drivers) sought a softer ride. All introduced problems that were not understood at the time. Softly sprung cars of that era handled poorly and often not predictably. It was not until the mid 1930s that it was understood that suspension geometry and its interaction with large cross section tyres substantially dictated on-road behaviour. In 2014 it still does.

The above is particularly true of caravan and tow vehicles, especially in evasive manoeuvres etc. In extremis, a rig’s ultimate behaviour is a function of its tyres’ behaviour. This is not understood by all caravan makers, nor caravanners.

Caravan and tow vehicle dynamics – tyre basics

A correctly inflated tyre does not roll over a surface. It lays down an elongated oval of tread (called its footprint). The footprint’s length and stability is substantially proportional to tyre construction and tyre pressure.

Slip Angle Graph good

A typical tow vehicle tyre (green) increases ‘cornering power’ as it slip angle increases, but then levels off and starts falling away sharply between 12-20 degrees. The latter will introduce major and possibly terminal oversteer, resulting in a  jack-knife.

The footprint grips a hard surface with tenacity but, unlike classical Coulomb friction, that grip is not linear with weight. When cornering, or reacting a yawing caravan or wind gust, the now higher load on the outer tyres increases cornering power by only 0.8 or so of that increase in grip.   

Steering a tyre is akin to twisting an inflated rolling balloon. Torque is applied, via the wheels’ rims, to the tyres’ sidewalls. These flex, and via their stiffness and air compressed within, cause the footprint to distort (i.e. not turn) in the direction required. The footprint is thus a springy non-linear medium that has a partly molecular bond. It never fully takes up the angle the steered road wheels’ rims attempt to impose except on a totally flat and straight road with zero side wind.

Caravan and tow vehicle dynamics – slip angles

The ‘slip angle’ is the angular difference between where the wheel rims point and the vehicle actually goes. The greater the tyre width, and sidewall and tread stability (and the higher the tyre pressure) the lesser the slip angle.

tyres slip_angle3

The term slip angle can mislead because, in normal driving, the footprint does not slip. It is subjected, by torque applied by the sidewalls, to a diagonal-like stretching/distortion. It is only when side forces exceed total footprint grip, that the slip angle becomes infinite.

Interaction of slip angles

The relative interaction of front/rear slip angles substantially dictate how a road vehicle behaves. It lies right at the heart of caravan and tow vehicle dynamics.

Almost all vehicles now have front slip angles intended always to exceed rear slip angles by a slight maintained degree. This effect, called understeer, causes vehicles to veer slightly away from side-disturbing forces, as do correctly trimmed yachts, and all aircraft.

If cornered too fast, an understeering vehicle automatically runs slightly wide, thus taking up a marginally wider radius. This reduces the side forces and hence slip angles. Such vehicles (virtually all post-1960s) are marginally less responsive to steering input, but safer for most drivers.

If however the rear slip angles are caused to exceed that of the front, the vehicle turns into an ever-tightening spiral. This causes the rear slip angles to progressively increase. If not corrected (by applying opposite steering lock, and/or reducing speed) the rear slip angles increase until their footprints lose control. The vehicle then spins.

understeer and oversteer

Under steer and oversteer in extreme. In mild form, under steer adds stability. If the vehicle is corned too fast, it automatically adopts a  wider radius turn, thus reducing undesired forces. Too much understeer however can result in the above. The above is from www.driversdomainuk.com/img/oversteer.jpg (original source uknown).

This unstable condition, called oversteer, may occur if the tow vehicle’s rear outer tyre is overladen whilst attempting to react (say) the yaw forces of a strongly yawing caravan. Another cause may be a weight distributing hitch (see below) that over-corrects whilst the caravan is yawing or the rig is swerving.

A neutral steering vehicle, i.e. with identical front/rear slip angles maintains a precariously balanced state. This condition, known as neutral steer, is impossible to achieve in practice. This is because even minor changes in tyre pressure, loading, or minor road camber will cause it to understeer or oversteer.  A neutrally steering car requires constant steering corrections: it is demanding and tiring to drive, and unsafe for unskilled drivers.

Caravan and tow vehicle dynamics – maintaining footprint balance

Problems arise if caravan builders and (particularly owners) do not understand the above: that a rig’s dynamic behaviour depends ultimately on the tow vehicle maintaining optimum tyre footprint and slip angle behaviour.

Whilst it may seem blindly obvious, the above requires caravan tyres to be firmly on the ground. Despite this some trailer and caravan makers (and owners) maintain that trailers don’t need shock absorbers (energy absorbers). They argue that, if leaf sprung, that interleaf friction provides sufficient damping.

It does crudely, insufficiently and only on compression, where it’s least necessary. The problem-causing rebound is totally unrestrained – the spring leaves are not then held in sliding contact and release their energy instantly. This causes tyres to bounce on rebound and lessen or lose their grip. It also results in major shearing forces wheels studs and stub axles. See Wheels Falling off Trailers

Inadequate or non-existent spring damping can thus prejudice electronic stability systems that rely totally on caravan braking.

Caravan and tow vehicle dynamics – slip angles and load/tyre pressure etc

Cornering power decreases with load, and to some extent increases with tyre pressure. Higher loads thus require higher (tow vehicle) rear tyre pressures to restore/retain the slip angle/s required. Conversely load removed from the front of the tow vehicle often requires front tyre pressure to be reduced to restore otherwise lost understeer.

If the tyre’s weight balance is unchanged front/rear, front and rear slip angles increase proportionally with cornering, and the vehicle’s balance is maintained.

But if rear outer tyre loading only is substantially increased (as when a caravan yaws), the related tyre’s slip angles change accordingly. If that induces extreme oversteer, the footprint is likely to lose all grip and the rig jack-knifes.

The relative tyre loading front/rear (and hence slip angles) is not simply a function of weight distribution. It depends on how the suspension resists roll.

Stiffening the rear springs or adding rear airbags without doing so proportionally at the front, transfers more of the roll couple (roll resistance) to the rear suspension. This yet further loads up the outer rear tyre whilst cornering, thus increasing its slip angle. Here again if that footprint collapses or slides there is a very real risk of jack-knifing. This is not theoretical conjecture. It happens.

Because of the above, vehicle manufacturer’s intended front/rear slip angle relationship should not be changed. If the rear suspension is stiffened (or air bags fitted), so should the front to the same degree (or a stiffer roll bar added) to maintain the intended roll resistance balance.

Minor spring rate changes are rarely detectable in normal driving. This is why so many people claim it’s safe. But if understeer is reduced (by say, stiffening the rear springing alone) the then close to neutrally steering vehicle can be triggered into terminal oversteer by a sudden strong yaw force.

Generally speaking, passenger vehicle suspensions that appear to need upgrading are being used for loads other than they are intended to carry. Suspension changes require serious engineering expertise – not buying airbags on eBay.

Caravan and tow vehicle dynamics – towball weight

As with an arrow, caravan and tow vehicle dynamics is such that some frontal weight is essential for a caravan to keep straight. That traditionally recommended in Australia (until tow ball limits set by tow vehicle makers decree otherwise) was about 10 per cent of the gross trailer weight. That figure was arrived at 60-70 years ago when most caravans were around 4.0 m long and weighed 1000-1200 kg. It became virtually a mantra but was, at best, realistic only for centre-kitchen vans of these sizes. The British, for reasons unclear, settled on 7%.

As physicists and engineers will appreciate, a tow ball load based on a percentage of caravan weight alone ignores the more significant effects of moments along a beam (let alone the moment of inertia).

What matters far more with caravan and tow vehicle dynamics is the caravan’s length and where mass is distributed along that length. A 250 kg mass at the front or (worse) rear of a 7.0 metre caravan results in greater yaw and pitching forces than does the same mass on a 4.0 metre caravan. Because of this, 10% may be too low for a 7 metre end heavy van. This is an increasing problem as tow ball limits and tow vehicle weight decreases.

Caravan and tow vehicle dynamics  – weight distributing hitches

The tow ball download on an overhung hitch causes the tow vehicle’s rear to be thrust downward but, as with pushing down on the handles of a wheel-barrow, the front lifts. This undesirably shifts weight from the tow vehicle’s front (steering) tyres.

A weight distributing hitch (WDH) assists to remedy this. Its effect is to form a semi-flexible sprung beam between tow vehicle and trailer. This restores some of the otherwise reduced imposed weight on the tow vehicle’s front axle.

From a caravan and tow vehicle dynamics perspective, a WDH addresses a condition that would better not be there in the first place. It’s a bit like using a truss to support a hernia. It is preferable to design caravans such that they need less download (as is being done in Europe). Australian (and US) reality however is that the tow ball loading is too high for most tow vehicles to cope with unless a WDH is used – as the otherwise loss of weight on the steering wheels is not acceptable.

Recent (2013-2016) recommendations (SAE J2807) is to adjust them such that they correct no more than 50% of the tow vehicle’s rear end droop (not the full amount). It suggest’s that 25% may be better. The reason for doing this is that a WDH inherently reduces understeer – precluding the rig being able to withstand the specified yaw forces of 0.4 g. (FALR of 100% limits the rig to 0.3 g). Advice to this effect is now given by Cequent in the USA (parent company of Haymen Reese).

Caravan and tow vehicle dynamics – independent suspension

The rationale for independent (front) suspension stems from the early 1930’s move to softer suspension. That, plus the longer suspension travel resulted in beam front axles occasionally ‘tramping’ violently. The front wheels would alternately jump fully up and down, meanwhile (as with the steering) swinging often violently from lock to lock.

General Motor’s Maurice Olley established that it was partially to gyroscopic precession the motion of a spinning body, which if also rotated in an axis other than in which it spins, its axis of rotation sweeps out in a cone.

To experience this, hold the front of a bicycle off the ground, spin the front wheel and swing it in an arc. Or play with a gyroscope.

gyroscopic precessionHere, innovative (US) teacher Gary Rustwick demonstrates the effects of gyroscopic precession by swinging the spinning wheel in an arc whilst standing on a free-moving turntable (as he does so the precession forces will cause the turntable to rotate).

Precession thus occurs when a fast-rotating, steerable wheel rises and falls in an arc whilst traversing bumps.

Associated with is that a moving front wheel would, once the then so-called ‘shimmy’ cycle began the tyre would contact the road surface toed-in, in turn causing it to swerved out forcibly. This in turn caused a gyroscopic torque (due to the forced precession of both wheels). ‘This torque lifted the down wheel, and slammed its opposite number down on the road in the toed-in position, which continued and built up the cycle’, Olley, Chassis Design: Principles and Analysis p. 612.

This action happens particularly if the vehicle has poorly damped and/or too soft suspension long travel suspension. (It is an issue with early OKAs and beam axled 4WDs with worn dampers (shock absorbers).

Such precession is dangerous. If it builds up, the vehicle is unsteerable from a vaguely straight line. Worse, reducing speed (as one must) decreases the tramping frequency but increases the amplitude – to the extent of occasional breakages.

In the early 1930s General Motors’ Maurice Olley realised that this was only preventable via suspension geometry that ensured steerable wheels to rise and fall vertically, not in the arc forced upon them by a tilting beam axle. This required such wheels to be independently suspended.

The independent suspension concept was not new. It was used on a huge road going steam locomotive in the late 1800s, by Dr Lanchester around 1901, Morgan in 1911, Lancia and Dubonnet in the 1920s. But all did so to reduce unsprung mass and improve the ride, rather than out of fundamental necessity. Olley realised the fuller implications with soft long travel suspension.

Beam axled non-steerable wheels are subject to the same forces but, as they cannot swivel, such forces do not matter. This is why many cars, most trucks and many 4WDs retain beam axle rear suspension.

There is no need for independent suspension on caravans as such. Nor is there any need for long travel soft suspension. Passenger vehicle suspension is dictated almost totally by human physiological constraints. Passenger car suspension and concepts are not usefully extrapolated to non-human carrying trailers.

Caravan and tow vehicle dynamics – the swing of a pendulum

A fifth wheel caravan is like a horizontal pendulum. It pivots predictably from the tow vehicle’s hitch and its bob is the mass of the trailer. Side wind gusts etc may cause the trailer to swing slightly, but the forces are mostly small and quickly self-damped. Drivers are rarely aware of them.

As long as a fifth wheeler’s rear wheels are well back, the weight on the tow vehicle is within that vehicles’ limits, and the hitch is above or in front of the tow vehicle’s axle, negligible lateral forces are imposed by one part to the other. A well-balanced fifth wheeler is stable and unrelated to speed.

A conventional caravan’s action is totally different. In effect is a double pendulum. The upper one (the tow vehicle) in effect pivots from its centre of mass. Its bob is the overhung tow ball. From that bob hangs a second pendulum (the caravan).

double pendulum

Explaining the dynamic behaviour of double pendulums involves double differential equations called Lagrangian Hamiltonians (and way beyond most mortals, including me). For those interested, see https://www.myphysicslab.com/pendulum/double-pendulum-en.html

As the tow vehicle swings to the right, the overhung tow ball swings to the left. As it does, it takes the nose of the caravan with it. Likewise, if the caravan yaws to the left, it swings the rear of the tow vehicle to the right. This direction changing interaction is known technically as a phase change – and is the root cause of (conventional caravan) instability. The greater that hitch overhang, the greater the (undesirable) effect.

The pendulum interaction mainly annoys at low levels. Tyre hysteresis and tow ball friction etc usually dampens the yaw. It typically dies out after two or three cycles. It can also be reduced by friction and other sway dampers.

Yaw is of concern at low speed if it does not die naturally (i.e. without frictional aids) inside a few seconds. It is an indicator of instability. It needs resolving at source (not masked by a friction damper).

Such yaw is of serious concern if severe. If it occurs above a critical speed (specific to each rig and its loading) the yaw may self trigger into a so-called positive feed-back loop fuelled by the rig’s kinetic energy.

If the above occurs, a double pendulum’s behaviour suddenly changes. Instead of being relatively harmless, it irreversibly changes to so-called chaotic behaviour. The action is not random as such, but once ongoing it cannot  realistically be predicted from that before. It is impossible for the driver to correct.

(Musicians and public speakers experience this if their microphone picks up their own sound from the loudspeakers. The output suddenly develops into a full-on yowl. That yowl is fixable only by drastically reducing the volume (akin to braking the caravan) – or breaking the feedback loop by moving back from the loudspeakers (akin to instantly eliminating tow hitch overhang).

A conventional caravan is inherently unstable, but a sanely designed, laden and driven rig is nevertheless a realistically safe combination.

Caravan and tow vehicle dynamics  – critical speed

Both pendulum theory and ongoing controlled trials show beyond doubt that there is a critical speed for any specific combination of tow vehicle and caravan. Once above that speed, yawing above a certain level of acceleration irreversibly escalates into chaotic behaviour and a seriously wrecked caravan.

That critical speed, and degree of yaw acceleration, is directly associated with the van’s length, mass distribution. The vertical and lateral disturbing forces (of pitching and/or yawing) increase as speed increases. The critical speed is also a function of the tow vehicle’s mass relative to the caravan’s mass (and particularly mass distribution), length, hitch overhang, tyre type and size, sidewall stiffness and pressure etc.

It has been conclusively shown that the critical speed is scalable. All of the above (and more) is involved but the longer and more end-heavy the caravan (and the lighter the tow vehicle) the lower the speed at which criticality occurs.

The onset of critical behaviour is sudden. Because of this, forum and campfire suggestions to ‘accelerate to dampen yawing’ is risky except at low speed.

That there is a critical speed does not imply that a rig will jack-knife if that speed is exceeded. If however a rig that is travelling at or above that critical speed encounters a situation (such as a strong side wind gust whilst cornering, or needing a strong evasive swing) it is at risk. Few owners encounter that, and many thus dismiss its possibility.

A demonstration of one the causes (excess rear end mass) can be seen at www.towingstabilitystudies.co.uk/stability-studies-simulator.php

When a caravan yaws, it transfers the yaw force via an overhung hitch to the tow vehicle. The transmitted (phase-reversed) forces are resisted by the inertia (i.e. resistance to movement) of the tow vehicle and by the grip of its tyres.

Below that critical speed, specific to each rig, wind drag and rolling resistance dampens the phase reversed yaw that result in the oscillatory action. The yaw energy is dampened by tyre and tow ball friction (and friction sway limiters). This is assisted if the driver holds the steering wheel steady, eases the accelerator pedal and/or manually brakes the caravan at less than full braking.

Full caravan braking increases the caravan’s tyre slip angles, and thus worsens the yaw. Braking the tow vehicle increases its own slip angles and the yaw may trigger that vehicle into oversteer.

 Caravan and tow vehicle dynamics – cruise control risk

Cruise control will detect the inevitable minor drop in speed when yawing occurs and accelerate the vehicle to restore the set speed. This feeds energy into the already unstable system. It reinforces the feedback loop and increases  yawing forces. Meanwhile the tow vehicles tyres heat up and slip angles increase. . .

Whilst convenient, using cruise control whilst towing a heavy rig at speed is not a clever thing to do.

Caravan and tow vehicle dynamics – wind effects

A further cause of major caravan instability are lateral, diagonal and vortex wind forces from fast moving trucks, particularly those towing one or more trailers and even more so if the truck has a flattish front (rather than bonneted). That bluff front creates an ongoing strong bow wave plus a vortex along its side.

An overtaking caravan rig may experience minor buffeting along most of its length. Then, as the caravan’s tow vehicle approaches the rear of the truck cab, that side vortex will initially cause the the tow vehicle to be drawn toward the truck. As the tow vehicle draws closer to the front of the truck cab it is hit by the strong side-going bow wave. This will cause it swing slightly to the right. The overhung hitch causes the caravan to sway toward the truck- where vortex pulls it in further. This initiates a rapidly developing yaw cycle that is suddenly and strongly enhanced as the caravan is hit by the bow wave – swinging it in the opposite direction. As the overtaking rig is now travelling at a high speed, this is virtually a recipe for jack-knifing.

A generally similar but less common effect occurs when the truck and the caravan rig are approaching each other at speed on narror roads.

For a full explanation of this (from Caldwell Consulting Traffic Engineering Services) refer to ‘Articles from Others’ on this website (or Click here to be directed to it as from 14 August.)

Eectronic stability systems

The various electronic stability systems monitor the caravan’s yaw acceleration. The AL-KO system applies the caravan brakes only when the sensor detects a yaw force exceeding about 0.2 g. The maker sensibly warns that it is only an emergency aid to prevent an accident. It is not a way of enhancing stability.

The US Dexter and Tuson systems apply the caravan’s brakes asymmetrically (i.e. out of phase with the yaw) at lower yaw acceleration levels. As test results are shown at speeds below the critical it is not possible to comment on the ability (except as a yaw reducer) to prevent a catastrophic incident at speeds above the critical speed.

Caravan and tow vehicle dynamics – enhancing rig stability

[This section will be progressively expanded] 

The major factors include everything that affects front/rear tyre slip angles.

Those within owner control include:

Loading and load distribution of the caravan and tow vehicle.

Excess tow ball overhang caused by unnecessary hitch bar extension.

Speed at which the rig is driven (kinetic energy is related to the square of that speed).

Fitting and use of yaw control devices, WDHs etc.

Those outside direct owner control (but subject to choice of rig) include:

Length of the caravan, weight of the caravan.

Weight and stability of the tow vehicle.

Those determined by the caravan builder include:

Distance from tow hitch to axle centre/s.

Mass of the caravan.

Distribution of such mass along the length of the caravan (particularly at its rear).

Centre of mass in both planes.

Height of the roll centre and roll axis (as imposed by the geometry of the caravan’s suspension).

Moment arms about the roll axis, particularly at the far rear.

Magnitude of yaw inertia.

Radius of gyration.

Damping of yaw and roll.

Tyres with good sidewall stability (such as light truck tyres).

Caravan and tow vehicle dynamics – optimising towing stability (summary)

Tow vehicle behaviour is now well understood, both in theory and practice. A long-wheelbase tow vehicle with short rear overhang, and one that weighs more than the trailer, is more stable than one that lacks such attributes.

Reducing caravan perimeter mass and particularly end-weight mass is vital.

It makes sense to house the spare wheel(s) below the chassis and in front of or just behind the axle. Batteries should be located centrally between the axles. Ideally water tanks should be wide but not long, and located as centrally as possible.

Friction devices smooth minor snaking and reduce settling time. Trials show that such (Coulomb friction) devices have no effect on forces beyond that however.  Elastic energy held within such devices may suddenly be released when such devices are overwhelmed and ‘fed into the system’.

Lateral sidewall stiffness of all tyres assists.

The major factor of all however is speed.

Caravan and tow vehicle dynamics – driver reaction

Most big rigs feel stable in normal driving conditions. There is also usually some margin of stability that enables a reasonably experienced driver to cope with scary but not accident-resulting situations.

The major concern with a conventional caravan and tow vehicle is how it will behave in the rare, but sometimes encountered, situations that cause major yaw. These typically include a sudden strong side gust on a motorway, braking hard on a steep winding hill at speed, and an evasive swerve at speed.

It cannot be stressed too strongly (particularly with long end-heavy caravans) that, unless the rig is grossly unbalanced, it is not possible for a driver, no matter how experienced, to assess (by feel) how that rig will behave in an emergency situation.

Long, end-heavy caravans usually feel stable. Short vans are usually adequately stable, but often feel twitchy (particularly if twin axle) at low speed.

These impressions are illusory. Inertia confers initial stability, but that very same inertia is that which may cause it to jack-knife.

That ‘my rig always seemed so stable until that day it suddenly rolled over’ is the most common post-jack-knifing reflection. (Such apparent stability characterises giant container ships, until a rogue wave proves otherwise).

There is increasing belief that the probable safe speed for big rigs (when there is no head wind) is less than 100 km/h. Personally I would not tow any conventional caravan longer than 5 m at over 90 km/h, unless built following the guidelines outlined above. And even then at not over100 km/h even if laws permitted.

I also very strongly advise against the use of cruise control with any rig over five metres until more is known.

The above is a precis of some of the most relevant parts of a major fully referenced article in this area that I have been working on for some years  and hope to complete in late 2016. It is at a higher technical level than this one – hence this hopefully more accessible level.


My constantly updated articles in this area primarily summarise current thinking. They stem from my interest and involvement whilst employed by Vauxhall/Bedford’s Research Dept in 1953, and particularly by the influence of General Motor’s Maurice Olley.

Maurice Olley was born in Yorkshire as the first automobiles were being made. Following a time with Rolls-Royce he moved to the USA to work with the General Motors Research Division, later returning to Vauxhall Motors in Bedfordshire. I was privileged to attend Olley’s lectures during his own time with Vauxhall Motors’ Research Laboratory.

He was a very quiet, totally unassuming and gentle engineer. His work lives on in the 620-page Chassis Design: Principles and Analysis. The book was prepared from Olley’s notes, some 27 years after his death by Milliken and Milliken.

Thanks also to Nick Bosio, P.W.D., and others for checking this article for readability etc. Their suggestions and improvements have been incorporated.

Brief bio

Originally trained as an RAF ground radar engineer, I spent a brief time with de Havilland, before working at the Vauxhall/Bedford Motors Research Test Centre. I moved to Australia in 1963, where I designed and built scientific measuring equipment. In 1971 I founded what, by 1976, became the world’s largest-circulation electronics publication, Electronics Today International.

From 1982 to 1990 I was technology editor of The Bulletin and also Australian Business magazines and in 1999 started my two companies: Caravan and Motorhome Books, and Successful Solar Books.  My brief is here – Biography.

My current books are the all-new Caravan & Motorhome Book, Caravan & Motorhome Electrics, the Camper Trailer Book, Solar That Really Works (for cabins, caravans and motor homes) and Solar Success (for home and property systems)


This part of the website is open to informed engineering comment and contributions. Preference will be given to those using their own name (if not please advise that as no postings will be accepted from those unknown to me). Please note than comments will not appear until vetted and approved – please bear with me as this can only be done from time to time.

This article is Copyright: caravanandmotorhomebooks.com