Caravan Stability

A big THANK YOU to Collyn Rivers who has graciously supplied content for this page.

‘The wild thing flew from left to right and knew not which was which.
And the wild rose grew above him as he lay there in the ditch’

About 20 years ago, the caravan press began increasingly to report incidents of rigs going out of control ‘for no apparent cause’. This article attempts to show what is probably happening. All we are looking at is Newton’s Laws of Motion – and these describe things very well (as long as they travel at less than the speed of light!). 

All trailers pitch fore and aft and snake laterally. They do this because the effects of disturbing forces and resultant energy are added back into the system and reinforced. Once initiated, pitching and snaking may build up strongly and rapidly. If you push a child on a swing only a few times and then stop, friction and wind forces will quickly slow that swing. But if you continue to give even a little push at the right time the swing (oscillation) will be caused to grow in size (amplitude). After a few swings, the movement will be difficult to stop.

Such instabilities are found in all dynamic systems. Cars, aircraft, sailing boats, PA systems, economies – all tend to behave this way. Fixes include ‘inverting’ part of the disturbance and feeding it back so that it opposes the unwanted effect. This is how hi-fi amplifiers reduce distortion.

With trailers, inverting and feeding back energy is less practicable. It can be done to some extent but requires rods and springs and shock absorbers mounted in all sorts of places – and some extremely sticky tyres.

Instead then, a trailer’s snaking and pitching loads are handled by dissipating the pitching and snaking energy, in the form of heat, via the towing vehicle’s shock absorbers and tyres (and to a lesser extent by those of the trailer). A lesser contribution is also made via inter-leaf friction in leaf-sprung vehicles. This energy is transferred, via the towbar (and equalising hitch) to the towing vehicle.

Mass and Weight

At this point we need to bring in the effects of mass and weight. Mass is the amount of material in an object. Weight is the measure of that object’s mass and is expressed in kilograms. As Newton pointed out a long time ago, things that have mass have inertia; ie they dislike any change in motion. Kick a brick accidentally and you’ll grasp the concept very fast.

If a rod with a 5kg weight at either end is rocked to simulate pitching and/or swaying, it will be much harder to stop it moving than a bare rod of the same weight and length. This resistance to change in movement is called ‘rotational inertia’ – or in correct engineering terms ‘moment of inertia’. The rod with the weights on the ends has a larger moment of inertia than the bare rod of the same weight.

But if we now have a 5-kg weight at one end and a 4-kg weight at the other (see below), the rod will now rotate around a point away from the centre of the rod because its centre of mass is closer to the heavier weight.

Caravans are deliberately made front-heavy so that the centre of mass will be ahead of the van’s axle/s, and this alone assists in inhibiting the build-up of undesirable pitching and snaking motions.

A caravan can be perceived as somewhere between a rod on wheels and a barbell on wheels, but with its centre of mass hopefully in front of the rear axle. In other words if some ball weight is present the van is front heavy.

There is also a leverage effect (particularly in the case of the motorcycle slung on the rear of a long caravan seen passing through Broome last year). What may not necessarily be realised however is that this effect is not linear.


A thing’s resistance to change of motion (called inertia and, in this case, rotational inertia) is proportional to the square of its distance from its centre of mass. And it’s this square law that can wreak havoc with caravan stability.

A van with a kitchen at the front, a double bed at the back, and two spare wheels on the back bumper, has a hugely larger moment of inertia than a van with a centre kitchen (and no spare wheels on the back). Its mass distribution is well forward and aft of its centre of mass. That distributed mass will better resist the onset of pitching and swaying, but like a very heavy man on a swing, once it does start to pitch and snake, there’s an awful lot more energy to be transferred to and dissipated by the towing vehicle.

There may well be too much because there’s a very finite limit to the amount of energy that can be absorbed and dissipated. If the former exceeds the latter, the rig is likely to jack-knife.

Worse, it’s perfectly possible to have a van that is nose heavy when at rest, but rear heavy when pitching and snaking. Further, once the van starts pitching and snaking the forces on the towball change dramatically!

Up and Down

At rest you may have a 100-kg weight (1000 newtons) pushing down on the towball. But during pitching that force may change to the equivalent of 500 kg (5000 newtons) pulling the back of the towing vehicle upward by that towball.

A second or so later the van pushes the back of the car down again with a force equivalent to 500 kg. This may well be over and above the static towball loading.

Stabilising hitch or not – it’s no wonder the steering feels light. It feels like that because a large amount of weight is being periodically lifted off the front wheels – at the very time when you need it there most.

The towing vehicle has its centre of mass in the centre of the car usually a bit behind the front seats. Energy is being transferred into this vehicle, via the towball, a metre and a half or so behind that vehicle’s rear axle, causing it in turn to pitch and snake about its own centre of mass.

So what we are looking at now is a coupled system within which each part is different, frequently changing, but interrelating.

Meantime the only realistic mechanism for dissipating pitching energy is by the shock absorbers (and to a lesser extent the tyres) turning it into heat. There’s even less ability to dissipate snaking energy.

So we what have is the car and the van wobbling about as a coupled system with its combined centre of mass somewhere around the towball (which is not clever). But worse, this ‘combined’ centre of mass is itself moving all over the place (This, Fletch suggests, with considerable academic restraint, is ‘A VERY BAD THING’).

The motion of a pitching ‘van acts on the overhang of the towing vehicle. It rocks and rotates that vehicle about centre of mass, alternately adding to and taking away from the down force on its front tyres. Snaking pushes the rear overhang of the towing vehicle sideways, but because the force is applied behind the towing vehicle’s centre of mass it attempts to spin the car in the opposite direction around that centre.

If the snaking forces are greater than the vehicle tyres can react, front and/or rear wheels will slide sideways (in opposite directions as the car is rotated uncontrollably about its centre of mass). The rig is then all set to jack-knife.

Thus, because forces increase with the square of the distance from the van’s centre of mass, weight distribution in a caravan is far more critical than generally thought.

As noted above, a ‘van may be front-heavy from static measurement, but is perfectly capable of giving a huge upward pull on the towball once pitching or swaying is set off by road irregularities, wind gusts or driver error. If these effects are severe, spare wheels and/or a toolbox on the caravan’s back bumper can act as that ‘last straw that broke the camel’s back’ – except it’s not a camel (this time) that suffers.

Internal forces are also increased on the van’s structure.

Put a piece of plasticine or clay on a drinking straw a centimetre or two from your fingers and wobble it about to simulate pitching and snaking. Then move it a few more centimetres from your fingers etc. Sooner or later (usually sooner) the straw will break. So may your ‘van if you do things like that to it.

Wind forces too may be involved. Locating the axle behind the centre of mass enhances stability. But if a caravan’s side area is substantially greater forward of the axle/s than behind, side-winds will attempt to rotate it clockwise (as seen from above). Caravans like this are normally rock steady, but strong, cyclically gusting winds can induce seriously dangerous snaking.

Sailors know this effect: wind gusts cause a yacht to turn into or away from that wind. This is controllable by varying the fore/aft sail areas. And, as with caravan axle positioning, gross examples are caused by incorrect fore/aft positioning of the centre of (wind) pressure relative to the centre of mass.

Stabilising Hitches

Acting as a semi-flexible beam between car and trailer, a stabilising hitch partially compensates for rear overhang by transferring some of the imposed weight back onto the towing vehicle’s front wheels and to the trailer wheels. These hitches help compensate for weight distribution on the tyres of the van and the car, ie they help keep the front wheels on the ground: and that cannot help but be useful.

Off-road however they are a menace. By transferring weight to the towing vehicle’s front wheels, and the trailer wheels, they may (in some conditions) reduce the weight on the towing vehicle’s rear wheels – and thus cause it to lose traction.

Anti-snaking Mechanisms

Anti-snaking mechanisms assist in restraining minor snaking. They assist in preventing it building up, but they have their limits because they have little ability to dissipate the snaking energy.

Ultimately snaking must be restrained by the grip of tyres on the road. The front tyres have only a tenuous grip (because there may be the equivalent of 500-kg or more pushing down on the towball and lifting the front). If the van has begun to snake (because it’s been hit by a bullet of side wind), and the road is wet at the time – you’ve may have problems that no anti-snaking device in the world can fix.

We saw a not dissimilar phenomenon with onset of radial ply tyres. These restrain ‘skidding’ more effectively than do cross-ply tyres. But because more energy is built up before they ‘let go’. And because the onset of ‘letting go’ is more sudden, when a radial ply tyre loses grip it does so suddenly at higher speeds. The resultant effects (like hitting a tree backwards) tend to be more severe.

I am not opposed to these devices because they may enable snaking to be ‘caught’ before it has a chance to build up. But I feel that claims of emulating ‘fifth-wheel stability’ are hard to accept without scientific evidence.

In practice, few conventional rigs have major instability problems, but RV magazines worldwide nevertheless do carry reports of cars and vans going out of control (usually jack-knifing) for ‘inexplicable reasons’. But a gross imbalance of the centre of mass, and the mass distribution seem frequently involved.

In the few cases where it’s possible to obtain comment, the most frequent observation is that ‘the rig felt so stable up to that time’. But so may billy carts and bicycles just before you screw up.

Rear Overhang Contributes

A modern vehicle’s rear overhang is a major contributing factor. Imposing the mass of a heavy trailer a metre and a half behind an often-swaying vehicle’s rear axle is fundamentally unsound. Cars are made this way because people want more carrying capacity at the rear – but coupling a van that far behind the rear axle is not a sound engineering solution. As a matter of interest, even with front wheel drive the DS Citroen (the Golden Goddess) was claimed to be a very stable towing vehicle because of its negligible rear overhang.

To experience this effect, hold a pencil (lightly) about two and a half centimetres from its tip to simulate rear axle position. Pushing the tip gently up and down and from side to side partially illustrates the effect on the towing vehicle. This is not an exact parallel, as it does not involve the car’s centre of mass – but it gives a rough idea.

With a heavy towing vehicle and a relatively light caravan, with mass distributed close to the centre of mass, there’s no cause for alarm. But I for one would not consider an end kitchen – and unless it was built of aluminium.

For big vans the fifth-wheel approach simply has to be a fundamentally better solution. Here, the towed vehicle is attached just ahead of the towing vehicle’s rear axle where the loads are reacted far more effectively. That is the distance between the centre of mass of the towing vehicle and the hitch does not nearly have as much effect on rotating the towing vehicle when the trailer pitches and snakes. Partially aided by a wider choice of dual cab towing vehicles, Americans are moving to the fifth-wheel concept in droves.

To simulate the effect of fifth wheel attachment, compared with rear overhang, repeat the pencil experiment, but push it sideways only a millimetre or two in front of where you are holding it. It’s far steadier.

Negligible quantitative research appears to have been done into the effects of the dynamic behaviour of mass distribution and on towbar loading (except a little by the military). This currently precludes reliable alternatives to the current and generally safe ‘10% on the ball’. But it’s probable that vans with mass distributed closer to the centre of mass could safely get by with far less than 10%, and that many combinations are rejected that might in fact be dynamically safe. 

But if quantitative research is lacking, some of us have done some qualitative research at some point – using descriptors (as Fletch points out) such as: “Oh-Sh**!” as we get a close-up but often brief side view of the van out of the front driver’s window.

In conclusion I’d like to thank Fletch and Tony W. for their totally invaluable assistance with this article. I could more or less visualise the effects of what I was trying to describe, but could not have put it into correct engineering/physics terms without them.

Whilst copyright, this article may be reproduced subject to the following acknowledgment. Copyright, Collyn Rivers, Caravan and Motorhome Books, Broome.


Share this Page