What causes a trailer to fishtail on the highway?

[adaviel]

I had a somewhat unnerving experience recently.
I had bought a travel trailer (caravan for UK readers) and was towing it home when it started to sway from side to side. The oscillations built up to an extreme amplitude of 90 degrees, limited by the trailer hitting the side of the tow vehicle. I responded as I would to a rear-wheel skid in snow (steering in the direction the vehicle was traveling) and brought the vehicles to a stop.
Now I want to understand why it happened, why it doesn't happen all the time, and how not to have it happen again.

The tow vehicle was a mid-engine AWD minivan, which handles very well on its own. It weighs about 3800 lbs, has a 113" wheelbase and is rated to tow 3500 lbs (braked). The trailer weighs about 2800lbs and has electric brakes (drum brakes triggered by the brake lights of the tow vehicle). The pivot point is a 2" ball hitch about a foot behind the rear of the van, rigidly secured to the vehicle frame. Basically a normal caravan setup you can see anywhere in the UK (Canadians tend to use large pickups as tow vehicles, usually heavier trailers, and often a gooseneck hitch).

The incident occurred when I was descending a slight incline on straight smooth dry tarmac at about 100kph. There was some wind, but not enough to notice had I been driving the van alone. There had been some swaying earlier - as I understand the motion, the trailer moving from side to side on the road, pivoting around a vertical axis through the hitch as seen from the tow vehicle - but these oscillations died down so I presume there was a damping mechanism in play. I don't have any witnesses, but I believe in the downhill incident the same motion became amplified through some positive feedback mechanism, though the hitch shows evidence of a twisting strain also.

I have since fitted a friction damper (sway control bar) to the hitch, but I still want to understand the problem - this was a professionally built vehicle towing a professionally built trailer within the published specifications, at the speed limit on a good road in good conditions. I've seen countless vehicles with larger trailers traveling significantly faster in similar conditions, with no damper and apparently no problem.

 

[Sirsh]

Is it possible that it is a problem regarding the wheel bearings? that could cause 'speed wobble' and vibration etc. especially at high speeds.

 

[Bystander]

electric brakes

slight incline

Get thee to a mechanic. Now.

 

[DaveC426913]

A little Googling seems to strongly suggest that there is too much weight to the rear of your trailer. It is balancing on its axle.

Going up an incline would definitely reduce hitch weight, exacerbating the problem.

Then again, too much weight on the tongue will also cause sway.

Proper trailer safety requires the tongue be in the right range - seems to be about 12-15% of trailer weight.

Do not take this as advice or instruction. I know little about trailers, I merely have a black belt in Google-Fu (I Googled 'what causes trailer sway').

 

[Bandit127]

It sounds like harmonic oscillation of a similar type that causes motorbike tank slappers.

Some bikes are factory fitted with steering dampers and aftermarket versions are also available.

Following this train of thought, a Google search for "trailer sway damper" returns some links that may be interesting.

I also found a good FAQ by Witter that discusses towing stability.

 

[Ketch22]

I have to jump in with Dave426913. This in part from my engineering and part from the fact that I am a commercial driver and have run our trailers around a lot. The action that you are speaking of is due to a weight imbalance in the trailer. To bring the engineering side in. Anytime that the center of gravity is close or behind the center of pressure on a vehicle moving in a fluid the vehicle becomes unstable. Look at your trailer and your hitch this is a very fine line. The tongue can support a limited amount of weight usually in the realm of 300 -800 lbs. However if the weight expressed as a ratio of 50% of total to remaining becomes greater than 1 the instability increases. To create a stable trailer the weight needs to be close to 55% on the front of the trailer and 45% on the rear. This assures that it inherently tries to follow as opposed to trying to swap ends.

In your case it can be as simple as putting too much stuff in the back of the trailer or carrying additional cargo near the rear. If you have a load leveling system use it. A similar result in the towing vehicle can be obtained by shifting weight to reduce tongue weight. This is diametrically opposed to the transfer of weight by dynamic tension achieved by using the load leveling and maintaining appropriate tongue weight.

 

[Ranger Mike]

Having towed various race car trailers with many types of vehicle over the years this is a bad situation.

The tow behind trailer's big weak spot is the point that it is hitched to the tow vehicle. It is at a point four or five feet behind the axle. This means the trailer has the necessary leverage to move the tow vehicle's rear to one side or the other, which has the effect of steering the tow vehicle. This steering effect can go into oscillations, big time and this trailer sway is dangerous.

This fishtailing can be controlled in most cases. Accelerating quickly on a straight road can bring it out and then gradually slow down. Hitting the trailer brakes (NOT THE TOW VEHICLE BRAKES) can bring it out of a fish tail.

The friction sway control attachment is a must if you have semi tractor trailers coming at you on a narrow two lane road.

Worst thing you can do is hit the tow vehicle brakes in a panic.

Strong crosswinds may tend to push the tow vehicle/trailer combination laterally, and it may end up wandering out of the traffic lane a bit if the driver isn't paying close attention. But steering should be predictable, and the driver should be able to use corrective steering measures without fear of sway. Likewise, it should be possible to drive a mountain road aggressively while being able to keep the tow vehicle in the proper position on curves.

Speeding 18-wheelers present hazards to conventionally hitched trailers that don't handle well, particularly while descending mountain grades. A tow vehicle/trailer rig is most susceptible to destabilizing forces while descending a grade at highway speeds, and such conditions are the true test of inherent stability. It's natural for the bow wave (air pressure) of a speeding 18-wheeler to have an effect on a tow vehicle and trailer - an effect that requires steering correction. But the effect should not be destabilization that makes the tow vehicle feel like steering control is minimal and therefore unpredictable.

Sway is a fishtailing motion of the trailer, caused by external forces that set the trailer's mass into lateral motion with the trailer's wheels serving as the axis or pivot point. The motion is a sideways seesaw. All conventionally hitched travel trailers will sway slightly in response to crosswinds or the bow wave of an 18-wheeler semi tractor trailer rig passing you either coming or going. A poorly balanced trailer will continue to sway after the force that caused the instability has ceased. In a very bad balanced trailer, the sway motion may increase until control is lost. Beware - evaluations of sway problems focus on the hitch or the tow vehicle, CAUTION - the trailer's weight distribution often is the primary cause.

Sway is a fishtailing motion of the trailer, caused by external forces that set the trailer's mass into lateral motion with the trailer's wheels serving as the axis or pivot point. The motion is a sideways seesaw. All conventionally hitched travel trailers will sway slightly in response to crosswinds or the bow wave of an 18-wheeler overtaking from the rear. The good ones will need little correction by the driver and will quickly destabilize. Only poorly set-up trailers will continue to sway after the force that caused the instability has ceased. In fact, in poorly balanced trailers, the sway motion may increase until control is lost. Unfortunately, most evaluations of sway problems focus on the hitch or the tow vehicle, but the trailer's weight distribution often is the primary cause.

All trailers are designed to be stable based on the amount of weight in front of the axles vs. the amount of weight behind. The difference between these two weight masses is the amount of weight on the trailers hitch, which is called the hitch weight (trailer tongue weight).

Trailers with insufficient hitch weight have two deficiencies: The percentage of weight (mass) behind the axle(s) is too high, so when set in motion it acts as a pendulum; and the distance between the hitch ball and the trailer axles is insufficient.

Simply stated, trailers with a high proportion of hitch weight to gross weight usually have more of their length ahead of the trailer axles, and they handle better. The generally accepted industry standard is that hitch weight should be approximately 10 percent of gross weight. In fact, that is a bare minimum, and some trailers with 10 percent hitch weight do not handle well. Hitch weights of 12 percent or higher (up to the weight limits of the hitch and vehicle beings used) assure proper handling.

One big problem is the inherent stability of the tow vehicle, which is yet another variable. A truck or van with a long wheelbase, a relatively short rear overhang and stiff springs often will at least partially make up for a trailer's lack of inherent stability, whereas if the trailer is towed by a softly sprung vehicle with a long overhang, the trailer's shortcomings will be more obvious.

What ever you are towing with, find out your correct balance and weight. Find a set of commercial scales at the local moving and storage firm, or grain elevators. Gross weight and hitch weight should be recorded with the trailer loaded for travel. Gross weight is recorded with the trailer unhitched on the scale.

Hitch weight is determined by recording two trailer weights. For the first, weigh the trailer, unhitched, on the scale.

For the second, position the tongue jack off the scale (trailer unhitched and tongue height same as when towing) to weigh only the trailer wheels. Subtract the two figures for hitch weight. If hitch-weight percentage is 10 percent or less, it can cause unstable trailer behavior. If hitch weight is 10 to 12 percent, towing stability still could be a problem if the tow vehicle is marginally stable. If hitch weight is 12 to 15 percent, the trailer should handle well and should not be a contributor to any instability problem.

It is important that hitch weight not exceed the rating of the equipment.

One more point- I can not over emphasize tire and air pressure .Check ALL tires for correct air pressure. Tow trailer tires are extremely susceptible to UVs. Get tire covers that slip over the trailer tires. These tires can develop side wall cracks almost over night. Do not trust cracked side walls..danger.. DANGER. Thing two is trailer brake maintenance. Grease the wheel bearing and check the brakes every 10,000 miles towing.Make sure tow vehicle has heavy duty shocks ( dampers).

BTW, the fifth-wheel hitch or goose neck type hitch are both is centered over the axle, unable to move laterally, which makes the fifth wheel trailer virtually immune to this motion.

 

[Baluncore]

The way to understand the mode of oscillation is as two masses, (tow vehicle and trailer), connected by a two bar linkage with a single pin joint at the centre, that represents the tow point. When the system is under tension it is stable, but when the tow vehicle brakes, remaining straight has become the unstable state, compression of the linkages alternates between the two stable states on either side. That is why braking aggravates the situation.

Notice that for each cycle of sway there are two cycles of effective length change of the linkage between the masses. That makes a frequency doubler. Once the amplitude reaches a certain point the mode flips to a chaotic situation with a period change. The pumping due to the cyclic change in effective linkage length then gains a vertical component and lifts the front wheels of the tow vehicle off the road, which results in a jackknife.

The best way to minimise the problem is to have brakes on the trailer that can keep the linkage under tension. That requires sensitive hydraulic over-run brakes, or electric brakes. The oscillation will also be minimised by moving the tow point as close as possible to the rear axle, and by using a longer trailer bar. The problem should disappear completely if the tow point could be located at the centre of mass of the tow vehicle. In that situation, a change in the angle of the linkage does not result in a change in the distance between the centres of mass of the tow vehicle and the trailer.

 

[adaviel]

This fishtailing can be controlled in most cases. Accelerating quickly on a straight road can bring it out and then gradually slow down. Hitting the trailer brakes (NOT THE TOW VEHICLE BRAKES) can bring it out of a fish tail.

Thanks Mike (and everyone). More research needed - I'll try and find a scale to check axle and tongue weights. My memory of those few seconds is a bit fuzzy - I don't think I braked strongly, but maybe a bit, and tried to steer out of it. I didn't know about using the trailer brakes manually, or accelerating - that's counterintuitive, but may well have worked given that it was a straight road which leveled out.

Researching online on my phone after the incident, I found an article suggesting that a down slope, wind, and low tire pressures were confounding factors. I didn't have significant cargo in either the tow vehicle or the trailer, but the previous owner had stored fairly large wooden jack supports in the rear locker. The tow vehicle has coil springs (another possibly confounding factor) and "normal" shock absorbers (about critically damped, not stiff).

I had previously towed a smaller trailer some 2000km with the same vehicle over the same roads at the same speeds (or faster) with no problems, but I don't have a lot of experience with different trailers and this was definitely heavier with a larger cross-section (higher area to mass ratio).
In casual observation at campgrounds I have never seen anyone using a sway control device, but as I say, this is Canada with typically larger tow vehicles and I may have missed some.

 

[Baluncore]

adaviel, you are lucky to have only experienced the onset of the process. You did not experience the violent forward–backward oscillation of the combination that doubles the oscillation frequency at the entry to unrecoverable chaos. It is shortly after that doubling that you find out if you will leave the road backwards by the near shoulder, or backwards through the oncoming traffic lane.

You avoided catastrophe because at the onset of oscillation, your initial steering counteraction skill was too fast, and your braking reaction was luckily too slow. It is not humanly possible to out-steer or out-brake the oscillation once the doubling of frequency occurs as the wheels are no longer in a stable frictional contact with the road.

I have only been through the full process once, it will always remain in my mind. I went down a hill that had a bend where the hill steepened. The road then crossed a bridge before climbing again. It did a jackknife U-turn on the two lane bridge with the front shoulder of the trailer tearing the armco rails from both sides of the bridge. Tearing out the posts and bending the armco is where the energy went. I had to drive back up the hill to turn around and then proceed much more carefully on my way. It took several days for the repair gang to fix the bridge. They never did find out who the very lucky driver was that left the long rubber snake marks on the road and demolished both the bridge railings.

 

[adaviel]

I weighed everything, and got the following:
Tow vehicle unladen, front 1020kg rear 890kg
Plus trailer, front 980kg rear 1030kg, trailer 1180kg tongue 90kg. That's within specs, but a bit low on the tongue. I think the previous owner shifted the axle slightly when adding a raise kit, but that's only worth some 20kg.
Something like this: tow.rpod.png

I found some studies by J Darling at the University of Bath, viz.
An experimental investigation of car—trailer high-speed stability
and a paper "the dynamics of towed vehicles" by Christopher J Killer.
I haven't been though all the math yet, but some of it requires measuring or calculating the moment of inertia of the trailer which isn't so simple. Concentrating the mass near the axle to reduce moment of inertia raises the speed when instability sets in, but my trailer is not so bad - most of the heavy items like popout and bathroom are in the centre and there wasn't a lot of cargo in the rear locker or elsewhere.
Darling's paper has experimental results showing that nose weight, tire pressure and moment of inertial all influence the stability, but appear to show that instability is inevitable at higher speed as the damping factor becomes negative.
I'd still like to be able to calculate, or safely measure, the damping factor to see just how fast I can go with a given load and/or damper setting.

 

[Ketch22]

Thanks for getting some values for us. You are right 90 Kg is light for the tongue. If the trailer empty weight was 1180 kg, I would look for it to be stable at closer to 220Kg. Even more to the point is when the trailer is properly balanced as it is loaded the balance must be maintained. Working from your numbers one could see that it would only take a small percentage of the load weight to counterbalance the tongue and start to make it unstable again.

 

[adaviel]

I've been trying to instrument it, but haven't got anything I can publish yet - not enough resolution in the hitch sensor.
I think Darling's work showed the tongue weight effect as unexpected - more mass at the front of the trailer should increase the yaw inertia (bad) and be similar to more mass in the rear of the tow vehicle in terms of axle weight, but it might add damping (good) through straight friction on the hitch ball.
The previous owner denies moving the axle and says he was careful just to lift it. It may have been bad to start with; a friend who worked in the industry says half the units have some problems right from the factory. Anyhow, minor effect.
I'm moderately sure I can be safe in the future - I inflated the tires, replaced the rear shocks on the tow vehicle with new stiffer ones, moved the spare wheel from the rear of the trailer to the front to increase tongue weight, and added a commercial friction damper so if I stay under the "experimentally determined" critical speed I should be OK... .
Darling's work says that a friction damper is effective at controlling low-amplitude sway but not higher-amplitude oscillations, so might actually make things worse if a wind gust triggers a sudden excursion with no precursors to warn the driver. I've fitted a viscous damper in parallel, but given the geometry of the commercial hitch adapter (second ball) I suspect it's not nearly stiff enough. It should, however, respond more strongly to more rapid movement.

I'd still like to understand the phenomenon better, even in simple terms. Where's the energy coming from to drive the oscillations in unstable mode, for instance ? From the vehicle's motion forwards, clearly, but what's pumping the system in phase ? My hands on the steering wheel ?

 

[insightful]

The pivot point is a 2" ball hitch about a foot behind the rear of the van, rigidly secured to the vehicle frame.

That one foot spacing might also be contributing. Think of how the trailer wheels "steer" as the hitch flexes side-to-side. Any lateral looseness or flex at the hitch point originating from the tow vehicle or the trailer is a problem.

 

[Baluncore]

The fundamental cause is compression in the hinged tow linkage.

The energy that drives the oscillations comes from the KE of the system when the tow vehicle brakes. The linkage between the tow vehicle mass and trailer mass has a hinge at the tow ball joint. When under compression that linkage wants to shorten, so it will deflect sideways if possible. The restoring force is the springiness of the tow vehicle tyre sidewalls and the drivers instinct to follow the road. The masses are then thrown apart as the linkage straightens, inertia takes it over-centre and so it deflects to the other side.

The critical speed without tow vehicle braking is when a misalignment of the tow vehicle causes a slight braking effect and so, a push from the trailer. That may be amplified by the driver braking in panic. Do not panic. Do not brake. Keep some power on to maintain a slight tension in the linkage while allowing the combination to gradually reduce speed. Yes, you may have to accelerate down a hill in the anticipation that you can control it on the next climb.

A tandem axle trailer can be arranged to have more braking effect when deviated sideways than the tow vehicle. That can make the combination stable, independent of speed.
Both moving the trailer wheels back and lengthening the trailer tongue will significantly reduce, but not eliminate the problem by increasing the critical speed.
Also, fit brakes to the trailer so you can decelerate the combination while maintaining linkage tension while descending a hill.

Remember, you must keep the tow linkage under tension.