Why does dynamic rollover keep causing helicopter accidents despite a high theoretical level of awareness of both the situation, and the correct recovery actions?

Dynamic Rollover (DR) is a situation of which, hopefully, all helicopter pilots are aware. It is taught in groundschool and explained in detail during initial rotary training, and the recovery actions are often quizzed during examinations and check-rides. Despite this, we see accidents on a far too frequent basis in which Dynamic  Rollover is cited as the cause. It might be useful to look at why this apparent contradiction may exist. 

Years of experience
After many years of instructing, I believe there are two main reasons for the prevelance of Dynamic Rollover related accidents. The first is fairly straightforward – Dynamic Rollover recovery cannot be practiced effectively in an aircraft. It can be briefed and discussed, but an instructor is very unlikely to put an aircraft into Dynamic Rollover and then hand control to a student to complete the recovery actions, at least not more than once, as the need to take immediate action to prevent a crash precludes this. This leaves a pilot in the position of never having actually practised the actions that would be required immediately and without prevarication in the event they find themselves in this situation.

However, there are many emergencies that cannot realistically be practised in an aircraft, yet Dynamic Rollover seems to be more problematic than so many other flight regimes and is an area where incorrect recovery technique is all too often mentioned in accident reports. Why might this be?

The root cause for the continuation of these accidents is that the recovery actions could be considered counter intuitive

This information allows us to understand why our reflexive response to an undesirable roll – applying opposite cyclic – will be ineffective in the case of Dynamic Rollover. If we imagine an aircraft on a lateral slope, right skid upslope for example, we can see how Dynamic Rollover might occur.

Prior to lifting, the pilot applies right (into slope) cyclic to level the disc. Collective is then applied, and as the left (downslope) skid starts to break the ground and rise, the pilot should move the cyclic left to maintain a level disc attitude.

However, the pilot does not do so, which has the effect of the disc now starting to tilt to the right, towards the slope. Our lift vector is now trying to roll us into the hill. Crucially, the rate of roll, with the upslope skid still in contact with the ground, is controlled primarily by the collective, with its huge control power. The pilot continues to raise the lever to break free from the ground. However, as the disc is not being levelled, this has a two-fold effect, firstly the disc tilts even further towards the hill, increasing the roll tendency by directing the rotor thrust even further towards the slope, and the magnitude of this misdirected thrust is increased by the larger collective input.

As the rate of roll towards the slope becomes excessive, the pilot notices and instinctively applies left (out of slope) cyclic. The problem, as we know, is that the collective, with its huge control power, has generated this rate of roll towards the slope. The cyclic, with its comparatively weak authority, is powerless to overcome it – only the removal of the collective force has any realistic chance of preventing the aircraft from continuing into the slope, hence the correct recovery action of lowering the lever.

This, of course, brings the possibility of lowering the lever excessively, preventing the Dynamic Rollover from occurring, but then lowering the downslope skid onto the ground so rapidly that the aircraft rolls over in the other direction. 

Maintain a level attitude
As we can see from this scenario, the pilot is having to make an unfamiliar control input, which he has likely never practised, in a very short period of time. This is clearly an extremely undesirable situation, and unfortunately the solution is as obvious as it is unhelpful – do not get into this situation in the first place.

When lifting from a slope, or frozen ground, or a surface where a wheel or skid may have become ‘caught’, remember to fly slowly and smoothly, and as the lever is raised, maintain the disc in a level attitude. If the aircraft does not react as you expect, then gently lower the lever and investigate the cause.

Remember that although there are known problem areas where Dynamic Rollover is more likely to occur, it has happened on flat airfield surfaces as a result of mishandling, so it can actually be encountered any time the aircraft is in contact with the ground.

It certainly helps to ‘armchair’ fly this recovery procedure, and the level of simulator fidelity is improving all the time to the point that those with access to such devices can now practise in a far more realistic environment. A mental note of the recovery actions should be taped to the back of the mind of every pilot and refreshed every time operations are likely to include any scenarios in which the factors that may lead to a Dynamic Rollover situation are in existence. 

I hope this piece is of help to my rotary brethren. Safe flight everyone.