The John Miles method of cornering.

[Tosh]

Hi all,

I've been reading John Miles' book Expert Driving The Police Way. It is a fantastic book with a lot of very useful information and his explanations make it very easy to understand. Even though it was written in 1970 a lot of the text is still relevant to driving today.

I'm sure you know this already but John Miles was a Class 1 instructor at Hendon and taught Chris Gilbert so I'll not be one to discard his teachings lightly. My point is that in his book in Chapter 9 on cornering he gives some very specific advice on how to position and to use what is now referred to as the 'late apex' method and outlines in detail as why this is the safest and most efficient method of getting around that bend. He then goes into more detail about the 'constant radius' method' and citing reasons not to use this technique as it promotes early entry and has a line which runs wide towards oncoming traffic on the exit of the bend. Now, I know that it is no longer mentioned in Roadcraft but in the Blue Book (1977) and right up to the Yellow book (1997) the diagrams show what is the line of constant radius and no mention of taking a later, sharper line to apex late in the bend.

Considering that Roadcraft is the training manual for the Police driver and is written in consultation with Police training staff I'm a bit perplexed as this supposed good information is omitted and 'bad advice' included.

Who am I to believe or am I supposed to now follow the Roadcraft teachings of today and not have this information at all?

[TripleS]

I remember, several years ago, being advised on a drive with a very senior member of HPC to ask myself the question "on the exit of a left-hand bend, where's the nose of my car pointing?" If the answer was "towards the centre line", it meant I'd taken too early an apex. He clearly agreed with John Miles.

For what it is worth, my own preference is to do whatever is necessary to avoid running wide on the exit.

[Tosh]

I've added some scans of the books. The only other place that I seen this information about cornering is in John Lyons' book on advanced driving.

Jonh Miles: Expert Driving the Police way


Roadcraft 1977 (The Blue Book)


Roadcraft 1997 (Yellow cover)



Looking at the last photo it does show a little less than a constant radius curve approach but still no mention of any of what appears to be very sound advice from John Miles. It's a kind of half hearted constant radius approach but one that is still taking an early curve into the bend and potential danger, everything John Miles was advising against decades beforehand. I wonder were trainees at Hendon and other schools told to disregard the advice in Roadcraft and use a safer, faster method of cornering?

[Astraist]

Entering the corner late and touching a very late apex on bends with limited vision offers better vision around the bend and does not limit progress since it allows for earlier acceleration coming out of the bend.

In fact, coming into bends that are particularly blind I sometimes compromise smoothness for vision slightly by turning in very late and a bit more positivelly than otherwise needed for the turn.

However, if there is nothing limiting my view, I would go for the geometric apex as to reduce cornering demands to a minimum, as well as offsiding where possible, which for me means that I am perfectly alone on the road.

Entering and apexing early can also be needed to "shimmy" a long articulated vehicle around a tight and narrow corner. It all depends on the corner...

[fungus]

I think there is a lot to be said for taking a late apex. I know from the diagrams in Roadcraft that the authors seem to be refering to approaching bends from a higher speed than residential speeds, but the principal is still the same.

I can think of one good example in a village near me, where the road is about three car widths wide and there is a 90 degree right hand bend. As you approach the bend there is often a parked car on the left which is just visible as you are about 25 metres from the bend. Vision across the bend is restricted by trees and shrubbery in a garden on the right, so it is not possible to see whether there is a parked vehicle on the right as well. If you take the line of a constant radius, you would be on a collision course with an oncoming vehicle that is passing the parked vehicle on the right. By taking a late radius this is avoided.

[jcochrane]

The technique of the late apex line has many supporters. Despite what some say it is the preferred line in my experience, for bends, rather than fast curves, for many respected racing drivers on and off circuit. Amongst those who use it that I have driven with or talked to are Stirling Moss, Jackie Stewart, Derek Bell, John Surtees Davinia Galicia and many others. (When saying this imagine the width of the track to be one side of the carriageway, the centre line on the road being the edge of the circuit.)

Apart from the racing world there are a vast number of supporters from the road driving fraternity. In my view it is an ideal line with advantages of vision as mentioned by others but is also overall more stable and quicker. It requires a slower entry but achieves a faster exit as the wheels start to straighten earlier with a more stable car and more early acceleration can therefore be applied. The sharper turn in is done at slower speed and last for less time which makes it more stable.

When teaching road driving I refer to it as "selecting a line to clip the back of the apex" At race speed on the circuit though it's more a case of aiming at the apex but as the car drifts it will achieve the same resultant line and clip the back of the apex.

[waremark]

On a circuit or on the road with perfect vision I suspect that the late apex works better in a quicker car, where you can take advantage of being straighter sooner to use more power. In a lower powered vehicle there is more advantage to the constant radius line which allows more speed to be maintained on entry.

John Lyon taught me in the early eighties that it was vital to take a blind left bend in a manner which would allow you to be tight on the nearside on the exit if there was someone towards you in the middle of the road.

[waremark]

In a lower powered vehicle there is more advantage to the constant radius line which allows more speed to be maintained on entry.

The other issue I have with constant apex is that you have to be millisecond precise all the time for it too work. If you don't start the turn in exactly the right position then you'll miss the constant radius apex and thus your line will take you into oncoming traffic. I'm not good enough for that sort of precision 100% of the time, so I advocate some form of late apex line.

I did say 'on a circuit or with perfect vision'. Certainly not appropriate with the possibility of approaching traffic.

[TripleS]

In a lower powered vehicle there is more advantage to the constant radius line which allows more speed to be maintained on entry.

The other issue I have with constant apex is that you have to be millisecond precise all the time for it too work. If you don't start the turn in exactly the right position then you'll miss the constant radius apex and thus your line will take you into oncoming traffic. I'm not good enough for that sort of precision 100% of the time, so I advocate some form of late apex line.

I did say 'on a circuit or with perfect vision'. Certainly not appropriate with the possibility of approaching traffic.

It looks to me as if there is a major difference here.

If we're talking about circuit driving (of which I have no experience) you would presumably aim to be close to the limit of grip, in which case sooner or later you will exceed it and go off, harmlessly one hopes. On the road you can't push things to that degree. Therefore you should have a margin to allow you to compensate for misjudgements, and tighten the line so that no real harm is done.

[waremark]

If we are onto circuit driving now ....

Some of us have many times driven at Bruntingthorpe, an airfield proving ground. There is a very fast (say 120 mph in a sports car) and very wide left hand bend at the end of the control tower straight. Years ago I saw a nasty accident there when a car went off the outside at the exit of the bend. One of our revered masters demonstrated to me subsequently that if you take a late apex AND lose control you can spin without leaving the tarmac. And I have managed to do exactly the same thing myself on yet another visit - taken a late apex, lost control, spun, and not left the tarmac. No damage done except to my pride.

[stefan einz]

Another vote for taking a late apex here.

On the approach to a closed bend on the public road, it provides for the safer line, minimising the chance of running wide and sustaining vision for longer. As a Lyon graduate, he drummed into me many time on left handers to exit so as to have a wide margin away from oncoming traffic.

On circuit, a late apex, all else being equal, is faster than taking the geometric apex. This was reinforced at a Ferrari Sport driving school earlier this year, both with data logging and by the F1 trained instructors. I am with Dave on this one.

Cheers

[Astraist]

But on the road in an open bend, a geometric apex would be desired as to reduce demands on the car throughout.

Even on the track, fast curves (taken with maintainence throttle throughout) are taken at the geometrical apex.

As I mentioned, on the road even an early apex might sometimes be desirable!

N.B. The late apex on the road is usually far later than the late apex on the track! I often refer to the desired line on the road as the "last apex".

[jcochrane]

Another vote for taking a late apex here.

On the approach to a closed bend on the public road, it provides for the safer line, minimising the chance of running wide and sustaining vision for longer. As a Lyon graduate, he drummed into me many time on left handers to exit so as to have a wide margin away from oncoming traffic.

On circuit, a late apex, all else being equal, is faster than taking the geometric apex. This was reinforced at a Ferrari Sport driving school earlier this year, both with data logging and by the F1 trained instructors. I am with Dave on this one.

Cheers

Very true.
The late apex v constant radius cornering techniques were scientifically researched some years ago. The key findings were that the late apex technique was quicker overall, more stable and less load on the car. The added advantage for road driving is the extra vision gained from positioning for longer on the outside of the bend particularly where the view of the road surface is otherwise limited. Also it makes positioning, as circumstances demand, on a desired exit line possible/easier .

[stefan einz]

But on the road in an open bend, a geometric apex would be desired as to reduce demands on the car throughout.

Even on the track, fast curves (taken with maintainence throttle throughout) are taken at the geometrical apex.

As I mentioned, on the road even an early apex might sometimes be desirable!

N.B. The late apex on the road is usually far later than the late apex on the track! I often refer to the desired line on the road as the "last apex".

On track, that was not the view of the F1 drivers I spoke to at Ferrari. All else being equal (i.e. assuming a completely flat and consistent road surface), they would apex after the geometric apex in all curves. The faster the curve, the closer to the geometric apex they would get, but never quite actually on it.

On the road, there many other factors that have to be taken into account - road surface, where the view is, other traffic, presence of other hazards etc.

But, in general, I prefer to sit next to a driver who is taking a later apex as it reduces the risk of them running wide - so if they are pressing on, I consider it safer - again, ceteris paribus.

Cheers

[TheInsanity1234]

In my (exceedingly) limited experience, I find that I'm more comfortable when I position myself slightly to the right on the approach to a corner, then steer into the corner and try to get the car as straight as possible on the exit, rather than gradually following the curve.

If done at a suitable speed, you'll find that the quicker steering movement required to steer into the corner isn't very noticeable.