CTT National Council - Helmets on the Agenda

[SilverBadge]

You seem to be deliberately avoiding the point. And that is that the sliding you claim is evidence of no horizontal deceleration of the head is no such thing, Its just a consequence of the body's KE.

As you say, 100+g of deceleration is less than 5ms for 12mph to zero, so unless the rider comes to a longitudinal halt in under a tenth of a second the longitudinal deceleration is considerably lower in magnitude and adds very little to the resultant. Or are you suggesting there is a spike in deceleration on impact followed by a reacceleration from the torso's KE? At those levels of force the neck effectively decouples the head from the torso, and if it didn't, helmets would have a far greater task in providing the required controlled deceleration.

All speculation as helmets are not specified or tested for higher speeds and how they behave and what the consequences are is unknown. But rather irrelevant to the point that in assessing the impact when falling off the bike the horizontal velocity is a significant factor.

TRL were once again very guarded in their words on this subject. The soundbite is that a helmeted head is equivalent to a bare head dropped from just one quarter the height, an asterisk and some small print explaining that this applies up to helmet 1.5m/12mph, bare head 0.375m/6mph. Now, since they've done this work, it was a real missed opportunity to demonstrate how good helmets are, debunk the argument that most serious injuries are in circumstances more taxing than the 12mph "simple fall" by testing helmets higher heights (I've seen the test labs, they have high enough ceilings) for higher speeds. Which tells me they either guessed the result wouldn't be pretty and ducked the issue, or performed the tests and didn't publish the results because they weren't pretty. It speaks volumes that some helmets need different versions to pass CPSC and CE 1078.

[Spinners]

On this issue, I'm against compulsion so I'm glad it got rejected.

[TonyR]

As you say, 100+g of deceleration is less than 5ms for 12mph to zero, so unless the rider comes to a longitudinal halt in under a tenth of a second the longitudinal deceleration is considerably lower in magnitude and adds very little to the resultant. Or are you suggesting there is a spike in deceleration on impact followed by a reacceleration from the torso's KE?

The latter. There will be a momentary spike in the vertical impact caused by horizontal velocity because of the square law I explained. It may sound counter-intuitive but its a consequence of resolving the velocity vector squared rather than squaring the resolved velocity vector. As I said the analogue is a side wind. Intuitively one would think a wind blowing from the side (or even slightly from behind) would not slow you down because it has no velocity vector in the direction you are travelling but because drag is proportional to the velocity squared, the component of drag in the direction you are travelling is increased by a side wind.

TRL were once again very guarded in their words on this subject. The soundbite is that a helmeted head is equivalent to a bare head dropped from just one quarter the height, an asterisk and some small print explaining that this applies up to helmet 1.5m/12mph, bare head 0.375m/6mph. Now, since they've done this work, it was a real missed opportunity to demonstrate how good helmets are, debunk the argument that most serious injuries are in circumstances more taxing than the 12mph "simple fall" by testing helmets higher heights (I've seen the test labs, they have high enough ceilings) for higher speeds. Which tells me they either guessed the result wouldn't be pretty and ducked the issue, or performed the tests and didn't publish the results because they weren't pretty. It speaks volumes that some helmets need different versions to pass CPSC and CE 1078.

I don't think there has been published testing of helmet beyond their specification and the current standards are self certified by the manufacturers anyway but the head of the main UK testing lab is on record as saying that many of the helmets they test struggle to meet the test standard implying that most would definitely fail at a smidgen greater height than the test.

The head equivalent fall height is also an interesting one. Apart from the coup type of injury with a direct impact on the head being relatively benign and skull fracture taking something like 7-10 times the impact that a helmet is designed for (which makes sense with evolution designing the skull to withstand the most common impact of the head hitting the deck from head height in a trip or fall), it has also been found that the skull is in fact far from rigid in such an impact and will flex and bend on impact and bounce back. This has raised concerns that the rigid helmet shell holds the skull and prevents that flex, effectively disabling the protective mechanism nature has evolved for such impacts.

[pjclinch]

I think his head would very possibly be a damn site worse off in that situation than if there wasn't a 20 cm vertical kerb there. Are you suggesting otherwise?

You seem to be deliberately avoiding the point. And that is that the sliding you claim is evidence of no horizontal deceleration of the head is no such thing, Its just a consequence of the body's KE.

The point is that with the kerb the head is constrained much more definitely in the horizontal plane so a lot more deceleration takes place, and that's Not Good. That some KE is subsequently added back by the body rolling over the top is something of a moot point, because we've already damaged the head more than if the kerb wasn't there. Which illustrates that it isn't the total impact KE that matters, but how much you lose and how quickly. Putting a 20 cm kerb there makes the head stop suddenly and then start again, rather than a continuous roll. If it were all about total impact KE then the kerb wouldn't make any difference.

All speculation as helmets are not specified or tested for higher speeds and how they behave and what the consequences are is unknown. But rather irrelevant to the point that in assessing the impact when falling off the bike the horizontal velocity is a significant factor.

The horizontal is vastly less important than you make out, as noted above. But saying all speculation about > 12 mph impacts is silly is, errrr, silly. The 12 mph is working backwards from the vertical fall, and what really matters is not that speed figure but how much and fast KE is scrubbed (if you hit a constrained anvil in a crash test, the speed will be 12 mph. If you hit something unconstrained it will be more). If I hit a sheet of paper at 20 mph we can safely predict I'll be a lot better off than if I hit a very solid wall at 12, because I won't actually lose very much of the 20 mph and consequently not much of the total impact KE needs to be absorbed.

But it's precisely the fatalities and serious injuries that are highlighted by DfT, BHIT, Brake etc as the reason everyone should always wear a helmet except when not cycling. DfT blurb talks about ksi then states helmets are effective across a broad range of injuries. That's about as honest as pointing out that wearing trousers reduces leg injuries compared to shorts and then talking about nothing less severe than falling down a ravine. If helmets aren't putting a measureable dent into the number and severity of hospital admissions then their effectiveness is very limited indeed. And in theory there are fatal impacts that can be reduced to no more than concussion with CE1078.

I think we're largely in pretty furious agreement here. The population studies show us that for A to B cycling helmets make no useful dent in serious injuries across all riders, and methodologically the case control stuff is so full of holes it puts a sieve to shame. However, while I'm very wary of duff information to encourage us to wear something we don't need I'm also wary that duff information the other way (like it's useless to speculate of them being any use if the rider is going over 12 mph, as if an MTBer doing 20 knocking an overhanging branch with his head can't possibly benefit from wearing a helmet) is counter-productive. What we can really tell from the population studies is very useful on the one hand but actually pretty limited on the other. It's important not to over-egg the pudding in the case against forcing us in to them.

Pete.

[drossall]

If our notional TT-ist is doing 40 mph along the way, unless they hit a fundamental constraint like a solid wall they're not actually going to decelerate from 40 to 0 in effectively damn-all space, but they will still decelerate from 12 to 0 in effectively damn all space. So in a lot of cases saying a helmet is useless over 12 mph is misunderstanding the issue.

That is an oft quoted claim but a misunderstanding of the physics. You can't just resolve it into the vertical component of velocity and then square it because impact is proportional to the square of the velocity. You need to calculate the velocity vector, square it and then resolve it into the vertical direction not resolve the velocity into the vertical direction and then square it... Taking your figures of 12mph vertical and 40mph horizontal the velocity squared is (144 + 1600). So the vertical component of the impact is 12/40ths of 1744 - 3.6 times the 144 you are assuming - the equivalent of 23mph not 12mph.

I'm catching up on this thread, and trying to get my head around this argument.

I'm not sure what you mean by "the vertical component of the impact", because impact is not a defined physical quantity as such, and certainly not a vector one, so I'm not sure what you are calculating. There are impact forces, which are vector, and proportional to the vertical deceleration (or the horizonal deceleration if you are unlucky enough to hit a tree or a wall). There are impact energies - these do depend on the square of velocity.

However, as far as I can see, the cyclist's kinetic energy is dissipated partly by a vertical impact with the ground, and partly by frictional forces acting horizontally. The amount dissipated by the vertical impact does not depend on the cyclist's actual velocity, because it is determined by the vertical velocity, which is always 12mph according to pjclinch's description.

[SilverBadge]

I'm not sure what you mean by "the vertical component of the impact", because impact is not a defined physical quantity as such, and certainly not a vector one, so I'm not sure what you are calculating. There are impact forces, which are vector, and proportional to the vertical deceleration (or the horizonal deceleration if you are unlucky enough to hit a tree or a wall). There are impact energies - these do depend on the square of velocity.

However, as far as I can see, the cyclist's kinetic energy is dissipated partly by a vertical impact with the ground, and partly by frictional forces acting horizontally. The amount dissipated by the vertical impact does not depend on the cyclist's actual velocity, because it is determined by the vertical velocity, which is always 12mph according to pjclinch's description.

The forces and accelerations from impact surely break down into vectors. If the head velocity before impact is forward 40mph, vertical 12mph, at nadir is forward 40mph vertical 0mph, and for simplicity's sake we'll argue that whatever "bounce" is probably lower in acceleration magnitude and maximum rebound velocity than the deceleration, due to hysteresis, then I would argue that peak deceleration vector is the same as for forward velocity 0mph, vertical velocity 12mph. The rate of helmet deformation is governed by the perpendicular velocity.
If there is also a change in longitudinal velocity then yes the two decelerations should be merged into the resultant vector, but I assume the argument is that in practice this hypotenuse is barely longer. I'm guessing that the brain is roughly of radial symmetry so the direction of the vector is not hugely important.
Fall at 0mph = 12mph velocity at 90deg to surface
Fall at 12mph = 17mph velocity at 45deg to surface
Fall at 38mph = 40mph velocity at 17.5deg to surface
All three have the same vertical velocity.

[drossall]

Yes, wouldn't disagree with any of that. Still don't understand the post on which I commented.

[MartinC]

In my view trying to resolve it into vectors is pointless. A cyclist crashing is a chaotic event that happens over a period of time in which many things may happen. The thing that confounds and attemot at calculating 'vectors' is that the body (and bike) will tumble and rotate around it's c of g. This means that different parts of the body will have a rotatational acceleration too. The head is very unfortunataely placed in this respect being relatively heavy and placed as far as possible from the c of g. Also the body isn't goping to behave as a single mass and vectors will need to be calculated for all the parts separately

All that you can say with any certainty is that all of the kinetic and potential enerrgy (assuming no other moving vehicle involved) available will be dissipated through the bike and rider. Trying to predict what absorbs what is confounded by far too many unpredictable and unknown factors affecting the event. The potential energy is a relatively small and fixed amount the kinetic energy a larger amount that will be proportional to the velocity of the rider.

Edit: ...................and that the amount of energy a helmet can absorb is fairly trivial compared to that available.

[TonyR]

I'm not sure what you mean by "the vertical component of the impact", because impact is not a defined physical quantity as such, and certainly not a vector one, so I'm not sure what you are calculating.

Impact seems to be the term usually used with helmets but the normal name for it in physics is Impulse - force x time or force x distance - which is a vector since its the integral of force (which is a vector) over time or distance. And the impulse to bring something to a halt is proportional to the velocity squared and the mass.

[TonyR]

In my view trying to resolve it into vectors is pointless. A cyclist crashing is a chaotic event that happens over a period of time in which many things may happen. The thing that confounds and attemot at calculating 'vectors' is that the body (and bike) will tumble and rotate around it's c of g. This means that different parts of the body will have a rotatational acceleration too. The head is very unfortunataely placed in this respect being relatively heavy and placed as far as possible from the c of g. Also the body isn't goping to behave as a single mass and vectors will need to be calculated for all the parts separately

All that you can say with any certainty is that all of the kinetic and potential enerrgy (assuming no other moving vehicle involved) available will be dissipated through the bike and rider. Trying to predict what absorbs what is confounded by far too many unpredictable and unknown factors affecting the event. The potential energy is a relatively small and fixed amount the kinetic energy a larger amount that will be proportional to the velocity of the rider.

Edit: ...................and that the amount of energy a helmet can absorb is fairly trivial compared to that available.

I don't disagree with that which is also why helmet standards testing which assumes a bodyless head in a nice simple slow linear fall onto a surface is pretty meaningless too.