Charles Basenga Kiyanda

Going Down Hard: A Bicycle Helmet’s Perspective

I had a little bicycle accident today. Nothing really bad. I was riding back from a meeting, going to my office. This is normally a 15 minutes ride, from my advisor’s office to mine… had it not been for a hole. It appears to have been left by a road crew at some point. It’s a perfectly rectangular hole, blocking about half the shoulder. It’s conveniently located right in the middle of the nice wide shoulder there, exactly where you’d want to ride. To make it even worse, the hole is not very deep and not very wide, so you don’t tend to see it, so you don’t tend to see it far ahead of time. So I went right over it, had a short discussion with gravity and lost. Three cars were nice enough to stop, wondering if I was ok. This was nice. Nothing incredibly bad happened, only annoying.

More importantly, this made me think I should write this post about whether one should wear a bicycle helmet. I had been, for years, a strong supporter of bicycle helmets. Then I read a well formulated opinion which made me doubt bicycle helmets. I ran numbers on the back on an envelope two days ago which are making me reconsider again. I’m now pretty convinced bicycle helmets are indeed useful for most people riding bicycles. The justification below the fold.

First off, let me say I’m fine. My pants are thorn in two places, so they’ll head for the garbage. (To be fair, they’re they were very light, single nylon layer, type of pants, so really not meant for this kind of abuse.) My GPS was clipped on my handlebar and it took a pretty rough beating, but the screen is almost pristine (given the circumstances, nothing to cry about, really more dirty than scratched) and all the buttons and functions seem to work. I’m impressed, it seems these Garmin eTrex units are quite durable. I hope Garmin never starts cheapening out on the toughness of this series. My front brake is a little broken. One metal piece started shearing, so it’s unusable right now, but I don’t expect it to be very expensive. Five to ten dollars maybe. I’m also impressed at my Arkel bag (made in Quebec). It’s a office-type bag with a clip to hook it to a bicycle rack. Very useful if a little expensive. (About 200$CDN.) The bag was on the side that fell towards the pavement and it ended up with some light scuffing on a buckle, but really nothing bad. This bag is marketed as indestructible and I’m starting to believe it. On the personal front, I came away with some nice scratches on the palm of my hands, an extreme example of a rug burn (I’ve got a 3-4 square inches of skin missing on my knee) and a slightly bruised rib on the right, but I don’t think anything is broken really. The worse part was really cleaning the wound with rubbing alcohol. Now, I’m just having wine to weather through the worst part of the pain and tomorrow, I’ll be fine.

Ok. So now onto business. Common knowledgee is that bicycle helmets are useful. Some people out there do say that bicycle helmets really aren’t that useful. To be fair, the arguments are quite well formulated. So, for a start, here they go.

  1. The often quoted bicycle study saying that helmets can prevent 85% of head injuries isn’t all that good.
  2. Bicycle helmets are tested for low velocity impacts, not very representative of the speeds at which adults cycle. They are representative of kids’ riding conditions, so kids should wear them, but the importance for adults is dubious.
  3. Wearing a helmet can actually be detrimental. The sound of the wind going by your ears seems sort of amplified, which means you can’t really hear coming cars.
  4. Wearing a helmet might prevent head injury, but it could increase neck injuries, by giving your head a larger moment arm. So you might save your head, but you might snap your neck.

To be fair, arguments 3 and 4 are not often used by skeptics of bicycle helmets, and when they are, they’re usually described as unproved, slightly kuckoo ideas. So let’s not care too much about 3 and 4. Number 1, I remember looking at and I was slightly convinced. It’s quite possible that the study wasn’t all that conclusive and/or that the conclusions were used a little liberally. As far as this is concerned, I’m really concerned about argument 2.

It helps to know how they test bicycle helmets. Basically (there are some subtle variations, but not too important for this post), they strap the helmet on a 5kg weight and drop it from a height of about 1.5m. Roughly that’s it. The helmet critics/skeptics basically say “Ah! You see, alduts usuallly fall from more than 1.5m and usually have some speed, so they’re rather ineffective. Children on the other hand don’t go very fast and don’t fall from very high, so they’re more effective for kids.” There’s a problem with this. It doesn’t really matter which height your head starts from or what speed you start at, what matters is the velocity at which your head hits something. So what’s this maximum velocity? Easy. Just match the initial potential energy (mass * g * height) with the end kinetic energy (0.5 * m * V^2) and solve for V. Now this assumes that no energy is dissipated in the crash by something else than the helmet. If you go through the calculation, you get a maximum velocity of about 12mph.

That’s not very fast. One would be quick to point out that many riding adults end up going faster than that. But that’s ok. The point is not how fast you’re going, but how fast your head hits the pavement at. So back to my crash today. I was riding along at about 25-27mph just before I fell. This is more than twice the critical speed we calculated. But like I said, the important piece of information is not how fast you go, but how fast your head hits the ground. It turns out that a lot of energy is dissipated by your body slamming into things. It took quite a bit of energy to drag my bike on the floor, break a metal part, twist the gps on it’s holder, rip my pants, bruise my ribs and tear my skin off. Probably enough energy so that had my hit the ground, it would have been moving at less than 12mph and so well protected by the helmet.

But what about head-on collisions with cars? Would you put on a bike helmet and ride a motorcycle at a brick wall at 25mph? and you’d be right, I wouldn’t ram my head in a wall at 25mph wearing only a bike helmet. The trick is, a literally head-on collision with a car probably doesn’t happen that often. For one thing, it’s probably pretty rare that a cyclist will run into a car with both going towards each other. Someone will react and try to get away, so that you end up with a side impact. Someone running into the back of someone else is also probably quite common so that both the car and the bicycle can be going quite fast. What matters is the relative velocity.

Very important though, is that there’s a lot of stuff between you and the car. For one, your car is in the way and will probably be hit first. Then your body will smash into the car. You may break a bone or more. All this takes energy. You’ll put your hands up to protect your face. (It’s a reflex really.) Hence your hands and arms will probably hit the car before your head does. Energy loss again. All that contributes to getting your head down to a manageable 12mph speed.

So, in fact, while it’s true that bicycle helmets don’t protect your head at very high speeds, impacts where your head slams against something at very high speeds are probably quite rare.

Conclusion: bicycle helmets are useful.

[Update 1: I had a comment from the owner of a site I had seen a while back. His view is that bicycle helmets are not very effective. I’ll say right now that I agree with his position that teaching safe riding practices to people probably prevents a lot more bruises than simply wearing a helmet. I still think that helmets can play a role in certain types of accidents. The question is whether such accidents will actually occur when you cycle. He submitted a link to two documents. I have read the first one and I’ll work my way to the second one.

Richard Keatinge’s site:

And from his comment:

The main problem is that in real accidents helmets dont seem to work as designed at any speed. See the senior engineer of Bell at
and the Australian Federal Bureau of Transport at


3 comments to Going Down Hard: A Bicycle Helmet’s Perspective

  • The main problem is that in real accidents helmets don’t seem to work as designed at any speed. See the senior engineer of Bell at
    and the Australian Federal Bureau of Transport at

  • Richard, thank you for leaving this comment. I’ve just finished reading the first document at:

    Basically, engineer Sundahi argues against the use of a heavier test mass (5kg was proposed for the standard instead of about 3kg) for infant helmets. The justification is that helmets need to decelerate the head at impact respecting a certain maximum acceleration. To do so, the helmet crushes. (Sundahi says that energy absorption is not the critical factor, but acceleration/time is the critical parameter to watch. While it may be through, most materials used to decelerate impacts do so via crushing which absorbs impact energy, so really the helmet does both, it decelarates the head and absorbs energy as crushed materials don’t tend to rebound very hard, but nonetheless, that’s not the point.) The analysis engineer Sundahi goes through shows that a helmet designed with a 5kg test mass would end up being stiffer (the liner material is made more dense) and a lighter head would not allow the liner to crush, but really just to rebound. So the helmet would decelarate the head, but make it reaccelarate the other way. The analysis also shows such a helmet in such a condition would also produce more acceleration than it was designed for.

    It’s quite interesting, but this letter doesn’t tell me whether this has caused actual head trauma in children wearing helmets designed for heavier heads. Also, this document is from 1998. I haven’t looked but is the standard still at 5kg for infant head proxy?

    I’m not saying the Sundahi letter is useless, I’m just wondering if there’s more information on this particular issue. From the analysis in there, it sounds like something to be investigated further. Could it be that in the end, maximum acceleration isn’t the critical factor in head damage in children? This is an interesting discussion (and I think an important one).

    I haven’t read the second document you sent, but let me do so today and comment back here.

  • Richard,
    I just saw the second document is 173 pages, so it might end up being tomorrow. 🙂
    Just before I keep going in this discussion, let me point out that I’ve seen your site (I had been there and read a lot of the material some time ago as well). I agree with you that there are other factors that play an even bigger role than helmet in bicycle safety. Learning how to ride a bicycle safely probably leads in a greater reduction of cycling injuries than simply wearing a helmet. I’m a strong advocate of teaching kids how to ride a bicycle safely. I would also advocate teaching safe riding practices to children in a formal teaching environment (i.e. not at home, but in a place which makes the child conscious that this is something important to remember). Safe riding practices are number 1. We agree on this.

    The questions I’m posing here are:
    1-Can modern bycicle helmet actually help your head in a crash and to what point?
    2-Are typical crash conditions which involve your head within the design bounds of bicycle helmets or do they occur at a point where it would make no difference?

    Just trying to clarify the scope of what I’m discussing here. It also seems like the documents you suggested in your comment deal with that point.

Leave a Reply




You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>