Myth 10: Stiffer Forks Steer Better

To celebrate Bicycle Quarterly‘s 15th anniversary, we are looking at myths in cycling: things we used to believe, but which we’ve since found out not to be true. This week, we have a ‘double feature’ that looks at fork blades. In the first post, we looked at whether they flex enough to improve comfort. Here we examine the belief that stiffer fork blades make the bike steer better.

Looking at Hahn cornering hard on the comparatively flexible Kaisei ‘TOEI Special’ fork blades (above), you can see that the wheels are perfectly aligned as he scythes around this fast downhill corner. His bike steers no differently from a bike with ultra-stiff fork blades. This goes against the widely held belief that a stiff fork offers more ‘precise’ steering.

Stiffer setups improve the steering response in cars and tricycles, where the forces of cornering flex the suspension components. On a two-wheeler, those forces are aligned with the centerline of the bike, otherwise, the rider would fall over (above, from Bicycle Quarterly‘s article on balancing and steering). Even when you corner hard, the centrifugal forces don’t cause the fork to flex.

At low speeds, the front wheel turns at a greater angle, which puts small lateral loads on the fork, but since you aren’t going fast, you won’t notice the little flex this causes. The one exception is tandems: Due to their long wheelbase, you can feel the flex of the front wheel when cornering very hard on tight mountain hairpin turns.

At high speeds, as in the top photo, the fork turns very little even when you corner hard, so the flex is insignificant – even on a tandem.

The only time you put significant side loads on the bike is when you ride out of the saddle. What you realize then is that the lateral flex of the front wheel is far greater than that of the fork.

Climbing out of the saddle on the same bike with two different front wheels confirmed this: On a wheel with a ‘narrow’ SONdelux generator hub (right), the rim rubbed on the brake pad. With the ‘Wide-Body’ model (left), the rim never touched the brake pads, because the wider flanges make the wheel laterally stiffer. The difference in the flex of the wheels was very noticeable.

By comparison, the fork blades flex only very little: Riding the same ‘Wide-Body’ wheel on bikes with different fork blades (standard vs. flexible Kaisei ‘TOEI Special’), we never had the rim rub on the brake pads – even flexible fork blades are plenty stiff for riding out of the saddle. You can easily test that by pulling sideways on your front rim while holding the handlebars steady. You’ll see the wheel flexing, but there won’t be any visible flex in the fork blades.

Even a flexible wheel corners just fine – I haven’t heard anybody talk about the poor handling of aerodynamic wheels with few spokes (and low lateral stiffness). Still, there is no benefit to a laterally flexible wheel… We asked Schmidt to make the ‘Wide-Body’ hubs for us, because they don’t just reduce brake rub – they also make a much stronger wheel. That allows you to use fewer, thinner spokes, which improves the wheel’s shock absorption, weight and aerodynamics.

The one time when you don’t want your fork blades to flex is when you are braking hard. The longer the lever, the greater the force – so fork blades taper as they move away from the hub. The tops of fork blades are ovalized for the same reason – they have to resist the loads of braking, rather than side-to-side flex. Thanks to these features, the flexible ‘TOEI Special’ fork blades are more than stiff enough for hard braking. Their flex is concentrated at the bottom of the blade, where the forces of braking are small.

Disc brakes are a different matter, as the forces of braking are fed into the fork blade where the caliper attaches. That changes the requirements on fork blades. Using flexible forks with disc brakes will require some more thought and testing.

Contrasting with the myth, slender fork blades actually improve the cornering, because they absorb bumps that would otherwise unsettle the bike. If your bike skips over bumps, the tires lose traction…

If you are a powerful sprinter, the suspension of the fork blades can be a disadvantage: When you work the bike extremely hard while sprinting out of the saddle, slender fork blades can feel a bit like a suspension fork that bobs with your pedal strokes, albeit to a much smaller degree. Even track sprinters don’t need ultra-stiff fork blades; they might choose something like the Kaisei ‘Standard’ fork blades. In fact, since they also rock the bike much more from side-to-side and don’t have brakes, they often use round fork blades.

For the rest of us, we’ve found that the Kaisei ‘TOEI Special’ fork blades (above) take the edge off large bumps to improve comfort, tracking in corners and speed. Steel is a great material for fork blades, because it can flex without affecting its durability: our ‘TOEI Special’ fork blades have been 100% reliable over tens of thousands of miles. I consider them a key component of our bikes, making them faster and more enjoyable at the same time.

More information:

About Jan Heine, Editor, Bicycle Quarterly

Spirited rides that zig-zag across mountain ranges. Bicycle Quarterly magazine and its sister company, Compass Cycles, that turns our research into the high-performance components we need for our adventures.
This entry was posted in Framebuilding supplies. Bookmark the permalink.

41 Responses to Myth 10: Stiffer Forks Steer Better

  1. Brad Lay says:

    I’d be interested in your thoughts on titanium as a material for forks. If flex in steel forks is considered to be acceptable, or even beneficial as you say, why has flex in titanium forks been framed in the negative?
    I don’t have first hand experience in riding titanium forks myself, but have wondered why titanium forks are not well received, beyond the obvious issue of aesthetics.

    • Tom G says:

      My impression is that Ti does not flex enough… and maybe is challenging to weld at the crown? So makers seem to spec a carbon fork on the Ti frames. Carbon does a wonderful job of absorbing vibration. I recently saw a steel fork on a Ti frame, but I’ve not seen a Ti fork on any bike.

      • Beau says:

        Ti flexes a lot more than steel. This is why tube diameters on ti frames are larger than on steel frames. They have to be larger to make them as stiff as a normal steel frame. Litespeed made a bike ages ago with smaller diameter ti tubing and it was known for being a noodle of a bike. Ti is also much more difficult to work with than steel and more prone to cracking than steel. Ti forks are not impossible, but the limitations of larger diameter tubing and the difficulty of working with it has deemed not worth the effort by most builders. I saw one on a 20″ folding bike once and it seemed to be fine due to it’s tiny size. That’s the only functional example I’ve ever seen or heard of but would be happy to hear otherwise.

      • I saw a British hillclimbing bike in a magazine once, and it had a titanium fork, too. The article said that since the bike was designed for uphill riding mostly, fork flex under braking was less of a concern.

    • Oscar says:

      At least one titanium fork (26″ tires, cantilever brakes) is famous in the MTB world : the Kona P2 titanum, it´s said to work really well, it was light….and in was costly: like 800$ in the first 90´s.

  2. What is the centrifugal force? When cornering I lean in to generate the centripetal force needed to accelerate into the apex of the corner.

    • We can argue semantics all day (whether the centrifugal force is a real or ‘imaginary’ force), but the fact is that the bike leans to balance the centrifugal force by using gravity to create a centripetal force.

      • ayjaydee says:

        the force of gravity is always straight down and hence perpendicular to centrifugal force. Centripetal force is parallel to centrifugal force and is generated by tire friction against the road. . It’s moment is resisted by gravity due to the lean.

  3. DaveS says:

    A few years back, we had the opportunity to compare 2 different forks on a tandem bicycle. The first fork was a Co-Motion T390 carbon fork that is very stiff. The second fork was a steel Co-Motion fork. Dimension wise, the two forks are identical. They have the same rake, the same crown to axel distance, same tire clearance. They both utilize disc brakes. They differ in material (the carbon was stiffer) and the carbon utilizes a taper steer tube, while the steel has a straight steer tube. Obviously the steel fork is not super flexible because of the disk brake on it.
    We have some regular routes that we take and noticed that the steel fork was faster on these routes. The only change on the tandem was the fork. At first this didn’t make sense because the steel fork is 2.25 pounds heavier than the carbon fork and the route we took is hilly (16 miles – 1000 feet of climbing). The time difference was significant so I’m confident this is not experimental variation.

    When going over bumpy roads, I can see that the steel fork flexes more. When corning over uneven pavement at higher speeds, the more flexible fork allows the tire to stick to the road better. The tire doesn’t chatter like when the carbon fork was used. With this added confidence, this allows us to go faster in these conditions. Other than this, I did not notice any difference in handling between the two forks. This includes during heavy braking which happens often on a tandem bicycle.

    The added flex also makes the ride more comfortable when riding over bumpy roads, although I’ll have to say that wide subtle tire have made a bigger difference here.

    That said, I still have problems talking to folks explaining how it is possible that a heavier steel fork could possibly be faster than a light carbon fork. I’m glad that BQ helps justify this.

    BTW, my real reason for going to a steel fork is because the carbon fork developed a crack in it.

  4. ayjaydee says:

    Your force diagram does not include the horizontal and vertical forces of the road on the tire at the contact patch. There is a resulting bending moment in the forks.

    • You are correct – the diagram is simplified. There is a small bending moment that is mostly flexing the tires. When you look at the top photo, with the tires inflated to about 35 psi, you can see that even at that low pressure, the force isn’t large enough to significantly flex the tire, much less the fork.

  5. Phil Brown says:

    The only way to begin to test these two theses is to use high speed video to actually show the flex, if it’s occurring.

    • You are right, dynamic forces can be different from static ones, due to inertia and other factors. If you ride two different forks, one with stiff blades and one with flexible ones, you’ll notice the difference in shock absorption right away. Clearly, the inertia of the front wheel isn’t so great that the fork can’t flex. Otherwise, suspension forks wouldn’t work, either.

      We might repeat the experiment in the video and set the bar of the low-rider so it touches the bag-support rack when the fork flexes a certain amount. Then ride the bike and we’ll can hear if the racks touch. By changing the distance between the racks, we can quantify the fork flex over a set of bumps, like the expansion joints in concrete road sections, to get quantifiable results.

      Or we can just continue to ride our forks, since they work so well. Really, our goal is to improve the design of bikes, not to win over skeptics.

    • Tim Cupery says:

      The prior blog post about suspension effects of steel fork blades includes a video, although it’s showing flex from weighting and un-weighting the bars, not while riding. But it does show part of what you’re looking for.
      https://janheine.wordpress.com/2018/05/17/myth-9-fork-blades-dont-flex/

  6. gcziko says:

    You wrote: “Stiffer setups improve the steering response in cars and tricycles, where the forces of cornering flex the suspension components. On a two-wheeler, those forces are aligned with the centerline of the bike, otherwise, the rider would fall over . . . ”

    Isn’t that the case only when the cyclist leans his or her body to stay aligned with the bike, as in your diagram? It is possible to lean the body more than the bike, or lean the bike more than the body while cornering without falling over. It seems to me that those cornering styles would put some side forces on the fork and wheel.

    • Yes, you can lean our body without leaning the bike. When you look at Hahn in the top photo, you can see that he is instinctively pushing the bike a bit into the corner, leaning the bike a bit more than his body. Some riders flip their inner knee outward from the bike when they corner, perhaps in an attempt to keep the bike more upright. (This doesn’t help with cornering.) In all those cases, the effect is very small.

      With motorcycles, you sometimes see racers hanging off the bike to the inside of the corner. I don’t know enough about motorcycling to understand why and how it affects the forces on the motorbike.

      • gcziko says:

        Moving the inside knee away from the bike could help the rider lean a bit less than the bike while keeping his or her center of mass in line with the bike.

        My understanding about motorcycle cornering is that the machine’s maximum lean angle is limited by the width of the frame at the bottom. So the rider has to move more body weight to the inside to compensate to make a hard corner. Fortunately, bicycles can lean much more than motorcycles.

      • Andy Stow says:

        It lets you keep the motorcycle more upright, which helps in two ways. The suspension gets to do its job, and the hard parts don’t scrape the pavement.

      • Thank you for the clarification. I forgot about the suspension! That makes sense – it’s a big problem in motorcycle design that when the motorcycle leans over at extreme angles, the bumps also occur at extreme angles to the direction in which the suspension moves. Even on smooth tracks, that limits the cornering capabilities of motorcycles.

      • I’ve always thought that one (unconscious) reason people flip their inner knee towards the corner is that it helps to turn the hips (and thus the bike) in the direction of the corner. Eyes turn—head turns—knee turns—hips turn—bike turns: all at once. It’s not necessary to do it, but it may help—either literally pulling the hips around, or directing the body/mind (perhaps in the way many people point with their non-throwing arm when throwing). I know that when I get tired on all night rides I sometimes use my knees like this to keep myself focused on my cornering. Just a theory, I may be ludicrously wrong.

  7. Conrad says:

    I strongly prefer the feel of a good steel fork over a carbon fork. As you said, on rough roads you can watch the fork blades actively flex like the leaf springs on a truck. The tendency to install carbon forks on steel frames bugs me. The fork is the last place you should look to save weight. If the fork fails, serious injury usually results. If bikes are raced and ridden hard and occasionally crashed… I dont know why everyone is okay with rolling the dice on a carbon fork. I have had steel forks straightened after hard crashes. I watch for cracks but nothing ever happens. My road racing bike is 25 years old. I have yet to render any of my steel bikes unrideable.

  8. Bill Lindsay says:

    The cliffhanger (in my mind) from yesterday was an assessment of the tradeoffs of a RockShox. Yesterday you showed a RockShox was faster and more comfortable than even a Singer fork, even on smooth pavement. That was a shock to me, and I suspect most of your readers don’t take that result seriously. Conventional wisdom is that shocks are wrong for the road. Today I was looking for the trade-offs for why 99.9% of ‘road’ riders don’t use and should not use a RockShox for road riding, but true suspension forks aren’t even mentioned today. Do you think actual suspension forks (Lauf Grit, for example) could be conceivably ideal for road riding? Or are there unpleasant trade offs, like weight, bobbing, and aesthetics that over-ride the speed benefits of actual suspension?

    • The difference between the RockShox and the Singer fork was not statistically significant. There is some noise in every test, and the road with the rumble strips wasn’t quite as perfectly flat as the track we use for our tire testing, so there was a little more noise.

      But yes, generally, it’s worth testing whether a Lauf Grit fork wouldn’t be faster on a smooth road than an ultra-stiff carbon fork. Or better yet, whether some compliance could be designed into carbon forks just like it was designed into the Kaisei ‘TOEI Special’ fork blades, to improve both comfort and speed.

      • Bill Lindsay says:

        From my perspective, the Lauf Grit is exactly that: a carbon fork with some compliance built into it, just like the Kaisei blades.

  9. Rick Thompson says:

    My new bike (Fitz) has the TOEI blades, they are more flexible than on any of my other bikes. Pressing down on the bars, or rocking the bike with the front brake on, they feel very soft and noodle-y. Handling is great, though, precise cornering and no bouncing that I can feel riding out of the saddle.
    You call it shock absorption, but surely do not mean they are dissipating energy like an auto shock absorber. These steel blades must be acting as springs, transferring the vertical deflection back into horizontal force to reduce speed loss. Perhaps the tire is pushing on the back side of a bump, then the force is directed forward.
    If they are springs, then there would be resonances and conditions where they perform better or worse. There is a path with a rough pattern of chip seal near me, at typical speed the ride is much rougher than I would expect and I feel it in the bars. This may be where the forks are hitting a bad resonance. Most other surfaces the combination of wide tires and flexible forks make the ride smoother.
    I have tried riding on the rumble strips, just to see what you guys did. Even with 44 mm tires that is no fun, I would really want some absorption at the rear as well to keep that vibration out of the seat!

    • You are totally right – the fork blades act as springs, not shock absorbers. The resonances are interesting. You see that with generator hubs, too, where their ‘notchiness’ can get in harmony with the bike’s fork at certain speeds and become really bothersome. The bike should be designed that these resonances don’t occur at the speeds you usually ride. Fortunately, the SON Delux hub and the Kaisei ‘TOEI Special’ blades work well in that respect. The aluminum fork of my Alan ‘cross bike worked very poorly with a generator hub when I used that bike for randonneuring for a short while. The vibrations were worst around 19 mph – exactly the speed I usually ride in brevets.

  10. Mackenzy says:

    Having recently switched from an old 80’s touring bike (Mikado Gaspe) with cantilever brakes and 27x32c tires (Paselas) to a new bike (Crust Romanceur) with a 650x48mm tire (switchback hills) with disc fork I’ve been thinking about how with the disc fork is built to withstand the braking. However it seems the switchback hill takes up a lot of that slack of less fork flex. Obviously, riding purpose and style aside, it seems that fork flex would be hard to look at independently from tire size/quality?

    It’s fascinating reading through all of these write ups, as bikes should ideally viewed as a complex system which can accommodated for an individuals and type of riding, ergonomic needs, and fitness level.

    • You are right, tires are the biggest component in your suspension. Especially with a tire as wide as the 48 mm Switchback Hill, you’ll get enough suspension to handle almost any bump you’ll usually encounter. With the 42s on my bikes, the flexible fork blades are more noticeable.

      When we tested on the rumble strips, we used a stiff Bontrager Hardcase 25 mm tire to isolate the effects of the fork. What that means is that with a more supple and wider tire, the power savings of the fork are probably smaller.

      • Tim Cupery says:

        It stood out to me that the padded handlebar tape appeared to make a difference, perhaps enough to be statistically significant if you pooled the smooth+rumble data for the Trek, and used multivariate analysis with two variables (surface and padding).
        It’s good to know that this was tested with stiff tires and any real effect would likely be smaller (and more difficult to detect from the noise) with larger, more supple tires.

      • You make some good points, but the suggestion to change the statistical analysis to ‘improve’ the results leads down a dangerous path, because it ‘unblinds’ the analysis and allows you to work toward getting the results you expect. We’ll never do that, but the only way to avoid it is to decide beforehand what analyses we’ll do, and then stick to that.

      • Tim Cupery says:

        Good point – re-doing the analysis based on “it looks like this may end up statistically significant” in search of publishable findings amounts to p-hacking.
        My initial thought wasn’t in-search-of-significance; I wondered if you’d taken a multivariate analytic approach from the start, as is normal in social science research, what the effect of the bar tape would have looked like. But I agree that you’d need to start out with that plan to be kosher.

      • From the accompanying article about the testing methods, written by our statistics expert Mark Vande Kamp:
        “a univariate analysis of covariance was used to test: a) the effect of the road surface, b) the effect of the equipment changes (e.g., tires or forks), and c) whether the effect of the road surface varied for the different equipment. The speed measured at the time of each power reading was included in this analysis as the covariate.”

      • Tim Cupery says:

        Thanks for the conversation here. A potential critique of my own suggestion: quantitative social science typically uses multivariate statistical analyses because we are constrained to working with observational (non experimental) data. We must statistically “control for” potentially spurious variables since we can’t alter the experimental design to vary only one factor at a time. I am less familiar with the statistical considerations of experimental design (we usually teach it as the cleaner, “ideal” contrast to analysis of observational data).

        My suggestion ends up being little more than a way to increase the sample size (or number of trials, in the case of experimental research). This could be done in univariate context by simply increasing the number of trials.

        Thinking through the mechanisms, softer bar tape is effectively a small amount of suspension for the hands contact point and a small proportion of the feet-and-seat contact point. Given what we know about suspension and suspension losses generally, I would expect it to meaningfully boost speed if tested in a large enough number of trials.

        But as you say, this is likely to be fairly minimal (even if tested in a large enough sample size to find statistical significance) when using supple tires with more air volume.

        Thinking on this topic raises a related question: carbon fiber handlebars and seatposts are claimed to substantially reduce road buzz reaching the rider at two main contact points. Thinking theoretically here, about mechanisms, might we expect carbon bars and seatposts to result in a speed advantage over aluminum, not due to weight but due to decreased suspension losses? Or does the vibration damping of carbon fiber differ enough from actual suspension flexing, as to have no effect on suspension losses?

      • The handlebar tape is less effective because the unsprung masses are so large. Once the entire front of the bike is vibrating, it takes much more to absorb those vibrations than at the tire level, where just a tiny contact patch is moving, or at the fork level, where only the front wheel is moving.

        If carbon posts and bars reduce the ‘road buzz,’ it’s because they are more effective at absorbing energy than metals. However, this might just move the suspension losses from the rider’s body to the bike component – better for comfort, but possible with the same amount of energy absorbed.

  11. Tom G. says:

    Ok this has me asking again why there isn’t a suspension fork for road / cross / gravel bikes? The only thing I’ve seen that’s close is Cannondale’s Lefty on it’s Slate gravel bike. Not sure it is adaptable to other frames and the left only support kind of freaks me out.

    A review in a MTB focused website convinced me that the Lauf fork is just a spring that could, under certain conditions, be more dangerous than a regular fork (flying into some deep washboard perhaps), The reviewer said it’s because it could cause the bike and rider to “pogo” due to the total lack of dampening. I recently looked at claimed weights for Rockshox and Fox forks and despite being marked to guys who dream of “hucking big air” some are down in the 3 pound range. That means a fork intended for road and gravel could be even lighter. I’d buy one!

    • Rick Thompson says:

      Just speculating here, but any dampening in a road shock would be soaking up power and losing speed. Too much travel without dampening leads to pogo. My hardtail mountain bike converted to road cruiser is great for riding over curbs but pogos when standing. I suspect the springiness of a curved steel fork is close to maximum useful range, and is the simplest solution (for rim brakes). The Lauf looks like a complicated way to achieve similar springiness with stiffer legs for disc.

    • Jacob Musha says:

      Rock Shox made a suspension fork for road bikes back in the 1990s designed around 700c wheels and side-pull caliper brakes called the Paris Roubaix. It never caught on, of course. Which surprises me since it must have been a huge advantage on the cobbles, especially with the ridiculously narrow tires they were running back then. My guess is that racers disliked its sprinting behavior.

      In any case, if you want to build a road/cross/gravel/whatever bike with a suspension fork, just build one around 26″ wheels. There are plenty of 1990s Rock Shox MTB suspension forks available that have cantilever studs and are a lot lighter/less travel than the big downhill ones they sell today. They even pop up in decent numbers NOS for not a lot of money, since no one seems to want them anymore.

  12. David Snyder says:

    I tested RockShox Roubaix front forks both with and without the air-sprung rear suspension on a Boulder Paris Roubaix full-suspension road frame, and overall I was less than enthused. Cornering clearance at the pedals is grossly affected unless the air preload pressure is very high, which then makes the suspension useless except on big hits. The air fork also bobbed quite noticeably while climbing while out of the saddle, even with the high pressure. It also made it more difficult to detect a slow-leaking front tire, which once caused me to have a rare slide-out crash in a fast corner.

  13. John Duval says:

    I have ridden several aluminum forks, none of which seemed to have good characteristics. The first was an ultra rigid Cannondale unicrown fork with a punishing ride, which was recalled for breaking. A flexible road fork, a middling cantilever fork, and stiff disk fork all had scary handling characteristics, and none with a smooth ride. Each time I switched to a steel fork, handling and ride quality improved dramatically, whether is was flexible, or super stiff. For whatever reason, aluminum doesn’t make a good fork for me regardless of design.

    • Jacob Musha says:

      My SR Litage road bike (lugged, glued aluminum) has an aluminum fork. I really like it. In fact, the frameset reminds me of a super-light steel frameset. The tubes are small in diameter so it’s not stiff like a Cannondale or most “modern” aluminum frames. As to your experiences, it is possible to design a bad fork out of any material.

Comments are closed.