Nerding Out: Why You Shouldn't Worry Too Much About Weight

No bike review is complete without a weigh-in and most will attract comments arguing whether the weight is acceptable or not. It's easy to see why: weight is a simple, objective metric that's easy to understand and quantify. Lighter is better - simple.

But how much better? After all, saving weight is insanely expensive - most weight-saving upgrades cost several dollars per gram saved. Picking Shimano's XTR drivetrain over SLX, for example, saves 322g but costs almost $1,000 more (about $3 per gram).

Within a category (XC bikes, enduro bikes etc.), the gap between a light or a heavy one is about 1kg (2.2lb), and that's about the amount of weight you could save from a typical mid-range bike by throwing money at it. So how much difference does 1kg make?


Climbing

When climbing, the energy required to overcome gravity is equal to the height of the climb times the system weight - the weight of the bike, kit and the rider combined. Let's say an average rider weighs 80kg, plus 5kg of kit and a 15kg bike, which gives a total system weight of 100kg (these are just rough numbers). So adding 1kg to the bike increases the total system weight by 1%. The time taken to complete a climb depends on the power to weight ratio, so adding 1% to the system weight means it will take 1% longer to complete a climb, or 1% more power to maintain the same pace. So for a half-hour climb, it would take 20 seconds longer carrying an extra kilogram at the same power.

What's more, that's an upper limit.

Bike calculator uses an algorithm to model the forces acting on a bike and rider in the real world, including gravity, rolling resistance and air resistance. It's generally considered to be an accurate representation of real-world cycling while eliminating all unwanted variables. If you plug in numbers for an 85kg rider and a 15kg bike, a 10-kilometre course and a 10% grade, it predicts it will take 77.86 minutes. Change the bike weight to 16kg and it will take 78.62 minutes. That's 0.97% more time, very close to the 1% figure you'd expect from power-to-weight alone. But change the gradient to 1% and the extra kilogram adds just 0.38% to the time; on a 0% gradient, the time difference is 0.2%.

So what's going on? On very steep gradients, air resistance and rolling resistance are pretty negligible, so most of your power is going to overcoming gravity. That means lifting 1% more weight takes (almost) 1% more time (or 1% more power). But on a flatter gradient, speeds increase so these other factors take up a bigger share of your power. Aerodynamic drag doesn't depend on weight, so the faster you go the more aerodynamics matter and so the less that weight matters in percentage terms. With rolling resistance, it's more complicated; adding weight will increase rolling resistance but not necessarily in proportion to the system weight. This partly depends on whether tire pressure is increased in proportion to system weight or not. Either way, rolling resistance does increase with weight; this is why Bike Calculator predicts a slightly slower average speed on a flat course with more weight.

The bottom line is that adding 1% to the system weight (about 1 kg) will impact climbing speed by at most 1% on steep climbs, but less on flatter terrain. For typical riding with a mix of flat pedalling and climbing, adding 1% to the system weight might affect the average speed by significantly less than 1%, perhaps closer to 0.5%.

To put that in more context, the difference in power transmitted to the rear wheel can vary by as much as four Watts between different chain lubes at an output power of 250W. That's a 1.6% difference in the power reaching the rear wheel. On a steep climb, speed is proportional to power, so the choice of chain lube could make a bigger difference to climbing speed than a kilogram of extra weight. That makes high-end chain lubes look like a bargain next to carbon components.

In the latest efficiency tests, the times varied by a huge amount (up to 11% between the fastest and the slowest bike), and only a small part of this variation can be explained by differences in weight. Suspension efficiency (pedal bob), drivetrain efficiency and the presence of an idler could make a bigger difference to climbing speed than weight.


Accelerating

You're probably aware that rotating weight has a greater effect on acceleration compared to non-rotating weight. This is true because when the bike accelerates, the wheel needs to accelerate in the direction of travel (this is called translational acceleration) as well as increasing how fast it's spinning (rotational acceleration). Accelerating anything requires adding kinetic energy to the system, but with a rolling wheel, you have to provide both translational and rotational kinetic energy (usually by pedalling).

If your bike is in a work stand you'll notice it takes energy to turn the cranks just to accelerate the rear wheel up to a high rotational speed - this is the rotational kinetic energy. How much kinetic energy depends on the mass of the wheel: a heavier wheel has more translational and rotational kinetic energy for a given speed.

The energy you apply through pedalling is equal to the power you supply to the pedals (minus all sources of drag), multiplied by the time that power is applied for. So, to get from one speed to a higher speed, you'll need to supply more kinetic energy if the wheels are heavier; this will take more time at a given pedalling power (so slower acceleration).

But how much, exactly?

Because wheels roll, they have a fixed relationship between their translational speed and their rotational speed, and this gives them a fixed relationship between the rotational and translational kinetic energy. The vast majority of the rotational kinetic energy in a bicycle wheel comes from the mass at the outer edge (the rim and tire) - the hub and spokes are pretty negligible. A typical tire weighs about 1kg and a typical rim, about 500g, so for a pair of wheels, that's 3kg of rotating mass that matters.

I've included the relevant equations in the box below, but the gist is that the rotational kinetic energy for this 3kg of mass at the outer edge of the wheel is equal to the translational kinetic energy when the wheel is rolling. That means the rims and tires need twice as much energy to get them rolling at a given speed as the same amount of non-rotating weight on the frame or rider. In other words, the mass of the rim and tire count double in terms of their kinetic energy and therefore have twice as much of an impact on the time taken to accelerate.


In other words, every gram in the rim or tire has twice the kinetic energy of a gram that's not rotating. The time taken to accelerate from a standstill to a given speed is given by the kinetic energy required divided by the (net) power supplied by pedalling, so every gram on the wheel has twice as much impact on the time as a non-rotating gram on the frame or rider.

But, remember we're only talking about 3kg here. So for a bike and rider weighing 100kg, the rotational component of the kinetic energy is only about 3% of the translational kinetic energy.

Let's say you swapped to lighter rims which were 200g lighter for the pair. Because this affects both the translational and rotational kinetic energy, the mass counts double, and so it will have the same effect on acceleration as saving 400g from the frame or rider. In other words, it will quicken acceleration by about 0.4% for a 100kg system weight. Hardly a noticeable amount.

What if you went the other way and swapped your 1,000g trail tires for 1,300g DH tires, plus 200g inserts in each wheel, so adding a total of 1kg of rotating weight? For a 100kg system weight, that would slow acceleration by about 2%. If you've noticed that swapping to DH tires feels like it has a bigger effect than this, that's probably down to the extra rolling resistance, not the weight. This suggests that fitting inserts instead of DH tires may be wise in some situations: in either case, the weight isn't noticeable, but the rolling resistance of a stiffer tire is.

One other thing to note is that the rotational kinetic energy of a wheel doesn't depend on its radius, because although a larger radius results in more rotational kinetic energy for a given angular speed (due to higher rotational inertia), the angular speed is lower for a given trail speed, and these two factors cancel out. This means that if you had a 29" wheel/tire that weighed the same as a 27.5" wheel and tire, the acceleration would be the same. Of course, a like-for-like wheel or tire will be heavier in 29", but only by about a hundred grams or so.


Descending

Another reason to want lighter wheels is to do with unsprung mass. This is the weight of the wheels and the other connected components that aren't held up by the suspension. When your wheel hits a bump, it has to accelerate upwards rapidly, then accelerate downwards again as it moves over the bump and the suspension rebounds. How rapidly the wheel accelerates back down is limited by the mass of the wheel but also the stiffness of the suspension spring, which in turn is determined by the sprung mass of the rider, the frame and everything else that is held up by the suspension.

The heavier the wheel, or the softer the spring, the slower it will accelerate downwards on the backside of a bump. The slower the acceleration, the more often it will lose contact with the ground on fast and severe bumps, resulting in less traction and a harsher ride. This is why engineers in all wheeled vehicles refer to the sprung to unsprung mass ratio - the higher the ratio, the better the suspension can perform.

Calculating the unsprung weight of a bike is pretty complicated because it includes a lot of components (wheels, tires, cassette, derailleur, brake calipers, rotors, fork lowers/swingarm etc.) and because the swingarm's contribution to the unsprung mass depends on where the mass is distributed along its length. It can be measured, however, by putting the bike in a stand horizontally, disconnecting the shock and putting the rear wheel on a scale.

Using rough numbers again, the unsprung mass is about 4kg on the rear and about 3kg on the front. Let's say you were to save 100g per wheel by buying lighter wheels - that's about the difference in weight between DT Swiss' EX 1700 and EXC 1200 wheels, which have a cost difference of $1,964. That would reduce the unsprung mass by 3.3% on the front and 2.5% on the rear. It's hard to quantify how much this will affect suspension performance because that depends so much on the terrain, speed and suspension settings, but I've tested different wheelsets with similar or greater weight differences than this back to back and I can't perceive a difference by feel. It seems to me there are better ways of spending $1,964.

For big changes in unsprung mass, you'll need to make compromises, not investments. Swapping trail tires for DH tires with inserts might add 500g per wheel (about a 12.5-16% increase in unsprung mass) but the extra cushioning, damping, and ability to run lower pressures in the tires will almost certainly outweigh any downside in terms of suspension performance. Similarly, going from 200mm to 220mm rotors will add 25g for the rotor plus about 25g for the adapter, so 50g per wheel. That's a negligible difference to the unsprung mass (about 1.4%) but a noticeable (10%) increase in brake power which could reduce arm pump and increase enjoyment and speed.


Could a heavier frame be a good thing when descending?
If making the wheels lighter is impractical, what about deliberately increasing the sprung mass of the frame in order to increase the sprung to unsprung mass ratio and thereby improve the suspension performance? I investigated this at my previous employer, and while the results were inconclusive, the bike definitely felt calmer and less harsh with 3kg of lead strapped to the frame.

At the time. I put that down to the increase in the sprung to unsprung mass ratio. But now I realise a more relevant factor is the ratio of the rigid sprung mass (the frame and other rigidly connected components) to the total sprung mass (the frame plus the rider).

Because the rigid sprung mass of a bike is so light (around 9kg), it accelerates upward rapidly when the bike hits a bump. This acceleration continues until the rider's bodyweight gets involved, which is only after the bike is moving upward towards him/her. This disconnect is what causes MTB suspension to perform so poorly compared to motorcycle suspension; the suspension needs to be stiff enough to support the total sprung mass of the bike plus rider (perhaps around 90kg), but when the bike first hits the bump, the rider is too loosely connected to the frame to provide much resistance. So the 9kg rigid sprung mass starts to accelerate upwards rapidly, without much suspension movement, until this lag is overcome and the rider starts to accelerate with the frame and create enough resistance to compress the suspension. Increasing the mass of the frame reduces this initial acceleration and forces the suspension to move earlier, and so decreases vibration experienced by the rider. This is why e-bikes are far more comfortable to ride on rough terrain, even when running less sag.

I'm not necessarily saying heavier bikes are better overall when descending - both Jack Reading and the Santa Cruz Syndicate experimented with adding weight to their race bikes, but as far as I know, neither stuck with it - but I am saying there are tangible benefits to heavier frames when going downhill. I certainly wouldn't stress about having a lighter downhill bike, unless you're carrying it to the top!


Don't miss out

One common argument is that, yes, a few grams here or there isn't much on its own, but saving weight where you can throughout the bike adds up to a noticeable saving. This may be true, but allowing for burlier parts in your weight budget adds up to big benefits, too. For example, bigger tires offer more grip and cushioning, bigger brakes reduce fatigue and allow you to stop later, stiffer forks can improve suspension performance through chunky terrain. These factors in isolation might not make much difference, but between them, they create a more confidence-inspiring ride.


Some good counter-arguements

My overarching argument here is that the importance of weight is overblown, largely because the weight of the bike is pretty small compared to the weight of the rider, so an extra kilogram on the bike might add only 1% to the amount of weight you have to shift up and down hills. But there are some exceptions where this logic doesn't apply. In particular, when bunnyhopping, pumping, carrying your bike, or technical climbs involving manoeuvring the bike up and over obstacles; a kilogram on the bike may count more than a kilogram on the rider here.

I've also been assuming throughout this thought experiment that the system weight is about 100kg, but of course, if you're a lot lighter any additional weight will have a bigger effect.

To conclude, I'm not saying that a 16kg bike is indistinguishable from a 10kg bike, or that lightness isn't a good thing. I am saying that when you run the numbers the benefits of saving a realistic amount of weight are smaller than you might think, while the advantages of heavier components can be far more tangible in terms of descending performance, not to mention reliability and cost.