Why sci-fi uber-mechs look cool but are really dumb

Your city’s being attacked by gigantic alien monsters, so you leap into your own 120-metre high mech and go out to do battle. Cool.

It all began in the 1950s with Inoshiro Honda’s Gojira (Godzilla) – a 120-metre tall reptile with radioactive breath – and has been a staple ever since.

Just one problem. It won’t work. Yup, I’m on my hobby horse again. You all think I’m a humble historian and writer with 50 books under my belt. True, and I’ll be on to writing from the next post.

But my other enthusiasm’s science, especially physics. And Hollywood movies…don’t…

Progress, nineteenth century style; bigger, faster, heavier... more Mordor.
OK, so this isn’t actually Bumblebee disguised as a Haulmax…but it’s one big sucker of a truck. For scale, I’m on the right, and I’m 182 cm tall without the hat, which puts me above average for Kiwis but not in the truck league. Click to enlarge (the photo…)

In Pacific Rim, the mech issue was controllabilty. Actually the problem is scalability. When you scale an object the mass goes up by a power of three. That’s why horses have skinny legs compared to elephants.

The average Kiwi bloke is 1.77 metres tall and drinks beer. Scale our statistically average beer-drinker up to 120 metres (a factor of 67.79) and you’ll be scaling mass by 311,619. Sure, our mech’s going to be made of carbon fibre and titanium (or something) – but you get the idea.

Mass brings other issues into play too – inertia, momentum and kinetic energy. You can swing your legs and arms pretty easily. But with 311,619 times the mass it’s a different story – and more than just by that figure. The physics are clear:

Momentum – this is linear, sure – it’s MV (mass times velocity)

Kinetic energy – (the energy contained in a moving object) isn’t linear at all – is ½ MV <exp>2.  The killer is the exponent.

Moment of rotational inertia (the energy needed to START an object moving, assuming you’re rotating the leg about a pivot in the hip) – also not linear, it’s I = mr <exp>2 – yup, that annoying factorial again, and this time the product isn’t even halved.

See how fast a hummingbird flaps its wings? No. I can’t either. Compare that with an elephant – they can’t flick those legs with the instant acceleration of a hummingbird’s wings.

There is no way you’ll get snappy high-speed mech punches or sprints – well, you could, but the energy requirements and tensile strength of the materials you’d need are huge, and the leverage needed means your mech, design-wise, wouldn’t be a scaled up Kiwi beer drinker at all (actually, this might be a good thing – think of the beer you’d save).

So that’s the physics. Mechs are cool. But they won’t work. Not if they’re shaped like humans. Because – as Sir Isaac Newton tells us – we don’t scale.

So is there a way of creating an exciting super-mech? Sure is…and – well, how can I put it? The mech I have in mind would trash the shiny metal ass of anything in Pacific Rim. Or anything else.

Next week…

Copyright © Matthew Wright 2013

Coming up next: Serious stuff (promise) – sixty second writing tips, tips for National Novel Writing Month, and ‘Write It Now ‘ – a regular exploration of everything about writing.


14 thoughts on “Why sci-fi uber-mechs look cool but are really dumb

  1. So glad you wrote this. Whenever I criticize the Star Wars AT-ATs, people look at me cross-eyed. They consider it the most brilliant tank concept ever. I consider it a farce. I agree with you that they “look cool.” I would even say they look awesome, but they are completely impractical. I was even looking at it from a different angle. Since ww2, a key element in tank design has been to lower the profile. The taller a tank gets the easier it is to hit. A revolutionary advancement in tanks was sloped armor, making it more difficult for a round to penetrate. This is quite impossible in a giant anthropoid tank. And why does a tank have to look humanoid anyway? Seems nihilistic to me. Current tank design stays pretty much the way it is because it’s the most practical form for a tank. There’s no point in mounting a giant sword on the arm of a giant tank when the bloody thing is so big it can be targeted from space! Why prepare the thing for melee when it can easily targeted by hundreds of weapons, hundreds of miles away? Walking mechs are cool for movies, but we’ll not see giants, like in the movies, ever. At best, we’ll see man-sized, tactical reconnaissance mechs, and they’ll likely be quadrupeds.

    1. My next week’s post is on exactly this! MBT design really hasn’t changed much since the late 1940s, apart from detail, the whole lot inspired as far as I can tell by the T-38. From which we can extrapolate the ultimate mech (following Stanislaw Lem…). More in a week… 🙂

      1. Looking forward to that. I never heard of Stanislaw Lem’s tank concept, but I have read extensively on Keith Laumer’s Bolo series. Now there was a tank!

  2. I think where Hollywood came close to getting it right was with the various walkers in the original Star Wars trilogy. They were slow, clumsy and incredibly tough. They were big enough to contain a small nuclear reactor (which is the only thing I can think of that could power such a monster, at least with our present knowledge). They had destructive firepower (though their lasers did go “pew”) which meant you got blown up before you had a chance to retaliate and if you managed to knock one down it wasn’t getting up again without the help of a very big crane. But Optimus Prime? Not likely, even if we could get our hands on some energon.

    1. Regarding movie mechs, I agree – the AT&T’s looked really cool and as quadrupeds were way better than a bipedal mech. Definitely based on elephants. But as combat devices, not so good…will explain in next week’s post (it’s written, I’m frantically sorting out photos of my model tank collection…)

      Did you ever watch Thunderbirds? Apropos quadrupedal mechs, I just had a flashback to the ‘Pit of Peril’ episode with the walking mech ‘Sidewinder’ – very clearly suspended on strings, but hey…

  3. I WANNA KNOW. *ahem*

    Okay, not going to lie, I’m a hardcore mecha fangirl. The first ever anime I watched was Gundam Wing. I eventually want to write a mecha novel series. So yeah. I get the scalability issue. Most of my mechs are more Patlabor sized than Pacific Rim sized. It just makes more sense to me I guess, but I know even that can wreck (no pun intended) people’s suspension of belief.

    Also humanoid shaped anything mech-wise is just ridiculous. I don’t think anyone truly believes it’s practical, but with the science and given the issues of the fantasy (often anime) worlds they’re in, man, it’s awesome. Escapism at its finest. But for those who enjoy more realism and less fantastic-ness, I can see why it’s not forgivable or something they’ll let slide on believability.

    But mechs look awesome and that’s what people like me enjoy. Also exosuits, but that’s in a different category.

    I have a lot of design concepts mech-wise, so I’d LOVE to see what could kick the pants off of the humanoid-shaped battle-bots of doom. 😀

    1. The elephant, which is maybe twice the height of a horse, weighs on average anything up to 5 tons. Whereas a horse, on average, might weigh only half a ton. That’s reflected in the thickness of their legs and the way they move.

      The reason for it is the ‘square cube law’ – the ratio between area and volume (which, in animals, translates to mass). Roughly, if you double the height of some poor creature, you’ll cube the weight, meaning that its legs have to be increased to suit. J B S Haldane wrote about it way back, calculating that if a human was scaled up to 60 feet tall, in proportion, they’d break their thigh bones. No giant at the top of the beanstalk I’m afraid!

      It’s worse than that in a way – elephants also have special leg bone structure, without which their legs would be even thicker. And I suppose if we go back a while and look at the dinosaurs, we see the same thing in the sauropods, on an even greater scale.

      Hope that’s helpful!

      I am an inveterate geek…the equation (hopefully the HTML code will work in a comment) is: v_2=v_1\left(\frac{\ell_2}{\ell_1}\right)^3 🙂

Comments are closed.