Where to start‽ (These are my recollections, perhaps wrong of course):

- The wafer thin margin of actual payload that the booster put up now being shared with landing fuel

- Engine restart after launch

- Re-entry forces on the engines

- The fact (perhaps since addressed) that a single Merlin puts out more thrust at near minimum throttle than the dry mass of the booster itself.

Think about that last one. If you start the burn too early you'll miss the ground .. for a while. The lesson here is trust calculus!

> - The fact (perhaps since addressed) that a single Merlin puts out more thrust at near minimum throttle than the dry mass of the booster itself.

It's addressed in the sense that they live with it. The rocket still does a suicide burn.

I'm not sure I've ever thought it was impossible -- I've learned to be more circumspect about that kind of thing as I'm nearing 40 -- but I tried imagining the fine detail engineering that would be required to do it and was totally overwhelmed.

I mean, take a yardstick or short rod or something, and try to balance it upright on the tip of your finger. That's basically what they're doing, but with a -rocket-. It's amazing.

The rocket has its center of mass very low during landing, as it's mostly an empty shell by this point, with heavy engines at the bottom. So take a meter tape, pull it out a bit, lock and try to balance that. Much easier, right?

The magic is in a) ensuring the whole construct survives reentry, and b) timing the landing burn - even a single Merlin engine is too powerful for the almost empty rocket to hover, so they need to zero out their velocity at precisely the moment they're hitting the ground. There's no second attempt.

> even a single Merlin engine is too powerful for the almost empty rocket to hover

Indeed. For their last launch, they decided to try landing the booster with three engines rather than one, which brings the whole 'hoverslam' concept to the next level.

This was considered so risky that they didn't try landing it on the barge, for fear of damaging it. Instead, they soft landed on the ocean.

But much to everyone's surprise, it actually worked! There's a picture of the Falcon 9 booster floating awkwardly. Musk said they'll try to tow it to shore.

And all that is really the core of why I love SpaceX: continuous improvement. It's a grand thing!

> So take a meter tape, pull it out a bit, lock and try to balance that. Much easier, right?

While you drop your hand toward the ground at mach 5.

This actually helps stabilize the rocket. Lawn dart, and all that ;).

I decided I had no way to guess on the mass distribution of fuel, engine, and airframe, plus wind shear, so I went with the hard scenario. If anybody knows the numbers for this, I'd love to hear it!

Not parent, but I was surprised when I first heard of it. I thought it must be infeasible - in retrospect, for no reason other than vertical-standing rockets are associated with kitschy cover arts of old sci-fi stories. It didn't fit the more modern sci-fi aesthetics I grew in.

Since realizing that I became more wary of unconscious biases picked up from environment, and how they affect what I think is technologically possible.

Perhaps I grew up in a different era, but it's precisely that same aesthetic which makes me see these landings as futuristic because they align so well with those images.

The ascent into space looks like a massive show of raw power, and in contrast these vertical landings look so graceful and controlled.

> The ascent into space looks like a massive show of raw power, and in contrast these vertical landings look so graceful and controlled.

Personally, I think both are a "massive show of raw power" - both literally and figuratively.

Literally goes without explanation. But what I mean of the word "figuratively" is that it shows our capability as a species to fully harness our intellect, knowledge, and emotions to do this kind of engineering, despite knowing it can all go horribly wrong (especially in the case of manned space flight).

Certainly there are many other drivers to why we do this (profit in the case of SpaceX certainly is a great one), but part of it is also to show ourselves we can push beyond what seemed or seems impossible at times, and make it routine, and eventually both practical and relatively "safe" (always knowing it might never be perfect).

The Space Shuttle was perfectly capable of carrying its main fuel tank into orbit... I always wondered why they didn’t do that as standard, then later connect up all the now empty tanks, pressurise them and hey presto, instant space station, far bigger than the ISS. Such a waste to just drop them.

Nasa offered to take the tanks into orbit if a private company would step up to do something useful with them but nobody ever did.

I was sure for a while that they'd have to revert to using nets to catch the rocket at the final moments of landing.

That is exactly what I thought.

I know that many aerospace engineers were skeptical it would be possible to achieve ignition when facing "backwards" during hypersonic flight. No way to get the right mixture in exhaust chamber.

Even engines that "startup" more than once were pretty "out there". Over the years watching along, I've gotten the feeling that getting the engines to restart at all, and then restart in the right order and with enough consistency was one of the major breakthroughs.

I found the NASA blog entry on engine ignition pretty interesting (https://blogs.nasa.gov/J2X/tag/ignition/). Doing that when you have mach 6 winds blowing always seems amazing to me.

to me it was having to carry the fuel for the landing burn. Also, an empty booster is very fragile, they're basically metal balloons. I figured the landing legs, extra support needed, and the required fuel would make it infeasible in the end.

> an empty booster is very fragile, they're basically metal balloons.

If you think about it that's not really the case. Post launch the boosters are empty and must sustain their own weight and that of the payload sitting upright on the launch pad prior to fueling.

During the launch the full force of the engine is exerted on everything and it must withstand that force even as the tanks empty. If the boosters were not able to withstand these forces then as the rocket climbed into the sky you'd see it start to compress like a tin can the higher into the atmosphere it climbed.

I thought it was kind of unrealistic as well. The reasons were that now you have to store fuel for the return trip of just an "empty can", as well as (possibly) even more launch trajectory restrictions to ensure it can reach the landing site.