Superheavy landing mechanism

[Slarty1080]

Musk's recent tweet:
"We’re going to try to catch the Super Heavy Booster with the launch tower arm, using the grid fins to take the load"
https://forum.nasaspaceflight.com/index.php?topic=47352.msg2173732#msg2173732


provides an interesting example of "the best part is no part" concept. But how are they going to make this work?

[Slarty1080]

Here's one interesting animation showing a concept using rotating hoops:

[steveleach]

There's another variant shown in



from 2:35.  I'd like to link to the original from C-Bass productions, but I can't find that.

This one just uses arms and a single tower.

A similar option from TijnM is at

Joris Robijn goes for two half-rings on arms at

Erc X had a variant with a u-shaped catching platform suspended from cables, but I can't find it now.

[Burninate]

For a land or shallow-water landing:

The only thing that makes sense to me if they're doing a "Catcher" is some combination of this:



and this:

https://www.reddit.com/r/EngineeringPorn/comments/lea0je/extremely_fast_cabledriven_parallel_robot/

Using towers surrounding the pad to hold up cables that position a landing ring can be very fast, with low moving mass, can give it a large maneuvering area, can cushion the landing and the rotation with some degree of spring constant (or catenary slack in a slower system), and with 6D control can even go off-horizontal to catch a Starship that's reached the desired spot, but which isn't vertical yet.  Guyed mast towers are somewhat cheap, and you can build this system so that it's got some redundancy (more towers than necessary for 6D control), and with engineered safety releases, you can protect the towers and winch mechanism while sacrificing the cables & ring if something goes very wrong.

[Thrustpuzzle]

The thickest wire ropes ever made are 18 centimeters thick. Even they would not be able to support the dynamic impact and decelleration of an empty Superheavy.  The wire ropes wouldn't be stressed like the pure tension of a crane's support cable; you're catching a very, very heavy dynamic load on the side of a cable, perhaps at a very focused point, in an instantaneous impact.

Using a "ring" as Burninate proposes, helps enormously but that ring itself would need to be massively strong, and therefore massive, and need even more cables to support just the ring.  It's a geometry problem.. materials like to be in direct compression or tension, so sideways forces are much harder to support. The landing leg problem has similar geometry issues with supporting off-axis loading.

I really liked the elegant idea of a cable catching system too, until I realized that at the scale of Superheavy's mass,  mechanical engineering and material science doesn't give us thick, strong, robust cables that bend as flexibily as our mental models.

[Burninate]

Tension is dramatically easier to engineer for than compression in most ways, because compression buckling failure modes demand so much more structural reinforcement.  Wire ropes pretty much get stronger in proportion to their cross-sectional area, but a rigid structure under compressive loading requires higher than geometric mass scaling.

I don't know where you're getting that about wire ropes, but suspension bridge cables of 1 meter diameter are routine. The breaking strength of a steel cable about 6cm in diameter is sufficient to hold up 200 tons of static load, but I would imagine given the tension and dynamic loads involved, you're better off examining, say, twelve cables 10cm in diameter attached to 6 different towers, and actuated with winches that have high torque, but which have some adjustable give (eg direct-drive motors rather than gearboxes).  I'm not very familiar with them, but the first place I would look for COTS analogues would be ship anchor windlasses, which deal with an anchor chain that has to keep a 200,000 ton container ship stationary against high winds and the force of millions of tons of water.

If you need some elastic tensioning, that doesn't need to be a consistent property of the cable, it can be placed at any point on the cable, and it's doable with steel springs or rubber or any variety of mechanisms.  Elastic tensioning is undesirable for maneuverability, but may end up being desirable for impact loads depending on how the engineering trades work out.  Or you might be able to perform that impact damping electromechanically, or hell, hydraulically.

If you had a high degree of control over landing precision, and dynamic maneuverability proved unnecessary, most of these benefits still apply to cable held in loose catenary tension like the FAST telescope scheme above, or a similar system at Arecibo with a much heavier (1000 ton) equipment ring using 39x 9-ton cables; the flexibility of the cables in bending (rather than stretching) and the potential energy associated with raising slack cables against gravity into a higher-tension arrangement, absorb energy as well.

[CorvusCorax]

A thing to keep in mind is reliability. You don't want a "Rube Goldberg" superheavy catcher, the mechanism must be simple, and most importantly failsafe. More so than legs on the booster itself.

Any mechanism that need real time synchronization with the landing booster, and time synchronized movement  of an actuator adds a potential cause of failure. Any additional actuator that needs to work in sync, multiplies this chance of failure.

The booster is coming down under thrust. This introduces plasma and a lot of ionization into the atmosphere above the landing pad. Add to that an extreme accoustic environment, and your chances of having a less than 100% reliable telemetry link to the returning booster in the critical stage goes up significantly.

Actuating a multi-hundred ton mobile structure - such as a grabber arm - both precisely and quickly - is also not trivial. The Falcon9 transporter-erector is an example of a structure that can do both precise and quick - but so far not at the same time. You will have noticed that it does precise motions - close to the core - very slowly, and does the quick "throwback" only from a pre-inclined position.

I wouldn't be too surprised if separate actuators are responsible for the large scale quick motion and the exact tower alignment.

As mentioned above, steel cables could easily hold the booster - including dynamic loads - if you go to sufficient diameters, such as used for suspension bridges. There is a catch though: Cables that large typically do not move. You cannot winch them - or if, you'd need a reel with a radius of 200 ft or more, which would move at a snails pace to not collapse under the dynamic loads of the cable itself which would weigh thousands of tons.

Again there's exceptions to the rule. The anchor chains of a super-tanker or aircraft carrier might be a good example

but all things combined, I would think the best solution would be that, which has no moving parts.

my 3d modeling skills aren't great, so i uploaded an image of something with a similar basic shape.

The actual "ring" could be lifted above a support structure via springs and passive dampeners to absorb the boosters remaining energy and adjust to off-angle landings.

This assumes that the on board control of the booster is capable of "threading" the ring perfectly - even if that means coming in at a slight angle. Having the legs angled outwards not only allows the structure to absorb sideways forces better but also creates the necessary clearance if the booster doesn't come down straight due to strong winds.

Edit 1: This is also in sync with Elon's credo "the best part is no part" - this is especially true for moving parts.

Edit 2: I would place this thing some 100 feet away from the primary launch tower and launch pad at least. then lift the landed booster out with a tankcilla style crane -- it could be sitting on a fixed mount, but you want an angled boom that can be rotated far away from the incoming booster.  Any solution that has the booster land in close proximity to a large solid tower as some solutions have shown in this thread makes me cringe. If the booster hits that tower oncoming down because of a sudden wind-gust at the worst moment, you have months of downtime to rebuild it.

The ring-tripod on the other hand could be built sturdy enough to survive even a booster crash - repaired quickly, and anything sensitive (crane, launch tower, GSE and propellant lines) would be at least one standard fireball size away from it

[steveleach]

Any solution that has the booster land in close proximity to a large solid tower as some solutions have shown in this thread makes me cringe. If the booster hits that tower oncoming down because of a sudden wind-gust at the worst moment, you have months of downtime to rebuild it.

The ring-tripod on the other hand could be built sturdy enough to survive even a booster crash - repaired quickly, and anything sensitive (crane, launch tower, GSE and propellant lines) would be at least one standard fireball size away from it

On this specific point, I believe that SpaceX are a lot less concerned about this than you might think.

Remember that SpaceX have a lot of experience in dealing with the results of rocket landing crashes. They have a pretty good idea of how much, and what kind of, damage is suffered by various materials at different distances from the crash site.

They also (unfortunately) have some experience of the results of an explosion before take-off, with a fully-fuelled vehicle. And also the problems of dealing with rocket exhaust during a successful take-off, of course.

The landing explosions look very dramatic, but there is a lot less energy involved than there is when a booster is full of propellants.

[Blueshift]

Tension is dramatically easier to engineer for than compression in most ways, because compression buckling failure modes demand so much more structural reinforcement.  Wire ropes pretty much get stronger in proportion to their cross-sectional area, but a rigid structure under compressive loading requires higher than geometric mass scaling.

My thinking too. Someone in an other thread posted a link from a Arianespace landing concept using cable towers. The concept is shown below in step 6.


https://www.esa.int/Enabling_Support/Space_Transportation/ESA_plans_demonstration_of_a_reusable_rocket_stage

It can’t get much simpler and would work with less than perfect landing accuracy. The cable attachment could be movable, like on sliders. As long as the booster comes down within the rectangle defined by the 4 towers, the cable control can adjust to catch it.

6cm thick wire rope has a breaking strength of about 150 tons. Four of them should be enough for a 200-300t booster coming in for landing.

[steveleach]

Tension is dramatically easier to engineer for than compression in most ways, because compression buckling failure modes demand so much more structural reinforcement.  Wire ropes pretty much get stronger in proportion to their cross-sectional area, but a rigid structure under compressive loading requires higher than geometric mass scaling.

My thinking too. Someone up thread posted a link from a Arianespace landing concept using cable towers. The concept is shown below in step 6.

It can’t get much simpler and would work with less than perfect landing accuracy. The cable attachment could be movable, like on sleds. As long as the booster comes down within the rectangle defined by the 4 towers, the cable control can adjust to catch it.

6cm thick wire rope has a breaking strength of about 150 tons. Four of them should be enough for a 200-300t booster coming in for landing.

Interesting concept. That image shows legs on the vehicle, so the cable mechanism clearly isn't designed to remove the need for them, so what is it for?

[Crispy]

Tension is dramatically easier to engineer for than compression in most ways, because compression buckling failure modes demand so much more structural reinforcement.  Wire ropes pretty much get stronger in proportion to their cross-sectional area, but a rigid structure under compressive loading requires higher than geometric mass scaling.

My thinking too. Someone up thread posted a link from a Arianespace landing concept using cable towers. The concept is shown below in step 6.

It can’t get much simpler and would work with less than perfect landing accuracy. The cable attachment could be movable, like on sleds. As long as the booster comes down within the rectangle defined by the 4 towers, the cable control can adjust to catch it.

6cm thick wire rope has a breaking strength of about 150 tons. Four of them should be enough for a 200-300t booster coming in for landing.

Interesting concept. That image shows legs on the vehicle, so the cable mechanism clearly isn't designed to remove the need for them, so what is it for?

Stabilisation. The legs on that booster are quite short, so need extra stabilisation on a moving sea platform

[steveleach]

Tension is dramatically easier to engineer for than compression in most ways, because compression buckling failure modes demand so much more structural reinforcement.  Wire ropes pretty much get stronger in proportion to their cross-sectional area, but a rigid structure under compressive loading requires higher than geometric mass scaling.

My thinking too. Someone up thread posted a link from a Arianespace landing concept using cable towers. The concept is shown below in step 6.

It can’t get much simpler and would work with less than perfect landing accuracy. The cable attachment could be movable, like on sleds. As long as the booster comes down within the rectangle defined by the 4 towers, the cable control can adjust to catch it.

6cm thick wire rope has a breaking strength of about 150 tons. Four of them should be enough for a 200-300t booster coming in for landing.

Interesting concept. That image shows legs on the vehicle, so the cable mechanism clearly isn't designed to remove the need for them, so what is it for?

Stabilisation. The legs on that booster are quite short, so need extra stabilisation on a moving sea platform

So more of an octagrabber alternative than the catching mechanism SpaceX are talking about then?

[hallmh]

6cm thick wire rope has a breaking strength of about 150 tons. Four of them should be enough for a 200-300t booster coming in for landing.

I'm afraid that wouldn't be so - the force along the tensioned horizontal cables would be something like ten times the vertical load. It's a 'triangle of forces' problem.

But I do think the concept is very interesting - could the cables be replaced by I-section beams, like a set of fast-acting bridge cranes? One pair would have to be offset below the other, which would affect the placement of the grid fins.

[Blueshift]

So more of an octagrabber alternative than the catching mechanism SpaceX are talking about then?

It's an illustration to demonstrate the principle of wire catching a stage at the top (gridfins). For SH the towers need to be higher and the cables able to support the weight of the booster obviously.

I'm afraid that wouldn't be so - the force along the tensioned horizontal cables would be something like ten times the vertical load. It's a 'triangle of forces' problem.

The cables can give when the load increases. They don’t need to be horizontal once the booster is caught. When the full load rests on them they can be at quite an angle (depending on ground clearance).  It's the same principle as arresting cables on aircraft carriers.

[steveleach]

So more of an octagrabber alternative than the catching mechanism SpaceX are talking about then?

It's an illustration to demonstrate the principle of wire catching a stage at the top (gridfins). For SH the towers need to be higher and the cables able to support the weight of the booster obviously.

I'm afraid that wouldn't be so - the force along the tensioned horizontal cables would be something like ten times the vertical load. It's a 'triangle of forces' problem.

The cables can give when the load increases. They don’t need to be horizontal once the booster is caught. When the full load rests on them they can be at quite an angle (depending on ground clearance).  It's the same principle as arresting cables on aircraft carriers.

While cable-catching has some interesting properties, it feels like it also has a number of drawbacks. More towers are needed, it is harder to dampen oscillations, etc.

Also, Musk did specifically say "catch the Super Heavy Booster with the launch tower arm".

Personally, I'm imagining a single tower with a single forked arm that is used to both lift SS onto SH before launch, and catch SH on landing before lowering it onto the launch mount.

[rickyramjet]

My concept for a SH catcher: the view is from above the tower looking down and the SH landing in 3 different positions.  The arms pivot from the tower and the catchers can move along the arms.  The arms are kept out of the way until the until the flamey end of SH goes by then they close in on the SH and the catchers move to position themselves under the grid fins.  This allows for landing position errors (which will happen), high wind tolerance, etc.   Apologies for the hand drawn sketch!

[Ionmars]

Whether cable catching or employing arms, we would use ground-based navigation to guide the catcher. We would begin moving the catcher towards the catch point before the vehicle gets there.

This procedure would be implemented when the booster descends to a pre-selected position above the catch point. At this moment its onboard GNC should make mo more adjustments, but should calculate an accurate pathway to touchdown.  Ground-based GNC would pick up control authority to guide the catching devise to the same predicted catch point at the exact catching moment. For example, at T-5 the catcher could arrive at the general area of touchdown but leave a 1m safety margin of error around the vehicle. At T-2 the margin closes to 30cm;  at T-1.5 margin is 10cm; at T-0.5 the catcher lightly touches the side of the ship. 

This capability is particularly important for a mechanical arm or similar slower-moving device.

[Ionmars]

Re: Cable catchers

Cables provide excellent capabilities to move quickly during the catching procedure and resilience to absorb momentum.

The reason that SpX may not choose this approach is that they may want to integrate the catcher with a launch mount for a faster relaunch. For example, they may want to catch the vehicle and immediately carry it over a launch mount. Alternatively, they may build the catcher directly over a launch mount and just lower the vehicle onto the mount after catching it. This would be more difficult with a cable system.

[AC in NC]

While cable-catching has some interesting properties, it feels like it also has a number of drawbacks. More towers are needed, it is harder to dampen oscillations, etc.

Also, Musk did specifically say "catch the Super Heavy Booster with the launch tower arm".

Personally, I'm imagining a single tower with a single forked arm that is used to both lift SS onto SH before launch, and catch SH on landing before lowering it onto the launch mount.

Perhaps it's a lack of imagination on my part but I'm not seeing how a multi-tower cable-catching system can satisfy known requirements.  Before anyone goes too far down that path, it would be good to consider how you are going to get the "arm" to the SH and return it to the pad for launch within an hour.

That single tower with arm(s) does seem like a much stronger approach.

[meekGee]

I'm also not in the cables camp.

A pair of arms that extend from a torso will do the trick.

When the arms swivel in the same direction, they compensate for any lateral deviations in the approach. 

Because they're long, they naturally allow for deviations along the flight path.

As capture happens, they swivel towards each other to ensure they grab the fins near their roots.

The torso is mounted on a linear slide, providing vertical adjustability and shock absorption. It also allows later placement of the SH on the mount

No communication between rocket and tower is necessary IMO.