SpinLaunch: General Company and Development Updates and Discussions

[jongoff]

One of the commenters over at Ars Technica, Wickwick, has insider knowledge about SpinLaunch. Although they're under NDA so they can't share too many juicy details, they've been giving hints about SpinLaunch's plans. I've compiled the most useful ones below. The original comments can be found in the comments of this article starting on page 11 or so (bold mine).

It's enough that it's not out of the question that instead of releasing a counterweight simultaneously with the projectile, you can wait half a turn and release a second projectile out the same exit tube.

I'm glad he mentioned that. I had heard the same thing. Two for the price of one. :-) Extra awesome for lunar ISRU launch w/o needing to toss a counterweight every time. Though I do wonder for terrestrial launch if that would complicate closing the door to the vacuum chamber, since you'd have to wait an extra ~65ms for the second projectile.

ivekadi wrote:
Re: SpinLaunch:

I'd like to point out that the aforementioned exit velocity of 2 235 m/s is ~140 m/s short of lunar escape velocity. As for (lunar) orbit, minimal extra dV will be required to raise the perigee. As such, this is very promising device.

They are very aware of that.

Yeah, I know I mentioned it to them before, but I'm sure others have as well.

~Jon

[jongoff]

Ok, I think having the last five comments in a row is a sign I've said enough. I love the idea though and am glad they're making steady progress.

[Lars-J]

The videos are cool. But I’m still skeptical that this will be useful for space launch, and I don’t think SpinLaunch will ever launch something to orbit.

They may have retired a lot of individual risks, but that doesn’t mean that all or the combined risk is retired. I think building a practical 2nd stage that A) survives 10k Gs for a long time, B) survives the the atmospheric shock and ascent and is able to start and run after all that, and finally C) has enough performance to put anything in orbit is going to be a step too far. Individually these are very solvable problems, but together? Good luck. (Prove me wrong, SpinLaunch!)

I still think (like many others) that there is some other aspect of this that we are not fully seeing that is driving the funding, be it as a weapons system or something else. Until there is real work and hardware in progress for the rocket stage, this remains a cool way to throw things pretty far.

[TrevorMonty]

Critical point of failure is release mechanism. If timing is out slightly due to sticking mechanism it goodbye LV and launch pad. Would put a whole new spin on RUD.

Sent from my SM-G570Y using Tapatalk

[TrevorMonty]

If they can get close to F9 $ per Kg it would be perfect for supplying new small fuel depots that a planning to launch in next few years. These depots won't be needing tons of fuel initially not at least till market picks up.

Sent from my SM-G570Y using Tapatalk

[jongoff]

Critical point of failure is release mechanism. If timing is out slightly due to sticking mechanism it goodbye LV and launch pad. Would put a whole new spin on RUD.

They have a clever solution for this that makes this type of failure extremely unlikely. But it's not public info yet AFAICT.

But trust me, the fact that most people home in on the release mechanism reliability and timing being absolutely critical isn't something that a group of ~100ppl working on this for several years somehow overlooked. :-)

~Jon

[jongoff]

If they can get close to F9 $ per Kg it would be perfect for supplying new small fuel depots that a planning to launch in next few years. These depots won't be needing tons of fuel initially not at least till market picks up.

Potentially. The price point I've seen (not sure if they've publicized it or not) would have it competitive on $/kg with current reusable F9 prices (with the F9 fully loaded). The only problem is that at this small of a size, the percentage of launch mass that ends up as net payload/delivered propellant might not be as good -- interface overhead can eat your lunch if you get too small. That said, it's an interesting possibility.

It could be a good solution for launching propellant to refuel servicer/tug spacecraft though. In LEO you don't want to have to move the tug around a bunch to get to a refueling point, so having something that could on-demand get a decent prop load right to where a tug needs it to be, could be pretty interesting.

~Jon

[niwax]

If they can get close to F9 $ per Kg it would be perfect for supplying new small fuel depots that a planning to launch in next few years. These depots won't be needing tons of fuel initially not at least till market picks up.

Sent from my SM-G570Y using Tapatalk

There is one aspect here that gets lost in the rapid-fire projections is the constraints of orbital mechanics. Unless you want to have hundreds of tiny depots, there is at the very most one launch opportunity a day.

[meekGee]

Critical point of failure is release mechanism. If timing is out slightly due to sticking mechanism it goodbye LV and launch pad. Would put a whole new spin on RUD.

Sent from my SM-G570Y using Tapatalk

That's an immensely complex and high risk pun right there.

[Yggdrasill]

There is one aspect here that gets lost in the rapid-fire projections is the constraints of orbital mechanics. Unless you want to have hundreds of tiny depots, there is at the very most one launch opportunity a day.

Not necessarily. If the propellant depot has a 0 degree inclination and the launch site is on the equator, it's more like 16 opportunities per day.

[Welsh Dragon]

There is one aspect here that gets lost in the rapid-fire projections is the constraints of orbital mechanics. Unless you want to have hundreds of tiny depots, there is at the very most one launch opportunity a day.

Not necessarily. If the propellant depot has a 0 degree inclination and the launch site is on the equator, it's more like 16 opportunities per day.

Who  needs a depot at 0 degrees?

[Craftyatom]

There is one aspect here that gets lost in the rapid-fire projections is the constraints of orbital mechanics. Unless you want to have hundreds of tiny depots, there is at the very most one launch opportunity a day.

Not necessarily. If the propellant depot has a 0 degree inclination and the launch site is on the equator, it's more like 16 opportunities per day.

Here we come across the problems with small depots (IMO).

Basically, if the depot only provides a small amount of propellant relative to the satellite's mass, it's more efficient to simply launch the satellite with more propellant, or a kick stage.  You can improve this balance by making the satellite smaller, which makes propellant more valuable per kg - but if you make your satellite smaller, then existing cheap launchers can already probably give it enough dV to get it to where it's needed, especially with a kick stage.  You could also put your propellant depot in a higher/more inclined orbit to increase its value, but the further out it is, 1) the more dV you need to refuel it in the first place (ruling out some small launch options), and 2) the more it constrains your launch window.  In that sense, what you'd like to do is have an on-demand depot launched by a larger launcher into exactly the orbit you need: this is called distributed launch.

As the depot size grows, you can start to justify siting it in lower orbits and using it for heavier satellites, at which point it begins to overtake single-launch options, even with kick stages; sending a 500 ton monolith to Lunar orbit is worth paying for a depot in LEO (which may take the form of multiple-launch upper-stage refueling without any permanent infrastructure) because there simply isn't a rocket powerful enough, even with kick stages, to send it all in 1 or 2 launches.  But that's beyond the scope of SpinLaunch, at least at present, and thus I personally am not sold on any "bulk propellant" projections as of yet.

What seems more promising to me is if they make Satellite-as-a-Service offerings.  They already have to be good at satellite manufacturing because nobody else will make/certify components for the appropriate G-loads.  Their launches are small and cheap, which is exactly what you want out of a tech demo or initial use-case.  Their launch schedule can be quite tight, which is great for prototypes, especially if you want to fail fast - heck, launching two rockets in one shot would be a great redundancy deal!  The only downside is that some cutting-edge technology might not be suitable for these G-loads, so you might not be comfortable testing your ultra-fine hyperspectral optics or whatever with them.

[Asteroza]

In LEO you don't want to have to move the tug around a bunch to get to a refueling point, so having something that could on-demand get a decent prop load right to where a tug needs it to be, could be pretty interesting.

~Jon

This is that in between market between propellant-as-a-service and deltaV-as-a-service isn't it.

Which has the interesting corollary that if you can do beamed power delivery, you can also do hotrod tug deltaV-as-a-service.

[Asteroza]

There is one aspect here that gets lost in the rapid-fire projections is the constraints of orbital mechanics. Unless you want to have hundreds of tiny depots, there is at the very most one launch opportunity a day.

Not necessarily. If the propellant depot has a 0 degree inclination and the launch site is on the equator, it's more like 16 opportunities per day.

Who  needs a depot at 0 degrees?

Technically nobody needs a LEO equatorial depot, right now anyways.

What it does provide is easier bulk delivery to LEO due to launch site revisit rate. You would then need a OTV refueler to haul propellant that would go higher to GTO, then an apogee plane change where it's cheaper deltaV wise to your actual customer if not GEO. You may be able cheat with an aerodynamic plane change if your OTV is so equipped.

The basic question becomes, is the bulk propellant Kg to 0 degree LEO depot plus the costs associated with last mile delivery from the depot by OTV better than direct launches to the customer. If launching from Al Cantara, you do have direct access to many customer orbits. But you would need to demonstrate a refueling OTV that can survive those accelerations. Arms to grapple a DogTag or similar that can survive launch seem like a hard engineering problem. Still, a spot refueling regime may be attractive for long duration commercial assets. Even if the amount delivered isn't large, if you can expect yearly delivery, it could work out.

Unfortunately with the recent dogma of many satellites with shorter lifetimes, the spot small refueling commercial market may not grow much. Spot refueling LEO/MEO military surveillance sats would be a promising market, like the SDA constellation, since there could be a need to expend a lot of deltaV regularly due to tasking (but that argument crumbles when you have enough sats for 24/7 global high resolution surveillance without gaps).

But then that goes up against the basic question, if SS/SH is also launching from Al Cantara, you probably lose the bulk propellant kg  to LEO argument. Then the fight shifts from Spinlaunch delivering propellant direct to customer, versus a simpler reusable last mile OTV refueler delivering from a SS/SH supplied 0 degree LEO depot.

[eeergo]

They chose Spaceport America because they can lob rockets over the mountains into White Sands Missile Range, not because they were afraid of sea level pressures. Compared to vacuum 12psi isn't that different from 14.7psi. Based on what I've seen (which is a lot more than is public), I'm pretty optimistic they can get their system to work.

[...]

They've already done tests where they've gotten up to and released subscale test articles at the full 2.3km/s tip velocity they're planning for their full-scale system, in their subscale chambers in the Bay Area and I think down in Long Beach as well. There's stuff that'll need to be dialed in to get the system working reliably for operational missions, but they've retired most of the show-stopper risks at this point.

[...]

I don't think people realize how little this thing decelerates -- it has a very high ballistic coefficient. Lots of mass per unit frontal area with a very aerodynamic shape. And it gets out of the "soup" really, really darned fast. But yeah, hopefully sloshing has stopped by the time they go to light the engines? Worst case they could use cold gas to deliberately settle the tanks before lighting. It wouldn't take much cold gas thrust to do that.

But I want to see them make orbit, if for no other reason than making all the cocky naysayers eat some crow. :-D

~Jon

Hi Jon, let me start by saying I really appreciate you taking time to answer to this conversation with your very valuable (and better-informed-than-most) input. Let me address a couple of issues I still remain unconvinced of after your answers:

- My comment of the altitude difference between White Sands and sea level was not because of the pressure difference between the vacuum in their centrifuge and ambient, but because of the extra kilometer, and atmospheric density, to clear with their projectile, during which time it will accumulate the largest drag losses, heating and dynamic stresses. That right there is a substantial risk that is not being cleared or even studied with these tests.

- An even-smaller-scale short-range laboratory test for this kind of system doesn't seem like an ideal confidence-builder for the actual release mechanism, much less to claim its risk has been retired... and that is just the very split-second, easily-repeatable and conveniently-iterationable first step without which the system cannot even be tested. Either way, I was discussing the relevance of the public tests for risk retirement, not other eventual developments that remain completely under wraps.

- Honest question: how can they make it to dissipate to drag less violently than, say, Sprint, which clearly had enormous drag losses that made it incandescent, even when launching from a 1km-high mesa? In fact, it should be going much faster much lower, and the shape of the projectile they showed isn't that different...

Anyway, I'd love for the concept to work too - seems much more environmentally-friendly and generally "cleaner" than atmospheric rockets. It still seems like it's running against fundamental huge physical inefficiencies.

[fatjohn1408]

- Honest question: how can they make it to dissipate to drag less violently than, say, Sprint, which clearly had enormous drag losses that made it incandescent, even when launching from a 1km-high mesa? In fact, it should be going much faster much lower, and the shape of the projectile they showed isn't that different...

Perhaps counterintuitively by making the outer shell heavier.
The heavier your package is the more kinetic energy it will pack at a certain speed.
The drag force will remain unchanged as long as the diameter doesn't change much but the drag acceleration or drag loss will thus be lower as your package gets heavier.

Also the key is to make it as slender as possible.
For the rest, the casing just needs to be able to deal with the heat but since it's not at all bad for it to be heavy that should not really be an issue. You could just smack an ablative coating on it and be done with it.

I did an optimization of different assisted launch speeds and altitudes once trying to optimize both trajectory and (single stage) vehicle parameters for maximum payload ratio and the difference of launching a 2 km/s single stage from sea level compared to 5000 meters altitude is about a 7% higher payload ratio.

So not a too large difference, for comparison, the difference between a 2 km/s assist and a 1.5 km/s assist was about a 22.5% boost in payload ratio.

So launching it at sea level or 2 km up is not a big deal, provided they are able to hit 2 km/s exit velocity.
If they would need to drastically reduce their system's exit velocity to say below 1 km/s I think launch altitude will start playing a larger role from a performance point of view.

[fatjohn1408]

SpinLaunch expands New Mexico test site

At a presentation last month during a meeting of the board of directors of the New Mexico Spaceport Authority, Scott McLaughlin, acting executive director of Spaceport America, said the company is building an evacuated centrifuge 30 meters in diameter. That will be able to accelerate objects to Mach 5 before “catapulting” it out a door.

The facility is intended for use in suborbital tests, which McLaughlin said will be done in cooperation with nearby White Sands Missile Range. Objects launched from the centrifuge will go to an altitude of about 100 kilometers before landing at White Sands. Those tests will begin some time in 2021, he said, the same time frame the company stated in the announcement of its expansion there.

a 30 m diameter vacuum chamber flinging a spacecraft at mach 5 into a wall of atmosphere rushing in at mach 1. That could be very exciting footage.

There is already footage of similar stuff.



Mach 5 would be a bit tame in comparison.

Where is the vacuum chamber in that video? I didn't see it.

Behind the sled.

[eeergo]

- Honest question: how can they make it to dissipate to drag less violently than, say, Sprint, which clearly had enormous drag losses that made it incandescent, even when launching from a 1km-high mesa? In fact, it should be going much faster much lower, and the shape of the projectile they showed isn't that different...

Perhaps counterintuitively by making the outer shell heavier.
The heavier your package is the more kinetic energy it will pack at a certain speed.
The drag force will remain unchanged as long as the diameter doesn't change much but the drag acceleration or drag loss will thus be lower as your package gets heavier.

Also the key is to make it as slender as possible.
For the rest, the casing just needs to be able to deal with the heat but since it's not at all bad for it to be heavy that should not really be an issue. You could just smack an ablative coating on it and be done with it.

I did an optimization of different assisted launch speeds and altitudes once trying to optimize both trajectory and (single stage) vehicle parameters for maximum payload ratio and the difference of launching a 2 km/s single stage from sea level compared to 5000 meters altitude is about a 7% higher payload ratio.

So not a too large difference, for comparison, the difference between a 2 km/s assist and a 1.5 km/s assist was about a 22.5% boost in payload ratio.

So launching it at sea level or 2 km up is not a big deal, provided they are able to hit 2 km/s exit velocity.
If they would need to drastically reduce their system's exit velocity to say below 1 km/s I think launch altitude will start playing a larger role from a performance point of view.

I wasn't referring to the  heat load being prohibitive (which might also be a factor), I was referring to the enormous drag force it is experiencing while it's flying under such conditions. For your 7% figure, you're assuming the assisted launch will get you at 2 km/s once in space. They're having a 2.1 km/s goal for *exit* velocity, which works out to 1.7 km/s at altitude *with no drag losses* - which I'm arguing in my post will be substantial, just by looking at things like Sprint, because that thing was accelerating from 0 to manyMach while in the lower atmosphere, and still glowed red because of friction, while still accelerating. SpinLaunch's projectiles will have their maximum speeds at release, and decelerate from then until vacuum altitudes, so their losses will start at maximum (much beyond Sprint's max losses, I postulate) and decrease exponentially. Can't see how that won't be a substantial %, which will get you into (I suspect) 1.5 km/s territory by the point the chemical stage can ignite.

Regarding the possibility of making it heavier so that it has larger momentum (and potentially helping with the heat load dissipation), sure - but then your centrifuging issues also scale up.

[Robotbeat]

That confirms the shock-resistance of SMT, for long duration loading that's down to creep-resistance of the solder alloy (and/or the conformal coating if you assume that will take the majority of the load). Some existing research (e.g. https://doi.org/10.1108/eb037718) at kilogee levels shows that some lead-free solder alloys are highly creep resistant (e.g. 96.5Sn/3.5Ag having around half a percent creep rate), and test electronics can be placed within an ultracentrifuge if you want to perform direct testing under design g loads for long periods.

If the problem is creep due to long spin up time, then reduce the spin up time. Let’s say it takes 120 minutes to spin up to 2km/s. If the projectile is 10 tons and the whole mass including the projectile is 100 tons, then the effective energy is ~.5*(100000kg)*(2km/s)^2 ( less than that since most of the mass is relatively close to the center of rotation), so over 120 minutes it takes 28MW. Over 12 minutes it takes 280MW. One Tesla Model S puts out 1MW peak power, for scale.

A cheap brushless motor, driver, and backing battery cost about 10˘/Watt, so you’re looking at about $30 million in electric power train costs vs $3 million.

$30M is small for aerospace, BUT this launcher is supposed to try to get to well under $500,000 launch costs. So that’s ~60 launches in revenue.

[fatjohn1408]

I wasn't referring to the  heat load being prohibitive (which might also be a factor), I was referring to the enormous drag force it is experiencing while it's flying under such conditions. For your 7% figure, you're assuming the assisted launch will get you at 2 km/s once in space. They're having a 2.1 km/s goal for *exit* velocity, which works out to 1.7 km/s at altitude *with no drag losses* - which I'm arguing in my post will be substantial, just by looking at things like Sprint, because that thing was accelerating from 0 to manyMach while in the lower atmosphere, and still glowed red because of friction, while still accelerating. SpinLaunch's projectiles will have their maximum speeds at release, and decelerate from then until vacuum altitudes, so their losses will start at maximum (much beyond Sprint's max losses, I postulate) and decrease exponentially. Can't see how that won't be a substantial %, which will get you into (I suspect) 1.5 km/s territory by the point the chemical stage can ignite.

Regarding the possibility of making it heavier so that it has larger momentum (and potentially helping with the heat load dissipation), sure - but then your centrifuging issues also scale up.

"For your 7% figure, you're assuming the assisted launch will get you at 2 km/s once in space."
No this was a master thesis for my personal master in Space Engineering, not a back of the enveloppe thing i just thought about this afternoon.
It included atmospheric models, gravity model, rotating earth etc, etc.
I optimized 11 different parameters of both the vehicle (chamber pressure, oxidizer to fuel ratio, nozzle exit pressure, nozzle exit diameter, length of the tanks, diameter of the tanks) as well as the trajectory (initial flight path angle and 4 parameters to model the pitch behavior).
I did this for several combinations of initial velocity and initial altitude. (i.e. 2km/s, 0m altitude or 2km/s, 5000m altitude) for both kerolox and hydrolox single stages.
So no I was not assuming the assisted launch would get me 2 km/s once in space. I assume it will get me 2 km/s at sea level.
At which point optimum inclination or initial flight path angle would be 39.9 degrees for example.
Also, since I did not limit mass but simply optimized for highest payload ratio in orbit, my optimization selected a 243 ton kerolox stage and a 233 ton hydrolox stage as the optimal solutions for this type of assist. Much heavier than for example the 54 and 30 ton vehicles it spat out when I optimized an assist of 2 km/s at 40 km altitude. This difference was clearly due to trying to limit drag losses by sizing up the vehicle.

Now the drag loss for kerolox was calculated to be 238 m/s whereas the hydrolox vehicle had a drag loss of 341 m/s.
Since spin launch will probably require an order of magnitude smaller vehicle the drag losses will become a lot more extreme than that.
Drag losses follow the cube square law meaning that as your mass decreases with the cube of your dimensions, drag force only decreases with the square of your dimensions. However the drag coefficient will be a lot better due to the shape of to cocoon they launch their vehicle in.

I'd guesstimate that for a 20 ton vehicle, the drag losses of a full kerolox system would be in the 350 m/s range, whilst for hydrolox in the 500 m/s range. It might be prudent to consider higher density propellants (solid) that could get it down to 250 m/s though.

It is worth noting that small launch vehicles also have considerable drag losses of probably in excess of 200m/s and that the launch assist will likely help keeping gravity losses low.
So I don't think it will be a dealbreaker. Just make it slender, dense, aerodynamic and as heavy as possible.