Relativity Space: General Thread

[HMXHMX]

Here's the image to go with that tweet.

Is it just me or is Relativity Space solving a problem that isn't a problem? 

Manufacturing tanks – conventionally – is probably one of the cheapest parts of space launch.  In fact, the hardware cost of a launch vehicle pales to insignificance against other costs (especially if it is reusable).  Labor (not manufacturing touch labor but testing, program management, regulatory & safety, etc.) is not easily automated, at least in the development stage.  For small vehicle, range costs (using gov't ranges) could approach the hardware costs.  Mission assurance costs add at least 25% to the cost of F9 launches. 

[ringsider]

Here's the image to go with that tweet.

Is it just me or is Relativity Space solving a problem that isn't a problem? 

Manufacturing tanks – conventionally – is probably one of the cheapest parts of space launch.  In fact, the hardware cost of a launch vehicle pales to insignificance against other costs (especially if it is reusable). 

I like this comment so much. This is like noticing the emperor isn't wearing clothes.

[mlindner]

People seem not to be posting recent updates.

https://twitter.com/relativityspace/status/1072546950233382914

Interesting to note how they censored the tool print heads in that image.

https://twitter.com/relativityspace/status/1072866091578286080

https://twitter.com/relativityspace/status/1072866629720010755

https://twitter.com/relativityspace/status/1072867062416953344

https://twitter.com/relativityspace/status/1072882261278617600

[mlindner]

https://twitter.com/relativityspace/status/1073806607660703744

In response to Robotbeat's question if sanding was used:

https://twitter.com/relativityspace/status/1073818455499956229

[HMXHMX]

Following on my earlier post, here's an image of the AirLaunch QuickReach Stage Two tank (LOX-Propane) which appears to be very similar to the Relativity tank show in the updates above.

Relativity suggested in their tweet it takes 12-18 month to fabricate a similar tank.  In fact, the clock time to build this tank was a few weeks (spinning at Spincraft, trimming, and E-beam welding of four parts).  The cost was a few hundred thousand dollars for the prototype and would have been less in quantity.

An amusing historical note: our former test site in the second image is now the Virgin Orbit LauncherOne engine test facility.

[Prettz]

https://twitter.com/relativityspace/status/1072866629720010755

This is an interesting aspect.

[ringsider]

https://twitter.com/relativityspace/status/1072546950233382914

Maybe I am wrong, but 3 weeks printing seems like a long time to make a small tank? Rocket Lab were saying they can do a tank in 2-3 days if I recall. And what about the Stage 1 tank - if that tank took three weeks they are looking at 2-3x that for a bigger tank - 6-9 weeks per Stage 1 plus 3 weeks per stage 2 - so between 9-12 weeks per vehicle? One vehicle every quarter?

I buy the story about reduced part count but not sure of the big time benefit. Am I wrong?

Interesting to note how they censored the tool print heads in that image.

Indeed, could be because it is very special or because it is very agricultural.

[Nomic]

https://twitter.com/relativityspace/status/1073806607660703744

In response to Robotbeat's question if sanding was used:

https://twitter.com/relativityspace/status/1073818455499956229

Odd response as the surface finish is different on the finished item to the in process pictures, so they've clearly done something.

[Bananas_on_Mars]

Since it's most likely black aluminium oxide on the tank, washing with some light acid or using some special etchant(?) might have done the trick...

[LtWigglesworth]

https://twitter.com/relativityspace/status/1073806607660703744

In response to Robotbeat's question if sanding was used:

https://twitter.com/relativityspace/status/1073818455499956229

Odd response as the surface finish is different on the finished item to the in process pictures, so they've clearly done something.

They've also done some post-machining to the top flange.

[TrevorMonty]

https://twitter.com/relativityspace/status/1072546950233382914

Maybe I am wrong, but 3 weeks printing seems like a long time to make a small tank? Rocket Lab were saying they can do a tank in 2-3 days if I recall. And what about the Stage 1 tank - if that tank took three weeks they are looking at 2-3x that for a bigger tank - 6-9 weeks per Stage 1 plus 3 weeks per stage 2 - so between 9-12 weeks per vehicle? One vehicle every quarter?

I buy the story about reduced part count but not sure of the big time benefit. Am I wrong?

Interesting to note how they censored the tool print heads in that image.

Indeed, could be because it is very special or because it is very agricultural.

To scale up just add more 3D printing machines, which is what Rocket Labs do to with their engines.
Print time is still same length but they are printing mutliple engines at same time.

[ringsider]

https://twitter.com/relativityspace/status/1072546950233382914

Maybe I am wrong, but 3 weeks printing seems like a long time to make a small tank? Rocket Lab were saying they can do a tank in 2-3 days if I recall. And what about the Stage 1 tank - if that tank took three weeks they are looking at 2-3x that for a bigger tank - 6-9 weeks per Stage 1 plus 3 weeks per stage 2 - so between 9-12 weeks per vehicle? One vehicle every quarter?

I buy the story about reduced part count but not sure of the big time benefit. Am I wrong?

Interesting to note how they censored the tool print heads in that image.

Indeed, could be because it is very special or because it is very agricultural.

To scale up just add more 3D printing machines, which is what Rocket Labs do to with their engines.
Print time is still same length but they are printing mutliple engines at same time.

Yes, but that is still one complete tank set per quarter per machine. How is that more  efficient than Rocket Lab's carbon tube system? Plus it is basically aluminum from what they say, so I really struggle to see the gain. What is missing?

[TrevorMonty]

https://twitter.com/relativityspace/status/1072546950233382914

Maybe I am wrong, but 3 weeks printing seems like a long time to make a small tank? Rocket Lab were saying they can do a tank in 2-3 days if I recall. And what about the Stage 1 tank - if that tank took three weeks they are looking at 2-3x that for a bigger tank - 6-9 weeks per Stage 1 plus 3 weeks per stage 2 - so between 9-12 weeks per vehicle? One vehicle every quarter?

I buy the story about reduced part count but not sure of the big time benefit. Am I wrong?

Interesting to note how they censored the tool print heads in that image.

Indeed, could be because it is very special or because it is very agricultural.

To scale up just add more 3D printing machines, which is what Rocket Labs do to with their engines.
Print time is still same length but they are printing mutliple engines at same time.

Yes, but that is still one complete tank set per quarter per machine. How is that more  efficient than Rocket Lab's carbon tube system? Plus it is basically aluminum from what they say, so I really struggle to see the gain. What is missing?

You missed point, they can build multiple tanks at same time, just need more 3d printers. 1st and 2nd stage tanks would be printed at same time.

No tooling modifications required if they want to change tank design, just a different  print file. As they said can easily add braffles and other internal features.

Not sure how RL make their tanks but it is possible to print large carbon fibre pressure vessels. US Navy demo this with mini sub.

[playadelmars]

Having come from Blue Origin previously, and a lot of team from SpaceX, seems like they are playing the long game. Printers will get faster over time, and no fixed tooling let’s them change designs faster than possible otherwise while simultaneously lowering part count and automating away a good deal of labor costs. If they add reusability in the future then printing could both come up with better and more iterative designs, and also help replace spare parts/upper stages quickly. The bet then is SpaceX and Blue Origin or others making smaller reusable rockets (if Rocket Lab goes that way) won’t be able to figure out rapid low cost reuse quickly or effectively enough precisely because their manufacturing methods have slower iteration cycles. I could easily see something with wings or a slightly non-circular tank cross section for lifting body being possible with Relativity tech in 5 years.

Relativity’s approach also drives toward their long term vision of printing on Mars and other planets, since those benefits are all needed to take advantage of ISRU one day.

As to whether that approach will work or not, we shall see, but I’m glad to see at least one company in the small launch space have a big vision that extends BEO!

[Bananas_on_Mars]

There's also the possibility that through process monitoring and perhaps a pressure test, they can forego Non-Destructive Testing in the future. That's something that's often forgotten about additive manufacturing, once you have your process worked out and are able to monitor all relevant parameters, there's no need to check for hidden defects as you  check every layer of material laid down.

Less assembly also means less chances for human error.

They might be able to upgrade their printers performance simply by adding more robots with printheads.
And they don't have to train their workforce as they scale up, or maybe also scale down. Because if demand for their rocket is low, they even might be able to survive on a handful of rocket launches per year if necessary.

[TrevorMonty]

Having come from Blue Origin previously, and a lot of team from SpaceX, seems like they are playing the long game. Printers will get faster over time, and no fixed tooling let’s them change designs faster than possible otherwise while simultaneously lowering part count and automating away a good deal of labor costs. If they add reusability in the future then printing could both come up with better and more iterative designs, and also help replace spare parts/upper stages quickly. The bet then is SpaceX and Blue Origin or others making smaller reusable rockets (if Rocket Lab goes that way) won’t be able to figure out rapid low cost reuse quickly or effectively enough precisely because their manufacturing methods have slower iteration cycles. I could easily see something with wings or a slightly non-circular tank cross section for lifting body being possible with Relativity tech in 5 years.

Relativity’s approach also drives toward their long term vision of printing on Mars and other planets, since those benefits are all needed to take advantage of ISRU one day.

As to whether that approach will work or not, we shall see, but I’m glad to see at least one company in the small launch space have a big vision that extends BEO!

Even if their LV falls by wayside their 3D printing  technology should find another market, which makes them a better investment that most small LV startups. There are few examples of companies developing a new technology for inhouse problem to eventually switch directions and make new technology their main business focus.

NB Blue have developed automated drilling machine for building NG tanks, which they spin off as separate business line.

[Ronsmytheiii]

https://twitter.com/relativityspace/status/1078719280437436416

[Asteroza]

https://twitter.com/relativityspace/status/1078719280437436416

So people don't get confused, as a followup tweet points out, these are conventional test stand tanks acquired elsewhere. These storage tanks were not 3D printed.

[gongora]

[CNBC] Air Force grants 3D rocket printer Relativity Space a ‘premier’ launch pad in Florida

Relativity Space signs five-year agreement to use Launch Complex 16 at Cape Canaveral Air Force Station.

https://twitter.com/thesheetztweetz/status/1085929959028387841

[gongora]

[Ars Technica] Relativity Space to launch from historic Florida site

Under terms of the competitively awarded agreement, the site will officially be a “multiuser” facility for five years. However, if Relativity meets certain milestones and begins regularly launching rockets, it will be able to convert the agreement into a 20-year, exclusive right to use the launch site.