Relativity Space: General Thread

[ShawnGSE]

Came across plans for Relativity Launch Complex 16 which I thought might be of interest to followers of this thread.  Documents were submitted to the St Johns River Water Management District yesterday - they can be found here: https://permitting.sjrwmd.com/epermitting/jsp/Search.do?theAction=searchDetail&permitNumber=162674

I've attached a couple of plans showing the site location and layout.

So, are they building the launch complex with "zero human labor?"

While that's worth a giggle they are currently hiring for pad staff in FL.  They will operate much like SpaceX with a TE that lifts the rocket vertical at the pad.  Going to be a lot of work even at half the size.

[HMXHMX]

Came across plans for Relativity Launch Complex 16 which I thought might be of interest to followers of this thread.  Documents were submitted to the St Johns River Water Management District yesterday - they can be found here: https://permitting.sjrwmd.com/epermitting/jsp/Search.do?theAction=searchDetail&permitNumber=162674

I've attached a couple of plans showing the site location and layout.

So, are they building the launch complex with "zero human labor?"

While that's worth a giggle they are currently hiring for pad staff in FL.  They will operate much like SpaceX with a TE that lifts the rocket vertical at the pad.  Going to be a lot of work even at half the size.

One thing that most folks don't realize is this fairly simple truth: it takes the same number of people to launch  vehicle that places 100 kg in orbit, as one that places 100,000 kg into orbit.  And, to a first order, the same number of people to design either vehicle, as well.  Labor is 80% of the production and operational cost of a launch vehicle, so it is always better to go larger rather than smaller, when you can.  Up to the point where infrastructure gets in the way...

[Asteroza]

Came across plans for Relativity Launch Complex 16 which I thought might be of interest to followers of this thread.  Documents were submitted to the St Johns River Water Management District yesterday - they can be found here: https://permitting.sjrwmd.com/epermitting/jsp/Search.do?theAction=searchDetail&permitNumber=162674

I've attached a couple of plans showing the site location and layout.

So, are they building the launch complex with "zero human labor?"

While that's worth a giggle they are currently hiring for pad staff in FL.  They will operate much like SpaceX with a TE that lifts the rocket vertical at the pad.  Going to be a lot of work even at half the size.

One thing that most folks don't realize is this fairly simple truth: it takes the same number of people to launch  vehicle that places 100 kg in orbit, as one that places 100,000 kg into orbit.  And, to a first order, the same number of people to design either vehicle, as well.  Labor is 80% of the production and operational cost of a launch vehicle, so it is always better to go larger rather than smaller, when you can.  Up to the point where infrastructure gets in the way...

The lower limit usually being about 2 people for solid rockets, at least for direct launch ops...

I remember an anecdote that the japanese Epsilon required 5 people for launch ops. For some ICBM TEL's it's 2-3 people. I suppose how much hand holding a payload needs probably starts to add to the number. What's a general lower bound for liquid fueled rockets?

[HMXHMX]

One thing that most folks don't realize is this fairly simple truth: it takes the same number of people to launch  vehicle that places 100 kg in orbit, as one that places 100,000 kg into orbit.  And, to a first order, the same number of people to design either vehicle, as well.  Labor is 80% of the production and operational cost of a launch vehicle, so it is always better to go larger rather than smaller, when you can.  Up to the point where infrastructure gets in the way...

Hold up, you changed subjects in the middle there. You started talking about fixed launch costs and design costs not changing much with rocket size. You then switched to talking about the overall cost of a rocket launch being 80% labor, now suddenly including production costs as well. For non reusable rockets the production cost is much more than the fixed launch costs (and design costs are divided by the total number of flights ever, so that has to do with price and ROI timescale, it is not a production or operational cost.) And at least for rockets with comparable build methodologies then the production cost will scale with rocket size.

Relativity is aiming for major reductions in the amount of labor required to build the rocket, since building the rocket is the driving cost for most rockets, and you just stated that most  of the build cost is labor, so that seems like a reasonable plan addressing the biggest cost driver

Fair enough.  Let's focus only on build costs, since they dominate the calculus (and also shows why reusables that don't require much refurbishment post-flight are game-changing).  I maintain that labor is by far and away the largest contributor to launch cost at "OldSpace" companies (as well as "NewSpace" firms that revel in high head counts).  But it is possible to reduce labor costs by properly managing a supply chain and exploiting a vendor's existing capability without vertically integrating one's development by fabricating your own structures.  I managed to build a prototype common bulkhead tank set almost exactly the same as Relativity's recent build in four months from design to delivery with one engineer on staff responsible for the work for around $300k.  Subsequent articles would have been half the price and half the time to delivery.  The vendor used custom tooling and two-four people to fab on a part-time basis.  It worked fine, was lightweight and functional and required no post-processing to be integrated into the stage assembly ready for testing. How is Relativity better than that?  They're solving a problem that doesn't exist and gullible investors haven't done their due diligence to realize that.

[TrevorMonty]

I don't see how they are going dramatically reduce labour cost on build and compared to competitors. Everybody is 3d printing engines, still need same labour input to mount and plumb the engines, build and fix wiring harnesses, launch it.
Don't know if RL 3d print Electron tanks and fairings but they've automated machining.

[whitelancer64]

I don't see how they are going dramatically reduce labour cost on build and compared to competitors. Everybody is 3d printing engines, still need same labour input to mount and plumb the engines, build and fix wiring harnesses, launch it.
Don't know if RL 3d print Electron tanks and fairings but they've automated machining.

Rocket Lab's tanks and fairings are made of carbon fiber composite, they aren't metallic so they can't be 3D printed.

Various articles I've read give a number around 100x lower parts count for the Terran rocket than if it were made conventionally. Rockets normally have many tens of thousands of parts, Relativity says that Terran's parts count is less than one thousand. They achieve this by 3D printing a lot of sub-assemblies as one unit, IOW one of their components is often the equivalent of dozens of component parts in other rockets.

As you note, the entire industry is going in that direction with 3D printing of engine parts, but Relativity is going a step further and 3D printing a lot of the tank's structures as well. So all the tank's internal struts, mount points, etc. are already going to be part of the assembly, and will basically be ready to bolt together once completed. So consider how struts are installed, for example, at SpaceX. The Falcon 9 tank, once fabricated, needs the strut mount points welded in place. The struts then use bolts, nuts, and washers to be installed in the tank. That could be a dozen parts per strut, and there are something like a hundred struts in the Falcon 9, and all of those parts and all of that labor required to install them is eliminated for the Terran rocket.

So yeah they will have some plumbing and wiring work to do, along with assembling things that (as yet) can't be 3D printed as one unit, but it will be much less than if it were designed and built normally. They are targeting less than 60 days from raw material arriving into the factory to completed rocket shipped out.

[whitelancer64]

*snip*
I managed to build a prototype common bulkhead tank set almost exactly the same as Relativity's recent build in four months from design to delivery with one engineer on staff responsible for the work for around $300k.  Subsequent articles would have been half the price and half the time to delivery.  The vendor used custom tooling and two-four people to fab on a part-time basis.  It worked fine, was lightweight and functional and required no post-processing to be integrated into the stage assembly ready for testing. How is Relativity better than that?  They're solving a problem that doesn't exist and gullible investors haven't done their due diligence to realize that.

Relativity would eliminate the vendor, the custom tooling, and the 2-4 people needed to fabricate the assembly. Ideally, they would incorporate the bulkhead assembly into the tank itself during the 3d print.

[ringsider]

... design costs are divided by the total number of flights ever, so that has to do with price and ROI timescale...

This is not the correct way to think about this. Many do, but they are wrong.

Even if we accept that it is the right way it's still an awful method because nobody knows the denominator of the equation (R&D costs/lifetime # flights) until the final flight is flown, at which point you could be positively or negatively surprised.

Here's an exercise: how much of Falcon 9's R&D cost is recovered on every flight, do you think? What's the number SpaceX uses? $1m? $5m? $20m? More? Less? Put a figure on it.

[meberbs]

... design costs are divided by the total number of flights ever, so that has to do with price and ROI timescale...

This is not the correct way to think about this. Many do, but they are wrong.

Even if we accept that it is the right way it's still an awful method because nobody knows the denominator of the equation (R&D costs/lifetime # flights) until the final flight is flown, at which point you could be positively or negatively surprised.

Here's an exercise: how much of Falcon 9's R&D cost is recovered on every flight, do you think? What's the number SpaceX uses? $1m? $5m? $20m? More? Less? Put a figure on it.

You appear to have entirely missed the point of my post. As I said, accounting for development costs has to do with ROI timescale, but the exact assumptions and way you account for ROI is not something I attempted to describe because it is irrelevant. The point in the context that you left out was that HMXHMX bringing up development costs was not actually relevant to the final point he made which was that the majority of launch costs typically are labor involved in building the rocket. In context I was just briefly mentioning the development costs to point out that he changed subject from development costs to instead talk about total per launch costs, in a sense that excludes the development costs, so his original mention of development costs simply did not apply to what he said at the end.

[playadelmars]

The way VC works, development costs are sunk costs and the value of the business has more to do with discounted future cash flows. Relativity’s printing tech, if it works as they say, would greatly lower labor costs for all future rockets (and anything else they choose to print) meaning the discounted future value of the company is already large in theory, just not proven yet. Development costs themselves do not matter as long as investors are willing to fund them at increasing valuations with an eye toward taking a slice of those very lucrative future cash flows while they hit proof points.

Also talked about less, but 3d printing is such a nascent technology still in its early days of possibility. Traditional manufacturing has largely been honed over 50-100+ years and is already so refined that it’s not going to grow as quickly in progress as printing. Terran 1 as a first version doesn’t have to be light years better itself. Printing will get far faster and cheaper over time, and if Relativity is at this forefront for rockets then it will be positioned to keep compounding upon its success more quickly than others. It’s a new paradigm, and since businesses are valued on this future potential that’s why investors are buying into it. Doesn’t hurt that several sophisticated customers, the US government, and a pretty impressively experienced team all agree there’s potential here enough to join. This is similar to why Tesla is worth what it is, or really any tech company - the future is electric/software/robotics/etc, and these companies are building that future.

[edzieba]

The actual assembly is only part of the question, there's also test and inspection which can also be automated. Here 3D printing has the nice advantage that you can nondestructively inspect the entire cross-section of the product in the manner you could normally only achieve with a destructive slice. e.g. you can verify dimensionality as material is laid down rather than after-the-fact with laser probing, you can detect porosity and inclusions during material laydown rather than after the fact with X-ray or ultrasonic testing, etc. You can do this in real-time and catch defects at moment of manufacture, either rejecting a part before even needing to complete manufacture or potentially repairing the defect in place.
Most of Relativity's work is not going to be in the additive manufacture process itself (while direct metal deposition is not quite trivial, it's not novel or unique either) but in the test and verification during manufacture and the backend modelling that allows for design and simulation. This is mature in subtractive manufacturing, but much less so when it comes to additive manufacture and often 'bodged on' (e.g. taking an AM part and subjecting it to the same testing regime you would a milled part).

[playadelmars]

https://twitter.com/relativityspace/status/1271175937309020160?s=21

[meberbs]

Fair enough.  Let's focus only on build costs, since they dominate the calculus (and also shows why reusables that don't require much refurbishment post-flight are game-changing).  I maintain that labor is by far and away the largest contributor to launch cost at "OldSpace" companies (as well as "NewSpace" firms that revel in high head counts).
...
They're solving a problem that doesn't exist and gullible investors haven't done their due diligence to realize that.

These statements from you are contradictory. The first one states the exact problem that they are trying to solve, high labor costs for building rockets.

And no, your anecdote does not demonstrate this problem being solved. Actual rocket stages that fly, even from companies that focused on massively reducing production cost, still cost millions of dollars. Your statement about "subsequent articles would have been" implies that this was abandoned for some reason and will never fly, there are dozens of launch startups that think they have a better way, flying is what will ultimately prove whether their plans work or not. Relativity is at least still progressing towards flight.

As we both agree, re-use is a proven way to address this problem, but that doesn't necessarily exclude other solutions, especially ones that don't necessarily conflict with reuse.

[john smith 19]

Fair enough.  Let's focus only on build costs, since they dominate the calculus (and also shows why reusables that don't require much refurbishment post-flight are game-changing).  I maintain that labor is by far and away the largest contributor to launch cost at "OldSpace" companies (as well as "NewSpace" firms that revel in high head counts).
...
They're solving a problem that doesn't exist and gullible investors haven't done their due diligence to realize that.

These statements from you are contradictory. The first one states the exact problem that they are trying to solve, high labor costs for building rockets.

They actually don't.

His point is twofold.
1) The bulk of costs for LV's in design and mfg are staff costs.
2) Those staff costs are not an inherent law-of-nature, they are determined by management choices, and better management choices can substantially reduce them to the point where the supposed benefits of no human labor" are eliminated.

OTOH
Integrated in  house mfg means you can tweak designs continuously should you run into a "unknown unknown."
But you are left carrying a lot of staff once the design is complete.
Unitised (as NASA used to describe it) eliminates joints and hence either weak points or points where you have to add extra metal (colloquially "doublers"). So instead of welding pipes onto tank ends make the tank ends thick enough (power spinning or casting) and machine those parts from solid.

but unitised construction could be sub contracted out anyway to delivery most (all) of the benefits of fully automated construction.

This raises the question are ELV's Relativity's end game or is this merely a means to fund their version of a "Christmas machine" or as nanotechologiests put it a "universal assembler."

 

[meberbs]

They actually don't.

His point is twofold.
1) The bulk of costs for LV's in design and mfg are staff costs.
2) Those staff costs are not an inherent law-of-nature, they are determined by management choices, and better management choices can substantially reduce them to the point where the supposed benefits of no human labor" are eliminated.

He quite clearly agreed with me about focusing on build costs not development costs. If management insists on keeping pure design engineers around during production at the same level as initial development, without them working on future projects, that is purely bad management that is simply not what is relevant here.

What is relevant is the cost of the people actually required to perform the build, technicians, test techs, engineers to review data, etc. (This includes their full labor rate which includes management overhead proportioned out, but not the costs of people working other projects.) That is the bulk of the cost based on what HMXHMX said, and I agree.

It seems clear that Relativity is attempting to reduce these costs with 3D printing, since they can have fewer parts, less integration, more optimal design in certain ways, and it also changes test needs with the ability to observe things during the print (Again something Relativity believes will result in less work.)

I don't know if this will ultimately work, but considering the resumes of some of their team members, I expect at least a relatively reasonable chance of success.

(Not replying to the rest of your post, because I don't have any particular disagreement with it or anything relevant to add.)

[FutureSpaceTourist]

https://twitter.com/thesheetztweetz/status/1272997520881393666

Here’s a sneak peek at what’s going on at our Long Beach HQ. The lifts for #Templar have been successfully installed. Once the robots are mounted, they will print, machine, and inspect parts for Terran 1. We’re getting closer and closer to finishing our #factoryofthefuture.

[playadelmars]

https://www.cnbc.com/2020/06/24/3d-rocket-printer-relativity-signs-deal-with-iridium-and-plans-to-build-a-california-launchpad.html

Relativity just announced they’ve secured a launch site at Vandenberg on the southern tip of the base, and a launch contract with Iridium to fly 6 of their satellites. As well as a Chief Financial Officer with investment banking/fundraising background. Pretty big news! That is some solid, very solid, business progress. Methinks as long as the printing tech is actually working, they will definitely make it to orbit just a matter of time now... and they’ll probably have the capital to do it if they don’t already given this kind of business traction which is catnip for investors.

[TrevorMonty]

https://www.cnbc.com/2020/06/24/3d-rocket-printer-relativity-signs-deal-with-iridium-and-plans-to-build-a-california-launchpad.html

Relativity just announced they’ve secured a launch site at Vandenberg on the southern tip of the base, and a launch contract with Iridium to fly 6 of their satellites. As well as a Chief Financial Officer with investment banking/fundraising background. Pretty big news! That is some solid, very solid, business progress. Methinks as long as the printing tech is actually working, they will definitely make it to orbit just a matter of time now... and they’ll probably have the capital to do it if they don’t already given this kind of business traction which is catnip for investors.

The Iridium contract is for spare satellites which will only be launched if needed, so not a launch they can bank on. Given reliability of current satellites, these may not be needed for +10yrs.

[Asteroza]

https://www.cnbc.com/2020/06/24/3d-rocket-printer-relativity-signs-deal-with-iridium-and-plans-to-build-a-california-launchpad.html

Relativity just announced they’ve secured a launch site at Vandenberg on the southern tip of the base, and a launch contract with Iridium to fly 6 of their satellites. As well as a Chief Financial Officer with investment banking/fundraising background. Pretty big news! That is some solid, very solid, business progress. Methinks as long as the printing tech is actually working, they will definitely make it to orbit just a matter of time now... and they’ll probably have the capital to do it if they don’t already given this kind of business traction which is catnip for investors.

The Iridium contract is for spare satellites which will only be launched if needed, so not a launch they can bank on. Given reliability of current satellites, these may not be needed for +10yrs.

Still, that they were selected instead of RocketLab for ground spare emergency replacement standby launch services is interesting.

[ncb1397]

https://www.cnbc.com/2020/06/24/3d-rocket-printer-relativity-signs-deal-with-iridium-and-plans-to-build-a-california-launchpad.html

Relativity just announced they’ve secured a launch site at Vandenberg on the southern tip of the base, and a launch contract with Iridium to fly 6 of their satellites. As well as a Chief Financial Officer with investment banking/fundraising background. Pretty big news! That is some solid, very solid, business progress. Methinks as long as the printing tech is actually working, they will definitely make it to orbit just a matter of time now... and they’ll probably have the capital to do it if they don’t already given this kind of business traction which is catnip for investors.

The Iridium contract is for spare satellites which will only be launched if needed, so not a launch they can bank on. Given reliability of current satellites, these may not be needed for +10yrs.

Still, that they were selected instead of RocketLab for ground spare emergency replacement standby launch services is interesting.

Given that each satellite masses 860 kg, that isn't really surprising.