SpaceX Raptor engine - General Thread 4

[FutureSpaceTourist]

https://twitter.com/elonmusk/status/1382363126117539847

69th Raptor engine coming soon

[cuddihy]

https://twitter.com/elonmusk/status/1382363126117539847

69th Raptor engine coming soon

this makes me think SpaceX actually did have a significant slowdown at the beginning of the year in Raptor production. I saw a raptor RN in the mid 40s under manufacture when I was at Hawthorne in early November. That one was at Boca by Feb, so I expected them to be through the 70s by now. They must have either expanded testing, retooled, or otherwise slowed down the line for a period of time somewhere in there. Maybe some of it was both adding the vacuum Raptors and the updated piping configuration.

Doesn't mean they're not ramping up again soon. But they must have quite a bit of production to get through before they're ready for 28 on the bottom of a SH, I would estimate that won't happen till we're coming up near RN -115 or 120.

[wannamoonbase]

this makes me think SpaceX actually did have a significant slowdown at the beginning of the year in Raptor production. I saw a raptor RN in the mid 40s under manufacture when I was at Hawthorne in early November. That one was at Boca by Feb, so I expected them to be through the 70s by now. They must have either expanded testing, retooled, or otherwise slowed down the line for a period of time somewhere in there. Maybe some of it was both adding the vacuum Raptors and the updated piping configuration.


Doesn't mean they're not ramping up again soon. But they must have quite a bit of production to get through before they're ready for 28 on the bottom of a SH, I would estimate that won't happen till we're coming up near RN -115 or 120.

They have changed Raptor recently.  At a minimum they've rerouted a lot of the spaghetti the newer version looks really tight and ready for prime time.

If there is a slow down I think it was likely those changes.  Or people moved over to Merlin for a bit.

[Chris Bergin]

Oof! Trimmed the thread as it was all kicking off.

A few things:

1) It's really cool when L2 members note L2 is great as L2 is great and without it there would be no site to say it's great!

2) However, that riles people who (and usually for very valid reasons) can't join L2. So best not to post "Oooh, L2 has info on....but I can't tell you" as that's annoying.

3) But also don't post "OMG, paywall" whilst running around on this very forum that is actually free because of L2 funding the huge hosting costs behind being able to host a super large and very busy forum. That's like running around a zoo on a free ticket, but complaining there's a donation box that also allows you to private tour of the elephant enclosure by way of thanks.

I notice people don't complain about other site's Patreon areas? That's a toned-down version of what L2 is, by purpose.

4) It is worth noting that we turn a lot of L2 info around into articles (which everyone can see), as we have been doing with Starship articles to make them far more unique and interesting - including Raptor via our cool McGregor stuff as was used in the most recent article: https://www.nasaspaceflight.com/2021/04/starship-sn15-rollout-raptor-testing/. And that is again going to be the case with the next Starship article!

Back to your regular programming.

[BT52]

I had one question. What is normally used for bearings in turbopump assembelys? I assume hydrodynamic is way to go, right. But for raptors swing and precooldown secquences are quite all over the place. That is quite tricky to cover all bases. All loads and resonances. What u think guys?

[edzieba]

Hydrostatic bearings. Hydrodynamic works at operating speed, hydrostatic works even when stationary.

[steveleach]

this makes me think SpaceX actually did have a significant slowdown at the beginning of the year in Raptor production. I saw a raptor RN in the mid 40s under manufacture when I was at Hawthorne in early November. That one was at Boca by Feb, so I expected them to be through the 70s by now. They must have either expanded testing, retooled, or otherwise slowed down the line for a period of time somewhere in there. Maybe some of it was both adding the vacuum Raptors and the updated piping configuration.


Doesn't mean they're not ramping up again soon. But they must have quite a bit of production to get through before they're ready for 28 on the bottom of a SH, I would estimate that won't happen till we're coming up near RN -115 or 120.

They have changed Raptor recently.  At a minimum they've rerouted a lot of the spaghetti the newer version looks really tight and ready for prime time.

If there is a slow down I think it was likely those changes.  Or people moved over to Merlin for a bit.

Going on public tweets, they were on Raptor 46 on the 16th November 2020 and are on 69 on the 14th April 2021. That's 23 engines in 149 days, or one every 6.5 days. Basically one a week.

This matches the rate they built them between June and August last year.

So there isn't a slowdown, but the rate at which we are seeing them appear isn't increasing either (acknowledging that there may be a lot in the production line that we don't see).

For most manufacturers, building a rocket engine a week would be fantastic, but I believe SpaceX need to increase that dramatically, hence Musk tweeting about Raptor production rate being one of his top 3 priorities back in Feb.

[sdsds]

That's 23 engines in 149 days, or one every 6.5 days. Basically one a week.

It wasn't likely your intent but expressing production rates as intervals, i.e. of the form "one every X," subtly implies serial production. Is there a general sense this is the case for Raptors, or are they produced in batches?

[cuddihy]

So there isn't a slowdown, but the rate at which we are seeing them appear isn't increasing either (acknowledging that there may be a lot in the production line that we don't see).

For most manufacturers, building a rocket engine a week would be fantastic, but I believe SpaceX need to increase that dramatically, hence Musk tweeting about Raptor production rate being one of his top 3 priorities back in Feb.

Good point, I conflated “rate of seen raptors” at different stages of production, with overall rate of Raptor production...who knows if Elon meant Raptor 69 will be complete fabrication soon, or tested and arriving at Boca China soon?
That would be a major difference and could mean they’re actually early into manufacturing RN80+, even if RN 69 is the RN arriving at Boca Chica next.

[steveleach]

That's 23 engines in 149 days, or one every 6.5 days. Basically one a week.

It wasn't likely your intent but expressing production rates as intervals, i.e. of the form "one every X," subtly implies serial production. Is there a general sense this is the case for Raptors, or are they produced in batches?

Yep, that wasn't my intention at all. They are rolling off the production lines at a rate of about 1 a week, but they could have 20 on the go at once, each taking 20 weeks to build.

I'm reasonably confident that they are doing parallel manufacturing. Some components will be produced in batches (start 20, work for a while, complete all 20) but for final assembly I think it will be a production line. One at step 1, one at step 2, all the way up to one pretty much complete at step 20.

[wannamoonbase]

That's 23 engines in 149 days, or one every 6.5 days. Basically one a week.

It wasn't likely your intent but expressing production rates as intervals, i.e. of the form "one every X," subtly implies serial production. Is there a general sense this is the case for Raptors, or are they produced in batches?

Yep, that wasn't my intention at all. They are rolling off the production lines at a rate of about 1 a week, but they could have 20 on the go at once, each taking 20 weeks to build.

I'm reasonably confident that they are doing parallel manufacturing. Some components will be produced in batches (start 20, work for a while, complete all 20) but for final assembly I think it will be a production line. One at step 1, one at step 2, all the way up to one pretty much complete at step 20.

I think they are building them at a pace they need them.  They are not going to be flying multi super heavys with a full engine compliment anytime soon.

Design, Build then  feedback into design and manufacturing
Test in TX, feedback into design
Repeat.

They have no need to be more than 1 per week at this point.

When they need more Raptors, they will pick up the pace.

[steveleach]

That's 23 engines in 149 days, or one every 6.5 days. Basically one a week.

It wasn't likely your intent but expressing production rates as intervals, i.e. of the form "one every X," subtly implies serial production. Is there a general sense this is the case for Raptors, or are they produced in batches?

Yep, that wasn't my intention at all. They are rolling off the production lines at a rate of about 1 a week, but they could have 20 on the go at once, each taking 20 weeks to build.

I'm reasonably confident that they are doing parallel manufacturing. Some components will be produced in batches (start 20, work for a while, complete all 20) but for final assembly I think it will be a production line. One at step 1, one at step 2, all the way up to one pretty much complete at step 20.

I think they are building them at a pace they need them.  They are not going to be flying multi super heavys with a full engine compliment anytime soon.

Design, Build then  feedback into design and manufacturing
Test in TX, feedback into design
Repeat.

They have no need to be more than 1 per week at this point.

When they need more Raptors, they will pick up the pace.

That's entirely possible, but we do know that Musk was worried about production rate back in February. I suspect, with no real evidence, that the current assembly process is far too manual and "cottage industry" for Musk's liking, and he's struggling with something like the "production hell" he saw at  Tesla.

But this is all nowhere near on-topic so if anyone wants to debate this, quote me in a new thread.

[oldAtlas_Eguy]

That's 23 engines in 149 days, or one every 6.5 days. Basically one a week.

It wasn't likely your intent but expressing production rates as intervals, i.e. of the form "one every X," subtly implies serial production. Is there a general sense this is the case for Raptors, or are they produced in batches?

Yep, that wasn't my intention at all. They are rolling off the production lines at a rate of about 1 a week, but they could have 20 on the go at once, each taking 20 weeks to build.

I'm reasonably confident that they are doing parallel manufacturing. Some components will be produced in batches (start 20, work for a while, complete all 20) but for final assembly I think it will be a production line. One at step 1, one at step 2, all the way up to one pretty much complete at step 20.

I think they are building them at a pace they need them.  They are not going to be flying multi super heavys with a full engine compliment anytime soon.

Design, Build then  feedback into design and manufacturing
Test in TX, feedback into design
Repeat.

They have no need to be more than 1 per week at this point.

When they need more Raptors, they will pick up the pace.

That's entirely possible, but we do know that Musk was worried about production rate back in February. I suspect, with no real evidence, that the current assembly process is far too manual and "cottage industry" for Musk's liking, and he's struggling with something like the "production hell" he saw at  Tesla.

But this is all nowhere near on-topic so if anyone wants to debate this, quote me in a new thread.

BTW 1 per week is 50 per year. 50 per year equates to just 2 SH (20 engines each) and 3 SS. SpaceX needs to increase that to 100 or 150 per year. 4-6 SH (20 engines each) or 3 - 4 SH (28 engines each)  and 6-9 SS per year. Building 1 less SH equates to 3 to 4 more SS being built. So the big roadblock to accelerated orbital testing and operations is engine production and test. One item that should enable at least doubling the testing rate is the vertical test cell at McGregor. The question would be when will the test cell be operational to start a stream of engines tested and out the door. As noted earlier since the hard mount engines will only operate in the vertical G  direction there is no reason to even test at horizontal these engines. Also there will be 3 times the number of these engines than those that need to be started in 0 G or in a horizontal G orientation.

Just doubling the production rate means that soon 2 Raptors a week will be rolling out of McGregor to Boca Chica. At the end of June that is likely to be as many as 16 Raptors delivered to Boca Chica.

[Alberto-Girardi]

That's 23 engines in 149 days, or one every 6.5 days. Basically one a week.

It wasn't likely your intent but expressing production rates as intervals, i.e. of the form "one every X," subtly implies serial production. Is there a general sense this is the case for Raptors, or are they produced in batches?

Yep, that wasn't my intention at all. They are rolling off the production lines at a rate of about 1 a week, but they could have 20 on the go at once, each taking 20 weeks to build.

I'm reasonably confident that they are doing parallel manufacturing. Some components will be produced in batches (start 20, work for a while, complete all 20) but for final assembly I think it will be a production line. One at step 1, one at step 2, all the way up to one pretty much complete at step 20.

I think they are building them at a pace they need them.  They are not going to be flying multi super heavys with a full engine compliment anytime soon.

Design, Build then  feedback into design and manufacturing
Test in TX, feedback into design
Repeat.

They have no need to be more than 1 per week at this point.

When they need more Raptors, they will pick up the pace.

I agree. We didn't see any slowing, but we saw a not increase of the productio, IIUC this is due to the design changes. But IMO Raptor production is (if there is any) bottleneck of Starship,  not because it is slow in absolute terms it is slow compared to the Starships production rate.

[oldAtlas_Eguy]

I reactivated the Raptor production thread for this discussion on Raptor production and particularly production rates.

https://forum.nasaspaceflight.com/index.php?topic=50955.new#new

[capoman]

Yeah, I think there is still a lot of flux in the design. They won't ramp up production until they get a mature design. Higher production means automation, and there no sense putting all that automation effort and cost in until you get to a reasonably fixed design. It's possible that they are doing batches of iterations right now, testing the entire batch to get good handle on data and performance, then tweaking and making another batch. I'm sure there are certain components that are reasonably settled on, but Raptor output really won't ramp up until all or most components are mature.

[Alberto-Girardi]

Yeah, I think there is still a lot of flux in the design. They won't ramp up production until they get a mature design. Higher production means automation, and there no sense putting all that automation effort and cost in until you get to a reasonably fixed design. It's possible that they are doing batches of iterations right now, testing the entire batch to get good handle on data and performance, then tweaking and making another batch. I'm sure there are certain components that are reasonably settled on, but Raptor output really won't ramp up until all or most components are mature.

I agree, since they plan the full stack for NET July (if this is true this summer will be simply beautiful) they have enough time to have 28 engines, only making production a bit faster, which IMO is happening now. Maybe with no payload they need less engine on SH, but I don't think a few engine will be the proble, since the production is probably in batches.

[robot_enthusiast]

Yeah, I think there is still a lot of flux in the design. They won't ramp up production until they get a mature design. Higher production means automation, and there no sense putting all that automation effort and cost in until you get to a reasonably fixed design. It's possible that they are doing batches of iterations right now, testing the entire batch to get good handle on data and performance, then tweaking and making another batch. I'm sure there are certain components that are reasonably settled on, but Raptor output really won't ramp up until all or most components are mature.

I agree, since they plan the full stack for NET July (if this is true this summer will be simply beautiful) they have enough time to have 28 engines, only making production a bit faster, which IMO is happening now. Maybe with no payload they need less engine on SH, but I don't think a few engine will be the proble, since the production is probably in batches.

I would just like to point out that according to the information Chris received, the target for an orbital flight is by July 1st, not NET July 1st.

[oldAtlas_Eguy]

With data gathered through most of the actual flight modes except that of actual vacuum. The design of the engine from this point forward should be significantly mature to accomplish its tasks of orbit and even landing of the SH and SS return from orbit if it makes through reentry in one piece. Which is significant in that there are now 2 branches from the main engine design the VAC branch and the no gimbal higher thrust SL branch in production and test. Just need 3 VAC engines but need 8 to 12 or even more of the no gimbal engines to do orbital.

Posted here because of the design maturity assumption. If I am wrong about the level of design maturity then please tell me why.

[copper8]

I've been thinking about longevity of a Raptor engine.  I would expect that, with time, they would settle into a maintenance schedule of some sort for an engine (similar to the inspection/maintenance schedule that an airline adheres to for a jet aircraft).

I've sort of been assuming that a start/stop cycle probably places more stress on an engine than a period of steady state thrust (similar to how they weigh takeoff/landings more heavily that flight hours when determining maintenance schedules for a jet).  Can someone with more knowledge than I confirm that is the case (or explain why it isn't)?