Starship Users Guide

[su27k]

v1.0 can be downloaded from https://www.spacex.com/sites/spacex/files/starship_users_guide_v1.pdf, found by reddit user Russ_Dill

[DreamyPickle]

> 21ton to GTO

That's somewhat unexpected. I remember claims that GTO payload was zero without a separate third stage.

[su27k]

Some interesting tidbits:

1. "The uncrewed Starship allows for the transport of satellites, large observatories, cargo, refueling tanks": separate tank for tanker?

2. "SpaceX is initially planning for two launch sites for the Starship vehicle: Kennedy Space Center LC-39A, Boca Chica launch pad": Well at least this part is clear now.

3. "Payloads are integrated into the Starship fairing vertically in ISO Class 8 (Class 100,000) cleanrooms.
Then the integrated payload stack is transferred to the launch pad and lifted onto the Starship vehicle, while maintaining the same vertical orientation throughout the entire process.": Are they implying the fairing part can be detached from the tank section and re-attached? Not quite sure how else to read it

4. "GTO 21t for single launch": Nice to have this confirmed too

5. "Fully-reusable Starship and Super Heavy systems are expected to allow for space-based activities that have not been possible since the retirement of the Space Shuttle and Space Transportation System or have never been possible before.": Good to see they're not afraid of comparing Starship to Shuttle, this is really the Shuttle 2.0++ folks.

[ThatOldJanxSpirit]

The images are already out of date - leg fairing covers are so last week.

They are still showing the big picture window on the manned version. We’ve sort of assumed that the lox header in the nose killed that off, but that is not necessarily the case. The nose header is needed to bring CoG forward during entry with an empty fairing. The manned version will always have a lot of mass forward so the header could be mounted further aft.

[[email protected]]

> 21ton to GTO

That's somewhat unexpected. I remember claims that GTO payload was zero without a separate third stage.

Except in this article, it's already stated that Starship can lift 20 tons to GTO, but no additional third stage mentioned (and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO)
https://spacenews.com/spacex-targets-2021-commercial-starship-launch/

And Elon has stated (replies to my tweet) that it can lift up to 40 tons to 27°

[soyuzu]

> 21ton to GTO

That's somewhat unexpected. I remember claims that GTO payload was zero without a separate third stage.

Except in this article, it's already stated that Starship can lift 20 tons to GTO, but no additional third stage mentioned (and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO)
https://spacenews.com/spacex-targets-2021-commercial-starship-launch/

And Elon has stated (replies to my tweet) that it can lift up to 30 tons to 27°

More noteworthily, this is RTLS performance. Previously I thought barge landing mentioned in KSC environmental impact assessment has relationship with requirements for single launch to GTO.

These performance numbers assume full Starship reuse, including Super Heavy return to launch site.

Another interesting point is the payload integration method seems explain the insufficient VAB Height

Payloads are integrated into the Starship fairing vertically in ISO Class 8 (Class 100,000) cleanrooms. Then the integrated payload stack is transferred to the launch pad and lifted onto the Starship vehicle, while maintaining the same vertical orientation throughout the entire process. Conditioned air is delivered into the fairing during encapsulated ground processing

[[email protected]]

> 21ton to GTO

That's somewhat unexpected. I remember claims that GTO payload was zero without a separate third stage.

Except in this article, it's already stated that Starship can lift 20 tons to GTO, but no additional third stage mentioned (and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO)
https://spacenews.com/spacex-targets-2021-commercial-starship-launch/

And Elon has stated (replies to my tweet) that it can lift up to 30 tons to 27°

More noteworthily, this is RTLS performance. Previously I thought barge landing mentioned in KSC environmental impact assessment has relationship with requirements for single launch to GTO.

It's more for early launches, so when something goes wrong (you know, landing a 9 m stainless steel vehicle for the first time) they won't be risking the launch site

[mmeijeri]

...and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO...

Why is that logical? What Isp and dry mass are you assuming? I'm getting only ~1700m/s delta-v from LEO assuming 120mT dry mass and 3750 m/s Isp.

[ThatOldJanxSpirit]

...and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO...

Why is that logical? What Isp and dry mass are you assuming? I'm getting only ~1700m/s delta-v from LEO assuming 120mT dry mass and 3750 m/s Isp.

The vehicle launches to LEO with its max 100+ Te payload, it then refuels and boosts it to GTO.

[mmeijeri]

The vehicle launches to LEO with its max 100+ Te payload, it then refuels and boosts it to GTO.

Sure, with refueling, but the 21mT is for a single launch without refueling. Maybe they're still expecting to get the dry mass way below 100mT?

[FutureSpaceTourist]

For ease of reference, a couple of key figures from the user guide

[pochimax]

Payloads are integrated into the Starship fairing vertically in ISO Class 8 (Class 100,000) cleanrooms. Then the integrated payload stack is transferred to the launch pad and lifted onto the Starship vehicle, while maintaining the same vertical orientation throughout the entire process. Conditioned air is delivered into the fairing during encapsulated ground processing

I can believe it. 

[wannamoonbase]

Payloads are integrated into the Starship fairing vertically in ISO Class 8 (Class 100,000) cleanrooms. Then the integrated payload stack is transferred to the launch pad and lifted onto the Starship vehicle, while maintaining the same vertical orientation throughout the entire process. Conditioned air is delivered into the fairing during encapsulated ground processing

I can believe it. 

Seems easier than trying to figure out how to do it horizontally.  That would be a very large hangar and complicated payload handling process to rotate and place and mount into Chomper.  Vertical is a big hangar too but should be able to drop in the payload vertically with a bridge crane.

I admit to at first thinking SS/SH were likely a step to far for SpaceX, or at best much longer in development than proposed.  But Raptor being ready and flight hardware progress of the last year it doesn't seem crazy to think that before the end of 2021 there could be payloads flying to LEO on SS.

[tbellman]

The vehicle launches to LEO with its max 100+ Te payload, it then refuels and boosts it to GTO.

Sure, with refueling, but the 21mT is for a single launch without refueling. Maybe they're still expecting to get the dry mass way below 100mT?

During the presentation in Boca Chica back in September, Elon said something to the effect that he would be very happy if they could bring the dry mass down to 110t, but 120t was more realistic.  If they since have managed to shave it down to less than 100t, that would be a big change.  So the alternative then is that we have underestimated the amount of propellant left when reaching LEO.

Throwing out some ideas, without thinking very detailed about them:

I note that the userguide defines LEO as 500km circular orbit up to 98.9° inclination for the 100+ tonne payload.  That presumably consumes a bit more propellant than a 27° 185km orbit.  Is that enough to explain the different results?

Have calculations for GTO payload assumed stopping in circular LEO first?  Injecting into GTO directly without first circulising in LEO ought to save some Δv.

A couple of days ago, Elon announced that SuperHeavy would be 2m longer.  That's roughly 100t more propellant in the first stage.  How much difference does that make?

Could there be something other than propellant that sets the 100t limit?

(Pet peeve: "mT" means milli-tesla; the proper abbrevation for the metric tonne is "t".)

EDIT: fixed formatting markup

[mmeijeri]

(Pet peeve: "mT" means milli-tesla; the proper abbrevation for the metric tonne is "t".)

Lol, it always felt weird to me too, but I thought that was what US-ians used if they didn't want it confused with short tons.

[Tulse]

I thought this was noteworthy:
"An extended payload volume is also available for payloads requiring up to 22 m of height."

So this and the way the payload gets integrated indicates (as others have noted) that the payload section of Starship will be separated from the propulsion section. Have we seen anything at Boca Chica with the current SNs that would be features involved in the mating and de-mating of these two pieces?

[equiserre]

Does the load factor chart include the accelerations of other than launch? if you compound the accelerations in the lower right corner (2g lateral; -1.5g vertical), the max g during (unpowered) decelerations would be 2.5g and equals an angle of attack of 53 degrees.
Since this is designed to fly at angles of attack up to 90, there are several phases of high lateral acceleration. Is 2g or 2.5g enough for hypersonic bellyflop at 10 km/s? earth point-to-point? high altitude launch abort? especially these last two, with less than orbital speed, usually imply high g´s

[guckyfan]

> 21ton to GTO

That's somewhat unexpected. I remember claims that GTO payload was zero without a separate third stage.

Can someone extrapolate from there if Dear Moon is still possible without refueling? When announced the mission profile did not show any tanker flights.

[waveney]

I thought this was noteworthy:
"An extended payload volume is also available for payloads requiring up to 22 m of height."

This would be a taller Starship!  Maybe another 5-6 Meters tall.

[equiserre]

...and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO...

Why is that logical? What Isp and dry mass are you assuming? I'm getting only ~1700m/s delta-v from LEO assuming 120mT dry mass and 3750 m/s Isp.

I get the same results. 100t to LEO is possible even with a 120t starship. Now 21t to GTO I can´t get it even with an 80t starship.
what is the delta V to a direct GTO as described?

[gustavohigon]

If S tarship is going to have a "dettachable" fairing or cargo bay, I think that's going to bring back the debate about a LAS.

Also I wonder how they will manage the TPS on the joint line....

[livingjw]

If S tarship is going to have a "dettachable" fairing or cargo bay, I think that's going to bring back the debate about a LAS.

Also I wonder how they will manage the TPS on the joint line....

TPS are fragile and segmented. I don't believe they will be a problem if they do want to detach in flight.

John

[rcoppola]

I thought this was noteworthy:
"An extended payload volume is also available for payloads requiring up to 22 m of height."

So this and the way the payload gets integrated indicates (as others have noted) that the payload section of Starship will be separated from the propulsion section. Have we seen anything at Boca Chica with the current SNs that would be features involved in the mating and de-mating of these two pieces?

We haven't seen the Interface design for the Fairing to tank section yet. Maybe we'll get a sense of an initial interface design when/if they attach the current in production fairing to SN-3. (If not SN3, then SN4)

[soyuzu]

...and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO...

Why is that logical? What Isp and dry mass are you assuming? I'm getting only ~1700m/s delta-v from LEO assuming 120mT dry mass and 3750 m/s Isp.

I get the same results. 100t to LEO is possible even with a 120t starship. Now 21t to GTO I can´t get it even with an 80t starship.
what is the delta V to a direct GTO as described?

Please note the LEO payload is actually

Up to 500-km circular orbit at up to 98.9-deg inclination

[OTV Booster]

A couple points of interest:

- “An extended payload volume is also available for payloads requiring up to 22 m of height.”

With a drawing showing a payload space of 17.24m that means they’re willing to add ~2.5 rings to the cargo bay. The first sign that SX is willing to customize variants.

- the chomper is still in play.

Phil

[TrueBlueWitt]

8 meter diameter works for LUVOIR A.. would it need the 22m length?

[oiorionsbelt]

8 meter diameter works for LUVOIR A.. would it need the 22m length?

[equiserre]

...and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO...

Why is that logical? What Isp and dry mass are you assuming? I'm getting only ~1700m/s delta-v from LEO assuming 120mT dry mass and 3750 m/s Isp.

I get the same results. 100t to LEO is possible even with a 120t starship. Now 21t to GTO I can´t get it even with an 80t starship.
what is the delta V to a direct GTO as described?

Please note the LEO payload is actually

Up to 500-km circular orbit at up to 98.9-deg inclination

No problem with that. What I dont understand is the GTO 21t. Anybody know the delta v required for that GTO?

[gosnold]

8 meter diameter works for LUVOIR A.. would it need the 22m length?

LUVOIR with a monolithic, non-lightweighted mirror could be a life-saver for the budget of the NASA astrophysics division. You don't want a repeat of JWST.

[mmeijeri]

No problem with that. What I dont understand is the GTO 21t. Anybody know the delta v required for that GTO?

Since they're saying GTO - 1700m/s, I'd guess its 2300-2400m/s.

[Robotbeat]

...and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO...

Why is that logical? What Isp and dry mass are you assuming? I'm getting only ~1700m/s delta-v from LEO assuming 120mT dry mass and 3750 m/s Isp.

I get the same results. 100t to LEO is possible even with a 120t starship. Now 21t to GTO I can´t get it even with an 80t starship.
what is the delta V to a direct GTO as described?

Please note the LEO payload is actually

Up to 500-km circular orbit at up to 98.9-deg inclination

No problem with that. What I dont understand is the GTO 21t. Anybody know the delta v required for that GTO?

the User’s Guide says 1800m/s to-go. So that GTO is approximately 2.5km/s from LEO.

[equiserre]

No problem with that. What I dont understand is the GTO 21t. Anybody know the delta v required for that GTO?

Since they're saying GTO - 1700m/s, I'd guess its 2300-2400m/s.

Plus how much from the ground? 9000?

[Robotbeat]

No problem with that. What I dont understand is the GTO 21t. Anybody know the delta v required for that GTO?

Since they're saying GTO - 1700m/s, I'd guess its 2300-2400m/s.

No. 1800m/s, not 1700.

[TrueBlueWitt]

8 meter diameter works for LUVOIR A.. would it need the 22m length?

That looks like LUVOIR B  which would be a piece of cake for starship. Do they explicitly say LUVOIR A fits? I don't remember.

[whitelancer64]

8 meter diameter works for LUVOIR A.. would it need the 22m length?

LUVOIR with a monolithic, non-lightweighted mirror could be a life-saver for the budget of the NASA astrophysics division. You don't want a repeat of JWST.

The mirror is not the problem with JWST. The mirror cost was something like 14% of the cost of JWST.

[haster123]

So the loads for starship are the *same* as F9? 

I assume this is just placeholder, until they actually do some test flight and update?  Interesting nonetheless

[Kazioo]

8 meter diameter works for LUVOIR A.. would it need the 22m length?

That looks like LUVOIR B  which would be a piece of cake for starship. Do they explicitly say LUVOIR A fits? I don't remember.

From The LUVOIR Final Report:

The final fairing dimensions are still being determined but SpaceX did conduct a preliminary analysis of a fairing whose shape was altered to fit LUVOIR-A (based on this study’s final concept models) and they reported that it was a viable option. Without modification, LUVOIR-B can fit into the currently plannedStarship fairing with room to spare as shown in Figure 10-9.

[rakaydos]

8 meter diameter works for LUVOIR A.. would it need the 22m length?

That looks like LUVOIR B  which would be a piece of cake for starship. Do they explicitly say LUVOIR A fits? I don't remember.

From The LUVOIR Final Report:

The final fairing dimensions are still being determined but SpaceX did conduct a preliminary analysis of a fairing whose shape was altered to fit LUVOIR-A (based on this study’s final concept models) and they reported that it was a viable option. Without modification, LUVOIR-B can fit into the currently plannedStarship fairing with room to spare as shown in Figure 10-9.

So the answer is "no, but we can work something up for you."

[Kazioo]

For comparison with the fairing of SLS Block 2.

Also an interesting quote about luvoir-B:

the SpaceX Starship fairing and the Blue Origin New Glenn fairing offer substantially more volume than the COTS 5-m faring. This opens up the option to launch LUVOIR-B partially deployed, substantially reducing its complexity, and therefore reducing its cost and risk.

[Robotbeat]

So the loads for starship are the *same* as F9? 

I assume this is just placeholder, until they actually do some test flight and update?  Interesting nonetheless

It's a design spec. They're designing the vehicle to fit the same load profile as F9.

[gaballard]

Saw this bit at the bottom of the guide (emphasis mine):

For more information on additional capabilities or to conceptualize new ideas, please contact [email protected]

Well, fellow NSFers, you know what we have to do 

[ZachF]

Starship will get a small boost for GTO payloads when launching from Boca Chica vs the Cape (28.6d vs 26.0d), about 65-70m/s.

Not a huge boost, but enough to theoretically add ~3 tonnes to the GTO-1800 payload vs the cape.

[Thunderscreech]

Can someone extrapolate from there if Dear Moon is still possible without refueling? When announced the mission profile did not show any tanker flights.

If you have a life support & habitat mass of 21 tons, then you can do a free return to GTO.  If you drop the mass of your hab in half, that's another 10.5 tons of propellant that can be used to push your mass further on the translunar injection burn.  So how far does 10 or so tons of propellant get you?  If the numbers I put into http://www.strout.net/info/science/delta-v/ are right, that's just 300 m/s (assuming a pre-hab dry mass of 110 tons) more impulse on that trans lunar burn.  If the hab and life support adds 5 tons onto a 110 ton baseline, then that goes to GTO+450 m/s. 

I might be messing something up really badly, but I can't figure out a way to shave the habitat mass down enough to get into the 3-4km/s that I think is needed for a free-return trajectory around the moon and it doesn't really even leave a reasonable amount of room for a landing fuel budget back on earth either.

Maybe expending the first stage (yikes) could help enough, but doesn't sound like it's an easy move from 21 tons to GTO->Lunar Free Return without refueling.

[acsawdey]

Starship will get a small boost for GTO payloads when launching from Boca Chica vs the Cape (28.6d vs 26.0d), about 65-70m/s.

Not a huge boost, but enough to theoretically add ~3 tonnes to the GTO-1800 payload vs the cape.

Does that include the losses from the dogleg it has to do to get out of the Gulf of Mexico without overflying anyone?

[Sciencefan]

Please note the LEO payload is actually

Up to 500-km circular orbit at up to 98.9-deg inclination

I don't like statements like this because it don't mean up to 500 km orbit with 98.9-deg inclination.
On the other hand,

These performance numbers assume full Starship reuse, including Super Heavy return to launch site.

Landing somewhere else can add some perfomance for Dear Moon launch.

[FutureSpaceTourist]

twitter.com/erdayastronaut/status/1245052949170094082

Looks like @SpaceX is seriously underselling their LEO capability of #Starship. If Starship can send 21 tonnes to GTO-1800, assuming 120t dry mass and 380 ISP, that’s 156t to LEO 🤯 so either dry mass is less, ISP more or LEO will be over 150 tonnes 🤯 @elonmusk @flightclubio

https://twitter.com/elonmusk/status/1245063992361406464

Mass of initial SN ships will be a little high & Isp a little low, but, over time, it will be ~150t to LEO fully reusable

[TrueBlueWitt]

8 meter diameter works for LUVOIR A.. would it need the 22m length?

That looks like LUVOIR B  which would be a piece of cake for starship. Do they explicitly say LUVOIR A fits? I don't remember.

From The LUVOIR Final Report:

The final fairing dimensions are still being determined but SpaceX did conduct a preliminary analysis of a fairing whose shape was altered to fit LUVOIR-A (based on this study’s final concept models) and they reported that it was a viable option. Without modification, LUVOIR-B can fit into the currently plannedStarship fairing with room to spare as shown in Figure 10-9.

So the answer is "no, but we can work something up for you."

Thanks, was just reading through that myself. The picture was with B version.

I think A is where the "altered" longer 22m fairing comes in..  Can't think of anything else that would need it offhand?

[OTV Booster]

(Pet peeve: "mT" means milli-tesla; the proper abbrevation for the metric tonne is "t".)

Lol, it always felt weird to me too, but I thought that was what US-ians used if they didn't want it confused with short tons.

Hey, them there is good ol merican tons. We don’t short nobody, and we don’t throw no extra letters in to make them heavier. 


Peace, Bro.


Phil

[TheRadicalModerate]

My initial thoughts:

1) Payload volume of the standard bay is 660 m³ if you model the nose as a frustum (which it isn't, but pretty close).

2) This "extended payload volume" comment is interesting. I'm guessing that that's for an expendable system without header tanks. I can't imagine that they can stretch the nose with the canards there. I guess that it's possible that they're planning on adding an extra 5 m to the cylindrical section but that's not shown in the renderings.

3) If I'm right about #2, then that makes the header tanks have a volume of about 12 m³. That's about 10.6 t of prop, which would be 290 m/s of delta-v on an empty ship if the average Isp is about 345 s. That might work to land empty on Earth, but it's nowhere close on Mars, especially when landing cargo. I wonder if the whole header tank idea has gone bye-bye. (It certainly has if that crew rendering is to be believed...)

4) They're sticking with the chomper! I simply don't understand this.

5) I especially don't understand the chomper in light of them offering the F9- and FH-compatible clamp-band payload attach fittings. Seems like you'd wind up with separation debris wandering around in the payload bay. Also, how does a system designed to move payload axially away from the vehicle suddenly wind up lifting them out to avoid the fixed clamshell at the bottom of the bay?

6) Vertical integration only!

7) They're advertising 100+ t payload to 500x500x98.9º polar orbits. I assume that that's with the KSC-based dogleg trajectory over Cuba. That implies that the normal 200x200x28.5º LEO capacity must be well over 100 t.

8) I'm a bit surprised to see the crew version with windows. I was almost certain that the first crew modules would be essentially payloads integrated into the regular payload bay. I suppose that the crew variants could just be "aspirational", to use one of Elon's favorite words. The point of this document is to drive non-crew payload planning.

9) Take a look at the legs in the lunar cargo picture. They're still just in-line telescoping legs. I've thought for a while that the only way you can do rough-surface landing is with self-leveling legs with linear brakes that clamp when all legs show that they're on the surface, and this seems to be headed in that direction. Wide-base fold-out legs have to be huge to handle the likely set of terrain cases.

10) I'll be very interested in seeing how SpaceX provides Shuttle-like supports for trunnions on co-manifested payloads. Given the enormous space, this seems like something where a lot more details will emerge as they get deeper into the program. It's not good practice for SpaceX to rely on payload integrators to devise and test that kind of hardware. Also, the "just mount stuff side-by-side on multiple PAFs" recommendation is interesting.

11) Note that the surface cargo payload bay is divided into shelves to offload separate modules. I wonder if this is just fanciful or whether there's engineering behind this.

12) It appears that for surface offload, they're using part of the fairing as the elevator platform. That's... interesting. I'll be curious to see how well they do packing it back in place for the return trip. Doesn't seem like a super-reliable plan to take part of your fairing, load it with random stuff, plunk it on the ground, roll stuff off of it, and then reposition it back in place for reentry.

[whitelancer64]

8 meter diameter works for LUVOIR A.. would it need the 22m length?

That looks like LUVOIR B  which would be a piece of cake for starship. Do they explicitly say LUVOIR A fits? I don't remember.

From The LUVOIR Final Report:

The final fairing dimensions are still being determined but SpaceX did conduct a preliminary analysis of a fairing whose shape was altered to fit LUVOIR-A (based on this study’s final concept models) and they reported that it was a viable option. Without modification, LUVOIR-B can fit into the currently plannedStarship fairing with room to spare as shown in Figure 10-9.

So the answer is "no, but we can work something up for you."

Thanks, was just reading through that myself. The picture was with B version.

I think A is where the "altered" longer 22m fairing comes in..  Can't think of anything else that would need it offhand?

Two Kibo modules welded end-to-end?

[OTV Booster]

I thought this was noteworthy:
"An extended payload volume is also available for payloads requiring up to 22 m of height."

So this and the way the payload gets integrated indicates (as others have noted) that the payload section of Starship will be separated from the propulsion section. Have we seen anything at Boca Chica with the current SNs that would be features involved in the mating and de-mating of these two pieces?

Mating different upper sections has been a point of debate, as has customized variants. If this doesn’t change both questions have been answered.

The difference between 17.23m and 22 meters comes out to about 2.5 rings.

We haven’t seen anything like a detach mechanism, but then we really haven’t seen the upper section permanently mounted on any SS. Give it time. They need to fly before nose swap comes up high enough on the list to get detailed attention. I’d guess an orbital flight before it shows up.

Phil

[salpun]

Could someone please  add Version 1.0 to the thread title. We all know its going to change sooner rather then later. 

[whitelancer64]

My initial thoughts:

5) I especially don't understand the chomper in light of them offering the F9- and FH-compatible clamp-band payload attach fittings. Seems like you'd wind up with separation debris wandering around in the payload bay. Also, how does a system designed to move payload axially away from the vehicle suddenly wind up lifting them out to avoid the fixed clamshell at the bottom of the bay?

Based on the deployment render, I think the PAF must be hinged to angle the payload prior to deployment.

[Coastal Ron]

Many of us have been working on space station designs, and no doubt all of us have been wondering how large of payloads we could fit inside of a Starship. Now we know, and I marked up the diagram in the Users Guide to make it easier to see what size cylindrical modules could fit.

As a note, for some of my rotating space station designs, I had been hoping for a minimum of 8m in diameter by 10m in length, but I can get by with 7m in diameter and 10m in length if it means transportation costs are significantly less.

[ZachF]

Many of us have been working on space station designs, and no doubt all of us have been wondering how large of payloads we could fit inside of a Starship. Now we know, and I marked up the diagram in the Users Guide to make it easier to see what size cylindrical modules could fit.

As a note, for some of my rotating space station designs, I had been hoping for a minimum of 8m in diameter by 10m in length, but I can get by with 7m in diameter and 10m in length if it means transportation costs are significantly less.

It looks like there will be an "XL" variant that is ~5m longer if you need it.

[TheRadicalModerate]

> 21ton to GTO

That's somewhat unexpected. I remember claims that GTO payload was zero without a separate third stage.

It can't go direct to GEO without refueling, but GTO is another story.  I get almost exactly 21 t to GTO in my model (which is a first; it's hardly ever that accurate).  Note that the GTO that they're specifying is GEO-1800 m/s.

[TheRadicalModerate]

My initial thoughts:

5) I especially don't understand the chomper in light of them offering the F9- and FH-compatible clamp-band payload attach fittings. Seems like you'd wind up with separation debris wandering around in the payload bay. Also, how does a system designed to move payload axially away from the vehicle suddenly wind up lifting them out to avoid the fixed clamshell at the bottom of the bay?

Based on the deployment render, I think the PAF must be hinged to angle the payload prior to deployment.

After a second read, that's what I think as well.

Anybody know about what kinds of separation debris you get coming off of a clamp-band PAF?

[TheRadicalModerate]

The images are already out of date - leg fairing covers are so last week.

They are still showing the big picture window on the manned version. We’ve sort of assumed that the lox header in the nose killed that off, but that is not necessarily the case. The nose header is needed to bring CoG forward during entry with an empty fairing. The manned version will always have a lot of mass forward so the header could be mounted further aft.

I'm wondering if they've abandoned the header tanks in the nose.  Both the crew and chomper renders are inconsistent with header tanks.  I was also wondering whether the 22 m "extended" version was simply a version without header tanks, but I don't think the geometry works out as I look at it a bit more.

[tbellman]

Based on the deployment render, I think the PAF must be hinged to angle the payload prior to deployment.

That's even explicitly stated, on page 2 (the third page in the PDF):

To deploy the payload, the clamshell fairing door is opened, and the payload adapter and payload are tilted at an angle in preparation for separation.

[gpm]

My initial thoughts:

8\) I'm a bit surprised to see the crew version with windows. I was almost certain that the first crew modules would be essentially payloads integrated into the regular payload bay. I suppose that the crew variants could just be "aspirational", to use one of Elon's favorite words. The point of this document is to drive non-crew payload planning.

I don't think windows are optional for dear moon considering the goal is a pleasure cruise around the moon for a rich patron and some artists. I suppose technically this could be done with a windowed vessel inside a chomper and have the chomper open, but that sounds like a big mass penalty for not much benefit.

[Twark_Main]

2) This "extended payload volume" comment is interesting. I'm guessing that that's for an expendable system without header tanks. I can't imagine that they can stretch the nose with the canards there. I guess that it's possible that they're planning on adding an extra 5 m to the cylindrical section but that's not shown in the renderings.

Expendable seems unlikely, even considering Starship's low (aspirational) build cost. I would bet on just a different weight distribution, or possibly smaller canards.

4) They're sticking with the chomper! I simply don't understand this.

5) I especially don't understand the chomper in light of them offering the F9- and FH-compatible clamp-band payload attach fittings. Seems like you'd wind up with separation debris wandering around in the payload bay.

Just thrust backwards to clear it out.

7) They're advertising 100+ t payload to 500x500x98.9º polar orbits.

They're advertising "up to" those parameters. I presume the math indicates that it can do both at the same time, but the text is ambiguous.

8 ) I'm a bit surprised to see the crew version with windows. I was almost certain that the first crew modules would be essentially payloads integrated into the regular payload bay. I suppose that the crew variants could just be "aspirational", to use one of Elon's favorite words. The point of this document is to drive non-crew payload planning.

And I was just as certain that you were mistaken. Especially given the revelation about swappable payload sections (which offers largely the same advantages), I hope we can finally put that "box in a box" concept into a pine box.

It's not good practice for SpaceX to rely on payload integrators to devise and test that kind of hardware.

Why not?

Furthermore, with the low price per launch, it's not clear if the complexity of co-manifesting can even pay for itself. If the payload integrators can't do it economically, why should SpaceX waste their own money?

11) Note that the surface cargo payload bay is divided into shelves to offload separate modules. I wonder if this is just fanciful or whether there's engineering behind this.

12) It appears that for surface offload, they're using part of the fairing as the elevator platform. That's... interesting. I'll be curious to see how well they do packing it back in place for the return trip. Doesn't seem like a super-reliable plan to take part of your fairing, load it with random stuff, plunk it on the ground, roll stuff off of it, and then reposition it back in place for reentry.

This is an old rendering. The document also says that "various payload bay configurations are available and allow for fully autonomous deployment of cargo to Earth (!), Lunar, or Martian surfaces with one example shown in Figure 7 (ie the image you're referencing)."

[dlapine]

So, maybe I'm misunderstanding something here but, for the GTO with 21tons, can't SpaceX simply replace the 79tons of potential cargo with additional methalox? Yes, the tanks would need to be sized to accommodate a larger fuel/O2 load.

Perhaps the simpler question is what's the dV of Starship in leo with 79 tons of propellant to burn and 21 tons of cargo? Can that make GTO -1800m/s?

<edited for spelling>

[Ludus]

Starship will get a small boost for GTO payloads when launching from Boca Chica vs the Cape (28.6d vs 26.0d), about 65-70m/s.

Not a huge boost, but enough to theoretically add ~3 tonnes to the GTO-1800 payload vs the cape.

What does it get from the equator?

[OTV Booster]

Many of us have been working on space station designs, and no doubt all of us have been wondering how large of payloads we could fit inside of a Starship. Now we know, and I marked up the diagram in the Users Guide to make it easier to see what size cylindrical modules could fit.

As a note, for some of my rotating space station designs, I had been hoping for a minimum of 8m in diameter by 10m in length, but I can get by with 7m in diameter and 10m in length if it means transportation costs are significantly less.

The XL version will get you 8x12.76

[OTV Booster]

My initial thoughts:

5) I especially don't understand the chomper in light of them offering the F9- and FH-compatible clamp-band payload attach fittings. Seems like you'd wind up with separation debris wandering around in the payload bay. Also, how does a system designed to move payload axially away from the vehicle suddenly wind up lifting them out to avoid the fixed clamshell at the bottom of the bay?

Based on the deployment render, I think the PAF must be hinged to angle the payload prior to deployment.

After a second read, that's what I think as well.

Anybody know about what kinds of separation debris you get coming off of a clamp-band PAF?

None if you do it right

[Cheapchips]

The document also says that "various payload bay configurations are available and allow for fully autonomous deployment of cargo to Earth (!)..."

There's also this on the following page

The cargo version can also be used for rapid point-to-point Earth transport.

I don't think we've had cargo P2P specifically mentioned by SpaceX before, although it has been speculated on NSF as a P2P starter business.

[TheRadicalModerate]

My initial thoughts:

8\) I'm a bit surprised to see the crew version with windows. I was almost certain that the first crew modules would be essentially payloads integrated into the regular payload bay. I suppose that the crew variants could just be "aspirational", to use one of Elon's favorite words. The point of this document is to drive non-crew payload planning.

I don't think windows are optional for dear moon considering the goal is a pleasure cruise around the moon for a rich patron and some artists. I suppose technically this could be done with a windowed vessel inside a chomper and have the chomper open, but that sounds like a big mass penalty for not much benefit.

Windows on the module with the chomper open was exactly what I've been thinking.

I'm extremely skeptical that you can have that many windows withstanding the aero and thermal loads of launch, to say nothing of EDL.  It's a lot easier if the windows are inside a fairing, i.e., the chomper.

Update:  Is it a big mass penalty?  As a cargo system, the Starship payload bay doesn't have to be a pressure vessel.  ISTM that integrating a pressure vessel directly into the structure vs. just sliding it onto a PAF or trunnion are pretty close in overall mass.

[Coastal Ron]

I don't think we've had cargo P2P specifically mentioned by SpaceX before, although it has been speculated on NSF as a P2P starter business.

SpaceX has talked about point-to-point for many years, and though they show the passenger version in their for it, I don't think they have ever ruled out cargo P2P.

[equiserre]

2) This "extended payload volume" comment is interesting. I'm guessing that that's for an expendable system without header tanks. I can't imagine that they can stretch the nose with the canards there. I guess that it's possible that they're planning on adding an extra 5 m to the cylindrical section but that's not shown in the renderings.

Expendable seems unlikely, even considering Starship's low (aspirational) build cost. I would bet on just a different weight distribution, or possibly smaller canards.

or, move the bulkheads back to give you 5m bigger payload bay. You end up with a 900t propellant starship which can still launch 50t to LEO. 50t/22m lenght payload anyone?

[Keldor]

My initial thoughts:

8\) I'm a bit surprised to see the crew version with windows. I was almost certain that the first crew modules would be essentially payloads integrated into the regular payload bay. I suppose that the crew variants could just be "aspirational", to use one of Elon's favorite words. The point of this document is to drive non-crew payload planning.

I don't think windows are optional for dear moon considering the goal is a pleasure cruise around the moon for a rich patron and some artists. I suppose technically this could be done with a windowed vessel inside a chomper and have the chomper open, but that sounds like a big mass penalty for not much benefit.

Windows on the module with the chomper open was exactly what I've been thinking.

I'm extremely skeptical that you can have that many windows withstanding the aero and thermal loads of launch, to say nothing of EDL.  It's a lot easier if the windows are inside a fairing, i.e., the chomper.

Update:  Is it a big mass penalty?  As a cargo system, the Starship payload bay doesn't have to be a pressure vessel.  ISTM that integrating a pressure vessel directly into the structure vs. just sliding it onto a PAF or trunnion are pretty close in overall mass.

I mean, if you can have one window, you can have many.  It's just a matter of mass penalty since they're quite a bit heavier than plain siding.  (Anyone have numbers for this?  Have materials improved since the Shuttle?)

Obviously you'd need to make sure to keep them sheltered from the worst of reentry, so maybe they'd want to push the edge of the big window section back from the nose tip a little bit, but otherwise the locations seem to make sense to me.  But SpaceX has excellent modeling software.  They'll be able to figure out what works and what doesn't work.

[Keldor]

The document also says that "various payload bay configurations are available and allow for fully autonomous deployment of cargo to Earth (!)..."

There's also this on the following page

The cargo version can also be used for rapid point-to-point Earth transport.

I don't think we've had cargo P2P specifically mentioned by SpaceX before, although it has been speculated on NSF as a P2P starter business.

I mean, if you can have P2P passengers, you can have P2P cargo, assuming you have anyone interested in using it.  If they can, say, shave 8 hours off a trans-Pacific shipment, it could be a big win for some critical part in a factory that has the place shut down hard and loosing $1000s per hour, even if it doubles shipping costs.

But autonomous unloading is useful for passengers too.  Imagine them starting unloading the luggage section while the flight personel are still rushing over from outside the exclusion radius immediately after landing.  Could save 15 minutes on the turnaround from this alone!  Significant savings when you're trying to reach airline type turnarounds.

A really big win might be possible if they can load up the entire passenger module on the ground like they're proposing for satellite launches.  Board the passengers into the module while they're still on the boat going out to the launch platform, and then all they have to do on the pad is raise the module up and mount it into the vehicle.  And the reverse when disembarking.  There are, of course, a lot of GSE concerns to this approach, such as making sure the fiery liftoff doesn't damage the lift tower, but it might be feasible.  Some sort of mobile lift tower might be possible, which could roll a safe distance away during launch and landing, and would have the added benefit of accomodating the vehicle landing a meter or two off target without having to pick it up and reposition it for loading/unloading.

[Robotbeat]

> 21ton to GTO

That's somewhat unexpected. I remember claims that GTO payload was zero without a separate third stage.

It can't go direct to GEO without refueling, but GTO is another story.  I get almost exactly 21 t to GTO in my model (which is a first; it's hardly ever that accurate).  Note that the GTO that they're specifying is GEO-1800 m/s.

Maybe it can do direct to GEO fully reusably by using a modified Dragon XL as a Chomper-recoverable third stage:

Separate topic: Dragon XL (without the pressure vessel, with some body-mounted panels and stretched tanks) would make a good recoverable third stage for Starship.

Would allow Starship to do single-launch fully recoverable launches to direct GSO.

The Dragon XL with the payload on top would stage at GTO out of Starship, do the burn into GSO, separate payload, then go back to GTO (phasing may take time?), and dock with Starship/Chomper and come back to Earth. Full reuse, single-launch direct to GSO. Could probably do escape missions this way, too.

Likewise, this is how Dragon XL could be launched and recovered from Starship for Gateway missions. Those folding solar panels (necessary for independent power on Gateway—body panels aren’t enough) will need to be either foldable or expended.

[Ludus]

21t to GSO on a launcher that’s fully reusable, returns to launch site, needs no extra in orbit propellant, would make putting really big satellites into GSO cheap and easy. They could be 5 or 6X the size. Much bigger solar arrays and antennas would mean comparably smaller cheaper ground antennas. Global demand for HDTV isn’t going away despite streaming. If nobody else is doing next generation GSO satellites maybe SpaceX will do them itself.

[tbellman]

Update:  Is it a big mass penalty?  As a cargo system, the Starship payload bay doesn't have to be a pressure vessel.  ISTM that integrating a pressure vessel directly into the structure vs. just sliding it onto a PAF or trunnion are pretty close in overall mass.

On the other hand, the cargo hold not being pressurized during entry and descent, means that it needs additional reenforcements to provide stiffness, stability, and buckling resistance.  Integrating the crew pressure vessel with the outer shell means that the one bar pressure inside helps with that stiffness.  Having a separate pressure vessel inside the normal cargo hold, you get the mass penalties of both.  I don't dare make estimates as to how large those mass penalties are, though.

Except for the possible weaknesses of windows, I don't see any advantages to doing the crew compartment as a separate cargo payload.  I can't imagine there would be much, or any, savings of development or manufacturing costs.

There are some things that having a separate crew compartment inside a chomper makes more difficult as well.  E.g, egress and ingress on the surface of Mars and the Moon, or docking to space stations like ISS or the Lunar TollboothGateway.  An open chomper clamshell would be in the way, or at the very least severely constrain where you can dock.  Even if Starship would only need those capabilities later on, making a separate crew compartment as a cargo module initially, would be a detour on the way to what SpaceX really want to achieve.

[niwax]

Update:  Is it a big mass penalty?  As a cargo system, the Starship payload bay doesn't have to be a pressure vessel.  ISTM that integrating a pressure vessel directly into the structure vs. just sliding it onto a PAF or trunnion are pretty close in overall mass.

On the other hand, the cargo hold not being pressurized during entry and descent, means that it needs additional reenforcements to provide stiffness, stability, and buckling resistance.  Integrating the crew pressure vessel with the outer shell means that the one bar pressure inside helps with that stiffness.  Having a separate pressure vessel inside the normal cargo hold, you get the mass penalties of both.  I don't dare make estimates as to how large those mass penalties are, though.

Except for the possible weaknesses of windows, I don't see any advantages to doing the crew compartment as a separate cargo payload.  I can't imagine there would be much, or any, savings of development or manufacturing costs.

There are some things that having a separate crew compartment inside a chomper makes more difficult as well.  E.g, egress and ingress on the surface of Mars and the Moon, or docking to space stations like ISS or the Lunar TollboothGateway.  An open chomper clamshell would be in the way, or at the very least severely constrain where you can dock.  Even if Starship would only need those capabilities later on, making a separate crew compartment as a cargo module initially, would be a detour on the way to what SpaceX really want to achieve.

An idea I've been playing around with for a bit is commonality with the cargo version as you say, but also replacing the chomper door with a different insert that contains the windows and airlocks. The door is a seperate component anyway and shouldn't be much harder to replace than the cargo itself.

[mclumber1]

For the crewed version, I wonder if they could simply remove the chomper portion of the starship and replace it with a crew module (with windows).  This module would be self contained and theoretically be capable of sustaining a crew without the starship.

[soyuzu]

...and it's logical anyways for 100 tons LEO launch vehicle to be able to launch that amount to GTO...

Why is that logical? What Isp and dry mass are you assuming? I'm getting only ~1700m/s delta-v from LEO assuming 120mT dry mass and 3750 m/s Isp.

I get the same results. 100t to LEO is possible even with a 120t starship. Now 21t to GTO I can´t get it even with an 80t starship.
what is the delta V to a direct GTO as described?

Please note the LEO payload is actually

Up to 500-km circular orbit at up to 98.9-deg inclination

No problem with that. What I dont understand is the GTO 21t. Anybody know the delta v required for that GTO?

High inclination, high altitude LEO requires more delta-v, so the payload to 28.5 degree 100nm LEO can actually be significantly higher. 500km SSO requires some 500-600m/s more delta-v, which can translate to 40t more payload to minimum LEO like GTO Parking orbits.

[ZachF]

21t to GSO on a launcher that’s fully reusable, returns to launch site, needs no extra in orbit propellant, would make putting really big satellites into GSO cheap and easy. They could be 5 or 6X the size. Much bigger solar arrays and antennas would mean comparably smaller cheaper ground antennas. Global demand for HDTV isn’t going away despite streaming. If nobody else is doing next generation GSO satellites maybe SpaceX will do them itself.

Or you could even throw up 10x 2-tonne all-electric sats if you build them flat and stackable like Starlink.

[niwax]

SpaceX has already partnered with Momentus on their LEO rideshares. It's not hard to see how they could offer GTO rideshares with GEO and lunar options through a COTS third stage.

[TorenAltair]

21t to GSO on a launcher that’s fully reusable, returns to launch site, needs no extra in orbit propellant, would make putting really big satellites into GSO cheap and easy. They could be 5 or 6X the size. Much bigger solar arrays and antennas would mean comparably smaller cheaper ground antennas. Global demand for HDTV isn’t going away despite streaming. If nobody else is doing next generation GSO satellites maybe SpaceX will do them itself.

Or you could even throw up 10x 2-tonne all-electric sats if you build them flat and stackable like Starlink.

Imo it would be better to change the sat design itself. Instead of using the best of the best components, go down 2 steps in capability (for example solar array efficiency) and make the sats cost only a fraction of today‘s expensive sats. Of course they would weigh then perhaps the double, but with cheap transportation, it could reduce the cost of space „stuff“ significantly. No longer the need to go for ultra lightweight with expensive components and manufacturing. That‘s the real benefit imo of (if successful) Starship and perhaps New Glenn.

[baldusi]

Actually, GSO satellites are cheap, ground terminal deployment is really expensive. Just adding bigger antenna and solar panels will reduce the overall mission cost. But I do see a different possible future: let's say that you build really big GSO platforms, with standardized payload slots. A lot of money is currently being spent on software defined payloads. But what if you could very cheaply send new payloads to standardized GSO platforms?
The platform would offer GNC, power and telemetry. May be even antennas (but just maybe). Sending payloads each 10 years to a platform might be the new economic model. And that's where a reusable MEV on Starship might shine.

[Benjamin_m5]

Just saw the new Starship Guide published by SpaceX: https://www.spacex.com/sites/spacex/files/starship_users_guide_v1.pdf
Awesome to read this infomations - it's getting a lot more realistic!

I think the most interesting thing is at the end of this guide:

So it's possible to contribute new ideas directly to SpaceX? How cool would it be...

[Slothman]

Just saw the new Starship Guide published by SpaceX: https://www.spacex.com/sites/spacex/files/starship_users_guide_v1.pdf
Awesome to read this infomations - it's getting a lot more realistic!

I think the most interesting thing is at the end of this guide:

So it's possible to contribute new ideas directly to SpaceX? How cool would it be...

I think that is more for customers who have certain requirements instead of internet fans with funny ideas.
Like you are a satellite owner and you want to capture some dead satellites and you want spacex to develop a grabbing arm or hold down system for inside the cargobay. Or your project definitely requires a different way to release the payload than standard.

[Benjamin_m5]

I think that is more for customers who have certain requirements instead of internet fans with funny ideas.
Like you are a satellite owner and you want to capture some dead satellites and you want spacex to develop a grabbing arm or hold down system for inside the cargobay. Or your project definitely requires a different way to release the payload than standard.

Yeah I think it's ment the other way - so SpaceX can conceptualize new ideas for special needs an not we for them. But would have been cool if someone with a great idea could give them directly to SpaceX.

I just thought about how cool it would be, being the one who can design the crew interior for starship...

[haster123]

So the loads for starship are the *same* as F9? 

I assume this is just placeholder, until they actually do some test flight and update?  Interesting nonetheless

It's a design spec. They're designing the vehicle to fit the same load profile as F9.

Yes, its a design spec, but it is for the end user (ie satellite manufacturers).  They are basically saying the load environment users should design their payloads for starship are the exact same enviroments as for F9, and I just don't see how that can be true.

Of course if you were really building a payload for either you would depend on Coupled Loads Analysis to tell you what your specific payload would experience, but still.....

[Kazioo]

The document also says that "various payload bay configurations are available and allow for fully autonomous deployment of cargo to Earth (!)..."

There's also this on the following page

The cargo version can also be used for rapid point-to-point Earth transport.

I don't think we've had cargo P2P specifically mentioned by SpaceX before, although it has been speculated on NSF as a P2P starter business.

The first time Elon talked about the Earth-To-Earth idea was actually about cargo, not passengers.
It was in 2016 (ITS presentation at 54:30):

https://www.youtube.com/watch?v=H7Uyfqi_TE8&feature=youtu.be&t=3270

[tbellman]

Yes, its a design spec, but it is for the end user (ie satellite manufacturers).  They are basically saying the load environment users should design their payloads for starship are the exact same enviroments as for F9, and I just don't see how that can be true.

Of course if you were really building a payload for either you would depend on Coupled Loads Analysis to tell you what your specific payload would experience, but still.....

This is not SpaceX saying the load environment is intended to be exactly the same as on Falcon.  They are saying that it is intended to not be worse than on Falcon.  Customers can then design for the Falcon environment and know that they can easily swap to Starship without risking their payload being damaged during launch.

The first paragraph under the "Environments" section says:

SpaceX will ensure that Starship environments meet or improve upon those of the Falcon Heavy launch system. To aid in the design of space vehicles capable of flying on Starship, SpaceX is providing the following preliminary payload environments.

Presumably they have made a bunch of modelling and computer simulations on the Starship design so they believe that they can achieve at least as good as Falcon.  Calling them "placeholders" is probably a bit too harsh, but they are definitively preliminary.

[RotoSequence]

By publishing this guide now, SpaceX is creating some level of commitment to a baseline specification for engineers and designers to engineer for. Deviating from these figures in a detrimental manner would be harmful for their business.

[steveleach]

By publishing this guide now, SpaceX is creating some level of commitment to a baseline specification for engineers and designers to engineer for. Deviating from these figures in a detrimental manner would be harmful for their business.

Harmful for their business in the sense that it would drive business to their competitors?

I suspect it would have to deviate quite a lot for that to be a major issue, once SS/SH is actually flying.

[Twark_Main]

By publishing this guide now, SpaceX is creating some level of commitment to a baseline specification for engineers and designers to engineer for. Deviating from these figures in a detrimental manner would be harmful for their business.

Harmful for their business in the sense that it would drive business to their competitors?

I suspect it would have to deviate quite a lot for that to be a major issue, once SS/SH is actually flying.

Once Super Heavy/Starship is actually flying, its credibility goes up a lot. This document is intended to build credibility (and thus attract customers) before that happens, given that launch services purchases tend to be planned well in advance.

[RotoSequence]

By publishing this guide now, SpaceX is creating some level of commitment to a baseline specification for engineers and designers to engineer for. Deviating from these figures in a detrimental manner would be harmful for their business.

Harmful for their business in the sense that it would drive business to their competitors?

I suspect it would have to deviate quite a lot for that to be a major issue, once SS/SH is actually flying.

Most of the promise of Starship comes from making things possible that were never possible before. If people start trying to build something now to take advantage of those capabilities, only to walk back the vehicle's specifications because they were not able to meet them, they will have caused a great deal of anguish for anyone who needs Starship's published capabilities. SpaceX isn't trying to attract customers away from their competitors, they're trying to create an expanded customer base. If SpaceX screws them over, they will not come.

[JamesH65]

By publishing this guide now, SpaceX is creating some level of commitment to a baseline specification for engineers and designers to engineer for. Deviating from these figures in a detrimental manner would be harmful for their business.

Not enough detail to 'engineer' to anything. Total size and mass, vertical integration, and some stuff about being similar to F9. That isn't enough to start designing anything line a PAF, but is enough to say "yes, I can use that" and start designing the payload itself, and worry about integration details later as they get more solid.

[steveleach]

Once Super Heavy/Starship is actually flying, its credibility goes up a lot. This document is intended to build credibility (and thus attract customers) before that happens, given that launch services purchases tend to be planned well in advance.

Agreed, but the customers in question must already have quite a high tolerance for risk to plan on using SS before it has flown. 

Lets say a future version of the user manual states higher acoustic loads than you've designed your 20t GEO satellite for. Where do you take your business?

Or do you just shake your fist at SpaceX a bit, sigh, and give tell your engineers to find a way to deal with it. Who probably expected something like this to happen anyway.

[Twark_Main]

Once Super Heavy/Starship is actually flying, its credibility goes up a lot. This document is intended to build credibility (and thus attract customers) before that happens, given that launch services purchases tend to be planned well in advance.

Agreed, but the customers in question must already have quite a high tolerance for risk to plan on using SS before it has flown. 

Lets say a future version of the user manual states higher acoustic loads than you've designed your 20t GEO satellite for. Where do you take your business?

Or do you just shake your fist at SpaceX a bit, sigh, and give tell your engineers to find a way to deal with it. Who probably expected something like this to happen anyway.

That's why, as RotoSequence said, SpaceX is putting some level of commitment behind these specifications. Obviously the SpaceX engineers think these (rather conservative) specifications are achievable, or they wouldn't publish them (for the exact reasons you just outlined).

[inaccurate_reality]

Think about missions like JWST or LUVOIR, which have been in development for 5+ years if not decades. The initial design phases were extremely abstract; things like payload integration are reasonably flexible, and things like maximum accelerations at launch will be designed with some factor of safety anyhow.

SpaceX is gambling somewhat by putting out figures, but if designers spend their time and money designing for a ship that's barely through the proof-of-concept phase, it's their fault if the ship's capabilities shift and their mission is scrapped.

[ulm_atms]

Starship will get a small boost for GTO payloads when launching from Boca Chica vs the Cape (28.6d vs 26.0d), about 65-70m/s.

Not a huge boost, but enough to theoretically add ~3 tonnes to the GTO-1800 payload vs the cape.

Does that include the losses from the dogleg it has to do to get out of the Gulf of Mexico without overflying anyone?

What dogleg?

[steveleach]

That's why, as RotoSequence said, SpaceX is putting some level of commitment behind these specifications. Obviously the SpaceX engineers think these (rather conservative) specifications are achievable, or they wouldn't publish them (for the exact reasons you just outlined).

I suspect we're all violently agreeing with each other here. Yes, putting all that in the document gives everyone (SpaceX and their customers) a common frame of reference to work towards. And yes, its a certain level of commitment because they know they will, if nothing else, annoy those customers if the back away from those numbers later.

All I was saying is that, given the vast difference between the capabilities of Starship and its competition, I didn't think it would be that harmful to their business if they did.

[ulm_atms]

That's why, as RotoSequence said, SpaceX is putting some level of commitment behind these specifications. Obviously the SpaceX engineers think these (rather conservative) specifications are achievable, or they wouldn't publish them (for the exact reasons you just outlined).

I suspect we're all violently agreeing with each other here. Yes, putting all that in the document gives everyone (SpaceX and their customers) a common frame of reference to work towards. And yes, its a certain level of commitment because they know they will, if nothing else, annoy those customers if the back away from those numbers later.

All I was saying is that, given the vast difference between the capabilities of Starship and its competition, I didn't think it would be that harmful to their business if they did.

I personally think SpaceX released it so we on this forum would have something to talk about for quite some time during the current world events.  They saw how desperate we were for new materials from them. 

[steveleach]

I personally think SpaceX released it so we on this forum would have something to talk about for quite some time during the current world events.  They saw how desperate we were for new materials from them. 

Personally, I can go for literally minutes at a time without checking in on the forums.

[baldusi]

Yes, its a design spec, but it is for the end user (ie satellite manufacturers).  They are basically saying the load environment users should design their payloads for starship are the exact same enviroments as for F9, and I just don't see how that can be true.

Of course if you were really building a payload for either you would depend on Coupled Loads Analysis to tell you what your specific payload would experience, but still.....

Little nit pick: Falcon Heavy environment. Acoustic environment go to 130dB at low frequencies. Low frequency shock is actually lower 20G vs 30G). Integration is vertical exclusively. Basically, they have one equal (axial acceleration), two almost the same (acoustic and shock), and one very different (integration).

[lrk]

Starship will get a small boost for GTO payloads when launching from Boca Chica vs the Cape (28.6d vs 26.0d), about 65-70m/s.

Not a huge boost, but enough to theoretically add ~3 tonnes to the GTO-1800 payload vs the cape.

Does that include the losses from the dogleg it has to do to get out of the Gulf of Mexico without overflying anyone?

What dogleg?

I don't understand that trajectory.  It seems like it is curving to the north?  Or maybe that is just a trick of perspective. 

What orbital parameters did you use for this visualization?

[TheRadicalModerate]

What dogleg?

I don't understand that trajectory.  It seems like it is curving to the north?  Or maybe that is just a trick of perspective. 

What orbital parameters did you use for this visualization?

Yeah, I was going to point out the same thing.  I think that's just Google Earth's version of a shortest-path line, not a great circle.  The real trajectory would arc down a bit.  However, there's still a decent path through the Florida Strait.  Note that there's also a path through the Yucatan Channel, but the Strait is almost certainly the lowest-inclination one.

Update:  I now realize that it can't be a shortest-path line, because that's a great circle, which would bend south.  No clue how that line was derived.

[OTV Booster]

The document also says that "various payload bay configurations are available and allow for fully autonomous deployment of cargo to Earth (!)..."

There's also this on the following page

The cargo version can also be used for rapid point-to-point Earth transport.

I don't think we've had cargo P2P specifically mentioned by SpaceX before, although it has been speculated on NSF as a P2P starter business.

IMO P2P cargo makes more sense than passenger. People are a PITA to hurd.

[Barley]

Update:  I now realize that it can't be a shortest-path line, because that's a great circle, which would bend south.  No clue how that line was derived.

That line probably does bend south.  The map has an odd perspective.  Note the relative positions of the southernmost part of the Gulf of Mexico, Jamaica and Puerto Rico, which have the same latitude.  Not horizontal, and not a straight line.

[OTV Booster]

Update:  I now realize that it can't be a shortest-path line, because that's a great circle, which would bend south.  No clue how that line was derived.

That line probably does bend south.  The map has an odd perspective.  Note the relative positions of the southernmost part of the Gulf of Mexico, Jamaica and Puerto Rico, which have the same latitude.  Not horizontal, and not a straight line.

It’s probably a lambert conformal conic projection. This projection shows of lines of latitude as concentric arcs and longitude as rays. The latitude measurements are not linear, nor do they change in a simple way. I grabbed this link because it had a clear visual representation of how this projection is done. The image below is courtesy of eye4software.com.

https://www.eye4software.com/hydromagic/documentation/map-projections/lambert-conformal-conic-projection/

Google can play with this projection in many ways to optimize for the viewed latitude. This will screw up other latitudes. They may even be using another projection. In short, I wouldn’t depend on google for rocket navigation.

Phil.

[livingjw]

Update:  I now realize that it can't be a shortest-path line, because that's a great circle, which would bend south.  No clue how that line was derived.

That line probably does bend south.  The map has an odd perspective.  Note the relative positions of the southernmost part of the Gulf of Mexico, Jamaica and Puerto Rico, which have the same latitude.  Not horizontal, and not a straight line.

It’s probably a lambert conformal conic projection. This projection shows of lines of latitude as concentric arcs and longitude as rays. The latitude measurements are not linear, nor do they change in a simple way. I grabbed this link because it had a clear visual representation of how this projection is done. The image below is courtesy of eye4software.com.

https://www.eye4software.com/hydromagic/documentation/map-projections/lambert-conformal-conic-projection/

Google can play with this projection in many ways to optimize for the viewed latitude. This will screw up other latitudes. They may even be using another projection. In short, I wouldn’t depend on google for rocket navigation.

Phil.

A globe and a piece of string works really well. :^)

[rsdavis9]

Update:  I now realize that it can't be a shortest-path line, because that's a great circle, which would bend south.  No clue how that line was derived.

That line probably does bend south.  The map has an odd perspective.  Note the relative positions of the southernmost part of the Gulf of Mexico, Jamaica and Puerto Rico, which have the same latitude.  Not horizontal, and not a straight line.

It’s probably a lambert conformal conic projection. This projection shows of lines of latitude as concentric arcs and longitude as rays. The latitude measurements are not linear, nor do they change in a simple way. I grabbed this link because it had a clear visual representation of how this projection is done. The image below is courtesy of eye4software.com.

https://www.eye4software.com/hydromagic/documentation/map-projections/lambert-conformal-conic-projection/

Google can play with this projection in many ways to optimize for the viewed latitude. This will screw up other latitudes. They may even be using another projection. In short, I wouldn’t depend on google for rocket navigation.

Phil.

A globe and a piece of string works really well. :^)

Google maps also has a globe icon that changes to north up globe projection.

[Twark_Main]

"Globe projection" is nothing of the sort, it's just a general perspective projection centered on the viewing point. It has size and/or angle distortions just like any other map projection.

[rsdavis9]

"Globe projection" is nothing of the sort, it's just a general perspective projection centered on the viewing point. It has size and/or angle distortions just like any other map projection.

Well the globe icon in google does seem to show the straight lines you see for great circle routes the same as the string on a globe.

[Twark_Main]

"Globe projection" is nothing of the sort, it's just a general perspective projection centered on the viewing point. It has size and/or angle distortions just like any other map projection.

Well the globe icon in google does seem to show the straight lines you see for great circle routes the same as the string on a globe.

Agreed. A map projection doesn't necessarily distort everything. But it's not a "globe projection" (there's no such thing), it's just a general perspective projection.

I never understood this worry from TheRadicalModerate that Google might not be using great circle distances/plotting. What else would they use, rhumb lines? There's no rational reason why they would do that. The math for great circles is not hard or expensive.

Anyway, we're drifting O/T. TL;DR, there is no dogleg required.

[TheRadicalModerate]

I never understood this worry from TheRadicalModerate that Google might not be using great circle distances/plotting.

I wasn't worried, just trying to figure out what had been done in that first map image.  I still don't know.

Anyway, we're drifting O/T. TL;DR, there is no dogleg required.

Two things that are not quite O/T, or at least things that a payload user could conceivably want guidance on:

1) Is there a performance advantage to launch to GTO or GEO from BC over KSC?  At maximum, you get 7 m/s more rotational boost, but it'd be nice to know if they can launch due east out of BC.  It appears to be a near thing:  Launching at 90º azimuth should wind up with a trajectory that takes Starship over the northern Keys (Plantation Key, specifically) and Andros Island in the Bahamas.

2) Is it possible to do tanker launches out of BC to 28.5º inclination, allowing them to service Starships launched out of KSC?  You can obviously do a plane change from 26º to 28.5º or vice versa, but that costs about 340 m/s to do in LEO.  If BC launches can go straight to 28.5º, it'll significantly reduce the total campaign time for a heavy BEO mission, which you'd like to do to minimize boiloff.  You can do that a lot faster if you can alternate tanker launches between KSC and BC, but the BC track to 28.5º goes over Central Florida.

[warp99]

I never understood this worry from TheRadicalModerate that Google might not be using great circle distances/plotting.

I wasn't worried, just trying to figure out what had been done in that first map image.  I still don't know.

Anyway, we're drifting O/T. TL;DR, there is no dogleg required.

Two things that are not quite O/T, or at least things that a payload user could conceivably want guidance on:

1) Is there a performance advantage to launch to GTO or GEO from BC over KSC?  At maximum, you get 7 m/s more rotational boost, but it'd be nice to know if they can launch due east out of BC.  It appears to be a near thing:  Launching at 90º azimuth should wind up with a trajectory that takes Starship over the northern Keys (Plantation Key, specifically) and Andros Island in the Bahamas.

2) Is it possible to do tanker launches out of BC to 28.5º inclination, allowing them to service Starships launched out of KSC?  You can obviously do a plane change from 26º to 28.5º or vice versa, but that costs about 340 m/s to do in LEO.  If BC launches can go straight to 28.5º, it'll significantly reduce the total campaign time for a heavy BEO mission, which you'd like to do to minimize boiloff.  You can do that a lot faster if you can alternate tanker launches between KSC and BC, but the BC track to 28.5º goes over Central Florida.

There is an available clue that the SpaceX property purchase company for Boca Chica properties is called Dogleg Park LLC. 
It is called that because a small dogleg is required once the rocket clears Key West on the southern tip of Florida to run just north of Cuba over the St Nicolas Channel and avoid the Bahamas.  The track does run over the Inagua Islands in the West Indies but that is far enough downrange and with a low enough population (~913) that it should not be a major issue. 

The benefit of the dogleg is that it should bring the orbital track closer to that from Cape Canaveral so a small further dogleg up range would put it on an inclination of 28.5º while not costing much in terms of delta V.

[ZachF]

I never understood this worry from TheRadicalModerate that Google might not be using great circle distances/plotting.

I wasn't worried, just trying to figure out what had been done in that first map image.  I still don't know.

Anyway, we're drifting O/T. TL;DR, there is no dogleg required.

Two things that are not quite O/T, or at least things that a payload user could conceivably want guidance on:

1) Is there a performance advantage to launch to GTO or GEO from BC over KSC?  At maximum, you get 7 m/s more rotational boost, but it'd be nice to know if they can launch due east out of BC.  It appears to be a near thing:  Launching at 90º azimuth should wind up with a trajectory that takes Starship over the northern Keys (Plantation Key, specifically) and Andros Island in the Bahamas.

Most of the dV savings for lower latitudes comes from inclination changes. The savings on 28.6d vs 26d is about 57m/s.

This normally wouldn't be huge, but on a GTO trajectory the Starship itself will be like 85%+ of the mass.

[TheRadicalModerate]

There is an available clue that the SpaceX property purchase company for Boca Chica properties is called Dogleg Park LLC. 
It is called that because a small dogleg is required once the rocket clears Key West on the southern tip of Florida to run just north of Cuba over the St Nicolas Channel and avoid the Bahamas.  The track does run over the Inagua Islands in the West Indies but that is far enough downrange and with a low enough population (~913) that it should not be a major issue. 

The benefit of the dogleg is that it should bring the orbital track closer to that from Cape Canaveral so a small further dogleg up range would put it on an inclination of 28.5º while not costing much in terms of delta V.

Isn't a dogleg at Key West pretty far downrange?  Won't that cost at least hundreds of m/s?  Note that a simple on-orbit inclination change costs about 340 m/s.

[TheRadicalModerate]

Most of the dV savings for lower latitudes comes from inclination changes. The savings on 28.6d vs 26d is about 57m/s.

This normally wouldn't be huge, but on a GTO trajectory the Starship itself will be like 85%+ of the mass.

I thought the only dv savings were the differences in rotational boost when you're launching due east.  Doing the latitudes a bit more carefully, I get 10 m/s.  How do you get 57?

[warp99]

There is an available clue that the SpaceX property purchase company for Boca Chica properties is called Dogleg Park LLC. 
It is called that because a small dogleg is required once the rocket clears Key West on the southern tip of Florida to run just north of Cuba over the St Nicolas Channel and avoid the Bahamas.  The track does run over the Inagua Islands in the West Indies but that is far enough downrange and with a low enough population (~913) that it should not be a major issue. 

The benefit of the dogleg is that it should bring the orbital track closer to that from Cape Canaveral so a small further dogleg up range would put it on an inclination of 28.5º while not costing much in terms of delta V.

Isn't a dogleg at Key West pretty far downrange?  Won't that cost at least hundreds of m/s?  Note that a simple on-orbit inclination change costs about 340 m/s.

Yes that is around 1000km down range and probably half way to orbital speed.  Therefore the dogleg will cost around half the delta V of the same change done at orbital velocity so around 170 m/s on your figures.  Compared with total delta V for a launch of around 9300 m/s this is an acceptable loss. 

Edit: Using FH trajectory as an approximation for Starship I get 1000km downrange at 420s after liftoff at which point horizontal velocity is 5100 m/s.  So the cost of the plane change is a bit higher than I estimated at around 230 m/s so 2.5% of the total required to get to LEO.

References:
Downrange distance vs time: https://i.imgur.com/yh4OleU.png
Horizontal velocity vs time: https://i.imgur.com/90hNWyf.png

[tbellman]

I thought the only dv savings were the differences in rotational boost when you're launching due east.  Doing the latitudes a bit more carefully, I get 10 m/s.  How do you get 57?

If you launch due east (or due west), following the great circle that pass through the launch point (and is parallel to the equator at that point), then you will get an inclination that is exactly the same as the latitude of the launch point.  If you are OK with that inclination, then all is fine and dandy, and the difference between latitudes will be the rotational boost.

But if you want some other inclination, then you need an inclination change.  And geostationary orbit is by definition at 0° inclination.  GTO can in principle be any inclination, but you will end up with different amounts of Δv remaining to take the payload to GEO.  This is exactly why Kourou in French Guiana is so nice for launching satellites that are destined for GEO.

If, instead of following the great circle, you continue straight eastward, parallel to the equator, as you increase speed and altitude, you won't be in any better position.  You will just be expending energy to move to other great circles with the same inclination, but with other longitudes of ascending node (i.e, it will be turned some amount around the Earth axis).  When you reach orbital speed and altitude, and stop your engines, you will start following that great circle, with the same inclination as the latitude where you stopped the engines (and as you flew parallel to the equator the entire time, you would still be at the same latitude as where you launched from).

[OTV Booster]

Update:  I now realize that it can't be a shortest-path line, because that's a great circle, which would bend south.  No clue how that line was derived.

That line probably does bend south.  The map has an odd perspective.  Note the relative positions of the southernmost part of the Gulf of Mexico, Jamaica and Puerto Rico, which have the same latitude.  Not horizontal, and not a straight line.

It’s probably a lambert conformal conic projection. This projection shows of lines of latitude as concentric arcs and longitude as rays. The latitude measurements are not linear, nor do they change in a simple way. I grabbed this link because it had a clear visual representation of how this projection is done. The image below is courtesy of eye4software.com.

https://www.eye4software.com/hydromagic/documentation/map-projections/lambert-conformal-conic-projection/

Google can play with this projection in many ways to optimize for the viewed latitude. This will screw up other latitudes. They may even be using another projection. In short, I wouldn’t depend on google for rocket navigation.

Phil.

A globe and a piece of string works really well. :^)

Ai, ai cap’n. If it was good enough for Nelson...

Meanwhile back at Mission Planning...

[TheRadicalModerate]

But if you want some other inclination, then you need an inclination change.  And geostationary orbit is by definition at 0° inclination.  GTO can in principle be any inclination, but you will end up with different amounts of Δv remaining to take the payload to GEO.  This is exactly why Kourou in French Guiana is so nice for launching satellites that are destined for GEO.

Got it.  I was only looking at launch to initial orbit insertion and not including the plane change at GTO apogee.  Thanks.

[ulm_atms]

Starship will get a small boost for GTO payloads when launching from Boca Chica vs the Cape (28.6d vs 26.0d), about 65-70m/s.

Not a huge boost, but enough to theoretically add ~3 tonnes to the GTO-1800 payload vs the cape.

Does that include the losses from the dogleg it has to do to get out of the Gulf of Mexico without overflying anyone?

What dogleg?

I don't understand that trajectory.  It seems like it is curving to the north?  Or maybe that is just a trick of perspective. 

What orbital parameters did you use for this visualization?

I used google earth with a P2P ruler line between those two points and adjusted the atlantic point so that the line overflew basically nothing.  The curve is just a projection artifact.

I changed the angle to make it look straighter :-)

EDIT:  Added picture.  Now it is straight for everyone.

[ZachF]

Most of the dV savings for lower latitudes comes from inclination changes. The savings on 28.6d vs 26d is about 57m/s.

This normally wouldn't be huge, but on a GTO trajectory the Starship itself will be like 85%+ of the mass.

I thought the only dv savings were the differences in rotational boost when you're launching due east.  Doing the latitudes a bit more carefully, I get 10 m/s.  How do you get 57?

Because to get to geostationary, you need to do an inclination change to 0 degrees.

This is why Ariane is such a good launcher to GTO.... going to a 200 x 35,786 from 6 degrees only takes a further 1500m/s to get to GEO, while 200 x 35,786 from 27 degrees takes 1800m/s. The boost from rotation is small compared to the gain from needing a smaller eventual inclination change.

https://gtocalc.github.io/

[warp99]

Starship will get a small boost for GTO payloads when launching from Boca Chica vs the Cape (28.6d vs 26.0d), about 65-70m/s.

Not a huge boost, but enough to theoretically add ~3 tonnes to the GTO-1800 payload vs the cape.

Does that include the losses from the dogleg it has to do to get out of the Gulf of Mexico without overflying anyone?

What dogleg?

I don't understand that trajectory.  It seems like it is curving to the north?  Or maybe that is just a trick of perspective. 

What orbital parameters did you use for this visualization?

I used google earth with a P2P ruler line between those two points and adjusted the atlantic point so that the line overflew basically nothing.  The curve is just a projection artifact.

I changed the angle to make it look straighter :-)

EDIT:  Added picture.  Now it is straight for everyone.

Is there not an issue with the fact that this trajectory is at 96 degrees from true North so at 6 degrees inclination to the equator when it needs to be at least 26 degrees when launching from Boca Chica at 26 degrees North?

[ZachF]

Starship will get a small boost for GTO payloads when launching from Boca Chica vs the Cape (28.6d vs 26.0d), about 65-70m/s.

Not a huge boost, but enough to theoretically add ~3 tonnes to the GTO-1800 payload vs the cape.

Does that include the losses from the dogleg it has to do to get out of the Gulf of Mexico without overflying anyone?

What dogleg?

I don't understand that trajectory.  It seems like it is curving to the north?  Or maybe that is just a trick of perspective. 

What orbital parameters did you use for this visualization?

I used google earth with a P2P ruler line between those two points and adjusted the atlantic point so that the line overflew basically nothing.  The curve is just a projection artifact.

I changed the angle to make it look straighter :-)

EDIT:  Added picture.  Now it is straight for everyone.

Is there not an issue with the fact that this trajectory is at 96 degrees from true North so at 6 degrees inclination to the equator when it needs to be at least 26 degrees when launching from Boca Chica at 26 degrees North.

A direct east trajectory from BC is going to look something like this:

[cuddihy]

So you’re going to cross US-1 somewhere in the Keys, then. Where would the dogleg take place to avoid this?

[freddo411]

So you’re going to cross US-1 somewhere in the Keys, then. Where would the dogleg take place to avoid this?

Is this necessarily a problem?   You are fly over shipping lanes as well that might have a higher person-density than the roadway connecting the keys.

[TheRadicalModerate]

A direct east trajectory from BC is going to look something like this:

Yup, that's what I got, too.  Note that it goes over the keys, and it's not incredibly far from Homestead.  So this trajectory may have range safety issues.  However, if you make the trajectory go just south of Key West, the azimuth is about 93º, and the extra delta-v to GEO is 5 m/s.  I suspect that that's tolerable.

BTW:  If you go by the FH trajectories that warpgg posted, you reach orbit at about the same longitude as the Keys, so it's possible that the impact point for late-stage failures is moving so fast that you can fly over the Keys and hit the 1/10,000 risk to the public with no problem.

The other issue, though, is matching 28.6º for refueling launches from both pads.  That requires an azimuth of 103º, which requires overflying western Cuba before burnout, or azimuth 77º, which overflies Tampa before burning out almost exactly over Canaveral.  However if you launch due east (90º) and then dogleg south about 670 km out, you should be able to skirt the northern coast of Cuba and hit 28.6º.  (Note that warpgg had also looked at a dogleg just south of Key West, but I think you can do better than that.)

Note:  I'm doing all of this graphically, drawing lines in Google Earth, based on knowing the needed azimuth from BC for a given inclination.  So I have no idea how much a dogleg 670 km out costs.  (FH speed that far downrange is about 4000 m/s.)  I'm guessing it'll be close to 200 m/s.  It might be more optimal to instead launch both the KSC and BC stuff to 30.8º, which would allow the BC launches to go south of Cuba with no doglegs.

Another obvious question is whether you actually need refueling launches from both pads.  Building up prop in your aggregation tanker (i.e., the tanker that actually docks with the payload Starship and does one refueling operation) will incur some amount of boil-off, so aggregating as quickly as possible counts for something.  There's also the military to consider, but I suspect that they're going to hate the idea of refueling until the risk gets quantified to within an inch of its life.

[warp99]

Is there not an issue with the fact that this trajectory is at 96 degrees from true North so at 6 degrees inclination to the equator when it needs to be at least 26 degrees when launching from Boca Chica at 26 degrees North.

A direct east trajectory from BC is going to look something like this:

OK - so it looks like the dog leg is the other way around so launch south east into the Gulf at something like 110 degrees until you are roughly level with a line 30 km off the north coast of Cuba then turn to a heading close to 90 degrees to follow a similar trajectory to the one you have shown but further south.  The change in heading is larger at 20 degrees but it occurs well before MECO and the velocity is much lower at perhaps 2000 m/s so the overall loss of delta V will be lower than doing the turn further out close to the Keys. 

This misses all the population centers by a wide margin and can track over the center of Inagua Island which is a wildlife sanctuary so almost completely unpopulated. 

To match the inclination of a launch from Canaveral do not fly quite so far south by launching at say 107.5 degrees and then do a dogleg to 92.5 degrees.

[niwax]

Note:  I'm doing all of this graphically, drawing lines in Google Earth, based on knowing the needed azimuth from BC for a given inclination.  So I have no idea how much a dogleg 670 km out costs.  (FH speed that far downrange is about 4000 m/s.)  I'm guessing it'll be close to 200 m/s.  It might be more optimal to instead launch both the KSC and BC stuff to 30.8º, which would allow the BC launches to go south of Cuba with no doglegs.

That distance poses the question what actually counts as a dangerous overflight. Boca Chica - Florida is ~1500km. During normal launches, the NOTAM goes nowhere near the 600km droneship. For the polar dogleg, they're avoiding highly populated Cuba at ~700km but then have no problem overflying South America at ~2000km.

[warp99]

Note:  I'm doing all of this graphically, drawing lines in Google Earth, based on knowing the needed azimuth from BC for a given inclination.  So I have no idea how much a dogleg 670 km out costs.  (FH speed that far downrange is about 4000 m/s.)  I'm guessing it'll be close to 200 m/s.  It might be more optimal to instead launch both the KSC and BC stuff to 30.8º, which would allow the BC launches to go south of Cuba with no doglegs.

That distance poses the question what actually counts as a dangerous overflight. Boca Chica - Florida is ~1500km. During normal launches, the NOTAM goes nowhere near the 600km droneship. For the polar dogleg, they're avoiding highly populated Cuba at ~700km but then have no problem overflying South America at ~2000km.

That makes sense as they are reaching orbit around 1500 km out.  Yes there will be a brief moment when the instantaneous engine out trajectory paints across South America but operating the self destruct would ensure that very few sizable pieces of S2 made it through re-entry.  Probably just the engine minus the extended bell. 

There is a similar situation during a normal geosynchronous launch when the engine out trajectory paints across Africa very briefly.

[ulm_atms]

Ok, try this.

Here is all the three western hemisphere GTO launch trajectories based on what I have seen (red lines).

Boca has two.  One (green) would be normal if flyover didn't matter in Florida and you were going for lowest inclination.  Purple is the path with minimal flyover of land.

For green, the inclination is slightly less then the Cape.  The purple, it slightly more then the Cape due to the angle.

If I did something wrong please explain...but these orbits are perfect circles and do go back over the starting spot if drawn around the globe completely.

NOTE:  If this needs to be in a different thread mods...please move..this is just the thread I saw this come up in.

[philw1776]

A direct east trajectory from BC is going to look something like this:

Yup, that's what I got, too.  Note that it goes over the keys, and it's not incredibly far from Homestead.  So this trajectory may have range safety issues.  However, if you make the trajectory go just south of Key West, the azimuth is about 93º, and the extra delta-v to GEO is 5 m/s.  I suspect that that's tolerable.

BTW:  If you go by the FH trajectories that warpgg posted, you reach orbit at about the same longitude as the Keys, so it's possible that the impact point for late-stage failures is moving so fast that you can fly over the Keys and hit the 1/10,000 risk to the public with no problem.

The other issue, though, is matching 28.6º for refueling launches from both pads.  That requires an azimuth of 103º, which requires overflying western Cuba before burnout, or azimuth 77º, which overflies Tampa before burning out almost exactly over Canaveral.  However if you launch due east (90º) and then dogleg south about 670 km out, you should be able to skirt the northern coast of Cuba and hit 28.6º.  (Note that warpgg had also looked at a dogleg just south of Key West, but I think you can do better than that.)

Note:  I'm doing all of this graphically, drawing lines in Google Earth, based on knowing the needed azimuth from BC for a given inclination.  So I have no idea how much a dogleg 670 km out costs.  (FH speed that far downrange is about 4000 m/s.)  I'm guessing it'll be close to 200 m/s.  It might be more optimal to instead launch both the KSC and BC stuff to 30.8º, which would allow the BC launches to go south of Cuba with no doglegs.

Another obvious question is whether you actually need refueling launches from both pads.  Building up prop in your aggregation tanker (i.e., the tanker that actually docks with the payload Starship and does one refueling operation) will incur some amount of boil-off, so aggregating as quickly as possible counts for something.  There's also the military to consider, but I suspect that they're going to hate the idea of refueling until the risk gets quantified to within an inch of its life.

In winter I'm just south of Tampa on FL West coast. Issuing written permission for SpaceX to overfly my property.

[TheRadicalModerate]

OK - so it looks like the dog leg is the other way around so launch south east into the Gulf at something like 110 degrees until you are roughly level with a line 30 km off the north coast of Cuba then turn to a heading close to 90 degrees to follow a similar trajectory to the one you have shown but further south.  The change in heading is larger at 20 degrees but it occurs well before MECO and the velocity is much lower at perhaps 2000 m/s so the overall loss of delta V will be lower than doing the turn further out close to the Keys. 

There's no dog leg. But if you launch due east (90º), you go over the Florida Keys (see his map).  If you launch at 93º, you miss the populated parts of the Keys, and likely are in orbit before overflying populated parts of the Bahamas.

[jthygesen]

Just a small note about the Users Guide: The author is registered in the meta data as Jessica Jensen, whom I believe is director of Dragon Mission management.

Could be a real sign of SpaceX moving personnel from Dragon to StarShip. Or not.

[[email protected]]

Just a small note about the Users Guide: The author is registered in the meta data as Jessica Jensen, whom I believe is director of Dragon Mission management.

Could be a real sign of SpaceX moving personnel from Dragon to StarShip. Or not.

FYI she's indeed assigned to the Starship program when she inaugurated the B1035 core on Space Center Houston. Her position is the director of Starship mission hardware & operations

https://m.facebook.com/groups/2387776317?view=permalink&id=10158427094176318

And I'm sure the workforce transition from Falcon & Dragon to Starship is already happening. For instance, Lars Blackmore who's the main person behind the amazing Falcon landings, his future projects is leading the Starship entry & landing (which is cool, because he already have an experience on Mars landing in JPL)

http://www.eng.cam.ac.uk/news/alumni-stories-meet-principal-rocket-landing-engineer-spacex

[Twark_Main]

OK - so it looks like the dog leg is the other way around so launch south east into the Gulf at something like 110 degrees until you are roughly level with a line 30 km off the north coast of Cuba then turn to a heading close to 90 degrees to follow a similar trajectory to the one you have shown but further south.  The change in heading is larger at 20 degrees but it occurs well before MECO and the velocity is much lower at perhaps 2000 m/s so the overall loss of delta V will be lower than doing the turn further out close to the Keys. 

There's no dog leg. But if you launch due east (90º), you go over the Florida Keys (see his map).  If you launch at 93º, you miss the populated parts of the Keys, and likely are in orbit before overflying populated parts of the Bahamas.

The criteria isn't about "overflying," it's when the instantaneous impact point passes over populated areas.

[TheRadicalModerate]

There's no dog leg. But if you launch due east (90º), you go over the Florida Keys (see his map).  If you launch at 93º, you miss the populated parts of the Keys, and likely are in orbit before overflying populated parts of the Bahamas.

The criteria isn't about "overflying," it's when the instantaneous impact point passes over populated areas.

Fair enough, but the impact point is moving downrange so quickly near burnout that the probability of impact anywhere in the Bahamas is extremely low.  Beyond that, the only areas of the Bahamas along the track are the mangrove swamps of South Andros, the southern tip of Exuma and a mile-wide strip of Long Island, half of which is a golf course.

[john smith 19]

Imo it would be better to change the sat design itself. Instead of using the best of the best components, go down 2 steps in capability (for example solar array efficiency) and make the sats cost only a fraction of today‘s expensive sats. Of course they would weigh then perhaps the double, but with cheap transportation, it could reduce the cost of space „stuff“ significantly. No longer the need to go for ultra lightweight with expensive components and manufacturing. That‘s the real benefit imo of (if successful) Starship and perhaps New Glenn.

An observation first made in the 70's at several shuttle conferences on space servicing.

Why do you need triple redundant systems when if it fails just recover it and bring it in for servicing?

[john smith 19]

By publishing this guide now, SpaceX is creating some level of commitment to a baseline specification for engineers and designers to engineer for. Deviating from these figures in a detrimental manner would be harmful for their business.

Correct.

[john smith 19]

I'm late to this party as I simply didn't expect them to have written a user manual this early.

So what' we seem to be looking at is a replaceable nose, rather like the SR71, which had several.

Fairchild did this in the late 40s/early 50s and the Sikorsky Skycrane helicopter also had a cargo module (never sure how many of those were actually built).

Obvious benefit is the "payload module" (it's a lot more than a fairing now) is a lot lighter than the fully load SS to crane up to the top. Say 10t for the module. That's 160t to move about 140m into the sky.

Two variants, cargo and passenger carrier with the implication that the cargo carrier becomes a tanker with additional tanks in the cargo area.  I'm guessing a trade off between what's left in the main tanks to refuel the payload carrying SS and what additional can be carried in the payload bay tanks (perhaps pressurant gas to speed up the expulsion?).

Vertical integration has to be in because they'll never get the biggest (most expensive) NSS payloads without it.

This all makes perfect sense to me.

As for the passenger carrier worries about windows SS comes in a lot like shuttle, very high AoA. I suspect it will fly that way to the moon as well to protect the window. Of course the downside is the belly TPS will get hit by MMOD, but I'm guessing the plan a short period in LEO before setting off and (I'm guessing here) MMOD levels drop a lot once reasonably far away from earth (like above GEO for example).

I also suspect there will be fast acting cover system to seal a window if it's blown in. Operating with the speed of an airbag during a car crash (might even be a similar design, but non deflating, to give time to evacuate the area and plan repairs.

The replaceable nose also makes P2P cargo or passenger management (in principle) much simpler. Load everyone on at ground level then lift the whole package to the nose.  A 20m long PM is nothing in terms of height to putting several hundred passengers on at full height one at a time. You can leverage most of what's been learned about the process in the last 90 years of commercial airline operations.


And if the wind picks up those aero surfaces will make great sails. Being inside the module will get quite interesting.  Like being in a cable car at a very windy location for example. I've never had that experience, but I'll bet someone here has.

And then of course we get to the join between the PM and the rest of SS. That will have to be made and broken for every flight.

BTW that plan with the trunions mounting to the walls of the PM? Failure to standardize those was a major cause of why shuttle payload integration was a PITA,

Perhaps SX should contact Hempsell Astronautics. They did quite a lot of work on this subject.