Amateur rocket to orbit

[ChrisSpace]

On 6 July 2017 at 7:18 AM, sevenperforce said:

Obviously I'm not suggesting that anyone attempt to build an orbital rocket in their backyard. This is more a design question -- what would be the simplest, cheapest, most dependable option for getting a rocket into orbit at an amateur level?

This exact question is pretty much the basis for... actually 2 novel ideas I have, plus a sub-story of Steel Sky (which is currently in the "not cancelled but postponed development for the tim being" phase of development). Unfortunately all my info is on my computer back at home so for now I'll have to go off memory.

On 6 July 2017 at 7:18 AM, sevenperforce said:

Things to consider include:

• Staging: Serial or parallel?
• Fuel: Liquid, hybrid, or solid?
• Guidance: Gimbal, differential thrust, or aerodynamic?

For staging, remember that even SpaceX had trouble with a multi-core rocket design. Best to use a simple "one stage above another above another" approach. Fuel depends on what can be afforded, but "R-Candy" and top-tier model rocket engines would be a good place to start. Guidance would best be done using a combination of reaction wheels and aerodynamic control surfaces, I think.

On 6 July 2017 at 9:33 AM, Thor Wotansen said:

I saw a video on YouTube by Copenhagen Suborbital where they use vanes in the exhaust to vector the thrust for control.  This could be done fairly easily with some beefy servos , however you might need to make the vanes out of tungsten or something else heat resistant.

CS is rather low-budget so it's probably a good "case study" regarding propulsion and navigation.

On 8 July 2017 at 11:12 AM, RedKraken said:

The engineering approach to amateur sub-orbitals : http://watzlavick.com/robert/rocket/index.html

Wonder if he will roll out a methane (or just LNG) engine for better performance?

Maybe change pressure-fed to electric turbopumps that are 3d printed?

Then comes adding a 2nd stage to try for orbit.

Depending on your payload, that could be an engine for any stage, although I'd be curious to see the price tag.

On 10 July 2017 at 5:13 AM, TheEpicSquared said:

I think that before going any further, we need to decide on the staging. After all, parallel and Arian stages rocket do work differently. Will we use parallel or serial staging? Both? If serial, 2 or 3 stage design? Have both parallel and serial, like Falcon Heavy? IMO we need to get that sorted before we go any further.

Depends on your propulsion capabilities, for example I have a simple example described here:

https://forum.kerbalspaceprogram.com/index.php?/topic/165955-everyday-space-drives/

On 10 July 2017 at 5:23 AM, wumpus said:

http://www.aerotech-rocketry.com/resources.aspx sells high-power ammonium perclorate.  One rather large engine (12.6kg!) appears to have an Isp of 188s (I'm uneasy about my calculation).
http://www.aerotech-rocketry.com/customersite/resource_library/Catalogs_Flyers_Data_Sheets/ldrs-27_prod_data_sheets.pdf (bottom rocket).  Note that adding a vacuum bell might help.

Those engines only burn for a few seconds at most.

On 12 September 2018 at 4:43 AM, BillKerman1234 said:

The consensus of everyone on the rocket design was to have ether a single core with 4 boosters and a upper kick stage, or, if that didn’t have enough delta v, 4 cores and 4 boosters with the same upper kick stage. The cores and boosters would be identical, using fins to steer in the atmosphere, while using RCS to steer in vacuum. The motors would be hybrids, running of a propellent that most people agreed would be some sort of htp/gasoline (I think). The RCS would be built into the core, and the side boosters would use the empty holes where the RCS would be on the identical core stage to mount gas thrusters to act as sep motors. The boosters and core would be recovered with parachutes. The upper stage would be a cluster of ether solid or hybrid motors. For simplicity it would have no guidance system other than gyros and accelerometers.

All seems to make sense to me.

[ChrisSpace]

On 12 September 2018 at 4:43 AM, BillKerman1234 said:

I’ve included some diagrams of the proposed rocket, and some calculations for it.

Have you tried simulating the system in Realism Overhaul? After that, someone should start calculating the cost of each component.

21 hours ago, BillKerman1234 said:

Also for consideration is Australia since sounding rockets have been launched from there every now and again and even also orbital rockets (the black arrow). Plus Australia is outside the US (although if something like ITAR still apples I’m not sure).

This is where I'm going to place my totally-not-at-all-biased vote. But seriously, it seems to be the best option.

20 hours ago, BillKerman1234 said:

Also, on another topic, does anyone here know how to 3D model stuff?

"Openrocket" is a useful software tool for designing model rockets. Besides that there's CoaDE, but it doesn't simulate engine performance in atmosphere, and Realism Overhaul which would be useful if the specific rocket parts could be "custom-made" by modders.

20 hours ago, James Kerman said:

There has been some talk recently of a launch facility being built by Equatorial Launch Australia in the Northern Territory (closer to the equator) that intends to provide services to commercial, research and government organizations.

https://www.news.com.au/technology/science/space/australias-first-commercial-space-base-to-launch-rockets-within-a-year/news-story/eb7841c5b39e04fd31302e8b1056e3ab

Exactly what I was thinking.

8 hours ago, BillKerman1234 said:

Think about it, what is a rocket? The fuel tanks are metal tubes with caps on the ends, the motor is a sphere with a funny shaped cone on the end, the fuel pumps are the same as you get in fire trucks (in fact there was a famous incident where von Brown was designing the V2 rocket and asked a bunch of pump engineers to design him a pump to some specifications only to have them all tell him that he was describing exactly a fire engine pump), and the guidance system can just be fins. Okay, I know this is vastly simplified, but when you think about it, there’s no real reason rockets have to be so expensive.

As I said, we need to find the cost of each sub-component as accurately as possible.

[Mad Rocket Scientist]

1 hour ago, qzgy said:

Wait you're already thinking about reusability? From a skeptical point of view, I doubt reuse is possible with at least the early designs. Any stage reuse means adding more equipment and weight to a design we want as light as possible. Even back in the early days, stage reuse was basically a thing that didn't happen because of the extra complexity.

The way I see the current project is that its an experimental prototype rocket, not a viable launch system. If it becomes expensive, so be it.

It seems like it would be worth it to budget for at least first stage parachute return, even if it is too damaged to reuse. After all, that way we can see what happened to it better, plus (either as a design or an practice) this is just for fun, and that sounds fun.

[James Kerman]

I've been thinking about the payload and mission as we would probably need a ballpark mass before we could design a rocket to launch it.

I propose we design a payload that can transmit Kevin Macleod's Arcadia (the orbital theme from KSP) from orbit, assuming the vessel will be capable of transmitting telemetry.

Tentative list of PL Components:
Nosecone/frame
Electronics
Guidance including Gimbal and reference
Battery and solar panels
Transmitter and Antenna
De-orbiting system

What else would the PL need?

[BillKerman1234]

Quote

For staging, remember that even SpaceX had trouble with a multi-core rocket design. Best to use a simple "one stage above another above another" approach. Fuel depends on what can be afforded, but "R-Candy" and top-tier model rocket engines would be a good place to start. Guidance would best be done using a combination of reaction wheels and aerodynamic control surfaces, I think.

First off, a mult core design is vastly simpler than a ‘one on top of the other’ design (which is one of the hardest things to pull of in rocketry). Also, R-candy and model rocket engines. Seriously? The isp of those would be less than 200s! And while building our own solids might work, it would be very difficult. The easiest thing to do would be hybrids, but liquids would be even better. 

Also, just for the record, reaction wheels in real life are much less powerful than in ksp (even RO), so there not good for ascent guidance. RCS or exhaust tabs/fins are the way to go.

Edited September 13, 2018 by BillKerman1234

[ChrisSpace]

28 minutes ago, James Kerman said:

Tentative list of PL Components:
Nosecone/frame
Electronics
Guidance including Gimbal and reference
Battery and solar panels
Transmitter and Antenna
De-orbiting system

What else would the PL need?

Try breaking those into smaller sub-categories, then I can get cost estimates as well as mass/volume estimates for them

[BillKerman1234]

I came up with a few ideas:

I’d say if we give ourselves 20kg of payload, that should be enough for whatever we want to do.

[BillKerman1234]

I would say if we’re going to make it recoverable, we should go with the ball shaped RV, but if it’s just going to be a normal satterlite, the a cubesat would be the best.

[BillKerman1234]

2 hours ago, James Kerman said:

I've been thinking about the payload and mission as we would probably need a ballpark mass before we could design a rocket to launch it.

I propose we design a payload that can transmit Kevin Macleod's Arcadia (the orbital theme from KSP) from orbit, assuming the vessel will be capable of transmitting telemetry.

Tentative list of PL Components:
Nosecone/frame
Electronics
Guidance including Gimbal and reference
Battery and solar panels
Transmitter and Antenna
De-orbiting system

What else would the PL need?

 

2 hours ago, ChrisSpace said:

Try breaking those into smaller sub-categories, then I can get cost estimates as well as mass/volume estimates for them

 

- Aluminium sheets for the frame, 1mm thick and, say, 1*1m total surface area

- Aluminium rods for the structure, 5mm wide, 5m long in total, cut into sections. 

- Aluminium spheres for RCS fuel, 20cm diamater

- gold foil for insulation, 2*2meter sheet

- Kale wool (a highly insulated material), 1*1m 

- Solar cells, 125mm^2 each, 20 - 30 pieces

- ham radio, very high power

- Li-po Battery, highest specs commercially available

- RPi, gen 3 b+, with vacuum prof case (probably a cast resin block that we incase it with)

- 2 Cameras (GoPro 6) incased in clear resin to space proof them

- custom CNCed metal RCS thrusters

- fuel tubing and fittings

- spark generator (modified taser) for RCS igniter 

- RCS fuel valves and servos

- accelerometers and gyroscopes

- Wireing

- a bunch of miscellaneous stuff I forgot

[ChrisSpace]

Okay, that is not what I had in mind. I was expecting that our payload would be a cubesat, or at least constructed out of cubesat parts besides the structure itself. To see what I mean, feel free to browse the following sources:

https://www.cubesatshop.com

https://www.clyde.space/products?page=1&subsystem=

https://www.isispace.nl/products/

Using satellite parts already designed to be used in space takes a lot of uncertainty and guesswork out of the payload itself, and I can even see a few modules in the above that could be used for the rocket itself (flight computers, comms systems and decouplers in particular).

[James Kerman]

20 125mm^2 solar cells @2.8W (18% efficiency) only gives us 56W and that's if they're all seeing sunlight.  Power is going to be a big issue.

 

[ChrisSpace]

2 hours ago, BillKerman1234 said:

I’d say if we give ourselves 20kg of payload, that should be enough for whatever we want to do.

 

1 hour ago, BillKerman1234 said:

I would say if we’re going to make it recoverable, we should go with the ball shaped RV, but if it’s just going to be a normal satterlite, the a cubesat would be the best.

 

Okay, so here's what I was thinking our roadmap should be:

Launch 1: Test launch vehicle, deliver relatively cheap test payload into LEO (Cubesat, Tubesat or Pocketqbe)

Launch 2: Launch much larger, more complex payload into LEO to confirm maximum lift capacity into LEO

Launch 3: Test a satellite capable of reentry, landing and recovery

Launch 4: Add new, small, upper stage to boost a small payload into deep space (exact destination TBD)

 

It's now that I'll ask the obvious questions: How are we going to afford this, and who will purchase all the required parts and materials?

Edited September 13, 2018 by ChrisSpace

[BillKerman1234]

If we manage to make blueprints for a launch vehicle and satterlite, the getting then money is easy: just ask a university to sponsor us, and In return they get to have their students work on building an orbital rocket! 

I’m also a member of the British Interplanetary Society, so I could see if I can turn this into one of their technical feasibility studys.

Edited September 13, 2018 by BillKerman1234

[BillKerman1234]

 

https://n-avionics.com/platforms/6u-cubesat-bus-m6p/ - 6u prebuilt

https://n-avionics.com/platforms/3u-cubesat-platform-plt3/ - 3u prebuilt

https://n-avionics.com/platforms/cubesat-2u-platform-plt2/ - 2u prebuilt

https://n-avionics.com/subsystems/ - individual components

How about this? I’m thinking the 6u would be good, but it might be too expensive, so we might need to go for the 2u.

Edited September 13, 2018 by BillKerman1234

[BillKerman1234]

Ok, so I’ve been thinking, and from what I can see, this is what we need to do:

1) figure out what type of satterlite our payload is gonna be

2) figure out the cost and mass of the payload

3) figure out what type of rocket propulsion system we’re going to use

4) do the math on the size, shape, and mass of the launcher

5) design the basic concept of the launch vehicle around the required propulsion system and payload

6) use ksp realism overhaul to simulate it

7) 3D model the vehicle and do cfd on it

8) get me to persuade some guys I know at the British Interplanetary Society to let me use their proper rocket simulation software 

9) 3D model the vehicle in much more detail 

10) reach out to university’s and organisations to sponsor us

11) build it

 

...sound simple enough! *sarcastic facial expression*

Edited September 13, 2018 by BillKerman1234

[Steel]

20 minutes ago, BillKerman1234 said:

Ok, so I’ve been thinking, and from what I can see, this is what we need to do:

1) figure out what type of satterlite our payload is gonna be

2) figure out the cost and mass of the payload

3) figure out what type of rocket propulsion system we’re going to use

4) do the math on the size, shape, and mass of the launcher

5) design the basic concept of the launch vehicle around the required propulsion system and payload

6) use ksp realism overhaul to simulate it

7) 3D model the vehicle and do cfd on it

8) get me to persuade some guys I know at the British Interplanetary Society to let me use their proper rocket simulation software 

9) 3D model the vehicle in much more detail 

10) reach out to university’s and organisations to sponsor us

11) build it

 

...sound simple enough! *sarcastic facial expression*

Just my quick thoughts:

If this gets to a serious point of asking for money from people, people will expect serious documentation. There are 1000's of people who have ideas for low cost rockets, but very few will have detailed sourcing plans, every single part designed, costed and sourced and a genuine plan a out how they'll build and fly a rocket. All of that stuff will be needed before anyone will even have the conversation about money.

One other thing. With a project so complex, you have to factor in the fact that there will be a lot of testing involved, some of it destructive. I can almost guarantee the first attempt (as well as the second, third and maybe fourth) at designing and building some parts will fail, and that all needs to be factored into the project.

[ChrisSpace]

42 minutes ago, BillKerman1234 said:

https://n-avionics.com/platforms/6u-cubesat-bus-m6p/ - 6u prebuilt

https://n-avionics.com/platforms/3u-cubesat-platform-plt3/ - 3u prebuilt

https://n-avionics.com/platforms/cubesat-2u-platform-plt2/ - 2u prebuilt

https://n-avionics.com/subsystems/ - individual components

 

Did you.. did you go out of your way to avoid using the sources I listed? Actually you've done well, since having a pre-built cubesat makes literally everything way easier.

44 minutes ago, BillKerman1234 said:

How about this? I’m thinking the 6u would be good, but it might be too expensive, so we might need to go for the 2u.

2U for the rocket's maiden flight, we won't really need anything more while testing the launch vehicle.

31 minutes ago, BillKerman1234 said:

Ok, so I’ve been thinking, and from what I can see, this is what we need to do:

1) figure out what type of satterlite our payload is gonna be

2) figure out the cost and mass of the payload

1) A simple cubesat on the first launch

2) If we use the NanoAvionics cubesat design, finding the cost/mass/volume should be easy

34 minutes ago, BillKerman1234 said:

3) figure out what type of rocket propulsion system we’re going to use

For this we first need to have a full rundown on all our options for propulsion systems. Here's the best sources I can find on what we might be able to buy or build:

http://aeroconsystems.com/cart/index.php?p=home

http://www.watzlavick.com/robert/rocket/regenChamber/index.html

http://www.watzlavick.com/robert/rocket/regenChamber3/index.html

http://pages.total.net/~launch/ss67b3.htm

http://www.maxentropy.net/rocketry/liquidproject/design/design.htm

39 minutes ago, BillKerman1234 said:

4) do the math on the size, shape, and mass of the launcher

5) design the basic concept of the launch vehicle around the required propulsion system and payload

6) use ksp realism overhaul to simulate it

7) 3D model the vehicle and do cfd on it

8) get me to persuade some guys I know at the British Interplanetary Society to let me use their proper rocket simulation software 

 9) 3D model the vehicle in much more detail 

Perhaps you could upload the simulations onto youtube or something for extra publicity perhaps?

40 minutes ago, BillKerman1234 said:

11) build it

Actually it's:

11) Purchase all required materials, parts, modules etc

12) Deliver everything purchased in 11 to construction site

13) Construct rocket

14) Transport rocket to launch site

15) Launch

[BillKerman1234]

@Steel Yes I know. Every component on the vehicle, down to what type of rivets we use, will need to be carefully planned out. In aerospace, every component needs to be traceable back to the ore in the ground. But right now, thats very far off. We still haven’t even completed step 1, find out the mass of the payload we want to put into orbit.

[Steel]

I'd also suggest that a forum perhaps isn't the most efficient method of planning something like this. Maybe put a Slack channel (or similar) together with interested parties, then post periodic updates on progress here to update people and generate interest.

[BillKerman1234]

Okay, I’ve just created a slack account, you can join here: 

https://join.slack.com/t/amaturerockettoorbit/shared_invite/enQtNDM0NjgxMTcxMjM2LWY5NmM5YWEwYTY2ZDU1M2ZhYWJiYzA5MmNhNDBiNTJmNTA4NDMwZjE4MWMzYzA3ZWFkY2JiNzNmNWQ1OGE2NjU

Anyway, the main propulsion system choices we have are:

1) Hybrid

- jellied petrol / htp

- htpb / nitrous oxide

- htpb / lox 

2) Liquid 

- methane / lox

- kerosene / lox

- alcohol/ lox

 

I think ether methane or kerosene for the core stage, and hybrids for boosters, though they too could be liquid

Maybe something like this:

Core stage: kerosene/lox

4 boosters: kerosene/lox

the core stage would be slightly larger than the boosters, think of the r-7 for example

Edited September 13, 2018 by BillKerman1234

[wumpus]

On 9/12/2018 at 3:03 PM, BillKerman1234 said:

Wumpus, I aggre with everything you said, except for the bit about it costing 6 figures. If we do it right, we could probably do it for under 100k, maybe 50k. The main cost driver in rocketry is the cost of hiring designers, the testing equipment, and most of all, the paperwork. Everything in a rocket (conventionally) has to be checked 500 times and must be traceable back to the ore in the ground.

The cost was taken directly from the poster Riven who was claimed his[?] group needed $60k to fling 1kg at 500m/s.  Checking their website it turned out that their final goal was $10k.  The $10k was raised but no reports on building the rocket (or successful launches and/or explosions) were listed on the site.  The biggest cost would be manufacturing, unless somebody has access to a CNC machine (or perhaps a lathe and gear that can approximate a CNC machine given enough time), especially if an iterated design process (design, build, explode, redesign).

On 8/26/2017 at 5:02 AM, Hesp said:

@wumpus Basically you need the know-how. To build that you need engineering time, prototypes, measuring lab and testing equipment, and failures. Lots of them. Any failure you investigate correctly gives you further knowledge of the topic.

In other words, learning rocket science is expensive (see design, build, explode, redesign) .  There are comments about "re-use".  Any "built on the cheap" orbital rocket won't have the slightest delta-v budget for reusability, but strong parachutes (possibly adding soyuz-style retro rockets) might make a great dummy load for testing the first stag e.  Still expect to lose a few boosters while getting them to work.

Things to consider include:

Staging: Serial or parallel?

Fuel: Liquid, hybrid, or solid? Guidance: Gimbal, differential thrust, or aerodynamic

Staging: Primary boosters should be parallel.  This removes a lot of "how do you light the second stage" (procedure is almost certainly ignite verniers [if any], ignite sides: if everything is burning, ignition and liftoff).  Expect to need plenty of serial staging after that simply for want of Isp.  A third stage might by pressure fed nitrous into some sort of hybrid propane [might need to be pressure fed as well] as this combo is said to be hypergolic.  Final solid stages would be ideal for sufficiently tiny satellites, presumably simply purchased ammonium chlorate in a carbon-fiber tube (keep costs low by making it small) [this should be available in the USA, no idea about exporting it].

Fuel: Liquid should be avoided as hybrids should reduce complexity by half (nitrous/propane might be an option for a stage lit in flight).  Hybrid is pretty much the booster of choice.  Solids make ideal rockets that can be arbitrarily small and lit in flight.  Turbopumps are likely out (no way to reduce the cost enough) even if they are electric.

Guidance: Expect legislated laws to drive the design more than physical laws, as rocket+guidance=missile.  Gimbal seems a nightmare.  Differential thrust might be done better by a set of verniers (like the Soyez).  Note that any "vernier" we would use is almost certainly a small hybrid rocket with a throttle: it would have to be able to run at more than 1/2 throttle the whole way up (where full throttle is wide open and minimum throttle is as little oxidizer as you can get away with and keep the rocket burning).  Aerodynamic guidance sounds good, but probably just left to "having fins" in the end (small rockets have even greater aerodynamic issues than the big boys, so such a rocket likely makes a pitchover well past where fins really work).

 

Edited September 13, 2018 by wumpus
get rid of annoying strikethrough bug

[sevenperforce]

A few major points:

• Veto on recovery of the upper stage or payload. Minimum possible payload.
• Veto on expendable side boosters. The point is to be able to do this on the cheap and so we need to do non-destructive testing. We can test-fire the boosters as sounding rockets and recover as chutes. Maybe do test suborbital spaceflights using a single solid motor on top of a recovered booster.
• Veto on any air-started liquid rocket. Ignition is a tough enough problem on the ground, and true hypergolics are far beyond what we could pull off. The only way we could conceivably manage an air-started hybrid motor is that decomposed HTP is hypergolic with hydrocarbons.

Along with many other advantages, parallel staging allows us to use differential throttling in place of gimbal for in-atmosphere and exoatmospheric attitude control. It is the only way to do attitude control without drag or cosine losses, and it is essentially "free" because we need fine throttleability of our boosters already. Boosters can either have fixed fins (to help hold prograde) or controllable fins for roll control.

Additional hybrid advantage is venting steam from the pressure chamber as automatic RCS for attitude control of the core.

If we can get away with fixed fins, then ideally the only moving parts on the rocket are little valves actuated by a small controller. It will save money if every single valve is identical.

It may be cheaper to synthesize HTP than it would be to try and source it.

[sevenperforce]

Ablative paint on the inside of the nozzle (can probably be sprayed on) makes the nozzle reusable.

No need for regenerative cooling.

Parachute recovery of a true bipropellant stage is a complete non-starter. A hybrid motor is, by design, tough enough to land via chute.

[BillKerman1234]

Plus, hybrids are difficult to refuel

Although there’s no reason why a biprop stage can’t be landed via chute

[Steel]

No, but I think a hybrid is a much easier motor to work with. Gives you the benefits of throttling without the inherent difficulties of a liquid bi-prop engine (fuel mixing, two sets of plumbing e.t.c)