Amateur rocket to orbit

[qzgy]

32 minutes ago, sevenperforce said:

If you use a bolted-together four-core sustainer cluster rather than a single-stick core sustainer, you get 8.65 km/s cumulative dV and slightly lower gravity drag losses.

What about a 2 or 3 core sustainer cluster?

[sevenperforce]

6 minutes ago, qzgy said:

What about a 2 or 3 core sustainer cluster?

Somewhere between 8.32 km/s and 8.65 km/s, though symmetry becomes more complicated at that point.

I'm unsure whether even 8.65 km/s is enough to get into orbit once gravity drag, pressure drag, and aerodynamic drag are pulled out. For a small launcher, aerodrag is high. If we could push to around 9.2 km/s, I'd breathe easier.

Increasing the diameter of the rocket grows propellant mass quadratically while growing stage mass linearly, so I can look into exactly how wide it would need to be in order to achieve 9.2 km/s, either with a single-stick or a four-core. And all this is dependent on the assumption that I can use the propellant fraction of the HEROS-3 rocket as a viable starting point. Finally, I really wish I had a better grasp of what isp I can reasonably expect. Peroxide + kerosene ranges from 230 seconds to 319 seconds, depending on pressure and expansion ratio, which is really broad. For all I know, the aluminum salts used to gel the gasoline could push specific impulse considerably higher. 

Edited July 11, 2017 by sevenperforce

[TheEpicSquared]

@sevenperforce

Maybe just two cameras on the center core would suffice. Using VHF, we could set the cameras to take pictures every X seconds, and relay them down to earth. This way we could visually confirm strap-on separation, and chute deployment, and finally landing.

About the center core, it seems better, in terms of Dv and TWR, to have a quad-stick core design, with single-stick strap-ons. The increased TWR would mean we could use a single-stick for the second stage as well (minus recovery hardware), instead of having to miniaturize the hardware for a second stage. Just make some modifications to the ablative nozzle and you'd be good to go. Control could again be maintained via the use of HTP RCS thrusters near the nozzle, and those same thrusters could be used for spin stabilization just before kick motor separation. I doubt that the longer second stage would push the Dv to 9.2km/s, but it would provide some excess Dv, which is always useful. We could make the tanks wider to hold more propellant, however this causes its own issues as discussed earlier. 

One idea for raising Dv, which is... very kerbal, to say the least, is to use two quad-stick cores as strap-ons.

Actually, this equals to 8 single-sticks, so instead of having two quad-stick strap-ons (which would weigh a lot due to the structural attachments, and provide less control), we could have 8 single-stick strap-ons. Configuration is below:

     A  B

C  D  E  F

G  H  I  J

     K L

So DEHI is the quad-stick core, and A, B, C, G, F, J, K and L are the 8 single-stick strap-ons. AB would be next to each other, as would CG, FJ, and KL. The throttling and staging would be the same.

This is getting on the complicated side now however.

 

EDIT: Picture for clarification (sorry that it's terrible, I drew it on my phone)

Edited July 12, 2017 by TheEpicSquared
Added picture

[sevenperforce]

4 hours ago, TheEpicSquared said:

Maybe just two cameras on the center core would suffice. Using VHF, we could set the cameras to take pictures every X seconds, and relay them down to earth. This way we could visually confirm strap-on separation, and chute deployment, and finally landing.

It should be easy enough to have a videocamera that stores high-quality video locally while simultaneously capturing still frames every few seconds and sending them back via VHF. This gives us nearly-live visuals but also preserves a video feed for later recovery.

4 hours ago, TheEpicSquared said:

About the center core, it seems better, in terms of Dv and TWR, to have a quad-stick core design, with single-stick strap-ons. The increased TWR would mean we could use a single-stick for the second stage as well (minus recovery hardware), instead of having to miniaturize the hardware for a second stage. Just make some modifications to the ablative nozzle and you'd be good to go. Control could again be maintained via the use of HTP RCS thrusters near the nozzle, and those same thrusters could be used for spin stabilization just before kick motor separation. I doubt that the longer second stage would push the Dv to 9.2km/s, but it would provide some excess Dv, which is always useful.

To clarify my earlier proposition: I was already working under the assumption that the second stage was a full-size single-stick core, so there's no extra dV to be gained there.

4 hours ago, TheEpicSquared said:

We could make the tanks wider to hold more propellant, however this causes its own issues as discussed earlier. 

To a point, there are really no significant disadvantages in making the stages slightly wider; it doesn't make them significantly harder to handle, and the increase in drag is more than counterbalanced by the linear increase in propellant fraction.

The maths, in case anyone is curious:

Spoiler

Consider a hollow cylinder of length L and radius R. The weight of the metal skin making up the cylinder is proportional to the surface area of the cylinder, which is the circumference times the length, or 2*π*R*L. The propellant capacity of the cylinder is proportional to its volume, given by 2*π*R²*L.

If you increase the length of the cylinder (e.g., from L to 2L), then both the surface area and the propellant capacity increase together (surface area becomes 4*π*R*L and volume becomes 4*π*R²*L), with no change in propellant fraction. 

However, if you increase the radius of the cylinder (e.g., from R to 2R), then the square term kicks in. The surface area doubles (4*π*R*L), but the volume quadruples: 2*π*(2R)²*L = 2*π*4*R²*L = 8*π*R²*L. So the ratio of propellant mass to container mass increases linearly with the increase of cylinder radius.

There's a slight increase in skin thickness in order to maintain structural integrity, but this is such a small value that it's negligible for relatively small increases in cylinder radius. And the increase in drag is also counterbalanced by the increase in overall vehicle weight. Another reason hobby rockets need high fineness ratios is that their low isp means their mass drops rapidly, but with our higher isp we hang on to our propellant for longer and so drag has less of an impact.

I think it will be much simpler to make a slightly wider stage than it would be to double or triple the number of first-stage cores.

4 hours ago, TheEpicSquared said:

One idea for raising Dv, which is... very kerbal, to say the least, is to use two quad-stick cores as strap-ons.

Actually, this equals to 8 single-sticks, so instead of having two quad-stick strap-ons (which would weigh a lot due to the structural attachments, and provide less control), we could have 8 single-stick strap-ons. Configuration is below:

     A  B

C  D  E  F

G  H  I  J

     K L

So DEHI is the quad-stick core, and A, B, C, G, F, J, K and L are the 8 single-stick strap-ons. AB would be next to each other, as would CG, FJ, and KL. The throttling and staging would be the same.

This is getting on the complicated side now however.

Yes, this is definitely getting complex.

I'd like to be able to model it with aerodrag, gravity drag, and pressure drag losses included, but the only thing that can be directly calculated is gravity drag, and even there it only holds for the initial part of the ascent (the kick stage should have no gravity drag, as it will already be in free-fall, but I'm not sure about the hybrid upper stage). I suppose I could estimate aerodrag from this analysis of the Lambda 4S, but 66 m/s seems ridiculously low, and it only calculates gravity drag for the 1.5 stages, out of 4.5 stages total. Then again, with its very high thrust maybe it really does clear the atmosphere that fast.

If anyone can find or come up with an isp curve for a pressure-fed HTP+petrol+gel engine, I'd deeply appreciate it. 

[sevenperforce]

Oh, while we're at it...any ideas on a name?

A few possibilities:

• Legion (since it uses clusters and parallel staging)
• Jebediah (in honor of KSP)
• EELOO (in honor of KSP, could backronym to something)
• Kilgore (character in Apocalypse Now who delivers the famous line, "I love the smell of napalm in the morning")

[TheEpicSquared]

1 hour ago, sevenperforce said:

It should be easy enough to have a videocamera that stores high-quality video locally while simultaneously capturing still frames every few seconds and sending them back via VHF. This gives us nearly-live visuals but also preserves a video feed for later recovery.

Yeah, a video of the first stage core would be nice, and a camera that records and takes pictures simultaneously shouldn't be too difficult to get.

1 hour ago, sevenperforce said:

To clarify my earlier proposition: I was already working under the assumption that the second stage was a full-size single-stick core, so there's no extra dV to be gained there.

Ah, ok.

1 hour ago, sevenperforce said:

To a point, there are really no significant disadvantages in making the stages slightly wider; it doesn't make them significantly harder to handle, and the increase in drag is more than counterbalanced by the linear increase in propellant fraction.

The maths, in case anyone is curious:

  Reveal hidden contents

Consider a hollow cylinder of length L and radius R. The weight of the metal skin making up the cylinder is proportional to the surface area of the cylinder, which is the circumference times the length, or 2*π*R*L. The propellant capacity of the cylinder is proportional to its volume, given by 2*π*R²*L.

If you increase the length of the cylinder (e.g., from L to 2L), then both the surface area and the propellant capacity increase together (surface area becomes 4*π*R*L and volume becomes 4*π*R²*L), with no change in propellant fraction. 

However, if you increase the radius of the cylinder (e.g., from R to 2R), then the square term kicks in. The surface area doubles (4*π*R*L), but the volume quadruples: 2*π*(2R)²*L = 2*π*4*R²*L = 8*π*R²*L. So the ratio of propellant mass to container mass increases linearly with the increase of cylinder radius.

There's a slight increase in skin thickness in order to maintain structural integrity, but this is such a small value that it's negligible for relatively small increases in cylinder radius. And the increase in drag is also counterbalanced by the increase in overall vehicle weight. Another reason hobby rockets need high fineness ratios is that their low isp means their mass drops rapidly, but with our higher isp we hang on to our propellant for longer and so drag has less of an impact.

I think it will be much simpler to make a slightly wider stage than it would be to double or triple the number of first-stage cores.

Yeah, is also agree that making the stages wider is easier. That would increase the Dv by quite a bit, especially if we retain the quad-stick core with four single-stick strap-ons. I never really thought my 8 strap-on idea would be feasible anyways.  

So, what type of diameter are we looking at? I know we're using the HEROS-3 as a base, does anyone know what the diameter is? From a picture, the diameter looks roughly 20cm, so ours could be slightly wider - 35 or 40cm? 

5 minutes ago, sevenperforce said:

Oh, while we're at it...any ideas on a name?

A few possibilities:

• Legion (since it uses clusters and parallel staging)
• Jebediah (in honor of KSP)
• EELOO (in honor of KSP, could backronym to something)
• Kilgore (character in Apocalypse Now who delivers the famous line, "I love the smell of napalm in the morning")

I was thinking Chimera, because in mythology a Chimera was a mix of other animals, like our rocket uses a mix of different propellant types. 

[sevenperforce]

Just now, TheEpicSquared said:

Yeah, is also agree that making the stages wider is easier. That would increase the Dv by quite a bit, especially if we retain the quad-stick core with four single-stick strap-ons. I never really thought my 8 strap-on idea would be feasible anyways.  

So, what type of diameter are we looking at? I know we're using the HEROS-3 as a base, does anyone know what the diameter is? From a picture, the diameter looks roughly 20cm, so ours could be slightly wider - 35 or 40cm? 

HEROS-3 is 7.5 meters long and 223 mm in diameter. 

I'm going to go ahead and set up a cleaner version of the same spreadsheet I used yesterday, but factor in stepwise drag losses and allow for some deeper throttling, and tie it all to a single diameter multiplier. This will allow me to baseline total dV at a HEROS-3 equivalent size and then increase diameter gradually until we have enough dV to reach orbit. If I can't do it with a single-core I'll try again with a quad-core.

Just now, TheEpicSquared said:

I was thinking Chimera, because in mythology a Chimera was a mix of other animals, like our rocket uses a mix of different propellant types. 

Interesting. Hydra would also be a neat name, since the base will have "many heads". Or we could call it the ARDYH since it's an upside-down Hydra...maybe backroynm that.

[TheEpicSquared]

8 minutes ago, sevenperforce said:

HEROS-3 is 7.5 meters long and 223 mm in diameter. 

I'm going to go ahead and set up a cleaner version of the same spreadsheet I used yesterday, but factor in stepwise drag losses and allow for some deeper throttling, and tie it all to a single diameter multiplier. This will allow me to baseline total dV at a HEROS-3 equivalent size and then increase diameter gradually until we have enough dV to reach orbit. If I can't do it with a single-core I'll try again with a quad-core.

Could you share the spreadsheet here once you're finished? 

9 minutes ago, sevenperforce said:

Interesting. Hydra would also be a neat name, since the base will have "many heads". Or we could call it the ARDYH since it's an upside-down Hydra...maybe backroynm that.

HYDRA: HYbriD Rocket by Amateurs

Could work  

[TheEpicSquared]

@sevenperforce How's that spreadsheet going?  

[sevenperforce]

According to page 2 of this study, vacuum isp for peroxide+kerosene ranges from 304 seconds at 2 bars chamber pressure to 313 seconds at 30 bars chamber pressure. If you look, you'll notice that the O/F fuel ratio changes based on chamber pressure as well. One advantage of a pressure-fed engine is that highest combustion chamber pressure is at launch, where pressure losses are highest; as it ascends and pressure losses decrease, tank pressure drops. 

OTRAG's pressure-fed design, which we are aping, calls for 600 psi ullage pressure or around 41 bars. So I think it is safe to aim for the high end of chamber pressures and O/F ratios. Higher chamber pressure helps tremendously with pressure losses at launch. I think I'll set launch isp at 250 seconds, just under the SL isp of the 47-bar Gamma 8 engines on the Black Arrow kerosene+peroxide LV, increasing stepwise to 310 seconds by core burnout. This requires an O/F mass ratio of 7.4:1, which I'll factor back into my equations.

The "simulated" gravity drag and air drag for the Lambda 4S don't really make sense at all, because we determined earlier that there were 1.2 km/s of losses in aerodynamic and gravity drag. Gravity drag isn't too hard to calculate; the time from launch to orbital insertion was 104.8 seconds, 28.2 seconds of which was the terminal stage and 19.4 seconds of which was the second-to-last stage. Operating under the assumption that gravity drag pretty much dies out over the course of the second-to-last-stage burn, I'll use 67 seconds of gravity drag for a total of 656.6 m/s of gravity drag losses, leaving 543 m/s of aerodynamic drag losses. That's a little more reasonable. It's a tossup whether the Lambda 4S would have higher or lower aerodynamic drag than our launch vehicle; higher acceleration means greater speed and lower isp means a lighter stage, both resulting in more severe initial drag losses, but our vehicle will have more parallel boosters and will spend more time in the atmosphere, which means more drag losses for us. I'll just set total aerodynamic drag at 550 m/s for conservatism, and split it among the first few acceleration phases. Notably, I estimated 600 m/s of aerodynamic drag on the HEROS-3, which is very close.

Recall the layout (only two boosters shown, but four planned total):

A screengrab from the video of the HEROS-3 launch shows that the recovery interstage (containing chutes) is 0.48 meters long and the nose cone is 0.97 meters long, meaning that the actual tankage length is 6.05 meters. This means the volume of the tankage portion of the vehicle is 236 L, but based on earlier calculations (using N₂O and paraffin density, stated propellant mass, and known O/F ratios for N₂O+paraffin hybrid rockets), total propellant volume is 128 L. This means volumetric efficiency is just 54%. We should be able to squeeze slightly better volumetric efficiency out of our design...perhaps around 65%.

Once I get all these numbers into the spreadsheet, I'll post it.

[Rocketscience101]

I've actually thought about this a ton. And I've devised a plan, that only requires funding.

the first stage and its two side boosters are solid, to simplify, and is the largest segment of the rocket. It allows me the easiest access to resources and the flight profile is relatively simple. The second stage is a hybrid. This is because at this stage we should be higher in the atmosphere where we can no longer rely on the air around the vehicle to provide oxygen for ignition, and would allow for thrust control, but it restartability for the engine, the final stage, which will be the payload will be a small liquid fuel satellite who's only job is to take photos, read temperatures and perform small experiments that neither take up a lot of weight, nor require the return of a specimen, i.e. The data can be returned digitally.

My first mission will be a proof of concept. Only the camera will be launched as to return images. My preferred launch sight is in the Pacific Ocean, either directly in, or close to the equator to maximize on the energy of the earths rotation. He first stage would take an almost straight up flight path, to simplify to launch profile, leaving the first major maneuvers like the gravity turn to the second stage while it is higher in the atmosphere. From there the second stage would fire and guide the satellite into a high long suborbital trajectory, allowing for it to either get into orbit, or take and transmit as many pictures as possible in the event of failure. Finally the finale stage will fire it's rather simple and as light as possible liquid fuel engine, to bring it into a final orbit. It would then take photos and transmit them back.

the satellite itself would be simple, powered by solar panels, with a camera on the front for its most simple model. The fuel would preferably be liquid hydrogen and oxygen, but if I am unable to acquire that, gasoline and oxygen will do. The second stage will use a hybrid fuel system, using a solid based fuel with a liquid oxidizer. This will allow me better control of the burn rate of the fuel, and therefor thrust of the second stage. I haven't picked a fuel system for the second stage. I know of one that uses a highly flammable hydrogen based oxidizer and what is similar to car tire rubber as a fuel. But I have yet to find a way to acquire such a thing. Finally the first stage will be solid fuel, for both eas of the flight profile and control, of the rocket. Not the burn. I aim to get as much thrust from the first stage as possible to achieve the highest of ceiling altitudes for the stage. It will have 2-3 solid fuel side boosters that will fire at the same time and cut out at 2 separate milestones for the flight, then the final stage will stop burning, and release the second stage. 

[qzgy]

33 minutes ago, Rocketscience101 said:

I've actually thought about this a ton. And I've devised a plan, that only requires funding.

the first stage and its two side boosters are solid, to simplify, and is the largest segment of the rocket. It allows me the easiest access to resources and the flight profile is relatively simple. The second stage is a hybrid. This is because at this stage we should be higher in the atmosphere where we can no longer rely on the air around the vehicle to provide oxygen for ignition, and would allow for thrust control, but it restartability for the engine, the final stage, which will be the payload will be a small liquid fuel satellite who's only job is to take photos, read temperatures and perform small experiments that neither take up a lot of weight, nor require the return of a specimen, i.e. The data can be returned digitally.

My first mission will be a proof of concept. Only the camera will be launched as to return images. My preferred launch sight is in the Pacific Ocean, either directly in, or close to the equator to maximize on the energy of the earths rotation. He first stage would take an almost straight up flight path, to simplify to launch profile, leaving the first major maneuvers like the gravity turn to the second stage while it is higher in the atmosphere. From there the second stage would fire and guide the satellite into a high long suborbital trajectory, allowing for it to either get into orbit, or take and transmit as many pictures as possible in the event of failure. Finally the finale stage will fire it's rather simple and as light as possible liquid fuel engine, to bring it into a final orbit. It would then take photos and transmit them back.

the satellite itself would be simple, powered by solar panels, with a camera on the front for its most simple model. The fuel would preferably be liquid hydrogen and oxygen, but if I am unable to acquire that, gasoline and oxygen will do. The second stage will use a hybrid fuel system, using a solid based fuel with a liquid oxidizer. This will allow me better control of the burn rate of the fuel, and therefor thrust of the second stage. I haven't picked a fuel system for the second stage. I know of one that uses a highly flammable hydrogen based oxidizer and what is similar to car tire rubber as a fuel. But I have yet to find a way to acquire such a thing. Finally the first stage will be solid fuel, for both eas of the flight profile and control, of the rocket. Not the burn. I aim to get as much thrust from the first stage as possible to achieve the highest of ceiling altitudes for the stage. It will have 2-3 solid fuel side boosters that will fire at the same time and cut out at 2 separate milestones for the flight, then the final stage will stop burning, and release the second stage. 

Not to annoy you, but you do already know that the thread has a plan right? Its right above you.

Solid fuel was ruled out since making really perfect castings to actually go orbital is probably out of the realm of amateur rocketry. Restartability is very hard to do reliably, so most of our stages are unlikely to be restarted, only throttled down. LOX is annoying, hard to get, hard to work with, and all that other fun stuff.

Questions

- Can you give numbers on any of these specifics?
- How are you controlling it? In our current design, we are using the HTP, the oxidizer, also for an RCS system.
- I would also say that pure LF engines are quite hard to make work right. Can you or have you come up with a successful and reliable design?

You, to me, have just said "I need money, it might work". I am very skeptical at the moment, not least because of "This is because at this stage we should be higher in the atmosphere where we can no longer rely on the air around the vehicle to provide oxygen for ignition, and would allow for thrust control, but it restartability for the engine," I hope you do realize that rocket engines, by definition, have oxidizer, so I'm not sure where you are getting the notion that the atmosphere provides oxygen in any rocket system, whether solid or liquid or even hybrid.

 

8 hours ago, sevenperforce said:

I'll just set total aerodynamic drag at 550 m/s for conservatism, and split it among the first few acceleration phases. Notably, I estimated 600 m/s of aerodynamic drag on the HEROS-3, which is very close.

I would, in the case of not totally knowing the numbers, put it a bit too high and aim for that. Better to have excess rather than to little. So maybe a better number to just put conservatively high is 650 m/s. That, I realize, is quite a decent amount, but if we have that amount and we lose less than expected, we still make it.

Edited July 13, 2017 by qzgy

[TheEpicSquared]

I agree with @qzgy - if we can still make orbit with higher than expected aerodynamic losses, we have some wiggle room and some room for error. 

Here's some recovery ideas. I'm thinking drogues and then main(s), obviously. For the strap-ons, we could pack the chutes into the nosecone, like the Shuttle SRBs. However, it's a bit trickier with the core, since there is no nosecone space to pack the chutes into. 

I say that we put the chutes near the rear of the core. This is because the core will be traveling quite fast on its parabolic trajectory, and having the chutes deploy from the front would require the stage to flip 180 degrees, which would put a lot of stress on both the stage and the chute itself. I'm pretty sure the stage would survive the stresses, but I'm not sure the chute (and the ropes that connect it to the stage) would. This wouldn't be as much of an issue with the strap-ons since they would be traveling much slower upon separation. And since pressure-fed hybrids are inherently much stronger than liquid or solid fueled rockets, we can just let them land on their sides.

Also, do we really need separation motors on the strap-ons? I have a spring-based idea, which would save a lot of mass and be simpler. We would use conventional explosive bolts as the primary separation mechanism, but we'd have a compressed spring at the top, attached to the strap-on and compressed against the core. When the explosive bolts release, there is nothing keeping the spring compressed, and it can un-compress, pushing away the strap-on to a safe distance without the use of a heavy separation motor. Then the strap-on can land on its own. The spring itself could be made out of any lightweight metal, like aluminum. Theoretically, even the springs in a car's suspension would work if they are not too heavy. 

 

[sevenperforce]

9 hours ago, Rocketscience101 said:

the first stage and its two side boosters are solid, to simplify, and is the largest segment of the rocket. It allows me the easiest access to resources and the flight profile is relatively simple. The second stage is a hybrid. This is because at this stage we should be higher in the atmosphere where we can no longer rely on the air around the vehicle to provide oxygen for ignition, and would allow for thrust control, but it restartability for the engine, the final stage, which will be the payload will be a small liquid fuel satellite who's only job is to take photos, read temperatures and perform small experiments that neither take up a lot of weight, nor require the return of a specimen, i.e. The data can be returned digitally.

I'm gonna stop you right there. Solid-fueled rockets do not "rely on the air around the vehicle to provide oxygen for ignition".

Otherwise, everything @qzgy said. Not to rain on your parade, but this is rocket science. Nothing wrong with giving your input, of course.  

8 hours ago, qzgy said:

I would, in the case of not totally knowing the numbers, put it a bit too high and aim for that. Better to have excess rather than to little. So maybe a better number to just put conservatively high is 650 m/s. That, I realize, is quite a decent amount, but if we have that amount and we lose less than expected, we still make it.

I'll set it at 650 m/s but make it one of the independent variables in the spreadsheet so we can tune it down a little later on if needed.

3 hours ago, TheEpicSquared said:

Here's some recovery ideas. I'm thinking drogues and then main(s), obviously. For the strap-ons, we could pack the chutes into the nosecone, like the Shuttle SRBs. However, it's a bit trickier with the core, since there is no nosecone space to pack the chutes into. 

If you watch the HEROS-3 launch, you can see that they packed the chutes into an interstage between the N₂O tank and the payload module. That should work well enough.

Do we necessarily need both drogues and mains? It would seem to increase the number of possible failure points. What about a reefed main that opens fully at lower altitudes?

3 hours ago, TheEpicSquared said:

Also, do we really need separation motors on the strap-ons? I have a spring-based idea, which would save a lot of mass and be simpler. We would use conventional explosive bolts as the primary separation mechanism, but we'd have a compressed spring at the top, attached to the strap-on and compressed against the core. When the explosive bolts release, there is nothing keeping the spring compressed, and it can un-compress, pushing away the strap-on to a safe distance without the use of a heavy separation motor. Then the strap-on can land on its own. The spring itself could be made out of any lightweight metal, like aluminum. Theoretically, even the springs in a car's suspension would work if they are not too heavy. 

I don't think we need dedicated sep motors; I think we can repurpose an RCS thruster to act as both pneumatic decoupler and sep motor. The thruster would be fitted inside an airtight cylinder attached to the core. At the separation signal, the pressure valve opens and pushes residual HTP through the catalyst bed and RCS thruster, filling the cylinder with high-pressure gas until the force overcomes the spring on the clamps that hold the stages together. After separation, the RCS thruster continues to fire, pushing the nose of the booster away from the core just as a typical separation motor would.

Not sure about attachments on the base, though. We cannot very well just strut them together.

[TheEpicSquared]

@sevenperforce I don't really follow your separation idea. Diagram maybe? And about the chutes, I'm all for less chutes. If a reefed chute would work, then we'll go with that. Less mass is always good. 

[qzgy]

Just now, TheEpicSquared said:

@sevenperforce I don't really follow your separation idea. Diagram maybe? And about the chutes, I'm all for less chutes. If a reefed chute would work, then we'll go with that. Less mass is always good. 

I think what he means is that the RCS thruster is placed to aim into a tube of some sort and used to push it away instead of a normal seperation motor.

 

What I don't understand is why not use a tiny-ish hobby motor (like C or D size) as a separation motor.

[TheEpicSquared]

Makes a bit more sense now, but pics would still be appreciated. 

Hold on just a minute. Were we planning to have RCS thrusters on the core and strap-ons in the first place? I don't recall so, we said that we'd use differential throttling for control. From what I understand, RCS would only be on the second stage. If that's the case, it seems more complicated to add a single RCS thruster to the strap-on, and configure it to work as per @sevenperforce's idea. Opposed to just attaching a compressed spring to the strap-on and letting that push it away, I think my proposal seems easier to accomplish, and more reliable. It's only one component, versus @sevenperforce's several, which all need to go exactly right for separation to work.

@qzgy I think the reason we're not using hobby motors as separation motors is that we want to keep things as simple as possible, with as little ignitions as possible. 

Also, @sevenperforce, how do you draw so symmetrically in MS Paint? I'm trying to make a diagram of what I'm thinking, but symmetry is hard.

[sevenperforce]

8 minutes ago, qzgy said:

11 minutes ago, TheEpicSquared said:

@sevenperforce I don't really follow your separation idea. Diagram maybe? And about the chutes, I'm all for less chutes. If a reefed chute would work, then we'll go with that. Less mass is always good. 

I think what he means is that the RCS thruster is placed to aim into a tube of some sort and used to push it away instead of a normal seperation motor.

What I don't understand is why not use a tiny-ish hobby motor (like C or D size) as a separation motor.

It's because we need an actual separation mechanism, and the RCS thruster can provide both mechanism and reaction. An explosive bolt disintegrates when it receives the electrical separation signal, but if we were using any kind of clamping mechanism, then you need a servo or solenoid or some other electromechanical system to actually push the clamp open via electrical current. With an RCS thruster in an airtight tube, the normal action of the RCS thruster can build up enough pressure to force open the clamps without needing any electrical signal, and then automatically push the booster away.

Spoiler

The trouble with a hobby motor is twofold: first, the plume would tend to damage the stage; second, ignition is not instant and so the whole function is potentially compromised.

5 minutes ago, TheEpicSquared said:

Hold on just a minute. Were we planning to have RCS thrusters on the core and strap-ons in the first place? I don't recall so, we said that we'd use differential throttling for control. From what I understand, RCS would only be on the second stage. If that's the case, it seems more complicated to add a single RCS thruster to the strap-on, and configure it to work as per @sevenperforce's idea. Opposed to just attaching a compressed spring to the strap-on and letting that push it away, I think my proposal seems easier to accomplish, and more reliable. It's only one component, versus @sevenperforce's several, which all need to go exactly right for separation to work.

All the cores will be more-or-less identical for the sake of commonality. Differential throttling is more efficient and more powerful than RCS, but each of the cores will be plumbed for RCS, so adding a single RCS thruster in the existing port shouldn't be difficult.

And we'll need RCS on a standard single core when we are testing, as well.

5 minutes ago, TheEpicSquared said:

Also, @sevenperforce, how do you draw so symmetrically in MS Paint? I'm trying to make a diagram of what I'm thinking, but symmetry is hard.

Lots of experience, haha! I make the selection box transparent and do a lot of copying and flipping and manipulating with the arrow keys, if that helps.

[zeta function]

For the separation of the first two boosters, an air-brake attached to the the side could be deployed, this would cause downward and lateral force to push it away from the first stage. I am thinking spring deployment, that way it is fast. Since It would be at supersonic speeds it would cause a lot of drag, but it would be hard to maintain the structural integrity of the air brake at those speeds.

[sevenperforce]

Okay, so here's the spreadsheet I've been working on.

The values in red are independent variables, the values we can play with at will. We can increase base thrust, or alter throttle for a given ascent phase, etc., and see where things come out. Of course, we still have to be careful. We can't go below the minimum throttle rating, and cranking up the TWR too high can cause problems, and so on.

The values in orange are fixed independent variables. These are the numbers that are determined by systems outside our control, but which may change as we acquire more information. For example, we may only be able to push 300 seconds of vacuum specific impulse instead of 313, or we may only need to factor in 575 m/s of aerodrag rather than 650 m/s.

The values in blue are the output of the equations. Obviously, the number at the bottom right in blue-bold is the total dV; that's the number we want to maximize. But we have to keep an eye on the rest of them, as well; we don't want TWR to go too low or too high at any point.

For aerodrag, I used the simplifying assumption that the percentage of total aerodrag for a given ascent phase will be proportional to the fraction of ascent time spent in the atmosphere. So I split total aerodrag (a fixed independent variable) up based on the duration of each of the first three ascent phases. Aerodrag after that point should be negligible. 

For conservatism, I have calculated the dV for each phase based on the starting isp, rather than the average isp over that phase. This should give us a significant buffer.

The nature of the spreadsheet is such that changing any one of the independent variables produces an entirely new optimization curve for the other variables, so it takes some work. Here's what it looks like when I increase the size of the booster and then tweak all the numbers to their optimal values:

So it definitely looks doable!

If anyone would like to play around with the spreadsheet, let me know and I'll try to upload it.

 

[zeta function]

10% throttling is pretty low (on the core stage)... How are you planning to do that?

[TheEpicSquared]

@sevenperforce Your idea seems quite complex. What if the air-tight chamber isn't air-tight? What if the spring fails? What if the clamps themselves fail? It looks like there are too many failure points, IMO. A few simple explosive bolts with a spring up top seems more reliable.

@zeta function That's an interesting idea. It could work, and the airbrake could just be a curved piece of metal. However, it would have to be light, which would compromise its structural strength.

@sevenperforce That's one awesome spreadsheet! I, for one, would love playing around with it. Maybe transfer it to google sheets and upload it that way? And yeah, 10% throttle seems pretty low... 

And I guess HYDRA is the name then? I don't mind, I like the backronym - HYbriD Rocket by Amateurs. And I like "Micropayload Orbital Vehicle":

HYDRA MpOV

 

[qzgy]

I wonder if its possible to make a variation on the airbrake, and instead use a lifting surface. Should have the same effect, but 1) would increase drag and 2) possibly cause it to become unstable.

HYDRA as a name works. I'm okay with it, but was thinking of proposing JEB (Joint Experimental Burn). Doesn't really make sense, but eh.

[TheEpicSquared]

I think we need to figure out the booster sep mechanism. As you all know, I'm in support of conventional explosive bolts, with a compressed spring up top to push it and the booster away. It's simple, it's been tried and tested, and it should be reliable. 

Also, I shall reaffirm my liking with the name HYDRA, because of its mythological relevance as well as its relevant backronym.  

[kerbinorbiter]

3 minutes ago, TheEpicSquared said:

HYDRA

and because of its involvement of Project odyssey