Amateur rocket to orbit

[wumpus]

On 8/6/2017 at 0:10 AM, DerekL1963 said:

No - we do not know it works.  It's a completely new engine requiring a fuel with...  let's just say 'unusual' requirements.  Claiming that the difficulty is on par with a conventional engine is ludicrous at best.

I was refering to a bog standard hybrid (N₂O+rubber).  But even then there would be the issues of removing the boosters from the sustainer.  The big problems is even if it "works", it might be impossible to scale to the multi-km/s requirements (presumably at least 1km/s of velocity + 1 or more km/s of aero and drag losses.  And that is pretty bad for an first stage (especially in a booster/sustainer configuration).

[sevenperforce]

On 8/3/2017 at 6:14 PM, DerekL1963 said:

Quote

Not sure what geometry you're imagining, but neither gravity nor acceleration require time or distance to "accelerate" the fuel. The downward force on the fuel column will be several gees from launch to burnout.

I'm not imagining anything, I'm using real world physics.  For your fuel to flow, it has to accelerate from zero speed - and that takes time and distance. And the force is only one gee at ignition, which just adds further difficulty - the force flowing your fuel varies radically from ignition to burnout.

Sorry, wasn't trying to use "imagining" pejoratively; I probably should have said "visualizing". Anyway, I'm still not quite sure what you're suggesting. Flow in any fluid is a function of the instantaneous force vector, not time or distance. The rocket comes from zero acceleration (one gee) to full acceleration (4+ gees) instantly at ignition. Of course it takes time and distance for that acceleration to be translated into airspeed, but that's inconsequential to the fuel flow; the fuel doesn't care how fast it is going, only how forcefully it is accelerating. 

Moreover, the use of parallel staging means that for (at least) the lower stages, the acceleration only varies slightly over the course of the burn. Obviously the lowest-acceleration state is at ignition, but ignition is the least of our worries since it is by definition not a steady-state situation.

On 8/3/2017 at 6:14 PM, DerekL1963 said:

o.0  A few messages back, it wasn't a gel...  and now it's a gel again.  Except when you're mistakenly relying on the properties of oobleck to make it once again, not a gel.  Seriously, your fuel magically changes properties to magically address whatever objection is being made. And that's setting aside that there's no mechanism creating droplets, let alone driving entrainment and mixing.

The viscosity of a non-newtonian fluid may vary in response to shear force, static pressure, and temperature. A steady-state combustion would involve a range of shear forces and temperatures across the fuel column. It is not difficult to conceive a fluid composition which would take advantage of the force and temperature gradient to achieve the desired flow, melt, and vaporization behavior. 

On 8/3/2017 at 6:14 PM, DerekL1963 said:

Quote

The only reason I'm still proposing this design is that IF it turns out to work in preliminary tests, and works well, it offers significant advantages.

Feel free to enumerate them.

High combustion stability, wide throttling range, extremely high thrust-to-weight ratio, relative ease of fuel acquisition, higher reusability, and lower tank/structure dry mass.

On 8/4/2017 at 3:28 AM, shynung said:

Our goal is a paper design of an amateur-accessible orbital rocket, right? I agree with the others that we should start with a proven design, to lighten the load on engine R&D. We can scale the design as needed later on.

If it turns out that the gel-hybrid approach doesn't work, we can always use a more heavily congealed cook of napalm and make it a traditional hybrid rocket.

On 8/4/2017 at 11:04 AM, wumpus said:

I'm still curious to see if there are any ways to salvage this design (if only for an upper stage).  One quick way to do it would be to find a sufficiently solid hydrocarbon (I'm guessing coal is too heavy, but perhaps a conglomerate of anthracite and napalm could be held back by a thin screen) and a traditional hybrid.  Solving this could well be easier than coming up with a new high Isp rocket + ignition.

More heavily gelled gasoline -- to the point that it is basically rubberized -- would work just fine.

On 8/4/2017 at 11:04 AM, wumpus said:

I'd recomend going straight to a traditional hybrid for stages 1.5 (first sustainer plus side boosters).  While N₂O might have lousy Isp, that isn't terribly important when you have the dry mass of two more stages on top of the stage.  Furthermore, the early stages have to be *big* and will make up most of the mass of the rocket (slightly bigger with poor Isp, but they would be big even with LOX+RP1).  This type of rocket is relatively easy for the amature to create, and presumably scale up and no materials really draw huge oversight flags (the scale involves reduces sourcing the HTP for stage 2 to a much less significant problem than doing the first stages as well). [PS. After looking at the real delta-v requirements, I'm less sure.  I just really doubt amature construction of "real rocket fuel" at the size these boosters would need for a few km/s of delta-v.]

The isp for N₂O+PTB is so low that it IS problematic. I looked at this before I looked at HTP; if you imagine strapping together dozens of the HEROS-3 rockets you still never get into orbit...you run into this problem.

On 8/4/2017 at 11:04 AM, wumpus said:

Have we even looked at a delta-v budget?  Last I looked, the project was to throw a satellite massing a handfull of grams into orbit (I suspect that recent 4g "picosat" supplier could supply/sponser one).  I've found in KSP that starting with equal budgets per stage for delta-v is a good starting point (other methods involve geometrically reducing mass.  If using N₂O in one stage and LOX in another, I'd happily use vastly more N₂O than LOX).  Knowing that you need at least 2km/s in the first stage or 5km/s in your second coudl cull plenty of design options.

Yeah, I looked at delta-v budget a few pages back; there's a whole spreadsheet and you can download it yourself if you like and vary the constants.

 

On 8/5/2017 at 1:15 PM, wumpus said:

AMMENDUM (and possible helpful breakthrough):

I was poking around for background on another thread when I came across a paper detailing a simple bipropellant mixture: both easily obtainable and non-cryogenic (hopefully ideal for pressure-fed engines), hypergolic with an Isp~300 (in vacuum).  Sorry, no spoilers.
https://tfaws.nasa.gov/TFAWS06/Proceedings/Aerothermal-Propulsion/Papers/TFAWS06-1026_Paper_Herdy.pdf

That's really cool (literally and figuratively)! Would be much easier to get a working engine prototype than the HTP design. One problem I do foresee is the whole pressurizing-the-fuel-with-the-oxidizer-via-bladder issue. Dunno how I feel about that.

But as far as getting your hands on the propellants is concerned, it is definitely hard to beat propane and nitrous.

On 8/6/2017 at 0:10 AM, DerekL1963 said:

No - we do not know it works.  It's a completely new engine requiring a fuel with...  let's just say 'unusual' requirements.  Claiming that the difficulty is on par with a conventional engine is ludicrous at best.

We do know it works, to some degree, since the first liquid engine ever flown by the Soviets used gelled petrol and a liquid oxidizer.

On 8/5/2017 at 7:21 PM, wumpus said:

I'm really wondering what it would take for the N₂O+Propane pressure-fed system.  It looks like a pretty straigtforward "build in a garage" setup, I'm wondering if there is a great big catch (exhasut products are lethal? exhaust temperature melts the nozzle? something else?).  It appears to have the same requirements as any non-crygenic liquid fuel, less the crazy catalyst requirements from HTP.  Of course you have two to liquids to deal with instead of only one as in a hybrid, but everything else looks pretty good.

Melting the nozzle would definitely be a potential issue; this is going to be VERY hot. I don't know whether regenerative cooling would be beyond the capacity of amateurs. Propane would be a great heat sink but you still have pressurization problems.

[shynung]

2 hours ago, sevenperforce said:

If it turns out that the gel-hybrid approach doesn't work, we can always use a more heavily congealed cook of napalm and make it a traditional hybrid rocket.

What I mean is, maybe we should start off using a traditional hybrid in the first place, to save R&D trouble off whoever ends up using our plans later on.

[Steel]

2 hours ago, sevenperforce said:

The viscosity of a non-newtonian fluid may vary in response to shear force, static pressure, and temperature. A steady-state combustion would involve a range of shear forces and temperatures across the fuel column. It is not difficult to conceive a fluid composition which would take advantage of the force and temperature gradient to achieve the desired flow, melt, and vaporization behavior. 

Be careful here, being able to conceive a fluid with these properties is very different to being able to engineer a fluid with those properties in the real-world. If John D. Clark has taught me anything in Ignition!, it's that additives that you think will help a fuel more often than not have the opposite effect. 

2 hours ago, sevenperforce said:

We do know it works, to some degree, since the first liquid engine ever flown by the Soviets used gelled petrol and a liquid oxidizer.

Let's not pretend this is a shining beacon of a demonstration that this technology works. The launch record of the GIRD-9 [1] - which is the hybrid version of the engine - was 4 failures and one partial success where the rocket flew 400m before the engine failed. After that they seem to have completely abandoned the concept. Not to mention that the geometry that was used is totally different from what's been proposed here, we can pretty much say that it's a completely untried concept. All we can really take from the Russian engine is that the propellants burn when put together and heated.

[1] http://www.astronautix.com/g/gird-09.html)

Edited August 8, 2017 by Steel

[DerekL1963]

 

2 hours ago, sevenperforce said:

High combustion stability, wide throttling range, extremely high thrust-to-weight ratio, relative ease of fuel acquisition, higher reusability, and lower tank/structure dry mass.

Relative ease of fuel acquisition?  There's no supplier (at least not in the US) that will provide HTP in small quantities to non professional organizations (read: folks who aren't schools or governmental bodies.)  This means brewing your own.  
Compare with LOX, which can be had with a phone call.  Higher reusability?  Not as much as you think - cat packs are difficult to re-use because of problems with catalyst poisoning and stripping.  (One downside of concentrating down your own HTP from commercial grade?  It concentrates the stabilizers too - and they're almost always catalyst poisons.  This further compounds your acquisition problems.)   And that's setting aside the problems of designing a cat pack in the first place - even one without the novel configuration required by this engine.  Not to mention that, AIUI, cat packs are fairly expensive.

And seriously, unless you're launching a dozen of these on a regular basis, reusability shouldn't even be something that figures into your planning.  Reusability is cool, and hip, and hot...  But it also imposes costs, and those costs have the be balanced by payback.

I've said it before, but it bears repeating  - lowering costs isn't just an exercise in engineering, it's an exercise in beancounting too.
 

3 hours ago, sevenperforce said:

I don't know whether regenerative cooling would be beyond the capacity of amateurs.

From what I read it's not, people build regeneratively cooled engines regularly.  (And it's getting even easier as 3D printing improves and you're no longer limited by what can be machined by conventional means.)
 

1 hour ago, shynung said:

What I mean is, maybe we should start off using a traditional hybrid in the first place, to save R&D trouble off whoever ends up using our plans later on.

Agreed.   If you're on an amatuer budget, it seems like a good idea to spend your money on development (that is, coloring within the lines of what is known) rather than research (going outside the lines into the unknown).  Development of a conventional motor is going to be difficult and costly enough, researching and developing a complex new motor just cubes that.  (And to make things worse, the new motor is an evolutionary dead end, so it doesn't even make a good training ground.)

[shynung]

Expanding @DerekL1963 's response on catalysts, I think we should consider using a starting slug, something like a solution of potassium iodide in water, to decompose only the initial charge of the HTP. The rest of the undecomposed HTP can sustain combustion by itself.

[monophonic]

21 hours ago, sevenperforce said:

That's really cool (literally and figuratively)! Would be much easier to get a working engine prototype than the HTP design. One problem I do foresee is the whole pressurizing-the-fuel-with-the-oxidizer-via-bladder issue. Dunno how I feel about that.

Ethane has a vapor pressure around the same magnitude as N2O, if I read the internets properly. It is also the precursor to ethylene, which is widely used in plastics production, so it is commercially produced in large quantities. It's not as easy to come by as propane, but you might ask the same suppliers for relevant quantities of propane anyway...

[wumpus]

On 8/8/2017 at 8:33 AM, sevenperforce said:

That's really cool (literally and figuratively)! Would be much easier to get a working engine prototype than the HTP design. One problem I do foresee is the whole pressurizing-the-fuel-with-the-oxidizer-via-bladder issue. Dunno how I feel about that.

I'm not sure that is any worse than the HTP design.  Anything hypergolic is going to have serious issues keeping fuel+oxidizer apart (although it might be something to remember to ignite a rubber hybrid.  Just a small cartridge of propane should get it started.

On 8/8/2017 at 0:06 PM, DerekL1963 said:

From what I read it's not, people build regeneratively cooled engines regularly.  (And it's getting even easier as 3D printing improves and you're no longer limited by what can be machined by conventional means.)

Considering that most 3d printers work by melting the feedstock, I really don't think it is an option for 3d printing.  Yes, I'm furiously ignoring that welding is typically involved, but I still greatly suspect 3d printing for regenerative cooling (unless it is "only" used to connect loops around the nozzle to the rest of the plumbing).

I strongly suspect that Copenhagen Suborbitals uses alcohol to avoid needing this type of thing, and as far as I know they use turbopumps.  With LOX + RP1 you can get away with only one turbopump.  I suspect that isn't the case with LOX + alcohol.

Edited August 13, 2017 by wumpus
notes about regenerative cooling

[DerekL1963]

On 8/9/2017 at 7:39 AM, wumpus said:

Considering that most 3d printers work by melting the feedstock, I really don't think it is an option for 3d printing.  Yes, I'm furiously ignoring that welding is typically involved, but I still greatly suspect 3d printing for regenerative cooling (unless it is "only" used to connect loops around the nozzle to the rest of the plumbing).

You're furiously ignoring a lot of things - like the existence of major components and even flown engines using significant amounts of 3D printing. SpaceX is also flying 3d printed hardware. Lawrence Livermore has produced a prototype engine, regeneratively cooled, with no welding.

 

On 8/9/2017 at 7:39 AM, wumpus said:

I strongly suspect that Copenhagen Suborbitals uses alcohol to avoid needing this type of thing, and as far as I know they use turbopumps.  With LOX + RP1 you can get away with only one turbopump.  I suspect that isn't the case with LOX + alcohol.

Why wouldn't it be possible?

[wumpus]

9 hours ago, DerekL1963 said:

You're furiously ignoring a lot of things - like the existence of major components and even flown engines using significant amounts of 3D printing. SpaceX is also flying 3d printed hardware. Lawrence Livermore has produced a prototype engine, regeneratively cooled, with no welding.

Why wouldn't it be possible?

None of those examples are amatures.  If you have access to that type of equipment, why don't you have access to a cheaper and much more common CNC milling machine to churn out turbopumps?

Everything done professionally is possible, but it might have a NASA level price tag.  I suspect that a lot depends on if you can merely increase the temperature of laser sintering and thus use appropriate materials at a much lower speed.  This would still be pricy, but hopefully the cost would be somewhat linear.  I would still be expecting to switch to turbopumps with this type of access.

[DerekL1963]

1 hour ago, wumpus said:

If you have access to that type of equipment, why don't you have access to a cheaper and much more common CNC milling machine to churn out turbopumps?

The sound you heard was my point whooshing over your head - 3D printing is getting cheaper and more capable almost by the day.  Your presumption that since "melting is involved" 3D printing is suspect is off base.  (If for no other reason than melting underlies all metal forming processes - even conventionally machined engine parts.)

[wumpus]

2 hours ago, DerekL1963 said:

The sound you heard was my point whooshing over your head - 3D printing is getting cheaper and more capable almost by the day.  Your presumption that since "melting is involved" 3D printing is suspect is off base.  (If for no other reason than melting underlies all metal forming processes - even conventionally machined engine parts.)

Then by all means just print an F1 engine.  It (both the original and printed edition) are proven tech.  It certainly is more likely to get into orbit than any other suggestion so far.

You might also want to look up some brand new technology called "forging".  It makes metals stronger than if you simply melt bits of pieces together (although not quite needed for said printed F1 engine, it will still make for stronger metals of the same material/mass).

[Vanamonde]

No need to get tetchy with each other, please. 

[wumpus]

While "just print a F1" might seem unreasonable, it likely is a valid strategy for trying to get into orbit.  Back when Moore's law was still enforced by physics and economics, there was alleged to be a "slacking law" that said any program that required more than 2 years to run could be accelerated by simply doing nothing for n years and then running the code on a computer with the advancements made during that time.  I could easily imagine one group simply waiting for availability of the class of printer NASA* used to build their engines could beat a group using our various pressure fed liquid and hybrid rockets to orbit.

The problem is that once you posit such machinery, there is no longer much point to this thread.  I'm also of skeptical the rate of availability of such printers being available to the public, although there have been print shops willing to do work far beyond consumer printers for at least a decade.  I'm curious what is available now (don't forget to check your local machine shop for more traditional competition, at least where topologically possible).

The importance of forging and the cost and availability of subtractive (CNC milling) vs. additive ("3d printing) shouldn't be ignored.  While 3d printers get more press and appear magical at first glance, CNC milling devices are the actual workhorses of prototype and low volume (and sometimes high volume if such precision is needed) production.  I'd expect them to always be able to make a superior product for anything topologically possible to produce (even if this requires blasting bits off with a laser).  One catch is that plumbing is obviously topologically unsuitable for milling (and this not only allowed improvements in the F1 design, it also made things possible without needing to teach welders 1960s welding skills).  I'd still expect that things like turbopumps (and presumably housings) be done by milling.

* I'm not sure they contain actual engine blueprints, but the University of Maryland received a "complete copy of all NASA records from [~Apolloish era- ~shuttleish era] around 1990.  They are probably in the technical library unless the US archives (on campus) snagged it when it was built.  Since 9/11 NASA campuses (at least the one I've been on) are mostly closed to visitors and it might be easier to get to UM (a subway ride from Washington DC).

[Hesp]

22 hours ago, wumpus said:

The problem is that once you posit such machinery, there is no longer much point to this thread.  I'm also of skeptical the rate of availability of such printers being available to the public, although there have been print shops willing to do work far beyond consumer printers for at least a decade.  I'm curious what is available now (don't forget to check your local machine shop for more traditional competition, at least where topologically possible)

The technology you are looking for is called DMLS (Direct Metal Laser Sintering), a branch of SLS (Selective Laser Sintering).

While the cost for one of these machines ranges from 50k to 500k€, there is plenty of existing companies able to "3d print" your parts.

These printed parts usually need to be finished on a CNC center to achieve the required tolerances, anyway.

[wumpus]

2 minutes ago, Hesp said:

These printed parts usually need to be finished on a CNC center to achieve the required tolerances, anyway.

Unless it is cheaper than starting with the CNC, that only works for the available surfaces (there will likely be exceptions, but since the main point of printing an F1 engine was to get the plumbing without all the difficult and critical welds, you don't get to CNC the insides).  Hopefully your 3d printed plumbing doesn't have significantly more resistance than NASA's printer.

[Hesp]

Just now, wumpus said:

Unless it is cheaper than starting with the CNC, that only works for the available surfaces (there will likely be exceptions, but since the main point of printing an F1 engine was to get the plumbing without all the difficult and critical welds, you don't get to CNC the insides).  Hopefully your 3d printed plumbing doesn't have significantly more resistance than NASA's printer.

I wasn't referring to the F1 engine, specifically. To 3d print it you'll need the technical drawing and the 3d model of every single component, good luck getting those from NASA. And yes, internal surface roughness will be different from original since they are coming from a different process.

Plumbing is not a big issue today with hydroforming and a certified welder doing the job.

Metal 3D printing is increasingly common in prototype manufacturing compared to other processes like casting and forging, it's not an alternative to CNC milling. No need to laser sinter a part that can be obtained only by machining.

[wumpus]

16 minutes ago, Hesp said:

I wasn't referring to the F1 engine, specifically. To 3d print it you'll need the technical drawing and the 3d model of every single component, good luck getting those from NASA. And yes, internal surface roughness will be different from original since they are coming from a different process.

I doubt NASA would bother handing out the plans (although you can try asking the team that printed it, it is probably technically publicly available [but you might have to license it from an agency tasked with maximizing profit]).  You might find them in that UofM dump (there was at least one other place that has a copy now, but I have no idea where it is), but I suspect it would be the proverbial needle in the haystack.

If these methods are more precise than what you could do with 1940s slave labor, I suspect that pressure fed rockets to orbit (or even the karman line) are obsolete.  Is there enough knowledge in this forum to even bother with a new thread on amature turbopump-driven rockets?  Plenty of us (well I did, but a few others pointed out how it would cause waves through the pipes that turbopumps wouldn't like) thought "how hard could it be" with respect to asparagus staging, but it didn't take that long for Spacex to abandon that idea.

[Hesp]

56 minutes ago, wumpus said:

Is there enough knowledge in this forum to even bother with a new thread on amature turbopump-driven rockets?

I have some work experience in simulations and design of turbochargers... Making a functional turbopump is something way beyond the amateur/hobbist territory.

[DerekL1963]

2 hours ago, Hesp said:

Plumbing is not a big issue today with hydroforming and a certified welder doing the job.

3D printing and CNC are used not to eliminate welding, but to reduce the amount of welding.  The F1 engine required an enormous amount of touch labor to assemble components from individual weldments that we'd machine as single piece today.  (No CNC or six-axis machines back then.)
 

2 hours ago, Hesp said:

Metal 3D printing is increasingly common in prototype manufacturing compared to other processes like casting and forging, it's not an alternative to CNC milling. No need to laser sinter a part that can be obtained only by machining.

No, 3D printing is not an alternative to machine for things that can be machined.  The point of 3D printing is create things that cannot be created with more conventional methods or which it would be prohibitively expensive to create with more conventional methods.  Such as this prototype engine manufactured at Lawrence Livermore.

 

[Hesp]

58 minutes ago, DerekL1963 said:

No, 3D printing is not an alternative to machine for things that can be machined.  The point of 3D printing is create things that cannot be created with more conventional methods or which it would be prohibitively expensive to create with more conventional methods. 

We said the same thing, read again.

In my current job 3d printing it is an alternative to metal CASTING or FORGING for single or very small batches of prototypes.

It makes you save the cost and the design time of all the casting/forging toolings and fixtures, really prohibitive if you need only a small lot, quickly.

[DerekL1963]

40 minutes ago, Hesp said:

We said the same thing, read again.

In my current job 3d printing it is an alternative to metal CASTING or FORGING for single or very small batches of prototypes.

 

1 hour ago, DerekL1963 said:

No, 3D printing is not an alternative to machine for things that can be machined.  The point of 3D printing is create things that cannot be created with more conventional methods or which it would be prohibitively expensive to create with more conventional methods.

No, we aren't really taking about the same thing.  You're discussing 3D printing as an alternative for existing conventional processes.  I very specifically stated that it's not just an alternative.

Though I linked to a prototype as an example of something that can't be done with existing processes, I have specifically otherwise avoided using the word prototype because 3D printing is rapidly moving out of producing prototypes in the lab and into the world of manufacturing.

[Hesp]

2 hours ago, DerekL1963 said:

No, we aren't really taking about the same thing.  You're discussing 3D printing as an alternative for existing conventional processes.  I very specifically stated that it's not just an alternative.

Though I linked to a prototype as an example of something that can't be done with existing processes, I have specifically otherwise avoided using the word prototype because 3D printing is rapidly moving out of producing prototypes in the lab and into the world of manufacturing.

Got it, what you say regarding the new capabilities of the process is true and i completely agree. If SLS is the only way you can manufacture a part, it qualifies automatically for production rather than just prototypes.

My remark was that we both stated the same thing, when saying that it's not an alternative to machining for things that can be machined.

 

[wumpus]

On 8/18/2017 at 11:45 AM, Hesp said:

I have some work experience in simulations and design of turbochargers... Making a functional turbopump is something way beyond the amateur/hobbist territory.

Is this an issue of "you need the right background", or a matter of "even with a team of experienced engineers it still takes extended amounts of time and lots of prototypes and expensive lab equipment?" 

A lot of the controls issues (even for things like trying to build something like Rocket Lab's electric turbopump motor) isn't out of reach of many electrical engineers.  Building any circuit board tends to add a million dollars to the cost of the project (although typically such boards are larger than anyone wants to use in amature rocketry), but that has more to do with engineer-hours and costs (note that it should be pretty easy to get an engineer to build a one-off.  About 90%+ of the time is spent for reasons that are really only necessary for corporations and can be ignored for this type of project).

Any project like this is going to need some pretty specific backgrounds.  Of course, if you were starting from scratch (instead of starting with a KSP forum plan) you would presumably plan your rocket around your team's skills.  So the idea is to require only reasonably common skills (or at least common among those willing to spend the time needed to get a rocket into space.  That likely includes more people with turbine knowledge).

[Steel]

30 minutes ago, wumpus said:

Is this an issue of "you need the right background", or a matter of "even with a team of experienced engineers it still takes extended amounts of time and lots of prototypes and expensive lab equipment?" 

The latter. There's so much to go wrong with things spinning at tens of thousands of RPM, and that's before you add in heat and vibrations to the mix. There's also so much to design in intricate detail if you want to make one from scratch. Just the topic of turbine blade geometry is something that people spend their entire careers researching.

Edited August 22, 2017 by Steel