Reusable LOX/Kerosene SSTO with drop tanks

GW Johnson · 2016-09-20 09:08:27

Well, sometimes it's a little hard to distinguish what is really happening in a concept from the names, and from the pre-test hype. Estimates used to sell the program usually turn out to be quite high compared to what is found in test.

I know "air augmented rocket" and "ducted rocket" as synonyms for something more properly called a "gas generator-fed ramjet", which is distinct from "solid-fueled ramjet".

But I've also heard those same two terms applied to a shroud ring about a rocket engine, which is not the same thing at all. At low (subsonic) flight speeds, adding a ring about a rocket amounts to an ejector air pump, increasing the massflow in the combined plume and reducing its velocity. You do not need to afterburn in the ring, that just reduces Isp, and captured airflow.

But, done right, this adds greatly to plain rocket Isp, because you get increased propulsive efficiency if your jet speed is closer to your flight speed.

However, the ring is heavy to install, which is why most vehicles have never used this device. Not worth the weight penalty, not even for missile cruise flight, much less a climbing accelerator.

At higher (supersonic) flight speeds, you need to burn fuel with the ring airflow to (theoretically) better match jet speed to flight speed. You still get the jet pump effect, but reduced, because the influences inside the ring cannot propagate upstream to influence the size of the captured airstream. That and the fuel flow reduce Isp far below the subsonic flight case. It's still an improvement, just not as much.

And it's still heavy. Which is why no one ever did this.

That's about all that I know. Joe Bendot knew a lot more. He reversed this concept, and came at it as a full-blown ramjet engine, just one in which the fuel supply was largely a very high-speed fuel-rich rocket exhaust, packaged way down inside the ramjet. That's the ejector ramjet. Theoretically, it could develop static thrust. In practice, there were very serious flameholding issues. This thing required specialty hypergolic fuels to work.

What I and others found in the solid gas generator-fed ramjets with ordinary flameholding-type fuels was that the fuel injection Mach number should never exceed 1. We saw a slight increase in the total pressure ratio across the combustor due to the incoming fuel momentum, and still maintain usable operating limits from a flame stability standpoint. Raise the injection Mach, and the operability limits narrowed very quickly, if it was ignitable at all.

You have to understand that a gas generator-fed ramet's fuel is quite unlike kerosene. It comes from a fuel-rich propellant, and is a mixture of carbon soot and carbon monoxide to burn, polluted with gas generator combustion products. This two-phase fuel supply imposes some very unique flameholding geometry requirements. Most of the common concept geometries will not work. If I told you what did, I'd get into trouble.

Unless you use hypergolic fuel, you cannot get away from that requirement. In the SA-6, the fuel was hypergolic magnesium vapor, which eliminated flameholding entirely. All they had to do was mix to burn.

Do it with a fuel-rich liquid rocket motor, and you still face the the operability limits versus injection Mach problem, just along a different tradeoff curve. That was the sort of thing Joe got around with hypergolic fuels.

GW

Last edited by GW Johnson (2016-09-20 09:17:52)

kbd512 · 2016-09-20 13:40:25

GW Johnson wrote:

Well, sometimes it's a little hard to distinguish what is really happening in a concept from the names, and from the pre-test hype. Estimates used to sell the program usually turn out to be quite high compared to what is found in test.

The figures given were from the PR-90 rocket that was actually tested, although the Soviets were also known to embellish test results.

GW Johnson wrote:

I know "air augmented rocket" and "ducted rocket" as synonyms for something more properly called a "gas generator-fed ramjet", which is distinct from "solid-fueled ramjet".

Hewitt Dissertation

GW Johnson wrote:

But I've also heard those same two terms applied to a shroud ring about a rocket engine, which is not the same thing at all. At low (subsonic) flight speeds, adding a ring about a rocket amounts to an ejector air pump, increasing the massflow in the combined plume and reducing its velocity. You do not need to afterburn in the ring, that just reduces Isp, and captured airflow.

We're going to accelerate this thing to Mach 1 first.

GW Johnson wrote:

But, done right, this adds greatly to plain rocket Isp, because you get increased propulsive efficiency if your jet speed is closer to your flight speed.

Thanks for the confirmation.

However, the ring is heavy to install, which is why most vehicles have never used this device. Not worth the weight penalty, not even for missile cruise flight, much less a climbing accelerator.

GW Johnson wrote:

At higher (supersonic) flight speeds, you need to burn fuel with the ring airflow to (theoretically) better match jet speed to flight speed. You still get the jet pump effect, but reduced, because the influences inside the ring cannot propagate upstream to influence the size of the captured airstream. That and the fuel flow reduce Isp far below the subsonic flight case. It's still an improvement, just not as much.

What applicability does this have to a solid fueled ramjet?

GW Johnson wrote:

And it's still heavy. Which is why no one ever did this.

With modern materials or what was available around 1960?

GW Johnson wrote:

That's about all that I know. Joe Bendot knew a lot more. He reversed this concept, and came at it as a full-blown ramjet engine, just one in which the fuel supply was largely a very high-speed fuel-rich rocket exhaust, packaged way down inside the ramjet. That's the ejector ramjet. Theoretically, it could develop static thrust. In practice, there were very serious flameholding issues. This thing required specialty hypergolic fuels to work.

Does flow turbulence cause the flame holding issues or the ignitors? Continuous burn pyrotechnic ignition (it's only operating for about a minute or so)?

GW Johnson wrote:

What I and others found in the solid gas generator-fed ramjets with ordinary flameholding-type fuels was that the fuel injection Mach number should never exceed 1. We saw a slight increase in the total pressure ratio across the combustor due to the incoming fuel momentum, and still maintain usable operating limits from a flame stability standpoint. Raise the injection Mach, and the operability limits narrowed very quickly, if it was ignitable at all.

If the temp is high enough the fuel will ignite, won't it? Was electrical ignition ever tried?

GW Johnson wrote:

You have to understand that a gas generator-fed ramet's fuel is quite unlike kerosene. It comes from a fuel-rich propellant, and is a mixture of carbon soot and carbon monoxide to burn, polluted with gas generator combustion products. This two-phase fuel supply imposes some very unique flameholding geometry requirements. Most of the common concept geometries will not work. If I told you what did, I'd get into trouble.

If the ignition temperature is high enough, will RP-1 or Jet-A ignite anyway?

GW Johnson wrote:

Unless you use hypergolic fuel, you cannot get away from that requirement. In the SA-6, the fuel was hypergolic magnesium vapor, which eliminated flameholding entirely. All they had to do was mix to burn.

Ok

GW Johnson wrote:

Do it with a fuel-rich liquid rocket motor, and you still face the the operability limits versus injection Mach problem, just along a different tradeoff curve. That was the sort of thing Joe got around with hypergolic fuels.
GW

Any experimentation with electric ignition?

GW Johnson · 2016-09-20 16:43:57

Hi kbd512:

I knew Pat Hewitt. Back then he knew a lot less than I did about the physical chemistry of fuel-air combustion and flameholding. Based on that dissertation, that statement is still true today.

None of the CFD codes have ever modeled what is really going on there. The key to the gas generator-fed ramjet (what Pat calls the ducted rocket ramjet) is centrifuging-out the solid carbon from the carbon monoxide plus products in the generator effluent. Raise the carbon/monoxide ratio in your effluent, and your ramjet burns poorly, if at all. Quite often not at all. Only certain geometries work, and none of them are in the textbooks.

This is true because the reaction rate of solid phase carbon with air is at least factor 100 to factor 1000 slower than almost any fuel gas species one cares to name. This is true essentially regardless of carbon soot size, right down to atomic scale. It is why finely-pulverized coal only burns at residence times approaching 1 full second, when we had about 0.7 msec in the flameholding zone of the 7 inch form of the ducted rocket ramjet. I understood this, Pat never did. Apparently still doesn't.

It is why every gas generator fuel propellant that I offered to the old USAF AMRAAM VFDR program back in the early 1990's performed way better than anything Pat and his group ever offered. Those were USAF-refereed tests. I let “theoreticals” go by the board to raise monoxide/soot ratio, when formulating propellants. This was well-proven in tests. Still true today, by the way.

Pat and his group offered boron because theoretically it looked better, but they knew nothing about how to burn it efficiently. I did know, and I did burn it very efficiently, and in two different ways. And I did it with more boron than they could use. And I did it without much smoke, unlike them. Insanely-stupid corporate politics closed my location before I could pass that technology on to USAF, though.

Electric ignition in ramjet combustors is not very feasible, because the high flow speeds literally blow the spark column out from between the spark plug electrodes. Been there and done that. In ramjets, cold flow approach speeds to the igniter vary between 0.6 and 0.7 Mach, in a decently-designed system, higher at higher altitudes. That's quite unlike turbine combustors, which typically run under 100 ft/sec or thereabouts in cold flow. On the ground, or close to it. Period.

Hypergolic ignition of kerosene-air requires air total temperatures on the order of 2000 F. That's well over Mach 3-4, even in the stratosphere. That's because tactical-size engines have overall residence times in the 2-6 msec range, and flameholder residence times about a third of that. Or thereabouts. You have to look for “fast” or “instantaneous” autoignition data in the literature to compare. That 1-sec stuff is just not applicable. Depends critically on design details as well. It's art, not science. Sorry.

The key features to flameholding are a recirculation flow pattern that feeds hot combustion gases back to the initial mixing point, and a recirculation residence time on the order of the same size as the time constant to react fuel and air. That goes with an overall effective air-fuel kinetic rate expression of the Arrhenius type, something usually completely lacking for these fuels in air at these conditions. Without recirculation, there is no flameholding, and thus there is no burning, unless your fuel is hypergolic with air at flameholder-residence-time timescales.

If the flameholder residence time is too long, combustion is too complete in that zone to maximize effective gas temperatures. Likewise, if too short, there is not enough combustion available to heat the zone gases up at all. It's a dynamic process, it has to be "right" to work, and it's art, not science. Not science yet, anyway. The closest thing to a workable theory is "perfectly-stirred reactor theory" (which real combustors are not!!!!), and it's way too limited and incomplete to serve as a design tool. It sort-of predicts troubles with short flameholder residence times, but does not account well at all for what happens at large scales where those residences times get too large.

None of the CFD codes have anything at all like this built into them. Not then, not today. They usually take the fuel and air in an element, and “burn” it to a user-input efficiency level, without any regard at all to what is ignitable and what is not. As I said, it's art, not written down for others to use out of a book. Nobody ever wanted to pay for that.

For the last couple of years, I've been trying to write as much of this down as I can in the time I have left (which ain't much). Hopefully, this will be to ramjet propulsion engineering what Hoerner's "Fluid Dynamic Drag" was to aerodynamic engineering. Sort of a smarts-and-data-bible.

GW

PS - I went back and got a PhD too, but I did it 8 years before Pat did. Mine's in general engineering, with the dissertation in alternate fuel effects in aircraft piston engines. I actually know why overall energy conversion efficiency is higher in test (!!!) with ethanol than gasoline in piston engines. That’s the point of my dissertation. However, airplanes, cars, lawnmowers, no difference. Turns out a piston engine is a piston engine, period.

kbd512 · 2016-09-20 22:33:49

GW Johnson wrote:

Hi kbd512:
I knew Pat Hewitt. Back then he knew a lot less than I did about the physical chemistry of fuel-air combustion and flameholding. Based on that dissertation, that statement is still true today.
None of the CFD codes have ever modeled what is really going on there. The key to the gas generator-fed ramjet (what Pat calls the ducted rocket ramjet) is centrifuging-out the solid carbon from the carbon monoxide plus products in the generator effluent. Raise the carbon/monoxide ratio in your effluent, and your ramjet burns poorly, if at all. Quite often not at all. Only certain geometries work, and none of them are in the textbooks.

No working CFD modeling codes? Bummer. From what you say, it seems as if a lot of this can only be determined experimentally and many of the people designing ramjets really don't have a very good understanding of what they're doing.

GW Johnson wrote:

This is true because the reaction rate of solid phase carbon with air is at least factor 100 to factor 1000 slower than almost any fuel gas species one cares to name. This is true essentially regardless of carbon soot size, right down to atomic scale. It is why finely-pulverized coal only burns at residence times approaching 1 full second, when we had about 0.7 msec in the flameholding zone of the 7 inch form of the ducted rocket ramjet. I understood this, Pat never did. Apparently still doesn't.

It sounds like this requires some sort of stagnation region behind the flame holder. If RP-1 is that much of a pain to work with, then what kind of problems would you encounter with LH2?

GW Johnson wrote:

It is why every gas generator fuel propellant that I offered to the old USAF AMRAAM VFDR program back in the early 1990's performed way better than anything Pat and his group ever offered. Those were USAF-refereed tests. I let “theoreticals” go by the board to raise monoxide/soot ratio, when formulating propellants. This was well-proven in tests. Still true today, by the way.

I'm just starting to read up on this. Recent tests would indicate that there is substantial variability in RP-1, never mind Jet-A.

GW Johnson wrote:

Pat and his group offered boron because theoretically it looked better, but they knew nothing about how to burn it efficiently. I did know, and I did burn it very efficiently, and in two different ways. And I did it with more boron than they could use. And I did it without much smoke, unlike them. Insanely-stupid corporate politics closed my location before I could pass that technology on to USAF, though.

I thought the French were still messing with boron. I also thought some sort of nano particle boron is being utilized to address a variety of problems with boron slurry.

GW Johnson wrote:

Electric ignition in ramjet combustors is not very feasible, because the high flow speeds literally blow the spark column out from between the spark plug electrodes. Been there and done that. In ramjets, cold flow approach speeds to the igniter vary between 0.6 and 0.7 Mach, in a decently-designed system, higher at higher altitudes. That's quite unlike turbine combustors, which typically run under 100 ft/sec or thereabouts in cold flow. On the ground, or close to it. Period.

Laser ignition?

NTRS 20030062091

Basic Laser Ingnition Experiments

GW Johnson wrote:

Hypergolic ignition of kerosene-air requires air total temperatures on the order of 2000 F. That's well over Mach 3-4, even in the stratosphere. That's because tactical-size engines have overall residence times in the 2-6 msec range, and flameholder residence times about a third of that. Or thereabouts. You have to look for “fast” or “instantaneous” autoignition data in the literature to compare. That 1-sec stuff is just not applicable. Depends critically on design details as well. It's art, not science. Sorry.

I figure I could pick the brain of the Rembrandt of Ramjets to try to determine what concepts pass the test of reasonability. I'd love to find a better / faster / cheaper way to get our astronauts and small cargo to orbit.

GW Johnson wrote:

The key features to flameholding are a recirculation flow pattern that feeds hot combustion gases back to the initial mixing point, and a recirculation residence time on the order of the same size as the time constant to react fuel and air. That goes with an overall effective air-fuel kinetic rate expression of the Arrhenius type, something usually completely lacking for these fuels in air at these conditions. Without recirculation, there is no flameholding, and thus there is no burning, unless your fuel is hypergolic with air at flameholder-residence-time timescales.

Is it really just flame holding, is it flame propagation, or something more sophisticated than that that's required for the engine to function as intended?

GW Johnson wrote:

If the flameholder residence time is too long, combustion is too complete in that zone to maximize effective gas temperatures. Likewise, if too short, there is not enough combustion available to heat the zone gases up at all. It's a dynamic process, it has to be "right" to work, and it's art, not science. Not science yet, anyway. The closest thing to a workable theory is "perfectly-stirred reactor theory" (which real combustors are not!!!!), and it's way too limited and incomplete to serve as a design tool. It sort-of predicts troubles with short flameholder residence times, but does not account well at all for what happens at large scales where those residences times get too large.

Well, thanks anyway for the added reading material to run down.

GW Johnson wrote:

None of the CFD codes have anything at all like this built into them. Not then, not today. They usually take the fuel and air in an element, and “burn” it to a user-input efficiency level, without any regard at all to what is ignitable and what is not. As I said, it's art, not written down for others to use out of a book. Nobody ever wanted to pay for that.

I would really like to know if lasers could meaningfully effect what is and is not ignitable.

GW Johnson wrote:

For the last couple of years, I've been trying to write as much of this down as I can in the time I have left (which ain't much). Hopefully, this will be to ramjet propulsion engineering what Hoerner's "Fluid Dynamic Drag" was to aerodynamic engineering. Sort of a smarts-and-data-bible.

I'll be waiting in line to get my copy autographed, GW.

GW Johnson wrote:

GW
PS - I went back and got a PhD too, but I did it 8 years before Pat did. Mine's in general engineering, with the dissertation in alternate fuel effects in aircraft piston engines. I actually know why overall energy conversion efficiency is higher in test (!!!) with ethanol than gasoline in piston engines. That’s the point of my dissertation. However, airplanes, cars, lawnmowers, no difference. Turns out a piston engine is a piston engine, period.

I'm working on building a Molt Taylor Mini-IMP. Any reading material you could provide about contending with torsional vibration would be most helpful.

GW Johnson · 2016-09-22 09:51:23

combustion processes in airbreathing combustors:

The way I conceptualize airbreathing-combustor engine combustion, there's a recirculation zone or zones that is (are) the so-called flameholder, really just a point of continuous ignition for incoming fuel and air. Those zones maximize hot gas output at a partial level of combustion efficiency. Hot gas output is what the rest of the engine needs.

Outside the recirculation zone(s), the problem is one of flame propagation throughout the incoming charge, before it reaches the nozzle contraction. Since turbulent flame speeds are usually in the vicinity of 100 ft/sec or so, while ramjet or afterburner flow speeds usually exceed 1000 ft/sec, you can see the importance of proper distribution. Plain old turbulence just cannot hack it, all alone.

And that's for pre-mixed fuel and air coming in the inlet. If you do direct fuel injection into the combustor (which you just about have to do, with gas generator-fed systems), there's another layer of mixing that has to get done, all in the short available residence time (small handful of msec at tactical sizes).

Burning liquid fuels, there's a liquid droplet vaporization problem that must be solved before the recirculating flow makes its reverse turn, or you will quench the flameholder with a heat absorbing liquid spray. The solutions are different in each geometry, but these fuels can be burned in any conceivable ramjet engine or afterburner geometry. Since chemistry rates don't scale, there's a min size below which you cannot make it work, for each geometry class. More volatile and more reactive fuels equates to smaller min sizes.

In the gas generator-fed solids, there's no phase change, but the analogous problem is the two-phase nature of the fuel effluent to be burned. It's actually the harder of the two to address. Most effluents are a mix of carbon monoxide gas, solid carbon soot, and various gaseous and condensed combustion products. These all show up at about gas generator flame temperature, which is around 1000 F colder than the feasible range of temperatures in the flameholder. So the solid soot and the combustion product species both act as cold quench agents.

Of the combustible species, the gas CO reacts around 100 to 1000 times faster than the soot can, even for near atomic-scale soot particle sizes. The soot simply cannot react at short residence times, so it is doubly a quench agent. It merely dilutes the CO concentration further. You are only burning CO in the recirculation zones. Which means that your fuel effluent needs to equal (or better yet exceed) about 25-30% CO by mass. If you don't, you will not burn at all. Formulating propellants for high theoretical heating value always drive soot up and CO down, which is exactly why you do not want to formulate that way.

To make this work, there has to be a strong, well-organized vortex in the recirculation zone of a size that fills the zone. This centrifuges the solids to the outer edge of the vortex, leaving the CO and product gases "in the clear" in the interior of the vortex, where they can burn. You get this vortex only in an asymmetrical side entry inlet class of geometries. Those typically have far higher mass entrainment ratios into the flameholder zone than baffle, V-gutter, or even coaxial dump schemes, so flameholder residence times are around a third of the overall residence times. In those other geometries, they are more nearly equal.

Short residence time makes the CO- vs C-burning problem worse in gas generator-fed systems, balanced against the centrifuge effect that makes effluent burning feasible at all. There is also a short residence time vaporization vulnerability in asymmetric side entry with liquid fuels, but it is nowhere near as fatal as the effluent soot content with the solids.

That is why I typically say that you can burn liquid fuels in any ramjet geometry you want, but you can burn non-hypergolic gas generator effluents only in asymmetric twin side inlet geometries. If you have to use non-hypergolic effluents in a 4-side inlet configuration, you must offset one asymmetric pair downstream.

It's also why I was able to get good burns with non-hypergolic effluents down to about 4.5 inch combustor diameters, but not below, and no one else ever did, either. The flameholder residence times are just too short in practical flyable geometries. And I only got good burns at 4.5 inch size by stretching the combustor diameter a little bit, relative to properly-scaled inlet and throat diameters. That's a symptom of residence time distributions being too short.

I also had to stretch L/D a little to make scaled-down performance look like full scale tactical. That's the overall residence time effect, which is much easier to "fix".

torsional vibrations:

I know a lot less about this, but I do know that engine crankshafts can get broken by it. If the prop or flywheel has a big rotational inertia, and/or the distance between prop or flywheel and center of the crankshaft is long, the inertia terms get large compared to the stiffness, and the system becomes very vulnerable to torsional vibration. As the crankshaft twists, this can show up as oscillating errors in spark timing, varying from cylinder to cylinder. It also leads to rapid fatigue failure in the crank. That's about all I know about that topic, really. The P-39 and P-63 aircraft designs were vulnerable to this, because of the long shaft between nose prop and behind-the-pilot engine.

GW

Last edited by GW Johnson (2016-09-22 10:19:17)

RGClark · 2016-10-03 08:06:14

kbd512 wrote:

SSTO is mostly a pipe dream, but propellant tank staging should actually work.
GW has a pretty good article on why a LOX/RP-1 SSTO is infeasible:
SSTO Trade Studies
Kirk Sorensen has a pretty good article on why NERVA SSTO is infeasible:
SSTO is a bad idea, but NTR SSTO is worse
...

You should read the comments to the GW post. The analysis in the post is based on assumptions GW made for the possible Isp and dry mass fraction. Both of these have been exceeded for real world rockets. When you take into account what's actually doable now you find the payload for a SSTO is significantly higher so that you have sufficient payload capacity to add the systems needed for reusability.

Bob Clark

Last edited by RGClark (2016-10-05 01:41:27)

kbd512 · 2016-10-03 11:40:54

RGClark wrote:

You should read the comments to the GW post. The analysis in the post is based on assumptions GW made for the possible Isp and dry mass fraction. Both of these have been exceeded for real world rockets. When you take into account what's actually doable now you find the payload for a SSTO is significantly higher so that you have sufficiently payload capacity to add the systems needed for reusability.
Bob Clark

Antius wanted to determine the feasibility of a reusable LOX/RP-1 SSTO. I wanted to determine the feasibility of a non-reusable but inexpensive assembly-line-produced LOX/Jet-A SSTO or TSTO. How cost effective can an air launch or surface launch of a non-reusable SSTO or TSTO be for delivery of 4t to 6t payloads to LEO? SSTO would be ideal, but as GW pointed out there are also some practical limitations to consider with respect to engine throttling and vehicle acceleration near burnout.

I would like to know whether or not an aerospike nozzle with injector sets that can be shut off as the engine throttles down is a potential solution to that problem. I would also like to know whether or not the Russian GNOM missile concept can be made to work with solid rockets to dramatically boost the Isp of the rocket between 40K and 100K to provide 2 - 3km/s acceleration prior to burnout followed by a small LOX/Jet-A upper stage. By all accounts, the Russian tests were fairly successful, but the concept was dropped when the chief designer died during the project.

I think a subsonic air launch from 40K feet on the correct trajectory is the way to go for flexibility. I have a thread about redesigning StratoLaunch which indicates why our military needs a small fleet of multi-purpose super heavy lift cargo transport aircraft with air launch capability for drones, missiles, and 5t payload class rockets to deliver low-cost, replaceable comms and spy satellites. I've pushed that concept pretty hard because nothing we presently have meets those general requirements. Every launch is a hand-wringing event and obscenely expensive. Unfortunately, StratoLaunch I is a one-off airframe that will be unaffordable to operate and unsuitable for the other purposes I've assigned for reasons I've already noted in that thread.

How we can resupply ISS with cargo every other month for $10M per mission or less? How do we can resupply ISS with astronauts for $50M per mission or less ($10M for the rocket and $30M or less for a two crew member capsule)? At those price points, there's no justification for not maintaining the only orbital laboratory and technologies testing facility in existence. ISS resupply should be noise in NASA's overall budget so that more funding can be directed towards CL-ECLSS, ZBO cryogen storage, ISPP, MCP suits, and other crucial enabling technologies for real space exploration.

RGClark · 2016-10-03 13:54:40

GW Johnson wrote:

Well, sometimes it's a little hard to distinguish what is really happening in a concept from the names, and from the pre-test hype. Estimates used to sell the program usually turn out to be quite high compared to what is found in test.
I know "air augmented rocket" and "ducted rocket" as synonyms for something more properly called a "gas generator-fed ramjet", which is distinct from "solid-fueled ramjet".
But I've also heard those same two terms applied to a shroud ring about a rocket engine, which is not the same thing at all. At low (subsonic) flight speeds, adding a ring about a rocket amounts to an ejector air pump, increasing the massflow in the combined plume and reducing its velocity. You do not need to afterburn in the ring, that just reduces Isp, and captured airflow.
But, done right, this adds greatly to plain rocket Isp, because you get increased propulsive efficiency if your jet speed is closer to your flight speed.
However, the ring is heavy to install, which is why most vehicles have never used this device. Not worth the weight penalty, not even for missile cruise flight, much less a climbing accelerator.
At higher (supersonic) flight speeds, you need to burn fuel with the ring airflow to (theoretically) better match jet speed to flight speed. You still get the jet pump effect, but reduced, because the influences inside the ring cannot propagate upstream to influence the size of the captured airstream. That and the fuel flow reduce Isp far below the subsonic flight case. It's still an improvement, just not as much.
And it's still heavy. Which is why no one ever did this.

I'd like to see this actually be tried. For instance the famous Centaur upper stage, used for example on the Atlas V, has enough delta-v to be SSTO, but it's engine does not have enough thrust to take off from the ground. Also, it already has a nozzle extension attached that is used to increase the area ratio in vacuum. Perhaps this nozzle extension when not extended could act just like ring duct for a ducted rocket.

It would need though to be able to double the thrust for it to work though.

Bob Clark

GW Johnson · 2016-10-03 14:49:49

Resupply the space station for a month or two at $10M?

Well, the most egregiously simple and bone-headed thing would be to reprise the old "Scout" launch vehicle that had such a fabulously safe launch history in the 1960's and 1970's. It was an utterly simple 4-stage solid that delivered about 200 lb to GTO. I'd guess around 500-600 lb to LEO.

Those solid motors are no longer available, and we'd want to scale up, anyway. But there ought to be some motors used as SRB's and some ex-ICBM stage motors we could use.

No complicated vectoring nozzles, just plain, simple, fixed-geometry missile stuff. Just add peroxide or hydrazine monopropellant thrusters for attitude control, which is what Scout had. Maybe a hydrazine-NTO propulsion thruster to get it close enough to ISS for arm grab and docking.

If you put a heat shield on it as the payload shroud, and add a recovery chute, you could salvage and re-fly the guidance system, and you could return stuff from ISS. But no more -- you have to think like an artillery round designer, or costs will explode.

Other than that, it's the same very-low logistics-tail type of wooden-round solid system that made Minute-Man practical and affordable as an ICBM, when liquid-propellant Atlas and Titan were not, and could never be.

Just a concept. No numbers. No motors identified. Nothing. But it just might fill the bill of sending ~1000 lb to ISS once every month or two.

GW

Last edited by GW Johnson (2016-10-03 14:52:27)

kbd512 · 2016-10-03 16:45:50

GW Johnson wrote:

Resupply the space station for a month or two at $10M?

$5M is the goal, but $10M is the plan. How can we affordably and continuously keep feeding consumables, experiments, tools, and satellites to the astronauts, industries, and military that we've paid so dearly to develop?

Military satellites should cost $25M max, with a design life of five years instead of ten to fifteen years. Every five years, satellite constellations should be replaced. The frequent replacement ensures that better electronics, power, and propulsion technologies are regularly incorporated into our constellations.

I want to launch every week and I want swappable payloads. If, for whatever reason, your payload isn't ready to launch, then we're sending a different payload and you have to wait another week. The service needs to be so routine that corporations, universities, and the military can simply expect it to work like air mail. You deliver a payload to the airport, pay the military for the use of their aircraft and personnel, and they deliver your payload. I proposed using LOX and Jet-A because they're inexpensive and readily available, but solids pose less of a danger to the launching aircraft and have fewer special handling requirements. Perhaps a solid first stage followed by a LOX and Jet-A upper stage is the way to go.

GW Johnson wrote:

Well, the most egregiously simple and bone-headed thing would be to reprise the old "Scout" launch vehicle that had such a fabulously safe launch history in the 1960's and 1970's. It was an utterly simple 4-stage solid that delivered about 200 lb to GTO. I'd guess around 500-600 lb to LEO.

This appears to be what I want, only smaller than what I had in mind.

GW Johnson wrote:

Those solid motors are no longer available, and we'd want to scale up, anyway. But there ought to be some motors used as SRB's and some ex-ICBM stage motors we could use.
No complicated vectoring nozzles, just plain, simple, fixed-geometry missile stuff. Just add peroxide or hydrazine monopropellant thrusters for attitude control, which is what Scout had. Maybe a hydrazine-NTO propulsion thruster to get it close enough to ISS for arm grab and docking.
If you put a heat shield on it as the payload shroud, and add a recovery chute, you could salvage and re-fly the guidance system, and you could return stuff from ISS. But no more -- you have to think like an artillery round designer, or costs will explode.
Other than that, it's the same very-low logistics-tail type of wooden-round solid system that made Minute-Man practical and affordable as an ICBM, when liquid-propellant Atlas and Titan were not, and could never be.
Just a concept. No numbers. No motors identified. Nothing. But it just might fill the bill of sending ~1000 lb to ISS once every month or two.
GW

I like stupid simple, but not so stupid that there are stupid requirements attached to its use. Like I've said numerous times before, small and efficient is beautiful. Let's service 90% of the requirement instead of over-engineering the solution to service 99% of all possible use cases.

RGClark · 2016-10-06 07:46:42

RGClark wrote:

GW Johnson wrote:
Well, sometimes it's a little hard to distinguish what is really happening in a concept from the names, and from the pre-test hype. Estimates used to sell the program usually turn out to be quite high compared to what is found in test.
I know "air augmented rocket" and "ducted rocket" as synonyms for something more properly called a "gas generator-fed ramjet", which is distinct from "solid-fueled ramjet".
But I've also heard those same two terms applied to a shroud ring about a rocket engine, which is not the same thing at all. At low (subsonic) flight speeds, adding a ring about a rocket amounts to an ejector air pump, increasing the massflow in the combined plume and reducing its velocity. You do not need to afterburn in the ring, that just reduces Isp, and captured airflow.
But, done right, this adds greatly to plain rocket Isp, because you get increased propulsive efficiency if your jet speed is closer to your flight speed.
However, the ring is heavy to install, which is why most vehicles have never used this device. Not worth the weight penalty, not even for missile cruise flight, much less a climbing accelerator.
At higher (supersonic) flight speeds, you need to burn fuel with the ring airflow to (theoretically) better match jet speed to flight speed. You still get the jet pump effect, but reduced, because the influences inside the ring cannot propagate upstream to influence the size of the captured airstream. That and the fuel flow reduce Isp far below the subsonic flight case. It's still an improvement, just not as much.
And it's still heavy. Which is why no one ever did this.
I'd like to see this actually be tried. For instance the famous Centaur upper stage, used for example on the Atlas V, has enough delta-v to be SSTO, but it's engine does not have enough thrust to take off from the ground. Also, it already has a nozzle extension attached that is used to increase the area ratio in vacuum. Perhaps this nozzle extension when not extended could act just like ring duct for a ducted rocket.
http://www.alternatewars.com/BBOW/Space … utaway.jpg
It would need though to be able to double the thrust for it to work though.

This idea of "air entrainment" was used to make an easily inflatable sleeping pad:

Amazing Air Pad Inflates in Seconds w/ NO power or pumping.
http://www.youtube.com/watch?v=JydG8Iyr7Kw

Since this is based on the well-known and well-studied Venturi effect it should be calculatable how much extra thrust can be generated by using it:

https://en.wikipedia.org/wiki/Venturi_effect

A key application would be the SLS core stage. It's irritating that we have these high performance engines in the SSME's designed to be reusable, and we would throw four of them away with each SLS flight. But if you look at the specs on the SLS core, it has the delta-v to be SSTO, but it could not lift off from ground with the four SSME engines. But you need only 50% more thrust for it to be able to lift off. Could using a duct ring allow it have the extra thrust needed?

Bob Clark

GW Johnson · 2016-10-06 09:03:17

Putting a ring shroud around a rocket engine can induce airflow through the shroud that increases thrust, yes. At very subsonic to static speeds. If the friction of the airflow past the rocket engine can be reduced, you might get factor 1.4 more thrust statically, less as your speed increases. Smooth shapes and proper venturi internal profiles are required to make this work at all. The ducted propellor is another example of the same device with a different prime mover. Same sort of airspeed restrictions apply. Static to about 200 mph.

The air ejector pump takes this to extremes by going to a geometry that is a whole lot more complicated (and heavy) than a simple venturi shroud ring. It only works statically, and nobody is interested in its thrust, only the pressure increase it can supply without any moving parts. As a pump, its efficiency is very low. There's a whole lot more massflow coming from very high pressure in the driving jet, than any low pressure airflow it can induce. The thing essentially works by fluid friction between the two streams, which is inherently wasteful. Applications I am familiar with include taking a rocket propellant mix bowl to hard vacuum conditions for mixing without air entrainment.

I'm not at all sure this thing would ever be worth the extra weight to add it to a launch rocket. Using the skirt extension for your shroud ring has the incorrect venturi geometry, you will not get much of the static thrust multiplier of 1.4.

GW

Last edited by GW Johnson (2016-10-06 09:04:09)

SpaceNut · 2016-10-06 19:02:59

Gw is right about the solid rocket motors that we could repurpose from disassembly of the military rockets as is being done for the use in the Minotaur but there are others that we can use as well. It is simple light it and go to which we only need a liquid fuel for the last little bit to catch and dock with the station.

https://en.wikipedia.org/wiki/Solid-fuel_rocket

Due to reliability, ease of storage and handling, solid rockets are used on a number of missiles and ICBMs.
Air-to-air missiles: AIM-9 Sidewinder
Ballistic missiles: Jericho
ICBMs: LGM-30 Minuteman, LGM-118 Peacekeeper, RT-2PM Topol, DF-41, Agni-V
Orbital rockets
Solid rockets are suitable for launching small payloads to orbital velocities, especially if three or more stages are used. Many of these are based on repurposed ICBMs.
Scout
Athena
Mu
Pegasus
Taurus
Minotaur
Start-1
PSLV - alternating solid and liquid stages
Shavit
Vega
Long March 11
Larger liquid-fueled orbital rockets often use solid rocket boosters to gain enough initial thrust to launch the fully fueled rocket.

elderflower · 2016-10-07 04:22:19

Can anyone explain why we are launching from sea level. Would it not be more efficient to launch from a high plateau near the equator so that the rocket starts as far as possible from the earth's axis?

GW Johnson · 2016-10-08 08:59:36

Hi Elderflower:

I think the answer to your question is two-fold. (1) there really aren't very many sites like that available, those being the Andes Mountains in South America, and the mountains in Eastern Africa. Those aren't really plateaus, they're real mountains or mountain ranges, a less useful topography. (2) the logistics for building anything significant, and supply the rockets to them, are next-to-impossible. Both locations are far from the industrialized world, and lack any significant transportation. In particular, the most practical means of transporting very large rocket stages has proven to be by barge on the ocean. You can't do that with those two places, because there's no way to haul big stuff up the mountain.

As far as I know, the ranking in order of importance of things needed to reach orbit are (1) speed, (2) path angle, and (3) altitude. Near-equatorial gets you the most speed effect from the Earth's rotation, but being within about 25 degrees of the equator seems to be "good enough". (Not having that is why the Russian rockets were so much larger than ours for so long.) The path angle thing relates to zero-lift ballistic trajectories that minimize drag. That's why vertical launch rockets seem to work so very well. Altitude at launch doesn't buy you very much, but being at sea level buys you barge transport for your logistics. That seems to be far more important.

GW

Last edited by GW Johnson (2016-10-08 12:14:04)

Antius · 2016-10-08 12:22:04

Musk's ITS rocket system appears to be a reusable TSTO with huge scale economies. In reusable configuration it is capable of 300 tonne to LEO lift capability. I am curious as to what the practical limits of this sort of configuration could be. Could we build 1000 tonnes to LEO launchers? At what point do diseconomies of scale make increasing size counterproductive?

GW Johnson · 2016-10-08 12:52:57

Antius:

I honestly don't know the answer to your question. Musk's giant ship design already looks alarmingly large to me.

He's using carbon fiber structures, but there are limits to what materials can do in the way of strength for the weight. Weight more-or-less scales up as dimension cubed, while part strengths only scale up more-or-less as dimension squared. Sooner or later you hit material limits you cannot pass.

There can be serious, even fatal, unexpected problems when you start using new materials. Such as room-temperature aging in beta-phase titanium that catastrophically cuts strength and increases brittleness. The aluminum-lithium alloys that the rocket makers have jumped on has a somewhat similar low-temperature aging problem, just not as catastrophic, percentage-wise.

Even the third generation lithium-aluminum alloys still suffer from it. It's awfully easy to take a piece of that stuff and accidentally heat it to only 85 C. That's the aging temperature, even for the third generation alloys. You lose around 20-30% of your ultimate strength, and about half your ductility.

Carbon fiber stuff exists in a lot of forms, the most common of which is carbon cloth or fiber in an epoxy matrix as an organic composite. That stuff is junk at only boiling water temperatures. Even the "best" epoxies are beginning to actually char (and literally fall apart) at only 290 F. Most are completely junk at only 200 F.

In contrast, you can take aluminum 2024 and 6061 to about 300 F. Steel case bombs and rockets sitting out in the sun at ammo dumps have been measured at 180-190 F. Even painted white. The usual military spec is serviceability at 160 F, reduced in recent decades to 145 F, because so few manufacturers could successfully comply. But they saw 180-190 F as recently as Desert Storm.

I worry about things like that, in structures exposed to space and entry environments. Not worrying enough about that sort of thing is what killed shuttle Columbia (a gas leak is inevitable; what do you do then?).

GW

Last edited by GW Johnson (2016-10-08 13:00:48)

Terraformer · 2016-10-08 15:42:21

I find it striking that the ITS, in expendable configuration, is meant to throw as much mass as Sea Dragon. Which is why I wondered if he was considering sea launch for it. There were also proposals to reuse the Sea Dragon first stage.

RGClark · 2016-10-09 10:51:17

GW Johnson wrote:

Putting a ring shroud around a rocket engine can induce airflow through the shroud that increases thrust, yes. At very subsonic to static speeds. If the friction of the airflow past the rocket engine can be reduced, you might get factor 1.4 more thrust statically, less as your speed increases. Smooth shapes and proper venturi internal profiles are required to make this work at all. The ducted propellor is another example of the same device with a different prime mover. Same sort of airspeed restrictions apply. Static to about 200 mph.
The air ejector pump takes this to extremes by going to a geometry that is a whole lot more complicated (and heavy) than a simple venturi shroud ring. It only works statically, and nobody is interested in its thrust, only the pressure increase it can supply without any moving parts. As a pump, its efficiency is very low. There's a whole lot more massflow coming from very high pressure in the driving jet, than any low pressure airflow it can induce. The thing essentially works by fluid friction between the two streams, which is inherently wasteful. Applications I am familiar with include taking a rocket propellant mix bowl to hard vacuum conditions for mixing without air entrainment.
I'm not at all sure this thing would ever be worth the extra weight to add it to a launch rocket. Using the skirt extension for your shroud ring has the incorrect venturi geometry, you will not get much of the static thrust multiplier of 1.4.
GW

A thrust multiplier of 1.4 is quite close to the 1.5 needed for the SLS core stage to lift off on its own. Do you know of references for the ducted rocket? Perhaps with tweaking we could get that extra 10% performance. The ducted nozzle extension doesn't have to be of the type on the Centaur; we could make it of similar form to the duct rings already used.

In regards to the velocity limit, as a rocket goes faster it is also burning off propellant and progressing towards lower air density with increasing altitude. Then the required thrust will be less because of the reduced mass, but the rocket thrust itself will be increased because of the reduced air density. So it still might work

Bob Clark

elderflower · 2016-10-09 13:15:28

Thanks GW
I suppose that there are political stability issues as well for a lot of possible places.
The mining industry seems to be able to open large undertakings in very hostile terrain, such as the mountains of New Guinea and South America despite the drawbacks, so I supposed that there would be a technical reason. Good enough for the task is usually a good enough reason.

GW Johnson · 2016-10-09 16:13:50

Elderflower:

Glad what I said helped. Looks to me like there's a pretty good reason most launch sites except the Russian ones are +/- 25 degrees equatorial, and pretty much sea level.

Bob Clark:

The very best you can get is T/Tplain = 1.4. It is seemingly not possible to exceed the 1.4, but it is very possible to do worse. I saw this in a report about unsteady augmentors associated with a pulsejet from long, long ago.

For the unsteady case, T/Tplain was nearer 2. At least at the time of the reports (1959-1964), no one had ever, ever shown T/Tplain > 1.4 in steady flow. Factor 1.1-1.2 is reasonably likely, though.

Sorry.

GW

kbd512 · 2016-10-09 22:46:52

GW Johnson wrote:

Antius:
I honestly don't know the answer to your question. Musk's giant ship design already looks alarmingly large to me.
He's using carbon fiber structures, but there are limits to what materials can do in the way of strength for the weight. Weight more-or-less scales up as dimension cubed, while part strengths only scale up more-or-less as dimension squared. Sooner or later you hit material limits you cannot pass.
There can be serious, even fatal, unexpected problems when you start using new materials. Such as room-temperature aging in beta-phase titanium that catastrophically cuts strength and increases brittleness. The aluminum-lithium alloys that the rocket makers have jumped on has a somewhat similar low-temperature aging problem, just not as catastrophic, percentage-wise.
Even the third generation lithium-aluminum alloys still suffer from it. It's awfully easy to take a piece of that stuff and accidentally heat it to only 85 C. That's the aging temperature, even for the third generation alloys. You lose around 20-30% of your ultimate strength, and about half your ductility.
Carbon fiber stuff exists in a lot of forms, the most common of which is carbon cloth or fiber in an epoxy matrix as an organic composite. That stuff is junk at only boiling water temperatures. Even the "best" epoxies are beginning to actually char (and literally fall apart) at only 290 F. Most are completely junk at only 200 F.
In contrast, you can take aluminum 2024 and 6061 to about 300 F. Steel case bombs and rockets sitting out in the sun at ammo dumps have been measured at 180-190 F. Even painted white. The usual military spec is serviceability at 160 F, reduced in recent decades to 145 F, because so few manufacturers could successfully comply. But they saw 180-190 F as recently as Desert Storm.
I worry about things like that, in structures exposed to space and entry environments. Not worrying enough about that sort of thing is what killed shuttle Columbia (a gas leak is inevitable; what do you do then?).
GW

GW,

What about the new resins for composites?

Resins for the Hot Zone, Part I: Polyimides

GW Johnson · 2016-10-10 09:22:04

Well, what I had experience with was the older materials. Epoxies, vinyl esters, and polyesters. The first one works pretty good with carbon fiber, the other two work very well with glass and kevlar fibers.

The new polyimides in that link you provided are intriguing. I'm surprised and pleased they go so hot. It'll be interesting to see these used in missile fins and skins.

There is a problem with higher glass transition temperatures, though. It makes them too brittle to be useful as cryogenic tanks. Just something else to worry about, I suppose.

GW

kbd512 · 2016-10-10 13:05:36

GW Johnson wrote:

Well, what I had experience with was the older materials. Epoxies, vinyl esters, and polyesters. The first one works pretty good with carbon fiber, the other two work very well with glass and kevlar fibers.
The new polyimides in that link you provided are intriguing. I'm surprised and pleased they go so hot. It'll be interesting to see these used in missile fins and skins.
There is a problem with higher glass transition temperatures, though. It makes them too brittle to be useful as cryogenic tanks. Just something else to worry about, I suppose.
GW

Is there any way to use the new polymer aerogels to insulate the inside of the cryogen tanks and then to use the high-temp resins for binding the fibers in the composite tank structures, or maybe the other way around (normal binders for the tank to withstand cryogenic temperatures and ablative spray-on aerogel polymer foam on the outside of the tank)?

GW Johnson · 2016-10-10 17:09:01

Kbd512:

I dunno. I'd start with a double shell separated by some sort of insulation, which could be your aerogel. The inner shell is a cryo-tank structure. The outer shell needs to be able to take some aeroheating. They are unlikely to be the same material.

Most organic composites are quite porous, actually. This thing will need some sort of liquid seal liner layer, which adds to inert weight. I do not forsee practical cryo-tanks survivable upon entry at Mars or Earth, with only 4-5% inert weights.

That's where Spacex is right now using aluminum-lithium alloys, uninsulated. I think without proof that they have already bumped into new-material aging problems with it, too.

Considering how they do not hire folks over about 40-45 years age, it's hard to see how the necessary materials expertise is actually on their staff to deal with such oddball materials problems.

Sorry, it just ain't all science. It's about 40% science (written down), it's about 50% art (never written down, passed-on one-on-one on the job if there's old guys on your staff), and about 10% blind dumb luck. That's in production. In development, the art and luck percentages are higher. Pretty much true across engineering, not just "rocket science".

GW

Last edited by GW Johnson (2016-10-10 17:10:31)

New Mars Forums

Announcement

#26 2016-09-20 09:08:27

Re: Reusable LOX/Kerosene SSTO with drop tanks

#27 2016-09-20 13:40:25

Re: Reusable LOX/Kerosene SSTO with drop tanks

#28 2016-09-20 16:43:57

Re: Reusable LOX/Kerosene SSTO with drop tanks

#29 2016-09-20 22:33:49

Re: Reusable LOX/Kerosene SSTO with drop tanks

#30 2016-09-22 09:51:23

Re: Reusable LOX/Kerosene SSTO with drop tanks

#31 2016-10-03 08:06:14

Re: Reusable LOX/Kerosene SSTO with drop tanks

#32 2016-10-03 11:40:54

Re: Reusable LOX/Kerosene SSTO with drop tanks

#33 2016-10-03 13:54:40

Re: Reusable LOX/Kerosene SSTO with drop tanks

#34 2016-10-03 14:49:49

Re: Reusable LOX/Kerosene SSTO with drop tanks

#35 2016-10-03 16:45:50

Re: Reusable LOX/Kerosene SSTO with drop tanks

#36 2016-10-06 07:46:42

Re: Reusable LOX/Kerosene SSTO with drop tanks

#37 2016-10-06 09:03:17

Re: Reusable LOX/Kerosene SSTO with drop tanks

#38 2016-10-06 19:02:59

Re: Reusable LOX/Kerosene SSTO with drop tanks

#39 2016-10-07 04:22:19

Re: Reusable LOX/Kerosene SSTO with drop tanks

#40 2016-10-08 08:59:36

Re: Reusable LOX/Kerosene SSTO with drop tanks

#41 2016-10-08 12:22:04

Re: Reusable LOX/Kerosene SSTO with drop tanks

#42 2016-10-08 12:52:57

Re: Reusable LOX/Kerosene SSTO with drop tanks

#43 2016-10-08 15:42:21

Re: Reusable LOX/Kerosene SSTO with drop tanks

#44 2016-10-09 10:51:17

Re: Reusable LOX/Kerosene SSTO with drop tanks

#45 2016-10-09 13:15:28

Re: Reusable LOX/Kerosene SSTO with drop tanks

#46 2016-10-09 16:13:50

Re: Reusable LOX/Kerosene SSTO with drop tanks

#47 2016-10-09 22:46:52

Re: Reusable LOX/Kerosene SSTO with drop tanks

#48 2016-10-10 09:22:04

Re: Reusable LOX/Kerosene SSTO with drop tanks

#49 2016-10-10 13:05:36

Re: Reusable LOX/Kerosene SSTO with drop tanks

#50 2016-10-10 17:09:01

Re: Reusable LOX/Kerosene SSTO with drop tanks

Board footer