Quite a bit. I could probably figure out a first-order estimate by hand.
Maybe Aero's claim that basic theory will get you to 3Km/s (about M8.8 ) is correct but given the results of the X30 programme what I'm suggesting is a flight programme.
Quote from: john smith 19 on 08/16/2014 11:28 AMMaybe Aero's claim that basic theory will get you to 3Km/s (about M8.8 ) is correct but given the results of the X30 programme what I'm suggesting is a flight programme.And really what DOES air-breathing to @Mach-5+ get you if its "cheap" and "easy" enough over the alternatives?(I'd suggest hitting up the "Ex-Rocketman's Take" blog to see some of the work he's done as one of "those" engineers )
I have downloaded the copy of the Glen Olson's site, including the ramjet design tool (although working out how to use it looks tricky. I'm guessing it's imperial units, but I'm not sure ).
My suggestion is a pure research project to settle an apparently simple question (it's one of the outstanding question's on Olson's site). In the same way the X1 was designed to answer the question "could you build a vehicle big enough to carry a person that could exceed M1".
My impression is that it seems that historically a missile requirement has been stated and people have built a ram jet to meet it but no one seems to checked how fast can such an engine go before you have to start looking at full supersonic combustion or pure rocket systems. In the US upper speeds have only been established by stuck open valves (and I'll note they seem to have been still accelerating so thrust > drag and lift > mass) while the French report you cited indicated that a 10 flight M5 test programme seemed to be fairly straight forward.
Olson suspected the operating range of 3 Mach numbers of most actual missiles was a convenience for the designers (and a perfectly valid trade off in a weapon system). The implication being that with more attention to detail and possibly accepting you'd lose some payload that range could be much wider.
I'll note the Russian/Indian "BraMos" anti ship missile is good to M2.8-3.0 and their planned 2nd generation is to M7. They don't seem to be talking about supersonic combustion to do it but they do seem to want liquid fuel.
As a side note I will note that M7 is 1/4 of the velocity to orbit including losses and M13 (using M1=340 m/s) is about 1/2 to orbital velocity.If LEO is "half way to anywhere" then I guess at 1/2 LEO velocity you'd be a quarter way to anywhere?TBH that's beyond what the comments of the people Olson talked to thought possible but the truth is that despite the first one running in the early 1930's we still don't know.
I'm still playing with it when I remember I have it
Anyway, in some of his notes Glenn mentiones that during his research there was enough written data to support the conslusion that "someone" had flown a subsonic ramjet engine at speeds to around Mach-8 and most seemed confident that Mach-10 was possible and still generate thrust. But as you note, we really don't KNOW what the limits are.
As far as I know there were only a few actual "test" ramjet vehicles the best known one being the X7. Most engines were "point-designs" specifically designed to operate at a restricted range speeds.
I got from his site and other sources that if you're willing to put up with the expense and compelxity of moving inlets and exhaust AND intergrate into the vehicle (the majority of early work concentrated on podded engines as experiance with imbedded jet engines hadn't been as good as hoped, turns out different engines have different operational experiances... Who knew ) then you're main barrier should JUST be generating thrust.Of course that in itself can cause some issues to be worried about... (below)A LOT of folks "talk" SCramjets but operationally its all been ramjets and everything I've seen so far points to subsonic combustion ramjets being perfectly capable of doing the job.
Mach-7 with a subsonic combustion ramjet seems dooable, Mach-13...The main point as you say is we simply don't know. Designing a vehicle to be able to test up to such speeds isn't easy and the need for variable intake/exhaust system drives up the system cost. I like Glenn Olson's idea of "cheap" testing of subscale ramjets but the need to intergrate them works very hard against the "cheap" part.Something along the lines of the X7 test vehicle (wikipedia I know but, for general info: http://en.wikipedia.org/wiki/Lockheed_X-7) that's both robust but able to take a variety of engines to test.(Funny finding, again its wikipedia so take it with a grain of salt but I'd actually heard the "target-drone" version of the X7 was almost impossible to shoot down even with the advanced systems of the day. http://en.wikipedia.org/wiki/AQM-60_Kingfisher)But probably something more along the lines of the "ASTROX" RBCC booster (http://forum.nasaspaceflight.com/index.php?topic=22610.0) or the more "sugar-scoop" inward-turning design so that the exact nature of the "active" inlet and exhaust systems are less of an overall design issue?Might be "nice" to have take-off and landing capability but in general I'd settle for the "job" being the main point of the design so "flight" rather than being able to land and take off from a certain point.
Finding the actual speed limit of a conventional ramjet powered vehicle.Sorry, but the clarification must be asked... African or European?
He (Olson) mentions that there are enough old documents with graphs running up to M6.5 to suggest someone had flight tested something up to that speed. "Aero" suggested the thermodynamics of the fuel suggested M8.8 was possible (if you ignored drag, which would be a very silly idea). Checking the...
That's what I suspected. On the upside it demonstrates that ramjet designs at multiple point speeds and altitudes are viable.
The designs I suggested in the OP were just outlines. My starting view would be a "simplest possible" approach with fixed geometry, which got the French and the US designs to M5. My instinct is for a reusable design to allow (relatively) gradual expansion of the flight envelope and the ability to reuse the basic structure. OTOH a fully expendable design skips the landing gear and could be built more lightly. I think modern materials open up some interesting options. PICA demonstrates high temperature entry. There are also various ceramic materials that can offer light weight(ish) high temperature protection.
You seem to have cut off there
One of the reasons I mentioned the X7 was when looking at the "inward-turning" ASTROX design I noted the twin "horns" on the front an thougth "a duel spike X7" and imagined a very robust "recoverable" test vehicle using the X7 landing method
I want to see an actual flight vehicle to anchor some of those CFD predictions.
Quote from: john smith 19 on 08/27/2014 07:02 AMI want to see an actual flight vehicle to anchor some of those CFD predictions.No super-/hypersonic wind tunnel tests in between?
Quote from: R7 on 08/27/2014 09:55 AMQuote from: john smith 19 on 08/27/2014 07:02 AMI want to see an actual flight vehicle to anchor some of those CFD predictions.No super-/hypersonic wind tunnel tests in between?I'll quote 2 data points on that.1)The Navaho cruise missile was designed to fly with a pair of 40 000lb ramjets in cruise at M3 in the late 1950's. IOW the people of the time were confident enough of their ability to deploy this tech that they could fit it to a large operational vehicle and make it work.
I think wind tunnel tests up to about M3 should be used to test the separation dynamics but I think by then the test ramjet should also have ignited.
And lest we forget there was PLUTO (http://www.merkle.com/pluto/pluto.html)Unshielded nuclear reactors at 300ft AGL at Mach-3 for the win! (And by "win" I mean in the same sense anyone "wins" a game of Nuclear War )
Seperation of what? And where?
Quote from: RanulfC on 08/27/2014 05:49 PMSeperation of what? And where?One of the outlines in my OP was the idea of a 2 stage Parent/Child design. The parent being a small RPV designed specifically to carry the test vehicle (whatever it turned out to be) to launch height, accelerate to launch speed and release it.
The firebee supersonic drones showed that if your thurst is more like 50% of your GTOW rather than 30% powering through the sound barrier is not that difficult. What I have in mind is a small vehicle big enough to carry the main test vehicle whose design is driven by the requirements to make it easy to control (especially during the separation phase) with AFAP minimal danger of the test vehicle hitting any control surfaces.
Conceptually I'm thinking of a low supersonic White Knight 2, although I'd expect the actual layout to be very different. No crew should allow significantly larger payloads or smaller engines (as an aside does anyone make pure turbojets these days which aren't for expendable drones or missiles?) and only enough structural endurance to survive at M1+ (probably nearer M2) to get the ramjet firing and separated.
I'll note that WK2 has had quite a successful life hiring out to various people who want to drop test stuff off it. Perhaps VG are already in profit on their operations? Sadly I doubt the ramjet carrier would be so popular.
Oh sure, like "I" ever read what you actually wrote.. My bad I missed that. Beside one of the reasons I keep bringing up the ex-rocketman's blog is he's done a lot of work, (given it was his "real" job for a long time) on hybrid rocket-ramjets using solid boosters inside the ramjet that transition to ramjets engines using liquid fuel for the rest of the flight.
And my own idea was that the vehicle could be launched using a liquid/solid/hybrid booster vertically before switching to ramjet power.
The above mentioned Crossbow ALV isn't a bad idea for something of a similar nature:http://thehuwaldtfamily.org/jtrl/research/Space/Launch%20Vehicles/Air%20Launch/Air%20Launch%20To%20Orbit%20-%20ALTO%20-%20Crossbow-concept,%20MSFC.pdf
"Biggest" issue for seperation is the carrier vehicle having a postive lift factor after release, the more the better which argues towards something like a box or joined wing. Since the ramjet is just running flat-out (pun-intended) there's no need for fancy manuevers at seperation.
Subsonic would require some sort of booster to get the ramjet up to speed I'm thinking. (And "yes" technically the ONLY folks that make small jet engines these days make "pure" turbojets but they're pretty much all centrifugal compressors and turbines coupled with one or more "fan" blades for added cruise efficiency. The problem is they are all pretty small and what you'd need comes closer to a small fighter engine like a J85 or some such: http://en.wikipedia.org/wiki/General_Electric_J85. However I'll point out there are a number of studies on usign the "standard" cruise missile type engines, the F107/F112 types, http://en.wikipedia.org/wiki/Williams_F107/http://en.wikipedia.org/wiki/Williams_F112, and fitting them with afterburners to get supersonic performance out of them similar to early centrifugal turbojets)
Depends on what people want to use it for The ALTO/Crossbow study mentioned above was partially refined and driven by the idea of an RPV space launch vehicle after all according to some folks here
Quote from: john smith 19 on 08/16/2014 11:28 AMMaybe Aero's claim that basic theory will get you to 3Km/s (about M8.8 ) is correct but given the results of the X30 programme what I'm suggesting is a flight programme.I think I noted that Glenn Olson (of the old alt-accel website) had spoken to enough ex-ramejt engineers to come away pretty confident that a well-designed subsonic combustion ramjet could reach speeds in a bit excess of Mach-8 and for the most part (unlike many of the folks riding "theory" till it auguered into the ground in the form of the SCramjet ) couldn't see many "good" reasons to go faster even if most of them thought Mach-10 was possible given the right propellant and design And really what DOES air-breathing to @Mach-5+ get you if its "cheap" and "easy" enough over the alternatives?(I'd suggest hitting up the "Ex-Rocketman's Take" blog to see some of the work he's done as one of "those" engineers )Randy
The concept of Scramjet is over hyped over ramjets, on a false compare between engines with very different levels of construction technology.
On the side of low cost, French STX got to M4.9 with simple and cheap stainless tubes, also compatible to standard balloon tank rocket airframe.Going above M5 need to increase cost on airframe, M8 needs heavy TPS or regen cooling of airframe.So that for launch vehicles it maybe wise to cutoff at M5, similar to Skylon.
Make it easy. Assume body drag = 0, prop is kero, mdot (air)=1 reaches stagnation pressure within the engine, then thrust = ram drag = V*mdot (air) andthrust = ram drag = (mdot(air)+mdot(kero))*Ve andE = thrust*Ve/2 = 1/2 * total mdot *Ve2How much kero do you want to burn? Well, you have about 0.2 O2 and if you use a oxygen/fuel mix of, I don't know, 2.2 maybe (for Pc = 25 bar) gives about 0.09 kero. Choose your units, then what is your energy, in Joules = (watt/sec)? Plug in E and solve for ram drag which gives you velocity. But be sure to observe that Ve /= V.Oh, I looked it up for you. The heating value of kerosene is 43.1 MJ/kg.The big problem with this is that drag = 0 (excepting ram drag) is not a valid assumption so you won't get the ~3 km/s velocity this approach produces. Design your engine with a nice low frontal area and a good supersonic coefficient of drag, then extend my little model to account for that drag. Of course now you have to add ram and body drag together to equal thrust.
Quote from: 93143 on 08/14/2014 06:00 PMQuite a bit. I could probably figure out a first-order estimate by hand.I don't think anyone in their right mind would design a modern flying vehicle without CFD to begin with to converge on roughly what it should look like. One really interesting result of the RAND reviews of the X30 programme was the little factoid that if the CFD model data point count is >= R ^9/4 (R being the vehicle Reynolds number) you don't need a turbulence model and you can get the results directly.In an era of GPU's offering power in 1x10^12 FLOPS capacity I wonder if the time has come (at least for small vehicles, which this is meant to be) to brute force the problem? What I'm not so sure about is the number of equations at each point you'd be dealing with and the memory they'd need to store the parameters.Historically we know of at least 2 designs (the US ASALM and the French M5 ramjet programme) that got to M5+ with fixed geometries [EDIT and subsonic combustion]Maybe Aero's claim that basic theory will get you to 3Km/s (about M8.8 ) is correct but given the results of the X30 programme what I'm suggesting is a flight programme. There just aren't that many (flight) data points for any design going through this range. [EDIT a page on the Talos missile programme http://www.okieboat.com/Ramjet%20history.htmlstated that later versions switched to what appears to be hydrogenated dimerised JP10 (tetra cylopentadiene). This increased the energy available per (US) gallon of fuel from IIRC 126 to 143 thousand BTU's, about a 13% increase Called RJ-4]Note that while I'd prefer to keep things as simple as possible and with most geometry fixed to side step problems with high temperature actuators I think active skin cooling may be a reasonable trade off to lower weight if necessary. However since the goal is research, not an operational vehicle flying a constant dynamic pressure trajectory (IE climbing, not accelerating at a fixed altitude) is the way to go. I also suspect sub sonic ramjet combustor design has been a rather neglected field and CFD has been not really been used on it.If I'm right this is an area where substantial improvements should be possible. While the inlet has to handle the full Mach range of the incoming air stream by the time it hits the combustion section things should be much more constant (after all the goal is to what can be done below M1). A key element here seems to be the "flame holder." It's been suggested that good designs retain certain combustion products in their lee which are critical to continued stable effective combustion. A similar comment has been made about the injector designs in rocket engines. I'm guessing bad designs end up in combustion cycles of flame outs, fuel accumulation, ignition, resulting in either a pulsing thrust like a series of car engine mis fires or a low efficiency pulse detonation engine if (somehow) you've hit the conditions for deflagration to detonation transition . I think the latter is unlikely but not entirely impossible. This suggests 2 things. 1) An effective CFD based design must incorporate both the aero thermodynamic and thermochemical aspects of the problem (including compressibility) and 2) Most o f the action will take place below M1. That much lower Mach range (and hence Reynolds numbers) coupled with much more powerful modern processing hardware, should make the problem small enough to fit into a relatively small processing environment. [EDITIt seems the key task of the flame holder is to ensure complete mixing and combustion of the fuel by slowing down it's passage through the duct by increasing turbulence, giving enough time to complete combustion. In the Talos design this is done by igniting the fuel/air mix and flowing it inward through a truncated perforated metal plate. My guess is that the size and spacing of those holes can make a significant difference in ramjet performance but IRL (around 1948-1953) they were chosen either by what materials were available or by a cut-and-try process that stopped when they got a result that was good enough to deliver the performance target. The challenge is to deliver enough delay without increasing the interior drag too much. AFAIK no ramjet flight vehicles have been built that used the current generation of non invasive diagnostics to measure temperature, pressure and combustion chemistry within the combustor to gather data for a more optimized design.
Quote from: Katana on 01/30/2017 11:45 PMThe concept of Scramjet is over hyped over ramjets, on a false compare between engines with very different levels of construction technology.Agreed. QuoteOn the side of low cost, French STX got to M4.9 with simple and cheap stainless tubes, also compatible to standard balloon tank rocket airframe.Going above M5 need to increase cost on airframe, M8 needs heavy TPS or regen cooling of airframe.So that for launch vehicles it maybe wise to cutoff at M5, similar to Skylon.There are 2 parameters. The absolute top speed a (subsonic combustion) ramjet can reach and the operating range it can do.AFAIK the usual rules of thumb for ramjets have been a) No more than 3 Mach numbers of acceleration and b) Slow the flow inside the combustor to about M0.5These seemed to have been arrived at as empirical rules, not absolute limits.So I wondered if you you slowed the flow down less how fast could you go EG M0.9 in the combustor. And at what point do the subsonic design rules break down (I'm pretty sure it's below M1 in the combustor,, but what is it?) Likewise if you have different injector sets within the combustor, or avoid a conventional flame holder, and use "thermal choking" to create "virtual" variable geometry withing the ramjet, how high could you go?On the materials side you could have relatively cheap spray coated ceramic layers inside to handle the heat, going on up to full superalloys. The driver for this BTW was the notion of a "launch assist" architecture, where every Mach number you can add to the operating Mach range means a larger payload for the terminal rocket powered stage. Increasing the Mach range from 3 (IE M2.2 to 5.2) to say 5 (M2.2 to M7.2) would definitely be worthwhile. Hence my question what if you ran an X programme just to find those limits.
Four missiles reached a flight speed of nearly 4,600 feet per second, i.e., about Mach 5, between the altitudes of 40,000 to 115,000 feet.
When flight mach nuber approaches upper limit, subsonic mach number inside combustor reduces (behind stronger shockwave) to ~0.15. Engine performance and stability becomes more forgiving to flameholder, while inlet and nozzle supersonic losses and shockwawe stability go critical and require good CFD.
Talos style flameholder can't survive heat of M5, French STX used a simple fingured shape similar to early cobra burner test vehicles in 1946 before Talos.
The temperature rise during inlet compression is restored during nozzle expansion. So the flight velocity may exceed 3km/s If effective energy release is greater than drag loss.
However with practical total thermal efficiency less than 50% (already better than best car engines), flight velocity is generally limited below 3km/s.At high speed, subsonic combustion ramjets suffer from high temperature chemical disintergration of air (N2 and O2 to N, O ,NO). If recombination is not complete during nozzle expansion, efficiency drop.
Scramjets reduce temperature by only partly convert kinetic energy to heat. But low efficiency of mixing (of air and fuel) at supersonic conditions becomes a bigger problem, long and thin flow path is required, with large drag loss.The timescale of macroscopic mixing is much larger than chemical recombination of air molrcules. So the problem of air disintergration is easier than the problem of fuel mixing.Low efficiency of supersonic mixing (slow growth rate of supersonic shear layer) is fully understood until CFD tools of late 1990s made X-43 possible (after failure of X-30). Already too late that Scramjets attracted too much attention over ramjets.
M0.9 in combustor goes to M1 thermal choking after little amount of heat addition (either subsonic flow or supersonic flow goes near to M1 with heat added), and you can't add heat anymore. The velocity inside combustor is inverse related to velocity outside. If the engine starts at M2.5 with M0.3 inside, when flying to top speed of M5, the velocity inside falls to ~M0.15.
For thermal choking to occur with suitable amount of fuel, Internal velocity should lies between M0.1 to M0.5, normally M0.1 to M0.3Relatively higher internal velocity M0.3~M0.5 at startup may widen the total operating range. Need to startup with lean fuel and have inlet and nozzle designed for a range of conditions instead of fixed design point.
For launch assist applications, reduce startup velocity may be more benificial, for low boost fuel consumption and low TPS weight. Maybe M1.5 to M5.5, even M0.8 to M4.
There has been some comments about a "dump combustor" that seems to have good properties in this area.
Which I presume is where we enter SCramjet territory.
I think it's usually referred to as dissociation. The NN and OO bonds are broken but the atoms are intact.Do you mean "was not fully understood" before CFD in the late 90's? I can certainly believe that.
That's a key factor that will set a ceiling on maximum Mach in a design. How high can you start, how low can you go.
Ideally a fixed geometry design is best but beyond that a moving spike inlet seems mechanically the next simplest, although having different sets of fuel injectors would need multiple valves to modulate the fuel flow and where the fuel, and hence heat, is injected.
M0.8-0.9 put it in in range of conventional turbofans but that would not give you any new capabilities for an application. The original concept that sparked this was a Launch Assist Platform carrying turbojets or pre cooled turbofans to get to M2.2 (IE Concorde speed) then RP1 (allowing common tankage) fueled ramjets mounted on the wing tips of a winged rocket stage. At M5.2 the ramjets are jettisoned for later recovery and reuse. The goal was a fully reusable design with all stages operating over speed ranges for which there is solid (decades in the case of Concorde, but also various bombers and fighters) operating experience. This thread however was to see what you could get as a pure X programme, where there is no goal except go as fast as possible or go over as many Mach numbers as possible. Widening the operating Mach range of the ramjet component seemed the best simple way to raise payload on the final stage, by cutting the delta V it has to supply. Lowering the starting velocity to M0.8-0.9 would widen the range of available jet engines but give you no increase on the top speed, which is what this concept would need to either lighten the final stage or raise the payload it could carry to orbit.
Pulse detonation jets can be more efficient than other jets as a higher compression ratio can be obtained. I dont know if they are suitable for such high speeds. In June 2008, the Defense Advanced Research Projects Agency (DARPA) unveiled Blackswift, which was intended to use this technology to reach speeds of up to Mach 6.https://en.wikipedia.org/wiki/Pulse_detonation_engine
Quote from: colbourne on 02/02/2017 04:27 AMPulse detonation jets can be more efficient than other jets as a higher compression ratio can be obtained. I dont know if they are suitable for such high speeds. In June 2008, the Defense Advanced Research Projects Agency (DARPA) unveiled Blackswift, which was intended to use this technology to reach speeds of up to Mach 6.https://en.wikipedia.org/wiki/Pulse_detonation_engineNo practical PDE exist up to now.The best flight one: long-EZ test platform with PDE modified from automobile engine front endhttp://www.airliners.net/photo/Untitled/Rutan-61-Long-EZ-PDE/4179773
Quote from: Katana on 02/02/2017 05:15 AMQuote from: colbourne on 02/02/2017 04:27 AMPulse detonation jets can be more efficient than other jets as a higher compression ratio can be obtained. I dont know if they are suitable for such high speeds. In June 2008, the Defense Advanced Research Projects Agency (DARPA) unveiled Blackswift, which was intended to use this technology to reach speeds of up to Mach 6.https://en.wikipedia.org/wiki/Pulse_detonation_engineNo practical PDE exist up to now.The best flight one: long-EZ test platform with PDE modified from automobile engine front endhttp://www.airliners.net/photo/Untitled/Rutan-61-Long-EZ-PDE/4179773I think the impressive thing about the PDE demonstration was how it was relatively easy to get it working.
The point is that pulse detonation engines are potentially more efficient. They might not be easy but the future savings in fuel costs will provide an incentive for research.
Quote from: RanulfC on 08/19/2014 10:21 PM>snipping my quote <Ramjets could reach speeds of Scramjets, restrictions on fuel energy density to max velocity of 3km/s are basically same. Main difference is ramjet burns hotter, Scramjets has low mixing efficiency.
>snipping my quote <
On the side of low cost, French STX got to M4.9 with simple and cheap stainless tubes, also compatible to standard balloon tank rocket airframe.
Going above M5 need to increase cost on airframe, M8 needs heavy TPS or regen cooling of airframe.So that for launch vehicles it maybe wise to cutoff at M5, similar to Skylon.
trying to help here. The Lockheed X-7 reached Mach 4.31 in 1960 with the same ramjets as the BOMARC (XRJ-43 Marquardt) The ONERA Stataltex may have broken that record and flew at Mach 5 but it remains unclear. http://xplanes.free.fr/stato/stato-17.htmlhttps://archive.org/stream/nasa_techdoc_19670008070/19670008070_djvu.txt
what is sure is that Mach 5 - Mach 5.5 remained unbroken until the 90's at least, and then it was "supersonic combustion" (Kholod, HyFly and on)
Quote from: john smith 19 on 01/31/2017 05:44 PMM0.8-0.9 put it in in range of conventional turbofans but that would not give you any new capabilities for an application. The original concept that sparked this was a Launch Assist Platform carrying turbojets or pre cooled turbofans to get to M2.2 (IE Concorde speed) then RP1 (allowing common tankage) fueled ramjets mounted on the wing tips of a winged rocket stage. At M5.2 the ramjets are jettisoned for later recovery and reuse. The goal was a fully reusable design with all stages operating over speed ranges for which there is solid (decades in the case of Concorde, but also various bombers and fighters) operating experience. This thread however was to see what you could get as a pure X programme, where there is no goal except go as fast as possible or go over as many Mach numbers as possible. Widening the operating Mach range of the ramjet component seemed the best simple way to raise payload on the final stage, by cutting the delta V it has to supply. Lowering the starting velocity to M0.8-0.9 would widen the range of available jet engines but give you no increase on the top speed, which is what this concept would need to either lighten the final stage or raise the payload it could carry to orbit.Using supersonic carrier aircraft (F-15?) is generally too expensive for either budget of small test program (private funded or SBIR) or meaningful satellite launcher (size of Concorde needed). For a small test program, French Stataltex launch from ground and reach maximum speed with minimum cost up to now, though boosted to M3 with solid rocket.For practical launcher, air-breathing propulsion with VTVL architecture is optimal. Low minimum speed allows for direct rocket boost without turbojet/turbofan engines. At external speed M0.8 the internal speed could be M0.3, rising to M0.5 when multiple shockwaves establish at M1.5, and drop to M0.15 at near M5.
Actual the ORIGINAL, orginal idea I think was a concept that expanded on a multi-engine turbojet launch assist system based on an original concept by Dani Edar when he was working at Boeing:http://yarchive.net/space/launchers/jet_first_stage.htmlhttp://forum.nasaspaceflight.com/index.php?topic=25095.0https://en.wikibooks.org/wiki/Space_Transport_and_Engineering_Methods/Human_Transport
Which used them in a VTVL configuration, again originally as booster pods and later as a "ring-wing" stage in and of itself to boost a TSTO configured vehicle to Mach-2 and 100,000ft. NASA-Dryden re-packaged (and patented)the concept in 2013 (https://www.nasa.gov/offices/ipp/centers/dfrc/technology/DRC-010-039-Ram-Booster.html) and added a ramjet second stage and a Centaur based third stage. My personal addition was the inclusion of Mass-Injection-Pre-Compressor-Cooling (MIPCC) to significantly increase the turbofan thrust though it also made it "possible" to raise the maximum Mach to around 4 I wasn't sure it would be worth the extra engineering for the LV though the Dryden concept uses the ramjet for Mach-2 through 4 acceleration so it seems like it would still work. As MIPCC LOX injection later in flight to stabilize the combustion chamber combustion and temperature it's baseline to operate to a bit above 100,000ft anyway so this cuts out the Dryden ramjet stage entirely. (Unless someone wants to use it to get from Mach-4 to higher speeds but again, air breathing past Mach-5 is problematical anyway...)
"Ex-Rockteman's" take was to avoid the use of turbojets at all and go with integral solid/liquid hybrid ramjet booster pods from zero-to-around-Mach-4 as per ASALM which are recovered down-range. I'll point out there's no reason you couldn't use hybrid solid/liquid ramjets in a 'stage' somewhat similar to the Dryden concept as well. (Again eliminating the turbofans entirely)
Of course the nice thing about the turbofan-stage is you get high-use/fast turn-around RTL-VTVL out of it where as down-range recovery requires getting the boosters back to the launch site for refurbishment/reuse. Edar/GW Johnson's pods didn't require much more than a couple of guys and a truck with a small crane, (assuming down-range land-landing which was predicted to be within a pretty small area, you can't of course recover turbofans in the ocean but you can the empty ramjets so it's similar in that respect) but the pods could be spread over the area and individual recovery would take some time. Recovering a large "stage" would of course require more infrastructure and personnel.Of course Edar, GW Johnson, and I are fine with VTVL architecture but a lot of people over the years have made convincing arguments for both VTHL and HTHL concepts as well so YMMV as per usual
PDW engines haven't been seen in the 'white' aerospace world because they are not as efficient or as easy to control as they were thought to be. Again, that's at least as far as the public record has it They were supposed to have a higher operating range than a 'standard' turbojet because they relied less on intake air but they were also supposed to be efficient 'ejector' engines which turned out to not be so true which called into question their actual utility as proposed. Like SCramjets they appear to have a lot of promise they haven't lived up to yet. (And "Blackswift" has been proposed with every type of advanced engine from PDW's to dual-mode-SCramjets and isn't even a real project as of yet so I wouldn't hold my breath )