Sunday, February 28, 2010

Preliminary Acceleration Margins for Baseline Pod

The baseline ramjet strap-on pod is illustrated in figure 1. It is a very clean cylindrical shape of length-to-diameter ratio near 8. The ramjet inlet is a translating-spike pitot type, all external compression, with a single 30o cone angle. Hidden in the air plenum are several necessary equipment items, and the wheel well for the steerable nose wheel. The pod carries its own fuel supply, and need not draw fuel from the core vehicle.

This design concept has evolved slightly from earlier concepts, mainly in terms of flight controls. The earlier wrap-around fins have been replaced by fixed fins, and dive brakes whose mechanisms are hidden in the ramjet nozzle recess. The lower pair of fins has replaceable skid tips, thus comprising the main landing gear.

Figure 1. – Baseline Ramjet Strap-On Pod Concept

The pivot wing is about 2/3 of the body length, and can support sea level stall speeds in the neighborhood of 120-140 mph. The design calls for fuel depletion and stage-off at maximum 100,000 feet Mach 6 conditions in near-vertical flight. The dive brakes are fully deployed for fastest deceleration to subsonic speeds, at which point the wing is deployed. Using the dive brakes differentially for steering forces, the pod then spirals down to a land landing, under remote control from the ground.

Estimates for component weights in this design are based upon the X-15 rocket plane, arguably the most inexpensively-reusable rocket vehicle in history. Its inert mass fraction was very close to the 40% used for this concept design. One can assume titanium structural members with Inconel-X skins and inlet duct hardware, and insulated steel fuel tankage located inside, away from the hot surfaces. Bladder fuel expulsion is assumed. This construction approach is adequate all the way up to Mach 6 without heat protection, as was shown by the X-15 program.

The ramjet combustor is ablatively-cooled with DC-93-104 silicone, and the integral boost propellant charge is case-bonded in place upon that insulation, using the bond technology developed for ASALM-PTV. The inlet port cover is frangible glass with insulation. The booster nozzle rides with an O-ring seal on the ramjet throat, held in place by finger struts slotted into the nozzle bell. These are released by an outward-opening clamp, similar to the SA-6 “Gainful” ejectable booster nozzle. A booster pressure sensor triggers clamp release, the shatter charge for the port cover, initiation of fuel flow, and a pyrophoric magnesium ignition flare.

The pod is modular, in the sense that the inlet section is separable from the fuel tank and duct section, in turn separable from the ramjet-and-booster chamber section. It is assumed that a freshly-insulated and propellant-loaded ramjet chamber is substituted for the fired unit. The fired chamber is, in turn, refurbished and reloaded, for assembly onto another vehicle later. Modular separability thus enhances access for refurbishment.

On a unit area basis (So = 1 square foot), the weight statement estimated for this design is as follows. The ejectable items are 5 lb for the booster nozzle and 2 lb for the port cover in the 13.54-inch diameter corresponding to So = 1 ft2. In the absence of better information, it is simply presumed that all of these items scale in proportion to So.

Wrjbo.......133 lb.......weight at ramjet burnout
Wf...........29 lb.......ramjet fuel weight (JP-5 density)
Wrjig.......159 lb.......weight at ramjet ignition
Wejecta.......7 weight of transition ejecta
Wbbo........166 lb.......weight at boost burnout
Wp..........184 lb.......weight of booster propellant
Wbign.......350 lb.......weight at booster ignition

Estimated booster performance is for a non-aluminized AP-HTPB propellant system that averages near 2000 psia chamber pressure. The nozzle is “perfectly expanded” at sea level, and under-expanded as it climbs. Average thrust per unit reference area is Fth/So = 5540 lb/ft2. Average Isp at sea level is 250.7 sec. The estimated exit area proportion Ae/So = 0.172, and the nozzle expansion area ratio is 15:1.

An actual trajectory estimate has not been made, in part because no core vehicle has been defined for this strap-on to assist. However, a constant-acceleration rough guide was calculated for vertical flight to Mach 6 at 100,000 feet on a standard day. The vehicle must average 5.5 gees to “hit” this endpoint condition. Ramjet takeover is presumed to be Mach 1.5, which for this rough trajectory, happens at about 7500 feet.

Ramjet engine sizing and performance mapping was done with RJ-5 (Shelldyne-H) dense synthetic fuel. However, figures should be quite similar with JP-5 (Jet-A) or RP-1, both petroleum-derived standard kerosenes. Standard kerosene density was used estimating vehicle and fuel tank dimensions. Takeover fuel flow rate per unit reference area was 1.67 lbm/sec-ft2. Peak flow flux was 1.75 lbm/sec-ft2, in the neighborhood of Mach 2.5 to 3.9, and 10,000 to 20,000 feet. Minimum flow rate flux was 0.11 lbm/sec-ft2, at the Mach 6, 100,000 feet condition. Thus fuel flow turndown ratio looks “reasonable”.

Using previous performance mapping results, plus a crude estimate of pod aerodynamic drag, the self-acceleration capability of the pod is then its thrust minus its drag (thrust margin), all divided by its weight. Net jet thrust was used, with all additive and spillage drag lumped into airframe drag. This self-acceleration figure was calculated for both the ramjet ignition and burnout weights, to “bound” the results. Calculations were made as a function of Mach number, parametric on several altitudes from 5000 to 100,000 feet.

One should note that this thrust margin-over-weight approach provides an estimate of level flight acceleration capability. For the ability to accelerate vertically, the calculated self-acceleration should exceed 1 gee. The following figures show that low altitude acceleration potential is quite good from Mach 1.5 to a little short of Mach 4. At the higher altitudes, there is at least some acceleration potential out to just over Mach 5. However, even the Mach 3 peak values of acceleration potential are inadequate for vertical acceleration at 80,000 and 100,000 feet. It is still capable of vertical acceleration at 60,000 feet to Mach 5, but not in the 5-gee range. For improved high-Mach number performance, the inlet spike may well need to be multi-angle, or even “semi-isentropic”.

Figure 2. -- Self-Acceleration Potential Near Ramjet Takeover

Figure 3. – Self-Acceleration Potential at 10,000 Feet

Figure 4. -- Self-Acceleration Potential at 20,000 Feet

Figure 5. -- Self-Acceleration Potential at 40,000 Feet

Figure 6. -- Self-Acceleration Potential at 60,000 Feet

Figure 7. -- Self-Acceleration Potential at 80,000 Feet

Figure 8. -- Self-Acceleration Potential at 100,000 Feet

Although this is neither a realistic design study nor a full trajectory calculation, four important conclusions can be drawn from these results.

First and foremost, this kind of high-impulse thrust assist actually looks very promising for launch applications of all kinds, both vertical stage rocket, and hypersonic carrier aircraft launch. In large part, the attractiveness of ramjet-assist derives from the integral booster technology employed in this study. Such booster technologies were unavailable in the 1950’s when launch technologies were first being developed, appearing for the first time in 1967 with the SA-6 “Gainful” missile system. Subsonic-combustion ramjet technology has been operational since about 1950, but was “dead weight” from launch to takeover without the availability of integral boosters. (Note also that there is no compelling reason why a liquid or hybrid integral booster could not be developed.)

Second, without a more sophisticated inlet pressure recovery design than the single-angle cone used here, attractive levels of vertical ramjet acceleration capability all the way up to Mach 6 speeds at high altitudes appears to be just barely out-of-reach. That is not to say that useful horizontal-flight capability at Mach 6 is unavailable, even with this fairly simple inlet design, for these numbers show positive thrust margins to about Mach 5.5 at 80,000 and 100,000 feet. Remember, the pod drag used here may be too high, as it corresponds to a gun projectile, and includes a lot of nose pressure drag and base drag contributions that would be missing with a flow-through nose-inlet engine. (That drag data came from Hoerner’s self-published “Fluid Dynamic Drag”, section 16 page 21, figure 30.) Ultimately, for vertical launch, Mach 5 may be a better “target” terminal speed, while Mach 6 is more feasible for horizontal flight of hypersonic carrier aircraft.

Third: thrust and drag decrease with altitude, while weights do not. There is an altitude above which the acceleration capability of the ramjet becomes unattractive, simply because the weights are too large. For the vertical launch assist, this altitude is lower, simply because higher acceleration levels are required (the carrier plane can essentially “cruise” horizontally). Given these results with these assumed numbers for drag and weights, that vertical launch limit may be nearer 70,000 feet than 100,000 feet. This can be seen in the peak thrust margins shown in the plots for 60,000, 80,000, and 100,000 feet. Remember, to accelerate at all vertically, this must exceed 1 gee, and to maintain attractive levels for this kind of launch, it should be nearer 5 gees. But, for the carrier plane scenario, all that is needed is a nonzero acceleration potential, available at 100,000 feet to Mach 5, even with the crude design concept explored here.

Fourth, the ramjet strap-on pod explored here could just as easily serve a hypersonic carrier plane as the vertical stage rocket. The only difference is flying a little higher and faster with the hypersonic carrier plane, and a little lower and slower with the vertical stage rocket. Speculatively speaking, appropriate design “targets” for pod ramjet burnout conditions might be around Mach 4.5 to 5 at 70,000 feet for vertical launch, and Mach 5.5 to 6 at 100,000 feet for carrier plane launch. These are only crude estimates, and could shift dramatically with better inlet designs and more appropriate drag and weight data.

Tuesday, February 23, 2010

snow-day driving

Today was another "snow day" in Central Texas. I wish folks would drive sanely in this stuff. It's not a lot different from driving sanely in the rain, just more extreme. Nobody around here drives sanely in the rain, either. Proof: local accident rates skyrocket whenever it rains.

Key: slow down, you need a lot more time to respond to trouble, and traction is bad.

Key: stay farther away from other vehicles, you need a lot more space to respond to troubles, and traction is bad.

Key: turn on your lights, all colors are "stealth" in poor visibility conditions, even reds and yellows.

Advice specifically for pickup drivers: reduce rear tire pressures in bad traction conditions unless you are fully loaded. "Softer" tires break away less. (I have a pickup, too. This really works!)

Key for icy bridges: slow way down BEFORE you reach the bridge. You have to see what's on it, and still have time to respond.

If the bridge really is icy, here is practical advice on how to move over it (in priority order):

(1) slow down under 30 mph (under 20 is even better),

(2) put your wheels in prior wheel ruts where the ice is thinner and softer, and

(3) cross the bridge in a straight line coasting (no power, no brakes).

This stuff worked for me driving without chains, in two 100-year-record Minnesota winters. I never, ever had a problem, much less an accident. That was a real education for a Texas flat-land boy, too!

However, I will admit that keeping my Texas plates as long as I could helped me stay farther away from all the other vehicles. (Everyone else was scared to death to come near me!)

Saturday, February 20, 2010

Ramjet strap-on pod point-performance mapping

These images are also posted primarily for my buddies. They are excerpted from the detailed performance maps I generated for the nose inlet/center duct ramjet strap-on pod, of two earlier posts.

I generated these maps with a series of Excel spreadsheet worksheets to calculate the engine balance correctly, in lieu of an actual performance-mapping code. This is the same ramjet strap-on pod design reported in earlier posts, operated at equivalence ratio 1 (stoichiometric mixture).

The images posted here are a subset of those generated in the actual mapping operation. I actually have data at 5000 feet, 10,000 feet, 20,000 feet, 40,000 feet, 60,000 feet, 80,000 feet, and 100,000 feet conditions. However, the subset posted here is sufficient to understand the results.

The first image is the thrust and impulse versus flight Mach number at 5000 feet. Like all, this was for standard day conditions. The selected fuel control law was constant equivalence ratio (ER) 1. It quite clearly shows a peak near Mach 4 for thrust potential, with severe falloff toward both Mach 1.5 and Mach 6.

These data were obtained for a pod frontal blockage area So = 1 square foot, and are thus, in effect, frontal thrust density, lb/sq.ft. The specific impulse is higher at the low speed end, and about half or less at Mach 6, although still substantially attractive. This curve shape is very close to the inlet pressure recovery vs Mach curve.

The following image is the same kind of data in the same format, at 20,000 feet. It shows the same story, except that frontal thrust is a bit lower. Impulse is about the same.

The following image is the same kind of data, for 60,000 feet. Thrust density is a lot lower, reflecting the thin air. Yet impulse is not a lot diferent. At Mach 1.5 only, the inlet unstarts by a small spillage percentage.

The following image is the same kind of data for 100,000 feet. Again, the thrust density is very low, reflecting the almost-vacuum conditions, while the impulse is still not a lot different. At 80,000 and 100,000 feet, I saw a rather flat thrust variation with Mach number. This allows a rather wide choice of terminal Mach number in the design for the 100,000 foot altitude. Again, there is a small inlet unstart at only Mach 1.5.

The solid propellant booster inside the combustor cavity features an AP-HTPB-aluminum composite propellant. I had to make a lot of assumptions, but I tried to be very conservative and way-underestimate what I could package into that space. I think the Isp estimate at sea level of 250.7 sec is pretty good, as is the effective frontal thrust density of about 5500 lb/sq.ft of So. The booster calculations I ran are based on an average operating pressure of 2000 psia, a burn time-to-hit of 8.3 sec, and a volumetric loading of about 60% in a wagon-wheel-type grain design. Expansion area ratio (15:1) was based on ideal expansion at sea level.

These data were calculated at odd times and off-hours over the course of a few weeks. Sequentially, they correspond to about 2-3 man-days of heavy-duty engineering, and in more than just one specialty. I hope you find them interesting and useful.

It was a lot of fun discovering that I really can still do this.

ramjet strap-on pod concept

These images have to do with my ramjet engine sizing for a ramjet-assist strap-on pod for space launch. This post is also for the benefit of my buddies.

This first image is the concept behind an integral rocket-ramjet engine pod that is supposed to be a fly-back, reusable unit. I'm still learning how to get images embedded where I want them in this text. Tough for an old slide rule guy.

For the first few seconds off the pad, a solid propellant booster housed within the ramjet chamber provides considerable thrust, until vehicle flight speed reaches somewhere about Mach 1.5 or so. Depending upon the acceleration levels, that could be as low as 5000 feet, or substantially higher for significantly lower accelerations.

At that takeover point, with suitable geometry changes, the pod transitions to ramjet thrust in a fraction of a second. The pod continues operating as a ramjet up to approximately Mach 6 at 100,000 feet, at least as maxima. The geometry changes are an ejectable nested booster nozzle, and an inlet port cover. These are similar to those used on SA-6 "Gainful", ASALM-PTV, ALVRJ, and more recently, the SS-N-22 "Sunburn".

I had an old DOS Basic language code I wrote some 15 years ago that I made run again on an old Windows 98 / 486 machine from my shop. I have had no success trying to run this ancient DOS / Basic code in Windows ME or Windows XP machines.

This old code was a ramjet engine sizing code. I used it to size the ramjet pod's engine for a takeover point of Mach 1.5 on a standard day at 5000 feet. I used equivalence ratio 1.1 (110% of ideal mixture strength) at inlet pressure margin 2% (to cover manufacturing tolerances). That result is depicted as follows:

As you can see, the pod features an air plenum beghind the inlet structure, feeding a center coaxial duct, in turn feeding a sudden-dump entry into the combustor. There are some very necessary equipment items packaged within the air plenum space. The sudden dump is the flameholder, very similar to the old ASALM-PTV. I assumed RJ-5 (a.k.a. Shelldyne-H, a synthetic substitute for kerosene that is denser than water), although the basic sizing would be just about right for JP-5 jet fuel, or RP-1 rocket fuel, both petroleum-derived kerosenes.

This pod features a pivoting wing, wrap-around fins, and retractable landing gear, for recovery and re-use. I'm thinking inert mass fraction about 40%, like the X-15, to ensure the structural robustness for 1000's, if not tens-of-1000's, of flights. The X-15 was the most inexpensively reusable rocket vehicle in all of history, after all. Depending upon whose data you believe, its inert mass fraction was about 40%, within about 2-3%.

Not shown are dive brakes for deceleration after staging. I may switch to fixed fins, and use the dive brake panels differentially, for steering control. Not sure yet.

In any event, the idea is to fly this thing back by remote control from the ground, like a big model airplane. It has a nosewheel and skids, like the X-15. That ought to provide sufficient capability for a runway recovery adjacent to the launch site. For the translating-spike inlet design, I would prefer a land landing, over ditching and towback at sea. Less risky.

I do not yet have a performance-mapping code here at home (or anywhere else, for that matter). I am attempting to re-write one from that old DOS-Basic sizing code, but I am nowhere near done. Meanwhile, I have been programming the basic engine balance into an extensive sequence of Excel-spreadsheet worksheets.

Using a constant 5.5 gee vertical acceleration as a rough guide to a fast ascent trajectory that reaches Mach 6 at 100,000 feet, I picked off several Mach-altitude points and ran point performance in the spreadsheet at a "sweep" of fuel/air equivalence ratios (relative richness, 1 = perfect mixture, under 1 is lean, over 1 is rich). All the plots looked exactly like this one at takeover, only the numerics were different. This one went from ER = 0.4 to ER = 1.5 by 0.1's. Rich is at lower right, lean is at upper left.

I did a manual graphical "aft tangent" analysis on each one of these plots. That effort determined the "best" compromise between impulse and thrust for a vertical accelerator engine to be equivalence ratio 1 (stoichiometric, or perfect mixture). That's quite different from the "best" ramjet cruising missile design, which is max Isp consistent with a thrust coefficient matching the drag coefficient for the design cruising speed, typically around ER = 0.7.

I am not yet at all sure that the nose inlet / center duct design approach is the best for this application. Because of severe aeroheating, even for a short transient, at Mach 6 conditions, this geometry offers "coking" potential for fuel residuals on the tank surfaces, both outer, and especially inner along the center duct. This is hard to avoid, and very much harder to make easily refurbishable and reusable.

Perhaps a side-mounted multiple-inlet geometry similar to SA-6 or ALVRJ would be better. At least the hot ducts would be outboard of any exterior tank insulation. However, inlet pressure and massflow recovery characteristics are a little different from what I used here. Although in the same basic ballpark as what I used, they are a tad lower, and far more sensitive to off-angle attitudes.

I just dunno. So, I will push this center duct "design" a few steps further before I give up. Watch this space for further details.

Inlet data for ramjet strap-on pod

This post is really for my on-line buddies at It is one of multiples to come that I hope they can use. For an old ramjet engineer like myself, this has been a most cathartic experience.

These images are from an inlet estimate that I did recently, for the concept design of a ramjet strap-on pod for space launch. I used the old 1953 publication NACA 1135 as the source (something most of us old compressible flow people acquired rather early in our careers).

The design is for a translating-spike, all external compression, axisymmetric nose inlet. Specifically, it is for angle-of-attack 0 degrees, but for axisymmetric nose inlets, performance is essentially invariant out to about +/- 15 degrees off-angle.

Definitions are:

The basic inlet recovery and spike position data are:

Inlet-related drag data are:
It was a lot of fun doing this. I haven't done this kind of thing for over 15 years. Nice to know I still can. This was all by-hand, pencil-and-paper work. I put it into an Excel spreadsheet for graphing. The graphics program I used for pictorials, and for data transfer to this site, was Microsoft's Windows accessory "Paint".