Tuesday, October 28, 2014

Not Nominal

A short while ago, an Antares rocket launched from Wallops Island exploded shortly after takeoff. The launch was supposed to be an ISS resupply mission, and had been scheduled to launch yesterday, before being canceled as a boat intruded into the downrange safety area.

At this point, we don't really know what caused this. It will probably be months before we do. Still, I'm going to engage in a bit of idle speculation.

From the video, it looks like the rocket broke apart very suddenly and spectacularly, and didn't deviate much from its course beforehand. Since it detonated extremely close to the pad, it's almost certain that this wasn't an activation of the range safety system, which would be used to destroy the rocket if it had suffered a failure in the guidance system.

Compare the Antares explosion to a recent failed Proton launch:

This article states that the crash was caused by a shutdown of one of the booster engines. The Antares rocket has two engines, compared to the Proton's 6, so in the case of an unplanned shutdown, it would have fallen back to the pad almost vertically. That it exploded in midair suggests that something caused the propellant tanks to rupture and ignite. In my opinion, the most likely thing to cause this would be some sort of overpressure event in the engine. Numerous faults in the engine could cause such an event, and I don't know enough to even make a guess at what it could be.

The Antares rocket uses a single AJ-26 engine in the first stage. This engine is better known as the NK-33, a development of the NK-15 engine used on the Soviet N1 lunar rocket. These engines were all built during the 1960s and 1970s, and subsequently refurbished by Aerojet for use in the Antares rocket. I'm certain that Aerojet has very robust quality control and inspection procedures, but it's not impossible that some microscopic defect could have slipped by. Additionally, until it was surpassed by a variant of SpaceX's Merlin engine, the NK-33 was the kerosene/liquid oxygen fueled engine with the highest thrust to weight ratio in the world; roughly 137:1 (the RD-275M had a slightly better thrust to weight ratio, but used UDMH/N2O4 propellants). That an engine with such performance was developed in the 1960s is a testament to the skill of Soviet engineers. It also suggests to me that the engine was very 'hot', basically squeezing everything possible out of the design with little room for error (I've also heard this about the RS-25). While this would allow higher performance, it (in theory) means that the engine could be more prone to failure. I'm not going to go so far as to say that this is definitively the cause of the launch failure, but it's my guess at this moment.

Sunday, October 26, 2014

Ravaged Skies

 Found this quite good track (Ravaged Skies, by Mitch Murder) while browsing SA recently. Good music, and this video gets bonus points for using footage from 'The Right Stuff', an excellent movie which everyone and their children should watch. The Original Music Video Here, which is slightly different.

A minor historical note, while 'The Right Stuff' depicts Yeager's flight (and subsequent accident) in the NF-104 as being an unauthorized joyride, it in fact took place as part of a regularly scheduled test mission. However, according to this summary, the accident was very much Yeager's fault.

Thursday, October 23, 2014

Soviet/Russian Spaceplane Concepts

During the Cold War, both the US and Soviet Union developed numerous concepts for reusable spaceplanes. Many of these concepts, such as the X-20, MiG-105, STS, and Buran, are well known. While flipping through my copy of Unflown Wings, I happened across a couple of other, lesser known spaceplane concepts developed in the Soviet Union and Russia during the past few decades. To my knowledge, none of these ever flew, but there's some interesting designs that were drawn up.

(credit for all pictures goes to Yefim Gordon and Sergey Komissarov)

 The first of these is the Tu-136, also known as Zvezda, which was developed during the late 1960s. As is apparent from the picture, it appears to be roughly analogous to the X-20, although there are some notable design differences. Like the X-20 it would have been vertically launched on top of an expendable launch vehicle, most likely an existing Soviet booster. Based on the weight given for the vehicle (7-9 tons), it is likely that an early version of the Proton rocket would have been the launch vehicle for the Tu-136. Like the X-20, the Tu-136 would have been intended for military usage in LEO, although the exact nature of these missions is unknown.
 Interestingly, it appears that this design was intended to be able to operate in both manned and unmanned modes.

The next design was the Tu-2000. This was not a single design, but rather a series of several conceptual spaceplane designs.  Unlike the Tu-136, they would have been launched horizontally, either from a runway or a purpose built (and truly mammoth) carrier aircraft. Given that the Tu-2000 was to function as an SSTO spaceplane, massive amounts of LH2/LOX fuel would have been needed. Some of the concepts would have had takeoff weights of almost 300 tons. Total orbital payload was projected to be in the 5-10 ton range.

In Unflown Wings, it is mentioned that the Tu-2000 would have featured up to six turbojet engines (in addition to liquid fuel rocket engines), in order to improve in-atmosphere performance. Interestingly, it is also mentioned that use of a nuclear rocket engine was considered, due to its higher specific impulse compared to chemical engines. This is probably the first time I have heard of a nuclear engine being proposed for such an application.

The MiG AKS (AKS roughly translates as 'aerospace system') was a proposal for a reusable spaceplane dating to the 1980s/1990s. As can be seen from the picture, it would have been launched from a high speed 'mothership' aircraft, likely travelling at hypersonic speeds. The mothership was planned to operate on conventional jet fuel at low speeds, and LH2 at hypersonic speeds. While this would have complicated logistics greatly, it was believed that the benefits of an air launch (reduced delta-v, more flexibile launch site) would have been worth the disadvantages.
There is also brief mention in Unflown Wings of a MiG designed spaceplane named 'Oryol'. This was intended to be ground launched, with one proposal having it launched by an electromagnetic mass driver. However, it was seen as less technically feasible than the AKS, and was not developed as far.

 One interesting design from the 1990s is the S-XXI, developed by the Myasishchev design bureau. Unlike most other Soviet/Russian concepts, which were intended to be fully capable of reaching orbit, and used for military missions, the S-XXI would have been suborbital, and used for space tourism. Launched from the back of a modified M-55 carrier aircraft, the S-XXI would have reached an altitude of just over 100 km on a ballistic trajectory, allowing the passengers to briefly experience weightlessness. Though it was intended to have flown in 2005, it was ultimately cancelled due to lack of funding.

Saturday, October 18, 2014

Advanced Cryogenic Expendable SSTO: The Future?

Research into Single Stage to Orbit (SSTO) rocket designs has been ongoing for decades. During that time, numerous designs, such as the X-33, DC-X, and Roton, promised to deliver exceptional performance, but failed. However, they all failed, due to technology issues, lack of funding, an inherently flawed design, some combination of the above, or other variant problems. Like fusion power, I have heard it said that SSTO rockets are always a few years away.

I recently acquired a paper (which is unfortunately behind a paywall) entitled "ACE: Practical SSTO" (written by Paul W. Gloyer, Tim S. Lewis and Zachary R. Taylor), which claims to present a workable design for an SSTO launch vehicle. Given that engineers at Lockheed Martin, McDonnell Douglas, and other companies probably claimed the same thing about their designs, I'm a bit skeptical. Still, after reading the paper, I'm a bit more optimistic, though not entirely certain that their design will revolutionize spaceflight.

Many SSTO designs wanted to achieve both SSTO performance and a reusable vehicle. There is some logic behind this; with a reusable SSTO design, the entire vehicle could be reused (unlike designs such as the X-37 or STS, where only a portion of the stack is recovered). The only costs would be fuel, and whatever (theoretically) minor repairs would have to be done to the vehicle between flights. However, reusability adds more complexity and mass to the vehicle, in the form of a heat shield, landing gear, and the like. Additionally, as demonstrated by STS, the amount of time and money needed for refurbishment between flights is nontrivial.

In the paper linked, ACE stands for Advanced Cryogenic Expendible. Since the design is not intended to be reusable, heat shields and similar parts are not required. Some might ask, then, what is the advantage of an SSTO over a conventional staged rocket if it cannot be reused? In the paper, primary advantages are identified as reduced complexity, and reduced operating costs due to lack of spent stages falling to Earth. While I am somewhat skeptical on how important the second point will be, the reduced complexity of a single-stage design is definitely a valid point. A single-stage, expendable vehicle has the potential to be a very simple design indeed, and anything which reduces the number of failure points is usually good.

A diagram of the vehicle described in the paper.

Another aspect of the SSTO described which differentiates it from most earlier SSTO designs is its relatively small size. While both the DC-X and X-33 were eventually planned to be developed into manned versions, the ACE is designed to only launch small satellites. The configuration in the paper is listed with a total weight of under 3500 kg (7500 lbm), and a payload of about 70 kg. Not developing a manned vehicle means that life support systems are not needed, and eliminates the need to satisfy NASA (or any other space agency's) requirements for man-rating a launch vehicle. The small size would also reduce support and transportation costs of such a launch vehicle. However, the small size does introduce new technical challenges; volume increases with the cube of size, but surface area increases quadratically. This means that larger launch vehicles have to devote a lesser portion of their mass to fuel tanks, typically the largest portion of a rocket's dry mass.

ACE dimensions, in Imperial units for whatever reason.

The team developing the ACE proposes to eliminate this problem by using advanced composite fuel tanks. Specifically, they plan to use a type of composites known as BHL. Those familiar with the X-33 program may recall how issues with composite fuel tanks were one of the issues that led to the scrapping of that program. However, given that the ACE is a geometrically simpler design (roughly cylindrical, compared to the X-33s shape), and the advancements in composites over the past decade, I think that using composite fuel tanks is a viable strategy. The ACE team plans to achieve quite low mass/volume ratios with composite tanks; on the order of .5 lb/ft^3 o (8 kg/m^3). Naturally, larger diameter tanks will be more efficient. Since larger tanks are required for LOX/LH2 (the only commonly used fuel which could provide reasonable SSTO performance), this could be a major breakthrough.

Comparison of BHL with aluminum-lithium metal tanks, which are some of the lightest metal tanks currently in use.

One interesting thing mentioned in the paper was the relatively large size of the vehicle (for its payload) in orbit. While the actual satellite payload might only be a few dozen kilograms in mass, there would still be several meters of vehicle remaining. As mentioned in the report, if covered in solar panels, there would be a very large amount of power available for the payload. It is also mentioned that scaling up the vehicle, and/or fitting an electric propulsion system for in-orbit use, could give the payload sufficient delta-V for transfer to GEO or beyond.

While the concept is, overall, very interesting, I do have a few minor nitpicks. For one, they assume use of a LOX/LH2 engine with a thrust/weight ratio of roughly 75. This is in excess of both the J-2 and RS-25 (I have heard that the latter was considered to be a very 'hot' engine). While a T/W ratio of 75 might be possible with new developments in rocket engine technology, I would prefer to see an actual test of such an engine, especially one of such a small size as the ACE would use, before making any conclusions.

Another one is the economic issues. While there may be a market for putting relatively small satellites into orbit, I am skeptical of whether the ACE would provide sufficient savings over simply launching them as a secondary payload on a larger rocket. Additionally, the STTO concept suffers from an inherent disadvantage compared to staged rockets; the entire rocket must be carried for the whole flight, while a staged rocket can simply dump useless mass (empty fuel tanks and associated structure) partway through the flight. How the ACE SSTO compares with a small two-stage design incorporating similar technologies would be an interesting thought exercise.

Despite these misgivings, I feel that the ACE SSTO concept has merit, and could potentially be quite good. At the very least, I hope that a couple test launches of the concept are made; that would put it far ahead of any of its peers to date.

Tuesday, October 14, 2014

A Brief Criticism of “A Rocket Drive For Long Range Bombers”

While going through my folder of pdfs that I have amassed on various topics, I came across a quite interesting one. Saved as “saenger.pdf”, it was in fact a translated copy of “A Rocket Drive For Long Range Bombers” (“Über einen Raketentrieb für Fernbomber”), by E. Sänger (Saenger) and J. Bredt, originally published in August 1944. 


Aviation and spaceflight enthusiasts will recognize Saenger’s name almost immediately from his work on the Silbervogel rocket-powered bomber. While the concept was originally formulated in the 1930s, the German military hierarchy did not become especially interested until late in the war, as the situation became increasingly desperate. Though work on the Silbervogel never progressed beyond a wind tunnel model and tests of a few individual components, quite a bit of theoretical work and calculations were done by Saenger in this paper.

I have provided a link to “A Rocket Drive For Long Range Bombers” here. Though I already had a passing familiarity with the Silbervogel concept, I found this paper to be quite interesting, and I learned very much from it. As it’s about 175 pages long, I won’t pick it apart line by line in this blog. However, I did get a few main impressions from it.

The first was the thoroughness of the work. Rather than simply stating that a certain material will be used, or a certain propellant combination will give a certain level of performance, Saenger and Bredt devote several pages to calculating values of various parameters. The downside of this is that it adds to the length of the paper several pages’ worth of equations and graphs. However, such work is necessary when exploring such an untested frontier as transatmospheric flight in the 1940s, and overall, I feel that the inclusion of so many calculations and data points was worth it.

(Only 30 pages in, and we’re already on figure 14).

Of minor annoyance was that various letters used for certain variables were different from those I have commonly encountered (for instance, using x for specific heat ratio instead of γ). While this was not a real problem, as there was an index of variables and parameters in the back of the report, it did force me to stop a few times. Not that any of this is Saenger’s fault by any means.

While Saenger’s report was very thorough, I do feel that it was somewhat optimistic in some areas. For instance, performance and flight paths are given for a rocket engine producing an exhaust velocity of 3,000, 4,000, and 5,000 m/s. The first two are within the typical range for engines using liquid oxygen as an oxidizer (which the Silbervogel was planned to do), though 4,000 m/s is honestly a bit high unless LH2 is being used, and the Silbervogel would not have used that fuel. 5,000 m/s is right out; it is beyond even LH2/LOX, and entering that realm of performance requires mucking about with fluorine (always a fun experience). However, there’s a bigger issue; the A4 rocket motor was only producing about 260 seconds of specific impulse at the time, equivalent to roughly 2,500 m/s of exhaust velocity. Assuming that a Silbervogel would have used a similar engine, its performance would have been reduced even further.

By far the most common flaw brought up in Saenger’s work is the heat flow calculations. Apparently, due to an error somewhere in the paper, the Silbervogel would have burned up on its first skip across the atmosphere due to higher than expected heat loads. While I’m not going to run the calculations at the moment, thermal loads at hypersonic speeds are certainly nontrivial. Additionally, at the time the Silbervogel was conceived, the hypersonic flight regime was very poorly understood. Compare, for instance, designs for reentry vehicles from the 1950s;

Considering this, it is likely that the Silbervogel design would have required significant refinement before actually flying. This might not have been insurmountable, but it would have required a lot of time and money, which Nazi Germany was not exactly flush with in 1944.

Despite all this, Saenger’s work is very technically sweet, and quite interesting for an aerospace nerd like me. However, I’m having trouble getting around a quite fundamental flaw in the Silbervogel concept.

Given that the spaceplane would have had insufficient delta-V to reach orbit, it would have had to land at an airbase elsewhere. At the time, the most likely candidate was various Japanese held territories, such as the Marshall Islands. This brings up the question of how one plans on getting the Silbervogel back to Germany. Launching it was planned to involve several A4 rocket motors, each of which would have required several tons of fuel, as well as a quite long launch track. All these components would somehow have to be transported to Japan through hostile waters; I severely doubt the ability of Japanese aerospace industry to natively manufacture the required components. Otherwise, the Silbervogels would be consigned to single use weapons; launching once, then spending the rest of the war rotting on a remote island.

Even more damning is the Silbervogel’s small payload. The rocket plane would have had a payload of at most 5 tons (larger payloads were discussed, but were not feasible without an exceptionally large vehicle). This is roughly comparable to Allied strategic bombers, which cost much less than the Silbervogel, and could be made of much greater quantities. Given that even thousands of bombers were insufficient to defeat Germany on their own, it is most unlikely that a few dozen rocket planes dropping 5 ton payloads on New York would force America out of the war. A nuclear weapon is one of the only viable payload with the Silbervogel’s mission profile, however, the poor state of the Nazi atomic program meant that no such weapon would be forthcoming.


While Saenger’s spaceplane was never built, the concept of a winged spaceplane lived on. Designs such as the X-15, X-20, MiG-105, and STS are all similar to the Silbervogel in some way, although it is doubtful that Saenger’s work was anything more than a secondary influence on these designs. Still, I personally find the Silbervogel, and “A Rocket Drive For Long Range Bombers” to be quite interesting, and worthy of discussion.

Thursday, October 9, 2014

Yes, the Aggregat 4 (and von Braun) was a Pretty Big Deal

Those of my readers with significant aerospace knowledge are probably slightly concerned by the title. Has this blog so quickly devolved into writing about obvious content? Will my next post be about how the sky is blue, or water is wet?
No, I will not be writing about the wetness of water or the blueness of the sky. However, in my recent travels on the internet, I came across a person who made the bold claim that “von Braun’s work was just scaled-up rip-offs of Goddard’s”. That said person, when pressed on this claim, chose to resort to insults rather than providing facts to back up their statement is a topic for another day.
The intent of this article is not to downplay Robert Goddard’s achievements in the field of rocketry. His work was an inspiration to aerospace engineers and scientists worldwide. Indeed, von Braun himself acknowledged Goddard’s achievements, saying in 1963; 

"In the first third of this century, interest was limited to a few lone-wolf scientists who were often labeled "crackpots." One such "crackpot," Dr. Robert H. Goddard, is now credited with being the first to fly a liquid rocket, complete with a "re generatively cooled" combustion system and a simple guidance system to keep it on course. Dr. Goddard, a truly great man, was a professor of physics at Clark University in Worcester, Mass. His rockets, which were flown starting in 1926, may have been rather crude by present-day standards, but they blazed the trail and incorporated many features used in our most modern rockets and space vehicles.”

Had Goddard received as much funding and resources as von Braun, he might well have been able to create a rocket like the A4. Alas, he did not, and such speculations will remain just that, speculations.
This post is also not intended to unnecessarily glorify Nazi engineering or practices. The usage of slave labor in the assembly of the V2 rockets was a horrendous atrocity. While the advances in aerospace technology caused by the A4/V2 are numerous, they were most certainly not worth the lives of thousands of innocent men. Also, to head off any wehraboos who might be reading this article, the A4/V2 was a major technical achievement, but useless as a weapon. The German state would have been better suited investing the time and resources spent on the V2 (according to some accounts more than the Manhattan Project cost) on simpler and more mundane projects with a greater chance of success. Not that it would have made a difference in the final outcome.

Now, onto the main topic. According to the Encyclopedia Astronautica, the largest rockets Goddard ever tested were the P-C series, which he built in 1939-1941. These rockets were fueled with liquid oxygen and gasoline; this propellant combination was a precursor of later kerolox rockets such as the Saturn V. This combination, when used in the P-C’s engine, resulted in just over 3 kilonewtons of thrust. With a total mass of roughly 220 kilograms, and a height of almost 7 meters, the P-C series were nothing to sneeze at. Certainly better than anything I could put together in my garage. However, let’s compare them to the A4.
The A4 was roughly 14 meters tall, twice the height of the P-C. Moreover, it had a mass of over 12,000 kilograms, more than an order of magnitude greater than Goddard’s rocket. Not only this, but the engine in the A4 produced nearly 100 times as much thrust as the one on the P-C, and used a different propellant combination (liquid oxygen and alcohol). 

If you are still unconvinced, here’s a visual comparison. First, a picture of the P-C, from Astronautix. 

Not the greatest picture, but it does show the internals clearly visible, as well as an indication of the relative size.

Now, the A4;


It is highly obvious that the two rockets are of completely different scale. Considering that the gulf between the A4 and Goddard’s earlier efforts is even larger, it is clear that the A4 is not simply one of Goddard’s rockets with the dimensions changed to metric (as one person on the internet claimed).
Of course, some might argue that the A4, and other rockets in the Aggregat series are simply scaled up versions of Goddard’s rockets. Since scaling up a rocket is trivial (according to these people), von Braun and friends didn’t actually do any new work. Ignoring that it is clear that the A4 is not a P-C (or any other of Goddard’s rockets) with the dimensions multiplied by some number, let’s take a look at whether scaling up a rocket to a bigger size is actually trivial.
During the 1950s and early 1960s, kerosene/liquid oxygen (kerolox) was a commonly used propellant combination in the American space program. Many rockets, including the Atlas ICBM and the Saturn I used kerolox engines in the first stage. However, the most famous usage of this propellant combination was the Saturn V, whose F-1 first stage engines were quite possibly the most powerful in the world (depending on whether you count the multi-nozzle RD-170).
By the time the Saturn V was developed, the US space program had several years of experience with kerolox engines. While none of them were as large as the F-1, it should have been trivial to scale an existing design up to that size, right? Wrong. The F-1 suffered from massive combustion instability issues, which took several years to resolve, and represented a major setback in the program. (More can be read about the development of the F-1 here; http://history.nasa.gov/monograph45.pdf).
In addition to issues with combustion instability in the F-1, the Saturn V had several other kinks that needed to be worked out during its development, including pogo oscillations that nearly resulted in the loss of the (fortunately unmanned) Apollo 6. Clearly, making a bigger rocket by simply scaling up existing designs is not trivial. This is due to the inherent nonlinearities and complexities in large systems such as space launch vehicles. In the 1960s, when computer simulation technology was in its infancy, it was impossible to predict exactly how a large system such as a Saturn V might behave (even today it is difficult). In the 1940s, when computer simulation technology was nearly nonexistent, and slide rules were the computational tool du jour, it would have been impossible.
What if it was argued that von Braun’s earlier work, such as the A1, was a copy of Goddard’s? This question is more interesting. Indeed, the rockets of the L-B series, which Goddard tested from 1930-1937, are comparable, or even larger than, von Braun’s efforts from the same period. Also, while Goddard’s L-B rockets flew successfully multiple times, the A1 was tested a single time, which ended in failure. But was the A1 merely a crude copy of the L-B, as has been claimed? Or was it a native, if somewhat inferior, design. Let’s compare pictures of the two.

This is a diagram of the A1.
And here’s the L-B:

As can be seen, the two rockets are distinct designs, with different proportions. While the A1 was undoubtedly influenced by Goddard’s work, it is a unique design, so far as I can tell. Of course, even if the A1 were a clone of a Goddard rocket (a dubious assertion), this would not discount von Braun’s later work.

In conclusion, Goddard’s contributions to rocketry and the aerospace sciences are numerous. They went criminally unrecognized, and had they received the attention they deserved, spaceflight could have been advanced several years. However, they do not diminish the magnitude of the accomplishments of later engineers, such as von Braun. While von Braun’s work may have been inspired and influenced by Goddard’s, being influenced by others is a natural part of science and engineering. Is Einstein’s work diminished because it was based on that of Maxwell and others? I think not.