I have compiled some data on the payload fraction (payload to LEO / Gross mass) of various rocket systems;
From this, several thing can be seen;
Solid rocket boosters utterly ruin your payload fraction. Despite having a significantly higher specific impulse than other engines (365 seconds for the RS-68 vs. 285 seconds for the RD-275), hydrogen-fueled launch systems only have a slightly better payload fraction than hypergolic systems, or are even significantly worse. Larger rockets generally have a larger payload fraction (Saturn I vs. Saturn V, Falcon 9 vs. Falcon Heavy). Titan II and Titan IV are not entirely comparable. STS is a stupid pile of trash. Kerolox first stage provide significantly better payload fractions in almost all cases, while avoiding the difficulties associated with liquid hydrogen. Hypergolics generally have inferior performance to both, but are significantly easier to handle, and the difference is not extreme.
Data via wiki, except where noted (the gross weights for Delta IV Heavy and Atlas V 551 were horribly off, especially for the latter). Encylopedia Astronautica data mostly agreed, but that site is severely lacking in info on the Falcon family.
Also posted here.
The history of American efforts to develop nuclear thermal rockets is relatively well known. Similar Soviet efforts have remained far more obscure. However, during the Cold War, the Soviet Union developed and tested an advanced nuclear thermal rocket engine, designated the RD-0410. Unfortunately, relatively little English-language information about the RD-0410 can be found (at least in easily available sources). Similar to the American NERVA program, development of Soviet nuclear rocketry began in the mid-1950s. Serious research began in 1955, with development of a rocket beginning in 1956 (the people working on this project included such notable people as Kurchatov, Keldysh, and Korolev). Initially, the Soviets planned to use the nuclear rocket to power an intercontinental ballistic missile, or possible a cruise missile. However, it was quickly realized that chemical rockets were good enough for suborbital flights. As a result, by the 1960s, it was decided to develop the engine for usage in space. The engine was developed by the KBKHA bureau, which had also developed engines such as the RD-0105 (used on some derivatives of the R-7). The goal was to develop an engine with a specific impulse of roughly 800-900 seconds, double what can be achieved with normal chemical rockets. Doing this would require creating a nuclear reactor that was both very light, and capable of withstanding very high temperatures around 3000 Kelvin. I have seen a few references to a program to develop a 2,000 isp engine, but this would require temperatures (over 15,000K) well in excess of what was possible in the 1950s (or even today) for a solid core design. The test site selected for the Soviet nuclear engine was Semipalatinsk in Kazakhstan, a remote location similar to Jackass Flats in Nevada. The Soviets had already tested numerous atomic weapons (including their first in 1949 there), so the place was no stranger to nuclear activity. It appears that tests of the engine were conducted in a mine shaft approximately 150 meters deep, unlike the American NERVA, which was tested aboveground. Most likely, this was due to concerns over radiation should the engine malfunction. At some point, the engine acquired the designation RD-0410, it is less commonly known by its GRAU designation 11B91. That the engine received a GRAU designation means that it was almost certainly considered for military applications. The American NERVA had a thrust of approximately 330 kilonewtons. This was much more than the RD-0410, which had about 35 kilonewtons. This was both by design, and due to political/monetary considerations. The Soviet government had somewhat lost interest in the project once it had become apparent that the nuclear engine was not usable as an ICBM upper stage. More importantly, by developing a lower power engine, the reactor assembly as a whole would be smaller. The RD-0410, including propellant, was planned to mass roughly 15 tons when completed; putting it well within the payload capabilities of Soviet launchers like Proton. The actual engine itself weighed only about two tons. In contrast, the American NERVA was much heavier, and could only be launched by a Saturn V or similar vehicle. There were other important differences between NERVA and RD-0410. The NERVA’s fuel elements were hexagonal in cross section, with several holes drilled in them for hydrogen to pass through. Hundreds of these elements (each about an inch wide) made up the NERVA’s reactor. NERVA Fuel Elements
It has been difficult to find exact information about the geometry of the RD-0410’s fuel rods, however, it appears that they had a complex shape. The fuel rods were twisted, and had a complex cross section, shaped like the petals of a flower. This was intended to lock the fuel rods together, and prevent fuel from falling out of the reactor if a few rods cracked or became dislodged. The fuel elements were made of uranium carbide, in order to better withstand the high temperatures of the core. Development and testing of the RD-0410 proceeded slowly. By 1973, America’s NERVA had already been test fired, then cancelled before actually flying. However, large scale tests of the RD-0410’s components did not begin until 1978. The test reactor was first started on March 27, 1978, and ran for 70 seconds. Gradually, the reactor was run for longer, and at higher temperatures. By 1981, the RD-0410 was running for an hour, its design duration. A specific impulse of 910 seconds was achieved; this was superior to that which was obtained with NERVA. The American Timberwind/SNTP project from the late 1980s planned to achieve similar efficiency with much higher thrust to weight, but it encountered numerous technical problems and did not reach the test stage. All accounts of the RD-0410 state that it’s testing at Semipalatinsk went very well. Originally, it was planned that the engine would fly in 1985 (likely replacing the Block D 4th stage on Proton). However, as the Soviet Union imploded during the 1980s, development slowed, then halted. Other Soviet nuclear rockets were planned, such as the RD-0411; a high thrust (~400 kN) engine that would have been used on a Mars mission, and an engine designated 11B97, which would have had the capability of either nuclear thermal or electric propulsion. However, like all other nuclear rocket programs, none of them came to be. via Astronautix, a concept for a Soviet Mars spacecraft, that likely would have used RD-0411
Important Stats: Unfueled Mass: ~2,000 kg Total Stage Mass ~14,000 kg Thrust: 35 kN ISP: 910 sec Maximum Run Time: 3600s Height: 3.50m Diameter: 1.6m Bibliography: http://www.astronautix.com/engines/rd0410.htm http://www.popmech.ru/made-in-russia/5983-k-marsu-na-reaktore-vzryvnaya-sila/ http://www.cosmoworld.ru/spaceencyclopedia/programs/index.shtml?yard.html