Why does Starship use so many engines compared to the Saturn V

Why Does SpaceX Use 33 Engines While NASA Used Just 5?

In Rockets, Videos by Paul ShillitoLeave a Comment


If you have seen any of the Apollo launch videos or even the new Artemis launch videos you will no doubt be familiar with the shot of the massive five F1 engine’s at launch on the Saturn 5 or the four RS-25 engines and boosters on the SLS rocket as they leave the mobile launchpad.

Now compare that to SpaceX ‘s starship with the Superheavy booster with its 33 Raptor engines at takeoff and you may be wondering why they need 33 smaller engine’s when the Saturn and SLS  only have five or four respectively.

Now yes, there is a difference in the thrust, the SpaceX Starship generates about twice the amount of thrust than the Saturn 5 did, but the end goal of new Artemis missions is the same, whether they are using the SpaceX Starship or the NASA SLS, and that is to take a lander with a human crew and place it onto the moon’s surface and then bring them home again. The Apollo programme did that six times.

In fact, in the 1960s and early 70s, the Apollo program and the later Skylab program launched 13 Saturn 5’s which included Apollo 4, 5 and 6, the unmanned test flights and used a total of 65 Rocketdyne F-1 engines with 100% success rate, although all of them ended up in the Atlantic Ocean because they were designed to be used once and then discarded.

Compare this to the SpaceX Starship and the Superheavy booster, both of which in time should be fully reusable in a similar way to the SpaceX Falcon 9 and Falcon heavy and this is where the two rocket systems diverge into separate paths.

The Saturn 5 was the 1st and so far only rocket to take men beyond low earth orbit but the  Saturn 5 didn’t just come out of nowhere. It was a development from the Jupiter Rocket family, a Research and development vehicle which in turn was part of the Redstone rocket family, which was in turn was an intermediate-range ballistic missile, in the 1950s even though it was being used for Space Research, just like the soviet R-7 ICBM which launched Sputnik into orbit.

The US Army ballistic missile agency designed the Jupiter under the direction of Wernher von Braun, who during the Second World War was the lead scientist and developer of the German V2 rocket.

As part of Operation Paper Clip, Von Braun, 1500 rocket scientists and a number of unused V2 rockets had been brought over from Germany to America after WW2 to help the US develop their rocket technology but they were closely controlled on what they could and couldn’t develop.

That changed in 1957 when the Soviets launched Sputnik 1 on top of an R-7 ICBM, this really put the wind up the Americans because if the Soviets could launch a satellite into orbit they could place a nuclear weapon anywhere in the US.

From then on, Von Braun was put in charge of creating a new heavy rocket which he would based on Jupiter in fact, he referred to it as “an infant Saturn”, and a Jupiter-C rocket was used to launch the first American satellite, Explorer 1 into orbit in 1958 and to match what the Soviets had done with Sputnik the year before.

From 1959 to 1962 the Marshall Space Flight Centre under the direction of von Braun designed a series of Saturn rockets,

The first 3 stage version was the C-3. This would be a three-stage launch vehicle that could lift 45,000 kilogrammes into low earth orbit and send 18,000 kilogrammes to the moon via a translunar injection. The design started with two F-1 engines for the first stage but by 1961 this had been increased to three and it was figured that it might take two or three launches to get a single landing on the moon, so a bigger rocket was planned that could lift a heavier payload.

This would be the Saturn C-4 which would use four F-1 engines and would only need to carry out two launches for each moon landing. But why not do it with just one launch rocket and save the costs of having two launches and two rockets, so the Saturn C-5 was planned, this would use five F-1 engines for the first stage, five J-2 engines for the second stage, and single J-2 engine for the third stage, this would be the design which would go on to take men to the moon for all the Apollo missions.

But as time went by and more testing was done a problem was revealed with the F-1.

The F-1 engine had been developed to replace the E-1 engine which was created to meet the 1955 US Air Force requirement for a very large rocket engine, while the E-1 worked it was seen as a technological dead end and the F-1 was its even larger replacement.

Back in the mid 50s, there was the ethos of making everything as large as possible, the biggest planes, the biggest buildings, the biggest cars and so it went with rocket engine’s.

But at the time there was no rocket that could use such a large rocket engine, so the development was halted. However, when NASA was created in 1957 it could see there might be an application for a very large engine and contracted Rocketdyne to complete its development.

However, early on in its development, tests showed that serious combustion instability could sometimes lead to catastrophic failure. These were oscillations around 4 KHz with harmonics up to 24 kilohertz in the combustion chamber,  they could be so strong that they could cause the combustion chamber to fail but they were also very intermittent and varied in strength.

This was something which had been seen in smaller engines but with the very large size of the F-1, it became a very big issue. The problem was so serious it took almost two years to find a solution and threatened to end the Apollo program if it couldn’t be fixed.

The solution was found by detonating small explosive charges of black powder in a tube just outside the combustion chamber while the engine was running. This would show how these powerful oscillations were moving through the chamber and possibly how to stop them.

In the end, after many design iterations created through trial and error, baffles were used on the injector plates to dampen the oscillations. This became so successful that the engine was stable enough that it would self-damp any artificially induced instability within a 10th of a second and producing a very reliable engine.

Meanwhile in the Soviet Union, engineers had also come across the same instability in larger rocket designs. Their solution was to split the engine up into a single turbopump which fed multiple smaller combustion chambers, usually four. 

If you look at the RD107 engine’s used on the R-7 rocket which launched Sputnik and many others later, you will see the four boosters around the racket base, each with 4 nozzles, but in fact each of these is just one engine with four separate combustion chambers and nozzles and this is how the Soviets got around the instability issue.

They would later use the same idea of many smaller engines for their version of a moon rocket which would be more powerful than the Saturn 5 and would be known as the N1.

This design used 30 smaller rocket engines but they were more efficient than the F1 engine’s with a specific impulse of 331 seconds, compared to be 263 seconds of the F1. They also used a different way of steering the rocket compared to the Saturn 5.

On the Saturn 5, the four outer engines used hydraulically powered gimbals to move engines and steer the rocket on take-off.

The N1 rocket used a method called differential thrust, here all the engines were fixed and they would increase the thrust of the engines on one side of the rocket compared to the other. The problem here was that they had to rely upon an early control computer called KORD to keep the thrust balance correct, if one engine on the left-hand side failed, its counterpart on the right-hand side would also have to be shut down to maintain the thrust balance

This meant and if you had two or more engine failures, you would have to shut down four or six or eight engines instead of just two or three or four.

One of the main reason why the design used many smaller engines meant that should you lose one or two then the launch would not be compromised.

If the same were to happen on a Saturn 5 and one of the five F-1 engines were lost it would lose 20% of its thrust straight away, and would probably mean the mission would have to be aborted as it would not have the thrust required to make into the correct orbit unless it was very late on and near the shutdown of the first stage anyway.

However, due to budget restraints and the lack of test facilities for such a very large rocket such as the N1, only about one in six of the engines was actually tested, the rest were taken straight from the factory floor and fitted to the rocket. The flight itself would be the test, very much like SpaceX and their “Move fast and break things” methodology.

The first four launches of the N1 all failed. Some failed due to fuel plumbing failures caused by an engine shut down and others by the engines themselves failing but both caused a cascade failure and the loss of the entire rocket.

The untimely death of Sergei Kareliov in 1966 the lead engineer and scientist, who unbeknownst to the US was literally the man in control of everything to do with the N1 also was a big shock to the Soviet moon programme. This left his deputy, Vasily Mishin in charge but he lacked Korolev’s political astuteness of the soviet system and influence and was reputed to be a heavy drinker.  By 1972 and already having lost the race to the moon to the Americans, with the failure of the 4th N1 rocket, the Soviet Communist Party lost patience with the N1 programme and cancelled it.

Now some 50 years later SpaceX is using a similar type of design for the Starship heavy booster which uses 33 Raptor engines. This is nothing new, the Falcon Nine uses 9 Merlin engines and the Falcon heavy uses 3 boosters each with nine engines per booster, so 27 engines in total and these have become the most reliable launch systems in space history, but there is another reason why they use so many smaller engines instead of just a few larger ones and that is re-usability.

The Saturn 5 with its F-1 engines was created at a time when reusability was not seen as an option, it wasn’t that they could not be reused, they could have been and were rated for 10 reflights, the problem was getting them back in one piece. In the 1970s, Rocketdyne did studies into a version of the F-1 which included a reusable flyback Saturn S-IC first stage but it did get any further than the drawing table.

Now we live in a world where the old way of spaceflight is seen as wasteful because very little if anything of the Rockets and spacecraft were reused. Elon Musk, CEO of SpaceX, has said you wouldn’t buy a car to use it for one journey and then throw it away, so why should you do the same with a rocket?

One of the main features of SpaceXs rocket designs is to make them reusable like the Falcon 9 which can land back at the launchpad or another landing zone after it has delivered its payload to orbit.

And this is why the Super heavy booster needs so many smaller rockets. Firstly to get the huge amount of thrust at take off, twice that of the  Saturn 5, but when it comes to landing the now virtually empty booster back on earth it requires much less thrust than it did to lift it off in the first place.

If we look at the Super heavy booster, its weight at launch is 3600 tonnes, 3400 that is fuel leaving just 200 tonnes for the dry weight of the booster plus some fuel for landing, so the thrust required to slow it down to landing speed is not much.

If the Super Heavy had say five F-1 equivalent engines each with 700 tonnes of thrust, then just one engine would be too powerful for it to land as the F-1 engines were unable to throttle down the power. Even if the engine’s had throttle control it would have to go down under 40% which only the most modern engine’s that have been designed with that in mind can do.

During the final landing of the Superheavy booster, 13 engines are used to slow it from high altitude and about 1200km per hour which then reduce to 3 when it is close to the ground and the speed drops to under 10km per hour.

Having the ability to use many engines to slow the descent and then switch to just three engines and using their throttle control and gimbling to guide the booster down allows it to land and be reused.

Big engines are also much more expensive to manufacture, the F1 engines were effectively handmade with thousands of welds holding it together. Today with new manufacturing techniques the number of parts required can be 80 to 90% less with the production of Raptor engines running at about one per day and with mass production levels, the cost of a Raptor engine is expected to be under $250,000.

The Rocketdyne F-1 engines were a simple design and relatively cheap to produce considering how labour-intensive they were to build.

They were estimated to cost the equivalent of  $15 million each in today’s money assuming that forty were purchased and 10 to 12 were manufactured per year,  in all  98 were produced and delivered to NASA in total.

Although each F-1 was going to be used just once for the launch and ditched in the Atlantic after, it was rated for 10 re-flights.

In testing, 2 engines were used with one performing 20 tests for a total of 2,256 seconds and the other for 34 tests and 2913 seconds. During the actual flights they were only used for between 159 and 165 seconds, if they could have been returned to earth safely then they could have been reused many more times.

In all, SpaceX’s model of using many engine’s not only gives it the thrust it requires at takeoff  but also the control for landing either the booster or the Starship back to earth or on another planet or moon if required.

It’s also a little ironic that the SLS rockets are using the RS-25 engines from the space shuttle which the designers went to great lengths to make reusable are now being used for a one way journey and being dropped into the Atlantic Ocean in the same way as their F-1 predecessors and that the reusable part of the SLS rocket are now the solid rocket boosters which are far simpler in their construction, so you’re saving the simple stuff and throwing away the complicated bits. Maybe the best brains in NASA have already left and are working for SpaceX because it’s beginning to look that way.

So thank you for watching and I hope you enjoyed the video and if you did then please thumbs up share an subscribe and a big thank you goes to our patreons for their on going support.

Paul Shillito
Creator and presenter of Curious Droid Youtube channel and website www.curious-droid.com.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.