Starship - flight test path

Table of contents:

Starship - flight test path
Starship - flight test path

Today, the attention of many is riveted to the launches of prototypes of the SpaceX SpaceX spacecraft by Elon Musk. What test strategy did Musk choose, what new flight features does Starship demonstrate, what remains behind the scenes? We present the views and perspectives on his tests.

Installation of the SN10 prototype at the launching position. in Boca Chica on January 29, 2021

Starship. Portrait in outline

SpaceX allows some confusion in the names of its products, so it's worth deciding which is which. Actually Starship is a reusable spacecraft and at the same time the second stage of the Starship rocket space system. In addition to the ship, the rocket includes a first stage called Super Heavy.

Geometrically, the Starship hull is a cylinder 50 meters long and nine meters in diameter (the dimensions of the Soyuz launch vehicle with a warhead). The front part is sharpened by an ogival shape. Four short, movable wings, acting independently, like the flippers of plesiosaurs and pliosaurs, are located in pairs in the nose and tail of the body. The propulsion system is formed by six Raptor engines running on liquid methane and oxygen. The three central motors are optimized for low altitude operation at full atmospheric pressure, hence the name Raptor Sea-level. They are able to significantly change the magnitude of their thrust and, on their movable suspension, can be consistently, quickly and significantly deviate to various positions, creating control moments.

Closer to the edges of the housing are three Raptor Vacuum high-altitude motors, optimized for vacuum operation and therefore with wide nozzles with a large expansion ratio. The prototypes currently being tested do not have these engines.


What is being tested now at the Boca Chica test sites and in the Texas skies above them? Simplified versions of Starship. What is the correct name or classification of this "simplified"?

During development and testing, many gradations of a simplified test product arise.

There are models, usually mass-dimensional - MMG (mass-dimensional model). These are non-functional products that repeat the geometry and mass parameters of the original. There are prototypes - working models that perform basic functions (or a separate function) in general terms, approximately. There are prototypes, more sophisticated prototypes, between which there are often no major distinctions. There are technology demonstrators. There are laboratory samples that also perform the main function or some kind of auxiliary (other laboratory samples can fire and perform interceptions). There are a bunch of pre-production gradations, experimental, technological (for example, a monitor sample) and experimental samples. Do not break spears in search of precise definitions: let's call the flying products tested in Boca Chica just prototypes.


SpaceX, in its rocket program, is making many prototypes for one or another phase of testing, and gradually becoming more complex. Simultaneously, several prototypes are at different stages of production and are numbered according to the assigned sequence of their tests. The last two flight tests of similar prototypes with similar flight and results were called SN8 and SN9. The prototype SN10 is already at the launch site, but the date of its testing is unknown.

To the bottom of the air ocean, or Flight of the steel Pegasus

The recent launches of the SN8 and SN9 prototypes have demonstrated an unusual (for atmospheric rocket technology) descent of Starship in the dense layers of the atmosphere - a flat fall. The drag force of the flow is proportional not only to the air density and the square of the falling velocity, but also to the area of ​​the apparatus across the flow - the area of ​​the silhouette that is "visible" to the flow. For a cigar-shaped body, the largest area will be when the body is located across the stream. This removes the maximum atmospheric resistance - and this is what SpaceX engineers want to ride. This is how a parachutist falls through the air: when falling flat, his speed is lowest.In SN8 and SN9, the falling speed reached a maximum of about 150 meters per second, rapidly decreasing as it descended, and in the last kilometers was small 80-70 meters per second.

The fall of Starship must be manageable. Just like a skydiver controls a free fall.

The parachutist's task is to fall steadily and evenly relative to the horizon. If you strain your arms and legs, trying to push them further in the stream, the speed of the fall slows down. If you relax, the stream will immediately squeeze the limbs closer to the body, and the falling stream will accelerate downward, falling in relation to the circle of soaring neighbors. Having picked up his arms, but spread his legs, the parachutist will get a failure of the front part of the body and, as a result, tilt the body with his head below the horizon, negative pitch. And with it sliding forward, planning. The parachutist, going down, will fly horizontally relative to exactly falling neighbors, fly sideways from one and fly up to another.

Such work of the governing bodies was also shown by the recent tests of the Starship prototypes. Its four steerable wings are aerodynamically similar to the parachutist's limbs when falling flat. Their movement is unusual. Unlike the usual ailerons and flaps, the deflection axis of which is perpendicular to the fuselage, these wings fold closer to the body or move away from it. The axis of their rotation lies along the surface of the fuselage.


By folding its wings like the flippers of a walrus or fins of fish, closer to the body, or spreading wider, Starship controls its fall in the lower atmosphere. Having folded the nasal wings tighter, he gets a sinking / tilt of the body forward under the horizon and gliding, sliding with a horizontal shift. By folding the nasal wings to different degrees, one more than the other, Starship can turn to the right or left relative to the horizon - to change course. And by pressing both wings more strongly on one side, front and rear, you can change the roll - tilt to one side.

A set of these capabilities allows Starship to control its movement and bring the landing to the target - the center of the site. As demonstrated by tests of prototypes SN8 and SN9. This is an aerodynamic guidance system, that is, the alignment of the trajectory line with the target point, near which the engines that regulate the landing speed should be triggered. For this, the entire Starship re-starts two of the three engines and is quickly deployed at a low altitude with the nozzles down.

Control wings of this type, folding with the fuselage and opening with short fins, have not yet been used. For these controls there is still no special name reflecting their mode of action. In some places they are already called "Ilonrons" - by analogy with the ailerons. In the news, they say “flaps,” but flap is different. This is a retractable and retractable trailing edge of the wing, in the case of retractable flaps, extending "beyond the wing". Relatives (not the closest) of Starship airfins can be deflectable flaps on the tail skirt of the warheads of ICBMs.

The same distant relatives and flaps of the aerodynamic brakes of supersonic fighters in the tail section of the aircraft, in the form of paired burdocks protruding from the fuselage into the stream with hydraulic cylinders. The burdock of the brake, which opens into the flow, receives the resulting drag force from the flowing stream.


In aviation, in general, there is a lot of control through an increase in the area of ​​the streamlined silhouette - from the drop-down wingtips of the Su-25 for steering along the course to any spoilers, up to the landing "wing-splits" of the B-2 and wing spoilers of passenger aircraft, the work of which we see through the window at each landing run of the aircraft. But they are all located on the wing consoles. The local controlled obstruction of the transverse flow around the hull is unknown in aviation. In aviation, there is no transverse flow around the hull outside of emergency situations.

Gliding control by controlling the angle of folding of the wings is in nature.This is how the butterflies plan, holding their wings in a motionless "boat" with a controlled angle between the wings: either folding them a little stronger and lowering, then slightly spreading and slowing down the decline. In the future, the appearance of aircraft with a similar flight control principle is quite possible.

Future supersonic flight

So far, the work of the wings has been demonstrated in subsonic fall and transverse flow around them - when the role of the wing is to create aerodynamic drag. When it comes to supersonic motion, the picture is different. The perpendicular flow around the fins will change to a longitudinal one, from the leading edge to the trailing edge. The resistance and lift force will separate and become perpendicular, and the resistance force will be several times less than the lift force. The fins will turn into wings that create lift in supersonic flight.

In them, the rocket will move in a little over a minute after the start, switching to supersonic as it accelerates. Starship at this time is attached to the top of the first stage and sticks the nose into the air, taking on supersonic aerodynamic loads. At the same time, its wings are unlikely to work, just being present on the body. When the first stage separates and Starship continues to accelerate, the density of the remnants of the atmosphere will become very low, and even higher, the concept of supersonic motion will begin to lose its meaning due to the deep rarefaction of the environment.

The supersonic flow increases on its return, when the Starship begins to descend from high altitudes. After all, the ballistic speed is very high - the cosmic level or close to it. Such a speed will give a large heating of the structure, which can quickly become the cause of destruction. And for the sake of reducing heat-shielding masses, they slow down the rate of decline, at least by a couple of times. We need optimization for thermal affairs (and power loads, which will also be there). It may be necessary to stretch the trajectory and make this path more gently.

Supersonic aerodynamic lift will help here. It will appear on the hull and wings of the Starship, which are very effective at supersonic with their fairly large area. Supersonic flow will require the heat resistance of the wing material in order to keep high aerodynamic loads hot. The material of the leading edges, where the highest temperature will be, must still be heat-resistant - not burn in the high-temperature flow. In this case, the body and wings will be covered with tiles of a heat-shielding coating from the side of the incident flow.

But this is all ahead: for now, for subsonic tests, the wings, like the apparatus itself, are made in a simpler design. With aerodynamically low accuracy, with uneven surfaces that do not cause significant resistance at subsonic levels.

Strategy and challenges

Test work is its own world. The objectives of the tests can be as different as the tests themselves. What to solve or create by the test stage? They can be more validating or search engines. What tasks are best solved by the method of field tests, within the framework of old schools, within the framework of new projects. Which areas and stages of the creation of technology are faster or more reliable to solve through tests. What to minimize when designing their stage?

“Strategy without tactics is the longest path to victory. Tactics without strategy is noise before defeat."

Sun Tzu

One way is a long and thorough modeling aimed at avoiding emergency events during the first or rare test flights, verifying everything at once comprehensively. Such calculations take up their own tangible set of resources. Time, personnel, calculation equipment and others. Well, a model is a model, the modeling reliability is not absolute. And the model will not cover all the details - there is a huge variety of complex interconnected processes in the product, simultaneously going on in different areas.

Loads and strength, flow dynamics, temperature fields, variable pressures and vibrations - both in construction and instability in processes.It will take many models to describe and predict the performance of a product piece by piece. How adequately will the complex of these models describe and predict reality?

But on the other hand, modeling saves on carrying out full-scale tests and flight copies. Making them is also costly and takes a lot of resources. Therefore, the metal itself should be saved, and the metal stage should be reduced. This is a more classic, traditional path, which is, for example, the creation of SLS.

Another way is to redirect resources from modeling to field testing and measurements. A path with a greater degree of empiricism. Less simulation, more prototypes per development period. Tests through an increased number of full-scale tests allow obtaining large amounts of factual information. Including one that is difficult or impossible to simulate, with dense measurements of real parameters. Roughly speaking, measurable breakdown of prototypes provides a ton of evidence and, at a price, saves other resources. One of these most important resources is time. The bottom line is a possible strategic gain in the form of earlier commissioning.

The strategic gain can be years - meaning the additional years of operation gained. And not somewhere in the back of the project. Musk seems to be going that route.

The empirical way of testing Starship involves numerous flights and the manufacture of a couple of dozen copies, which are continuously built during testing. Work is going on at two sites at once. Manufacturing many elements in small series reduces the cost of devices, but more measurements of real processes are accumulated. This knowledge can shorten the path and time to achieve the goal - the operation of the product.

In a similar way, the artilleryman, having made a short general calculation, makes two shots: sighting, test. Seeing the undershoot and overshoot of the actual falls, he will use this measured “fork” to determine the third correct shot - the hit.

Measuring fog

How exactly measurements take place, in what formats and volumes are taken and where information is received, SpaceX does not disclose. Therefore, it remains behind the scenes which measuring equipment is used in tests. What is included in the trajectory measuring complex, what telemetry systems, what is the measurement technology?

Indirect estimates are possible. In a recent test, the second engine failed to fire on landing, exhibiting a long burst of bright flame. This is one of the classic manifestations of combustion instability. More precisely, its consequence is that it is not known from the form what type of instability led to the ejection. There are several of them.

There is low frequency instability. These are fluctuations in the fuel pressure in the lines and the chamber, which form a single oscillatory circuit with the combustion process. Pressure pulsations occur with a frequency of 8-12 times per second. Instabilities of various kinds may occur during an emergency shutdown of the engine, with different rates of development of events. There is a high-frequency, acoustic instability of combustion in the combustion chamber, of a gas-dynamic wave nature. Its frequency is a hundred times higher, the first kilohertz.

To measure what is happening with such a frequency (otherwise there is a risk of leaving the cause in the blind zone, outside of measurements), you need to reliably track it in time. The sensors should be polled with a frequency of several thousand times per second. This data stream must be processed and transmitted by a telemetry system, which is therefore forced to be a high-speed radio telemetry system, BRS, and have at least several hundred measurement channels, optimally the first thousand. The higher the telemetry power, the greater the volume of measurements and the usefulness of the tests.

Engine operation

Prototypes SN8 and SN9 are powered by three Raptor Sea-level center engines. They allow a prototype to take off when it is less than half full of fuel components.For the troposphere upward (10-12 kilometers) and testing a controlled descent and landing, this is enough with a margin (clearly exploding at the finish). The launch, climb and controlled aerodynamic descent were flawless. Takeoff on three engines, then with a climb, one was planned to be switched off, later another, and at the top of the trajectory - and the third engine. After aerodynamic descent, two out of three engines were launched flat, on which the landing was made.


However, both tests ended in a landing crash that the same engines were unable to accomplish. In aviation, in official documents, such incidents are formulated as "collapsed, colliding with the earth's surface, followed by a fire." Elon Musk describes it as a joke "landed at high speed and experienced a quick unscheduled disassembly" (that is, RUD - rapid unscheduled disassembly, which quickly became a meme)

In the SN8 test, two engines were turned on during landing, which provided sufficient torque to steer the SN8 vertically and hold it in the landing position. Here, only the vertical deceleration was not enough, the speed reduction to landing, although the landing vertical position of the vehicle was fulfilled.

Why was there not enough thrust of a working pair of engines? Shortly before landing, the power of the engines was switched from the main tanks to the landing ones, and according to Musk's explanation, the methane pressure in the landing tank turned out to be lower than the calculated one, which reduced the fuel supply to the engines and reduced their thrust.

In the second start, one engine stably turned on for SN9, showing a normal jet stream - striped by supersonic shock waves. The second engine threw out an uneven curvature of bright flame, clearly not accelerated to supersonic speed. This indicates that the engine has not reached the design mode. After this outbreak of erratic mode, the engine quieted down.

An attempt to set the landing position with one running engine failed: there were not enough control forces and height for such small forces. The device went into swinging and fell obliquely, with a pitch angle of about 45 degrees. SpaceX itself writes: "During the landing roll-over maneuver, one of the Raptor engines failed to re-engage and the SN9 landed at high speed and experienced RUD."


There are different versions of why this happened. When turning quickly to a vertical position, the fuel components in the tanks are displaced in the form of a wave, which could change the mode of supply of the pipelines and, as a result, the operation of the engines. This sounds unconvincing. There were definitely no external gas-dynamic hindrances for restarting, at a low speed of the oncoming flow of less than 70 meters per second - in contrast to the starts of the Falcon-9 engines in the supersonic mode of descent, which pretty much locks the engine with a supersonic shock wave and its compression filling the nozzle.

There are discussions about the more complex behavior of methane engines and methane as a fuel. However, methane does not have the fundamental features of combustion and thrust creation, which fundamentally distinguish it from other cryogenic fuel pairs. What exactly and how the mechanism of the methane curse works is not stated anywhere. If the engine is made well, it works at all stages and starts. If it is done poorly, it will work that way or not at all. In the two passed tests, all the engines clearly and unanimously worked out the starting task. On landing, the engines started, but did not provide their function - a soft landing. On the next flight, the prototype caught an engine failure. And there was no longer a soft landing option.


Identical engines behaved radically differently - this is also an interesting situation. It is unlikely that the prototype has three different own methane and oxygen lines to three engines. Most likely, the highways are uniform and common. Therefore, the wave displacement of the fuel components in the tanks during verticalization of the prototype affects one common intake neck at the bottom of the tank.In addition, switching to small landing tanks significantly eliminates the problem of misalignment of fuel components.

The motors may have changed their state after being turned on for the first time. And this changed state counteracted restarting. Counteraction can be easily identified by measurement data: an unstable "sneezing" by a flame at a shutdown engine is only a consequence of the behavior of its units and systems, processes in them. Measurement of parameters in a couple of hundreds of channels on the engine (that is, a couple of hundreds of sensors measuring temperatures, pressures, flows, revolutions, voltages and currents at a hundred points) five thousand times per second will give a very detailed picture of what is happening, its development, features, attenuation, all essential components of manifestation.

However, the problem was not identified after the first launch, because on the second it was repeated. Therefore, a slightly longer stage is now possible in order to understand the reasons for repeated engine failures and eliminate them. In principle, this is normal work with a prototype, allowing you to quickly reveal a practical problem. As part of the strategy of empirical bias in the creation of technology. The solutions after the first crash were inadequate, the landing of SN9 was even more emergency. A comparison of two different pictures will probably give enough information and a practical "fork", after which a normal regular landing is possible. The SN10 test, already at the start, will show whether this is so or how serious the problem faced by the two predecessors.

Musk, in turn, said that SpaceX will make changes to the flight pattern and use three engines during landing. So that you can turn off the instability that has shown, and land on the two stably launched. Selection of engines by operating mode.

Civil Aviation Graters, or FAA

The explosions of the two crash landings also echoed in a controlled official setting, prompting claims against SpaceX by the FAA. The US Federal Aviation Administration (or the Federal Aviation Administration) has filed a violation of its SN8 test license by SpaceX.

The FAA (represented by its Space Division) is responsible for licensing commercial space launches by any American company. The agency's role is to ensure that launch operators comply with public safety regulations that limit risk to the general public. The FAA also requires launch operators to purchase liability insurance to cover potential property damage.

FAA regulations require companies with launch licenses for reusable launch vehicles, such as SpaceX and its Starship, to meet an “expected loss limit” for a “non-launch public” of no more than 0,0001 per launch, or one loss per launch. 10 thousand starts. The risk to any person cannot exceed one in a million.

SpaceX, apparently, did not want to pay for the excess of the risks posed by its tests over the levels allowed by the FAA. Therefore, the company, before the December test launch of SN8, sought from the regulator to refuse to recognize such an excess, for which it was necessary to pay, but did not receive consent. The FAA refused to license the G8 launch free of charge, “forgiving” exceeding its standards. And SpaceX launched SN8 without FAA approval. It's unclear if SpaceX was fined or otherwise charged for launching an SN8 test flight without FAA approval.

SpaceX did not receive a license for the next launch of the SN9 prototype. The planned launch date of January 28 was canceled due to the FAA's refusal to issue a license for it until SpaceX provides the FAA with the results of an investigation into the cause of the accident. Musk burst into an angry phrase that with such FAA rules, mankind would never go to Mars. The small war between SpaceX and the FAA ended with the issuance of a license to launch the "nine" less than a day before launch, on the evening of February 1, with the condition that such an accident should not be allowed again.As an FAA spokesman said, "corrective actions related to the SN8 incident are included in the SN9 launch license."

After that, a second explosion occurred immediately, following the same scenario. The FAA will now conduct its own detailed investigation. And here again good telemetry could help, taking a detailed picture of what is happening. High completeness measurement data can be provided to the FAA to expedite the acquisition of a new license for the tenth prototype launch. However, it is very likely that the FAA's functions will be adjusted to simplify the issuance of licenses for unmanned test launches of space technology.

Further path

Musk wants to reduce time as much as possible - this is at the heart of his strategy - so SpaceX announced acceleration of prototype testing. In addition to the SN10, there are six SN prototypes in varying degrees of readiness. The tests are becoming more complex: already on the crashed SN9 there were small black pieces of the TPS (Thermal Protection System) thermal protection coating. On SN10 they are larger and located in several places of the hull, on SN11 they are already taken up to the bow.

The first stage, Super Heavy, is being prepared for testing with two prototypes BN1 and BN2 currently being assembled at another site in Boca Chica. It is not yet known when their flights will begin, but SpaceX plans to launch Starship into orbit later this year. Test flights of both the first and second stages will require large volumes of fuel, which are being transported by sea today.

Equipment for the liquefaction of oxygen and the construction of a local oxygen plant is being brought in, while methane will be taken from local wells. All this speaks of the imminent strengthening of the flight test program, which means that there is a large series of interesting tests ahead with unpredictable results.


It is worth noting that Musk turned the tests themselves into an interesting show and a rich news topic that will occupy viewers and analysts for more than one year. A huge number of people are watching prototype launches, testing progress, and the creation of the SpaceX main vehicle. Someone with empathy, someone - skeptical, someone - making bets and predictions or fiercely debating in the spirit of "fly - not fly".

The previously unknown village of Boca Chica now attracts attention in all parts of the world. Soon, a quarter or a third of the information flow will be space topics. Musk involved a lot of people in the creation of his technology, and SpaceX made test work available to the public mass spectacle and the mainstream. As if all together are implementing this project, which is becoming a common one, sympathizing with the technology being created. And imperceptibly growing up in a more cosmic culture. SpaceX's approach is a forward-looking decision, perhaps going much further than the actual flights of steel prototypes.

Popular by topic