## Concept explanation A reusable rocket can land safely only if its **vertical velocity** and **horizontal velocity** both reach nearly zero at the exact moment the vehicle reaches the pad, while its **orientation** stays close to upright. Gravity is always pulling downward, so the engine must produce just the right amount of thrust and often a slight tilt to cancel sideways drift. That makes landing hard because the rocket is solving several correction problems at once: height, falling speed, sideways motion, and attitude. ## What you see You’re looking at a tall 3D landing scene with the rocket above a glowing pad and a broad ground plane for spatial reference. The large colored arrows show the main influences on the rocket: blue for vertical motion, gold for sideways motion, red for gravity, and teal for engine thrust. The glowing marker on the ground shows the predicted impact point based on the rocket’s current state, so you can immediately see whether the vehicle is heading toward the pad or missing it. ## Try it yourself - **Move the descent-condition slider** to switch between different starting altitudes and speeds, and compare how much harder the recovery becomes. - **Increase or decrease the engine-thrust slider** and watch how the predicted impact point shifts as the rocket changes how fast it slows its descent. - **Adjust the rocket-tilt slider** to trade vertical support for sideways correction, and notice how fixing drift can also make the downward problem worse. - **Press the reset button** after experimenting so you can return to the selected starting condition and test a different correction strategy. - **Orbit the camera by dragging** to check the rocket, vectors, and landing pad from different angles and build your 3D intuition. ## Concept explanation A rocket descending vertically is controlled by the balance between **thrust** and **weight**. **Weight** is the downward force caused by gravity, while **thrust** is the upward push from the engine. If thrust is smaller than weight, the rocket speeds up downward. If thrust matches weight, the forces nearly cancel and the rocket hovers or keeps descending very slowly. If thrust becomes greater than weight, the rocket still may be moving downward at first, but it begins to **decelerate** because the net force is upward. ## What you see You’re looking at a 3D rocket above a landing pad with a red gravity arrow pointing down and a green thrust arrow pointing up. The teal vertical line behind the rocket shows its current altitude relative to the pad. The readout in the corner shows altitude, vertical speed, and the current motion state, while the control panel lets you change the engine thrust and compare the thrust-to-weight ratio directly. ## Try it yourself - **Move the thrust slider below `1.00`** and watch the rocket fall faster as gravity becomes the stronger force. - **Set the slider close to `1.00`** and notice how the rocket approaches a near-hover condition with only small changes in vertical speed. - **Push the slider above `1.00`** and see the downward motion begin to slow because thrust now exceeds weight. - **Compare the arrow lengths** to connect the force balance with the rocket’s changing motion. - **Drag to orbit the camera** so you can inspect the straight-down descent and the altitude trail from different angles. ## Concept explanation When a rocket engine is **gimbaled**, the nozzle pivots so the **thrust vector** no longer points straight through the rocket’s **center of mass**. That offset creates a **torque**, which starts rotating the vehicle. This is useful during landing because you can steer the rocket’s attitude without aerodynamic surfaces, but it also means the same control input both pushes and twists the vehicle. If the gimbal angle is too large or held too long, rotation can build up and make the landing unstable. ## What you see You are looking at a 3D landing rocket standing above a pad. The nozzle under the rocket can pivot left or right, the teal dot marks the center of mass, the gold arrow shows the thrust direction, and the purple cue appears when off-center thrust creates torque. The attitude readout and rotation-rate readout help you connect what you see geometrically with how the rocket actually begins to turn. ## Try it yourself - **Drag horizontally across the scene** to pivot the nozzle and watch the thrust vector swing away from the centerline. - **Move the gimbal angle slider** to make precise left-right nozzle changes and compare small corrections with large ones. - **Increase the thrust level** and notice how the same nozzle deflection creates a stronger turning effect. - **Pause the motion** to inspect the geometry: look at whether the thrust arrow passes through the center of mass or misses it. - **Reset the rocket** and try a brief deflection versus a sustained one to see how rotation keeps building if you do not manage it carefully. ## Concept explanation During atmospheric descent, **grid fins** act like small aerodynamic control surfaces on the booster. As air rushes past them, their angled lattice structure generates **aerodynamic force**, which pushes sideways on the rocket and steers its path before the engines relight for landing. The key idea is that this steering depends on airflow: with dense air or high descent speed, the fins can make strong corrections, but with very little air moving past them, they produce almost no useful force. ## What you see You are looking at a descending booster inside a semi-transparent atmospheric column. The glowing streaks show the surrounding airflow, the colored arrows show sideways force produced by the fins, and the teal trail traces the booster's recent descent path. As the force grows, the path bends more; when the air is thin or the fins are retracted, the arrows shrink and the descent becomes much straighter. ## Try it yourself - **Lower the descent speed / air flow slider** and notice how the force arrows almost disappear, even if the fins stay angled. - **Raise the descent speed / air flow slider** to make the airflow streaks stronger and watch the curved path bend more clearly. - **Drag the fin-angle control ring** beside the booster to tilt the fins directly and see the steering direction flip. - **Move the fin angle through `0°`** and observe how the sideways force changes sign, switching the path from left-curving to right-curving. - **Retract the fins** and compare the result with the deployed case; the booster loses most of its aerodynamic steering authority. - **Press Reset scenario** and test combinations such as high airflow with small fin angle versus low airflow with large fin angle. ## Concept explanation A landing burn works best when the rocket begins thrust at the **precise moment** that lets its engine cancel the remaining downward speed exactly at the surface. If the burn starts **too late**, gravity has already built up too much speed and the rocket hits hard. If it starts **too early**, the rocket loses speed high above the pad, which means it must hover or climb slightly while still burning fuel. The most fuel-efficient case is the one where **velocity reaches nearly zero at touchdown** instead of before or after. ## What you see You are looking at a 3D side view of the descent. The **teal path** is the trajectory for the burn timing you choose with the slider, and the **blue ghosted path** is a near-optimal reference. The **gold marker** shows where the engine would first ignite on your chosen path. As the slider moves, the status panel updates to show whether the rocket is late, early, or well timed, along with the burn altitude and the speed or stop point at the end. ## Try it yourself - **Move the burn start slider later** and notice how the rocket reaches the pad with too much downward speed. - **Move the burn start slider earlier** and watch the rocket bleed off speed before reaching the surface, leaving it above the pad. - **Try to match the teal path to the blue ghost path** to find the timing that gives the softest touchdown. - **Watch the gold burn-start marker** as you adjust the slider and see how a small timing change shifts the ignition point by many meters. - **Drag the view slightly** to inspect the side-by-side paths while keeping the landing pad and burn marker in sight. ## Concept explanation A real reusable rocket landing depends on **closed-loop control**, which means the vehicle keeps measuring its own state and continuously correcting its motion instead of following one fixed script. During the final descent, **guidance** compares the rocket’s position, velocity, and **attitude** with the landing target, then the controller updates engine **throttle** and **gimbal** direction to reduce those errors. If crosswinds or small tilts push the rocket away, the feedback loop senses the mismatch and commands new corrections, so the vehicle bends its path back toward the pad. ## What you see You are looking at a 3D rocket descending toward a small landing pad while a live trail shows the path it has taken. The engine nozzle pivots to indicate gimbal commands, and the flame length changes with throttle response. The floating sensor card reports altitude, vertical speed, lateral miss distance, and total attitude angle, so you can watch the controller react as disturbances push the rocket off-center and the feedback system pulls it back. ## Try it yourself - **Drag the rocket sideways** before release and notice how the landing path curves back toward the pad instead of falling straight down. - **Press `Release`** and watch the engine gimbal turn toward the error while the throttle changes to manage descent speed. - **Turn off `Crosswind`** to compare a calmer descent with one that needs stronger lateral corrections. - **Turn off `Tilt disturbance`** and see how much smoother the attitude response becomes when the controller has fewer rotational errors to fight. - **Adjust the `Wind strength` slider** and look for bigger lateral deviations, followed by stronger guidance corrections. - **Adjust the `Tilt strength` slider** and observe how attitude noise changes the nozzle motion and recovery behavior. - **Toggle `Sensor overlays`** to focus either on the motion alone or on the measured state values driving the feedback loop. - **Use `Reset`** and repeat with different starting offsets to test how continuous feedback handles different initial errors. ## Concept explanation A reusable rocket landing is a **system-level trade-off**: the vehicle needs enough **fuel reserve** to keep firing late in descent, enough **control authority** from engine gimbaling to fight drift, a low enough **touchdown speed** to avoid structural damage, and a small enough **tilt angle** to stay upright on contact. If wind grows, engine response is delayed, or fuel is reduced, improving one goal often makes another harder, which is why landing strategy is really an optimization problem rather than a single “best” setting. ## What you see You are watching a rocket descend toward a landing pad in 3D while the dashboard tracks the most important margins in real time. The bars show how much fuel remains, how much of the gimbal limit is being used, how close the rocket is to the maximum safe tilt, and what landing speed the system is carrying. The outcome panel updates continuously so you can connect what the rocket is doing in space with why the landing ends as a successful touchdown, hard landing, tip-over, or fuel exhaustion. ## Try it yourself - **Lower the initial fuel** and watch how the rocket can still aim for the pad but may run out of propellant before it can fully slow down. - **Increase the wind strength** and notice how the gimbal bar rises as the engine works harder to cancel sideways drift. - **Raise the engine response delay** to see how late corrections can turn a clean approach into a hard landing or a tip-over. - **Switch between autopilot modes** and compare how a fuel-saving strategy differs from a precision-focused one under the same conditions. - **Look for combinations** where one metric improves while another gets worse, such as saving fuel but arriving with too much speed or too much tilt. - **Press `Reset approach`** after changing settings so you can test a new trade-off scenario from the top of descent.