Spacecraft systems

In short

A spacecraft is a system of interacting subsystems — structure, power, thermal, attitude control, propulsion, command and data handling, communications and payload — that together keep an instrument alive and correctly pointed in vacuum, radiation and extreme temperature cycling. The "bus" is everything that serves the payload; mission design is largely the trade of mass, power and pointing budgets between them. It is examined in GATE Aerospace (space dynamics and spacecraft engineering), university Space Technology/Satellite Engineering papers, and forms the core of technical interviews at ISRO centres (URSC Bengaluru, VSSC, SAC, LPSC, ISTRAC), and at HAL, Godrej Aerospace, L&T, Ananth Technologies, Skyroot, Agnikul, Pixxel and Bellatrix. ISRO Scientist/Engineer-SC written papers and private NewSpace design reviews both expect subsystem-level numbers, not just names.

Explained

The spacecraft as a system: bus, payload and budgets

• Payload: the mission equipment — imaging camera, radiometer, transponder, scientific instrument. Everything else exists to support it.
• Bus (service module): structure and mechanisms, electrical power system (EPS), thermal control (TCS), attitude and orbit control (AOCS), propulsion, command and data handling (C&DH or OBC), telemetry-tracking-and-command (TT&C).
• Mass budget: typical communications satellite — payload 15–25%, propellant 40–50% (GTO-to-GEO missions), bus dry mass the rest. Low Earth orbit (LEO) remote-sensing satellites carry far less propellant.
• Power budget: summed over operating modes (sunlight, eclipse, peak payload, safe mode) with 10–25% margin held at design review.
• Link, pointing, data and ΔV budgets are maintained the same way; margins shrink as the design matures (typical practice: 20% mass margin at preliminary design review, ~5% at flight acceptance).
• Redundancy: cold/hot redundancy of critical boxes, cross-strapping of interfaces; single-point failures are listed and justified in the FMECA.
• ISRO buses as reference: I-1K (IRS class, ~1,000 kg), I-2K, I-3K, I-4K/I-6K (GSAT class), and the small/microsat buses (IMS-1, IMS-2) and PSLV Orbital Experimental Module (POEM) platform. Exact power and mass figures differ by configuration — quote them as classes, not fixed numbers.

Attitude determination and control (AOCS/ADCS)

• Purpose: determine orientation, control it to the required accuracy, and reject disturbance torques.
• Environmental disturbance torques (LEO, descending order of importance with altitude): aerodynamic drag torque (below ~500 km), gravity-gradient torque, magnetic torque from residual dipole, solar radiation pressure torque (dominant above ~1,000 km and in GEO).
• Sensors: sun sensors (coarse, ~0.1–1°), earth/horizon sensors, magnetometers (LEO only), star sensors/trackers (arc-second class, best accuracy), gyroscopes (rate, drift needs periodic update), GPS/NavIC receivers for position and time.
• Actuators:
  • Reaction wheels — exchange momentum with the body; fine pointing; need periodic momentum desaturation (unloading) by thrusters or magnetorquers.
  • Control moment gyros (CMG) — gimballed spinning wheels, large torque; used on large platforms and the ISS.
  • Magnetorquers (torque rods/coils) — react against Earth's field, no propellant, low torque, LEO only; T = m × B.
  • Thrusters — high torque, propellant-consuming, used for slews, desaturation and safe mode.
  • Passive methods — gravity-gradient boom (stable along local vertical), spin stabilisation, dampers.
• Stabilisation types:
  1. Spin stabilised — whole body spins, gyroscopic stiffness; simple, but only one axis pointed and solar cells partly shaded (early Intelsat comsats, Rohini class).
  2. Dual-spin — spinning rotor plus despun platform carrying antennas.
  3. Three-axis (zero-momentum) — wheels on three axes, any attitude achievable; standard for imaging and modern comsats.
  4. Momentum-bias (one-wheel) — a single wheel gives stiffness about pitch; cheap three-axis-like performance.
• Rigid-body dynamics: Euler's equations, I ω̇ + ω × (I ω) = T. Major/minor axis rule: a spinning body with energy dissipation is stable only about its axis of maximum moment of inertia.
• Control law: PID or PD on quaternion/Euler error; stability margins (gain ≥ 6 dB, phase ≥ 30°) demonstrated with flexible-mode and fuel-slosh models included.
• Pointing terms: accuracy (absolute error), stability/jitter (error rate over an exposure), knowledge (how well attitude is measured after the fact — can be better than control accuracy, which is why ground-processed imagery improves geolocation).

Electrical power system (EPS)

• Chain: solar array → power conditioning (shunt/PPT or MPPT regulator) → bus (28 V, 42 V, 50 V or 100 V regulated) → battery charge-discharge regulator → loads.
• Solar cells: silicon (~14–17% efficient) largely replaced by multi-junction GaAs/GaInP (~28–32% at beginning of life). Output falls with temperature and with radiation degradation (typically 1.5–3.75% per year for Si; less for triple-junction).
• Array types: body-mounted (small satellites), deployable rigid panels with single-axis solar array drive assembly (SADA), flexible roll-out arrays.
• Batteries: Ni-Cd and Ni-H₂ historically; lithium-ion now standard (ISRO uses indigenous Li-ion cells, also supplied for terrestrial use through technology transfer). Key parameter is depth of discharge (DoD): LEO satellites cycle ~5,000–6,000 times a year in orbit, so DoD is held low (10–25%); GEO satellites see only ~90 eclipses a year and allow 60–80%.
• Other sources: radioisotope thermoelectric generators (RTG) for deep-space/outer-planet missions; fuel cells for short crewed missions (Apollo, Shuttle).
• Protections: latching current limiters, bus undervoltage load shedding, and automatic transfer to safe mode (sun-pointing, payload off) on anomaly.

Thermal control (TCS)

• Environment: direct solar flux (solar constant ≈ 1,361 W/m² at 1 AU), albedo (~0.3 of solar flux reflected by Earth), Earth infrared (~237 W/m² average), deep-space sink at ~3 K, plus internal dissipation. Vacuum means no convection — only conduction and radiation.
• Governing property pair: α (solar absorptivity) and ε (infrared emissivity); the ratio α/ε sets equilibrium temperature. White paint has low α/ε (cools), polished metal/gold has high α/ε (runs hot).
• Passive methods: thermal coatings and paints, second-surface mirrors (OSR), multi-layer insulation (MLI) blankets of aluminised Kapton/Mylar separated by Dacron net, thermal doublers, conduction isolators, heat pipes (constant-conductance and variable-conductance).
• Active methods: electrical heaters with thermostats/software control, louvres, pumped fluid loops, thermoelectric coolers and cryocoolers (for IR detectors, typically 60–150 K).
• Typical allowable ranges: electronics −10 to +50 °C, batteries 0 to +25 °C (narrowest, so batteries drive heater power), propellant lines above freezing (hydrazine freezes near 2 °C — hence line heaters), detectors cryogenic.
• Analysis: lumped-node thermal network solved for worst hot case (end of life, high α, summer solstice/full sun) and worst cold case (beginning of life, eclipse, payload off).

Propulsion and ΔV

• Function: orbit raising, station-keeping (GEO north-south for inclination, east-west for longitude), orbit maintenance against drag in LEO, attitude control and desaturation, and end-of-life disposal.
• Chemical options:
  • Cold gas (N₂, Isp ~50–75 s) — simple, used on small satellites.
  • Monopropellant hydrazine over a Shell-405/LCH catalyst bed, Isp ~220–230 s — the classic reaction control system.
  • Bipropellant MON-3/MMH or N₂O₄-UDMH, Isp ~300–325 s — the Liquid Apogee Motor (LAM) on ISRO's GSAT satellites; LPSC's 440 N LAM is the standard unit.
  • Solid apogee boost motor — single-burn, fixed impulse.
• Electric propulsion: gridded ion and Hall-effect thrusters, Isp ~1,500–3,000 s, thrust in mN, so burns last weeks — huge propellant saving. ISRO has flown electric propulsion on GSAT-9/GSAT-20 class platforms and the all-electric TDS-01 technology demonstrator; GEO operators now use it for full orbit raising.
• Feed systems: blowdown (pressure falls as propellant depletes) vs regulated pressure-fed (helium through a regulator, constant thrust).
• Green propellants: HAN/ADN-based and hydrogen-peroxide monopropellants under development in India (VSSC/LPSC and startups such as Bellatrix) to replace toxic hydrazine.

Command, data handling, TT&C and ground segment

• On-board computer (OBC): radiation-tolerant processors, watchdog timers, EDAC memory, stored command sequences (time-tagged), autonomy for safe-mode entry.
• Data handling: payload data compression, solid-state recorder for store-and-forward (needed because a LEO satellite sees an Indian ground station only ~10–12 minutes per pass), packet telemetry to CCSDS standards.
• TT&C link: S-band for telemetry/command (low rate, omni-coverage for safe mode), X-band/Ka-band for high-rate payload data, C/Ku/Ka-band for communication payloads. Ranging and Doppler tracking give orbit determination.
• Link equation elements: EIRP, free-space loss, G/T of the receiver, required Eb/N₀, coding gain, rain margin (large in Ka-band over India's monsoon).
• Indian ground segment: ISTRAC (Bengaluru, Lucknow, Port Blair, Thiruvananthapuram, Mauritius, Brunei, Biak, Antarctica), MCF Hassan and Bhopal for GEO satellites, NRSC Shadnagar for data reception, IDSN Byalalu for deep space.
• Radiation effects: total ionising dose (degrades electronics), single-event upsets/latch-up (mitigated by EDAC, triple modular redundancy, latch-up protection), displacement damage in solar cells and detectors. The South Atlantic Anomaly and Van Allen belts set the dose.

Structures, mechanisms, environment testing and regulation

• Structure: primary load path (central thrust cylinder or box of honeycomb panels — aluminium honeycomb with CFRP or Al face sheets), secondary brackets, launch-load design cases (quasi-static 6–12 g axial plus lateral, acoustic, random vibration, shock from separation).
• Stiffness requirement: the launcher specifies minimum first natural frequencies (commonly ~20–35 Hz axial, ~10–15 Hz lateral) so the spacecraft does not couple with launcher modes — check the PSLV/GSLV/LVM3 user manual for the mission's actual values.
• Mechanisms: separation systems (clamp band, pyro bolts), solar array deployment and hold-down release, antenna pointing mechanisms, despin bearings; all single-shot items are life-tested to many times the required cycles.
• Environmental test sequence: mass properties → sine/random vibration → acoustic → shock → thermovac/thermal balance and thermal cycling → EMI/EMC → antenna range/compact antenna test facility → magnetic cleanliness. ISRO's facilities are at URSC/ISITE Bengaluru and LPSC/VSSC.
• Standards and practice: ECSS and MIL-STD-1540/NASA-STD derived practice is followed; CCSDS for telemetry/telecommand packets; ITU Radio Regulations and WRC filings for frequency and orbital slot coordination, filed in India through the Department of Telecommunications/WPC.
• Indian law: the Indian Space Policy 2023 opened space activity to non-government entities, with IN-SPACe as the authorisation and regulatory body and NSIL as the commercial arm; ISRO retains R&D. Space activity by Indian entities needs IN-SPACe authorisation, and India follows the UN COPUOS Space Debris Mitigation Guidelines and the IADC 25-year post-mission disposal rule for LEO (now being tightened internationally toward 5 years in some jurisdictions). ISRO's Debris Free Space Missions (DFSM) commitment targets debris-free Indian missions by 2030, supported by the IS4OM centre in Bengaluru. A dedicated Indian Space Activities Bill has been under consideration — treat its status as pending rather than enacted.

Key points

• Bus vs payload — the bus is every subsystem that supports the mission instrument; payload mass is typically only 15–25% of launch mass.
• Budgets — mass, power, ΔV, link, pointing and data budgets, each with declared margin, are the real design currency.
• Disturbance torques — aerodynamic (low LEO), gravity-gradient, magnetic, solar radiation pressure (dominant at high altitude/GEO).
• Star sensor — most accurate attitude sensor (arc-second), needs a star catalogue; sun sensors and magnetometers are coarse backups.
• Reaction wheel — torque by momentum exchange, no propellant, but saturates; desaturated by magnetorquers (LEO) or thrusters.
• Magnetorquer — T = m × B; works only where there is a magnetic field, so LEO only, and gives no torque about the field direction.
• Spin stability — with internal energy dissipation, spin is stable only about the maximum-inertia axis.
• Three-axis stabilisation — standard for imaging and modern comsats; spin and dual-spin are legacy or small-satellite choices.
• Pointing accuracy vs knowledge vs jitter — three distinct requirements; knowledge can exceed control accuracy.
• EPS chain — solar array → regulator → bus → battery; sized for end-of-life, worst-case sun angle and eclipse load.
• Depth of discharge — low (10–25%) for LEO's ~5,000–6,000 cycles/year, high (60–80%) for GEO's ~90 eclipses/year.
• α/ε ratio — sets radiative equilibrium temperature; low α/ε coatings cool, MLI isolates, heaters protect batteries and hydrazine lines.
• No convection in vacuum — all spacecraft heat rejection is ultimately by radiation to a ~3 K sink.
• Specific impulse — Isp = F/(ṁ g₀); cold gas ~60 s, monopropellant ~225 s, bipropellant ~310 s, electric 1,500–3,000 s.
• Electric propulsion trade — enormous propellant saving against months of low-thrust burn time and high power demand.
• Launch environment — quasi-static g loads, random vibration, acoustics and shock, plus a minimum first-mode frequency set by the launcher.
• Radiation — TID, SEU/SEL and displacement damage; countered by shielding, rad-tolerant parts, EDAC and redundancy.
• Indian regulation — Indian Space Policy 2023, IN-SPACe authorisation, NSIL commercialisation, ITU filings via WPC/DoT, IADC debris guidelines.

Formulas and examples

Example 1 — Propellant for GTO-to-GEO apogee burn. A 2,000 kg satellite in GTO needs ΔV = 1,500 m/s using a 440 N LAM with Isp = 310 s.

• Effective exhaust velocity c = 310 × 9.80665 = 3,040 m/s.
• Mass ratio m₀/mf = exp(1500/3040) = exp(0.4934) = 1.6378.
• mf = 2000/1.6378 = 1,221 kg → propellant mp = 779 kg (39% of launch mass).
• Burn time = mp/ṁ, with ṁ = F/c = 440/3040 = 0.1447 kg/s → t = 779/0.1447 ≈ 5,383 s ≈ 90 minutes (so it is split over several apogee passes).
• With an electric thruster at Isp = 1,600 s: c = 15,691 m/s, mass ratio = exp(0.0956) = 1.1003, mf = 1,818 kg → mp = 182 kg — a saving of about 597 kg, but at 200 mN thrust, ṁ = 1.275×10⁻⁵ kg/s and the burn needs ≈1.43×10⁷ s ≈ 165 days.

Example 2 — Solar array sizing for a LEO satellite. Orbit period 95 min with 35 min eclipse (Td = 60 min, Te = 35 min). Loads: Pd = 600 W daylight, Pe = 400 W eclipse. Peak-power tracking: Xd = 0.85, Xe = 0.65.

• Psa = [(400 × 35/0.65) + (600 × 60/0.85)] / 60 = [21,538 + 42,353]/60 = 63,891/60 = 1,065 W required at the array.
• Cell output at beginning of life: Φ ηcell = 1361 × 0.30 = 408 W/m²; with packing factor 0.85 and incidence cos θ = 0.90 → 408 × 0.85 × 0.90 = 312 W/m² (BOL).
• Degradation 2.5%/year over 5 years: (1 − 0.025)⁵ = 0.881 → EOL = 312 × 0.881 = 275 W/m².
• Array area = 1065/275 = 3.87 m² (say two 2.0 m² wings, giving margin).

Example 3 — Battery sizing for the same satellite. Pe = 400 W, Te = 35 min = 0.583 h, DoD = 0.20, one battery, ηt = 0.90.

• Energy from battery = 400 × 0.583 = 233 W·h.
• Cb = 233/(0.20 × 1 × 0.90) = 1,296 W·h nameplate.
• At a 28 V bus: 1296/28 = 46 A·h. (At DoD = 0.40 it would be 648 W·h / 23 A·h — half the mass, but far shorter cycle life.)

Example 4 — Radiative equilibrium temperature of a radiator panel. A 1.0 m² panel faces deep space on one side and rejects 150 W of internal dissipation; it also absorbs sunlight on the same side with α = 0.14, ε = 0.84 (white paint / OSR class), Φ = 1,361 W/m², incidence 60° (cos θ = 0.5).

• Absorbed solar = 0.14 × 1361 × 1.0 × 0.5 = 95.3 W. Total in = 95.3 + 150 = 245.3 W.
• T = [245.3/(0.84 × 5.67×10⁻⁸ × 1.0)]^0.25 = [245.3/4.763×10⁻⁸]^0.25 = (5.150×10⁹)^0.25.
• (5.150×10⁹)^0.25 = 268 K ≈ −5 °C — acceptable for electronics.
• With gold-like finish (α = 0.30, ε = 0.04): in = 204 + 150 = 354 W; T = [354/(2.268×10⁻⁹)]^0.25 = (1.561×10¹¹)^0.25 ≈ 629 K — the α/ε ratio, not the heat load, dominates.

Example 5 — Reaction-wheel slew and sizing. A satellite with Iz = 60 kg·m² must slew 30° (0.5236 rad) about z in 60 s, accelerating for 30 s and decelerating for 30 s.

• Half the angle in the first 30 s: 0.2618 = ½ α (30)² → α = 2 × 0.2618/900 = 5.82×10⁻⁴ rad/s².
• Torque T = I α = 60 × 5.82×10⁻⁴ = 0.0349 N·m (35 mN·m — a small commercial wheel).
• Peak rate ω = α t = 5.82×10⁻⁴ × 30 = 0.01745 rad/s → momentum stored H = 60 × 0.01745 = 1.05 N·m·s, so a wheel of ≥1.05 N·m·s capacity (plus margin for disturbance accumulation) is needed.
• Desaturating 1.05 N·m·s with magnetorquers of m = 15 A·m² in B = 30 µT: max torque = 15 × 30×10⁻⁶ = 4.5×10⁻⁴ N·m → time = 1.05/4.5×10⁻⁴ ≈ 2,333 s ≈ 39 min of favourable geometry.

Where it is used

• ISRO satellite development: URSC Bengaluru (bus design, AOCS, EPS, integration), SAC Ahmedabad (payloads, communications), LPSC (propulsion, LAM, electric thrusters), ISITE (assembly, integration and test) — roles such as satellite systems engineer, AOCS engineer, thermal analyst, test engineer.
• Indian NewSpace: Pixxel (hyperspectral smallsat constellation), Dhruva Space (satellite buses and deployers), Skyroot and Agnikul (launch vehicles with avionics and propulsion), Bellatrix Aerospace (in-space propulsion), Digantara (space situational awareness).
• Industry suppliers: HAL, L&T, Godrej Aerospace, Walchandnagar, Ananth Technologies and Alpha Design build structures, propulsion hardware and integrate satellites under ISRO/NSIL contracts.
• Operations: ISTRAC and MCF Hassan for telemetry, commanding, orbit determination and station-keeping manoeuvres; NRSC Shadnagar for payload data reception and products.
• Applications downstream: INSAT/GSAT communication and DTH, IRS/Cartosat/RISAT/EOS remote sensing for agriculture and disaster management, NavIC navigation, Chandrayaan and Aditya-L1 science platforms, Gaganyaan crew module environmental control and life support.
• Analysis tools and skills employed: STK/GMAT/FreeFlyer for mission analysis, MATLAB/Simulink for AOCS loops, Thermal Desktop/ESATAN for TCS, NASTRAN/ANSYS for structural qualification, CCSDS-compliant software for telemetry.

Interview questions

• Q: What is the difference between the spacecraft bus and the payload? A: The payload performs the mission (camera, transponder, instrument); the bus provides structure, power, thermal control, attitude control, propulsion and communications so the payload can work.
• Q: Why do reaction wheels need desaturation? A: Reaction wheels only exchange momentum internally, so external disturbance torques accumulate as wheel speed; once near maximum speed the wheel cannot absorb more, and thrusters or magnetorquers must dump the momentum externally.
• Q: Why do magnetorquers not work in geostationary orbit? A: Earth's magnetic field at GEO altitude is roughly three orders of magnitude weaker than in LEO, so the achievable torque m × B is negligible.
• Q: Which attitude sensor gives the best accuracy and what does it need? A: A star tracker, to arc-second level; it needs an onboard star catalogue, a clear field of view away from sun/Earth, and a reasonable rate limit on the satellite.
• Q: Why is the allowable depth of discharge much lower in LEO than in GEO? A: A LEO satellite goes through thousands of charge-discharge cycles each year while a GEO satellite sees roughly 90 eclipses, and cycle life falls steeply with increasing DoD.
• Q: Why is the α/ε ratio central to spacecraft thermal design? A: In vacuum there is no convection, so equilibrium temperature is set by the balance between absorbed solar energy (α) and radiated infrared energy (ε); a low α/ε surface runs cold and a high α/ε surface runs hot, independent of size.
• Q: What does MLI actually do, and why does it need many layers? A: It blocks radiative heat transfer; each reflective layer reduces radiation exchange and the low-conductivity spacers minimise conduction, so effective emittance falls with the number of layers — but it works only in vacuum.
• Q: When would you choose electric propulsion over a bipropellant system? A: When propellant mass dominates the budget and time is available — station-keeping over 15 years or slow GEO orbit raising — provided enough electrical power exists and the long low-thrust burn is operationally acceptable.
• Q: What dominates the disturbance torque at 700 km versus at GEO? A: At 700 km gravity-gradient and magnetic torques dominate with some residual aerodynamic torque; at GEO solar radiation pressure dominates, especially with large solar arrays and antennas.
• Q: Why does a launcher specify a minimum first natural frequency for the spacecraft? A: To keep spacecraft modes away from the launch vehicle's structural and control frequencies, preventing dynamic coupling and amplified loads during flight.
• Q: What is a single-event upset and how is it handled? A: A heavy ion or proton flips a memory or logic state; it is handled by error detection and correction coding, memory scrubbing, watchdog timers, triple modular redundancy and latch-up protection circuits.
• Q: Why is a satellite's pointing knowledge sometimes better than its pointing accuracy? A: Attitude can be reconstructed on the ground from recorded star-tracker and gyro data far more precisely than the control loop can hold it in real time, which is why image geolocation improves after processing.
• Q: A satellite enters safe mode — what does it typically do? A: It switches off the payload, points the solar arrays sun-ward using coarse sensors and a simple control law, sheds non-essential loads, switches to the low-rate omni S-band link and waits for ground commands.

Common mistakes

• Treating g₀ in the rocket equation as local gravity — it is the fixed constant 9.80665 m/s² used to define Isp.
• Sizing the solar array for beginning-of-life, normal-incidence conditions, then finding the power budget fails at end of life with off-pointing and degradation.
• Using battery nameplate capacity as usable energy; only DoD × capacity is usable, and transmission efficiency must be applied.
• Writing convective heat transfer into a spacecraft thermal balance — in orbit there is none; only conduction and radiation.
• Assuming reaction wheels can hold attitude indefinitely, or that magnetorquers can produce torque about the local field direction (they cannot).
• Mixing up spin stability rules — with energy dissipation a spinner is stable only about the maximum-inertia axis, not the minimum.
• Quoting pointing accuracy without saying whether it is control accuracy, attitude knowledge or jitter, and over what time interval.
• Ignoring mass and power margins at concept stage, so every later growth has to be paid for by cutting payload.

Related topics

Orbital mechanics basics · Rocket propulsion and launch vehicles · Satellite communication and link budgets · Space environment and radiation effects · Aerospace structures and vibration · Control systems and state estimation