What thrust actually is.
Before any equation, the intuition. Thrust is not magic; it's not "the motor pushing harder." Thrust is the consequence of the propeller accelerating air mass in one direction. By Newton's third law, the drone gets pushed in the opposite direction. The drone hovers because it is constantly accelerating air downward at exactly the rate gravity is trying to accelerate the drone downward.
By the end of this section, you'll understand:
- Why hovering and accelerating air are the same physical event.
- The two variables that determine thrust: how much air, how fast.
- Why bigger propellers are dramatically more efficient than smaller ones for the same thrust.
- What "induced velocity" means and why it's the right way to think about hover.
- Why a hovering drone uses energy continuously even though it isn't moving.
Imagine a drone hovering motionless ten metres above a calm pond. The propellers are spinning hard. Below the drone, air is rushing downward — fast enough that you can feel the breeze on the pond's surface. The drone is not moving, but it is doing work: it is taking still air from above, accelerating it downward, and ejecting it below.
That acceleration of air mass is what produces thrust. Newton's second law says force equals mass times acceleration: F = ma. Newton's third law says every action has an equal and opposite reaction: if the drone accelerates a kilogram of air downward at some rate, the drone experiences an upward force of equal magnitude. The drone hovers when that upward force exactly cancels gravity.
Two variables determine how much thrust you produce: the mass flow rate of air through the propeller (how many kg/s of air the propeller is moving) and the velocity change (how much faster the air leaves than it arrives). More air, faster air — more thrust.
Here's the crucial insight: you can produce the same thrust by moving a lot of air slowly, or a little air fast. But the energy required is dramatically different. Kinetic energy is ½mv² — quadratic in velocity. Doubling the air's exit velocity quadruples the kinetic energy you have to put into it, which means quadrupling the power. Doubling the mass flow only doubles the energy. So moving more air slowly is far more efficient than moving less air fast.
This is why a 10-inch heavy-lift drone can hover with significantly less power than a 5-inch racing drone of the same weight. The bigger propellers move much more air per revolution at lower exit velocities. Same thrust, less power. Same battery, longer flight. If you remember nothing else from this page, remember this.
The technical term for the speed at which air leaves the propeller in hover is the induced velocity. It's "induced" because the propeller induces the airflow rather than encountering pre-existing flow (which is what happens when the drone is moving forward). In hover, induced velocity is what you're paying for. Section 2 will give us an equation that calculates it from the propeller's disc area and the thrust we want.
The 60-second mental model
- Thrust = mass flow rate of air × velocity change. More air, faster air, more thrust.
- Hover means continuously accelerating air down to balance gravity pulling the drone down.
- Same thrust, two recipes: lots of air slowly (efficient) or little air fast (inefficient).
- Bigger props = more air per revolution at lower velocity = far more efficient.
- This is why endurance drones use big slow props, and why small props are limited in payload.
- Energy goes into the kinetic energy of the air leaving the disc — which is why hover uses power even when the drone isn't moving anywhere.
The momentum theory equation.
Momentum theory was developed for ship propellers in the 1860s and applied to aircraft propellers in the 1920s. It treats the propeller as an idealised "actuator disc" — a perfectly thin disc that uniformly accelerates the air passing through it. The actual propeller blades are abstracted away. The result is an equation that predicts hover thrust to within ~15% of measured reality, using only the air density, the disc area, and the induced velocity.
The static thrust (T) of an idealised propeller in hover is:
ρ · air density (kg/m³, ~1.225 at sea level standard)
A · propeller disc area (m², = π·r² where r is half the prop diameter)
v_e · exit velocity of air below the disc (m/s)
This equation is just F = ma rewritten for a continuous flow. The mass flow rate through the disc is ρ·A·v (density × area × velocity), and the velocity change in hover is v_e (air enters at zero velocity, leaves at v_e). The factor of ½ comes from a careful derivation that splits the velocity change between the air arriving at the disc and the air leaving — the disc itself sees only half the total velocity change.
For a typical 5-inch propeller producing the cohort default thrust of ~620 g per motor (6.08 N) at hover:
Now let's compare to a 10-inch heavy-lift drone hovering at the same per-motor weight (so we can see the prop-size effect cleanly). Hypothetical: same 620g per motor target with a 10-inch prop.
This is the quantitative proof of the intuition from Section 1. A 10-inch prop hovering at 6 N of thrust requires about ⅛ the air-acceleration energy of a 5-inch prop doing the same job. (In practice, the savings are smaller — maybe 50–60% — because real propellers have additional losses, and there's also the propeller's own profile drag. But the directional truth holds.)
The induced power equation that explains the v³ scaling:
To halve power at constant thrust: disc area must increase by 4×.
To halve thrust required (lighter drone): induced power drops by ~65%.
Why momentum theory is approximate
The actuator-disc model assumes the propeller perfectly accelerates uniform air through a perfectly thin disc. Real propellers are finite blades that produce non-uniform flow, induce tip vortices, and waste energy in profile drag (drag on the blades themselves). Real-world hover thrust is typically 80–90% of what momentum theory predicts. The remaining 10–20% goes to losses momentum theory doesn't capture.
- Tip losses — the air slips around the blade tips instead of being accelerated downward. Costs ~5–10% of thrust.
- Profile drag — the blades themselves experience drag as they spin through the air. Costs ~5–8% of power.
- Non-uniform flow — the air's velocity through the disc isn't actually uniform; it's faster near the tips than the centre. Slight efficiency loss.
- Hub losses — the centre of the prop near the hub doesn't produce thrust. Small effect on small props, larger on big slow ones.
For a careful first-pass estimate: take momentum theory's answer and multiply by 0.85. That'll get you within ~10% of measured reality at sea level. For better than that, you need empirical data — which is Section 3.
Why we use real data.
Momentum theory gives you the right ballpark and the right physical intuition — but for actual build decisions, manufacturer thrust data is more accurate and easier to use. The right workflow: use momentum theory to understand why a build will perform a certain way, and use empirical data to predict what that performance will be.
Between momentum theory and the manufacturer's thrust chart, there's a body of more sophisticated propeller modelling called blade element theory. It treats each segment of the blade as a small wing with its own angle of attack, lift coefficient, and drag coefficient, integrating along the blade length. It's more accurate than momentum theory but requires detailed blade geometry data that manufacturers usually don't publish.
For practical drone design, the chain of usefulness is:
- Momentum theory — first-principles understanding of why bigger discs are more efficient. Good to ±15%.
- Blade element theory — better predictions when you have detailed blade data. Used by serious aerodynamicists; rarely needed for cohort-scale work. Good to ±5%.
- Manufacturer thrust charts — actual measurements on test stands at specific voltages with specific propellers. The most accurate single-source data available to most builders. Good to ±5–10% depending on QC.
- Independent test data — measurements from third parties (RC Groups, MiniQuadTestBench, Phaser Pony's reviews). Useful when you don't trust the manufacturer or want cross-brand comparison. Good to ±10%.
For Lumipad cohort work, the right workflow is: use the manufacturer's thrust chart for the motor + prop + battery combination you're considering. Sanity-check it against momentum theory (it should be within 10–20% of the theoretical hover thrust). Apply environmental corrections from Section 9. Don't over-trust any single number to better than ±15% in the field.
What momentum theory still gives you
Even though we'll mostly use empirical data for actual numbers, the momentum-theory intuitions matter every time you make a design choice:
- Want longer flight time? Bigger prop, lower disc loading. Momentum theory tells you why.
- Want more thrust margin? Either more disc area (bigger prop, more motors) or higher induced velocity (more power). The second is expensive.
- Adding payload? Disc loading goes up, induced power goes up faster than linearly with weight. A drone that hovers comfortably at 600g may struggle with 900g.
- Operating at altitude? Air density drops, exit velocity must increase to maintain thrust, induced power increases. Section 9 quantifies this.
The numbers come from data. The reasoning comes from theory. Use both.
Reading a thrust chart, properly.
The motors deep dive covered thrust charts at the surface level. This section goes deeper: how to verify a chart against momentum theory, how to interpolate when your prop or battery doesn't exactly match the test conditions, and how to spot manufacturer charts that don't add up.
Recall the EMAX 2207 2400 KV thrust chart, with HQProp 5×4.3×3 on a 4S battery at 16.0 V (under load):
Sanity-check against momentum theory
Take the 100% throttle row: 1,490 g of thrust = 14.6 N. The induced velocity required:
At the more relaxed 50% throttle row: 620 g (6.08 N), measured power 121 W. Theory predicts:
What's actually being measured
Reputable thrust chart measurements are taken on a static test stand: motor mounted to a load cell, propeller spinning in still air, battery at nominal voltage, ambient temperature ~22°C, sea-level pressure. The chart is the truth at those conditions. Your real flight is different from those conditions in several ways:
- Battery voltage sags under load — the chart shows nominal voltage; real flight sees 5–10% lower under high throttle.
- Forward flight — once moving forward, the propeller sees its own air-relative speed, which reduces thrust slightly.
- Air temperature — Mindanao midday at 32°C is ~3% less dense than 22°C standard.
- Air density at altitude — Section 9 covers this in detail.
- Pack age — older packs sag more, deliver less peak voltage.
A reasonable approximation for "real-world" thrust = chart × 0.90 in good conditions, × 0.80 in challenging conditions (hot day, slightly old battery, tropical altitude). Section 9 gives you the exact corrections.
How to spot dishonest charts
Most reputable manufacturers (EMAX, T-Motor, iFlight) publish reasonable, consistent thrust data. Some smaller brands publish charts that are "best case" or simply optimistic. Sanity checks:
- Efficiency at 50% throttle should be 4–6 g/W for a well-matched 5". If a chart claims 8+ g/W, doubt it.
- Peak thrust at 100% should be ~2.4× the 50% thrust. If a chart claims 100% is 3.5× the 50% thrust, doubt it.
- Current at 100% throttle should be roughly 4× the current at 50%. Doubling thrust roughly quadruples current; a linear current claim is wrong.
- Power = Voltage × Current. If the chart's power column doesn't match this calculation within rounding, the chart's been edited inconsistently and is unreliable.
For most cohort builds, sticking with EMAX, T-Motor, iFlight, and HQProp avoids these problems entirely. The charts from these manufacturers match independent testing within ~10%.
Calculating thrust-to-weight ratio.
Thrust-to-weight (T:W) is the dimensionless ratio of total static thrust at full throttle to total all-up weight. It's the best single predictor of how a drone will feel in flight, the practical upper limit of payload, and the manoeuvring envelope. The calculation is simple. The interpretation is what matters.
The formula:
W_total · all-up weight including battery, payload, and any sensors
Both in the same units (grams or Newtons).
For the cohort default Lumipad Quad v1 with the NDVI rig:
What different T:W ratios mean in flight
Why static T:W exaggerates real capability
The 9.6:1 we calculated for the Lumipad Quad v1 is the spec-sheet, sea-level, fresh-battery, full-throttle ratio. By the time you fly a real survey on a hot Mindanao afternoon with a battery that's already at 50% charge from the previous flight, real available thrust is closer to 75% of spec — making real T:W around 7:1 instead of 9.6:1. Still plenty, but the number is smaller than it looks on paper.
- Build with at least 2.5:1 spec T:W to maintain real-world ≥2:1.
- Build with at least 3.5:1 spec T:W for comfortable survey work in field conditions.
- Build with 5:1+ spec T:W if you'll be flying in strong wind or carrying variable payloads.
For most cohort survey work, the targets are: 3.5–4.5:1 spec T:W at the start, dropping to ~3:1 by the end of a long survey day as the battery and the air both warm up.
Payload & lift margin.
A common partner-org question: "Can we add a thermal camera to the existing build?" The answer comes from understanding how payload reduces T:W, increases hover throttle, and disproportionately reduces flight time. Adding a 100g sensor isn't just a 100g problem — it cascades through the energy budget.
The relationship between payload and performance has three coupled effects:
- Hover throttle increases — more weight means each motor has to produce more thrust, which means more throttle, which means more current and more voltage sag.
- Maneuvering margin decreases — with the motors closer to peak thrust just to hover, less "headroom" remains for climbs, sharp turns, or wind compensation.
- Flight time decreases more than proportionally — power consumption rises faster than linearly with thrust. Adding 20% payload doesn't reduce flight time by 20%; it reduces it by closer to 30%.
Calculating hover throttle from a thrust chart:
More precisely: interpolate the thrust chart to find the throttle where measured thrust = T_required.
For the cohort default with the NDVI rig (620 g all-up):
Adding a thermal camera
Now suppose we want to add a FLIR Lepton 3.5 thermal module — about 90g for the module + housing + interface board.
The payload margin formula
For a quick estimate of "how much more can I add," use:
The 100g safety margin gives space for wind, maneuvering, and battery sag.
For the cohort default: W_max ≈ (4 · 620) - 100 = 2,380 g. Empty is ~430g. Max sustained payload: ~1,950 g in theory.
Why "max payload" isn't really 1,950 g for a 5"
The math says 1,950 g is the theoretical max payload. In practice, no one would fly a 5-inch with a 2 kg payload. Reasons:
- Flight time becomes too short — at sustained 50% throttle, flight time drops to ~3 minutes per battery. Not useful for survey work.
- No margin for wind or maneuvers — a gust would push throttle above 50%, into the inefficient regime.
- Frame stress — 5-inch airframes weren't designed for 2 kg of payload. Arms flex, screws loosen, things break.
- Crash energy — a 2.4 kg drone falling from 50m makes a much bigger impact than a 620g one. Safety margins matter.
Practical guidance: for survey work, target hover throttle ≤35% with payload. For the cohort default 5-inch, this means total payload (battery + sensors) ≤ ~750g all-up. Anything beyond this, move to the 7-inch frame which is designed for it.
Top speed & drag.
Quadcopter top speed is determined by an interaction between two opposing forces: the propeller's pitch-limited maximum forward thrust, and the airframe's aerodynamic drag rising with the square of speed. Understanding both lets you predict top speed from a build, and explains why a drone with twice the thrust isn't twice as fast.
Two physical limits set top speed. The first is pitch speed — the maximum forward velocity at which the propeller still produces net thrust. Above pitch speed, the air is moving past the propeller faster than the prop can grip it; the prop "spins out" and produces zero or negative thrust.
RPM_max · max motor RPM in revolutions per second
η_pitch · pitch efficiency, typically ~0.7 for drone props
For the cohort default 5×4.3 prop on EMAX 2207 2400 KV motors with 4S battery:
The second limit is aerodynamic drag — the air resistance the airframe pushes through. Drag rises with the square of velocity:
v · forward velocity
C_d · drag coefficient (~0.5–0.8 for typical drone airframes — they're not aerodynamic)
A_frontal · frontal cross-section area presented to the airflow
Top speed is reached when forward thrust equals drag:
For the cohort default at full forward tilt (~30° from horizontal, putting most thrust into forward motion):
Why doubling thrust doesn't double speed
Drag is quadratic in velocity. To double top speed (94 → 188 km/h) requires quadrupling the forward thrust. Most builds can't do that — they'd need a much higher-pitch prop (which they can't run at the existing motor's torque) or much bigger motors and battery (which adds weight, increasing drag).
- From 85 km/h to 100 km/h: ~40% more thrust required.
- From 100 km/h to 120 km/h: ~45% more thrust required (compounding).
- From 120 km/h to 150 km/h: at this point pitch speed becomes the binding constraint regardless of thrust.
This is why race drones use radically different setups (5×5.0 or higher pitch on 6S batteries) — they're trying to push pitch speed up enough that thrust can keep pace with rising drag.
Top speed isn't usually the binding constraint
For survey work, top speed is rarely the limiting factor. Survey missions cruise at 5–15 m/s (18–54 km/h) for stability and image quality. You're nowhere near top speed. So the practical question for cohort work is rarely "how fast" but "what cruise speed gives the best efficiency?"
- The cohort default cruises efficiently at 5–10 m/s (18–36 km/h).
- Pushing past 15 m/s (54 km/h) starts to consume battery faster than the time saved is worth.
- For long-range 7-inch builds, optimal cruise can be 10–15 m/s — mostly because of better aerodynamic shape, not because the props can spin faster.
- For 10-inch heavy-payload, cruise stays low (5–10 m/s) — the bigger frontal area makes higher speeds prohibitively expensive.
Three worked builds.
Putting Sections 1–7 together: end-to-end thrust, T:W, hover throttle, payload margin, and top-speed calculations for the cohort default 5-inch, 7-inch, and 10-inch builds. These are the numbers behind the at-a-glance comparison strip in the parts primer — derived from the physics, cross-checked against manufacturer charts, corrected for typical Mindanao field conditions.
Build 1: 5-inch combo 5A (cohort default with NDVI rig)
EMAX 2207 2400 KV motors, HQProp 5×4.3×3 props, CNHL 4S 1500 mAh 95C battery, with the standard NDVI rig.
Build 2: 7-inch combo 7A (long-range standard with NDVI rig)
iFlight XING 2806.5 1300 KV motors, HQProp 7×4.5×3 props, CNHL 6S 2200 mAh 75C battery, with the standard NDVI rig.
Build 3: 10-inch combo 10A (heavy-payload standard)
T-Motor MN3115 900 KV motors, T-Motor 10×4.5 carbon props, CNHL 6S 5000 mAh 30C battery, with a multispectral sensor payload.
What the three builds tell you
- 5-inch has the highest spec T:W (9.6) but the lowest absolute thrust (5.96 kg). Great manoeuvrability, modest payload margin.
- 7-inch has middling T:W (2.8) and the most balanced overall envelope. The "all-rounder" of the three.
- 10-inch has the most absolute thrust (8.6 kg) but the highest static load. Designed for payload, not for performance margin.
Notice that absolute T:W ratios converge at the high-payload end (~2:1 real-world) but separate at the maneuvering end (5" is much more agile). This is the physics speaking: once you're loaded for the payload you actually have to carry, the maneuvering envelope is similar across frame sizes. The 5" is more agile not because of better motors, but because it's lighter — fewer kilograms of inertia to fight.
Environmental corrections.
Manufacturer thrust charts are measured at sea level, in cool air, with fresh batteries. Mindanao field surveys are at sea level (ish) but in 28–34°C heat, sometimes at altitude (Bukidnon's plantations sit at 600–1,200m), often with high humidity, frequently with batteries that have already done a flight that morning. Each of these reduces real-world thrust below spec. This section quantifies the corrections.
Thrust scales linearly with air density (T ∝ ρ), so all environmental corrections come back to "what's the local air density vs the chart's reference?". The reference is 1.225 kg/m³ at 15°C, sea level, dry air. Every deviation from that reduces thrust.
Bukidnon plateau (1,000m): −12%.
Mt Apo summit (2,950m): −30%.
Early morning at 24°C: 9°C above ref → −3.5%.
Late flight, older pack: −10%.
Aggressive forward flight at 25 m/s: −10%.
Combining the corrections multiplicatively for a typical Mindanao plantation survey:
What this means for build planning
- Build margin into spec T:W. If you want to fly with 2:1 real T:W, target 2.6–2.7:1 on paper.
- Do payload tests on the actual flight day. Conditions vary; spec calculations are starting points, not commitments.
- Plan flight time in real conditions, not spec. A 6-minute spec-sheet endurance is realistically 5 minutes in field heat.
- For altitude operations (Bukidnon, Cordillera partner orgs), the correction matters a lot more — 1,200m elevation is a 14% thrust reduction before any other factor. Build 7-inch instead of 5-inch when working consistently above 1,000m.
Build your own thrust model.
Everything in the previous nine sections fits into a one-sheet spreadsheet that takes a BOM as input and outputs T:W, hover throttle, payload margin, top speed, and corrected real-world numbers. Building it once, for your own preferred frames, lets you evaluate any future build decision in minutes instead of hours.
What the spreadsheet should compute:
- Inputs: motor model + KV, prop diameter + pitch + blade count, battery S + mAh + C-rating, sensor payload, frame size, target altitude.
- Lookup: motor max thrust at the relevant cell count from manufacturer chart.
- Calculate: all-up weight, T_max, T:W, hover throttle, max payload, theoretical pitch speed, drag-limited top speed.
- Apply corrections: air density at target altitude, temperature factor, battery sag, forward flight loss.
- Output: "Real-world expected" values that account for environmental conditions.
The Lumipad workshop maintains a reference spreadsheet built around this model, calibrated against the cohort fleet's flight log data. It's a 12-row Excel file that any alumnus can clone for their own builds. Three versions exist:
- Quick estimate (12 rows) — single-frame quick check; takes 30 seconds with motor and prop in hand.
- Multi-build comparison (60 rows) — compare 3 candidate builds side-by-side. Used when alumni are planning a new frame and want to see options.
- Mission planner (250 rows) — plug in survey area, altitude, expected wind, and battery count. Outputs total mission flight time and battery requirements. Used by partner orgs planning multi-day surveys.
All three are downloadable from the link below. They run in Excel, LibreOffice Calc, or Google Sheets. Total time investment to learn: 15–30 minutes. Total time saved across a year of build decisions: significant.
Three habits that pay off
- Run the calculation before buying parts. A $5 spreadsheet check prevents $300 of wasted parts on incompatible combinations.
- Calibrate against your own flight logs. After a few flights, you'll know the actual thrust correction for your specific environment. Update the spreadsheet's correction factor; future predictions get more accurate.
- Share the model with the cohort network. Alumni in different regions have different correction factors (altitude, climate). The spreadsheet improves when more people contribute their data.
This is what distinguishes "an enthusiast who builds drones" from "an engineer who designs flying systems." The math is identical. The difference is having a practiced workflow that turns physics into build decisions.