Centre of mass — definition for system of particles: Visual walkthrough

We will use only these plain ideas, each earned before use:

a position = a number telling you how far along a line something sits (in metres),
a mass = how "heavy" something is (in kilograms, kg),
an average = defined precisely just below.

Everything else we grow from those ideas.

Step 1 — One line, two dots, and the question

WHAT. Put two particles on a horizontal line. Call the left one mass $m_{1}$ sitting at position $x_{1}$ , and the right one mass $m_{2}$ at position $x_{2}$ . "Position" here is just a ruler reading in metres — how far right of a chosen zero mark (the origin) the particle sits.

WHY. Before any formula, we must fix what we are even looking for: a single point $x_{c m}$ that "stands in" for both dots. The simplest arena to find it is one straight line, so we start there and add dimensions later.

PICTURE. In the figure, the blue dot is $m_{1}$ , the orange dot is $m_{2}$ , the gray tick is the origin $0$ . The green triangle underneath is a pivot — imagine the line is a stiff rod resting on that pivot. Our whole quest is: where must the pivot go so the rod balances?

Each symbol, right where it lives:

$x_{1}$ — ruler reading of the blue mass (metres).
$x_{2}$ — ruler reading of the orange mass (metres).
$m_{1}, m_{2}$ — how heavy each is (kg); drawn as dot size.
$x_{c m}$ — the unknown pivot we are hunting for.

Step 2 — Why "just take the midpoint" is a trap

WHAT. The tempting first guess: put the pivot exactly halfway, at $\frac{x _{1} + x _{2}}{2}$ — the plain average of the two positions. The figure shows what happens when the two masses are unequal — the rod tips over toward the heavy side.

WHY. A balance is about turning effect, not distance alone. A heavy mass far out and a light mass near in can still balance — so distance-only averaging must be wrong the moment masses differ. We need the weighted average from the definition above, weighting each position by its mass.

PICTURE. Left panel: equal masses → midpoint balances (rod is level). Right panel: heavy orange mass → the same midpoint pivot tips down on the orange side (red arrow shows the tip). This picture is the seed of every mistake in the parent's "Common Mistakes" list.

Step 3 — The lever law: mass × distance-from-pivot must match

WHAT. Physics of a see-saw: a mass balances if its turning effect on one side equals the other side's. Turning effect = mass $\times$ its distance from the pivot. So balance means $m_{1} lever arm of m_{1} (x_{c m} - x_{1}) = m_{2} lever arm of m_{2} (x_{2} - x_{c m}) .$

Term by term:

$x_{c m} - x_{1}$ — how far the pivot is to the right of the blue mass (its lever arm).
$x_{2} - x_{c m}$ — how far the orange mass is to the right of the pivot (its lever arm).
multiplying each by its mass gives the two turning effects; balance demands they be equal.

WHY this tool (a product, not a sum)? Because a see-saw's tendency to rotate grows with both how heavy the kid is and how far out they sit — doubling either doubles the effect. Only a product $m \times (distance)$ captures "both matter, multiplicatively." This product has a name we meet again: a moment. (See Weighted Average and Moments.)

PICTURE. The two shaded rectangles have the same area $= mass \times lever arm$ . Balance = equal areas. When the orange mass is heavier, its rectangle is taller, so its arm must be shorter — the pivot slides orange-ward to shrink that arm. You can see the pivot being dragged toward the heavy mass.

Step 4 — Solve the balance for the pivot

WHAT. Take the balance equation and unwrap it to isolate $x_{c m}$ .

$m_{1} (x_{c m} - x_{1}) = m_{2} (x_{2} - x_{c m})$

Multiply out both sides (open the brackets): $m_{1} x_{c m} - m_{1} x_{1} = m_{2} x_{2} - m_{2} x_{c m}$

Gather every $x_{c m}$ on the left, everything else on the right: $m_{1} x_{c m} + m_{2} x_{c m} = m_{1} x_{1} + m_{2} x_{2}$

Factor $x_{c m}$ out of the left, then divide: $x_{c m} = \frac{m _{1} x _{1} + m _{2} x _{2}}{m _{1} + m _{2}}$

WHY. We derived the parent's formula from a see-saw alone — no momentum needed yet. It is exactly the weighted-average template with weights $=$ masses. Reading the box:

numerator $m_{1} x_{1} + m_{2} x_{2}$ — each mass tagged with its place, summed (this is the total moment).
denominator $m_{1} + m_{2}$ — the total mass $M$ , which turns the moment back into a plain position.

PICTURE. The figure animates the algebra as a slider: as the orange mass grows from equal to heavy, the boxed value of $x_{c m}$ slides from the midpoint toward $x_{2}$ . The dashed midpoint line stays put so you can watch the true COM peel away from it.

Step 5 — From two particles to $n$ : the same balance, more dots

WHAT. Add a third dot, a fourth, ... $n$ of them. The balance idea doesn't change: sum every mass's moment, divide by the total mass. $x_{c m} = \frac{m _{1} x _{1} + m _{2} x _{2} + \dots + m _{n} x _{n}}{m _{1} + m _{2} + \dots + m _{n}} = \frac{\sum _{i} m _{i} x _{i}}{\sum _{i} m _{i}} = \frac{1}{M} \sum_{i} m_{i} x_{i} .$

Reading the compact symbols:

$\sum_{i}$ — "add up over every particle $i$ " (a tidy shorthand for the long $+ \dots +$ ).
$m_{i} x_{i}$ — one particle's moment: its mass tagged with its place.
$M = \sum_{i} m_{i}$ — the grand total mass.

WHY the sum $\sum$ ? Because turning effects simply add — three kids on a see-saw contribute independently, so their moments pile up by addition. Nothing new is needed; the two-dot law scales up untouched. This is still the weighted average, now over $n$ numbers.

PICTURE. Five dots of different sizes on the line; each contributes a stack of area $m_{i} x_{i}$ . The pivot lands where all the stacks balance — visibly pulled toward the cluster of big dots.

Step 6 — Off the line: same idea per axis, then bundle into a vector

WHAT. Real particles live in a plane (or in space). Give each particle a pair of readings $(x_{i}, y_{i})$ — its ruler reading rightward and upward. Then $x_{c m} = \frac{\sum _{i} m _{i} x _{i}}{M}, y_{c m} = \frac{\sum _{i} m _{i} y _{i}}{M}, z_{c m} = \frac{\sum _{i} m _{i} z _{i}}{M} .$ The last one, $z_{c m}$ , is for full 3D (ruler reading "out of the page"); it's the identical formula on a third axis.

WHY can we do each axis on its own? Because balancing left–right and balancing up–down are separate equations in separate variables. Look at the two-dot balance: the $x$ -equation $m_{1} (x_{c m} - x_{1}) = m_{2} (x_{2} - x_{c m})$ contains only $x$ 's — no $y$ appears in it, and no $x$ appears in the matching $y$ -equation. Sliding the pivot sideways changes only $x$ -turning-effects and leaves every $y$ -term untouched. So the same weighted-average solution works axis by axis, and a third axis $z$ is no different — that is why the parent could write "each axis is handled independently."

Bundling into one symbol. Writing three separate averages is tidy but clunky. We collect a particle's three readings into a single object, its position vector $r_{i} = (x_{i}, y_{i}, z_{i})$ — the little arrow just means "this is a package of coordinates, not one number." Because each coordinate obeys the same weighted average, all three collapse into one line: $r_{c m} = \frac{1}{M} i \sum m_{i} r_{i} = (x_{c m}, y_{c m}, z_{c m}) .$ So $r_{c m}$ (with the arrow) is not a new idea — it is exactly the three scalar averages $x_{c m}, y_{c m}, z_{c m}$ stapled together. This is the header formula, now fully earned.

PICTURE. Three particles in a plane (the parent's example: $1$ kg at $(0, 0)$ , $2$ kg at $(2, 0)$ , $3$ kg at $(0, 2)$ ). We drop each onto the $x$ -axis to balance for $x_{c m} = 0.67$ m, then onto the $y$ -axis for $y_{c m} = 1.0$ m. The crosshair where they meet — the vector $r_{c m}$ , inside the triangle — is the COM.

Step 7 — Moving the origin: the balance point doesn't care

WHAT. Ruler readings depend on where you nailed the zero mark. Slide the whole origin left by an amount $a$ (so every reading becomes $x_{i}^{'} = x_{i} - a$ ). What happens to $x_{c m}$ ? $x_{c m}^{'} = \frac{\sum _{i} m _{i} ( x _{i} - a )}{M} = \frac{\sum _{i} m _{i} x _{i}}{M} - \frac{a \sum _{i} m _{i}}{M} = x_{c m} - a .$

Term by term:

$\sum_{i} m_{i} x_{i} - a \sum_{i} m_{i}$ — split the sum; the shift $a$ factors out because it is the same for every particle.
$\frac{a \sum _{i} m _{i}}{M} = a$ — since $\sum_{i} m_{i} = M$ , the whole correction is just $a$ .

WHY this matters. The computed $x_{c m}$ shifts by exactly the same $a$ as everything else. That is translation invariance: the COM is a physical balance point stuck to the particles, not an artefact of your ruler. Move the ruler, and the number labelling the pivot moves with it — but the pivot stays wedged among the same dots. (This is precisely why Step 3 in the parent could "absorb the constant of integration by fixing the origin.")

PICTURE. The same masses drawn twice: once with the origin on the left, once with the origin shifted right. In both, the green pivot sits at the identical physical spot; only its printed coordinate changes by $a$ .

Step 8 — Edge cases: the picture must never break

WHAT & WHY. A good derivation survives every extreme. Three to check:

Equal masses. Weights cancel → the weighted average becomes the plain midpoint. The pivot sits dead centre; both rectangles are equal by symmetry.
One mass dominates ( $m_{1} = 1$ kg at $0$ , $m_{2} = 99$ kg at $10$ ). The tiny mass barely tugs; $x_{c m} = \frac{1 ( 0 ) + 99 ( 10 )}{100} = 9.9$ m — a hair short of the giant.
Empty-space COM. Two masses at $x = 0$ and $x = 10$ with nothing between: COM sits at $5$ m where no particle exists. The pivot is a geometric balance point, not a physical object.

PICTURE. Three mini-panels, one per case, each showing the pivot settling exactly where the maths says — including the lonely pivot floating in empty space in panel 3.

Step 9 — Why this exact point is the one Newton loves

WHAT. We built $x_{c m}$ from a see-saw. But the parent derived the same formula from momentum. First, the plain-words meaning of the momentum symbols:

$v_{i}$ — the velocity of particle $i$ : how fast and in which direction it moves, i.e. the rate its position changes, $v_{i} = \frac{d r _{i}}{d t}$ ("how much $r_{i}$ moves each second").
$P$ — the system's total momentum, the mass-weighted sum of all velocities: $P = \sum_{i} m_{i} v_{i}$ . Momentum is "mass in motion" — what Newton's laws actually push on.
$v_{c m}$ — the velocity of the COM point itself, $v_{c m} = \frac{d r _{c m}}{d t}$ .

Now differentiate our earned box $r_{c m} = \frac{1}{M} \sum_{i} m_{i} r_{i}$ once with respect to time (masses are constant): $M v_{c m} = M \frac{d r _{c m}}{d t} = \sum_{i} m_{i} \frac{d r _{i}}{d t} = \sum_{i} m_{i} v_{i} = P .$

WHY the two approaches must agree: the mass-weighted average is the unique combination whose time-rate equals $P / M$ . So our see-saw point is automatically the momentum point — the place where the messy swarm behaves like one particle of mass $M$ (leading to Newton's Second Law for a System of Particles, Conservation of Linear Momentum and the clean Collisions — Elastic and Inelastic frame).

PICTURE. Left: many particles each with a little velocity arrow $v_{i}$ . Right: one big dot of mass $M$ at the COM carrying the summed arrow $P = M v_{c m}$ . The two panels are declared equivalent by the equals sign between them.

The one-picture summary

Everything above, compressed: dots of different sizes on a line, each casting a moment $m_{i} x_{i}$ ; a pivot dragged toward the heavy cluster; the boxed formula riding on top; and the promise "this point moves like one mass $M$ ."

Recall Feynman retelling — the whole walkthrough in plain words

Picture a broomstick with weights taped along it. You want the one spot where it balances on your finger. A heavy weight yanks that balance spot toward itself; a light weight hardly matters. To find the spot, you tag each weight with where it sits — multiply its mass by its position — add all those tags up, then divide by the total weight so you get back a plain "where" instead of a "mass-times-where." That "divide the tagged sum by the total weight" move is just a weighted average. Do it left–right and up–down (and in–out for 3D) separately, because sliding your finger sideways can't fix an up–down wobble — then bundle the three answers into one arrow $r_{c m}$ . Slide your whole ruler over and the printed number changes, but the balance spot stays wedged among the same weights. The answer can even land in thin air between the weights — it's just a balance point, not a thing. And here's the punchline: if you now toss the whole broomstick spinning across the room, that balance point flies in a clean arc as if the entire broom were one dot sitting right there. That's why we bother inventing it.

Recall Rapid self-check

Balance law for two masses on a line? ::: $m_{1} (x_{c m} - x_{1}) = m_{2} (x_{2} - x_{c m})$ — equal moments about the pivot. Solve it for $x_{c m}$ . ::: $x_{c m} = \frac{m _{1} x _{1} + m _{2} x _{2}}{m _{1} + m _{2}}$ . Why split into $x$ , $y$ , $z$ separately? ::: Each axis is its own balance equation in its own variable; a sideways shift can't cure an up–down imbalance. What does the arrow in $r_{c m}$ mean? ::: It bundles the three scalar averages $x_{c m}, y_{c m}, z_{c m}$ into one package. If you shift the origin by $a$ , what happens to $x_{c m}$ ? ::: It shifts by exactly $a$ — the physical balance point is unchanged (translation invariance). $1$ kg at $0$ , $99$ kg at $10$ : where is COM? ::: $9.9$ m — nearly at the heavy mass. Define $P$ in words. ::: Total momentum, the mass-weighted sum of velocities $\sum_{i} m_{i} v_{i}$ .

Connections

Parent: Centre of mass — definition (Hinglish)
Weighted Average and Moments — the see-saw "moment" $m_{i} x_{i}$ is the same maths.
Newton's Second Law for a System of Particles — Step 9 flows into $F_{e x t} = M a_{c m}$ .
Conservation of Linear Momentum — $v_{c m}$ constant when no external force.
Collisions — Elastic and Inelastic — cleanest in the COM frame.
Centre of Mass of Continuous Bodies — swap $\sum$ for $\int$ when dots become a smear.