Every derived quantity reduces to some combination of powers of the seven fundamental (base) quantities -- length [L], mass [M], time [T], electric current [A], temperature [K], amount of substance [mol], luminous intensity [cd] -- and this combination, written in square brackets, is called the quantity's dimensional formula. Writing that formula as an equation (e.g. [v]=[M0LT−1]) is the dimensional equation.
Worked derivation (velocity): velocity=timedisplacement=[T][L]=[M0LT−1] -- dimension 0 in mass, 1 in length, −1 in time.
Dimensional formulas of common quantities follow directly from their defining relation: force = mass × acceleration =[MLT−2]; work = force × distance =[ML2T−2]; Planck's constant h= energy/frequency =[ML2T−2]/[T−1]=[ML2T−1]; the gravitational constant G (from Newton's law F=Gm1m2/r2) =[force×distance2]/mass2=[M−1L3T−2]; torque, being force × distance just like work, shares work's dimensional formula [ML2T−2] -- so torque and energy, despite measuring physically different things, have the same dimensions.
Classifying by dimension. Every physical quantity falls into exactly one of four boxes:
- Dimensional variables -- have dimensions, and take variable values (length, velocity, acceleration).
- Dimensionless variables -- no dimensions, but variable values (specific gravity, strain, refractive index).
- Dimensional constants -- have dimensions, but a fixed value (G, h).
- Dimensionless constants -- neither dimensioned nor variable (π, e, pure counting numbers).
Principle of homogeneity of dimensions. Every additive term in a physically valid equation must carry the same dimensions -- you cannot add a length to a time. E.g. in v2=u2+2as, the terms v2, u2 and 2as all reduce to [L2T−2]. This principle is exactly what makes the checking use of dimensional analysis possible: if two sides (or two additive terms) of a proposed equation come out with different dimensions, the equation is definitely wrong -- though matching dimensions alone does not guarantee the equation is numerically correct (see 'Dimensional Analysis').