Classical-Mechanics · Unit 13 · Video 5 · Interactive Practice

Why Leaking Sand Changes Nothing and Blown-In Grain Changes Everything

IKey Formulas

FormulaNameWhat it takes
Fext=m(t)dvdt+dmdt(vu)F_{\text{ext}} = m(t)\dfrac{dv}{dt} + \dfrac{dm}{dt}\,(v - u)Equation of motion with mass flowuu = velocity of the transferred matter in the ground frame
vc(t)=Fbln ⁣[mc+msmc+msbt]v_c(t) = \dfrac{F}{b}\ln\!\left[\dfrac{m_c + m_s}{m_c + m_s - bt}\right]Emptying car — sand out at u=vcu = v_cConstant FF, leak rate bb, released from rest
vA(t)=btmA,0+btuv_A(t) = \dfrac{bt}{m_{A,0} + bt}\,uFilling car — grain in at uuNo external force along the track, starts from rest
(mA,0+bt)vA(t)=btu\left(m_{A,0} + bt\right)v_A(t) = bt\,uMomentum ledger of the filling carEvery kilogram of grain hands over the momentum it arrived with

Key Insight: p=mvp = mv was defined for a fixed collection of matter, so differentiating it with the product rule on a system whose mass is changing is illegal — the stray vdm/dtv\,dm/dt even changes value from one reference frame to another. Follow the same matter through Δt\Delta t instead, and the only survivor is dmdt(vu)\dfrac{dm}{dt}(v - u): mass rate times relative velocity.

IISand Out: The Leak the Equation Never Feels

Sand drops with the car's own horizontal velocity, so only a shrinking mc(t)m_c(t) survives in F=mc(t)dvc/dtF = m_c(t)\,dv_c/dt.

IIIGrain In: How Fast Once the Mass Has Doubled?

Grain arrives carrying uu; how fast is car A once that grain has doubled its mass?

💡 Integrating in time instead of in mass gives the identical curve: dvAuvA=bdtmA,0+bt\dfrac{dv_A}{u - v_A} = \dfrac{b\,dt}{m_{A,0} + bt}, where the antiderivative's factor 1/b1/b cancels the bb upstairs.

IVOne Term Decides

Every mass-flow problem turns on one term: the mass rate times the relative velocity uvu - v.

VQuiz Questions

Problem 1 · Emptying the Freight Car

Given: a car of mass mc=3000 kgm_c = 3000\ \text{kg} carrying ms=2000 kgm_s = 2000\ \text{kg} of sand is released from rest under a constant force F=2000 NF = 2000\ \text{N}, while sand streams straight down through a floor port at b=100 kg/sb = 100\ \text{kg/s}find the car's speed at the instant the last sand leaves.

✅ Correct! The same 2000 N2000\ \text{N} buys more acceleration every second as the mass falls from 5000 kg5000\ \text{kg} to 3000 kg3000\ \text{kg}: (F/b)ln(5/3)=20(0.511)=10.2 m/s(F/b)\ln(5/3) = 20(0.511) = 10.2\ \text{m/s}.
❌ That is the constant-mass answer. Ft/(mc+ms)=2000(20)/5000Ft/(m_c+m_s) = 2000(20)/5000 would hold only if the car never got lighter; the real car ends up faster than 8.0 m/s8.0\ \text{m/s}.
❌ That treats the car as empty the whole way. Ft/mc=2000(20)/3000Ft/m_c = 2000(20)/3000 is an upper bound — the true answer must lie between the full-mass and empty-mass estimates.
❌ Check what the denominator collapses to. At t=ms/bt = m_s/b the remaining mass is mc+msbt=mc=3000 kgm_c + m_s - bt = m_c = 3000\ \text{kg}, not ms=2000 kgm_s = 2000\ \text{kg}.
❌ Not quite. Evaluate vc=(F/b)ln[(mc+ms)/(mc+msbt)]v_c = (F/b)\ln[(m_c+m_s)/(m_c+m_s-bt)] at t=ms/b=20 st = m_s/b = 20\ \text{s}.
Show solution

The sand leaves through the floor with the car's own horizontal velocity, so no dm/dtdm/dt term appears and Newton's second law keeps its familiar shape with a time-dependent mass:

F=mc(t)dvcdt,mc(t)=mc+msbt=5000100tF = m_c(t)\,\frac{dv_c}{dt},\qquad m_c(t) = m_c + m_s - bt = 5000 - 100t

Separate the variables and integrate from rest:

0vc(t)dvc=0tFdtmc+msbt    vc(t)=Fbln ⁣[mc+msmc+msbt]\int_0^{v_c(t)} dv_c' = \int_0^t \frac{F\,dt'}{m_c + m_s - bt'} \;\Longrightarrow\; v_c(t) = \frac{F}{b}\ln\!\left[\frac{m_c+m_s}{m_c+m_s-bt}\right]

The sand runs out when bt=msbt = m_s, i.e. at t=ms/b=20 st = m_s/b = 20\ \text{s}, where the denominator collapses to the empty car mcm_c:

vc=2000100ln ⁣[50003000]=20ln ⁣(53)=20(0.5108)=10.2 m/sv_c = \frac{2000}{100}\ln\!\left[\frac{5000}{3000}\right] = 20\ln\!\left(\frac{5}{3}\right) = 20(0.5108) = 10.2\ \text{m/s}

Sanity check. The full car (5000 kg5000\ \text{kg}) would have reached 8.0 m/s8.0\ \text{m/s} and the empty car (3000 kg3000\ \text{kg}) would have reached 13.3 m/s13.3\ \text{m/s}; 10.2 m/s10.2\ \text{m/s} sits between them, exactly where a mass that shrinks from one to the other belongs.

Problem 2 · The Product-Rule Trap

Given: at one instant that same leaking car has mass mc(t)=4000 kgm_c(t) = 4000\ \text{kg} and speed vc=5.0 m/sv_c = 5.0\ \text{m/s}, with F=2000 NF = 2000\ \text{N} still applied and dmc/dt=100 kg/sdm_c/dt = -100\ \text{kg/s}find its acceleration.

✅ Correct! The sand separates at the car's own horizontal velocity, so vcu=0v_c - u = 0, the mass-flow term dies, and F=mc(t)dvc/dtF = m_c(t)\,dv_c/dt gives 2000/40002000/4000.
❌ That is the product-rule shortcut. Writing F=mdv/dt+vdm/dtF = m\,dv/dt + v\,dm/dt and solving gives (Fvdm/dt)/m=(2000+500)/4000(F - v\,dm/dt)/m = (2000 + 500)/4000. But p=mvp = mv was defined for a fixed collection of matter, and that leftover term is no thrust — it even changes value when you change reference frames.
❌ The same shortcut with the opposite sign: (F+vdm/dt)/m=1500/4000(F + v\,dm/dt)/m = 1500/4000. The term should be absent altogether, not merely re-signed.
❌ That is F/mcF/m_c with the empty-car mass 3000 kg3000\ \text{kg}. At this instant the car still carries 1000 kg1000\ \text{kg} of sand, so the mass in the denominator is 4000 kg4000\ \text{kg}.
❌ Not quite. With u=vcu = v_c the equation of motion is simply F=mc(t)dvc/dtF = m_c(t)\,dv_c/dt.
Show solution

Follow a fixed collection of matter — the car plus the sand still inside at time tt. The element that drops out separates while moving horizontally with the car, so at t+Δtt + \Delta t both pieces share the velocity vc+Δvcv_c + \Delta v_c and the mass changes cancel:

Δpsys=(Δms+mc+Δmc)(vc+Δvc)mcvc=mc(t)Δvc(Δmc=Δms)\Delta p_{\text{sys}} = \big(\Delta m_s + m_c + \Delta m_c\big)(v_c + \Delta v_c) - m_c v_c = m_c(t)\,\Delta v_c \quad (\Delta m_c = -\Delta m_s) F=mc(t)dvcdt=20004000=0.50 m/s2F = m_c(t)\frac{dv_c}{dt} = \frac{2000}{4000} = 0.50\ \text{m/s}^2

Why the shortcut fails. The general mass-flow term is dmdt(vu)\dfrac{dm}{dt}(v - u), the mass rate times the relative velocity. Sand falling through the floor has u=vcu = v_c, so the term is exactly zero. The product rule instead keeps vdm/dt=(5.0)(100)=500 Nv\,dm/dt = (5.0)(-100) = -500\ \text{N}, an invented force of 500 N500\ \text{N} — and its value would change if you measured vcv_c from a moving platform, a sure sign it is not a law.

Corollary. Set F=0F = 0 and the result is dvc/dt=0dv_c/dt = 0: a rolling, leaking car holds its speed exactly, however much sand it sheds.

Problem 3 · Filling the Freight Car

Given: car A starts from rest with mA,0=2000 kgm_{A,0} = 2000\ \text{kg} while grain blown from car B arrives at b=50 kg/sb = 50\ \text{kg/s} carrying car B's horizontal velocity u=12 m/su = 12\ \text{m/s}; no external force acts along the track.

What is car A's speed at t=20 st = 20\ \text{s}?

How much grain must be aboard for car A to reach vA=9.0 m/sv_A = 9.0\ \text{m/s}?

✅ Correct! vA/uv_A/u is the fraction of car A's current mass that arrived as grain: 1000/30001000/3000 at 20 s20\ \text{s}, and reaching 0.75u0.75u demands 6000 kg6000\ \text{kg} of grain inside an 8000 kg8000\ \text{kg} car.
uu is the limit, not the value. vA=[bt/(mA,0+bt)]uv_A = [bt/(m_{A,0}+bt)]u climbs toward car B's speed but reaches it only as tt \to \infty.
❌ The fraction is upside down. vA/u=bt/(mA,0+bt)=1000/3000v_A/u = bt/(m_{A,0}+bt) = 1000/3000 — the grain's share of the mass, not the car's share mA,0/mA=2000/3000m_{A,0}/m_A = 2000/3000.
❌ You divided by the initial mass. btu/mA,0=12000/2000bt\,u/m_{A,0} = 12000/2000 hands all the momentum to the original car; the grain aboard shares it, so divide by mA(20)=3000 kgm_A(20) = 3000\ \text{kg}.
❌ Check the ledger. mA(t)vA(t)=btum_A(t)\,v_A(t) = bt\,u with mA(20)=2000+1000=3000 kgm_A(20) = 2000 + 1000 = 3000\ \text{kg} and btu=12000 kgm/sbt\,u = 12000\ \text{kg}\cdot\text{m/s}.
❌ That is car A's total mass at that moment. The grain aboard is 80002000=6000 kg8000 - 2000 = 6000\ \text{kg}.
❌ Not enough grain. With 3000 kg3000\ \text{kg} aboard the mass is 5000 kg5000\ \text{kg} and vA=(3000/5000)(12)=7.2 m/sv_A = (3000/5000)(12) = 7.2\ \text{m/s}.
❌ That is 0.75mA,00.75\,m_{A,0}. The ratio 0.750.75 is a fraction of the current mass, which the grain has itself inflated.
❌ Not quite. Solve bt/(mA,0+bt)=9.0/12=0.75bt/(m_{A,0}+bt) = 9.0/12 = 0.75 for the grain mass btbt.
Show solution

The system is car A, the grain already aboard, and the element Δmg\Delta m_g about to arrive — the same matter at both instants. With no external force along the track,

0=mA(t)dvAdt+dmAdt(vAu)0 = m_A(t)\frac{dv_A}{dt} + \frac{dm_A}{dt}\big(v_A - u\big)

Separating against mass rather than time and integrating gives ln[u/(uvA)]=ln[mA/mA,0]\ln[u/(u - v_A)] = \ln[m_A/m_{A,0}], hence

vA=mAmA,0mAu=btmA,0+btuv_A = \frac{m_A - m_{A,0}}{m_A}\,u = \frac{bt}{m_{A,0} + bt}\,u

Part 1. At t=20 st = 20\ \text{s} the grain aboard is bt=1000 kgbt = 1000\ \text{kg} and mA=3000 kgm_A = 3000\ \text{kg}:

vA=10003000(12)=4.0 m/sv_A = \frac{1000}{3000}(12) = 4.0\ \text{m/s}

Ledger check: (3000)(4.0)=12000=(1000)(12)(3000)(4.0) = 12000 = (1000)(12) — the car carries exactly the momentum the grain brought in.

Part 2. Demand vA/u=9.0/12=0.75v_A/u = 9.0/12 = 0.75:

bt2000+bt=0.75    1500+0.75bt=bt    bt=6000 kg\frac{bt}{2000 + bt} = 0.75 \;\Longrightarrow\; 1500 + 0.75\,bt = bt \;\Longrightarrow\; bt = 6000\ \text{kg}

So mA=8000 kgm_A = 8000\ \text{kg}, reached at t=6000/50=120 st = 6000/50 = 120\ \text{s}. Ledger check: (8000)(9.0)=72000=(6000)(12)(8000)(9.0) = 72000 = (6000)(12).

Problem 4 · Sand from a Hopper at Rest

Given: a 2000 kg2000\ \text{kg} car rolls freely at 4.0 m/s4.0\ \text{m/s} on level track when sand starts pouring in from a hopper hanging at rest above the track at 50 kg/s50\ \text{kg/s}, with no applied force — find the car's speed 20 s20\ \text{s} later.

✅ Correct! The sand arrives with u=0u = 0, so vu=vv - u = v: the term is alive and brakes the car. Horizontal momentum is conserved, (2000)(4.0)=(3000)v(2000)(4.0) = (3000)v.
❌ That is the leaking car's answer. It worked only because the departing sand left at the car's own velocity; sand arriving at u=0u = 0 must be dragged up to speed, and only the car's momentum can pay for it.
❌ The mass ratio is upside down. v=v0m0/m(t)v = v_0\,m_0/m(t), so growing from 2000 kg2000\ \text{kg} to 3000 kg3000\ \text{kg} must make the car slower, not faster.
❌ Only 1000 kg1000\ \text{kg} was added, not 2000 kg2000\ \text{kg}. The mass goes 20003000 kg2000 \to 3000\ \text{kg}, so the speed falls by the factor 2/32/3, not 1/21/2.
❌ Not quite. Sand falling from a hopper at rest brings no horizontal momentum, so m(t)v(t)m(t)v(t) stays at its initial 8000 kgm/s8000\ \text{kg}\cdot\text{m/s}.
Show solution

Mass arrives, so the transfer term is present; what matters is the velocity it arrives with. Here u=0u = 0, and with no external force along the track:

0=m(t)dvdt+dmdt(v0)    mdvdt=bv0 = m(t)\frac{dv}{dt} + \frac{dm}{dt}\big(v - 0\big) \;\Longrightarrow\; m\frac{dv}{dt} = -bv

That is exactly d(mv)/dt=0d(mv)/dt = 0: the horizontal momentum of car-plus-sand is conserved because the incoming sand brings none.

m(t)=2000+50(20)=3000 kgm(t) = 2000 + 50(20) = 3000\ \text{kg} v=m0v0m(t)=(2000)(4.0)3000=2.7 m/sv = \frac{m_0v_0}{m(t)} = \frac{(2000)(4.0)}{3000} = 2.7\ \text{m/s}

The three cases side by side. Sand out through the floor has u=vu = v, the term vanishes, and the speed holds. Grain blown in from car B has u>vu > v, the term acts like a forward thrust, and the speed climbs toward uu. Sand poured from a hopper at rest has u=0<vu = 0 < v, the term brakes, and the speed decays like 1/m(t)1/m(t). One term, dmdt(vu)\dfrac{dm}{dt}(v-u), decides all three.

Solved: 0 / 4