Home / Courses / IIT JAM Physics / Practice / The variational method and WKB

The variational method and WKB

Quantum mechanics · 6 questions and templates for IIT JAM Physics, in order from warm-up to top rank. Check each answer, then read the full solution.

Loading questions… (they need JavaScript; nothing you type is sent anywhere)

Written questions with solutions

The same questions as a list, so you can print them or read without the practice box. Templates (fresh numbers) appear only in the practice box above.

  1. MCQWarm-up

    A student uses a trial function for the hydrogen ground state (exact energy −13.6 eV) and reports a minimised ⟨H⟩=−14.1 eV. The right conclusion is that
    1. the trial function is better than the exact one
    2. there is a mistake: the calculation is wrong or the trial function breaks the rules (not normalised, wrong boundary behaviour)
    3. the method has found an excited state
    4. nothing is wrong, since variational energies can lie on either side of the true value
    Answer and solution

    Answer: B

    For any acceptable normalised trial state, ⟨H⟩≥E0. A value below −13.6 eV is impossible, so something is wrong: a lost normalisation factor, an arithmetic slip, or a function that is not in the domain of H. Excited states lie higher, not lower.

    Why people get this wrong: The variational energy is one-sided: it can only be too high.

  2. NATStandard

    A harmonic oscillator has ℏω=0.40 eV. Bohr-Sommerfeld (WKB) quantisation, which is exact here, gives the number of states with energy at or below 5.0 eV (measured from the bottom of the well) as ______.
    Answer and solution

    Answer: 13

    ∮pdx=2πE/ω=(n+12)h gives En=(n+12)ℏω. We need (n+12)×0.40≤5.0, i.e. n≤12. With n=0,1,…,12 there are 13 states. (Each state takes up an area h of phase space.)

    Why people get this wrong: 5.0/0.40=12.5 suggests 12; counting must start from n=0 and include the zero-point offset.

  3. MCQStandard

    In Bohr-Sommerfeld (WKB) quantisation for a one-dimensional bound particle, the phase-space area enclosed between the orbits of two neighbouring levels is
    1. ℏ
    2. h
    3. h/2
    4. 2πh
    Answer and solution

    Answer: B

    ∮pdx=(n+γ)h, so neighbouring orbits differ by exactly h in enclosed area. Each quantum state "occupies" an area h of phase space, the rule used to count states in statistical mechanics.

    Why people get this wrong: Mixing h and ℏ; the area is 2πℏ=h.

  4. MCQExam level

    Two students apply the variational method to the same Hamiltonian with different normalisable trial functions and get −3.2 eV and −3.5 eV. What can be said about the true ground-state energy E0?
    1. −3.5 eV<E0<−3.2 eV
    2. E0≥−3.2 eV
    3. Nothing, without the exact answer
    4. E0≤−3.5 eV
    Answer and solution

    Answer: D

    Every trial expectation value is an upper bound: E0≤⟨H⟩ for each. Both bounds hold, and the stronger one is the lower value, so E0≤−3.5 eV. The trial giving −3.5 eV is the better estimate of the energy.

    Why people get this wrong: The two results do not bracket E0: the variational method gives bounds from above only.

  5. MSQExam level

    Which statements about the variational method for the ground state are correct?
    1. Enlarging the trial family (keeping the old one as a special case) cannot raise the minimum of ⟨H⟩
    2. The minimum of ⟨H⟩ is an upper bound on the true ground-state energy
    3. The best trial function gives an upper bound on the exact ⟨x2⟩
    4. If the family contains the exact ground state, the minimum equals E0
    Answer and solution

    Answer: A, B, D

    A larger family includes all the old trial states, so its minimum is at most the old one. Every ⟨H⟩≥E0, with equality only for the exact ground state. Nothing bounds other quantities: the optimised function can give ⟨x2⟩ too large or too small, and its error is first order in the wavefunction error while the energy error is second order.

    Why people get this wrong: The bound property belongs only to the energy (the quantity being minimised).

← All topics · Timed mock test →