# I Simple harmonic oscillator

#### koustav

• In the series solution of simple harmonic oscillator,why do we have a factor of 1/2 in the trial solution?

Related Quantum Physics News on Phys.org

#### BvU

Homework Helper
There's a lot more factors there. Could you be a bit more specific and post the trial solution you are referring to ?

BvU

#### drvrm

In the series solution of simple harmonic oscillator,why do we have a factor of 1/2 in the trial solution?
the factor is chosen from the asymptotic behavior of the solution of differential equation ...in the above situation as the 2nd order partial derivative should yield square of the variable times the function the factor 1/2 is chosen...
see for details
<http://ocw.mit.edu/courses/physics/8-04-quantum-physics-i-spring-2013/lecture-notes/MIT8_04S13_Lec08.pdf> [Broken]

however the trial functions are always chosen by anticipating the behaviour of the solution as r going to zero and infinity as it is easy to estimate. for bound states at both these extremities the wave function should go to zero.

Last edited by a moderator:

#### koustav

the factor is chosen from the asymptotic behavior of the solution of differential equation ...in the above situation as the 2nd order partial derivative should yield square of the variable times the function the factor 1/2 is chosen...
see for details
<http://ocw.mit.edu/courses/physics/8-04-quantum-physics-i-spring-2013/lecture-notes/MIT8_04S13_Lec08.pdf> [Broken]

however the trial functions are always chosen by anticipating the behaviour of the solution as r going to zero and infinity as it is easy to estimate. for bound states at both these extremities the wave function should go to zero.
But still we are left with the term e^-q^2/2

Last edited by a moderator:

#### PeroK

Homework Helper
Gold Member
2018 Award
You take $\psi = e^{-q^2 /2}$ because that's the solution. A better question is how do you get to that solution?

If you try $\psi = e^{-q}$ then $\psi'' = \psi$ so that doesn't work.

If you try $\psi = e^{-q^2}$ then $\psi'' = (4q^2 - 2)\psi$ so that doesn't quite work, but it should give you the idea:

If you try $\psi = e^{-q^2 /2}$ then $\psi'' = (q^2 - 1)\psi$ and, if you ignore the $-1$ for large $q$, then you have a solution.

#### koustav

You take $\psi = e^{-q^2 /2}$ because that's the solution. A better question is how do you get to that solution?

If you try $\psi = e^{-q}$ then $\psi'' = \psi$ so that doesn't work.

If you try $\psi = e^{-q^2}$ then $\psi'' = (4q^2 - 2)\psi$ so that doesn't quite work, but it should give you the idea:

If you try $\psi = e^{-q^2 /2}$ then $\psi'' = (q^2 - 1)\psi$ and, if you ignore the $-1$ for large $q$, then you have a solution.
But if we take the 2nd case still our assumption is correct for large value of q

#### PeroK

Homework Helper
Gold Member
2018 Award
But if we take the 2nd case still our assumption is correct for large value of q
Really? Like $4q^2 - 2 \approx q^2$ for large $q$?

#### koustav

Really? Like $4q^2 - 2 \approx q^2$ for large $q$?
Ok got the point.thanks for simplifying the point!

#### BvU

Homework Helper
So basically the trial solution is $\ e^{-\alpha q^2} \$ and the 'characteristic equation' yields $\ 4\alpha^2q^2-2\alpha q - q^2 = 0 \$, to expand a little on PeroK post #8.

"Simple harmonic oscillator"

### Physics Forums Values

We Value Quality
• Topics based on mainstream science
• Proper English grammar and spelling
We Value Civility
• Positive and compassionate attitudes
• Patience while debating
We Value Productivity
• Disciplined to remain on-topic
• Recognition of own weaknesses
• Solo and co-op problem solving