Finding a generating function for a canonical transformation

Click For Summary
SUMMARY

This discussion focuses on finding a generating function for the canonical transformation defined by Q = (sin p / q) and P = q cot p. The user initially attempts to derive a type 1 generating function, F_1, which depends on variables q and Q, but ends up with a type 2 generating function, F(P, q). The user seeks clarification on the relationship dF = pdq - PdQ and how to express the differential dF in a single form. The discussion highlights the importance of understanding the transformation equations in canonical transformations.

PREREQUISITES
  • Understanding of canonical transformations in Hamiltonian mechanics
  • Familiarity with generating functions and their types
  • Knowledge of differential calculus and integration techniques
  • Proficiency in manipulating trigonometric functions, specifically cotangent and sine
NEXT STEPS
  • Study the properties and applications of generating functions in Hamiltonian mechanics
  • Learn about the different types of generating functions and their specific uses
  • Explore the derivation and implications of transformation equations in canonical mechanics
  • Practice solving problems involving canonical transformations and generating functions
USEFUL FOR

Students and researchers in physics, particularly those focusing on classical mechanics and Hamiltonian systems, as well as anyone interested in advanced mathematical techniques in physics.

fluidistic
Gold Member
Messages
3,934
Reaction score
286

Homework Statement


I'm trying to find a generating function for the canonical transformation [itex]Q=\left ( \frac{\sin p}{q} \right )[/itex], [itex]P=q \cot p[/itex].

Homework Equations


I am not really sure. I know there are 4 different types of generating function. I guess it's totally up to me to choose the type of the generating function I want, since it's not specified in the problem statement?
So let's say I want a type 1. Thus F_1 depends on q and Q.
[itex]\frac{\partial F_1}{\partial q}=p[/itex]
[itex]\frac{\partial F_1}{\partial Q}=-P[/itex].

The Attempt at a Solution


From the relevant equations, [itex]\partial F_1=-P \partial Q[/itex] and [itex]\partial F_1=p \partial q[/itex].
I don't really know why, but apparently [itex]dF_1=pdq-PdQ[/itex]. It looks like [itex]dF_1=2 \partial F_1[/itex] for some reason unknown to me.
Now I think the main idea to get [itex]F_1[/itex] is to express [itex]dF_1[/itex] as a single differential and then integrate.
I have that [itex]p= \cot ^{-1} \left ( \frac{P}{q} \right )[/itex] and [itex]dQ=\frac{1}{\sin p } \left [ \frac{\cos p}{q}dp-\frac{\sin p}{q^2}dq \right ][/itex].
This gives me [itex]dF_1=(p+ \cot p)dq-q\cot ^2 p dp[/itex]. In a book I found that this is equal to [itex]d[q(p+\cot p)][/itex] (how?!), therefore [itex]F_1=q \cot ^{-1} \left ( \frac{P}{q} \right )+P[/itex]. But it seems that F depends on P and q instead of on q and Q. So I "found" a type 2 generating function instead of type 1... I think I know why, I found p instead of q. Anyway my problems remains.
1)Why is [itex]dF=pdq-PdQ[/itex]?
2)Why and how do you figure out that [itex](p+ \cot p)dq-q\cot ^2 p dp=d[q(p+\cot p)][/itex]?
I'd be very grateful if someone could answer these 2 questions.
Thanks!

Edit: I just answered the "why" of my question 2. I just don't know how do one figures this out.

Edit 2: I do not know why I found out a type 2, namely F(P,q) while I was looking for F(q,Q).
 
Last edited:
Physics news on Phys.org
I guess I "transformed" the problem back to the original one, before the transformation. Now I understand why dF=pdq-PdQ, that's because of the transformation equations. I guess now I just have to figure out the answer to my question 1. How do you figure out that (p+ \cot p)dq-q\cot ^2 p dp=d[q(p+\cot p)]?
 

Similar threads

Replies
19
Views
3K
  • · Replies 2 ·
Replies
2
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
Replies
27
Views
4K
Replies
1
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K
Replies
4
Views
3K
Replies
6
Views
4K