Scribd
Upload
2 views2 pages

Tutorial 2

The document outlines a tutorial for PH554 - CMBP at IIT Bombay for Spring Semester 2025/26, focusing on estimating π using Monte Carlo methods, analyzing the Ising model through single-spin flip and cluster updates, and performing scaling analysis. It includes specific tasks such as variance estimation, magnetization analysis, and the implementation of algorithms like Wolff and Swendsen-Wang. Additionally, it emphasizes the importance of scaling analysis to determine critical exponents and universality in phase transitions.

Uploaded by

sajal.techfest
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
2 views2 pages

Tutorial 2

The document outlines a tutorial for PH554 - CMBP at IIT Bombay for Spring Semester 2025/26, focusing on estimating π using Monte Carlo methods, analyzing the Ising model through single-spin flip and cluster updates, and performing scaling analysis. It includes specific tasks such as variance estimation, magnetization analysis, and the implementation of algorithms like Wolff and Swendsen-Wang. Additionally, it emphasizes the importance of scaling analysis to determine critical exponents and universality in phase transitions.

Uploaded by

sajal.techfest
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

PH554 - CMBP - Tutorial II

Spring Semester - 2025/26


Department of Physics,
IIT Bombay

(Dated: February 8, 2026)

A. Estimate π with Monte Carlo

1. Estimate by drawing random numbers N times uniformly on a square. Estimates the area fraction of
a unit circle inside the square (by counting the number of points hitting the circle), and hence derive
an estimator for π.

2. Estimate the variance of the error for given N . This requires you to do the measurement more than
once.

3. Plot the variance of the error versus N on a log-log scale. What can you say about the error of the
estomator with respect to N .

B. Single-spin flip updates for the Ising model.

1. Use the code given to you for this purpose to estimate the magnetization as function of temperature.

2. Check for a given L how magnetization is evolving with MC steps, below Tc , and above Tc .

3. Perform a binning analysis of the single-spin-flip algorithm to find out the auto-correlation time.

4. Changing L detect the tranisiion by plotting the ⟨|M |⟩ as a function of temperature T .

5. Now include a magnetic field H ′ = −h


P
i σi . Repeat the above question.

6. Calculate the following quantity, known as Binder cumulant (also learn about it in the notes by
Sandvik, linked in the moodle):

⟨M 4 ⟩
 
3
UB = 1− . (1)
2 3⟨M 2 ⟩2

Plot it for L, 2L as a funciton of temperature, and see that the crossing of this qunaity gives a more
precise definition of Tc .
2

C. Cluster updates for the Ising model.

1. Implement the Wolff algorithm or Swendsen-Wang. Check the energy vs time curve to see that the
algorithm is giving the same energy as previously found in the single-spin-flip algorithm.

2. Check the magnetization vs temperature curve for 3 different system sizes.

3. Calculate the autocorrelation time for magnetization using any of the above cluster update algorithm.

4. Design the Wolff algorithm for X-Y model in two-dimension, and check for a possible phase transition.

D. Scalng analysis

Whatever the algorithm you use, you should also perform a scaling analysis to find out exponents, which
is a signature of universality.

1. Binder cumulant has the following scaling

UB = Φ(τ L1/ν ), (2)

T −Tc
where Φ is an unknown universal scaling funtion, and τ = Tc . Plot UB versus τ L1/ν for various
L. Vary the unknown exponent ν until the curves all appear on a single line.

2. Similarly, check that the specific heat also has a scaling relation and Cv ∝ |τ |−α . Find α approxi-
mately.

You might also like