central difference approximation

Hermano · January 4, 2011, 09:49

Hi,

Where can I find the expression of the central difference approximation of the first and second derrivative (spatial) for a NON uniform grid?

otd · January 4, 2011, 19:00

For 3 points, (x1, f1), (x2, f2), (x3, f3), x1 < x2 <x3

Two methods (identical results for the approximation):

1. Expand f1 and f3 about x2 in Taylor series. Expand through the second derivatives.

You get two equations with unknowns df/dx and d^2f/dx^2 evaluated at x2. Solve those for your solution. Note: Expand through the 3rd and 4th derivatives, you have terms for the major errors you ignore by truncating the series after the 2nd derivatives.

2. Fit a parabola, f(x) = f2 + b(x - x2) + c(x - x2)^2 through the 3 points. b and c will be functions of x1, x2, x3 and f2 and f3.

Good luck.

OTD

DoHander · January 5, 2011, 18:23

Try this book:

Ferziger and Peric - Computational Methods for Fluid Dynamics

Do

January 4, 2011, 09:49	central difference approximation	#1
Hermano New Member Join Date: Jul 2010 Posts: 10 Rep Power: 16	Hi, Where can I find the expression of the central difference approximation of the first and second derrivative (spatial) for a NON uniform grid?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
First order in time and Central Difference Convergence problem	RameshK	Main CFD Forum	7	July 17, 2010 15:13
central difference method	sudhir	FLUENT	6	May 6, 2008 10:17
Specified Blend factor =1 Vs Central Difference	Kushagra	CFX	4	May 2, 2008 14:14
Central DIfference	Glen	Main CFD Forum	9	May 27, 2005 03:06
Central Difference vs Upwind Defference	S. Balasubramanyam	Main CFD Forum	5	January 16, 2002 02:58

January 4, 2011, 19:00		#2
otd Member private Join Date: Mar 2009 Posts: 74 Rep Power: 17	For 3 points, (x1, f1), (x2, f2), (x3, f3), x1 < x2 <x3 Two methods (identical results for the approximation): 1. Expand f1 and f3 about x2 in Taylor series. Expand through the second derivatives. You get two equations with unknowns df/dx and d^2f/dx^2 evaluated at x2. Solve those for your solution. Note: Expand through the 3rd and 4th derivatives, you have terms for the major errors you ignore by truncating the series after the 2nd derivatives. 2. Fit a parabola, f(x) = f2 + b(x - x2) + c(x - x2)^2 through the 3 points. b and c will be functions of x1, x2, x3 and f2 and f3. Good luck. OTD

January 5, 2011, 18:23		#3
DoHander Senior Member Join Date: Nov 2009 Posts: 411 Rep Power: 20	Try this book: Ferziger and Peric - Computational Methods for Fluid Dynamics Do