initialisation in CFX

littlelz · September 22, 2009, 08:12

I am just wondering if I don't set up any initialisation, what initial conditions will be used in CFX simulation. many thanks

littlelz

bharath · September 22, 2009, 10:35

If you don't initialize a problem it takes some more time to get converge.with out initialization you can run a problem

littlelz · September 22, 2009, 19:04

thanks. that's exactly what I experienced. I need more time to get converged result. I can run the problem without initialisation. however, I guess CFX must have some estimated initialisation in order to start the running although we don't know exactly what they are.

I am running one problem with M=0.25. If I set nothing for initialisation, it will fail because of overflow. however, if I set freestream conditions as initial data, it will converged very quickly.

for another case with M=0.8, without initialisation, it can be converged, but need longer time, with some initial setting, it can be converged quickly.

could anyboday give me some explaination on it?

many thanks

littlelz

ckleanth · September 22, 2009, 20:45

for any numerical method you need to provide a starting point for the calculations. obviously the closer you are to the solution the faster the process. CFD is no different and choosing realistic initial conditions for your problem will yield to faster convergence. I think the solver will use default initial values for variables if an initialization setup is missing; although I have to say I might be wrong as I dint test the above and I don't particularly try to make my life difficult on purpose.

ghorrocks · September 23, 2009, 08:33

If you don't specify anything CFX uses zero for all variables everywhere. For simple stuff this is OK, but tricky simulations need a better guess than this to get started.

littlelz · September 24, 2009, 07:09

Quote:

Originally Posted by ghorrocks

If you don't specify anything CFX uses zero for all variables everywhere. For simple stuff this is OK, but tricky simulations need a better guess than this to get started.

thanks guys,

your reply are very helpful. many thanks.

littlelz

September 22, 2009, 08:12	initialisation in CFX	#1
littlelz New Member Join Date: Mar 2009 Posts: 24 Rep Power: 17	I am just wondering if I don't set up any initialisation, what initial conditions will be used in CFX simulation. many thanks littlelz

September 22, 2009, 20:45		#4
ckleanth Senior Member George Join Date: Mar 2009 Location: Birmingham, UK Posts: 257 Rep Power: 18	for any numerical method you need to provide a starting point for the calculations. obviously the closer you are to the solution the faster the process. CFD is no different and choosing realistic initial conditions for your problem will yield to faster convergence. I think the solver will use default initial values for variables if an initialization setup is missing; although I have to say I might be wrong as I dint test the above and I don't particularly try to make my life difficult on purpose. __________________ Top 4 tips 1. Knowledge is everything and Ignorance is dangerous. 2. Understand your limitations and try to eliminate them. 3. Get yerself a bike and hoon the chuffer. You will soon learn why dogs like to hang their heads out the car window. 4. Please before asking any questions on how to run simulations in CFX, go though all the tutorials

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Pros and Cons for CFX, CFdesign, COMSOL	Val	Main CFD Forum	3	June 10, 2011 03:20
Importing solutions in CFX.	Alphonso	CFX	1	August 1, 2008 15:01
Fluent Vs CFX, density and pressure	Omer	CFX	9	June 28, 2007 05:13
PhD using CFX	Rui	CFX	9	May 28, 2007 06:59
CFX 4.4 installation problem	Pandu Sattvika	CFX	1	December 1, 2001 05:07

September 22, 2009, 10:35		#2
bharath Senior Member Bharath kumar Join Date: Apr 2009 Posts: 169 Rep Power: 17	If you don't initialize a problem it takes some more time to get converge.with out initialization you can run a problem

September 22, 2009, 19:04		#3
littlelz New Member Join Date: Mar 2009 Posts: 24 Rep Power: 17	thanks. that's exactly what I experienced. I need more time to get converged result. I can run the problem without initialisation. however, I guess CFX must have some estimated initialisation in order to start the running although we don't know exactly what they are. I am running one problem with M=0.25. If I set nothing for initialisation, it will fail because of overflow. however, if I set freestream conditions as initial data, it will converged very quickly. for another case with M=0.8, without initialisation, it can be converged, but need longer time, with some initial setting, it can be converged quickly. could anyboday give me some explaination on it? many thanks littlelz

September 23, 2009, 08:33		#5
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,841 Rep Power: 144	If you don't specify anything CFX uses zero for all variables everywhere. For simple stuff this is OK, but tricky simulations need a better guess than this to get started.