CFD Online Discussion Forums

CFD Online Discussion Forums (
-   Main CFD Forum (
-   -   The MAC Method (

Maciej Matyka October 30, 2000 05:36

The MAC Method
Hello everybody. At first - sorry for my pure english.

Im Polish student of Physic at University of Wroclaw. Im beginner in CFD, but im learning. Im working 2d program to simulate fluid motion. My C code base on algorithm presented in "The MAC Method for incompressible, viscous..." by Harlow, Welch (1965 Los Alamos). In general - program is finished, but didnt work correctly. I cannot change general algorithms myself, becouse i didnt understand 100% of that technique (e.g. pressure relaxiation). Anyway, my questions are:

1) Anybody work with algorithms prescribed in that Harlow's and Welch paper?

2) If yes - have you problems with that algorithm? Maybe there are some part missing???

3) I have also problems with free surface calculations, and boundary conditions (not described in that paper). At the moment im interesting only in NOSLP walls (i just wanna run that program to satisfy me at the moment - not solving complicated problems).

thanx, and sorry if i do anything wrong - its my first mail on that forum, i believe you undersand me

Maciej Matyka

ps. I spend lot of my time on that program (~4 months) and i cannot finish that program. help me if anybody work with that methods...

Ma October 31, 2000 19:52

Re: The MAC Method

I also worked on the MAC method for a few months, and I could not get the correct result.

There were many modifications(or improvements) after that orginal paper you mentioned. So maybe you can read some more paper on MAC.

As to me,I changed to another method(VOF)at last.

Good Luck!


Hush October 31, 2000 20:15

Re: The MAC Method
Hello Maciej

The MAC method solves a Poisson eq for pressure which is derived from the continuity eq. The source term of the Poisson eq. is then given from the NS eq.

So, 1) it finds first the source term with old velocity field. 2) solves Poisson eq till pressure field is converged 3) solives new velocity explicitly with the converged pressure. 4) moves marker particles with tne new velocity 5) go to the step 1)

I think you may find some information from other reports.

1) Chan, R.K.-C., Street, R.L. 1970, "SUMMAC- a numerical model for water waves", technical report no.135, dept. of civil engr., Stanford univ.
: it is a very kind report written on the MAC method. it treats the dynamic pressure boundary condition using 'irregular stars' when solving Poisson eq. for pressure.

2)C. W. Hirt, et al, 1975, "SOLA : a numerical solution algorithm for transient fluid flows"
: you can download this report from LANL it has the same idea as the MAC method except for pressure boundary condition.

3)H. Miyata, 1985, "Finite Difference Simulation of Nonlinear Ship Waves", JFM, vol 157
:The method is called TUMMAC which treats body boudary using some segments.

Maciej Matyka November 3, 2000 06:09

Re: The MAC Method
you thing VOF method is easier? my MAC solver works, and im still interesting in new methods of fluid calculations. where can i download any infos about VOF? (im asking, becouse i spend lot of weeks to search good papers about MAC method...).

btw. i didnt find any good examples of MAC method. for a couple of weeks (when i port my MAC solver for linux stations) i will put info here with url of my homepage with full source code of MAC program. i thing its very important for beginners in CFD (like me).

// maq

Ma November 4, 2000 13:36

Re: The MAC Method
It's very nice of you to put the source code on web.

As to the VOF, I remember there is a papar about Level Set and VOF method in JCP, 162.

:) I have changed to studying high speed flow.


Maciej Matyka November 6, 2000 04:08

Re: The MAC Method
"VOF method in JCP, 162" hmm... is any web page with JCP reprints? my Univeristy didnt buy it...

/// maq

Ma November 12, 2000 16:51

Re: The MAC Method
Maybe you can try


All times are GMT -4. The time now is 02:24.