|
[Sponsors] |
Does the workstation effects the output accuracy? |
|
LinkBack | Thread Tools | Search this Thread | Display Modes |
January 31, 2005, 00:51 |
Does the workstation effects the output accuracy?
|
#1 |
Guest
Posts: n/a
|
Recently I just developed a code using Matlab. The results need about 6000 time steps to converged. But the example shows that it should be able to converge at about 1200 time steps. So I checked the results for every calculation, and there is a problem. For grid point 16 or 17, the results for every calculation agreed very well. But there are some variation at grid point 15, and this small variation caused my algorithm take more time to converge. My lecturer told me that it maybe 2 reaons: 1)Round-off errors. 2)Caused by the machine itself.
What I know is that the workstation only affect the speed of computing. Will it affect the accuracy of the output as well? This really puzzled me. If this is not the reason, then is it because of my boundary condition? |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Increasing the accuracy of pressure head values in output data | Saeed1390 | FLUENT | 0 | June 13, 2011 02:17 |
Questions about data output | aki_yafuji | OpenFOAM Running, Solving & CFD | 3 | September 9, 2010 01:59 |
lift and drag on ship superstructures | vaina74 | OpenFOAM Running, Solving & CFD | 3 | June 8, 2010 12:30 |
[Other] Output Format of MetaMesh | t42 | OpenFOAM Meshing & Mesh Conversion | 0 | August 3, 2007 04:28 |
accuracy of Discontinuous Galerkin | Qu Kun | Main CFD Forum | 0 | September 6, 2004 09:19 |