Density Ranges

JM_Sarthak

Hey there !

I have a few of questions regarding the density ranges, which the user is required to input in the xml. The questions are as follows :-

1) What is the most optimum range, on either side of the Reference density, such as (+/-300,+/-400 etc) , in order to get the most accurate simulations?

2) In order to incorporate a wide range of values, can the minimum be kept at 0 ? and the maximum be 1000 or so units from the reference value If no what is the Physics behind these ranges, and how do they affect computation ?

Ahsry

what I understand: some particles might encounter high fluctuations in density during simulation. They are excluded if they exit the specified range by user. However, there is a re-normalization process each bundle of time steps (default = 40 time steps).

Alex

Dear all

Let me clarify this:

1) the range of density we define in the XML is only to fix some limits that are used to exclude particles. However if reference density is 1000, the values in the simulation should vary between 990 and 1010 for example... so 700 or 1300 are really very bad values that means that sthg wrong is happening so that we prefer removing those particles

2) there is no re-normalization process.... only Verlet time algorithm uses different set of equations every 40 time steps to avoid divergence in the computations...
Verlet is fast but not accurate, only recommended for tests. You should use Symplectic.
This is no relalted with density ranges

Regards

Ahsry

Thank you for clarification. What aboud shepard normalization, is it done while picking a point post-processing (I saw it in the post processing guide). isn't so?