This shows you the differences between two versions of the page.
Both sides previous revision Previous revision Next revision | Previous revision | ||
appendices:a5 [2019/06/25 11:26] helene |
appendices:a5 [2020/08/25 15:46] (current) |
||
---|---|---|---|
Line 1: | Line 1: | ||
====== Appendix 5: Data file for remeshing ====== | ====== Appendix 5: Data file for remeshing ====== | ||
This file is used by the remeshing programs CRECOU, CREDAT, INTERP, and REMAIL (see \\Obelix\c\Progs\sources\Remeshing\). These programs were created in the 90's and have not been modified since. It is likely that some of them do not work properly anymore. Some info on these programs can be found in the "REMAIL1.DOC", "REMAIL2.DOC", "REMAIL3.DOC" files on the server (\\Obelix\c\Progs\sources\Manuels). \\ | This file is used by the remeshing programs CRECOU, CREDAT, INTERP, and REMAIL (see \\Obelix\c\Progs\sources\Remeshing\). These programs were created in the 90's and have not been modified since. It is likely that some of them do not work properly anymore. Some info on these programs can be found in the "REMAIL1.DOC", "REMAIL2.DOC", "REMAIL3.DOC" files on the server (\\Obelix\c\Progs\sources\Manuels). \\ | ||
- | __Note__: By default, one can realize a remeshing with an empty .REM file, i.e. containing 8 blank lines. In this case, the default options are taken into account. | + | __Note__: By default, one can generate a remeshing with an empty .REM file, i.e. containing 8 blank lines. In this case, the default options are taken into account. |
^1st line (12I5)^^ | ^1st line (12I5)^^ | ||
|LCASE<sup>5</sup>|Choice of the transfer method for σ and κ for the internal elements| | |LCASE<sup>5</sup>|Choice of the transfer method for σ and κ for the internal elements| | ||
Line 69: | Line 69: | ||
|:::|= 2 the high density points will be replaced by the minor squares line| | |:::|= 2 the high density points will be replaced by the minor squares line| | ||
|:::|= 3 the density data of the father mesh will be taken into account (read on the file name_old.gml)| | |:::|= 3 the density data of the father mesh will be taken into account (read on the file name_old.gml)| | ||
+ | |||
+ | **Remarks** \\ | ||
+ | IDIW is smaller than IDIS in all cases. The low density points situated at a square distance lower than $L^2/IDIW$ of the strong points and the strong points situated at a distance lower than $L^2/IDIS$ of the other strong points will be eliminated. In other words, if the user wants density points at least separated by a distance $D_{critical}$ that he chose, he will have to choose the corresponding factor IDI(W or S) such that: \[IDI = \left(\frac{L}{D}\right)^2\] | ||
+ | Geometrically, if one considers each point successively as the center of a circle, one suppresses all the points (for example, the weak points) which are in the circle of radius R at this point with: \[R=\frac{L}{\sqrt{IDIW}}\] | ||
+ | {{ :appendices:a5_1.png |}} | ||
+ | For more information details relating to the computation mode of the density points, see the description of the routine RECDEN given in appendix. \\ | ||
+ | The POIDS will balance FMULT as a function of age according to the following rule: | ||
+ | \[FMULT -> \frac{FMULT}{POIDS^{age-1}}\] | ||
+ | It will be chosen: | ||
+ | * > 1 if the density has to decrease when the point becomes older | ||
+ | * < 1 if the density has to increase when the point becomes older | ||
+ | |||
+ | ^4th line (5G10.0)^^ | ||
+ | |EMULT|If IEMUL = 0 it is the multiplyer of the number of elements of the father mesh \\ If IEMUL=1, it is the absolute number of the elements of the son mesh| | ||
+ | |ANGLIM|value of the angle α∈[0,π] under which there is a line (solid angle of intersection) \\ Default value = 160° {{ :appendices:a_2.png?400 |}}| | ||
+ | |ANGPRE| Useful if and only if ANGLI=1 for parabolic element \\ Limit value of the angle β between the two segments going through the common node of two parabolic elements, for which a parabolic side is considered topologically linear \\ Default value =5° \\ A line will not be defined if β > ANGPRE {{ :appendices:a5_3.png?400 |}}| | ||
+ | |MAXELE|Maximum number of elements which may not be exceeded| | ||
+ | |RIGM|Penalty value of the new contact elements when a solid law is used| | ||
+ | ^5th line (I5)^^ | ||
+ | |IFDENS|Number of foundation nodes to which one wants to give a density (constant during the whole simulation) \\ Max 20| | ||
+ | ^5th bis line (I5, 2G10,0) - repeated IFDENS times if IFDENS ≠ 0^^ | ||
+ | |NF|Number of the foundation node| | ||
+ | |DPL|Associated density coefficient| | ||
+ | |CPL|Influence reduction parameter| | ||
+ | ^6th line (I5)^^ | ||
+ | |IFFAIBL|Number of density points generated between 2 different foundations nodes| | ||
+ | ^6th bis line (2I5, 3G10.0) - repeated IFFAIBL times if IFFAIBL ≠ 0^^ | ||
+ | |N1|N° of the node belonging to the foundation 1| | ||
+ | |N2|N° of the node belonging to the foundation 2| | ||
+ | |RATIO|Ratio a/b ∈ [0,1] {{ :appendices:a5_4.png?200 |}}| | ||
+ | |DPL|Density coefficient of the generated points| | ||
+ | |CPL|Influence parameters of the generated points| | ||
+ | ^7th line (2I5, 1G10.0) - Memorization parameters of the successive remeshings to limit the number of remeshings^^ | ||
+ | |The number of the step where the latest meshing has been realized is stocked in the file .LOA and the remeshing is effectuated only if the usual criteria are achieved and if the number of steps realized since the last remeshing > NDSTEPM|| | ||
+ | |ILOA|≠ 0 if one wants to use a file .LOA| | ||
+ | |NDSTEPM | Minimum number of steps between two remeshings| | ||
+ | |DIV | Divisor of the time step (1.0 per default)| | ||
+ | ^8th line (I5)^^ | ||
+ | |ILIC|=0 : no file .LIC. ETAGM represents then the **absolute percentage** of localized elements which provokes the stop of the solicitation for the remeshing.| | ||
+ | |:::|≠0 : one wants to print in the file .LIC the number of elements to be remeshed at each solicitation step. The remeshing will be effectuated only if the increment of the number of elements to be remeshed from step to step is sufficient, that is > ETAGM. In this case, ETAGM becomes the **relative percentage** of the localized elements since the latest converged step.| | ||
+ | ^9th line (I5, 2G10.0) - repeated INDMOD times if and only if INDMOD>0 (cf 1st line)^^ | ||
+ | |INOD | Number of the node| | ||
+ | |XNEW| New abscissa for this node| | ||
+ | |YNEW| New ordinate for this node| |