Difference between revisions of "Cometparams.ini parameter file"

Revision as of 16:47, 16 April 2009 (view source) Markdayel (Talk | contribs) m ← Older edit		Revision as of 14:47, 17 April 2009 (view source) Markdayel (Talk | contribs) m Newer edit →
Line 24:		Line 24:
	NUC_LINK_FORCE          2.0		NUC_LINK_FORCE          2.0
	NUC_LINK_BREAKAGE_DIST .237		NUC_LINK_BREAKAGE_DIST .237
		+
	When nodes are created, STICK_TO_NUCLEATOR defines whether they stick to their point of creation on the nucleator surface.  Stuck nodes exert a force proportional to NUC_LINK_FORCE multiplied by the distance from the surface stuck point until they are extended beyond NUC_LINK_BREAKAGE_DIST when the link breaks.  If RESTICK_TO_NUCLEATOR is true, ''unstuck''  nodes will re-stick if they come into contact with the surface again.		When nodes are created, STICK_TO_NUCLEATOR defines whether they stick to their point of creation on the nucleator surface.  Stuck nodes exert a force proportional to NUC_LINK_FORCE multiplied by the distance from the surface stuck point until they are extended beyond NUC_LINK_BREAKAGE_DIST when the link breaks.  If RESTICK_TO_NUCLEATOR is true, ''unstuck''  nodes will re-stick if they come into contact with the surface again.
Line 32:		Line 33:
	NODE_REPULSIVE_MAG      2.7		NODE_REPULSIVE_MAG      2.7
	NODE_REPULSIVE_POWER    2.0		NODE_REPULSIVE_POWER    2.0
		+
		+	[[file:mymodel_forces.png|right|thumb|350px|Figure 1: Forces acting on bead]]
	The repulsion force between nodes is of the form:		The repulsion force between nodes is of the form:

---

Revision as of 14:47, 17 April 2009

Here are the core settings in the cometparams.ini file (explained below):

Run Time

DELTA_T                 0.01
TOTAL_SIMULATION_TIME   5600.0
TOT_FRAMES              700

TOTAL_SIMULATION_TIME defines the run length in simulation time (uncalibrated, nominally seconds). DELTA_T defines the time step between iterations, i.e. for the given TOTAL_SIMULATION_TIME of 5600 and DELTA_T of 0.01, there will be a total of 560000 iterations. TOT_FRAMES defines the number of snapshots to be taken during the run, i.e.~700 snapshots would mean one snapshot every 800 iterations.

Nucleator

SHAPE                   SPHERE
ELLIPSOID_STRETCHFACTOR 1.5
RADIUS                  2.5
CAPSULE_HALF_LINEAR     2.75

SHAPE can be SPHERE, CAPSULE or ELLIPSOID. For SPHERE, only the RADIUS matters. For CAPSULE, RADIUS and CAPSULE_HALF_LINEAR are used, and for ELLIPSOID, RADIUS and ELLIPSOID_STRETCHFACTOR define the shape. (Arbitrary shapes can be defined in the code, given a function that for a supplied point, returns a vector normal to the nearest point on the surface to the given point.)

Nucleator attachments

STICK_TO_NUCLEATOR      true
RESTICK_TO_NUCLEATOR    true
NUC_LINK_FORCE          2.0
NUC_LINK_BREAKAGE_DIST .237

When nodes are created, STICK_TO_NUCLEATOR defines whether they stick to their point of creation on the nucleator surface. Stuck nodes exert a force proportional to NUC_LINK_FORCE multiplied by the distance from the surface stuck point until they are extended beyond NUC_LINK_BREAKAGE_DIST when the link breaks. If RESTICK_TO_NUCLEATOR is true, unstuck nodes will re-stick if they come into contact with the surface again.

Node repulsion function

NODE_REPULSIVE_RANGE    1.0
NODE_REPULSIVE_MAG      2.7
NODE_REPULSIVE_POWER    2.0 

The repulsion force between nodes is of the form:

<math>

F_R = M_R \left( \left(\frac{d_R}{d}\right)^{P_R} - 1 \right), \quad 0<d<d_R

</math> where <math>d</math> is the distance between nodes, <math> M_R </math> (NODE_REPULSIVE_MAG) is a magnitude scale factor, and <math> d_R </math> (NODE_REPULSIVE_MAG) is maximum range of the repulsive force.

The power factor <math> P_R </math> (NODE_REPULSIVE_POWER) is 2, so this is a simple inverse square repulsive force and is plotted in \fref{fig:repulsiveforces}.

Node links

P_NUC                   0.12
XLINK_NODE_RANGE        1.0
MAX_LINKS_PER_NEW_NODE  10
LINK_BREAKAGE_FORCE     3.0
LINK_FORCE              3.0
P_XLINK                 .700
VARY_P_XLINK            true 

P_NUC defines the rate of nucleation of new nodes per unit area per unit time. i.e.~for one iteration, the number of new nodes added over the whole of the nucleator surface is P_NUC * DELTA_T * surf_area, where surf_area is in {\micro}m<math>^2</math>. The nodes are added at random positions on the surface, with an even distribution unless the ASYMMETRIC_NUCLEATION variable is set.

New nodes are crosslinked to nearby nodes within XLINK_NODE_RANGE. The links then behave as Hookean springs, exerting a restoring force <math>

F_L = -{M_L} \left(\frac{d-d_L}{d_L}\right)

</math> where <math>d</math> is the distance between nodes, <math> M_L </math> is a magnitude scale factor, and <math> d_L </math> is the original length of the link when it was formed (\fref{fig:simulationdetails}). If the link is extended so that its force goes beyond a certain limit, the link breaks. (optionally this can be strain rather than stress, i.e.~a break occurs when <math>\frac{d}{d_L}</math> exceeds a certain limit rather than when <math>\frac{d-d_L}{d_L}</math> does)

Nodes are added to the surface and fixed there while their repulsive forces are ramped up linearly from 0 to full. This allows time for nodes already at the surface move and make room for the new node before it is crosslinked. The ramp-up occurs over CROSSLINKDELAY iterations. MAX_LINKS_PER_NEW_NODE limits the maximum number of crosslinks for each new node. LINK_FORCE is the spring constant, and when the extension forces reaches LINK_BREAKAGE_FORCE, the link breaks. P_XLINK is the probability of forming a crosslink to a node within range (still restricted by the MAX_LINKS_PER_NEW_NODE limit). The VARY_P_XLINK flag (normally on) also imposes a linear tail-off of this probability with distance. (see \fref{sec:parameterinfo} for more info).

Drag

FORCE_SCALE_FACT        0.3
NUCLEATOR_INERTIA       80
MofI                    0.5
VARY_INERT_W_RAD        false

This section relates the forces to the actual movement of the nodes and nucleator. FORCE_SCALE_FACT scales the movement of nodes (i.e.~effectively inverse of node drag). If you reduce this, you probably need to reduce DELTA_T as well. NUCLEATOR_INERTIA determines how hard it is to displace the nucleator and MofI determines how hard it is to rotate it. If VARY_INERT_W_RAD is set, inertia will be scaled by the size of the nucleator (see \fref{sec:parameterinfo} for more info).