An object-oriented cluster search algorithm Page: 3 of 14
This report is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to UNT Digital Library by the UNT Libraries Government Documents Department.
Extracted Text
The following text was automatically extracted from the image on this page using optical character recognition software:
CLUSTER SEARCH
I ,
Porenetwork
I ,
all _nodes all _links
I ,
'------- --------------------- -------J
onenode onelink
one node one node one link one link
FIGURE 1. The structure of a pore network object
2. NODES AND LINKS
In this section, we define the main objects operated by the algorithm. The
algorithm has been implemented in C++ flow simulation code NetSimCPP [13].
NetSimCPP, in turn, is a derivative of MatLab code ANetSim by Tad Patzek im-
plementing the conclusions of [12]. The objects in NetSimCPP were used for much
more general tasks than in the present paper, therefore we present them in a reduced
form.
Each node object belongs to the class onenode described in Appendix. Each
node has a unique integer index, which we call ID. There are also two integer
parameters: an openness flag equal to one if the node is open for a given fluid
phase flow or zero otherwise and cluster-o (which is equal to the index of the
cluster whose element the node is). The last parameter is reused in the second
algorithm described in Section 5 for labelling the "already checked" nodes. This
algorithm also requires information whether node is at the inlet or at the outlet
of the network. Two parameters are reserved for this purpose. Two pointers to a
node and to a link are used to store the information about the node and link, from
which this node was approached by the algorithm. Each node also includes a list
of pointers to the links connecting the node to the neighbors and a list of pointers
to these neighbors. Obviously, both lists have the same number of items, which is
equal to the coordination number of the node. Different nodes may have different
coordination numbers, so the length of the lists of links and neighbors is not fixed.
A node object also includes a method, which we will call addtocluster, which is
used in computing the clusters, and a method called next._step used in calculating
the inlet-outlet connection.
The link object properties include an integer ID, two node pointers to the nodes
connected by the link and an integer flag signifying the openness of the link, see
class onelink in the Appendix.
As we have already mentioned above, the entire network is represented by two
lists: the list of all nodes and the list of all links, see Fig. 1. Such an object is
defined by class porenetwork in the Appendix.3
Upcoming Pages
Here’s what’s next.
Search Inside
This report can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Report.
Silin, Dmitry & Patzek, Tad. An object-oriented cluster search algorithm, report, January 24, 2003; Berkeley, California. (https://digital.library.unt.edu/ark:/67531/metadc736846/m1/3/: accessed April 25, 2024), University of North Texas Libraries, UNT Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.