I’m going to take a break this week from graphs, and talk about a fundamental new concept introduced in GeoStudio 2007. Understanding why some changes affect all analyses while other changes affect only the current analysis will help you work smarter and faster and reduce the chance of introducing errors in your calculations.
Background
GeoStudio 2007 lets you combine as many analyses as you (and your hard drive) want in a single file.
In previous versions you could of course have as many files as you wanted in a particular folder and link them together by getting, say, Initial PWP Conditions from another file. But combining analyses in a single file is much more powerful than that, because you’re actually sharing much of the data.
In essence you are saying “these analyses are different views on the same slice of the earth.”
Global Objects
Most “objects” in a file are “global”, that is they are shared by all analyses. To get a feel for what I mean by an “object”, look at the KeyIn menu. Most things (or everything?) that you see in that menu are global objects.
The most important global object is the geometry: the regions, lines and points. Since every analysis is looking at the same “slice of the earth”, it makes sense that they all use the same geometry. The slope is the same angle, the soil layers are the same, the altitude is the same.
Another global object everyone uses is the material list. Create a material named “sandy clay” and any analysis that uses that material is using that exact material. You can’t use the same material in two analyses but say it has different physical properties. (But what if you want to compare the results as you change C? More on that later.)
Materials are interesting because different types of analysis care about different properties. For example, a Mohr-Coulomb material in SLOPE/W requires that you enter Unit Weight, Cohesion and Phi, but a SEEP/W analysis doesn’t care about those properties. From that we can see that the material “object” is global (you can see it in the list and you can assign it to a region in SLOPE/W and in SEEP/W) but each analysis has a different set of material properties it needs you to enter.
Some material properties are common across different analysis kinds. This makes sense because a soil’s conductivity, for example, is always the same no matter what you’re trying to analyze. Give a material a Hydraulic K Function in SEEP/W, and a VADOSE/W analysis will use that same function.
Other global objects include (but this is not an exhaustive list):
- boundary conditions
- functions
- climate data
- etc
Before you contradict me because you know you can have different materials or different boundary conditions in different analyses, read on about associations.
Associations
“Associations” is not a word you’ll see in our documentation, but we use it amongst ourselves and it’s a useful concept to understand the difference between “global” and “per-analysis”.
A region is global.
A material is global.
Which region is associated with which material is stored with the analysis. Primarily what differentiates one analysis from another is its list of associations.
Associations are typically made using the Draw commands. In Draw Materials, for example, you choose a material from a list and then click on a region. You’ve just associated that material with that region in the current analysis.
We asked the question earlier, how do you make two analyses identical in every way except a different value for C, if the material is global? You create two materials, give them the same properties except for C, and use one in one analysis and the other in the second analysis.
The same concept applies elsewhere. Draw (”associate”) the same boundary condition object in two different analyses, and if you use KeyIn Boundary Conditions and change the Action from 3 to 5, it has changed in both analyses. Draw one boundary condition in one analysis and a different one in another analysis, and changing the action only affects the one analysis.
Take the concept a bit further and you can bend the rules a bit. You can actually have different geometries in different analyses by simply failing to associate a material to a certain region. A region with no material (we call that a “null region”) is ignored by the solver, so you’ve essentially changed the geometry. That is how you do fill or excavation analyses in SIGMA/W, for example.
Conclusions
Objects are global. Analyses are associations of global objects.
Grok that concept and you’ll be able to effectively combine analyses, share data that needs to be shared and understand when a change made in one analysis will affect another.
{ 0 comments… add one now }