Introducing Parameterization and Parametric Tools
You can learn a great deal by running an Adams Simulation of a single configuration. You can learn even more by manually changing your model, and running simulations again and again, but the process quickly becomes tedious. Instead, you can use parameterization and the parametric tools that Adams View provides to automate your changes.
Using parameterization, you can make a single change and your entire model automatically updates. Using Parametric analyses, you can automatically run a series of simulations to see the effects of varying your model.
About Parameterizing Your Model
Manually updating your model can be time-consuming because rarely is it as simple as changing just one modeling object. Frequently, other objects depend on the object you are changing, which forces you to change those objects as well.
Therefore, the first step in creating a parameterized model is to select the critical design inputs that you want to vary to perform "what if" studies. When you change a critical design input, dependent model characteristics update automatically. You parameterize your model by creating parameters and defining how the model depends on them. You can parameterize your model as you build it, or build it first and then add the parametric relationships.
Adams View provides several parameterization methods. For example, you can parameterize your model using:
■Parameterization move tools (f(x) and f(
)that let you specify how one objects moves relative to another object.
)that let you specify how one objects moves relative to another object. ■Expressions, which are the basis of all parameterization.
About Parametric Analysis Tools
Parametric analyses help you investigate the influence of design variables on model performance. During a parametric analysis, Adams View runs a series of simulations with different values for the design variables and gives you feedback on the effects of the changes.
■Adams View has four types of parametric analyses:
■Design study, which shows the effects of varying one design variable.
■Design of Experiments (DOE), which shows the effects of varying several design variables simultaneously.
■Optimization, which adjusts design variables to minimize or maximize a particular aspect of your model's performance.
■Temporary settings sweep, allows one to specify and execute a number of trials, each of which is defined by applying one or more temporary settings files to the baseline model.
The first step in using parametric analyses is to understand design studies, DOEs, and optimizations, what they do for you, and how they can work together. Depending on your model and interest, you may use one, two, or all three to explore your model.
In all cases, you start by deciding which design variables to vary and how to measure the performance of your model. In the latch model from the guide, Getting Started Using Adams View, for example, the design variables are the coordinates of the pivot points, and the performance measure is the maximum spring force during the simulation.
Learn more about the different types of parametric analyses you can run and how you can use them together:
About Design Studies
A design study analysis is useful for understanding one design variable. A design study varies a single design variable across a range or series of values you specify, runs a simulation at each value, and reports the performance measure for each simulation.
Using a design study analysis, you can determine:
■How the performance varies with changes in the design variable.
■The best value for the design variable among the values simulated.
■The approximate design sensitivity of the variable; that is, the rate of change of the performance measure with respect to the variable.
About Design of Experiments
A Design of Experiments (DOE) helps you understand how several design variables interact. While design study analyses vary only one design variable at a time, a DOE varies as many variables as you want. You specify the range or series of values for each variable and which combinations of values you want to simulate. Adams View runs the simulations and records the performance measure for each run.
If your model is complex and involves many design variables, choosing the runs by intuition or trial-and-error can give you results that are more confusing than they are helpful. If so, you should consider using a structured approach based on DOE techniques.
The field of DOE (also called experimental design) offers a collection of procedures and statistical tools for planning experiments and analyzing the results. Although DOE techniques were developed around physical experiments, they work just as well with virtual experiments in Adams View. We've designed DOE analyses to make it easy to apply DOE techniques to your model. For more information on DOE techniques, see About Design of Experiments.
Using a DOE and appropriate DOE techniques, you can:
■Identify which design variables and combinations of design variables most affect the performance of your model (screening).
■Control the effects of variations due to real-world manufacturing and operating conditions (robust design or the Taguchi method).
About Optimization
An optimization adjusts design variables to minimize or maximize a performance measure. You can set ranges on how far to vary the design variables and add general constraints to keep the optimized design within overall limits. Using an optimization, you can find the best performing values for design variables.
For more information on optimization techniques, see About Optimization and Running Parametric Analyses.
About Temporary Settings Sweep
A temporary settings sweep allows one to specify and execute a number of trials, each of which is defined by applying one or more temporary settings files to the baseline model.
Using Design Study, DOE, Temporary Settings Sweep and Optimization Together
■A design study, DOE, or optimization are useful individually, but combining several parametric analyses can give you a fuller understanding of your model's performance. Design studies and DOEs help you explore variations and trade-offs in performance, while optimizations try to find a specific combination of design variable settings that achieve optimal performance.
■For example, you might start exploring your model with one or two design studies on design variables that you know are important. The design studies show you the major effects of changing your design and allow you to make a first guess at good values for these variables.
■You might then do a few more design studies or a screening DOE to find out if there are other important design variables or interactions between design variables that you should consider.
■When you have determined which design variables are the most important, you can use optimization to fine-tune their values.
■Once you have final values for the design variables, a design study or DOE can tell you what happens when those values vary. This is important if you are concerned about variations in real-world performance or you want to adjust your design to gain other advantages without sacrificing performance.
■A design study or DOE can also help you set up an optimization. If you do not have good initial values for the design variables, the optimization can fail, be slow, or may converge to a design that is not an overall optimum. A design study or DOE over a range of values can help find good initial values for optimization. It also can give you a polynomial approximation that you can optimize separately to find a starting point for a full optimization.
■A temporary settings sweep allows one to specify and execute a number of trials, each of which is defined by applying one or more Temporary Settings files to the baseline model.
Using Expressions
Expressions are the basis of all parameterization. You can specify most modeling data in Adams View as either a constant value or an expression that can change its value based on other objects and values in your model. When you specify an expression, Adams View stores the expression and automatically updates the value whenever a value in the expression changes.
For example, when you specify the mass of a part, you can supply a constant value, such as 5.0, or an expression, such as:
(2 * .model_1.part_1.mass)
Using the expression above, the new part mass is always twice the mass of part_1, even if you change the mass of part_1.
Expressions are always enclosed in parentheses and can include:
■Constants
■Standard mathematical operators and functions
■Special Adams View functions
■References to other object data in your model
You enter an expression directly in the text box for the value you want to parameterize. You can enter an expression when you create the object or modify it later to use an expression.
Adams View contains a Function Builder to help you construct expressions. You access the Function Builder by displaying the shortcut menu in a text box that accepts an expression, as explained in the next section.
To access the Function Builder:
■Right-click a text box where you want to place an expression, point to Parameterize, and then select Expression Builder.
For more information on creating expressions and using the Function Builder, see Adams View Function Builder online help.
To remove an expression, do either of the following:
■Modify the object and enter a constant value in the text box.
■Place the cursor in a text box containing the expression and hold down the right mouse button. From the shortcut menu that appears, point to Parameterize, and select Unparameterize.
Using Points
Points are the easiest way to parameterize the geometry of your model. Points let you specify important locations once and build other modeling objects from them. When you move a point, the related objects update automatically.
You create points using the Geometric Modeling Palette and Tool Stack on the Main toolbox. For information on points, also see the following sections:
You attach new modeling objects to points by selecting the points as you graphically construct the object. When you build objects on points, Adams View creates the necessary expressions for you.
Tip: | Right-click near the point to display a list of all objects in the area and then select the desired point from the list to ensure it gets selected. |
You can also attach existing objects to a new point by using the Attach Near option when creating the new point. In this case, Adams View creates expressions, using the function LOC_RELATIVE, to attach any nearby markers to the new point. With this option, you can parameterize model geometry, forces, and constraints as you need to rather than creating all points first.
If you later try to move an object that is attached to a point, Adams View warns you that doing so can break the parameterization and asks you how you want to continue. The warning prevents you from accidentally removing a relationship and also allows you to delete the relationship.
Usually you do not need to look at or understand the expressions that tie geometry to points. If you want to create more complicated geometric relationships, however, understanding how points work can help you write your own expressions.
If you draw a link between two points, for example, Adams View locates markers at each end of the link on top of the points. Adams View also creates expressions to keep the markers tied to the points. If you request information on one of the markers at the ends, you see something like the following for the location of the marker:
Location: -150.0, 250.0, 0.0 (mm, mm, mm)
(LOC_RELATIVE_TO({0, 0, 0}, .model_1.ground.POINT_1))
The first line is the current value of the location of the marker relative to the link part. The second line is the expression that Adams View created to keep the marker at the point. If you change the location of the point, Adams View automatically evaluates the expression and computes a new location for the marker. LOC_RELATIVE_TO is one of the Adams View functions that lets you locate points and markers relative to other objects in your model. For more on Adams View functions, see the Adams View Function Builder online help.