Running multiple simulation cases with short simulation times is something that is strongly connected with Simcenter Amesim. However, what is the benefit of producing simulation results on mass if the results are not taken care of properly and analyzed in a meaningful way?
In this week’s post we aim to cover a few useful plotting features which hopefully will allow you to get more out of your simulation results, and perhaps even find some hidden relationship which you haven’t uncovered before.
To start off with, plots inside Simcenter Amesim are created in three different ways. Firstly, an empty plot can be created from the plots menu by navigating to plots then selecting “new plot”. Secondly, plots can be opened while in Simulation Mode by double clicking a component, selecting the relevant variables, and then pressing plot. The third, and likely the much more used option, is to select the component of interest, and drag and drop variable results onto the Sketch area.
Additional variables may also be added to an existing plot by, for example, multi-selecting variables and dragging them onto a plot.
The Y Animated Curve feature, found among the plot’s tools, is another way of analyzing plotted results. By selecting this option, the time dependent plot is converted into a time dependent bar diagram including a slider for adjusting time. Apart from studying the variables as bars varying with time, the feature can also be used to conveniently create bar diagrams of end results. This is achieved by moving the slider all the way to the last simulated time-step.
In addition to being a very convenient way of creating bar diagrams of your final results, y-animated curves also provide some interesting possibilities for creating other types of graphs.
An appealing option when comparing results is the classical spider chart. In the example below results of a few KPIs are shown for an electric vehicle, where each metric is ranked on a scale from 0 to 5. This type of plot can be created by right clicking a Y-Animated Curve plot and then selecting Type, as seen below.
A similar alternative to spider charts is the 2D parallel coordinates plot. Here each metric is shown along its own vertical axis instead of radiating out from a center point. This option can perhaps make comparisons easier to understand, especially for an audience not used to the compact form of spider charts.
In Simcenter Amesim we mainly concern ourselves with dynamic simulations over time. Therefore, newly created plots generally show variable results with respect to time. However, a common situation in data exploration arises when we want to study the relationship between two or more variables. By plotting variables against each other we can observe trends or find patterns between them.
With scatter plots two or more variables can be plotted against each other. To create such plots the default 2D curve needs to be converted to either a XY curve or a XYZ curve, depending on the number of variables used. To achieve this, variables can be added to a new plot either by dragging and dropping each variable one at a time, or by multi-selecting variables and then dropping them onto a plot. Keep in mind that when multi-selecting variables the order in which you select your variables will determine which axis they are placed on. Once the variables are added, the axis conversion is done by navigating to Tools and then selecting XY curve, for the two variable case, or XYZ curve for the three variable case. The steps are illustrated in the figure below.
With a XY or XYZ plot different plot types become available. Below are some examples of what can be created using three result variables. Rotating 3D plots is done by selecting the zoom/pan/rotate button and then pressing down the right mouse button.
To conclude this post, a final post-processing feature should be mentioned. Cross-results in Simcenter Amesim allow you to compare end-results from different simulation cases (batch runs). By dropping either a parameter or variable onto the “Cross result” pan you create a cross result. By default, cross results are configured to provide the final values of a result-set, however you can choose to create a post-processing result of you variable first and then add this to you cross results to for instance show values at a specific simulation time, global maximum, or any other expression.
With cross result you can easily compare and analyze relationships between end-results from your simulation cases. In the example below, a contour plot created using multiple simulation cases is shown.
We hope you have found this article interesting. If you have any questions or comments, please feel free to reach out to us on support@volupe.com
Author
Fabian Hasselby, M.sc.
