SolidWorks Design Studies are a means to automate the creation and execution of multiple studies (static, natural frequency, linear buckling...). They are also the basis for design optimization studies from SolidWorks Simulation Professional. Optimization studies allow the user to drive towards an end goal for their design, such as minimizing the design's weight.
Design studies work very well and, frankly, I wish more of you would use them. At their barest they can be used to automate convergence studies by driving global mesh element size. But they also are good for testing a variety of parameters such as alternate materials (now made more easy with SolidWorks 2013!) or, for this blog, multiple gages. (thicknesses) I will document approaches to this that center around two modeling approaches.
Approach 1- native sheetmetal designs
This by far is the easiest method of the two. Native sheetmetal parts mesh automatically using shell elements. The thickness of the elements are automatically linked to the thickness of the sheetmetal feature (found when you first create a sheetmetal feature or when you edit the definition of the sheetmetal feature in the history tree). So one needs only create a parameter and link it to this thickness which is as easy as clicking on the thickness dimension from the model.
As an example, we'll use a simple single-bodied native SolidWorks sheetmetal part and we'll examine the sensitivity of its natural frequency with respect to it gage. Later, we'll return to this idea and look at the sensitivity with respect to the material.
We'll begin by creating a natural frequency study and restraining the mounting holes. The material was applied from within SolidWorks (but this won't be an issue when we want to return and test the material sensitivity) and, of course, being a native sheetmetal part the thickness comes over
automatically. Once the study has been meshed and solved we can then start a new design study by either right-mouse clicking the existing study's tab and choosing a new design study or through the Insert drop-down menu. Either way we must next create and link parameters.
Parameters can be model dimensions, Simulation features (such as a load's value or element's size), or the model's material. To create a new parameter one can start it through the design study or, again, from the Insert drop-down menu. Choose "Model Dimension" and then graphically select the thickness dimension of the sheetmetal part (as seen in this image). I renamed the dimension and showed Dimension Names so it would be graphically very obvious to us.
.
I control the thickness, or gage, using discrete values that represent common sheetmetal gages my company has access to. This link is an image of said setup.
View this photo Once you set up your Variables (the parameters) then you need to create a "Constraint."
Constraints are used to either monitor or control the direction the study takes. In our case we'll simply keep an eye on the maximum frequency across a requested mode. We'll visually plot the results later to get a trend with respect to our varying gage. SolidWorks Simulation will automatically filter the constraint type to the study type; for example, I chose to monitor the frequency of a mode shape and, therefore, only natural frequency studies (or Dynamic studies if using Simulation Premium) are available. Constraints are connected to "Sensors," something every SolidWorks user has access to. Sensors can be used to track changes in your model such as weight, center of gravity, or any type of dimensional aspect. For Simulation users one can easily watch over displacements, stresses, mode shapes, etc.
Click to see a screen grab of the sensor I used and use the Help to learn more about it.
With this approach I can capture and monitor (or even dictate through the Constraints) my model's behavior and gain insight as to what parameters have the greatest or least impact. Or course, this works for assemblies too. And with the automated nature of this I can plow through many different design ideas. Add on the Optimization and the process is further streamlined. Imagine saving hundreds of dollars in material costs alone before committing to even one prototype build and test.
Approach 2- Shell modeling with SolidWorks surfaces
I use this approach frequently when just testing an idea or when the data are imported solids and I'm not ready to convert it by inserting bends (to become native sheetmetal). The principle is to either model natively with surface bodies or use selected faces from imported solids as the basis for shell elements. When I use surface bodies I have a great deal of control over the design and the processes within it. I can still benefit from using Design Studies and replicate the gage sensitivity study from above. Let's look at a different model.
Like above, the material has been applied beforehand within the model and, as mentioned, that's not a problem for Design or Optimization Studies. But I'm getting ahead of myself. But what about the thickness? That's the meat-n-potatoes of this approach.
We'll again construct a natural frequency study, mesh, and solve it. And, like before, we'll create a Design Study. This time, however, we'll link a dimensional Parameter to an arbitrary dimension from a sketch that I constructed that has nothing to do with the geometry of the model. So it is very similar in concept to the first approach.
The parameter I linked to can be seen in this image.
Click here. And like before we'll drive the thickness of the shell linking the parameter to the shell definition. Edit the definition of the shell body within the Simulation tree and use the pull-down within the thickness field. Choose "Link value" and then select the Parameter we had made in the earlier step. We'll use discrete values of sheemtetal thickness to control the parameter for the Design Study, just as we had done previously. Create a sensor/constraint and run the Design Study.
But there is a pitfall to this approach. In Approach 1 we could make a sensor to track the Mass and hence optimize it. Since in Approach 2 you could use only surface bodies which have no calculated mass within SolidWorks (but Simulation does use the mass as an internal calc from the shell definition) you can not create a sensor to track Mass. Therefore, you can't optimize for mass in an optimization Design Study. You still have access to a potentially great deal of data representing the model's performance. So all is not lost.
Varying Material
Both methods, I've mentioned already, can take advantage of capturing the material and driving that. I find that using the Tab view of the Design Study to be easier than the Variable View.
One simply needs to make a Parameter for the material of the model and choose "Material" for the parameter type. So far this is no different from any other parameter.
Click the proceeding links for a look at both the Variable and Table views.
Click here for the Variable View and
here for the Table view.
You may be thinking at this point that Design and Optimization Studies could be a lot of work and that it would be a shame if you couldn't save this for future projects. The fine folks at SolidWorks Corporate have considered this. Design Studies can easily be saved as a raw text file and re-opened and applied to subsequent projects. So why aren't you using Design Studies?
