Can you integrate simulation automation and systems engineering? Let’s talk about that next with Phoenix Integration’s ModelCenter. All right. So let’s talk about the integration between systems engineering and simulation automation.
Now, this review actually ties into a series that we’re running right now that looks at how do you realize value from simulation-driven design? And one of those is to use simulation automation for engineers and also for simulation analysts. So this ties in directly.
What is ModelCenter?
Phoenix Integration is a solution provider. The solution is ModelCenter. Let’s take a quick look at kind of what capabilities it provides. Here’s how it works. Basically, you define a string of things that you want to have happen. For example, we had a 3D model and we want to take that geometry and feed it to a structural simulation. You also want to take some measurements off of it and feed that into a spreadsheet that does some fatigue calculations.
This is part of what the automation tools within ModelCenter does. You can kind of define these things as a black box. You can get an output of one of them, in this case, a measurement, parametric values of a 3D model to feed to the spreadsheet, as well as the geometry to feed to the simulation. And you can string a whole number of those things together.
So in this case we’re looking at fatigue life. We’re looking at structural simulation, but then following up with that you could feed the structural deformations from the analysis to a fluid dynamics simulation. And then you would want to tabulate that all together. And that would track what were the results out of those analyses and calculations.
Now, this gets interesting when you can kind of genericize the inputs, right? You can say here are the design variables at a high level that will feed into that 3D model and then kind of cascade all the way down through the rest of these things. At a basic level, this is what capability it provides.
Advantages of Simulation Automation
Simulation automation provides a number of advantages. You can automate the whole sequence of events and eliminate a lot of manual tasks to get the results of one thing into another. It can really shorten the time it takes to run that whole simulation process from end to end, but then you can automate it. You can pertubate it. You can change the input design variables and see what outcomes that has on performance. So that by itself is beneficial and advantageous.
There are some additional capabilities that are really interesting though. It’s not just about manually changing design variables and seeing what effect they have on the outcome. The next step here is to connect this with design exploration capabilities. Whether that is:
- a sensitivity study: progressively going through a series of changes to the design variables
- an optimization routine
- a Monte-Carlo kind of simulation: varies the design variables, so you have more complete coverage of the design space.
Those are all things that can be done within ModelCenter. It gives you these plots that let you look at the performance boundary. As I changed design variables, where is the best performing one, the worst performing one, but also the boundary of performance, so you can better understand what’s going on and kind of the limits of innovation within the design.
Automated Exploration of the Design Space
Let’s take a second to talk about that kind of automated exploration of the design space. It’s not going to be used in every case, but for highly engineered components or aspects of your products, it can make a big difference. This kind of design exploration can make a huge difference in making sure you hit your requirements, lowering costs, and shorter schedules.
If you’re faced with a set of really aggressive design requirements, it might take you a long time to find a feasible design. With this capability, it really shortened the time needed to do that. Now, where this really gets interesting and unusual is when you connect it in with a model-based system engineering (MBSE) definition. There’s a number of different tools out there that provides this capability.
You can pass the variables that you have defined in your MBSE model into this loop and run it with using those design inputs as at the beginning of your simulation automation. Then, you can track the results against that simulation. So there’s this two way integration.
So, first off the MBSE model passes the design variables to ModelCenter and then ModelCenter will pass the results back up to the MBSE model. Here’s what’s interesting about that. There’s this configuration that you have of the product for the system within the MBSE model. Then you get the results of how that configuration performed against the requirements that are very explicitly defined in the MBSE model.
It closes the loop here. Not only does it allow efficiency in terms of getting information from the MBSE model over the simulation, but you can also track how each configuration performs. You can imagine if you automate this, especially with design exploration capabilities you can really understand what affects what.
What are the takeaways here? There’s some simulation automation tools out there that are kind of like this. ModelCenter is a little bit unusual in that you have this drag and drop UI where you can develop these workflows or automations. The integration with MBSE tools is unusual too. I haven’t seen that anywhere else as of yet.
In closing the loop between these – getting simulation running alongside the MBSE model definition – there is a big advantage there, especially with the automation passing the design variables over, getting those validated or seeing how they perform and tracking that back and have results alongside the configuration of the system. That’s big, especially with the automation involved. This is a really powerful tool set.
If you have MBSE stuff in place, then this could be a really powerful tool to add in to close the gap between those two sets of capabilities.