3D Modeling & CAD for Motorsport: Simulation and FEA
Watch This Course
$199 USD
-OR-
Or
8 easy payments of only $24.88
Instant access. Easy checkout. No fees.
Learn more
Simulation and FEA
14.58
| 00:00 | - We discussed FEA briefly in the analysis section of this course and in this module we're going to take a deeper look at load simulation and finite element analysis. |
| 00:10 | This material isn't intended to teach you everything you need to know about these techniques and how to use them in CAD. |
| 00:16 | These are advanced processes and to get useful results and understand the results, you're going to need some sound engineering knowledge. |
| 00:24 | Covering all of this would require its own course so think of this module as more of a starter or introduction so you can get familiar and work towards some useful results. |
| 00:34 | On top of this, the software required gets fairly expensive so it's not the most accessible process. |
| 00:41 | Finite element analysis or FEA for short is the method of simulating how a model will react to real world forces. |
| 00:49 | It's also sometimes referred to as FEM or the finite element method. |
| 00:54 | We can think of FEA as a divide and conquer approach, where calculating the simulation for an entire system at once would be too difficult. |
| 01:02 | And to clarify, by system I mean the entire model which could be a single component or an assembly. |
| 01:09 | This method breaks a large system down into much smaller, more simple parts called finite elements. |
| 01:16 | The representation of the total system in the form of these smaller elements is referred to as the mesh. |
| 01:22 | Calculations are made for each element in the mesh and then the individual results are combined to find a final result for the total system. |
| 01:30 | FEA can be used in a range of different simulations including structural analysis to find the internal stresses, deflection and deformation of a part when subject to loads as well as heat transfer and fluid flow. |
| 01:44 | Now fluid flow actually falls under something that you may have heard when aerodynamics are discussed, CFD, which stands for computational fluid dynamics. |
| 01:54 | This is a way of modelling the flow of fluids, including liquids and gases and their interaction with solid surfaces. |
| 02:02 | It's a very complicated topic that again would require a pretty advanced course of its own so we won't cover any CFD but it's still good to know at least what it is. |
| 02:11 | In automotive and motorsport applications at a professional level, these simulation methods are commonly used to understand and optimise the strength and weight of designs as well as other physical factors like stiffness and deflection. |
| 02:26 | There are a range of programs that can be used for simulation. |
| 02:30 | Some are designed specifically for this work such as Ansys, or more more modelling focused programs that also have simulation functionality such as Solidworks or Fusion 360. |
| 02:41 | Generally a simulation specific program will have more functionality in this area but it's still possible to get useful results from either. |
| 02:51 | In saying that, getting useful results is very reliant on the setup of the simulation. |
| 02:55 | The phrase rubbish in, rubbish out definitely applies here. |
| 02:59 | Our simulation must be a good representation of the reality that we're trying to simulate. |
| 03:05 | There are always going to be inaccuracies because between each calculation point in the mesh, there are approximations. |
| 03:13 | In general, the finer the mesh, the longer the time to solve the calculation but the more accurate the result. |
| 03:20 | It's always important to validate any results we get in Fusion 360 with real world testing. |
| 03:26 | That doesn't mean the results can't be useful for making comparisons between different versions of a model though, we just need to keep as many things consistent as possible. |
| 03:35 | After the design has been modelled and the material defined, the steps to solve the simulation using FEA are generally the same across all platforms. |
| 03:45 | First, a mesh of the model needs to be generated, then the boundary conditions or constraints set and the loads applied. |
| 03:52 | The order of these steps isn't too important and can change depending on the program we're using. |
| 03:58 | After this, we can run the simulation and get the results. |
| 04:02 | Like I said, we won't be going too in depth on FEA in this course but there's definitely no harm at looking at how these processes work in Fusion 360. |
| 04:11 | For our example, we'll use this sheet metal bracket for mounting a catch can in the engine bay. |
| 04:17 | The aim here is to understand if the bracket is strong enough to support the catch can when it's full of oil. |
| 04:24 | Is using FEA to figure this out overkill? Absolutely but it's a nice example that'll work well to give you an idea of how the process works and the results. |
| 04:34 | We could also complete this study on the assembly with the catch can bolted to the bracket but the more components we include, the more complicated it gets so we'll just use forces to represent the catch can's weight instead. |
| 04:47 | With the bracket model open, we can navigate to the simulation workspace. |
| 04:52 | We're automatically shown the new study window. |
| 04:55 | Here we're given a range of different types to choose from and clicking on each of them gives a short description to the right. |
| 05:03 | In order to get the answers we want, we can choose static stress to analyse the deformation and stress in our model and click create study. |
| 05:11 | We might get a popup window here showing information about the cost to solve a simulation or complete other tasks. |
| 05:19 | Here's where it gets interesting, simulation is a Fusion 360 extension, essentially meaning it costs money to use, even if you're using a paid version of the software. |
| 05:30 | There are some options when it comes to paying for extensions like subscriptions, cloud credits or flex tokens. |
| 05:37 | The best place to find information on this is on the Autodesk Fusion 360 site. |
| 05:42 | At the time of this course being filmed, basic studies cost three tokens and more advanced studies are six tokens, so at $3 USD per token, we're looking at a minimum of $9 for each study. |
| 05:55 | Moving on, our toolbar is laid out so we can work from left to right, we'll skip the study and simplify tabs because we already have a study set up in the browser, and get started on the material tab. |
| 06:07 | If we select study materials, we can specify a range of different materials to use in the study. |
| 06:14 | By default this is set to the same as the model which happens to be steel but let's say we want to simulate stainless steel. |
| 06:22 | We can simply apply this change by editing the study material. |
| 06:26 | We can now select OK to make the change. |
| 06:29 | Next up are the constraints. |
| 06:32 | As the name implies, this is how our model is constrained, fixed, mounted or held in position. |
| 06:38 | It's what supports our model. |
| 06:40 | In this case we want to think about how it would be mounted to the engine bay. |
| 06:44 | Under the constraints tab, we don't need to worry about the two connection tools as these are for connections between component bodies and we're only dealing with one body. |
| 06:56 | The first constraint will be for the bolt holes. |
| 06:59 | We'll use the structural constraint tool and select the two front edges of the mounting holes. |
| 07:04 | Selecting the constraint type as fixed, select OK to finish. |
| 07:09 | To be clear, this definitely isn't the best way to represent fixings through these holes. |
| 07:14 | As you can imagine, the actual interface between a bolt and the hole isn't this simple but this will still give us a reasonable result and is enough for this example without overcomplicating things. |
| 07:25 | The other constraint we need to represent is the actual surface of the engine bay that the bracket is fixed against. |
| 07:32 | Again, we can use the structural constraint tool, select the back surface and then the frictionless constraint for simplicity. |
| 07:40 | This will simulate the back face being against the surface so it can't move back in that direction through the mounting surface. |
| 07:47 | But it also isn't preventing any other movement if the bracket were to flex. |
| 07:52 | For example, peeling off the mounting surface along the top edge. |
| 07:56 | Next we want to apply the loads to the part. |
| 08:00 | The load case will be a result of the weight of the catch can and whatever oil might be inside it. |
| 08:05 | So if our catch can is made from aluminium and weighs 2 kg and can hold half a litre of oil, we might expect the catch can to weigh no more than 3 kg with some fittings and partial weight of the lines connected to it. |
| 08:20 | It's safe then to assume a downward load of about 30 newtons. |
| 08:24 | But we'll assume the worst and call it 40 newtons. |
| 08:27 | Let's apply this load to the top edges of the three small mounting holes for the catch can. |
| 08:33 | Although it's not a perfect representation of reality, this will work fine for the purposes of this example. |
| 08:39 | We want to leave the force per entity preference deselected so when we enter our force, we can enter the total 40 newtons and it will be divided between the three load points. |
| 08:50 | Now we change the direction type to vectors and set -40 in the Y direction for our 40 newtons downward load, then click OK to finish up. |
| 09:01 | Moving along the toolbar, we have contacts. |
| 09:04 | Again we don't need to worry about this in our case as we only have one component body in the model. |
| 09:11 | However, if we were using a full assembly for the stress study, we'd define the contact limits between the parts here. |
| 09:18 | The final step before solving the study is to generate the mesh. |
| 09:22 | Under the solve drop down, we simply use the generate mesh function and we can see the mesh generate, showing the edges and nodes between the finite elements. |
| 09:33 | Under the browser or the manage tab, we can edit our mesh to refine it. |
| 09:38 | As we already touched on, generally the finer the mesh, the less approximations and the more accurate the result but the longer the calculation time. |
| 09:47 | It is possible to refine the mesh around key areas of the model and make it more coarse than others to get the best of both worlds. |
| 09:55 | However, meshing is a detailed subject in itself and can become quite involved so at the risk of overcomplicating things, we'll leave it at that. |
| 10:03 | Just understand that there's a lot that can be done with the mesh and it has some significant effects on the results of our simulation. |
| 10:11 | Before we move on, notice how all of these factors in our setup are captured in our browser. |
| 10:16 | We can come back to these and make changes if we need to and also change the units from here. |
| 10:22 | Let's also use the pre check tool under the solve tab to make sure we have all the required data for this study. |
| 10:29 | In this case we do, we can actually see this by the tick being green. |
| 10:33 | We are now ready to click the solve icon in the toolbar. |
| 10:37 | We should get a popup window notifying us at this point of the expense to solve the study. |
| 10:42 | After we confirm this, the model will take some time to solve and then display the results. |
| 10:48 | You'll notice that the workspace changes and we get this results toolbar to allow us to work with the results we've been given. |
| 10:56 | The first thing to look at is the minimum safety factor value in our results detail window. |
| 11:02 | Which basically tells us how much stronger our model is than it needs to be, to withstand the applied forces. |
| 11:09 | So as we can see here, we'd need to see around five times the applied force for the part to break. |
| 11:16 | We're also given this short note saying that our design isn't expected to permanently deform or break. |
| 11:22 | In the model space, our model is now illustrated with a colour map. |
| 11:26 | The colours are accompanied by a legend and they represent the intensity of whatever our legend is set to. |
| 11:34 | By default this is set to the safety factor but we can change this to show stress, displacement, reaction force and strain. |
| 11:43 | We also get the maximum and minimum for each of these shown on the model. |
| 11:48 | As we should expect from our knowledge of stress concentrations, the area of most concern is around the bend on the internal radius. |
| 11:57 | The deformation of the part is illustrated in the model, however by default it's scaled up with the adjusted setting to make it easier to visualise. |
| 12:05 | We can change this to show the true deformation or no deformation at all. |
| 12:11 | Back on the topic of safety factor, this is based on the yield stress of the material which is specified in our material preferences. |
| 12:20 | A minimum safety factor of five is pretty safe. |
| 12:23 | We could easily afford to change the design of the material to remove some weight but we also need to consider that this is just the weight force of the catch can while the vehicle is stationary. |
| 12:34 | Introduce some G force and it's a completely different story. |
| 12:38 | This design might be OK for a street car but it will most likely fail in a race car where the G forces are much higher. |
| 12:46 | That's not to mention impact force, fatigue or the wear and tear associated with removing, reinstalling and servicing parts more frequently in performance applications. |
| 12:56 | This is why there's so much more to FEA and stress analysis than what we just covered. |
| 13:02 | There's a wide range of other tools and processes here but we've just touched on the general process and ideas in order to get a basic understanding. |
| 13:09 | The next steps in our FEA process would be to finish with the results and revise the model to optimise the outcome. |
| 13:17 | Coming back to the results to make comparisons or adjust the settings of the study to try and increase the accuracy of the simulation. |
| 13:24 | Let's stop here before we end up going down a very deep rabbit hole and recap what we've covered in this module. |
| 13:31 | FEA is a method of simulating how a model will react to real world forces. |
| 13:36 | This method essentially breaks our model down into much smaller more simple elements that can be solved and then considered together to solve the overall result. |
| 13:46 | This method can be used to calculate structural stresses, thermal transfer, fluid flow and various other physical simulations. |
| 13:54 | In Fusion 360 we use the simulation workspace for these studies and solving studies does come at an expense. |
| 14:02 | When working through a static stress study in particular, we follow a general process from left to right on the toolbar of specifying the material, defining constraints, applying the loads and then generating the mesh. |
| 14:15 | After these are complete and we have used the pre check tool to confirm we've covered everything we need, we can then solve the study. |
| 14:23 | Our results workspace will provide us with the resulting safety factor, stress, displacement, reaction force and strain information. |
| 14:31 | Essentially showing us if our part will break, deform and by what amount. |
| 14:36 | From these results we can work to optimise our design and make comparisons between different revisions, materials or parts. |
| 14:44 | It's also important that we experiment with the setup of the study to understand the accuracy of the results and where possible, validate the results with real world testing. |
