Understanding Linear and Non Linear FEA Using Inventor Nastran

good morning everyone thank you for attending today's webcast understanding linear and nonlinear fva using inventor and Nast ran our presenter today is Keith Dorian solutions engineer with Hagerman and company before we get started I wanted to let you know that you are in listening only mode if you have any questions during the presentations you may type them in the question panel on the right hand side of your screen and they will be addressed at the end of the presentation as we close down the session today you'll be prompted to fill out a survey and we ask that you take a few moments to fill that out additionally all registrants will receive a follow-up email containing a link for the recording of this presentation and with that I will hand things over to Keith excellent thing thank you so the the way that I kind of athlean aimed this presentation was nature is nonlinear okay and the main goal of this is to try to validate design decisions before manufacturing or construction begins and before I kind of start into the entire process I did want to just make a couple of notes about the the overall range of the simulation software that Autodesk has either acquired or developed over the last several years and we've got a number of different tools depending on the real needs of the simulations that you're trying to accomplish if you're trying to do any kind of fluid flow down a dynamic simulation CFD is an amazing tool for that today the main focus will be inside of the mechanics side of upfront design in the Nash Tran in CAD software but there are also additional tools for dealing with composites or complex materials as well as dealing with plastics and all of the manufacturing process involved with them in the manufacturing side of varying the mechanics to side of the simulation there are two tools and again I mentioned where we're using a Nazarite in CAD which is a CAD embedded toolset so it gives us the ability to do advanced mechanical simulation right inside of our inventor session if we're looking for a little bit more robust command complete dataset then simulation mechanical would be a tool that you could use to further enhance the overall environment so being able to do more complex assembly analysis or more complex interaction of components or a multi physical type of simulation and obviously with the the Nastran in CAD and the simulation mechanical space you can use the Autodesk Nash transol ver which is a very high level industry standard solution solver for both of these tool sets so Nast ran is a tool that sits inside of inventor and it can do a number of different types of simulations in this case we're just kind of looking at some of the more basic level scenarios so just a linear static analysis to see how our parts might distort under standard loads critical buckling analysis so taking magnitudes of forces and seeing if it's going to cause a structural failure finding different basic vibration information in this case the normal modes of the harmonic frequency harmonic frequency vibration values what frequencies it would vibrate in that in nature and then other tools like C transfer working with composite materials dealing with more complex contacts as well as doing some thermal stress beyond these we can also do things like drop testings Trant heat transfer in a nonlinear transient State doing dynamic response as well as nonlinear transient response and then taking a lot a number of these different scenarios and running multi axial fatigue on those components in this case we're going to be working with the nonlinear statics portion of the software and the concepts are what we're going to try to cover today is the main usage of linear analysis some of the main differences between the two how Nastran can help us look at and perform those nonlinear analysis is how to take existing simulations and convert them and then review the changes in those results and then obviously all of these things how to take your designs far so what I wanted to do in this exercise are in this this presentation was I didn't want to you know throw a large huge complicated data set within this environment so we're going to see some pretty basic simulations and just see some different ways that those you know some of these nonlinear type of scenarios may come into play and the dominant reasons are the primary usage of linear analysis is you know when you know the forces on the component are not going to change direction or they're not going to change magnitude the other big aspect of linear analysis is when you have a static or a kind of a fused assembly right a well-knit would be a prime example of this or if you're trying to work with just simulating one component within a larger design you know as parts move and collide with one another those contacts are going to change and that in turn redistributes weight and force loads on those components the other big aspect about linear analysis is we don't expect the the model to deflect a huge amount right usually like one twentieth of the longest dimension is really all we're expecting a deflection or less and beyond that you know if this the deformation to that component you know if it does start to get into some of those larger values it's not contributing to the strength of that component so some of the general assumptions kind of as I was mentioning the model doesn't move in a way that changes contacts or they are already in contact with the components that they're going to be interacting with beyond that the materials that we're dealing with they're not going to pass beyond their material yield strength so they're they're maintaining their their linear function they're not going to become plastic and again the last piece of this is the forces don't change direction or magnitude or distribution so if a face starts to collapse on itself you know that would change the overall total force load on that face that in turn would be a nonlinear type of scenario so a basic linear analysis right all of these you know in this case just a single tangler beam in this case mounted on the left side simple ten inches long it's got a thousand-pound force on the far end all of these values are easily calculated all right we don't need anything other than these equations to figure out what the total amount of deflection in this beam is going to be as well as the total amount of stress applied to you the base or the far left edge or face of this component and so basically just kind of filling in the values right we've got one inch by two inch to get our total area we can take the base and the height divide that value by twelve get the moment of inertial point and then use those moment of inertia and our length and the position of this component to define the overall deflection as well as the maximum bending stress so nothing overly complicated inside of this space just really just plugging in the numbers that we know using kind of tried-and-true equations to define these values and the software can solve these exact same scenarios so in my case I've got just a beam that's 10 inches long that's 2 inches thick I'm going to jump into the nastran environment and in this scenario I've already gone through the steps of setting all of this stuff up right I've specified in this case a material that I want to use it's just a generic material there's no real settings in here as far as this component is concerned I did make it match my PowerPoint to use the modulus elasticity as well as my poisons ratio and if I jump into the material here I can see those values so my modulus elasticity and my poisons ratio with those values in place I've already defined the thousand pound load downward negative thousand pounds and I've grounded the far back end or fixed the back end and from here I can solve I don't need to worry about meshing it I've already done that the mesh in this case would have taken a very short amount of time in a few moments we get that linear analysis and from here if I check the displacement we can see that displacement matches what we're getting in our PowerPoint so again Ness Tran solves those basic equations in the linear scenario very very quickly you know obviously in this case we're dealing with a very simple piece of geometry but the work flow stays the same throughout any kind of geometry right you know one of the big advantages tune a strand in cat is that it can solve things with different types of meshing so you know in our scenario I simply went in and said make this a solid block of material use a solid mesh if this were say a very thin piece of material or a u-shaped piece that was very very uniform throughout its shape I could have you say a beam element and just simplify that mesh even farther and that's one of those big advantages that we have using nastran in CAD versus using inventor M so some of those things that we might need to interact with this beam that might affect the overall design of this right and the first one that I want to look at is just contact so you know in this case we just kind of represented that this part was mounted to the wall or mounted to a vertical surface but how is it attached to that vertical surface and in this scenario I've built a channel piece that we've used to hold this in component in place and what ends up happening here is in this scenario as this part starts to deflect the location of the end of the part if you will moves right so instead of being attached at this back left edge it's going to slowly event eventually come in contact with this bottom surface and that in turn will change the overall output over the total displacement and the overall stresses on this component and again it's because of a contact across this face so as soon as you start getting into that scenario where contacts are changing throughout a system or you know contacts are originally not in contact and then they become in contact and nonlinear scenario is becoming apparent if I were to solve this using an Allen analysis those contacts would not update properly and we'd get just bad output usually what you end up seeing when you run this sort of a scenario is you get what we call clipping inside of this area where essentially the two meshes overlap right obviously they're supposed to come in contact with one another so there's definitely something not working properly in that linear scenario and if we'd have time at the end of the day or the end of the presentation I do have this data set that I can open it up and show what little minor changes needed to be made in order to make this function work so contact is a definite proponent of a nonlinear scenario and you know the big thing is do these contact points change and if the answer is yes then a nonlinear scenario is the route you want to go so the next aspect of this is follower forces and what this means is essentially how do the forces on this component function within this object so in this scenario that I just ran the downward force that's being applied to that component is assumed to be a constant downward force in this scenario with a follower a true follower force that shape may change or that direction of that force may change along the path of the deformation of our component so you know not only staying in contact with that face but also you know always pushing towards the direction of the normal of that face and this is really really important for specific applications you know obviously in this case if it was some heavy mass that was pulling down on that component due to gravity this isn't really all that applicable but say for example I have a tank full of liquid and I want to show the pressure on the outside of the inside of that can that tank and use that information to see how the shape of that tank is going to distort you know in that scenario the forces are always going to push normal to the outside geometry and as that geometry deforms we need to make sure that the pressure is always maintaining that constant outward direction to our model and in our case we can see how this effect occurs in our environment this will also give us a good chance to see how how to take an existing linear analysis and simply change it to be a nonlinear analysis so in our case we're just going to see if having a follower force affects our model so I'm going to switch back over to inventor here and I've got my follower force part file and in my case I'm going to switch over to mine a strand environment and right now I basically have the exact same configuration I did make a minor adjustment here to this part in the CAD model I use the split tool so that I could specify a specific area for the force to be applied to nothing overly complicated just making sure that I have a face split up and if you've not seen that function before I'll just jump over here real quick all I did is I just created an offset work plane from the the rightmost end of this part and then I use the split command here to split this space into two separate faces and the main reason you'd want to do that in the simulation environment is now I can apply specific forces to a specific area of this model so back over to mastering so in this case the first thing that I need to do in this in this simulation is I need to change my analysis type because right now that is trying to run the linear static analysis to change the analysis type I simply right click and choose edit and then in this case I'll switch over to the nonlinear static type from here a number of options have changed and one of the main pieces of the functionality here on the options tab is I want to make sure that large displacements is turned on and this is kind of a misnomer because large displacement doesn't necessarily mean this parts going to distort some drastic amount right in our case we're getting some you know point zero six inches of material movement within the you know a ten inch component so large displacement simply being on is what essentially enables the follower forces to be applied so with this turned on follower forces will automatically hat take place once you've turned on the nonlinear functionality within the analysis you will add into your sub cases a nonlinear setup inside of that space and so in this case I'll go ahead and right click on that and go to edit and basically what this is asking me for is how many times based on the load or the the simulation specified does this system need to increment it so in this case if I do a 10 it's going to take the overall load in this case a thousand pounds downward and it's going to break it up into 100 than 200 then 3 then 4 and so on and so on and so on and it's going to compare those results or not compare them but stack those results in to the system so obviously the more increments I add the longer this simulation will take but potentially more accurate the results on this model the intermediate output with this option what it allows me to do is control whether or not I get all of the increments and then be able to look through them and in this scenario because I do want to see whether or not this is truly affecting the analysis I'm going to go ahead and turn that option on if I didn't leave that on all I would essentially get is the last increment as the the solution so I want to make sure to do that so in this case I want to go in and update my material so we'll simply grab this and change our values here to match the materials that we want to work with inside of our component go back to my other page here to make sure I don't lose my numbers I tried really hard to get rid of as in many of these little steps as possible so you don't watch me type in numbers I know that's super boring to watch here but some of these values do need to be defined so we'll go ahead and specify our structural values in this case I know that I'm not going to exceed the yield strength on this component and later on we'll start to look at changing the material properties to add the number garetty if we leave this alone we just essentially get a linear linear material but a nonlinear force analysis so in this case we've updated our material I'll go ahead and mesh this component real quick to do that and then from here I just need to add my constraints and loads so let's go ahead and fix the back wall and then we'll apply that downward load on this space so in my case I'm going to use a normal to surface load select the location and then in this case a negative thousand so now that I've got my load and my constraint and my material and everything else set up that I need I'll go ahead and run new results so in this case what we're seeing here is you know as I mentioned it's looking at each of those increments in this case the ten increments that we defined and it's doing all of the simulations within that increment and eventually we'll get to that last tenth one and we'll get a full solve all right so in this case I want to look at some of the results of this to do that I simply double click on the results button and then from here I have the ability to change out what types of data I'm looking at in this case I want to see the total displacement and right now we're looking at the first increment or a total load of a tenth right ten increments plunk one and I can jump through to the last load and seeing here that my maximum loading is slightly less than what I had started with my original calculation showed the displacement on the end of this component at point six six zero point zero six six and in this case right now we're now seeing a maximum value of point zero six three if I want to display that I can simply hit display and I can see that offset value some other things that I like to do inside of the Nastran environment the first thing is I go into object visibility and I like to turn off the mesh and the original undeformed edges just to get that stuff out of the way in my case I can see that downward force might also be helpful to take the loads out just to get your overall analysis just a little bit more cleaned up right if I'm going to take screen grabs of this stuff helps to kind of clear out some of that geometry I can also animate from this space so if I did want to animate this individual load I can do that one of the other neat tricks that you can do is if you run a multi-modal or a multi a multi increment if you right-click in here you can specify to run a multi set animation and essentially what this would do is it would show me the maximum loading from point one of the load to a full hundred percent so ten set up to 100 and the idea here is that I can see the part essentially distort based on that loading versus just animating from zero to a hundred percent in here so there's a little bit of different settings there depending on what I want to see now the last piece of this puzzle is I want to find out whether or not the load that I applied is truly nonlinear right did the follower force actually have a major effect on my component and in this case to check that value all I have to do is go into my XY plot and I can check this based on the maximum displacement versus the load scale factor and as I grab this and load it we can see in here that as the as the loading was set to point one and as it ramped up through the different increments we can see that we get a very very close to a perfectly linear effect right and this is letting me know that in this scenario I'm not exceeding any kind of material limitations I'm not exceeding how this component functions within this space and later on I'll take a look at another simulation where we'll definitely get you know different displacements based on the load or or a magnitude of displacement not linearly applied to the load and so the stress and the strain are not linearly functioned and this is kind of important right because you know sometimes we don't need a linear analysis and this is telling us right away that you know we just went through the nonlinear function and the linear work just fine right it would have output the same information so the next kind of instance that you can look at a non-linearity is the geometry itself and you know there are scenarios where the stiff the component will change based on the amount of deflection in those components and the example image that I've got in this picture here is the is essentially a paper tray and as that paper tray gets pulled downward the overall shape starts to deflect and gives us some strength back and another really common example would be like the bottom of a soda can and the whole point of it is that is the pressure inside the can increases the dome will essentially stiffen and become stronger to a point and we can simulate that effect within the Nastran environment another example inside of a pressure vessel right the pressure is constantly normal to the faces the shape of the chain the shape of the tank may change which will directly affect the stiffness of the tank shape so in our case we've got a quick model of this to kind of show so what I've got here is just a couple of very basic models and basically all I did here to kind of explain this model is I created a the full shape of the tank which in this case was a 48 inch tall two foot square I then took that and cut out the data that I needed and in this case I cut it at the middle of the component so what I ended up with was essentially 1/4 of the tank now the reason that I'm doing this is that I don't need to solve the entire tank to get the data that I'm after in all reality I can apply symmetry to this component and save myself the hassle of having to solve you know essentially four times as much geometry or four times as much mesh data I then again split this component at the where the purple and the gray phases are to represent the water line within this tank so with that in play I can then specify where a pressure inside of the tank is being applied lastly because I was dealing with a solid I went in and deleted out all the faces I didn't need so in this scenario you know the the six faces that make up a standard cube and I only really need three of those six faces so I just deleted out the ones I didn't want now that step isn't required by any means right you can take very very simple geometry and use it within this space you know beyond that I could have potentially thickened this to the actual wall thickness but in this scenario it's a very thin walled component in trying to use the solid elements right a tetrahedral shape to try to solve this might be very very difficult it would also take a long time to solve because in order to get a mesh refined down that small there would be tens of thousands or even millions of elements to make up this geometry and instead I can use a shell element to accomplish that same task so in the environment I've already done a lot of this set up in here so you're not watching me do super boring mouse clicks and other values but we will go through it so the first thing that I defined is a shell element and the shell element is this the stainless steel material that I want to make this tank out of and if I come in and take a look at the material there's nothing in here out of the ordinary in fact all I did was went to select material and then use the inventor material library to find my stainless steel material had I set this up front when I before I jumped into the Nastran in CAD environment that material would have been available to me and I would have already used these values and so in my case I'll cancel that out I can here and roads are numbers so it automatically populates these values for me with no no interaction for me required secondly I specified these solid elements to instead be a shell element and to do this all I had to do is just double click here fast double-click and then specify what type of element they are and now strand can do three different kinds right solid three-dimensional elements surface components shell elements and then if I'm doing things with like a frame structure I can use linear elements to define these these so she'll elements with shells I just specify how thick that material is I didn't specify the phases that I wanted to work with the bottom and then the two walls of this component next I had to go in and build up some of the rules about what I built and the basic constraints in this case is I needed double symmetry right so I created the X symmetry and there's really nice kind of macro buttons down here in the constraint dialog for that so I just choose X symmetry and then I chose my three edges that represented where the other half of this tank existed I then repeated that exact same process for the Z symmetry on the left side of this component and then finally I ground at the bottom of the tank as if it were you know sitting on the ground so it can't move at all once I have that done I needed to create the pressure load and this did take a little bit of calculation up front to get these values essentially defined I had to look at calculating the overall volume of the liquid that was inside the tank and then use that value plus a couple of equations to tell me what the overall pressure was at the base of the tank so going into the load here we've got a couple values that we need to define and the first thing that we want to do is we selected the faces in this case the three faces that were associated to the pressure so the purple quote unquote wedded faces and then I just defined a negative pressure to make sure that it knew that it was pushing from the inside out as opposed to the outside in now in this case I wanted to distribute the pressure in the way that fluid pressures are applied you know greatest at the bottom and slowly decreasing to zero so in order to do that I jumped into Advanced Options I specified that it's a variable load and then I just picked these two corners so here and there and when I did that it added those locations into the system at which point I then specified at what scale that I want to apply that load so in this case three point zero three two pounds per square inch at the bottom at bottom face of the selected faces and then nothing at the top so now that I have my pressure load defined my model has already been meshed as a general note here if I go into mesh settings I can change the element sizes in this case three inches is sufficient I'll go ahead and regenerate this mesh and then resolve now in this scenario I did not change the simulation type so in this case we did just run a linear static analysis on this component and if I check the displacement in here we get some very random kind of strange results and essentially what's happening here is it takes that pressure that we've applied and it simply applies it a constant right and the problem with this is that as of the shape deforms in this case we're getting a very large amount of displacement in this model that pressure is not updating given the shape of the new shape of this component so we're getting some very strange results inside of this environment and again remember that we're only looking at essentially one edge of a four sided tank so I want to compare these results against a nonlinear analysis of the same type to do this all I have to do is add a new analysis to this model so in this case I'm simply going to right click and say new analysis I'll give this a new title non linear pressure and then change the type again so in this case another not linear static from here essentially duplicates that data so we can see in here now I have my nonlinear static if I want to rename the analysis I can simply double click on it just to keep track of what types of analysis is I'm running inside of the space so from here I've got a couple things that I need to copy essentially all of the data that I've created is the same so I'm going to do is simply drag them from the model portion of my browser into the analysis that I want them to be apply I do so in this case I'm just going to grab our symmetry our ground face that pressure load that I'd created the pressure load and then also my material not idealizations this come on there we go so there's our our stainless steel component so from here I just need to remesh the model generate mesh we can see that the size is a little off so let's shrink this down to the correct size there we go and again just like the last model I ran with the nonlinear analysis you always want to go into the nonlinear setup just make sure that it's got the correct number of increments you want to run typically I would probably in this scenario with a pressure model I'd probably increased this to 15 or 20 but for the sake of the simulation solved I'm I'm going to leave it as 10 I do want to turn on the intermediate so I can see what the pressures look like at half the load and go from there let's go ahead and run this so right away in this scenario we can see that the overall shape of the distortion on this component is very very different than the original linear analysis in this case and once this completes we'll be able to check that same deformation inside of this environment so if I were to go to my results here results displacement at the final increment and then display that we can see here we have a very different shape and n shape of the model as well as the displacement happening in a very different location on the component and again this is coming from the stress stiffening stress on this component as the pressure redistributes and moves around within the model beyond that if I were to check things like our maximum displacement based on the load scale factor we can see in here we are getting a nonlinear fashion or nonlinear function right in the beginning we have a little bump once the pressure starts to increase and then as it gets to a certain point it starts to essentially taper off the amount of displacement so we can see in this scenario that one the overall way that the pressure is being calculated is very different from the linear type to the nonlinear type and then beyond that the stiffening or strengthening of the part based on the shape deformation is being taken into consideration due to the incremental system in play here and again right the way that the the way the tank changes based on the pressure is just completely ignored in that linear analysis the last piece of this puzzle is the linear materials or using a nonlinear material in the space and in a linear environment linear materials are essentially proportionate to the strain right and if we keep ramping up excuse me if we keep ramping up the load the stresses and deformations will simply proportionally change along with that value so in my case on that first analysis we had a deflection of point zero six seven inches at a thousand pounds if I were to multiply my load by a factor of ten I'm going to deflection multiplied by a factor of ten realistically though that's not really realistic to our models themselves right our components have a failure point a yield strength where they're going to start to become plastic and then again or in this scenario here you know in the linear material state or any linear analysis they're only functioning inside of our plastic range of our stress-strain curves for our materials which is not realistic to the real world right parts don't live in this space all the time right we start to exceed those strengths and we get some different functionality on those components on how they perform and beyond that we may not have a simple linear state on the a single Hookes law space on it on our stress-strain curve right this is pretty typical for metals but we may have specific soft materials for example that don't function that way and obviously we have to go into the system start to define those values so there's different types of materials different types of curves plastic for example has a Hookes law section in the in the in the elastic stage and then starts to become plastic after that specific yield point nonlinear elastic as we start to push and pull in those components it has different compression strength versus different tensile strengths we get into hyper elastic components where they change their elastic States based on the loading so in our case I'm going to use our ABS component and this is information that I pulled right from the mat web website if you've never been there it's a fantastic website for finding physical properties of you know something like seventy five thousand different materials based on manufacturer based on different processes on them so there's I think two or three thousand different steel alloys alone in their site and all the data that it came out with was using megapascals and so I went through and tried to adjust these values to continue using our English units as we've been using the system is intelligent to know the difference so if I were to go in and start specifying you know 43 mega Pascal's versus a PSI value it would just make the translation that was a new feature in 2017 it knows the conversion on its own but in my case I've already done all of that work ahead of time so the orange line in this case is what I'm going to use for our model to define essentially how I want this plastic part to function you know I could go through as far as defining the true material curve star strain curve for this model however you know that would take quite a bit more input and for simplicity sake I'm just going to use our our orange value so at this point let's break our model right we know some information about our original analysis we're going to get a total stress of 29,000 psi or about 200 mega Pascal's we know that this ABS material will not be able to handle that load right it's just going to fail and if I try to run an analysis with that level it's actually going to get into the fracture phase of that material so in this case I'm going to have the load and see if the part can withstand that so in this case again I've got another inventor part back to that same component we started with let's switch over to our Nastran environment and all I'm going to do in this case is change our analysis non-linear let me make sure that our large displacement is definitely turned on and it should be by default but always a good idea to check this you know given what we're about to do this model we're going to really rip it apart so our material here right now we're using a generic material we can see that it's associated to that component so in our case we're just going to add a new material and from here I'm just going to go select our ABS or our acrylonitrile butadiene styrene it's located in our Autodesk material library ABS in this case I don't need the G value and I'll jump into our nonlinear scenario in this case this is a plastic component so I'll go ahead and specify our plastic settings and again all I did is I'm going to use the I'm going to use some of the values from the mat website to fill in these values in my case let's go ahead okay there some of these are a little bit off to what I want to use - for 885 perfect we'll set our value here 2.3 for set our yield strength 255 84 psi all right so now we'll get into our nonlinear so from here I want to change our initial yield strength and again we've got some different values here that I want to use so we'll just clear up these numbers and then from here this upper right portion of this table is essentially showing us our stress-strain curve as points along the path or key points along that path and in this case I'm just going to define the second value here to keep the material plateaued at the initial yield strength so we'll define our value for the stress and then define 0.25 then we'll define our our strain value at that limitation and if I go in and check the plot for this I can see our stress versus strain plot inside of this space so now that I've defined the linear material data for this component the last thing I want to do is make sure we apply that load to this model and in this case I do want to apply it as a negative 500 or just a lesser value here we'll generate our mesh I think our constraint is already in place we're good there so again and the way that I always use the the NASS trans space is you know if I were to forget a step I always just go up top and check everything so these are kind of broken down into the order of what you do them usually right so the first thing I want to do is make sure my analysis is set up so I just come up here and touch this and see okay well I've got nonlinear static usually that means I need to come in here and check this in our case we've got the ten increments and we'll set this turn it on the next thing I want to do is make sure our materials are set up and our idealization Zoar set up so I created a material and in this case right now it's using generic so I'm going to grab our ABS and I'm just going to apply that to the solid one now in the model section it shows us just everything that's available to us and if we create multiple sub cases or multiple analysis and analyses all of the stuff we've defined will show up in this model section and then each analysis will then lay out what it's using and in all reality I could create a new solid material and call it ABS and then leave generic alone if I wanted to be able to swap back and forth but given what we're trying to do here I'm not all that deeply concerned about leaving the generic behind in fact I could just straight delete this since nothing is now using it so I've got my material we've got our idealization as a solid part I've got my constraint and my load in place I don't have any contacts and I'm dealing with a single part so I don't need to worry about that and then as far as my mesh control is set up the size here is sufficient so I'm just going to leave it alone Jen right the mesh and then run and this is a huge differentiator than a lot of other software I used to use ANSYS workbench some years ago and the workflow to get from point A to point B as far as start to finish within the environment was very very cumbersome and you basically had to memorize every single step and once you got those steps memorized it was cake but it's a lot of work to remember those and then if you're only doing these once a month to really remember them when you come back into the software okay I go months at a time without using a strand and I can still pick it up and run with it because I know that if I start at the left and work my way across I'm pretty much ready to rock so in this case let's run this it's going to take a moment to solve this and we'll see this part get really distorted I need to put some hold music on while it's solving get myself some some Nast Ran's solver music this won't take much longer another minute you let's see if there are any questions no I don't see any questions yet all right so there's our solve that took about two minutes to run in this case as I again that I load those results I have all ten of the increments that I can tap into and in my case I just want to look at the displacement here so we can see in here we're getting 200 millimeters of displacement this is controlled by our units right now so my case I was using CAD units which this case was inches I did change this honestly I don't remember why but it's using the modified SDI which uses our linear values as millimeters so we're seeing in here quite a bit of displacement and if I display that we'll see this part distorted to that value now right now it is giving us a kind of a representative of the deformation if I jump over to our deform options I can specify this to show me the actual deformation of this component in which case in this case we can see that it's pretty much deformed to the point where it's obviously going to fail right we knew that we were getting to the point of that excessive excessive load but one of the big things about using a nonlinear analysis in pulling all of the increments up out of it is that we can use those to find essentially where some of that yield is starting to happen so in this case again if I go into my XY plot here we can see that we have a massive amount of displacement shift after about half of the load that we apply so once we get to that point 5 to point 7 range that's giving us kind of an idea of where that true value was and in our reality if I were trying to figure out where this component was going to yield under the load I could do a couple of different things and the first thing that I could do is change my analysis type to a non-linear buckling and with the buckling I can then specify essentially weight how what frequency to magnify my load and then in this case I would probably just apply like 100 pound load as opposed to a thousand or 500 and then say increment this by a factor of 1 okay so it would just 1 2 3 4 5 and so on on this component and it knows all of its strength values from material properties and it would just go through the and try to find where that failure would occur I could also set it to a much smaller value to try to get a more precise understanding of where a buckling event was going to occur on this component but the big thing here that we're seeing via the displacement versus load scale is that at some point the load is exceeding what this components what this component can yield and it drops off giving us a large amount of deformation in this component so in this case we knew we are going to exceed the limitations of the part but we can see that with n we can see that within the results and again beyond that we can see the nonlinear effects within the XY plot within the you know deformation versus scale so in conclusion you know and the linear analysis don't discredit it it has very high potential force for simulating some real-world environments but it's very easy to step into a nonlinear environment based on a number of factors right how are the forces being applied to this component is it coming in contact with other models what types of materials am i using and do they have low yield points that I'm probably going to start stepping into and then lastly you know how do the forces on my component apply to this model is it a pressure load is it a you know force that's going to change over time right and in those sort of things we can affect we can effectively solve for in a nonlinear analysis and remember nature is nonlinear so with that just find out if we have any questions let's take a look at the panel here SN curves so I believe that sine curves are referring to a for fatigue analysis and yes absolutely so we would definitely want to switch the material type to a or switch the solution during the simulation type the analysis type to a fatigue analysis and just to kind of take a look at that here inside of the environment the first thing we would do is change the analysis type and one of our types within this space is our fatigue and then beyond that this quick nur let's do fatigue and then beyond that the material itself has a fatigue SN curve that we can build so you have the SN data or the en data depending on what pieces of information you know about the component what type of material you're dealing with how do I make this section bigger I want to make I can't see much at questions there we go I've only used I've only used linear static simulator simulator and inventor which you need to input the number of iterations and check conversions before you determine the validity of the study I notice that some of the steps in the nonlinear analysis we're iterating 20 to 30 times this is simply not until you reach some predetermined convergence data yes so nastran is doing some of that work on its own in the background there are some additional configuration settings inside of the inside of the the simulation here where is the button for it it's in no climate a fatigue that's why I'm not seeing where I want at it nonlinear static and then we go into here and we can get into the Advanced Settings and then specify all the rules about how we want to define convergence and whether or not we wanted to terminate based on bad data output and absolutely right when we did that last analysis in this case where we really distorted the part a long ways it went it ran the simulation something close to 38 or 40 times before it finally determined a convergent set of data so absolutely you can specify those settings up front I don't think that there's any more questions so this will conclude our broadcast if you have any additional questions simply reply to the confirmation or reminder email you received from GoToWebinar and we can wrap those to the appropriate parties to get those questions answered once again if you can take a few moments to answer the short survey we would appreciate it it will automatically appear when you close down the session thank you for attending today's webcast and have a great day

Channel: Hagerman & Company, Inc.

Views: 52,531

Rating: 4.9693093 out of 5

Keywords: nastran for inventor, simulation, autodesk simulation, fea simulation, nonlinear, fea

Id: OnD77vjsIc4

Length: 55min 56sec (3356 seconds)

Published: Fri Dec 09 2016