Name: Springback and Warpage Compensation
Uploaded: 2025-02-13T21:30:04.459Z
Duration: 52 min 36 s
Description: Springback and Warpage Compensation

Transcript for "Springback and Warpage Compensation": Hello, everyone, and, welcome to our session today on springback and warp compensation. My name is Jason Curtis. I'm the director of applications engineering for GoEngineer, and I will be both moderating and presenting today's, session. So, I would like to keep this interactive. I appreciate, those of you who've communicated to me already in the chat. If you'll notice on the right hand side of your, display, there is a chat area. There's also a question and answer area, and I'd like you to engage with me in those ways. I also have some polls to help keep it interactive today, and, then we'll get right into some live demonstrations on some real examples. So the first thing I wanna start off with is stating a little bit of the obvious. So, the the obvious is that no part, when we go to make it, comes out perfect. Right? We spend all this time and effort into making these perfect dimensionally accurate CAD models at nominal. And then as soon as we try to produce a part, whether it's through some stamping, if you're in sheet metal industry, or, perhaps it's injection molding for plastics, you know, even even in composites, laminate layered composites, subtractive manufacturing, additive manufacturing, every manufacturing process, no part comes out exactly as we design it. And this phenomenon requires us to sometimes compensate for that. When, effects like spring back and warp, the topics of our our webinar today, when they become so, prevalent in our parts that we need to compensate for it, we need to come up with some type of corrective measure. So this is the subject of, of our, session today. This is, the premise behind all of this is how do we correct the effects of springback and warp so that we can produce, good parts faster and reduce that time from engineering to good part. So, this brings me to my first poll. We're gonna get started right away. So let me I think if I stop sharing, I can show everyone results of this. So the first thing I'd like to, understand is what kind of techniques do you guys use today to compensate for spring back or warp in your parts? Right? So there's many different ways that we can, approach this, whether it's, creating a CAD model, with some compensation built in or, you know, some type of process controls where we're controlling time, temperature, pressure. And then, of course, we have, our expert journeyman who, are artists on the shop floor when they they just kinda take care of it and fix it. So see a bunch of votes coming in. Yeah. Thanks you. Thanks for participating in that. Okay. And this is pretty reflective of what I had expected to see and what I typically see. You know, I've been doing this about twenty years. In fact, I've been in tooling for most of my career, and I've seen this evolution from, you know, fixing it and process controls to you know, companies are starting to do more and more, creating a CAD model, a manufacturing model, to, to start to compensate for the effects of Springbak. K? Of course, all of those methods are valid, ways of of, controlling, springback and warp. K? So as we look at this, I I had mentioned there all of them are valid is there's typically there's been an evolution, that I've seen. And, you know, maybe you guys have seen this as well within your own organizations. You know, fixing on the shop floor, there's always gonna be some level of this, I think. But we're we're fixing it on the shop floor tends to have, some challenges is that it tends to be very iterative. Right? So you have expert, journeyman, you know, working on the die, working on the mold, taking material off, putting material back on, exchanging, you know, die steels or inserts. This can be really time consuming. It can be a process that we just loop through on and on. It can be expensive, especially when we have to, extend our delivery dates or, you know, buy new materials for it. And then probably the biggest cost that people don't typically think about is when you now have something that you have made, the real thing on the shop floor is not even close to what you designed. So when it comes to engineering changes, when it comes to making a repair, especially if that, tool isn't in your four walls, becomes a really challenging, and costly, process. K? So process controls is another way in which we control SpringBack. But anytime we start talking about tweaking process controls, we're usually talking about making compromises. We're slowing down the process. We are, you know, changing, the feed rate. We're, holding it longer. We're, changing changing the speed. Right? All of these, have an impact on your per profit, per part profit margins. And and, frankly, this is pretty challenging too because there are a lot of moving pieces with unintended consequences when we start playing around with process controls. So, you know, as the poll indicated, a lot of you are starting to create CAD models ahead of time to try to counteract as much as we can before we start making parts. And this actually has become the what I consider the gold standard, right, for those who are making the biggest strides in compensating, and in reducing their time from engineering to shipping. This is where, I call the gold standard. Now the downside of this is it typically requires a lot of CAD skill. Right? So someone who's very good with the software, very good, knows the process. There's a little bit of an art to it. This is open to a little bit of interpretation. You know, the, the downside is that there's some model quality. We're we're making compromises again. Right? There's a lot of manual effort and and overtime put into creating these models by cutting and rotating and bending and, you know, re manually remodeling it. And this this tends to become the bottleneck in the process, but there's a lot of value in doing this. So what if I told you that there was a way to get this compensated model faster, easier, more precise? You know, instead of from days of modeling time and a bunch of headache and only one person who knows how to do it, we could bring this down to minutes, and anyone in your design group could do this and do it based on data. K? So so this is where I'm gonna launch my, next poll question for everybody is get an idea of, how, you know, what tools you're currently using. So when you need to make a decision about how much to influence SpringBack, how much to influence Warp, how much of it is just based on past experience. I mean, a lot of times, that's where it starts. Some of you may have simulation software that you're using. There is, scan data. Right? So maybe we're gonna make a couple of parts. We know they're gonna be bad, but we're gonna scan them, and then we're gonna start to fix it from there. And then there's another one that I don't see on the screen. Hopefully, you guys can see it, but, fixture gauge, CMM, you know, empirical measurements, those types of things. Okay? Okay. So not as much simulation software as I would have thought or expected. Little bit of scan data. And, of course, the, the secret sauce is in, kinda looking at your historical parts that you've created and, and and making really, intelligent, decisions based on that historical data. So great. Yeah. Thank you for that feedback. That helps me, to craft where we're gonna go here in the presentation. And and this is not an easy challenge. So, you know, taking this into consideration, you know, there is a lot of, legacy knowledge that goes into compensating for our, springback award. Right? So, when we talk about creating this compensated model, which is gonna be the focus of our demonstrations today, this is where I want to show you a method that we can bring in your CAD data. So we're gonna combine the CAD data. We're not gonna ask you to change your design tools. No reason to. We can bring this in in step data. We can bring it in in a native format. And then from there, we're gonna use what else whatever else you may have. One of these options or a combination of these options, I'm gonna show you how we can incorporate simulation data. So if you're able to spring simulate the spring back, simulate the effects of warp, and get that in a simulation data, I'll show you how we can quickly and easily apply that to the CAD model. If you don't have that, I'll show you how we can use the empirical data that, that tribal knowledge, the physical data. Maybe it's the scan data, CMM measurements. So use all of these as inputs to very quickly compensate. Put that morph that CAD so that we have the compensated model. And then we'll be able to, export that back or out to whatever next place in the process is for you. So if that compensated model needs to go back into your design solution of choice, we just send that right back. If we just pass it right on down to, CNC, that's up to you. The idea here is that, you know, what would it mean for you? And and this is a a solution I'm really passionate about. So so I ask I ask this question a lot. What would it mean for your company if you could cut 80 to 9% 80 or 90% out of the design time to compensate the model or to cut that much out of the trial phase? And trial is a nice word. I love I love the word trial. It's a very nice way of saying we're gonna make bad parts over and over again until they're not bad anymore. Right? So, what would it mean if you were able to make your first shot your final shot? So, you know, on the top, we got center line in the sheet metal industry, who's a good reference of ours, who's using this as a proven solution. GHSP for plastic injection molding, using this as a proven solution. So what we're gonna talk about today is, what, you know, what would it mean if your company would, start to apply some of the things that I'm gonna show today? So we're gonna jump right into live examples. Thank you for, being interactive with me. We'll come back to that at the end of the the, our time together today. I am going to show you, methods that we can use simulation data to compensate our our tool surface or a part surface. We'll also show some examples of how scan data can be used for the same, same end goal as well as just any measurements or empirical data. And, of course, if you have questions throughout, I'll try to catch them on the q and a session, but we'll definitely have some time, at the end here, for for q and a. So to begin with, let's talk about simulation data. So I I saw a couple of you were using some simulation tools. There are some that, are probably have more name recognition than others out in the market. Today's presentation is not about simulation, but this is time for me to do a gratuitous plug, for some of the other softwares and solutions that GoEngineer does provide. So if you are not sure how you could use a simulation tool, if you could use one, if your current one isn't doing, quite the job that you expect it to, you know, feel free to reach out to me directly, check on our website. You know, we have some simulation options. But the idea here is that once you have a simulation result that, allows you to, predict the spring back or the warp, we can take those displacements and in one click directly morph the CAD geometry. Whether it's the part, whether it's the tool surface, that's up to you. And excuse me. And I note here that it doesn't really matter where that design originates. What we can do is bring in the simulation results, apply them to the imported geometry, do the compensation, and then we can pass that back to your design tool of choice. Of course, we have multiple verification steps along the way here, but very simple, kind of, circular workflow here to do iteratively. So instead of trying out on the shop floor, we can take the spring back prediction, apply it to the part, run it back through the spring back prediction, apply it to the geometry, you know, keep going and going until we are are satisfied, with the the geometries that we're creating. Okay? So let's take a look at this in action. So we're gonna take an example of an injection molded part first. This one has a, a material that uses a lot of glass content. So this is a very rigid part, and it is going to have quite a bit of flex out here on these ends. And, since we don't have a lot of time together today, I'm not gonna show the simulation. But these are flexing up to about a half an inch, on either side. And and being a very rigid material, we we can't we can't have that. It's not going to assemble. So, you know, there's a lot of features on this part, you know, a lot of lot of depths depth and complexity. It's not a symmetrical part. So, you know, using a simulation tool is really important for us to understanding, you know, the the warp on this particular part. Now how that started was I just simply imported the model, either from a STEP or IGES, some kind of neutral format. Of course, there are, additional licenses that can unlock capabilities like, direct translators to some of the more common, you know, leading CAD tools, but Steph and I just work just fine. And then from here, we use our morphing wizard to define what is it that we want to morph and select the, displacements file from the simulation tool. Now there's no magic behind this. It's just a a tab separated file that's an x y z and then an I j k, or you can call it distance and x, distance and y, distance and z. I can import it from that file format. I've already have this available inside of my software here. So I'm going to select it from a search. And then this is going to go through all of those nodes, again, where we have the x, y, and z kinda mapped onto the onto the part, the nominal, and the vectors for how each of those points has moved in the simulation space, and we can then take and apply that on our CAD model. So I'm gonna let this complete. And if you think about, you know, how I would have had to do this without this tool, maybe I try to cut it at the end with a plane and then rotate it and then try to blend it in. Maybe I, you know, do some other, step of, you know, modeling. Maybe it's parametric, but I'll bet nobody modeled in the angle there, so I can't, like, rotate it. You know, I'm making compromises. Eight hours of work, sixteen hours of work, how you know, forty hours of work, however much it takes for me to compromise and get something representative. Whereas what I've just done in, you know, about thirty seconds of processing is apply exactly what the simulation showed to the geometry. K? So the green was the original, and the yellow is my compensated model. Now, things to note here. There is no approximation of the surfaces going on in this case. We are using a the a very powerful surfacing engine that, if your part surfaces at nominal were tangent and continuous, and everything was nice and filleted and flowed, the end model is gonna be tangent and continuous, and all the fillets are gonna be nice and and flowing well. We are not merging faces. We are not dropping faces. It is the same number of faces, the same topology, just, morphed onto the new, result. And then this could be something that you could take that, compensated model directly to, to start be, doing your shutoff surfaces and and continuing the tool design. Now from here, it's just a simple matter of exporting this then again to your design system of choice. K? So looking at this from the standpoint of a sheet metal part, here, I'm looking at the, top die surface. So this is not the the part in the CAD geometry. Actually, we took the part and we extended the surfaces and we created the tool surface, and this is what I'm gonna morph. K? But we don't simulate the tool surface. We simulate the part. Right? So how does that work? So on the bottom, if, I want to add more detail or complexities or controls, we're just gonna work from left to right on the toolbar. So we have a wizard, but then we also have a logical workflow from left to right. So the first thing I'm going to define is what we call a shape constraints, where I can define the element that we are eventually going to morph. But if there is symmetry involved, if there are any areas that we want to to freeze so that they can still stay in contact or, you know, have any features that need to stay rigid, we can begin to embed a lot more of those controls. Here, I'm just gonna define the shape, that I want to deform. Secondly, I'm gonna create a field of vectors. Now if you remember correctly, in the plastic part example, I got those vectors directly from the simulation tool. Now I've worked with all kinds of simulation tools. Some can output that format, some can't, and that's okay. Because the other thing we can do is, choose our, vector fields from two points files or, in other words, two STL files. So we need to know the correlation between where the mesh start mesh point starts and the mesh point ends. So, typically, these both come from the simulation tool. They just send us the the mesh at nominal and the the warped or deformed mesh. So here again, I will instead of picking one file, I'll pick two. So this is gonna be my nominal, STL file to fill out the first input. The target is going to be, what I've called the the deformed mesh. And now we have a, a vector field that I will, apply some optimization on because I have surfaces that are flat. I have sidewalls and flats. Again, there's a lot of complexity here. We can optimally, optimize all of those vectors. Okay? And then this becomes our vector field. So as you can see, the, let me make sure that I didn't miss it. We're good. Alright. Didn't miss an input while I was talking. As you can see, we only have points or simulation data for the area where the part is. So what happens to all the rest of, this surface? Well, we're gonna find out. So I've created a field of vectors. Now the other thing I can do, even if I don't use it now, which I'm I'm probably not going to, is I can input a, filter or influence a filter. Now the filters, looks at, the, the shape and any filtering options that I would choose to use to eliminate any kind of distortions that I might get in the part. Now I'm not gonna choose any right here, so it's not gonna filter out any points, but this feature becomes parametric in my tree, so I can go back and edit it at any time if I don't like the result or if I need to refine the result even further. So we can start from imported b rep geometry, but all the features I'm adding have parameters, have history, are able to be modified and updated. K. So that filter, all that did was just put a feature in my tree. I don't have to use that. I just add a best best practice. I really like to put it in there just in case I need to go back and use it. Now we're going to define how this should morph. And, again, I'm just putting more control in here. I'm showing you more of the behind the scenes versus just the quick wizard approach. I will choose the the vector field, which automatically knows which, shape I'm going to morph. I can choose a scale. Now scale factor is going to be multiplied against every, vector, in this. So if I want to match the simulation results or need to, I would use a scale of one. If I wanted the opposite of that, I would put in a negative one. It's a multiplier. It can be some, you know, secret sauce recipe value that you decide. Right? It does not have to be an integer. So let's just start with a one. And, you know, this allows us to define any parameters that we want to use when we apply the morphing. And then the final step is just to go ahead and do that. Let's just apply the morphing now based on our morphing law. For the vectors that only span the part, and we are going to deform the entire tool surface. So where we have data, we are going to use that data to define the morphing. And you can see that happen really quickly. K? Where I don't have data, the surface just goes along for the ride. So it you know, sometimes a question comes up. You know, we do a simulation on the first iteration of the part, but as it evolves and we do minor changes, we move a rib here or there, maybe a new hole shows up. We don't go back and simulate it again. That's okay. If we don't have points where they're supposed to be on the model, the surface just goes along for the ride. And we can use some of our filters, you know, to to take out any areas that might be negatively influencing the model. K? But those are two examples, of using simulation data to directly compensate both a plastic part and a sheet metal part. K. Let's return to the presentation and take a look at what happens when we wanna use scan data. K. So scan data, the is, again, to state something obvious, not perfect. So the difference is simulation data, that's a very good set of points mathematically computed, physically correct. You know, the physics are correct in it, but, we don't have imperfections like, grease marks, drag marks, sink marks, wrinkles, scratches, you know, anything like that. So we need to reverse engineer a set of vectors that actually make sense to use for morphing. Now, again, another gratuitous plug. If you haven't figured out a good use for scanning, so you haven't really had a need to get a scanner, maybe you'd be interested in learning more about them now that you you know, once you see how we can utilize it in this process. But we don't have to you you know, this is not, any type of hardware or software vendor specific. Again, we're really talking about bringing in that STL from the scan. Now scanners, they they pride themselves on capturing as many points as humanly possible. A lot of this is redundant and unnecessary for the operations that we are going to be performing. So we have options again to to remove the noise, get rid of the imperfections, filter, optimize this. Because what we need to do is define our own vector field where we measure the distance between those points and our nominal, object that we wanna morph, so that we can get that vector field. And then from here, the process is actually the same. So we just morph that that CAD geometry. Now because we're working with, scan data and, typically, we are not using the entire scan, because it's it's overkill and and redundant, what we would typically do is verify that any morphing operation that we, that we did still conformed with the scan or the spirit of the scan. And then, again, we can go around and around with this. Now these two methods are not mutually exclusive. You can do both. You can do the simulation first, and then you get to the end of the process. And something is unique and it's still not quite right. We can scan it. Right? So let's take a look at one example of utilizing scan data, and we'll start with a plastic part scenario. Alright? So I had mentioned we have tools for working with scans. And, in this case, what we can do is simply import STL data. Oh, I, need to choose the right option. Yep. So we can import STL files directly from any any scanning software can create an STL. Okay? Now I have already, created this scan, imported this scan. So I'm just gonna search for it here. Repeating frame. Yes. That's the one. And I'll use the scan data, not the simulation data. And let's, see if that's the right scale factor. It is. Okay. So the gray was my nominal CAD data, nominal part. This thing's supposed to be nice and flat. You know, what are the hardest kind of parts to create? Flat parts, of course. So, you know, this is a very high tolerance or a tight tolerance part, very small part. So, you know, we're seeing some, areas where the the brown is actually the scan data and it's lifting. And this is actually a really high quality scan. Normally, they're not this, this good. This was done with, I think, a CT scanner, so it's very complete. But as such, it has a lot of, all of the imperfections of a RealWood part. So we're getting a little warp, a little twist. But, here's one method of filtering out the data that we don't need to use for the compensation. In fact, if I were just think about this. Probably this large flat surface on the inside of the part, those points on this scan are probably representative of what the entire part is doing in terms of its its twist. So if I wanted to make this very minimally invasive, I could just let me hide the nominal CAD for a minute and just look at the the scan data. We could just activate an area of this mesh to use. Now this activation tool is pretty nice. This is how we do a lot of the filtering. This is where I can use a, like, a brush, which gives me a a a diameter where I can just brush strokes the the triangles or the the mesh points that I wanna use. We can do a a trap, whether it's a rectangular or a polygon or a spline trap. We can just pick some triangles. I'm gonna use the flood method, and I'm also gonna use a flood with an angle. And what that means is if I pick a triangle, whatever is defined as my angle, it will look at the adjacent triangles. And as long as it's within that threshold, it will also select that one and so on and so forth. So this is a good way to select large areas without rolling up into a fillet. So here, I just selected one triangle, and you can see all of the red is all are all of the triangles on this particular scan that, I am going to desire to use. When I hit okay, you'll see the result is, you know, all the other triangles are hidden and just these are shown. And the point of me doing this is so that maybe I can extract this as its own little subset of data. You know, I don't wanna get rid of all of those other triangles because I still wanna use it to validate that my morph turned out correctly. So we'll just go ahead and show all of those triangles again. So now I have two. I have the original scan data, and then I have just the flat area that I'm going to use for my, for my, compensation. K. So from here, we go back to our set of tools for doing the morph, but I don't have any vectors. So I am going to do a quick deviation analysis to measure the distance between my nominal part and my extracted scan data. And if I wanted to, because a lot of this is gonna be redundant, and I know if I don't have points data in between the points, my surfaces go along for the ride. I still want to have a good part that my machine machinist isn't gonna come scream at me. If I put something out there that doesn't have nice, continuous surfaces or good blends. So, you know, again, working with scan data, you have the option to say, you know what? Let's just sample every point within a quarter inch. You know, optionally, I don't have to do that. I could use every single one. I'm just telling you from experience that, you know, typically, we don't need all of the data that CMM machine or, excuse me, scanners capture. Okay? Six looks like too much, gap, so maybe we'll do, three millimeters, and we get a denser, you know, field of points that were sampled from that set of triangles. K? So I think you understand kind of the principle here. So I'm I'm reverse engineering a set of vectors to do this morphing. So, again, we'll, we'll decide which what is it that we are in fact, morphing. We don't have a symmetry plane. Actually, I probably do, but, since this is warping or twisting, I don't wanna utilize that. I'll create a field of vectors instead of from STL files. This time, I'll do it from my measurement or my deviation analysis. And I'll just hide some things here as we, go along. And this vector field, I could continue to filter. We'll just skip it this time just to show you that you don't have to have it. We'll take my vector field. We're gonna take the shape. We'll start with a scale of one. So I'm gonna make a bad part. We scanned a bad part. Scale of one is gonna make a bad part. I'm going to do this so that I can judge whether or not the the points that I selected to use for my my warp, my morphing are good because when I invert this, I'm not gonna have any way to double check that. Right? It's gonna be opposite of the of the morph. So we'll use a morphing scale of one to start with, and, then we'll go ahead and apply this morphing. K? So pretty simple process, left to right. You know, we're gonna bring in the scan data. We're gonna filter the scan data. We're gonna find, you know, the distance between the scan and the part, and then we're going to apply that that, mathematically to the, shape. K? So let me hide these vectors so we can take a look. And now we have, the yellow is the morphed model. The gray is the nominal CAD model. And if I display the original scan, you can see we're we're we're close, but not exact. Right? Close, but not exact. Now I bet we're really, really close, almost exact on the plane that we selected, but this is why I wanna do a verification step. Right? Is this good enough? That's up to you to decide. And if it's not, then we go back and we choose more points, which use more things to control this warp with. The control is completely in your hands, but it's not, an art at this point. It is using mathematical data, the distance between the scan and the nominal CAD. So what I can do is I will take my, my my morph surface, and I'll perform that, distance analysis between my morph surface and the original scan data. And we'll just take a bun all of the points. What I'm looking for here is conformance. You know? If my tolerance, say, is plus or minus point one millimeters, you know, 92% of all of the points on the entire scan are within that, I did a pretty good job, you know, using just the points that I selected. You know, again, we're not trying to mimic exactly the scan because we know there's imperfections on it. We just wanna get the spirit of that scan. So for this purposes of today, I'm gonna say this is good enough. I'm happy with that. And then, of course, to counteract those effects, I change the scale to a negative one, and it inverse inverts that, morphing. So now when I compare my original scan or the scan part, which is the brown, to the nominal, which is the gray, to my compensation my compensated model, which is the yellow, you can see we have, completed this this task. Okay? So two examples of how we can utilize scan data. And then before that, we did the, simulation data. K. Coming back to the presentation. Now let's talk about what if I don't have simulation data? What if I don't have scan data? Is there still a solution for me? Yes. Absolutely. So we're talking about, literal data. We're talking about empirical data. We're talking about that person who's been in the company for thirty years, and they know if you push it down here six millimeters and you pull it up there three millimeters that we're gonna get a good part. Maybe it's a fixture measurement to, you know, any type of measurement data. The only difference here is that, you know, we, as a designer, would convert that into something that we can use. In this case, geometries. Right? So it might be a series of points along the part at known distances and then, you know, series of points that correspond with them. Maybe it's splines. Maybe it's lines on the part, those types of things. So we can use a a set of, reference and target elements to control how we're going to deform or or warp the parts. So let's take a little sidebar into, that as an example. I also had a, an example of a sheet metal part using scan data. If we can come back to that, if anybody has any questions, I'm assuming that the the other example, you know, made sense and, you know, just be more of the same. But, if anybody wants to see that, you know, please feel free to speak up. So going back to that empirical and literal data, here's an example where my objective is to overbend, that sidewall, you know, add more draft, add more, you know, whatever. Change the sidewall within that green boundaries that I've defined just to visually get everybody in alignment with me here. And here, what we've done is we've taken a series of measurements, and I've constructed reference elements, which are the yellow. Those are on the surface as it is. And the white dash lines are where I want it to go, my target elements, after the war. So for this, instead of using, digitized data or, you know, virtual digitization data, I'm gonna use some surface morphing tools of which we have a couple. So there are ways to create crowning and kind of bump things. There are wrapping curves, wrapping surfaces. The tool I'm gonna use in this case is a shape morphing tool where I choose the, the surface that I want to deform, and then I can choose a series of references and target elements and how those, you know, should correlate whether it's tangent, continuous point, whatnot. So I'm just gonna go down the line here and choose reference and target, reference and target. The only thing I have to make sure of here is that I'm picking references and targets that correlate. Right? So, you know, I've got those four. Now if I were going to not want to influence the rest of the tool surface part surface, whichever it happened to be that I was working on, that's when I in introduce a limit element. So the screen curve is more than just for your your visual. In this case, I'm actually gonna use it to contain the deformation, contain the morph inside of this, curve. And we're going to, make that make sure it's remains tangent. I'll just change the the continuity there. So let me go ahead and just accept that. And that may have happened kind of quick, so let me show you actually, the beginning surface, which was the gray surface, and the blue, which is the morphed. Okay? So everywhere outside of the gray boundary has stayed identical, exactly the same, and then we've just influenced the the shape inside of that green boundary curve. K? So just a great example of showing how, you know, we can use geometry based morphing as well if you don't have a scanner or if you don't have simulation data or you just need to make that one off tweak on something that, that you don't wanna influence the rest of the part on. Alright. So that brings us to the conclusion of the live examples that I wanted to show. I would be remiss, though, if I did not mention that, you know, we understand that springback and warp are very complex subjects. What I covered today was how we get the geometry quickly, and that's where we've had a lot of success with our customers cutting a lot of time out of the process. But there are other things that maybe you saw, that I didn't go into enough depth on that we also have solutions for at GoEngineer. So if you don't have a simulation software that's able to predict your warp or your spring back, you know, contact us. We can probably help with that or point you in the right direction. If you are working working in additive manufacturing, three d printing, metal metal printing, like laser laser deposition, you know, metal sintering, that type of thing, we have simulation tools for that as well. And we see a lot of tool toolmakers going to three d printed inserts and molds or some three d printed areas in in dies as well, to do things like conformal cooling lines for complex, interior shapes, or or even just to handle, scenarios that would, reduce a lot of, manufacturing CNC setups. Right? So if you're interested, check that out. We also have three d scanning solutions from hardware to software. I I should have mentioned that as well with three d printing. We have, you know, a group that works with the hardware, the materials, and the software. So from end to end process coverage for, working, with the effects of Springbak and Warp, we've probably, you know, got a lot of things to offer in terms of knowledge share and, experience to to help you make some informed decisions. So with that, I'd actually like to ask you about that. If there's something that you would like to see in a future webinar, I'm I'm I'm gonna go ahead and share this next poll. This is gonna help us really to define, you know, what we do next, where we go from here. Right? Like I said, maybe you saw something that you didn't consider before because, yeah, you know, simulation data would be nice, but how do we get it back into our design? Right? If you'd like to learn more, this will just help us focus our efforts. K. I'll leave this up for a second. Everybody yeah. And then with that, you know, I thank everybody for their time. I'll, we've got about fifteen minutes left on our scheduled time together. If you do have any questions that you'd like me to answer while we're still together, feel free to use the q and a panel. Great. Yep. Thank you for your feedback. I appreciate that. Alright. So let me close this out by sharing my contact information with everyone. So, you know, having done this for as many years as I have and worked with companies, I know that there's a bit of a learning curve here in terms of you're just trusting. Right? So I just compensated five different models in, you know, the half an hour that I started this. So, you know, it can seem magical. You know what? Put me to the test. I am so passionate about this. You know, let's, reach out to me. Connect with me. You wanna talk more about it? You wanna understand, you know, how it could, how what you saw today could be used in your company? We wanna maybe take a look at trying some things out on some of your models. You know? Let me know. I would love to connect with you, and answer any questions that I can. So with that, if there are no questions, you know, feel free to drop off. I will hang on in case anyone wants to, connect with me or ask any any any questions. So thank you very much for your time today. Alright. So quick question on the molded part example. Yeah. Had a lot of ribs, a lot of holes. When the deformed part is introduced, those features are no longer in the line of draw. Yes. How do you get those features back in the line of draw? So there are couple of things. So I I did show some of the shape constraints where we could freeze features. So if there are features that are are critical that we, you know, wanna, make sure that we hold them as the shape moves, so more translate them rather than morph them, there are some controls to do that. The other way that this is handled is, you know, typically in the simulation stage, we aren't too interested in the in the draft at that point. Right? So we wanna understand the shape, do that iterative analysis, that iterative loop. You know? Do we, are we are we getting it right? When we get to the point of wanting to, go to tooling with it, then that's when that's when the draft becomes critical. Right? So at that stage, there are some tools in this software to actually defeature, the the nominal part. And this is typically what I would do is only focus on the areas that are of are are no longer, moldable, like, get a dialogue condition. Defeature, like, take the draft off and make it a very, parallel linear, rib. Or, and then on the, morphed part, then we reintroduce the draft, using the, molding axis as the as the pole direction. So, there there are sometimes some areas that you need to do that on. So, yep, we we have some tools for removing, the draft features and then applying it back on. Question, what software package is this in? So we are using the, surfacing engine of the CATIA software. As I said before, it doesn't really matter where the design originates, and you do not have to use CATIA for your designs. So, we can import the, b rep geometries, perform these, morphing, operations, and then return the geometry back to your native CAD tool to to complete the tooling design. That's typical workflow. Are you able to change material properties in the simulation? Going from aluminum to tungsten, for example. Okay. So, with a simulation tool, yes, you are able to change the material properties. So, hopefully, if you ask that question, reach out to me. We might have some, further discussions around simulation softwares. But, yes, changing the materials definitely, is part of the properties that we can change during simulation. Now today, we didn't do anything in a simulation software. Today, what we did was we used the results of that to morph our geometries quickly without having to remodel them. Let's see. Are the morphing operations in SolidWorks simulation package, Simpro? No. What we saw today was, a CATIA geometry engine doing that morphing. That that capability is not in, in the SolidWorks, SymPro package. But SIMPRO may be able to give you the simulation results. I need to know more specifics about your parts and processes. But, if it did not, then we also have other simulation offerings that work with SOLIDWORKS that could achieve that goal. Let's see. Any other questions? Alright. Well, I don't see any other questions coming in. So with that, you have my contact information. Thank you for your time today. We'll end today's session, and, feel free to follow-up with me if you have any further questions. Thank you.