Increasing Mold Throughput with Conformal Cooling

KRISTEN KILROY: All right, so if you haven't noticed, we did change the title a little bit from the original post. But it's the same content that we described. Basically, if you sat in on Eric's presentation earlier today, going into a little bit more detail and a case study behind how to use PowerShape and Moldflow together to essentially create a better mold design and showing you how to use both of them.

So, today, I am joined by Jon. My name's Kristin Kilroy I have a background in plastics engineering from Penn State Behrend. And previous to Autodesk, I worked as a project engineer and facilities engineer for a custom injection molder. So I was actually in there creating the tooling designs for customers that we would work with. And with that experience, I took it over to Autodesk, joined the Moldflow technical support team, so really got the ins and outs of issues with the programs as well as issues customers ran into.

And now I am on the technical marketing team and product marketing team to help show customers the possibilities with Moldflow. And Jon?

JON HUNWICK: I think there's only three people in the room that don't know who I am. I'm Jon Hunwick. I joined Delcam in 1984. I was going to make a joke about saying, oh, come on, you can't possibly be that old. But I'm not going to do that now, because it's kind of spoiled, because everybody knows exactly how old-- anyway, we won't go into that. But I spent time, while I was at Delcam, in UK support and sales, international support, where I worked for a long time with Craig. Then, I moved into the marketing department.

And when we made the transition to Autodesk, I become the technical marketing specialist for PowerShape and, recently, for Netfabb.

KRISTEN KILROY: So, as I mentioned, today we're going to go a little bit in-depth into the challenges mold designers face when they're designing that tool, figuring out how to work with the customer, and so forth. We're going to look into what PowerShape is and what Moldflow is. And then we're also going to look into how the simulations between both of those help to reduce cost with tool designs. And then, finally, we'll do a case study-- a part, again, Eric used earlier today. We'll take a look at a conformal cooling example to show the validation behind simulation to see the different variations between cooling. And Jon will take it away.

JON HUNWICK: Yeah. Stealing a question from Eric's presentation earlier, what is the most expensive type of mold tool that you can make? There's only three people in this room-- two people in this room, actually-- that I think are qualified to answer this question, because the rest of you are cheating. So what is the most expensive kind of mold tool that you can make? Anybody?

KRISTEN KILROY: Anyone?

AUDIENCE: Same question.

JON HUNWICK: Same question. One that doesn't work. One that produces a scrap part. It's a question that we stole from Eric, but he probably phrased it better than I did. Mold makers are under a lot of pressure all the time. They're increasingly under pressure to improve part quality while shortening lead times. They're very often in that horrible catch-22 situation that they actually must start their mold design work before the part is completed. Because very often, parts design could change right up until the very last moment, just before it goes into production.

Often, the supply part geometry is not manufacturable directly. You've got to do quite a lot of work to make sure that you can make it. And it's often, unfortunately, still provided in formats such as IGES, which is notoriously unreliable. But even when it's not, even when it's provided as a proper native CAD data, there can still be features and problems with that model that make it so that you cannot physically manufacture it.

We've seen, particularly if you watched Eric's presentation this morning, part geometry is getting much, much more complex. Things are getting much more intricate. You look at a modern mobile telephone, the intricacy of the inside of that part are just getting more and more complex. New tools, new technologies appearing all the time. So everybody's got to keep on top of those, and that means that you've got to be continually retraining, continually gaining new skills.

And a big one is that mold makers very often get paid until the finished parts are being made. So if there's any snags, it's down to them to fix them. So it's crucial that they get parts right the first time as often as they possibly can. So, again, this is a largely redundant question, because we have loads of people from the ex-Delcam organization in this room. How many of you do not know what PowerShape is?

OK. All right. Well, PowerShape, in my view, is a suite of CAD tools that's specifically designed for dealing with the challenges of taking a CAD model and making it into something that you can physically manufacture. And that's where its strength lies. It's fully integrated surface, solid, wire, and mesh modeling. It's very tolerant of poor-quality data. You can get away with absolute murder sometimes in PowerShape, and you can do things that other CAD systems physically would not allow you to do. Because it's quite happy to work with models that have got holes and gaps. As long as that's not going to interfere with the machining process, PowerShape is quite happy to deal with that kind of data.

It also has specially integrated tools for designing whole mold stacks out of standard plates and parts. It's got an EDM module for designing and helping you to manufacture EDM electrodes that you need for making parts with very fine, small detail. And we've recently added a new ribbon-style interface to the program, which gives an instant air of familiarity. So if you use any kind of ribbon-based products in the past, you should be a pick PowerShape up fairly easy, because things are reasonably well organized.

There are several things that make PowerShape uniquely capable, I think, to doing this modeling for manufacture. The first is that we have a Solid Doctor, which works in a very different way than the solid healing tools that you get in most other CAD packages, in that you don't get just a single fix for each type of fault. But you can basically use any command inside PowerShape to repair that fault. The difference is that the Solid Doctor will categorize the fault, show you exactly what you're dealing with in a very easy-to-read tree format. And then, it'll allow you to go through and just choose whichever tools are most appropriate to fix that particular problem.

It's got direct modeling, which we all should be familiar with nowadays, the ability to just grab faces of a model, move them around, shift them offset, then rotate them. Add draft. It's got an automated die wizard. Oh, yet another Delcam person. Oh, goodness sake. They never end. But turning up fashionably late, very good.

It's got a die wizard, which will pull the part into as many separate cavity/core pieces as you need, complete with all the slides and side cores, side actions, as well. It doesn't make the assumption that a part is only going to open and close in z. It allows you to pull/pop faces out in any direction you need. The surface modeling is very, very easy to use. I'll show you a couple of examples in a few minutes. They're going to all be video. There's no actual live demonstrations in this, but most of the video is going to be shown real time.

So the surface modeling, basically, you just select the curves you want to use and say, make me a surface out of this, and PowerShape just gives you its best guess. You can tweak it if you want, change it if you need, but most of the time, you don't need to. It's a full hybrid system, which means that it can work equally well with any type of [INAUDIBLE]. You can work with solids. You can work with surfaces. You can work with any combinations.

So you can have a part that's mainly solids but with a few surfaces in it, or the other way around. Or all surfaces, all solids. Whatever it takes to get the job done, you can do. And very importantly, it has a direct connection to PowerMILL, which is our flagship CNC machining product, which will hand virtually any kind of machining tool out to five-- well, in fact, fully up to multi-axis robots.

KRISTEN KILROY: So, similar to what Jon had mentioned, how many of you know Moldflow? I'm guessing most of you at least have heard it around. How many of you have actually used it? Anyone? OK, very few of you. So, basically, Moldflow is a simulation tool that is used to simulate the plastic material flowing into the cavity. Now, that's not all it can do. It can do several other things, including predicting the deflection of the part, predicting cooling circuits, and looking into the thermal aspects associated with that material flowing into the cavity.

So with material databases, the material database within Moldflow is also a powerful tool. Because there's thousands of materials preloaded in Moldflow, in addition to materials that you can have tested on your own and import them into Moldflow. So it really gives that customization, being able to predict the exact material that you want to use in the long run when you're manufacturing the part with what you're using in Moldflow. Because slight variations could turn into catastrophic differences between the two.

In Moldflow, there's also an integrated CAD kernel, and that's used for direct import. So you can import Inventor models, you. Can import SolidWorks, if you're working with those. And then you can also import non-native files, such as IGES, as Jon had mentioned, and STL, STP, and so forth.

Again, the cool flow and warp simulations, in addition to 12-plus processes that you can do. So, earlier today, Eric was describing the capabilities of Moldflow. You can do injection molding. You can do gas assist. You can do mu cells through a microcellular injection molding, and several others. And then the advanced analyzes within Moldflow enable optimization tools, as well. So there's parametric and design of experiments analyzes that you can run to help with that design process, either if it's with the tool itself, if it's with, say-- if you're doing a compression molding, you can even alter the shape or the volume of the charge in a parametric study and see how much material you'd need for each of those shots.

And then, finally, the results in Moldflow. As you can see here, this is a fill-time result. Very graphic, very easy to look at. You can tell right away, you're going to have weld lines on that part.

JON HUNWICK: So the process a mold designer would typically go to, he'll start usually with a full design. And the mold design is usually based on his own experience. So a mold designer would look at a part, go, I've seen parts like this before. This is I'm going to make it. This is where I'm going to gate it. This is how I'm going to inject it, and so on.

Then you go on to step two, which is to manufacture the part. Then, you have to try to part out to make sure it works. And if necessary, and it sometimes is, you have to redesign and rebuild the mold. And Eric will tell you that he has customers that began a mold project over a year ago and still do not have a fully working tool. But they just accept that 1 part in 10 is going to come out scrap, something like that. And then, finally, take the part into production.

KRISTEN KILROY: Now, when we start to incorporate simulation, like simulation Moldflow, into the game, that's when you really see a difference. You're able to iterate between different designs, like I was mentioning with those optimization tools, as well as see those advanced options that you might be thinking about in the back of your mind but not necessarily sure if you should incorporate it into that tool. Bringing the model into Moldflow essentially saves time in the long run, just because you're adding the extra security to the design of the tool.

JON HUNWICK: So some bits of the process, obviously, never change. The amount of time it's going to take you to manufacture the tool will never change. But the bit where you get the real big benefit is where, in a traditional mode cycle, you do a test and a rebuild, particularly if that test actually happens off-site, as it very often does. Because you've got to take the mold down off the machine to build it. You've got to ship it out. You've got to test it. You've got to figure out, damn, it doesn't work. What am I going to do now?

You can try and change the parameters on the machine itself to see if you can sort things out. And that process can take several weeks, three to five weeks, or one to three weeks is not uncommon. Eric was telling us this morning about cases that he knows that have lasted considerably longer than that. It's a struggle to get a tool correct. And sometimes, it's so severe that your only option is to scrap it all completely and start again from scratch. So not only are you are losing a lot of time, you're obviously costing a lot of money during that process.

KRISTEN KILROY: Now, on the other end of the spectrum, again, with the simulation design tools, using PowerShape and Moldflow, you're able to work within PowerShape, send it straight over to a simulation, find out that, wait, that design didn't work. Send it back over PowerShape, tweak a few things, and save that amount of time with preventing cutting steel until you have that understanding for the simulation. And then, with that, you can also be doing other things during that, because that's all done on the computer.

We have cloud solving capabilities. You can have it run through Moldflow while you're doing other things, including maybe designing the tool at large. And then you're only looking at specific details within the simulation. That's where you can go in later, after the simulation is completed, where you have that understanding to go back and fine-tune the tooling designs.

Now, with Moldflow, there's two flavors, really. And when we're talking about the tool designer, I really wanted to show the capabilities in the eyes of the tool designer. The tool manufacturer basically just wants to get that tool cut, ready to go, and, like Jon was mentioning, get acceptable parts, even high-quality parts, for that customer. So with Adviser, it's more of a more basic tool but still has really powerful capabilities, such as the wall thicknesses.

If you're getting that model from someone else, you might not necessarily know if it's set up to understand the thicknesses behind there. And then you can also see undercuts, draft angles, and so forth, which, again, PowerShape can help to eliminate those. So if you do identify undercuts in the model, take it into PowerShape real quick, send it back into Moldflow, and you have it fixed right away.

Then, there's also a magnitude of advisers within Moldflow Adviser, including Design Adviser and Cost Adviser. If you want to provide that higher quality feedback to your customer or whoever is actually making the final decision on the tool build, you can run through those advisers and get an estimate for justification for the current setup of the tool.

And then gating locations and runner balances, that's something that usually falls on the tool designer. Sometimes they work with the customer, but it could be throwing a pebble into a pile of rocks. You might not know what you get out whenever you design a feed system. So this helps you to find the ideal system with lowest pressure and fuse the weld lines and so forth.

So, moving into the more advanced tool Moldflow insight, like I mentioned, you can get those design optimization tools, like parametric design studies. You're using that CAD geometry. So you design the tool or design the cavity in PowerShape, bring it into Moldflow, but want to tweak something real quick just to see what the differences might be. You can actually move features around the part, as well as increase and decrease thicknesses and so forth.

And then you can also simulate mold inserts and part inserts. That's really powerful when you're working with a tool designer, because they're trying to figure out, how do we get heat out of certain areas? You might have to add an insert. Cloud solve options, like I mentioned, set it and forget it, basically. Have it run through on the cloud rather than on your computer so you can work on other things while it's running.

Core deflection, as well. In addition to part inserts, maybe you want to analyze how a core is deflecting. If you have, say, a tube that you want to mold, that's going to have a narrow, long pin within it. And Moldflow can basically see how that is forming with the material pushing it back and forth. Advanced cooling and heating, that's where induction heating, conformal cooling, and so forth come in. You can model these within the simulation, and it provides that feedback, that thermal feedback on it, to let you know how that's actually working within the tool.

And then, again, advanced processes, like I had mentioned. You've got several others, including injection compression, compression molding, and so forth. And then, finally, valve gating and even hot runner systems, you can simulate those within Moldflow, as well.

So this is just showing the basics behind Moldflow Adviser. This was done real time, and I only sped it up 2 and 1/2 times. So all in all, from setup import to basically starting the analysis, it took me 7 and 1/2 minutes. So right here, we're just adding the injection locations. And then you're able to customize the tool size. If you're not sure what it is, you can play around with the different drag points or you can enter in an actual value for those. And you can see there, a grid really helps out to position that.

Now, with this tool, from Jon's animation earlier in this presentation, you saw the slides need to be formed-- or need to use to form those under-cut sections, the tubes. So that's why we created the tool that much larger than the actual part itself.

JON HUNWICK: In this particular case, because of the relative length of those tubes, we need to use hydraulic slides rather than pin-activated slides. Because otherwise, the angle of the slide would be so acute that the slide pit itself would probably break.

KRISTEN KILROY: Exactly. And here, we're just showing the creation of the feed system. So before manually drawing it here, I went through the wizard. That's another thing about Moldflow. We have wizards that, essentially, create that feed system for you, if you'd like. In this case, I wanted a specific design. I just wanted it relatively in the center of the three gates there. So I went through and just drew it on the grid there.

And again, with Moldflow Adviser here, it just clicks to the specific gate, so you're sure they're connected. And you can play around with that for however long you want, until you get that perfect setup you want. And right there, I just showed the blue tubes that went a little quick on the screen there. That was the wizard for the cooling system. So we have a wizard for that, as well. Here, the more square system there was for the import option within cooling.

So after setting up the tool, basically, all you do is run through the next steps within the study pane there. So we're choosing the material right now, showing the vast amount of materials within here. And an important thing to know is all those are tested. We have tested material data for those that we're able to use for it.

Next step would be setting up process conditions. So you can either leave them automatic, which allows the solvers more control to be able to show the estimate of the best process, or you can control it, if you know you need a specific cycle time. Maybe you can only inject it for two seconds. And then, next step is to let it run.

So again, this only took 7 and 1/2 minutes for me to do. Pretty quick. Only takes less than 10 minutes out of your day to design it for the toolmaker, because I know toolmaking, or tool designing rather, is pretty busy. So this is very helpful.

JON HUNWICK: The key fact is that if you've done this simulation up front, then you are 99% certain that when you actually build the tool, it is going to work. So, I presume that everybody really knows how molds are traditionally cooled. It's something like this. Through this steel plate, you just drill a series of holes that interconnect to form the pathway that you want the water to follow.

If you don't want it to go down a particular route, then you put in a plug. It's a screwed plug that you just put in with a very low Allen screw and seals off. You can see that happening at these particular points here. So the water comes in here and then follows that. You've got a hole here the water can get out. You've missed one here, but never mind. So this tool would leak, but never mind. We won't go into that. Or you can have a much more simplistic circuit where you just allow the water to go pretty much where you want it to.

So the answer is that, basically, the water simply follows some straight lines go around the shape of the part. In this particular case, for reasons we'll come into a bit later, we actually put a baffle in, in the middle, so we could force the water to flow up and over and give us much more cooling at the top of the part. But the reason that we chose to do that will become clear in a minute.

KRISTEN KILROY: So, looking at the sample part, the first thing I notice when looking at it is there's a lot of heat in that center region. If you look at the side there, it's pretty thick around that center rib on both sides, as well as the wall thickness just in general in that center cylindrical core there. So within Moldflow, just did a quick dimensional diagnostic. And you can see that it is kind of on the heavier side for wall thickness compared to the rest of the part.

So it's a good thing that in that adviser video that I'd showed earlier, we did gate towards the top, or in that center region, because it is the thicker region of the part. Traditionally, in injection molding, you want to gate thick to thin. That's just the way the material flows. It makes it flow easier into the part, and you're able to pack out the part. Because the material in thinner parts are going to basically freeze up much more quickly than the thicker regions.

So if you gate thin to thick, it's going to freeze before you're able to push it to the entire part or even be able to pack it out. So you could have surface defects. It won't look like a good part.

So we took what we had and brought it into PowerShape, and Jon designed this basic cooling system. And again, it's pretty machinable. You can basically add a couple of plugs, machine it, drill, what, four channels there, and be done with it. Within Moldflow, we actually analyzed it and saw there was a great amount of heat within that center area, what we predicted. So looking at that, we're thinking, how can we optimize this a little bit more?

So, in summary, looking at the ejection temperature. So the ejection temperature's essentially where it's taking that material property that you assigned to it and looking at how much time it takes for the part itself or regions within the part to reach the temperature where it solidifies. And with this part, it maxed out at about 248 seconds, so that's a long cycle time and most likely not going to be acceptable for a manufacturing situation where you want to create a lot of parts in a short amount of time.

So going back into PowerShape, like Jon had mentioned, we incorporated baffles into it. And they're a little tough to see here. But if you look at the inside of the part, you can-- let me just restart this here. There we go. You can see the baffles kind of protruding out. There's three of them in there-- one in the center core region and one on either side of the center. And with that, we were able to take it into Moldflow again.

And we did see a good increase in efficiency of the cooling system with pulling that heat out. You could see here, within the tool, you can see that cool spot. And each time it kind of cycles through there, where it's getting green and then going back to blue, that's representing heat from the injection. So it's injecting plastic. Heat's going into the tool, being green-- I'll just do that-- and then you're able to visually see it through a cross-section.

So looking at the decrease, rather, in ejection type temperature or time to reach ejection temperature, you can see here it's 134 seconds. So that kind of takes it into the range where it could be acceptable. Still a little bit high, but maybe we'll take another look at it.

And you can see with the arrow on the top right-hand corner of the screen there, it's pointing to that corner section, where it's really tough to pull heat out of there, because you're hitting the rib. You're hitting the wall thickness of the cylinder portion of it. So we took that, again, into PowerShape, and Jon will explain a little bit more about what we did with that.

JON HUNWICK: So this brings us into conformal cooling. So the first question to ask, who knows what conformal cooling is? You can keep quiet, Yan Willum, because I know you-- and Chris, yes. And you. Fine, fair enough. Well, the Wikipedia explanation is that conformal cooling channel is a cooling channel, a cooling passageway, that follows the shape or profile of the part to give you much more uniform cooling.

So what it does is to look a little bit like this. When I made this video, I suddenly realized that I've made something that looks a little bit like the game Lemmings. Anybody remember that? No, maybe not. OK, but what happens is the water comes in the bottom left-hand side and follows the channel all the way around the part and then comes out hot out the other end.

So the question is, how on Earth would you design something that looks like that? Well, there are actually lots and lots of different ways that you could potentially do it. I'm going to show you the way I did in PowerShape. And on the way, that will touch a few of the things that make PowerShape a little bit cranky in the traditional CAD world. Because there's some things that maybe take you by surprise if you don't know what PowerShape can do.

But basically, I divided the part into two separate sections. So we've got now an insert that we can use. Just view this in a shaded mode so we can see it. And then all I did was to take four or five surfaces from the top edge of that part. That's a single solid moment. Take those surfaces and basically just copy them out. I can blank that on the main bloke in there, because I don't need them at the moment. So this is just now a set of surfaces.

I don't want those little curvy gaps or holes at the bottom, so I can simply remove them by changing the trimming. This has got no base on it at the moment. This is completely open. But what I'm going to do is just to select the surfaces and say, from now on, as far as I'm concerned, that's a solid. I don't care if it's open or closed. Just make a solid from that.

And then I can take that solid and offset it inwards by seven millimeters, just to give me the shape that I want, the offset inwards. I don't like that sharp corner, so I'm going to put a radius on it. Just use the default radius of five. And now, we want to make that complex 3D helix that we're going to use. So we just take a simple helical curve. Just use the drag handlers to adjust the size. And in this case, this is really a bit too fine. You couldn't realistically print it in metal, which is something we're going to come to at the end, but I'm happy with that shape as it is.

So what I'm going to do now is just to draw a line. So now I've got a helix and the line. What I'm going to do next is to sweep that line around the helix to create a surface. This is fully automatic surfacing, literally one click. You just select the wireframe and PowerShape generates the surface.

So now I can select the surface and the solid, and I can find the curve that is the intersection of the two. So that's now generated for me a fairly complicated looking three-dimensional curve. I'm just going to rid of the bits of geometry that I don't need anymore. So there's that weird-looking curve. You can see it's got hundreds of points in it. I can smooth that out by taking some of the points out.

If it drifts out of tolerance during that process-- you can see some of those numbers up the side have gone red. That's just showing me that it's drifting out of tolerance, but I don't care, particularly, at this case. So I'm now going to make-- that's the cooling channel going up. I want one coming down. So I'm just going to take that part, rotate it 180 degrees, make a copy. And now, hidden away on another layer, I've got the basic flat portions of the cooling channel.

And all I'm going to do is join these four curves into one, just appending those four curves all together. So basically, just go around and select the bits of curves that you want to add.

[DISTANT APPLAUSE]

They liked it. Once you've got that, you can then make just a simple arc. In this case, it's going to be a radius two arc, and I'm just going to snap that into that bottom corner. And use that same trick I did before, make a simple surface, which is just a drive curve surface, again. You can see there's a little bit of a kink in some of the corners. I'm just going to use an advanced option to-- didn't it jump over that but? I missed a bit, I think. There are advanced options that you can go into take any kinks out.

So now I've got a solid, and I've got a very obviously open and rather bizarre-looking surface. So I'm going to select that surface and say, I just want you to take the surface away from the solid. So it's just doing a standard Boolean operation between an open surface and a closed solid. It takes it a second or two to do that. Once the little stopwatch at the bottom right-hand corner-- then I can take a dynamic section [INAUDIBLE] just to check that everything is OK.

And literally take that section, drive it through. And you see we've got this rather weird and wonderful looking helix actually built into that solid core. And it's got tapped holes defined in the bottom so that you can assemble the part. We can go a little bit further with this particular design and add the faces that we want to the machine. So if people wanted that outside profile machine, we could then just offset those a little bit to give ourselves a machine allowance so the panel has actually got something to cut. And that's basically how we do that. So what happens next?

KRISTEN KILROY: So, taking that model, I brought it into Moldflow. And essentially, with Moldflow, you can import the cooling system, since it's a CAD model, from PowerShape. We were able to easily import it into Moldflow, assign it the property channel 3D to basically represent that it's cooling channel elements rather than just general part elements. Because by default, it's always assigned as part 3D elements.

So after also importing the cavity section, which is just the plain basic beam elements that we used in the other two analyses, we can see here that, after meshing, we've got part and cooling system mesh. Now we have to go into creating the actual tool behind it. But before then, shown here, you also have to represent which inlet and outlet Moldflow is going to use when it's calculating that current going through the part channels there.

So with that set, we create a mold boundary around it. And again, it looks really large. But within Moldflow, you can see it's just encasing the entire cooling system and part in there. But when we mesh it, it'll actually take the cavities out. So, basically, you have a hollowed-out tool in there. After assigning the injection locations, the same ones we used at the other two analyses, we can go in again. Select the same material again for comparison.

And you can see here, you can also just easily search it by manufacturer or any other properties, as well. And again, showing the vast material database that we have. All those are materials that were provided by Total there. So we found the one we used on the other two. And looking at the material data, as I mentioned, it is tested. So anything that's not red highlighted is essentially measured data that we've collected, or else the manufacturer has collected.

Once we are OK with that, go ahead and close out. And next step is process settings. So you can see, it's a little bit more advanced than Moldflow Advisers. You have more advanced solver options, as well. So in here, you can create the gate diameter. Because we didn't create the feed system within Moldflow Insight Analysis here. So we wanted to be sure it's set up the same as Adviser. And you can see the other options available.

For this run, we did transient from production startup. So basically, it's going to simulate real life. So you loaded the mold. You're ready to run it. And if the mold is probably a little bit cooler, but throughout the simulation, you're actually going to heat it up. And this is going to actually simulate that, rather than doing a more ideal temperature standard temperatures within it.

And then, going in, you can select the deflection, if you're running a deflection analysis. In this case, we did a cool fill pack warp analysis. And here, we just ran locally, but like I mentioned, this is where you would choose between local or cloud solving. Now, this particular analysis because those channels were so fine, the mesh was very small in there. And it did take a little bit longer to run this analysis rather than the other two. However, looking at the results, we can see the similar trends that the other two had in there.

And then, once we go into the tool here, you can see through each cycle. Again, looking at the green as more hot, where the material is flowing into the cavity. And then the blue there is the cooling lines efficiency. So it's really pulling that heat out. I think that's the end of it. There we go.

So looking into the results of that, that brought the time to ejection temperature down to 124.4 seconds. Much, much better than the 200, what, 238 that it was.

JON HUNWICK: Yeah, six minutes down to two.

KRISTEN KILROY: Yeah. So looking at that, it's really convincing. It's eye-opening. But with this, I was still a little curious about it, so I actually compared all three of those results again and queried the same exact spot within a section of the part. So I just took a section kind of where the rib met the cylinder part of the part.

And we can see here that the basic cooling on the left-hand side here had a time to ejection temperature at about 185. So not quite that 248 that we have seen before, but this one is consistent with the spots we measured on the other two, as well. So looking at the more complex cooling, again, that's the one with the baffles. And we can see it's down to 100 seconds, so 85 second savings.

Now, bringing in the conformal study we ran, you can see it dropped all the way down to 75 seconds. So that's a drastic difference from the basic cooling. But with that, that's where it comes in where if you're the tool designer, does it, excuse me, quantify the cost for creating the conformal cooling versus just typical machining? Is that 25 seconds between the complex machinable cooling worth the conformal cooling cost? And that's where Jon's going to pick it up here.

JON HUNWICK: Now, there are some problems with the design itself in that the cooling channels that I've used are actually way too thin. But I just wanted to give you an idea of the sort of complexity that you can achieve quite simply. Now, if you were to really try to print one of these, you'd have a bit of a job getting the powder material out. But I've just taken that part and 3D printed it so that you can have a quick look and see the sort of complexity that you can achieve with this process.

If you were using it in real life, you would probably make the channels considerably thicker, which would improve the flow of water or coolant through the tool anyway. Oh, you were looking at querying it?

AUDIENCE: I am querying it.

JON HUNWICK: You are querying it. Fine. Fair enough.

KRISTEN KILROY: Well, the other part is it could-- water is not always as clean as you would expect it to be. So in a cooling system, it's bound to build up with calcium deposit or whatever is in the water or minerals. So a channel this thin wouldn't be practical with regards to running typical out-of-the-hose water onto it. But it was a good example for this particular class here.

JON HUNWICK: So, basically, if you want to make something like this, until somebody can invent a drill that can drill around corners, pretty much the only way that you're going to be able to make it is by printing it in metal. And that's becoming much more common these days. We're getting far and far more into industrial additive. You can print much faster than you used to be able to. It's still, right now, limited to the size of the parts that you can print.

So if somebody was to want to make a tool for a car bumper, you'd have to make multiple different inserts and join it together somehow in order to make it work correctly. But for smallish parts, this is feasible technology. With Netfabb as part of the Autodesk arsenal, you've got great ways of controlling 3D printing machines.

KRISTEN KILROY: So, to wrap things up, hopefully, today, we gave you a little bit more detail behind using Moldflow and PowerShape together. If not, please feel free to come to us and ask questions. But essentially, today, we went through how to ensure that the mold will work correctly. So making sure that cooling system is set to do what it's supposed to do before you cut the steel. Using that simulation to create options for both yourself and your customers or whoever, the OEMs that are concerned about cost as well as the quality qualification of the simulation.

And then, we've also discussed the different design options to choose the better mold possible. Again, mentioning is it really worth putting that extra cost into the conformal cooling mold rather than a machinable mold? And then, also, reduced lead time. So as we described earlier, instead of having to go back and forth between the tool shop and the manufacturer, this could potentially save that step in the process.

And again, costly rework. Kind of explains itself. With simulation, you're able to have that prediction and get that better understanding for that what-if scenario. And with that, we'll open it up to questions. Anyone have any for us today? Yeah?

AUDIENCE: [INAUDIBLE] around result time. What, in your opinion, is the thing that stops it being adopted?

KRISTEN KILROY: Do you want to take it?

JON HUNWICK: Probably the thing that stops it being adopted so far has been the cost of 3D printing. People don't see-- I mean, up to now, it's been a very, very expensive process. It's still very expensive, but it's no longer very, very expensive. And the price is going to be coming down over the next few years. The machines are getting faster. They're getting much more accurate. I think we are very rapidly approaching some kind of tipping point, where it's actually going to become feasible.

With any new technology, there's kind of a fear factor. There's that kind of valley of despair that Andrew was talking about this morning. Am I going to do this? Am I going to be the first person who jumps at this?

But once it reaches a point where it is economically viable and people are able to see a real tangible benefit and say, I can make parts 25% quicker than he can, because I can do this, then you're going to see people taking it-- Yan Willum is probably-- the guy sitting right behind him is a great guy to talk. Because he's been working with-- and Chris over here, as well. Working with conform cooling for 10 years, probably. Something in that order. Doing sort of research projects based around it and slowly increasing the feasibility of the technology.

KRISTEN KILROY: That's where the simulation comes in, also, introducing it to people who aren't really sure what conformal cooling is. This shows it to them visually. They can see that heat exiting the center of the tool. You're able to really justify that more costly option. Sure?

AUDIENCE: [INAUDIBLE]

KRISTEN KILROY: Exactly.

AUDIENCE: Certainly, the customers [INAUDIBLE].

KRISTEN KILROY: Very good.

AUDIENCE: I was sorry to see [INAUDIBLE] that sometimes you cheat for some reason from [INAUDIBLE]. So this kind of system will give you better [INAUDIBLE].

KRISTEN KILROY: Right.

AUDIENCE: It's expensive, but it solved another problem.

KRISTEN KILROY: Yeah, that's another good point that we didn't really touch base on this class in the warpage effect. A lot of people are worried about meeting that tolerance, getting that part completely flat, making sure it meets that spec. If you need to add that extra cost of having a conformal insert in there, then it could be worth it, in that respect, going back to, what, the second slide we had where what's the most expensive tool. Having a tool that produces a warped part is not going to be acceptable at all.

So having the simulation to be able to see it visually, predict the warpage behind it, incorporating that conformal cooling aspect, then it's eye-opening to some parts. All right. Any others? OK. Well, thank you all. And one final point is please don't forget to rate our class if you enjoyed it. Hopefully, you did. You can go to your Autodesk app. Or if you're online on your computer, please feel free to give us feedback. We're always open to it. If you have any other questions, we'll be around the rest of the week. Thank you.

[APPLAUSE]