Micro-Jet Engines With an Uninterrupted Print Process

00:00 Thank you so much for being here in this session. The title is only with additive micro jet engines within with an uninterrupted print process. So this will be a Co presentation. First of all, I would like to talk about the co developers of this jet engine is a co development between Technion and PTC from the Technion. Benny will be presenting the specifics of the design from the PTC side.

00:33 The co developer is Ronen Ben-horin, one of our colleagues at PTC. In my case, I have the honor to present the PTC portion in this presentation. I'm the product manager of the additive manufacturing functionality in Creo. So we have a very packed agenda. But just to start, I would like to say a few words about the Technion and PTC partnership. So in April 2021,

01:03 PTC entered into a long term partnership and long term strategic collaboration with the Technion in Israel. So part of it's a very it's, I will say, a very broad type of partnership. Even there are PTC offices inside the Technion in order to improve collaboration, etcetera. So there is a multi-million dollar budget allocated for this partnership. And the project that you will see today is one of such research activities.

01:39 So without anything else. Benny, thank you so much for being here. So maybe I'll say a brief few words about my background. I'm originally from Turkey. I got all my degrees here at Purdue University. My bachelor's, master's and PhD Go boilers. So after that, I spent around four years in Von Karman Institute for Fluid Dynamics in Belgium. So after that, in 2012, I had an opportunity to to be a faculty member in Technion.

02:11 And as a kid growing up in Turkey, it was my dream to be a faculty member there. So be careful what you wish for. It might actually come true. It's one of those stories. And since 2012, I've been the head of the Technion's Turbomachinery and Heat Transfer Laboratory. We conduct research and development in micro gas turbines. And on a lower technology readiness level, we deal with issues associated with basic and applied heat transfer and measurement technique development, which supplements our efforts in gas turbine activities.

02:44 And most of the applications of this research goes into propulsion and power generation of business jets, UAVs and drones. And these are beautiful building that's actually on the graffiti now. I'm really proud of that. If you if you're not impressed with the wonderful work that we are doing, this has got to impress you, right? So it's an amazing, amazing, talented artist called Solomon Sosa that did the graffiti for us. So in this building, we are around 30 research scientists from all backgrounds,

03:17 and we are running in parallel right around 12 projects. But recently we've been on a journey of additive. And you are quite familiar with the subject because everyone uses this term very vaguely. And the way that people are approaching it is that they're saying, okay, I was doing this part before by CXC and whatever, and now I'm going to do small batch production. So I'm now going to use additive manufacturing in order to replace that or combine a few parts. We don't take this approach. The approach that we are taking is how to actually use additive as

03:50 a research direction in itself. So that means from the get go, consider the constraints and the advantages of additive manufacturing into our design space. And this journey started right around. Six, seven years ago with a project that's on miniaturized turbines. And I'll try to take you on this journey of where we got to today. The very first one is miniaturized turbines. And here we made a proof of concept demonstrator for monolithic rotors.

04:22 That can be printed in one go from bottom up. This was actually one of the first activities as such for the Turbomachinery community, and we worked with porous media combustors that can be printed. And then from this we transformed our design space by removing the assembly requirements. So we started printing things of both the rotating and the stationary parts inside one another in one go. And this was applied to some emergency ventilator.

04:52 And I'll talk a little bit about that. And here we show the proof of concept of error driven hydrostatic bearings and pre-assembled design and support free design. And after that. We transition this. Into how to do this from metal. So not actually just toy parts that are made from plastic, but to be able to print it from Inconel 718. And this is what your. This is what we've ended up here today. The preassembled additively manufactured additively manufactured engine from metal

05:26 with fuel driven hydrostatic bearings. So now let's go to the beginning of the journey. The beginning of the journey started with these micro gas turbines that are two inch in scale. This is the actual scale of the generator is the world's smallest generator. And you know, the drones, they fly for about half an hour and then they can't really continue anymore. It's because the energy density of the batteries is really poor. Right. But we've already solved this problem many decades ago.

05:56 That's why all the large aircraft, they fly with fuel. So what if we had a mechanism of being able to capture the energy in the chemical bonds of the fuel into electricity in the smallest of scales and apply it to the drones? And that was the motivation of this project. So we started with a design architecture that's very unique. That's a monolithic router that contains maybe I'll say a few words before that Every gas turbine engine for propulsion and propulsion

06:27 or power generation contains three parts. You have a compressor, combustor and turbine. That's what completes your Brayton Cycle. So in this particular project, what we started with is combining all the rotating parts into one monolithic piece. And that's what you're seeing here. So we have a turbine that's at the bottom. And then a compressor and a shaft all manufactured in one go with internal cooling cavities and to be able to make it compact.

06:59 This whole rotor assembly spins at around 500,000 rpm, so it's right around 10,000 revolutions per second. So it's kind of easy to be able to. To be able to scale down the kinematic scales to get the velocities right. But it's very difficult to be able to actually get the combustion in a very compact form and to be able to do that. We started investigating a revolutionary technology that's called porous media

07:31 combustors. We are not the very first ones to deal with porous media combustors, but we are the first ones to apply it into the design space of a gas turbine. And by this way, we can actually have an extremely compact engine despite the fact that basically it's a miniaturized engine, despite the fact that conventional combustors would take basically the size of. I mean. A small, relatively much larger engine. So after that.

08:01 After proving that the rotor can spin and everything works. The Covid hit, it was right around 2019. And as you remember, there was a there was a shortage of ventilators. So everyone thought that the solution to the problem is just finding more ventilators. And we thought the same. And we tried to do our part. And. At the same time, you know, if the only tool that you've got is a hammer, every problem is a nail. We're turbocharger people, right? So we want to actually solve the ventilator problem in the turbocharger way.

08:33 So we designed a unique pre-assembled ventilator. That's actually plastic printed in turbocharger form. The idea was to use the oxygen supply of the hospital in order to be able to rotate a blower that sucks in the mixture of oxygen together with air into the lungs of the patient. And this is how this project came about. And the thing that we learned through this process is the key to making things pre-assembled is to make them support free in the design space.

09:07 And that's how we actually designed these ventilators. And this is the real part. This thing rotates at 40,000 RPM despite the fact that it's made from plastic. Supported by hydrostatic bearings because we cannot print bearings. And it actually supplies enough air not for one patient, but actually two patients at any given time. Look at the size of this thing. It's incredible, right? So here we learned two things.

09:37 One of them is how to design support free and how to be able to use the hydrostatic bearings in order to support a rotating structure inside a stationary structure. And the basic principles of that is exactly like an air hockey table. In the air hockey table, the puck is floating. So we're making this rotor float in the middle by fluidic pressure. And here you can see the print process of these and you can see the rotating

10:08 and stationary parts being built within one another without any internal supports. So then we try to combine all these technologies into the engine space. So we had already dealt with self supported additive manufacturing and hydrostatic bearings a little bit and porous media combustor and monolithic turbomachinery. So now what are the additional things that are missing in an engine? So you need to think about rotor dynamics and vibrations.

10:41 And heat transfer and the thermodynamic cycle itself. So typically, most engines are designed for the best aerodynamic efficiency of each component. But you cannot do that for small engines. So what we did is we built reduced order models for each one of these disciplines and optimize the result intuitively, inherently with the additive support, free additive manufacturing constraints.

11:11 So that was actually one of the major blocks within the design space. And the outcome of this is this funky looking jet engine that has all the same parts in terms of functionality, but the topology looks completely different. So this engine has the least number of parts possible on any engine because it has two parts. You have one rotating part and one stationary part. That's it. So the rotating part comprises of a turbine.

11:45 Some shell, basically a shaft, right? And then a compressor. And each layer of this rotating part is self supported. So each layer is inherently designed such that the lower layers support the upper layers. Similarly, on the stationary part, we have a heat exchanger on top of it. We have nozzle guide vanes, a combustor and bearings, and on top of it the diffuser space. And all of this is also self-supported.

12:17 So in reality, what we are doing is we are printing two separate parts, but just one happens to be inside the other. Not being aware of the other one at all. So that's why we didn't have to use any supports. So now this looks a little bit complicated. And I'll try to make it easy. Now let's follow the gas path, because we know how a jet engine works. I said there's a compressor combustor and a turbine. That's a jet engine. Right. So now let's follow the gas path. So the air comes in from the top through the compression process.

12:51 So here we actually mix a bit of fuel into the air. It goes through the diffusers here. So now we have static pressure rise. And then it gets diverted downwards and goes through a recuperator to pick up some excess heat from the exhaust comes into the combustion chamber and from the combustion chamber it goes through this nozzle guide vanes gets accelerated. In order to be able to pick up some mechanical shaft power on the turbine

13:23 and the remaining energy of the kinetic energy of the exhaust gases it goes through in this area from the nozzle and the remainder goes through the recuperator itself. So it's just like any other gas turbine. Just the architecture is funky. So the intended. So this thing this is the real size. So this is actually one of the questions that I get a lot is this real size. This is the real size. The main reason why it is in this scale is because of

13:54 the build volume available in today's 3D printers. So this was printed on EOS, M 300 machine, which means you have only 30cm as build volume in any given direction. And that's why this is the size of it. So this shaft is intended to spin at around 50 k rpm, supported by the same trick that we've done before. The hydrostatic bearings in this time powered by fuel rather than air,

14:26 and the residual fuel gets pumped back into the gas path to be burned in the combustor. And the intended thrust of this engine is right around 700 Newtons. So now let's take a look at what are the design considerations we had in this process. So we have a self supported rotor and a self-supported housing. And in the self-supported rotor. What you can see here is that the turbine looks a little funky.

14:56 So this is not how a normal turbine would look like, but the entire exhaust section of the turbine here is designed such that it can support the other parts of the turbine. Because if I don't grow the radius from the lower parts of it, I cannot build a larger radius. Largest radius part, right. So that's why we actually are we have an extended user here. We have overhang angles that are constrained to 35 degrees in every direction.

15:26 So in the three dimensional space. And if you look at the stationary part here, you see the same tricks. And this time we have self-supported parts. Basically in the nozzle guide vanes and in the diffusers such that each layer doesn't really collapse on one another. So if we were if we were to actually build pipe diffusers as circular pipes, they would collapse. Right. That's why you see these strange shapes that are diamond shaped, that are self-supported or teardrop shaped, that are self-supported,

15:56 that don't really require any additional struts. And at the end you actually can see the zig zag floor that we built here in order to be able to remove it from the base plate at relative ease. So we've done all this work. What is the impact? So today. In order to manufacture a gas turbine engine. The processes that take place are casting of the turbine, typically from nickel alloys.

16:28 Inconel. You do CNC milling of the compressor typically from titanium. You take some steel shafts and you have to do grinding, you have bearings, you have sheet metal work around the combustor and you have to do balancing and assembly and all that, right? And there are roughly around 200 parts that go into making an engine. The minimum part count that people talk about for small engines are in the range of 20 to 30 parts. So now we've taken that and brought it down to the true minimum global minimum,

17:01 which is two parts. And the main advantage of that is that it only requires a mental printer. We designed the whole thing from Inconel in order to actually consider the stresses that you have in the turbine and . Printed everything else around that. And there's no supply chain, so you don't have to worry about all these suppliers that are doing all this work for you. And now the intended application is that you would manufacture this at the platform. Producer You don't need engine companies, you need engine design houses.

17:33 But then the platform manufacturer itself would be able to make its own engines on demand, basically. And this is actually a snapshot of the print process that you're actually seeing out of the printer. And you can see each one of the layers that are being built on top of the previous layer without any support. And of course, we edited out the recorder passing back and forth in order for you to be able to visualize the laser melt pool as is going through in different layers.

18:05 So through this process, the fringe benefit is that the engine costs would be brought down to just the cost of the depreciation of the printer and the raw material. So with that, I'll give the stage to Jose. So now that we review what Ronen Ben-horin and Benny Cucurella did as a design for this jet engine, given the opportunity to talk about Creo,

18:38 which is my which is my domain, let's let's talk a little bit about Creo, which is the technology behind this design, you know, portfolio Creo Manufacturing Solutions portfolio covers, machining inspection, mold design, etcetera. In this case, we are using only Creo Parametric and the Creo additive manufacturing extensions. Everything else is is done. Everything is done using only this basic Creo

19:11 functionality and additive. That's it. Now, there is also a couple of things that I would like to to mention that. Or Creo manufacturing solutions are based on making functionality too lightweight. The designs of customers to make self-supporting geometries and to enable interoperability. In this case, the interoperability. It was done with the EOS EOS printer to do this job.

19:48 So basically the example here is having Creo additive manufacturing in a real project. I would say it's very difficult for us as product managers as just a reflection sharing, a reflection to showcase something that is a, I would say real use case because any of our Creo users that are doing very important things, very breakthrough advancements, obviously there are IP, IP considerations.

20:22 Okay. Now. Lessons learned. Very important. This is a research project in in any in any aspect. So my role as a project manager is to be close to these kind of research projects in order to make Creo better for for you or users. So. What are some reflections, some of the things that we learn of this project and that will drive the the roadmap in

20:55 the future additive manufacturing functionalities. First of all, in the first pillar that is lightweight designs, we want to work on intelligent transitions. So here we were experimenting with different lattices in order to define the combustion chamber. So one of the things that was very evident is it will be much better if instead of creating one

21:27 combustion chamber with only one lattice, we should have different lattices that can stitch together. So this is this is an area that we started working already on that this intelligent transition. Transitioning between one cell type to another cell type in in a in a very smooth way. The second on lightweight design is what about defining the pore size as a parameter that is driving the lattice geometry?

22:00 So instead of defining a lattice from the geometric perspective, like size of the beam size of the cell, how would you reinforce it? Okay. And then test it. Finite element analysis or fluid analysis, etcetera. And. Okay. No adjusted. No. Give me the porosity at 3D porosity map and then adjust the lattice to comply with this porosity map. So this is kind of the a some learning aspects that we will include in

22:34 the roadmap based on this jet engine. For sure. We need to improve how we interact with point cloud systems. So we need to make it a real end use product. Then we need to qualify it. Then we need to deal with all the information coming from the CT scans. So this is one area of work as well. For sure continue working on self-supporting geometries. Now we are talking about printability modifiers.

23:07 So there are several lessons learned in this regard that we will work on. Printability modifier we call modifiers which you define anything and then say, okay, you know what? There is some problematic area here, then it will give you a mechanism to adjust the geometry to make it self-supported. So the goal at the end is no supports. So and in the interoperability space for sure about the connectivity with key printer manufacturers like EOS,

23:39 which which by the way, it's a fantastic partner of us. EOS printed this jet engine and we are working on the second iteration, or maybe it will be in the third and we will be really firing this jet engine up and putting into a UAV, etcetera. So the connectivity piece is very important. So making the. Getting all the designs into the printer in an efficient way is for sure very important. We will continue working on that.

24:13 And all of this fully embedded into Creo having the same workflow for you. Just more advanced intelligent features, etcetera. Just the last remark here. So this example of this research project, we proved that we can produce today self-supporting geometries, that our lattice modeling is mature enough to make it happen. Right.

24:47 And the build preparation, it's also in in good, in good shape for us to do real world projects is not anymore in early stages or solutions are mature. This is a proof of concept. I think we all say make it happen. We made it happen.