0:01 Okay, so first of all, I’d like to thank Duann for organizing this fantastic event. It’s been a a great couple of days for both Lawrence and and myself. Lots of exciting presentations today. I really enjoyed actually the presentations from our colleagues in architecture. I think there’s a lot to learn from these guys, they’ve really nailed it when it comes to datadriven design. So my name is Manolis, and I’m the co-founder, together with Laurence Cook, at the back, of Metamorphic. We are a design engineering consultancy that specializes in AD manufacturing, and what we do, we help organizations navigate the complexity of ID manufacturing and make the most of the design freedom that it has to to offer. And we combine expertise in design for AM, which I approach from a computational design perspective, while Lawrence kind of grounds these crazy designs with multi physics simulation and more traditional engineering design.

1:11 So we specialize in early stage R&D projects. We operate as a flexible R&D resource that suits our clients or partners needs, and we typically help evolve or develop ideas that might be coming from fundamental research into rough working prototypes that can be tested in their intended environment or lab setting. So we bridge, in a way, we help kind of organizations cross this so-called Valley of Death, where ideas are tested and sometimes they fail. So yeah, that’s us. We founded Metamorphic about 2 and a half years ago, after having worked in the AM industry in for a number of years, and the aim, the mission, the motivation was to, is to elevate DfAM and make the, and help also in the adoption of ative manufacturing in in a wide range of industries, because there’s been fantastic developments in terms of materials and process in AM over the last few decades, but in the end what matters is what you actually make with this technology.

21:15 So we have a growing project portfolio in a range of industry sectors. We have focused primarily on emerging technology sectors such as nuclear fusion and and quantum sensing, because they they face major technical challenges that usually require multidisciplinary teams to come together and and solve these challenges so that the technology can become commercially viable. However, we are equally pursuing opportunities in established areas where there’s usually a drive to improve efficiencies. So I’ll give you a few, a few examples here. In the area of nuclear fusion, we’re working together with, we have this ongoing project with the University of Birmingham, in Tokomak Energy, we’re looking into new design and manufacturing methods for pure tungsten components. We have a s, a sample actually at the back, these can withstand extremely high heat fluxes.

In quantum sensing, this is the project Q team that I’ll be talking about today, in telecoms, together with U, a long list of partners, so BT, the University of Birmingham, M Squ, Helia Photonics, and 3T, we’ve been developing additively manufactured lightweight optical cavities that can be used as frequency reference es for timing, and that has loads of applications in telecoms, in 5G and beyond, in wearables, and kind of, this is touching also on medical devices. Medical imaging, we have designed this magneto craphy helmet for Circum Magnetics that holds quantum sensors in known locations and measures very small changes in the magnetic field caused by current that flows in different parts of the brain as they become activated. Now typical ME systems are huge, they weigh, they’re the size of an MRI, they weigh over a ton, and what CA have done, they managed to bring, bring all this functionality into a helmet, and that allows scientists to scan subjects while they move, and this also allows to scan children, children of course, who tend to move a lot, down to the age of two years old.

In chemical and and engineering and bioprocessing, we’re actively pursuing opportunities to innovate in this area with various partners. We believe there’s a huge potential for artive manufacturing in this field. So if you have any ideas in this space, or any other, please come and talk to us.

So today I’ll talk to you about project Q team, which is a collaboration between us, Ral Space, which is the UK’s space lab, the University of Nottingham, specifically the school of physics and astronomy, and A T Scientific company that specializes in ultra high vacuum components and also instrumentation components, based in the UK. So the the the project was partially funded by Innovate UK, which is the UK’s innovation agency, and the the aim was to provide alternative methods for designing and manufacturing components for the quantum sector that can improve the SWAP characteristics of quantum devices, that is their size, weight, and power consumption, and contribute this way towards this transition into more portable devices that can be used out in the field or in space.

So our focus, as Mr Morphic, was on the ultra high vacuum chambers that are at the heart of these systems. This is where all the magic takes place, in this case, the, we’ve, we designed this chamber for a quantum gravimeter that measures gravity, and the term ultra high vacuum refers to vacuums below 10us 6 Pascal, and to help you kind of visualize it, that means that you, you know, you’ve basically extracted most of the atoms inside this chamber, and maybe the few that remain, the possibility of them colliding with each other is so small that they would have to travel several kilometers before that happens. That’s how low the possibility is.

So one of the main objectives was to prove that AM is able to deliver bespoke, a highly functional components for these really stringent, that meet these really stringent requirements of the quant industry, and also this way kind of establish, help establish, a UK based additive manufacturing supply chain for the emerging quantum sector. So ultra high vacuum chambers are typically machined out of billet materials, and this poses limitations, as you can imagine, to their size and weight. Now, in addition to the waste that is generated through the machining process, there is also a lot of excess material, as you can see in these chambers, this is actually, so we based our design, Q team is based on this quantum gravimeter that was developed for the European Space Agency. It was a collaborative project between Ral Space, the University of Hanover, the University of Birmingham, and DLR, and you can see that there’s a lot of excess material in these components. It doesn’t serve any particular function, it’s just the result of the manufacturing process itself. We are surrounded by objects that have been shaped one way or another by the manufacturing process, while kind of trying to balance aesthetics, function, performance, usability, and cost of course.

In this case, however, these ultra high vacuum chambers, they they haven’t changed for more than three decades, and the reason for this is because they’ve been mainly used in labs. So the focus has been on proving the physics, but as they find applications out in the real world, there is this drive to actually reduce their weight and miniaturize them. Again, this is another thing that is linked to the conventional, to these limitations imposed by conventional manufacturing techniques. You can see that they consist of multiple parts, and this increases the the risk of leakage, because of multiple, the presence of multiple joints. But there are also limitations imposed by conventional CAD tools. So these are highly complex systems, they are really challeng, alling to design. Determining the optimum diameters and locations of the ports usually involves going through several iterations, they, the physicists have to run simulations for the optics, the lasers, the magnetic fields, and they could really benefit, so you, if you make a small change, basically, in one of the the ports, it can impact the entire assembly. So they can really benefit from tools that are more responsive, more dynamic, tools that allow them to easily visualize all these different scenarios and help them this way trade faster, and this is where a manufacturing, computational design come into play.

So Q team was really born out of this question, what would quantum devices look like if physicists were given more design freedom, and and what are the opportunities for innovation in this space using ative manufacturing, can we design these systems so that we get the best performance of the physics package out of it. So an important part of every multidisciplinary project is the knowledge exchange that happens at the start, this is where, when requirements are gathered, and when the team kind of tries to kind of identify areas that can benefit from design for AM. So within a short period of time, with the help of our partners, we had to really build an understanding of called atom interferometry, and also magnetooptical trapping, that basically means, means using a combination of lasers and magnetic fields that have specific shapes and gradients to trap and slow down atoms so that these can be observed. And also we went through the typical process that they follow when they designed these complex systems, and how they assemble them as well.

And at the same time, we demonstrated examples of previous work. Also, we went through some of the most common kind of techniques used in computational design, like field driven design, form finding, etc, to get them thinking really about the possibilities. And also we covered aspects of meto AM, including considerations for postprocessing and machining. And and it’s through this fusion really of these two very disparate fields that you get lots of ideas bouncing around, so you might, physicists asking, you know, we had this idea a while ago about combining these two things and maybe wrapping them around this surface, do you think this might be possible, and of course you know, sometimes the answer is no, because AM has its limitations, but there are cases when, where it is possible, so long as you know some compromises are made, for example, you know sometimes they have to relax some of the requirements. So, and this is when things I think get really interesting for us.

So in this project, are go into the technical details a bit. So we we followed a hybrid approach, we have a computational design script developed in Grasshopper for Rhino, and this script generates the chambers. It has access to a library of parts, mechanical interfaces, ports, that have been designed using trad additional C tools. So the user starts from defining, as you can see on the top left, from defining the optical paths and the corresponding port diameters, and the script calculates the optimal packing density of these ports, so the best arrangement of these ports, so that they don’t, they don’t intersect with each other, while at the same time ensuring that we have the minimum volume in of the resulting chamber.

So this is done through solving intersection events between cylinders, that that whose diameters correspond to the external diameters of the ports, and then the bullan union, based on these calculations, the bullan union of these cylinders is trimmed, and this, the resulting geometry can be, can be U, the mesh can be reparameterized, and through constrained mesh relaxation you can get a nice kind of external surface. Similar process followed for the, we follow for the the internal surface, in this case however the cylinders correspond to the actual ports, the the diameters of the ports, and this can be also kind of combined with a minimal surface to kind of minimize the internal, the surface area of the internal cavity.

So the first step was really to build a robust algorithm that allows the team, allowed the team, to explore hundreds of design variations with very minimal input, so the algorith them generates a nice watertight geometry of an ultra high vacuum chamber, and that corresponds, also gives us the minimum volume for any given combination of ports. And once you have that, you can actually have an assembly of chambers here, defined as nodes as you can see on the top left, and you follow, you can, the user can specify the relative orientation and position of of these chambers in relation to each other, and following the same approach you can actually get a nice consolidated assembly that can be used for the the operations that that follow in this workflow.

So at the same time we lightweighted a lot of these interfaces, so you can see here, I don’t know if you can see, my, no you can’t, so you can see one of the first iterations of the the chamber where basically we have all the ports populated around it, and the script essentially picks and places each port in the the defined location, and this is, joined, these ports are joined with a two mm thick skin. A lot of these ports eventually could be replaced with brazed windows, and this could also result in further reductions in in weight. However, scientists, they want to be able to assemble and disassemble the system, they want to be able to mount different different peripherals, different components, detectors and things onto these ports, so we we we we stuck to this kind of approach.

And on the far right you can see a cross-section of the the chamber, and you will notice there is this narrow tube connecting the upper section, which is the 2D magnetooptical trap, with a bottom section, which is the 3D magneto optical trap, and that previously these two sections had to be manufactured separately, because there’s no way of accessing this middle section with a drill bit. However, with manufacturing this can be built as one monolithic object, and that also helps reduce the the possibility of of leaks.

So so so now this slide shows you the workflow that we follow for generating the the external ltis, which serves a dual purpose, so it provides mechanical strength by connecting all these different ports, and at the same time it’s also the support structure used for printing. So after after the individual sections have been, of the chamber have been merged and smooth, the mesh that corresponds to the external surface is reparameterized, and the nodes of this mesh are then used for generating a network of lines, which is then filtered based on, so the user defines the printing direction, and any lines that fall between 0 and 45° are removed, and from there we have, also the script also identifies all the overhanging features, so we have all the ports basically populated, and without the feature that are machined, and by defining again the printing direction we can, the script identifies all the overhanging features and draws lines, draws members between these features and the ribs, as you can see these are the lines from the the original network extruded inwards, draws, so it draws members between these overhanging features and the closest ribs, and the user can specify the thickness of these members, and also the thickness of the ribs, which can also be varied locally, and through a series of Boolean operations you get a nice UHV chamber geometry.

22:16 But we’re not done yet, so these cold atom experiments, to be able to work, they need, they require magnetic fields with specific shapes and gradients, this is for, to essentially, to be able to trap the atoms effectively, and Naum, Uni, what they did, they went through a series of optimizations and they managed to determine the optimal size and location of these coils in relation to the atom cloud in the center of the chamber, and also the number of turns. So we had to basically trim the lattice so that we can fit these these coils, and by bringing them as close as possible to the chamber, the nice thing about it is that you can reduce power consumption in these systems.

So finally, the part that I’d say most people avoid talking about is machining, because most additively manufactured parts, they need to be machined in order to, I guess, meet the the manufacturing tolerances for assembly. So our script generates an STL, and pleas, one, don’t get too mad at us, that corresponds to, so that that’s essentially, on the far left, the part that we sent for printing, then another STL which corresponds to the as machin, and by overlapping them you can actually see which features are going to be machined away, and this is very helpful for actually planning the machining operations. However, we still need to produce technical drawings for the workshop, and for this reason we also output a BP, which is a a lightweight, let’s say, a a light representation of the chamber without the external lce structure, and this is a very time, like creating these drawings is a very time consuming process. So if anyone has solutions in this space, we’d love to hear about it.

So the verdict, we achieved, in compar to the original system designed for the European Space Agency, we achieved U approximately 50% mass reduction, and in our opinion, a design that better suits the next generation of ultra high vacuum, of well, of a portable quantum systems. The machines were printed in titanium 64 using laser padle bed fusion, you can imagine the machining was not easy, the chamber had, I think, almost 120 bolt holes. The print bureau, which is a well established bureau in the UK, they’ve got like more than 30 years of of experience, they said that this is one of the most complicated parts they’ve ever had to manufacture, they had to plan the the sequence of machining operations very carefully, and the chamber is currently being tested at Rous Space.

26:17 So early results suggest a range of, so a helium leak rate in the range of 10- 10 m l/s, which is in in line basically with most ultra high vacuum components. However, more testing needs to be done in the months that follow, the system basically needs to be baked down so that any trapped atoms are removed, and then it needs to be pumped down, which can take a number of weeks to reach a safe conclusion about its performance.

27:00 So I think Q team demonstrates very effectively how computational design combined with our manufacturing can really help teams approach technical challenges differently, and this unlocks opportunities for innovation in a whole range of both established and emerging technology sectors. Project Q team was just a start, it laid the foundations for the design of and manufacturing of lightweight, volumetrically optimized, ultra high vacuum chambers. Since, so the project finished a few months ago, since then we’ve managed to secure more funding to continue the development, and also make our computational design scripts more intelligent. And finally, we applied to TCT Awards with this design under the Aerospace and Defense category, and I’m delighted to share with everyone that we’ve been shortlisted alongside Boeing, Airbus, and Mark Forged.

28:13 So yeah, that’s all for me, please don’t hesitate to get in touch if you want to find out more about us, how we work, if you want to work with us. Thank you very much, everyone.

28:36 This is, okay, I’m going to give you the the CDFAM award here. Hey, this is the CDFAM Award for Best, Best Design. Thank you. Thank you. Right, oh yeah.