From Computational Design to Digital Fabrication using the Tissue add-on

hi hi everyone I'm uh Alexander zarelli I'm a computational designer and I'm also the developer of a nadon that is called tissue um I started to experiment with digital tools generative design procedural modeling you can call it how you want in a the field of experimental architecture and then using 3D printing technology I started to move to a smaller scale working on wearable devices and then some product design like medical devices and other type of product and I'm also involved in different teaching programs where I teach the tools that I use in my work so around 10 years ago I would say I started to collect some scripts some tools that I was using in my daily practice and I started to collect them in an Adon that I called tissue and this is something that was helping me in my work and I decided to share with a community because maybe someone else needed for his work and for me it was also a way to try to advertise blender that I was loving and I was thinking was a great tool I still think that um to the people that was in my field so for other computational designer architect it was just a way for me to convince them with a tool that they actually needed inside a software that was amazing already um but uh tissue in my opinion is something that can be useful for generic modeling purposes so what I would say is that if you're a lazy modeler maybe when you have to approach something like a Sci-Fi helmet you could approach it with a small component you define the features that you want in your final object then you define a topology with those different materials and the loops that you like more and these will inform the distribution of a component so you can work procedurally without the need of manually intervening in the modeling phases but at the same time everything is still related with actual 3D meshes that you can manipulate easily and U and of course uh it's something that they use in my teaching activities so for example we did a workshop in Saudi Arabia and it was about modeling architecture for the metaverse now the metaverse is a pretty fun playground for architect because you can have fun you can experiment with your architectural language and you don't actually care about how you're going to fabricate it because maybe you don't have to and U we use tissue for this Workshop so for example imagine uh you have a simple module like a wall a piece of ceiling and a piece of floor and uh you can apply that to a square face the same thing can be applied for example to the wedges of a polygon of a square and what happen is that that simple module becomes a column so this works also with generic polygons and also with system of polygons so the same thing when applied to just a bunch of random polygons can generate a kind of architectural system and the same thing can be applied to a smaller scale where you have your component that can be applied on the surface and that generate a more complex object and everything is procedurally linked so you can Define the feature of the main system with the feature of a smaller pattern component whatever you want to use and you can just deal with easy modeling without dealing with a a complex U node interface or scripting or whatever and this is just a video that showing practice how this works and I have to say that while teaching to this type of audience we are architect we are designers their goal is not to learn something complex we want to do nice design so for me that this is a way that that takes out of equation the complexity of the scripting of the nodes and just focus on the modeling and I have to say that this is already challenging sometime because if you're not familiar with mesh modeling understanding that the vertices should be align otherwise they don't weld it it's it's not that easy and sometime you spend a lot of time on very simple things and uh it's just better to keep other stuff that is not necessary specifically for the design so at the end they were modeling some spaces and the idea was you upload them on a platform and then you can share with colleagues friends or community and they can explore the space that you that you are creating and at the end of the workshop every participant was designing his own space and uh we designed a main one that was kind of govering some portals for the different spaces and uh this was like the the main hall that you were entering in while visiting the uh virtual installation and here you can see a video that shows shows a different workflow but based on the same principles so again very simple component in this case we added some materials because the materials can be used to inform different component in a second iteration and you just manipulate a little bit and then those are the component that are applied on top of that and if you play smart with very low poly geometry at the end if they combine well together then you can again some you can get something very complex at the end we added some texturing just because we wanted to reduce the number of total polygons and that was helping adding an additional level of detail and uh the other tutor was instead focusing on generative patterns using Houdini for those garment that was were exhibited there um so everything is easy for the metaverse but at some point if you're an architect a design designer you may want to actually fabricate something and this opens new challenges but what is interesting also for a computational designer is that a challenge in the fabrication process can become actually an opportunity to explore a specific language or a specific strategy in the design and uh I developed this uh additional tool in tissue which is the Contour tool and this just allowed to create slices of your geometry and you can play with the angle the distance the offset blah blah blah and also there are some other strategies that are just not planner slicing and everything is non-destructive so you can update the original geometry refresh and uh and just an easy one click way to generate those those curves something very basic but it works well for example with 3D printing if you want to actually design the printing path of your geometry so imagine you want to 3D print something usually you made a model then another software make the slices and then you print it but instead you can actually design the material that is going to be printed and uh usually when you slice a geometry you use horizontal planes and this is fine it works most of the time but sometime you have problem according with some material that you're using and the geometry because if you have a certain angle then the layers tends to add detach to each other and you lose some cohesion between them so an additional strategy that you can deploy is for example to use the jasic distance in order to have offset of the curves along the geometry in order to have a kind of uniform distance between them the version that is implemented in tissue is actually an approximation of a jesic distance but it works pretty fine for the job uh some other time if you want to instead ending your printing vase in this case with a continuous closed uh poly line or curve you can use instead the topology distance which is basically just a distance that count the number of segment that you have in your geometry so you have a very nice topology very regular this can create curves that that start from the bottom and end on the top that's it this is something that we use in this project that has been developed in collaboration with VCU arts in Qatar and it was about a system a simple system of smart blocks and uh what we did in this case was indeed to work with this strategy in order to start with a certain angle and then on a different angle and having the layers that were matching exactly the shape and the topology that we were making we call that smart block because it allowed to be combined in different way uh here you can see how how the layering of the printing is actually generating a texture that is part of a feature of the object that we wanted to create and when we are combined together there is a kind of sense in how the texture is related with the shape of the element and uh it's called smart block because you can rotate in different way and do different type of composition uh another iteration of this design is uh this other version always together with Vu arts and is currently hosted in Expo in Qatar and uh the idea in this case was to take inspiration from the mashrabia which is an architectural element in the Arabic culture where you have a poros wall that regulates the Sun the temperature and uh the environment and uh we wanted to create a simple shape that is always similar but playing with the position of the walls and the angles we wanted to achieve a kind of variable opening of those components it was a bit tricky because we had to 3D print and see if the angle was fine for the properties of the mat so it requires some iteration with Bruno deazi that was the camid that was printing it and we created a series of different component that we creating this kind of gradient we isolated some uh three different design with three different level of opacity and uh we were using them for the production of course when you 3D print something you can produce an infinite number of them but just for convenience we say okay let's let's choose three of them it's just easier to handle so Bruno deazi printed all of them and then we selected some of them and they are currently exposed as M panels in during Expo in Qatar and we created these uh two version the first version is a derivation of the first design that I show you and the second one is the one with the opening and closing uh another part of tissue that is pretty old now is the reaction diffusion uh for me was more like an experiment to see how I could Implement a simulation inside blender using scripting of of course now we have geometry node we have simulation node that would have saved me a lot of time a lot of headache at the time I didn't have it but I I learned a lot so it was it was worth the time and uh these are a couple of videos that he created with this sound designer that was adding some music to uh the videos that I was doing and for me it was like an experiment just to stress a little bit the the pattern and see what are the expressive possibilities of something that is uh extremely organic and also how to change the parameter over time over the surface just to to explore what I could get from that mhm so as you can see there are different areas with different effect and I I wanted really to try to get an effect that wasn't like the conventional uh reaction diffusion pattern one that you usually get right away and and also this is happening on the level of the vertex group and also the vertex colors so I was informing a display map together with some uh Shader process uh properties this second one is using a different workflow is a more simple pattern but I wanted to add a bunch of modifier just to add some extra geometrical properties to that this is something very simple at the end I think but this was actually the inspiration for another project uh that was a collaboration with a fashion brand which is called 3 S4 and strates which is one of the leading company in the 3D printing industry and strates developed this 3D printer that 3D printer that is able to print on fabric they can print different materials with different colors it's basically a raisin but they can change the opacity also the consistency of the material and uh 3 S4 had already quite some collaboration with them so they know how the material Works how how to deal with the design so they started to create a series of shapes and I was taking those shapes as input in order to generate my reaction diffusion Potter I was for example taking the outer edge in order to generate a gradient and having a variation in the scale according to that I was generating a vector field that was informing the direction of the reaction diffusion in order to have like a kind of twisting effect around uh the specific shape then using some geometry node it was just taking care of Extrusion of very simple triangles along a geometry and I generated a series of different shapes that then I sent to strategis for printing uh it's not visible here but there was also uh some tuning using V UV coordinates in order to use a texture for changing the color the opacity and that was possible with geometry node uh here you see a detail of the printed pattern and it's important to mention that all the different extrusions they shouldn't not touch each other so that the fabric can still move and be flexible so they put together all the different pieces and those are some details of a final garment where you can see a little bit the transparency a little bit the color grading of the material and this is the final piece I wanted to create to add the reaction diffusion to that because I think the shape was kind of suggesting this type of marine aesthetic underwater aesthetic um then I spent a couple of year in Denmark working uh with a create group in the University of Southern Denmark and they are working on experimental architecture dealing a lot with fabrication and computational design strategies and one of the project that they did together with them was this summer school 2022 true it was a project with other researcher with all the students involved and uh the idea was to design uh a couple of slabs and when you deal with structures usually you have to deal with stress and uh when you have some specific loading condition some support you have a vector field that represent how the stress is moving in your in your geometry according to this stress there are some stress trajectories that are generated and in order to make an efficient use of a material you may want want to distribute the material along those specific Direction sometime with a stress trajectory is a bit tricky because they converge they are not uniform so there should be a smart way to use that information but that generates something a bit more regular so when I see that I see the perfect food for a reaction diffusion pattern so the reaction diffusion can change the density can change its direction and because of his properties kind of already find the perfect way to occupy the space according to specific Direction so I did an implementation of the reaction diffusion that is based in this case on the texture uh painting the students were able to just provide different texture in order to control the scale of reaction diffusion the direction of reaction diffusion and then According to some parameters you can change if you want for example to have some specific walls or if you want a more interconnected structure that in structure it works a bit better so the student designed a couple of zaps one with so four supports another one with two supports with a different angle and they generated some maps that were used for generating the reaction diffusion so they took the direction of the stress and they created an RGB image from the X Y and Z and uh they just generate a grayscale image for the intensity of the stress and of course the pattern that was generated by the in interaction between them and using them they generated the thickness of this lab while the height is affected by is controlled by the intensity of the stress and at the end we needed to meal it uh when you meal with a robotic arm or in general where you're dealing with Milling there are some consideration that are not that far from 3D printing because for example if you want to reduce the the Ming time the time of the machine that reduce also the cost you may want to use a large step but this generate a quite visible pattern of the Milling path so you have to place Mar with that so one of the strategies that we used was to use actually the jasic distance in order to have a distribution of those curves that was kind of coherent with the shape and we figure out that if the curves are perpendicular to the pattern of of the reaction diffusion they were looking better there were no interference between the two pattern they were looking like they were a good match so this is the formwork that that we M there was some coating some rebars that were bent and inserted inside the formwork and of course the final piece that was produced with the casting and as you can see because of the meing setting that we used the layer is quite visible but at the same time is a texture that kind of makes sense with the with a type of object this is a different View and this is the second lap is uh the the pattern is pretty pretty similar uh we use a different strategy for the Milling but the concept is kind of very similar and the last project is this polyhedral structure that is is almost finished and uh polyhedral structures are structure that are interesting for an architectural point of view and an engineer istic point of view because they offer opportunities to optimize the use of a material According to some specific loading condition uh these open a lot of challenges on how you build those structures especially because you may have a lot of different components that have different shapes and also if you want to deal with actual structure sometimes it's recommended to use reinforced concrete and using reinforced concrete is complicated because you have to create a formwork for the concrete but you also need steel inside that ideally should be weld so we came up with a solution that was 3D printing those uh kind of three-dimensional frames that were taking some uh plates and rebars and converging them to the center here you can see how they align around the center and this was necessary for welding together all the rebars the frame that you see should be open so that the operator can of course weld the pieces together and also we had to 3D print some shells that were needed to close the forward now the pressure of the concrete inside those component while still liquid was quite high so we had to reinforce those shells by using this pattern that was kind of giving more St strength to the material without the need to increasing too much we wanted to reduce as much as possible the printing material and we were closing the formwork and this is the closed forwork then you pull the concrete from the top and you have the final piece that is produced as you can see all the feature of the concrete are coming from the formwork so when you design the piece you should think at the formwork and when you design the formwork this is kind of affecting the the piece that you're producing and you can see how the three-dimensional frame is generating those grooves along the geometry and the the curves on the shells are generating this pattern on the final piece and uh the formwork is quite complex and as I mentioned the structure has many different components so how do you model this in a smart way so that you don't have to manually remodel for every component and as they say if you are a number you see Nails everywhere so my solution was to use tissue and the so if you have like a topology which has some faces that are aligned with a feature that you want you may generate like just three simple components that contains the geometrical information for defining your formwork and as you can see there are three different components two of them are taking care of the overlap of the shells and one is just containing the wavy shape and the thing is that when you combine all of them together what you get are some different geometry that then can be separated easily what happened is also that in the component itself there are both the parts of the formwork and the surface of the concrete so what happen is that you see both at the same time every iteration so I was evaluating the aspect of a final piece of concrete and also the shape of a forward so this was an iterative process so I could evaluate exactly the consequences of both on both aspects in order to deal with that we needed a very uh nice topology and we tried like remeshing and was not good we tried like uh different strategies and we couldn't get a nice topology that was aligned exactly with our feature so for that I had to add another piece at tissue that was the polyhedral wireframe that is basically doing what the wireframe modifier is doing but dealing also with inner faces in your geometry all the things that usually messed up your wireframe are working find with a polyhedral wireframe we tried also the skin modifier that didn't work well because it's creat a nice topology but not perfect so we added some extra strategy extra rule just to have like planner intersections uh seams between the components because we have we had to use some uh flat uh steel and there are a lot of things that you don't consider while modeling because yeah you see a nice shape and that's it but then every aspect when you have to fabricate actually becomes a problem for you that you have to solve and there were a lot of challenge to solve in this case so this is a a system of different component that we produce with that strategy and we assemble a structure of 3 m more or less that unfortunately I cannot show you because it has not been disclosure yet so what I can show you are just some dramatic renders that I did with cycles that kind of give you the idea of uh a more complex system while assembled and uh I work on uh in different fields and on Research education and uh practice and I have to say that blender is an amazing tool for visualizing new words and new ideas but is also a valuable tool when you actually want to produce when and fabricate the word that we are living in and uh I want to thank of course the blender development developer for making this amazing tool we Community for all the valuable feedbacks that they gave to me and also the my patrons that are really supporting me and giving me really the enthusiasm for continuing this project thank you very much

Channel: Blender

Views: 5,522

Keywords: conference, BCON, BCON23, Blender Conference, Amsterdam, Felix Meritis, b3d, Blender 3D, Conferencia, bconf, blender conf, blendercon

Id: mLT2clOq0UQ

Length: 23min 43sec (1423 seconds)

Published: Sat Oct 28 2023