Dynam(o)ite Your Steel Design

DIETER VERMEULEN: My name is Dieter Vermeulen. I'm a technical sales specialist at Autodesk in northern Europe. So I'm based in Belgium. I flew over here with a few Belgians last Saturday. went to Death Valley on Sunday which is-- it's pretty dead over there, right?

And, yeah, I'm a structural engineer by trade. So I used to work for some engineering companies before I joined Autodesk. So I'm trying to use that experience I had from that moment to share all kinds of best practices and workflows to customers and resellers and so on. Yeah, I had to tell someone from AV that it should be pressed on a button to record or something.

So today in this Dynam(o)ite Your Steel Design Class I'm going to talk about steel, obviously. A lot of steel. Steel in design, steel in optimization, steel in conceptual design, steel in fabrication. So it's all about steel. So people among you who never work with steel and are not even interested in steel or hate it then, yeah, stay here. You might be afterwards.

So the key learning objectives for this class it's-- yeah, as I told you it's about designing something with steel. But I want to use computational power to get into an optimal steel design. So we are going to handle with some complex models.

And I will try to show you how you can simplify all of these things. So how simple it can be to work with optimization, to automate stuff to create very complex designs. And even to implement all kinds of algorithms in your designs to make very compelling stuff around it.

Now, let's first have a little introduction on that. I'm not going to tell you the basics of Dynamo today. So this is really something advanced. But I want to give you a little bit of insight in how we can position computational design and general world of design in engineering. So let's have a look at this guy.

The engineering expertise always has been in people's heads. So it's in our minds. That's something you cannot change. It's always there, and it will be there in the future as well. Except for if we believe about those robots, but that's another discussion.

Now, in the 60s the NASA wanted to use a-- they needed to find an optimal design of this quadrifilar antenna for their orbital missions. Now, they used decades with that engineering expertise to find an optimal solution for that. Now, in 2003 they did, like, a very groundbreaking project by making a special design around this. So by using computational power still combined with the engineering expertise, but then to find a way more optimal design.

So that was a first time where computational power, where a computational design actually got used into two very complex projects. It might look a little bit like a kindergarten project. But it's really something exciting to see that that time they used already the computer with that kind of power. I'm not really good in these hardware stuff.

But in 2003 they already had, like, a very powerful system to use. Like right now we are on that blue dot on top right. We have much more powerful computers that can be used for that.

Now, the buildings from back then are still buildings. And if you talk about the building, there's always a back structure that's behind it. It's not only about a nice facade and some windows and doors. There's always, like, a bearing structure. But the deliverables throughout the years have been changed.

So actually if we look at steel design, the questions we ask ourselves when are we creating something in steel there are always the same. They will be the same in 20 years. But deliverable will be different. You might want to do something automated. You might want to do something which is more optimal designed.

And this means that the expertise, it's still in our heads. It's still that the engineer that will use his mind to create a very complex thing. But the difference is now that we are combining this with computational power. And that's where, for instance products, like Dynamo come in when you need to create like more power into your design.

So if you have a look at the old way of designing, there is that one human that uses a computer. And you get a limited design, for instance, for a chair. And you might get a chair like this or this or another one. You're thinking in functions. You're thinking about, OK, this is what I want to have, and you document it into your design software.

Now, if you look at it from another perspective, this chair has a specific weight. It has a specific cost price to fabricate. Now, it might be possible to get an optimized version. And actually that's what Autodesk has done with Autodesk Research, which is like a department doing lots of R&D in these kinds of areas.

So they took it from a different perspective. They looked at this chair from a perspective of like we want to have something to sit on with a specific height, with a specific back height. It needs to be able to support that kind of weight of a human being, for instance.

Then this was the outcome. What we might discuss about its comfort-- I would love to sit on it. It's maybe even not aesthetic. But it's really about optimization. This is about finding a solution instead of documenting a solution. So in the new way of designing, it's still that one human. It's still that human brain combined with artificial intelligence algorithms and computational power. For instance, Cloud computing.

And then you get, like, hundreds or thousands of design options. At the same time that designer would need to only describe one in that old methods. So that's what it's all about when we talk about generative and computational design.

Now, yesterday we heard our CEO, Andrew, telling about more, better, and less. It's perfectly fitting within this presentation in m we are going to see how you can create, like, more steel structures, more-- where you have more possibilities to create. You will be able to create it better. And more specifically in this class we are going to learn to how you can save-- how you can reduce waste in fabrication of steel. So we are going to use less resources to create very complex models.

So I'm going to present two cases. They are no-- not real cases. So as technical sales, we don't use customer's data sets. We try to create our own data sets.

So if someone among you has like a lot of money and wants to build that thing that I'm designing, then please come forward after this meeting and then we can have a chat about it. I've called it the Mars Torus Arena, the first case. It's about a thing that-- yeah, you might have heard of the Dynamo design slam this evening. I will talk to you about this a bit later on in this presentation.

It's actually the thing I'm going to do in the design slam. And that's the one you see on the bottom right. It's actually a diagrid structure. It looks complex and simple at the same time. Actually to build it and-- or to create it as a design with products like Dynamo, it's very simple. If you want to create this with Auto-cad in a 2D drawing, then you might have some issues.

Now, the thing-- the project goals for this virtual project-- let's call it-- is that we want to create an economic design. A design that is such-- well, that is-- how to say it. A design that's build in a specific way that we can reduce waste, but also that we can find the cheapest solution in terms of cost, for instance connections, or the beam cuts production.

So the evaluation of the structure before we design it, it's something where the cutting costs will come in. And that's actually real cost pricing we have in fabrication of those structures. Of course, there is a structure weight. And we need to find a structure which is very lightweight, but which is still meeting the requirements for strength. So we will involve analysis in this one.

The steel connections they also have a specific cost, depending on which connection you are using. If you use a gusset plate, then you might have a specific connection plus connection price for it. And then, of course, cut length optimisation and material waste. That's something that goes along.

So I'm going to show you how you can use Dynamo, for instance, to group all the beams into specific bins to optimize their cut lengths. To see from which stock length are we going to cut that beam. And, in that way, trying to reduce waste.

So there are four phases for that project. And the first phase will be like a conceptual engineering, conceptual optioneering. A second phase is where we are going to use robot structural analysis to get to an optimized steel design. Then using Revit to document it with framing design in the third phase. And then in the fourth phase we are going to use advanced steel to create fabrication drawings from it.

Now, before I start I just want to do some little show hands to see what the knowledge is in this room in here. So who among you is using Dynamo already? Oh, wow, that's nice. And robots? OK. Revit? Of course. Advanced steel? OK, so Revit and Dynamo is quite popular in here. OK, good. Fractal? Great.

So actually the whole thing around these four phases, at Autodesk we call this Connected BIM. You might have heard it already in some presentations, keynotes, maybe other presentations in the past which are outside of AU. And Connected BIM is actually where your BIM data is streamlined between several products. But not just in a circular way, but like in a 360 degrees way. And that's by connecting our products with each other through Cloud services.

In this case, if you use Dynamo then it's not that's connected as we would like to be. It's still, like, some fragmented workflow that we will have in here. But you will see how fluent we get throughout these phases and re-use data from one project to the other.

So the products that are used in here-- so this is like a solution mapping. But beware, this solution mapping is not like the only possible way of how you can combine them. This is how they are combined for this project. It might be that for other projects you might use Revit, for instance, across the whole phasing. So starting from conceptual design and ending at the fabrication.

I started with the concept with a Dynamo Studio, Fractal, and Dynamo Customizer. And then got through with Robot, brought it to Revit, and then brought it to Advanced Steel to get some more streamlined design in here. Now, the Dynamo platform-- to avoid some confusions around this-- you saw that they used Dynamo Studio, Dynamo for Revit, and Dynamo for Advanced Steel, these are three separate products. But now with the new AC collections, they are all included.

So you have Dynamo Studio in the AC collection. You have Dynamo for Revit with your Revit license. And you get Dynamo for Advanced Steel throughout your Advanced Steel license. So it's all included in that collection right now. Where in the past you had, like, separated versions. And you had to buy Dynamo Studio, then had to buy Revit, and so on. So quite confusing in the past.

Now, the difference between those products is that Dynamo Studio is the only one that can connect to these Cloud services. So it can connect to Fractal and to Dynamo Customizer, while Dynamo for Revit and Dynamo for Advanced Steel they can not connect to that Cloud service. So that's something important to know in here.

There are also a few packages that are used in this class. One is called structural analysis for Dynamo. So it's an in-house package built by a software architect that used to work for the robot team-- or still working for the robot team at least. And then BIM4Struc.Productivity, that's my own package used in this thing to do all kinds of optimization stuff that you will see later on. Oops, sorry.

So you might have seen in the app that the handout is a very extensive handout of two pages. The reason for that is that I didn't build any huge handout like I did last year's, like. 120 pages because all is explained in this presentation through the videos. So the videos are recorded that way. That actually you can just follow them step-by-step. And now I'm going to talk through it how you can use it.

You can watch the recording of this session afterwards if you want to replay it yourself. And the data sets, it's like a huge bunch of data sets. Everything I'm using in here you can download it through that handouts. In the handout what you will find download links for all of that-- all of these things.

So the first workflow is a conceptual structure optionering. It's where we are going to start the basic layouts of that diagrid structure and bring it through Dynamo Studio, Dynamo Customizer, back to Fractal, and then use some Optioneering to find the ideal structure for it. Now, that ideal structure is based on an evaluation, on the concept evaluation. So there is an input, and there are some constraints.

Bringing these two together will generate us a specific evaluation of that structure. So the input is like the dimensions, of course, of that diagrid. The constraints are something like the beam stocklengths. So let's say that we are using a beam stocklength of 12 meter. And also there is an allowable waste percentage.

So the contractor in here, he doesn't want to have a waste of this exceeding, for instance, 4.5%. Let's just give it the number. Now, the evaluation of that structure is based on cost. So each of the beams needs to be cut so that machine is producing at cost. The connections-- there will be different connections, as I told you.

There is also a weight of that structure. So the structure itself-- the use structure-- will have a weight. So that will have a cost price. But also that waste will have a weight. And that's also something that costs money.

So based on all of these things we will have the total cost. So the idea behind this is that the Optioneering you will find a solution which is the cheapest one, but which still has less waste, which is also still using an optimized bin-packing, as we call it, for the cut optimization. Now, how is this cut optimization working? It's using an algorithm called the First Fit Decreasing Bin Packing Algorithm. Wow, that's a long name. So we just call it FFD.

So imagine you have this. It doesn't matter what it's representing. It could be, like, beam lengths. But it could be also, like, trash that you want to organize and you want to use as less as possible bins to trash it. Now, first you need to order them. You need to order them in a decreased order like this.

And then we can start this FFD bin packing. So we are saying that one of the-- each bin has, for instance, a maximum capacity of eight. Now, we need to put all these blocks in there. So how does this work?

Well, we start with the top one, eight. Then taking the smaller one and go down until we reached the smallest point. And actually it's just looping throughout all of these things, all of these bins. Finding do I have a place in bin number two? No, there is no space in there. So let's go to bin number three. And so on and so on.

So it's actually just a loop. And that's giving you a result waste, which is three. Now, this is almost the most optimal algorithm that you can use for these kinds of functionalities. There is one more. If you go-- if you Google on it. And I think it's even in the presentation. You will find a link in the speaker notes which is sending you to the specific page on the internet showing all the possibility-- all the possible bin packing algorithms.

So there is one more which is more optimal than this one. But it requires a lot of programming to get it in. While this is programmed in Dynamo in a very simple Python script, and you will see the results later on.

So let's start with the concept. So as you can see, it starts with inputs, a thorough surface, and a structure creation. So there are three parts in that script. And the input is, of course, our dimensions-- the dimensions of that arena.

Now, the area in here, it's based on a surface. I'm going to create like a Torus-- or like a donut. We are in an America so I need to use donuts, of course. But in this case, I'm using a 1/2 donut. It's just like a specific part. It's having a starting angle and an end angle. And this is the surface that will work as the base for this whole diagrid structure. Here you go.

So some points get generated on that. And these points are actually the connection points for the diagrids. Now, there are a few packages available in Dynamo such as, for instance, Lunchbox, which has a specific node called wire diagrids. You cannot use it in here, because it's not compatible with Fractal yet. Fractal can only read out-of-the-box notes and no custom notes, right?

So it can use Design Script, and it can use the basic nodes that are available inside of Revit. So that's why the diagrid structure in here has been recreated with nodes placing back to each other. Now, when we want to calculate the connection costs, then we also need to filter out the specific connection points.

So there are connection points at the side of that arena. There are connection points at the roof. And there are also connection points where you have four beams crossing each other. So these are the three types of connections in here, and each of them will have a different price.

So to get to that economic structure, it's not just like-- let's do it simple and let's say instead of using 200 beams let's use 100 beams. That will be the cheapest solution. No, because that might maybe produce more waste. It will be less strength, or the strength at least will be very reduced in here. So it's a very difficult thing to say, this amount of beams that's the perfect solution for all of these things. So you really need to Optioneer it to get to your optimal solution.

So to get to that point, we need to evaluate a structure. You need to build an evaluation, a score for that. So the score in here is, like I told you, the cost price. So let's add that part on the right. And we need some more inputs-- the stocklengths and the cost.

So this cost is just like effective value. Like, OK, let's imagine that this connection will cost that much. It doesn't even need to be something in dollars or euros. It just needs to have, like, a number for the impact. So that's how your score is working.

And that's always what you have to do in optimization. I've saw it already in previous years here at AU as well. If you're doing optimization, you need to find a way how to score your solution. And then the lower your score, the more optimal your solution is. So that's the way how it works.

So this first [INAUDIBLE] decreasing BIM packing note, it's one that's included in a BIM4Struc.Packaging. It's working very simple. You fill in the values, and you fill in the BIM sites. So the values are like, for instance, those blocks-- eight, five, six, there, two, and so on.

So the values in here are the length of each of the curves. So the diagrids created all these small lines and longer lines as well. Each of them has a length. And this length is then calculated with that FFD, put it into that bin of 12 meter. And then you see how it's getting ordered on these watch windows. And plus for each BIM, you get a waste.

Now, I'm doing just some formulas. It's just like very simple mathematics. Where you're taking that waste. You connect it with like, for instance, the bar unit weight. And then you get the total weight of waste. And you get the total weight of your structure. Plus in here just doing some adding and multiplying, and you get the connection cost.

Now, very important if you want to use Fractal is that we need to nickname our watch windows. So like the structure weight, the waste, material cost, and so on. You double click on the watch notes, and you give it a name. And that will appear as results in the Fractal platform later on.

Now, this structure has a waste percentage of-- what was it again-- 4.2% or something. So this structure is like, OK, now let's give it a different value. And we still want to find a solution which has less than 4.5% waste.

Now, if we change, for instance, the number of beams in here-- no, not a number of beams. It's like the division, the transversal division of that Torus. Now, it appears in red. This means that the waste percentage has been exceeded.

So we still have a solution, which has a specific cost. And it might be a very lightweight structure, but the waste is exceeding that amount. So we don't want to have that. We want to reduce that waste.

So this color is just like an indicator, and it will be visible in Fractal as well. You can use it in there to see which solutions are acceptable and which solutions are not. Another way of visualizing these results is-- I'm using in here T splice. And, for instance, I want to see each BIM packing group. Which beam belongs to which stocklength? So to see how it will be manufactured, right?

Because it might happen that you have like, for instance, a very scattered positioning of each of the beams belonging to one specific beam length, and this is how it's visualized. So the FFD custom nodes, it also gives us a list. So as you could see, we have the curve lengths at one hand. And then the FFD algorithm will organize all these curve lengths in the right order until we find an optimal solution.

And a third result you get from it is, like, the index. So where was it positioned in the original curve length list? And that's very easy to find then the beams that are belonging to a specific BIM group. So, for instance, BIM group in here, number 71, represents these beams that you can see on the top right in red.

So if you're working with Dynamo Studio you can send this script to Dynamo Customizer. Which means that if you want to share your design with people who are actually not-- well, let's say not capable. That might be a little bit straightforward. Not willing to work with Dynamo, or they just don't have the time for or are even scared of seeing all these blocks with these wires and so on and I think it will explode their computer, and so on-- then there is Dynamo Customizer.

So you upload it to your Dynamo reach server, and you share this whole thing with that customer or your other partner who you are working with. And they can use it just like this, without having to go through all that Dynamo magic. So they just have the sliders and the numbers. They fill in the values they want to have. And then see a result life in that web platform.

So that's the goal actually of using Dynamo Customizer. It doesn't optimize it yet. It's very manual. So you're optioneering actually your structure in a manual way. Now, when Fractal comes in that's where we are going to do the optioneering in an automatic way. So we are going to use the server to create all the possible design options for us.

So in here, again, it's exactly the same script that's used in here for Fractal. And as we already looked at, it's possible to use those colors to get like a visualization of that waste. Is it compliant to an allowable waste percentage? Yes or no.

Now, if we send that to the web give it a-- you give it a name and so on. Give a description and publish it. And then we are going to open up the Internet Explorer and get to www.fractal.life.

Now, as you see on the top right, over there we got our inputs. And the inputs are different the division parameters. You also get your constraints like the stocklength. And you get a constraint like the allowable waste percentage, and then started generating. Press that button and generate it and send it throughout all these possible options, let's say.

So it's calculating, it's executing that Dynamo script multiple times. Like in this case it's executing it for 100 a time-- 100 times at least. The results you get is on the left you see a black input. On the right you see that red input.

The black inputs are the parameters, so your inputs parameters. The red things are actually the output. These are the results. And in that parallel coordinate system it can filter all these results and find, like, the ideal solution for it.

So you filter them until you have like, for instance, line solutions. You pick one of them, you download the Dynamo script, and then you go on with it. Now, what do I mean in here with optimal? Just a second. The optimal end here is what-- which one has the lowest waste percentage. Which one has the lowest cost price? That's what optimal means in here.

And, yeah, I saw already some hands. And I'm really excited you have questions. But at the end of the session I will try to keep some 10 minutes time to have, like, all the other questions in here.

Now, let's go to phase two. Phase two is where structural analysis comes in. So we are going to reuse that whole diagrid, bring it to robot, evaluate the results. And then even more we are going to perform some what we call parametric optimization, right?

Now, in other years at EU maybe some of you went to one of my classes. I did something around genetic optimization where we introduced, like, the natural evolution theory into structural designs. And that could be added in here as well, but I didn't do it because we don't have time for it anymore.

So how do we can figure that analysis model? That's where the package structural analysis for Dynamo comes in. And as you can see, the left part of the script is exactly the same as the one we used for the Fractal calculation. It's exactly the same as the one we use for just the conceptual engineering.

And then at the end-- at the back end-- we are going to add all these notes in here to create analysis models and robots. So it's also a combination of Dynamo Studio with robots. So in robots itself you don't have, like, Dynamo for a robot that doesn't exist. So it needs to be connected throughout Dynamo Studio.

Now, it's quite a heavy part to explain everything around analysis. I will avoid that part in here. You will find that easy in that script. It's very readable, the script, how you need to setup.

You need to setup beams. You need to setup boundary conditions. You need to setup loads. And then you can start making your analysis. That's the basic idea behind it.

If we have done our analysis, we can get results. So the intention of this whole script is actually to make all these kinds of design options by changing parameters. Get results from it such as structure weight and a maximum deformation. And then make your decisions based on that.

So let's do it. So let's minimize all of this a little bit. And you will see immediately that the structure will be generated, and the robot will be calculated, and the results will be sent back to Dynamo. And these are the results. Well, let's go back to the results in a few seconds.

Let's first have a look inside of robot. What do we have actually as results? We got our load cases. And each of the load cases are reproducing, for instance, reaction forces. So if we want to know the self weight of that structure, we just take the reaction forces-- the total reaction forces for that self weight-- and then we have it, right?

But if you want to create, like, a deformation-- we want to see the total deformation in Dynamo. It took all the load cases separately, and then added the deformation values, right? Now, in robot what we are going to do is to add up all the load cases into one combination, calculate it, and then see if it's correct. That's actually what the engineers do.

If you perform-- if you execute, I mean, an analysis you want to verify with some other tool. So that's what we are actually doing in here, verifying the results from robot with Dynamo. Now, once this is verified you know that, OK, we're fine. We can use Dynamo for this whole exploration, let's say, of that structural analysis model. So we can start changing the parameters inside of that script.

You get new values. Plus each of the designs have been saved as a separate file. So we can open up that file afterwards in robot and then see, OK, maybe we should use this specific analysis model, tweak it a little bit, and then go on. That's a possible workflow.

Another workflow is the one that I prefer the most in here. It's something that we can call parametric run or manual optimization. And we are going to use that script, and just loop through it.

So we are going to say to Dynamo, OK, execute a script, like, hundreds of times. Generate all these robot models. Calculate them, and give me back the results for weight and deformation. And then I'm going to use Excel just with a simple diagram magic, and then find the ideal solution for that based on structure weight and maximum deformation.

Now, the way you do it is by creating a custom note of your analysis. You need to create like what I call a project custom note. No it's not like a custom note you are going to publish on a package. It's just for this specific project you're creating something.

And within that custom note there is all that analysis stuff. So the BIM model, the constraints, the boundary conditions, the load cases, the loads and so on, and also the results that come out of it. We are going to feed that custom note with these design options that you just saw. So the design options in here are like the transversal division of that arena, the lateral division of that arena, and then also some sections.

Because, of course, you can have multiple sections to create that structure. And based on that specific type of section, you will have a different deformation. So it's not just like changing the number of beams in there. It's also about creating a stronger structure.

Now, once all of this has been calculated-- you will see it immediately in here-- it takes like seven seconds to create a new model and calculate it. That's quite fast. If you need to do this manually, yeah, come to me if you do it in seven seconds. I would love to see that.

You get some results from it. And now I must say, if you will test it with the data sets yourself and you find out that after 15 iterations it stops, then it's normal. It's something I experienced myself. I just told you it's like hundreds of design options creating. But there is a problem with this specific model, and it stops after 15.

So I didn't found out yet what is the problem. It might be my computer, but there is something very strange in there. So I must be honest on that. You will notice if you try it yourself.

But if you look at the data sets from the other classes from last year, you will see that it goes on for like thousands of design options. So it's very weird. Now, the Excel note here-- the right to file-- will take all of these data-- so the inputs design options, like number of transveral division, lateral division, section name, and then also the results for structure weight and deformation-- and it puts it into that table.

And of course, yeah, Excel is still the most used engineering tool in the world. Just using some diagram magic, and then you'll find out that if we have a structure with a lot of weight, then the deformation is, of course, very low. Now, we need to find a solution where the weight is very low and the deformation is also very low.

So as you can see on the left, for instance, we have a really heavy structure for the first and second design option. The 15th design option looks like one which has a very low weight and also very low deformation. So that might be the ideal one if we look at it from a weight/ deformation perspective. If we look at it from, for instance, the conceptual perspective like costs and waste, then it might be that the cost/ waste has a different solution.

So actually, if you want to continue this workflow you need to go back to phase one. Go to Fractal. Fill in the values for option number 15, and then see what the conceptual engineering says. If that's one says, OK, you're good. You're under-- you're below that 4.5% waste. Then you have, like, your optimal solution in here.

Once you've got to that point, you go to phase three. That's where structural design comes in. We need to document it. And I want to document it with as less as possible paper. You've heard it already throughout several presentations in here. Maybe some of you have been to the exposition hall and saw the guys from Norconsult, how they did paperless.

This is also-- like we want to do its paperless in here as well, right? So to get to that point you need to have all your fabrication data. First, we start with an empty Revit screen, of course. And we need to create that Torus Arena inside of our Revit model. And what do you think, we are going to reuse the diagrid script again, of course.

So the first two parts-- input Torus surface and structure creation-- are the same every time. Throughout each script it's exactly the same one. It would be smart, maybe, to create a custom note from it, and then have less space on a Dynamo Script. But I didn't do it just for better explanation in here, right?

And the creation of that inside of Revit is actually very simple. There are only three notes that you can use-- column by curve, beam by curve, brace by curve. In here there are only two. There are no columns. It's actually beams and braces. And they act as how they act in the Revit. If you use a brace, then in Revit a brace is something that connects to a beam in a relative way, right? So that's very important to know.

Now, sometimes if you want to detail all of these things in the Revit, then you might see that it's using a start extension and an end extension. And it might be a little bit confusing sometimes how to get, like, to a very uniform design. So in this case, to document it in a better way I'm performing some specific operations by setting the parameters.

Like, for instance, the location point. How is the beam located throughout his location line? So in this case I want to align it in the set direction. I want to have the set alignment at least set to top.

Besides that I also want to have-- and I need to put in a little bit on pause, because I's talking too slow apparently. It's also-- what it also did a few seconds ago was disallowing the joints. You know if you create beams in Revit and you connect them with each other, you might have-- yeah, sometimes it's annoying that the beams are connecting with each other. But that's-- for normal structures-- necessary.

For these kinds of structures, you don't want to have them joined together. Because it might give you lots of side effects that give a difficult way to document it. So in this case, there is a custom note. There are a few custom notes available in the world that are automating all of this. Disallowing all these joints, and then you join them in a manual way by setting the parameters for your start and end extension. So just as simple as that.

Now, let's continue the video in here. I also want to have fabrication data in there. Fabrication data could beat the BIM mark. It could also be the BIM packing. So we did some BIM packing before. And maybe you want to know which beam belongs to which beam stocklength.

So, for instance, there is, like, that huge beam of 12 meter. Let's say, bin number 12 it comes in here. Which other beam for the production should be cut from it? So this information can be stored in there. So let's see how to do that, and continue that video. And now it's restarting it. That's fine.

So, yeah, the joints like I just talked you about. And here we have the marks assigned. So that's a very simple one. It's like count your beams, add some prefix to it, and then number them. So that's like automated documenting. It's not really it's not something like creating a sexy structure, it's just adding metadata to that structure. That's all about it.

And that's very important as well if you need to number them. Yeah, in this case they will be numbered in a very logical way, because the diagrid has been created from right to left, right? Now, the next part is where that's fabrication data for a BIM packing needs to get in.

Now, it's again-- it's actually it's a continuation of that same script. And it comes after the mark assignment. So what we are going to do in here is take the cut lengths from the Revit model, not from the Dynamo model. Because in Revit we use, like, start and end extension. Maybe you want to use other parameters in there to document these steel beams.

Now, we are taking the values from the Revit model, put them into that FFD algorithm, group them together in the right bins, and then assign the bins to each of the beams so that you can know which beams should be cut from which stocklength. And at the end we get or waste calculation.

So in this case, the structure generates a waste of 346 kilograms. Which is quite OK for that kind of structure. And then also in this case the structure weight has been calculated from what you are using in the Revit model. So it's actually Dynamo sending information to Revit. Revit sends information back to Dynamo. And Dynamo sends it back to Revit.

So the information that Revit sends to Dynamo is like the properties of each of the beams. So what is the unit weight of each of the beams? And the information that got sent back is the marks and the BIM packing.

Now, the last phase for this structure is where we are going to include fabrication in Advanced Steel. So you can guess it, the first part of the workflow reuse the diagrids again. Now, in this case we are going to reuse it in Advanced Steel. Now, there is also a Dynamo interface within Advanced Steel which makes it possible to create frame and plate models.

It's still a bit-- it's not as extensive as the possibilities you have from Dynamo with Revit. So it's not possible yet to create connections, or at least not possible out-of-the-box. I saw customers doing it already. They create connections with Dynamo, but it's more like something they coded themselves. So it's not out-of-the-box available yet in that.

So in this case, you need to open up Dynamo for Advanced Steel from within Advanced Steel. So it's not something that is using, like, Dynamo Studio. But as you see, the first three parts-- input, Torus surface, structure creation-- exactly the same. And now we need to bring it back to these Advanced Steel notes.

Now, there are two types of beams that are used in here. We have, like, the roof edges. We have the sight edges of that arena. And we have the diagrid beams.

The diagrid beams are the only ones that are straight. The other ones are curved. Now, if you want to do this in Advanced Steel, a curved beam is an arc. So you cannot use splines, for instance. You'll always-- if you have a spline then you need to convert it to an arc before you can get it into Advanced Steel.

The reason is very simple. In reality, you don't create splines. You are combining arcs to get to that spline, right, if it goes to fabrication. So this is the one you get in here.

And the sections got assigned from within Advanced Steel-- sorry-- from within Dynamo. And it's based on how they get documented into the database, into the Advanced Steel database. So they have a specific name in there. That name is set up and Dynamo for Advanced Steel. And you create like what we call a section strip, which is used then to assign them inside of the product.

Now, the fabrication details in here, as I told you, it's not possible yet out-of-the-box to do this with Dynamo. So this is pure Advanced Steel work. We could use, for instance, also that information of that bin packing and send that into the properties of each of these BIMs. But that's not necessary here, because the bin packing has been documented already in another part. So-- but could be possible.

So, yeah, there are lots. I'm not going to show you all the functionalities of Advanced Steel, or you don't get-- yeah, you will not even make the party tonight and we will still continue in discussion about it. But, for instance, in this case it's very handy to use some tools to cut them towards each other. So in this case, I want to have for instance a welded connection.

Maybe you want to cut these beams at some parts and then make modular grids, for instance, for this arena. So really depending on what you actually want to do. But very easy to work with. So you might think, OK, this is Auto-cad. Ugh, I don't want to use Auto-cad anymore. I'm a Revit user.