Steel Connection Design—Extend Your Advance Steel, Revit, and Robot Structural Analysis Workflow.

JURAJ SABATKA: Good afternoon. Welcome to our class. Let's get going. People might keep coming, but let's get started. We have something, about an hour.

My name is Juraj from my IDEA StatiCa. This is Lubos, also from IDEA StatiCa, and Graham from Arup. And we would like to, once again, welcome you to our class on steel connection design.

And just a quick intro, IDEA StatiCa is a solution associate of Autodesk. We develop software that we link with key Autodesk products. And one of the big topics is steel connection design. And that's why in today's class we will explain, introduce, and demonstrate how the design and check of steel joints can be done differently. We will introduce a new approach to do that, method behind it, the software itself, and the workflows with key Autodesk products.

In the class summary, you might have read that we'll be talking about steel joints of any topology. And we really meant it. We want to show that we can do very simple joints, standard joins, you all know. We can also do the slightly more complex, but also the devilishly complex ones. And hopefully we will show you how to-- that it actually is possible to design and check, according to AISC or the Eurocode, any type of steel joint in a reasonable time frame that is minutes-- that is not hours or days-- but do it in a regular part of your working day.

I do understand the limitations of currently-used methods. Methods behind design guides, you are well-aware of, and you use when choosing an appropriate joint for the project. You will see how we work with models from Advance Steel, Robot, and Revit, and how to do steel joint design of joints from these models.

And this approach is not new. It's been around for two years live. Before that, it was three years of development and another couple of years of theoretical preparations, so two years on the market. So, we were very happy that we can show off how it is working on real-life projects. And Graham from Arup will also go more into detail about how Arup uses that in their work.

So, why to even bother with steel connection design or steel joint design when the whole structural model is an immensely complicated structure? And we all have programs for that, a CAE or a finite element analysis program, if you want to. And we all use them. And they are great pieces of software.

The thing is that these programs, they focus on the model as a whole-- the whole structure. But the experience from the construction process is that the majority of problems that occur in steel projects are usually linked to a detail, to a malfunctioning or badly-designed joint. More often, this is mitigated by the engineer spending more time on it during the design process or putting more material there. Bluntly put, if you're not sure, you put more material there. That's why we want to do something dedicated to steel joints and to work with other programs to get the model of the structure as a whole.

In the projects, we have two types of joints. We've got standards and non-standard-- easy and difficult ones. For the standard, we have well-described and proven design guides-- whether it is AISC guides in US, or in Europe, we have prescribed formulas, examples, we can use those joints.

Well, the non-standard, it's a whole different story. When we step out from the territory of mapped joints and already-used joints, then we have to rely on putting together modeling on your own, using a scientific finite element analysis program, usually, then using a lot of Excel to do the code-checking-- honestly, use a little bit of estimation on the way. And that might lead that some engineers just decide to adjust the project a little bit to just avoid using that particular joint if it cannot be checked, using the standard formulas.

I'm pretty sure you all hold a university degree. And if you remember that you had to pass like 40 or 50 exams to get it, and not all of them are of similar toughness, of similar complexity. You probably spend 70% of time on 10 to 15 most difficult exams. And steel joint design is very similar. We spent most of our time on the smaller amount of more complex joints. And that's why we came up with the new approach to do all of them, not just the complex, not just the standard, but all of them. And Lubos will explain the theory and solution behind.

LUBOMIR SABATKA: Thank you. I will continue a little bit with the theoretical part of this presentation.

As Juraj said, now when we want to design a steel joint, now, something like this, two bolted plates, we have the option to use Component Model, which is method which is normally used in all design guides and in codes-- in American code, in European code, all modern codes are based on this Component Model. The model is-- in fact, when we want to calculate the internal forces in all parts of the joint, we have to prepare a specific model of such springs. And each spring can represent, for instance, bolt or [? welding ?] pressure, [? welding ?] tension, and a flange in the same.

So, this quite complex model has to be has to be derived. It's not for everybody. Normally professors and universities do it. And then, this model is composed, all internal forces in this model are calculated. It means that finally we know the force in each bolt, in welds. And for this, probably you can use check formulas, which are, again, in this code.

This method is frequently used. It's very sophisticated. It's very reliable. It's safe. It is not the problem from the point of view. From this point of view, the problem is that this model has to be prepared for each shape of connection specifically. And it means that it is not general. It is known for typical topologies. But it is not known for any shape of connection.

So, our idea was that if we have here a problem, we should get rid of this and replace it by something else. And the first idea, quite simple, is that we use finite elements. We use finite elements for analysis of building or for frames. So, it is, again, it is accepted method as good, reliable. So why not to use this finite elements also for connection design?

And it is this idea of a [INAUDIBLE] to this business. And it means that we calculate these forces in all parts of connections by finite elements. And then, they use the same formulas like in this component method. So, we came with this name, "component-based finite element model"-- "CBFEM."

It is quite a simple idea. But if we want to succeed we have to do some important things. One of them is that when we calculate the steel, we cannot use standard elastic analysis of the structure because steel is a plastic, and only in this case we can get good results.

Maybe you know these stress trend diagrams for steel, which are known from tests in labs. But these, we cannot use because we do not want to calculate reality. We want to check the connection, to make the check which is conformed with the code. And in the code, they speak about this elasto-plastic material. So, this is implemented. In this case, we cannot do the check of stresses because the stress cannot exceed this value. And the check is done for equivalent plastic strain, which is recommended in codes to 5%.

Well, this first thing which you need to do is plastic analysis. Then, we need high-quality finite elements [? software. ?] It is true, and some of you maybe remember that structural engineers were [? binaries ?] in using of finite elements in '80s, '90s, last century. Structural engineers were really on the top of this science.

But from that time, our industry was little bit sleeping. And the other industries moved faster. For instance, in aircraft industry, they not to make intersection between plates of their models, like when we modeled the building in Robot, or other [INAUDIBLE] you normally have to resolve intersection of this plate and this wall. And it brings a lot of problems. And it's not a nice feature of the finite elements.

In aircraft industry, they do not do not make this intersection. But they connect these plates by so-called "interpolation constrains." So, there are some specific links between them, between nodes. The big advantage of this is that we can mesh each plate separately. It's very easy, very fast, and the mesh is quite nice. And the time of analysis is much faster. And the results are better.

The bolted connection is not rocket science. There are two plates in contact and several bolts. It's nothing special.

But we need a good model of these bolts. In our method, to use quite a simple model of bolts, only a simple spring in one direction-- but if it's a specific force, the formation diagram. And this is [? proven ?] by real tests of bolts in labs. And all, again, these bolts are connected with this interpolation constrains.

Another thing is behavior of bolts in shear. This task was a little bit more difficult than the previous one. But we resolve it with a specific model of connection between this bolt spring and the steel plate. We do it with some contact elements. And the result is like that. You can see that when the bolt is in pressure onto the edge of the hole, only here we have pressure. And the opposite side is not loaded at all.

So then, the results of this model is quite nice. And this, we call it our masterpiece, that our blades on the top flange and bottom flange, we can see that the one plate is in pressure. This bottom in tension. And you can see where the bolts are working with the steel plates.

Another piece of story are welds. You can use two models. One is elastic, where we have only these interpolation constrains. And this method, which I described on the slide with mesh, has another advantage then-- when we have these constrains between two plates, you can directly have all forces in these constrains. And we can use these forces directly for design of weld.

So here, you can see the distribution of stresses in this weld. Recently, we came with even better solution for plastic model of welds and we model welds with small finite elements, as well. And then, the results are even better. And the behavior of weld is described very, very precisely.

Interesting feature of [? other ?] methods are contacts. I spoke about them with some bolted connections, but it can be used also in parts which are not bolted, like here. Now you can see that there is only pressure in between this lack of pleat and web of general section. Or how we can calculate such connections, which are used for secondary structures. Like for roofs, for instance.

Of course, we do also anchoring. We can calculate standard-based plates, but also very, very complex and strange columns, and these base plates. Or we can calculate anchoring to the walls.

When we came with this method, it was clear for us that at first we have to convince you that this method is useful for you and reliable. So, we had verification and validation of this method with universities. At first, we compared simple shapes with solutions which are known in design guides and the code. And then we made comparison with these advanced models, again, finite element models, which are much more precise than-- maybe not much more-- but they are closer to reality. Models where all bolts are modeled in this way. And compared the results of this too.

These models are known, but, in fact, we cannot use them in daily practice. It's the new one to model your connection with these elements, you need programs like [INAUDIBLE] or Abaqus, Nastran. And it takes days or weeks to calculate one connection. So, we use it only for verification of our CBFEM model.

And we also made the real tests in labs. We work together with two technical universities, and also some professors from Eurocode society. All this verification are summarized in the book of Professor Wald. It's here, this book. So, this verification is proved.

But maybe you'll say that it's too, too much science, it's too, too difficult for you. And it is true-- all these things which I described are quite enhanced and sophisticated. But it is, in our software, it is hidden. It is in the background. And the software is like detailing product, detailing software, where you compose, even so difficult solutions, you compose with the tools and method, in fact which are used in the real manufacturing process.

So, we say that the product is composed with manufacturing operations, and these operations are like [? god-like-- ?] stiffeners, like wideners, some [? ribs, ?] and plates, shifted end plates, this one, gusset plates, cleats, fin plates, all these simple things. But the difference from standard software is that you can compose them together in one model. So, you create this model by these things which you understand, and all the science is done automatically in background. You do not work with the machine and with its finite elements. You only see the calculated results.

And another thing which is very important for using of the software it is to put this software into your design workflow. Here we example of the link between Robot and our IDEA connection. I can set this on, on video.

This structure, made in Robot, you select only the node. Then call the plug-in IDEA Connection. And then fill some data. And then the data transferred, it's like, BIM solution. You will see this model, this joint that you select, you'll see here. And this moved into our program.

In the beginning, it is without these operations because this is like what you calculate in your frame model. But we have all the internal-- needed internal forces, or load cases. And then we can do the design with these manufacturing operations.

And you add step-by-step-- like here, end plate, the bolts, so you define precise position and precise size of end plate with position of bolts. And then continue with these connections of other elements. Finally, we have this, what is now a reasonable shape of this connection.

As I said, the analysis is done automatically. You don't define dimension and the things. And then we calculate it with our [? FEM algorithm. ?] And then we will have all our results, like in finite elements programs. It is one type of results which you can get.

And so, you see, in fact, weak parts of your connection. You see where the steel in the plastic status. You can see the forces in all bolts. Or here, there are areas of plastification. And we also, at some specific presentation of results, which I can show in the next example. And finally, we have results where all plates are checked, all bolts are checked, all welds are checked, according to formulas, according to AISC code or Eurocode.

So after this video, we will continue with some examples of real-life projects in our models. So, you can calculate stress and strain in the joint, but we can analyze also local buckling in connections, like here, of this gusset plate, but also on the webs of profiles. We can analyze stiffness of correction. And we do the overall check of the whole joint.

As Juraj said, we started with this product more than two years ago. And we have a lot of examples from our clients, from engineers worldwide. And some of them are very good, but we came across also with some examples which were designed in a very bad way, and where we have, in fact, helped engineers to make their design safer and solid.

This is example of electricity tower mast, it's from Poland. On the original design, you see that the whole plate is plastified. There are some red parts. And this detail was not satisfactory, according to code. And we proposed to rearrange position balls. And you see the difference between these two designs. It is the same amount of material, but the result is much better. For the same money, you can get much, much better-- better and safer joint.

This example from Germany, again, they made the design, which was good from the point of view of stress and strain. But after a buckling analysis, they realized that this web is quite thin, and it can be dangerous. So, we can simply add some stiffness. It is quite normal to put there the stiffener, but when you do it, you normally, in fact, estimate if it should be there or not. But with this method, you can decide, yes, I need it, or I do not need it.

This is about stiffness. It's from our country. We studied such a joint, this crossing of diagonals. And again, when we analyze it from the point of view of stress, it's safe. But you see what this whole section is doing. And there is no ovalization of this tube.

So much better is when you would do it in this way. And we calculate stiffness of this, and this one is 10 times stiffer than these ones. So again, for the same money, you'll get the much, much better-- much better part of your structure.

This is from Heathrow Airport. Our partners from UK made the design of the bridge in the Heathrow extension. And there were connections like that, which are not possible to design with standard guidelines or tools. So, we use our method, and the design was successfully done.

And now, there is the turn of Graham. And he will tell you about how they use the software in Arup.

GRAHAM ALDWINCKLE: Yup. Thank you. So, I've been in Arup now about 22 years as a structural engineer. And I've worked on a couple of these projects, in fact-- the Leadenhall Building in London, and also the High Roller here in Vegas.

And these steelwork projects have one thing in common-- the odd one out is the Opera House, by the way, because it's concrete-- but the steelwork here is on show. It's visible. It's architecturally exposed.

And in most cases, we don't get involved in the detailing of a connection as an engineering company. However, it is very valuable to us to know that what we're presenting as a structural solution is buildable, efficient, optimized, et cetera. So, it's really important to us, whether we're designing the connection or not, that we have gone through those sorts of analyses in our design, up front. So, I'm going to give you an example on the Leadenhall in a second.

So, what we were looking for in our tools-- and this is to spread across the Arup world, in giving our engineers the right tools for the right job-- is it a tool that allows for this workflow with interoperability with the various other methods and analysis tools that we use? So, we'll hear a bit more on advanced steel coming up from these guys. But looking at getting information out of one model and putting that into this tool to design the connections is a really valuable part of our workflow. Because the more you can make those things efficient, the more it gets done, and therefore, the better the design.

We want it to be simple to learn and easy to use. And that's pretty fundamental. But not too simple that it's overly-used because we need the engineers to know exactly what they're doing in the tool, as well.

We need it to be supported by code. Code-based design is good. And by validated theories-- so as the book that's been put forward here demonstrates, it has had a lot of validation. And we've been doing so internally, as well.

And we also want the tool to be usable at the full spectrum of the design stages. So that's concept design to do some checking, all through to the detailed design, when it's part of our deliverable, as well.

So this is one of the Eurocode aspects that it's actually based on. It's one of a few, actually, that it's based on. This is Eurocode 3, part 5-- design of plated structural elements. And there's also a part 6, which is shell elements.

And so, it's doing non-linear design in the tool. And that's what's demonstrated on the right, there. But I also wanted to highlight the note. I don't know if you can read it. The guidance is intended for engineers who are experienced in the use of finite element methods.

So I'd recommend this read. This isn't particularly long, this Eurocode. And there is a similar one in the AISC.

So the Eurocode Annex C gives a really good background to what's going on in the tool. But the theoretical background that's available with the software also covers it. It's just good bits to read.

So, as an example, 10 years ago the Leadenhall Building here-- this was one of the nodes, which was about three meters high, top to bottom, on a floor. And the workflow for this was actually quite complex and convoluted. And it didn't lend itself to the iterative process that we like to improve upon.

And what happened in this particular project, with various models that started off in tech. They're very basic. We had to mesh it, put it through Nastran, and understand a bit more about the topology of the connections and trying to optimize it. Now, that is quite a complex workflow, a long-winded workflow. And we were using Nastran analysis, but also LS-DYNA, which is one of our tools of choice for this sort of output to look at the weld design.

So, we were responsible for the plate sizes, the buildability, not the cost-- but we were still interested in that-- and the weld sizes. So, the process here, for one node out of quite a few hundred had to be done for each one. And as much as you can try and automate that, it takes time.

So, and this is one of the outputs, as well, which is a plot of the weld forces along the length of the weld for X number of load cases. So, quite an iterative process and fairly long-winded. But now with this tool, we've been looking at our improved workflows.

Because, for example, this is an LS-DYNA plot of the stresses of the connection for one set of forces. And we've taken this element here-- and I'll explain this in a second-- and run it through the StatiCa connection tool. And once the inputs are, in fact, slightly different between the two, so they're not directly comparable, it has been an immensely valuable exercise to us to compare it in a tool that takes a fraction of the time.

And also, one of the benefits is that you can run many load cases in the tool. So, one of the things that engineers grapple with is, do we provide a set of forces that are an envelope of forces? And that usually means that they are non-coexistent, as well.

So, unless it's a particularly simple structure, one of the problems is by providing fabricators with one set of numbers, we think it makes life easier. But of course, it doesn't. And the fabricators then struggle to make a connection work. So the ability to run through many load cases is a really, really valuable part of this.

And this particular exercise was a ring beam, supported on a strut here. Not a good image of it, but anyway, hopefully you can picture there's a strut here, which is this one. Now, it turned up to site with no stiffeners. And it turned out the connection design hadn't been done. It was actually just taking the design intent.

So we got involved to do some detailed analysis. And of course, with an envelope of forces, it didn't work because it was trying to capture lots of [INAUDIBLE] cases. And it just didn't work. So, by looking at each individual load case, we were able to justify that the design was fine. A highly valuable tool, we were both using LS-DYNA and StatiCa.

So, here's another example. One of our engineers was called by a contractor to try and help understand a problem they were having on the site. It wasn't our site.

So, what happened here? This was the connection we were looking at. And this top cord here led to a cantilever on the right. And they were finding that the cantilever was bowing a lot.

It was stretching because, well, it transpired is that-- this is the blow-up. This top cord, which was in tension, was being pulled to the right. And so, other engineers had a look on site. And it became very obvious what the problem to us was, which was the connection just simply wasn't stiff enough.

Let me explain a bit more. This stiffener here wasn't, in fact, welded to the web of the [? EZ. ?] So the load path to get to the brace here was just not stiff enough. So, we built the tool.

And by applying a few of the simple manufacturing operations, we built it to be as-built. And that process takes about 20 minutes, not very long. And then, we're able to run it. And so, analysis showed that with the forces, all the plates were, in fact, fine. So it didn't have a problem for strength.

And in fact, typically, when you look at these sorts of connections, you look at one connection at a time. So this might have had an engineer look at the connection on the right, another one on the left, another one for this brace-- a set of separate calculations, perhaps. But this tool allows it to be combined.

And so, when you then look at the deflected shape profile, you can begin to see that there's a lot going on in this connection. This is a PFC, a Parallel Flange Channel. And the gusset here-- let me go back a bit-- the gusset is on the middle of the bottom, which is the web. So it's not very stiff at all. And you can see that it's pulling down on that.

And also, the end plate on the right, here, has bolts at the bottom. So it's being pulled apart again. And structurally for strength it's sound. But clearly on stiffness it is no good at all.

And this is the stiffness plot. So the tool allows buckling checks, strength checks, and stiffness checks. And we were able to show you that the stiffness was just nowhere near where it needed to have been. And so, had the engineers and the original designer in the original design looked at or thought about this, they would realized very quickly that it wasn't able to do the job it was actually there to do.

And another example here of a braced element, which has some stiffness inside it. And so, we use the tool to try and justify no stiffness. We weren't able to get that far, in fact. But we were able to half the stiffness, the thickness. So, it's a tool to help optimize and minimize the amount of fabrication.

In an ideal world, you'd try and do it without stiffness at all. And as part of an engineer's tool kit, optimization of members has to include optimization of the connections. And so, there's an output from it with the stress results.

So, further examples of our use-- this is one we helped with to develop the connection with these guys. And that's tubular sections, but with bolted splices, as well. So, amongst all of this, there's a lot going on in a particular connection. And it's really helpful to understand where the failure modes might be, or where we can optimize it further, if the wall thicknesses, stiffeners, number of bolts, that sort of thing.

Here's an example we have. It is hot of the press this week. This is a cord of a massive truss of a stadium. And we're trying to design the connections as well in StatiCa. And so, I don't have any of [INAUDIBLE] yet, but that'll come.

So, this is all tubular sections, welded. And as part of the work we do, and we're looking at splice connections for it, as well. Where the optimal splice locations?

But, what's next? So, when I see engineers with drawings like these-- this one and this one-- I'm thinking, how much have they thought it through? And we like to show, or ensure, in our jobs that we've thought that sort of element through to the full extent, so that we're not providing a detail that's either unbuildable, unoptimized, et cetera.

So we put the effort into doing that, as engineers should. So, having a tool that really helps in that workflow when you can get this sort of geometry out of the model-- be it Advance Steel, or Revit, or Tekla-- then, that really helps us home in on what that optimum solution should be. Have we got the right number of bolts? Are they in the right location? Even, is an architect going to be happy with the end result? Because if it's visible, they have to know what to expect because often they only see the fabrication drawings. And then it's too late.

So, just a few examples there of our use of it. Happy to talk further. But going to hand over now to you to talk about Advance Steel.

JURAJ SABATKA: Thank you, Graham. Let's jump into workflow with Advance Steel.

We integrated it to a level that you launch a project, and you go to command-- it's called Concheck, that activates the link. And you simply select the node you want to forward, then the beams surrounded, and just all parts of [? joints. ?] And select all three levels of the information. And then it's automatically exported to IDEA StatiCa connection. It will create a new project. And the file is opened.

So, we can skip the geometry part and the design part in IDEA StatiCa. You just need to input loads to get the full picture for the analysis. Usually, we get the data from another program or from an Excel spreadsheet, so that's what we're going to do right here-- get the load cases in. And we're ready for the analysis of this joint, in pretty much a couple of seconds, as a majority of the work was done in Advance Steel.

So we run the analysis, taking into account all the iterations. And we get the results. We clearly see that all the bolts, plates, welds of this particular joint are OK. We can go through the checks of the individual components and potentially see reserves in the design. We can visualize the results in more ways to create the output report, as we want to do.

And during the whole process, we always know where we are. This is the sequence of how we gradually load in a joint. I will play that again. So we gradually load it up and see the plastification zones develop.

And it's quite easy. Green means passes the check. Red, it's not safe.

We've prepared also a quick example on the bold optimization. Lubos, will you go out and guide us through that?

LUBOMIR SABATKA: As we explained, we can analyze any connection of any shape. But we can also optimize the parts of these connections. We can decide where you should put stiffeners, where you should put [? ribs, ?] how many bolts you need, if all bolts are efficiently used, et cetera.

In this example, you see the specific colors of the connection. We call it a traffic lights presentation. And it's a simple idea. What is good is green. What is close to the limit, it is orange. What is over the limit is red. If there is something in gray color, it means that it is not loaded too much. And maybe it's not efficiently designed.

This is an example of a rigid joint-- bolted and the typical shape recommended for seismic regions. And we calculate it. And you see that some bolts are green. So then they are used enough. But there are also some gray, which are not used enough.

Here on this picture, you will see the forces in each bolt. And in this corner bolt, there is almost no force because this plate is not too much stiff. And so, the bolt cannot take too much load. So, it is the first step.

Second is that we try to get rid of these bolts in corners. You can do it with specific tool, where we can edit positions of all bolts, or delete some bolts. So, and the result is like that. And you see the checks is still very, very good.

So next step, we can delete also bolts in this second row. You see here, forces where all bolts now have almost the same force. But again, it satisfies very well. So we tried delete also these ones.

And this we should not do. You can directly see the red color in the presentation. So it means that these parts are now designed not safe. They do not pass the checks. So you can see red bolts. You can see this widener, also some welds are red as well. So, the good design was this. And this is too much.

So, we can continue with Revit. Again, a video. This is our last development. When we linked IDEA StatiCa also to the Revit, it is similar link like with Robot. The difference is that in Revit, there are already-implemented objects called a connection. And so, you can define your connection in Revit. And then you only click on this object.

And data are sent to the IDEA StatiCa, as we showed before. You will see all load cases, all combinations. You can define your combinations if you want. Or just take them from Revit. The rest is the same like for Robot.

You see your joint in the IDEA StatiCa. And again, the manufacturing operations are added here in downloads. Manufacturing operations and plates are here. You can see all of it. You can do these details, like dog bone, and dog bone details for seismic joints.

And again, traffic lights presentation, you'll see that the design is OK. There are some orange bolts, but they still pass the check. Here you can see a place of plastification or stresses in the members. And we generate output report, what you already saw in previous examples.

Now, we are close to finish of the presentation.

JURAJ SABATKA: Let me go through a couple more things before we go to Q&A.

We also working with Autodesk Forge, the cloud platform to share the [? mold ?] data. This is the preview from three weeks ago, when we were in the Forge Accelerator in Munich. It's actually the pilot joint push to the Fusion API. And this will help us to generate precise drawings that can use the cloud engine of that.

And once the Revit moves to this Fusion API, we will link it directly. So you will be able to take any joint from Revit, design it, and send it back. So this is what we are working with Autodesk right now. And hopefully it's an invitation for the next year class.

Please give us feedback. Let us know. And allow me to recap.

This is a thing to do design and check joints, any topology, any loading, according to the selected code-- and do it in minutes. And part of your workflow you already know with the programs you already have. Try this out. It's free for 14 days.

Visit our website-- idea-rs.com. Or just exchange business cards with us after we're finished. And you will see it by yourself.

First part of the Q&A will be a little competition and a question from us. We will project a question on the screen. And whoever feels comfortable to answer it, please raise your hand. And the first one to do that will receive the book we mentioned about the verification and validation. All right? So let's do that.

How would you compare steel joint design using Component method and CBFEM method? What is different? What is the same?

So who will be brave enough? The book is right here.

[LAUGHTER]

It's very close. [LAUGHS] All right, let's go.

AUDIENCE: I'll take a shot at it. For the joint design, what you did look at each component. I can use two examples of that. One of them was that [INAUDIBLE] you get piece by piece and try to build [INAUDIBLE]. That's the similarity when you [INAUDIBLE]. The other one was-- the example tha [? Arup had. ?] It was [INAUDIBLE] But that was the same. So may want to compare, one is [INAUDIBLE].

JURAJ SABATKA: So the difference is using the [AUDIO OUT].

Let's give it a--

[APPLAUSE]

Congratulations.

[APPLAUSE]

Any questions? Yes? Does it work?

AUDIENCE: It does. Yeah.

JURAJ SABATKA: OK, good.

AUDIENCE: So, question is scripting capabilities in the tool, or you have to go connection by connection? Or is it something you can script?

JURAJ SABATKA: [AUDIO OUT] joint. [AUDIO OUT]

--from Advance Steel or Revit. This is the [INAUDIBLE] topic for the future, to script it.

AUDIENCE: The envelope that you're using, is it a real envelope, or are you just looking for the highest forces for the connection in each direction?

JURAJ SABATKA: We're looking for the extremes and take--

LUBOMIR SABATKA: Extreme forces, yeah. In all [AUDIO OUT].

GRAHAM ALDWINCKLE: Any other questions? Come on over.

AUDIENCE: Hi. Great presentation. Sort of a question on what is the output of that software in [? calcs? ?] Can you actually export back to Revit or to Tekla? I just wasn't sure what the actual output was.

JURAJ SABATKA: The output is the actual output report with the schemes of drawings and all the checks and references to code. Pushing it back to Tekla, not yet. It will do that with Revit because it requires to recognize the macros on the way back, which is slightly more difficult. And what really was the main thing for structural engineers and fabricators was to do the checks first. So this is on the wish list.

AUDIENCE: Thanks.

JURAJ SABATKA: You're not alone.

GRAHAM ALDWINCKLE: The output report though, it has options to make it more detailed or less detailed as a PDF. Question?

AUDIENCE: I had a question. Do you have to use [INAUDIBLE] on custom families, or you do that just for--

JURAJ SABATKA: We can do that on any joint model in Revit, using--

AUDIENCE: With any steel type of-- even if it's like a custom family that you do on your own that's not coming from Autodesk?

JURAJ SABATKA: If it is designed as an object in Revit, well, as you see, it will be released in April because we have annual releases. But, yes, what you see, the workflow, it will work for any joint.

AUDIENCE: Because they presented for the 2017 that they have a section [? in them. ?] And right now, I think they're having 11 or 10 different types of sections in it.

GRAHAM ALDWINCKLE: 22 types, I think.

AUDIENCE: 22? OK, so what do you do when you have 23?

GRAHAM ALDWINCKLE: You can build your own section types in the tool. So you can make it a fabricated section of any shape. And they have a few functions to do that.

But in terms of importing the data, I guess that's a work in progress. But you can import the Revit families now that exist. So if you have the family types that are a member, that can come in-- the connection topology that may not exist in Revit-- it doesn't take much to build it in StatiCa, so that you can do it in that order.

Any other questions? Any volunteers? Anyone at the back?

AUDIENCE: Thank you. Actually, the first question I have is for you, Graham. The comparison you did between the Nastran and the StatiCa, the stress that you were using was principle stress?

GRAHAM ALDWINCKLE: Mm hm.

AUDIENCE: Or-- it was. Because the Nastran modeling seems to have more stress concentration?

GRAHAM ALDWINCKLE: So I had two tools being shown there. One was Nastran. And one was LS-DYNA. And the LS-DYNA plots have [? all these ?] stresses. And that's why it's not directly comparable to the equivalent stresses. And the Nastran tool diagram was only used on the Leadenhall example I showed, the node. And so, Nastran was used to look at plate sizing, and then LS-DYNA was used to look at the welds themselves.

So the equivalent stress that's output from the StatiCa tool, maybe you want to explain more about what the equivalent stress is? So you have various options of output. Equivalent stress is one. And you can look at the plastic stress plots, as well. The equivalent stress is defined in the literature. Do you want to explain more about that?

JURAJ SABATKA: Yeah. And you have an option to even shape the barrier. So, normally it's hidden, but you can have all the properties of the stress to adjust for analysis purposes.

AUDIENCE: Because in software, usually you can look at different type of stresses, just like on [INAUDIBLE]. And some of them fail. Like, for example, when I look at it, the braces-- it was more red, as it shouldn't. Probably you look at the stress, which puts tension on that. That's where it's compared. But you did a comparison, but-- which is great. I think it's a very useful software.

GRAHAM ALDWINCKLE: In my comparison, the direct comparison was not directly applicable because the input forces were different. So I had an LS-DYNA plot on the left of my diagram and then the StatiCa plot on the right. And they weren't showing the equivalent stresses. And they didn't have the same input forces. However, the end result was the same.

AUDIENCE: Got it.

GRAHAM ALDWINCKLE: They were showing which load cases of the 260 were failing and how much they were failing by. And they were very broad agreement for that and showing the stress plots-- that the same was not possible because I did have different input.

AUDIENCE: Got it. Got it. And then, the second question is that you mainly use the software for analysis aspect of it. Do you then use any output to develop the drawings? I think it was some question out there. But I didn't get a clear answer.

GRAHAM ALDWINCKLE: Yeah, so we haven't taken the output and put that back onto drawings yet. But as a direct workflow, we would then have to draw those or model those in Revit separately, until that functionality is available-- which would be good. But we have to know that what we're drawing and presenting to a fabricator is buildable and optimized. And so the tool certainly allows you to do that half of the equation.

AUDIENCE: Thank you.

JURAJ SABATKA: Thank you, all, for coming. And have a nice rest of the AU.

[APPLAUSE]