Precast Concrete Shop Drawings and Fabrication with Revit—Process, Content, and Strategy (Repeat)

ERICH BRETZ: All right, guys. So how about if we go ahead and get started. First off, I want to give you a little bit of information about who we are. My name is Eric Bretz. My background actually is in engineering consulting design.

I graduated from the University of Illinois with a master's degree in structural engineering. And I worked as a consultant for about 10 or 12 years, just as a typical structural engineering design consultant working on large commercial projects. And then about six years ago, Dan McCloskey and I spun off to form MB BIM Solutions where our primary clients are general contractors and especially subcontractors, like precasters, which is what we're talking about now.

STEVEN RICHTER: Good afternoon, guys. Thank you for coming. My name is Steven Richter. My background is primarily in architecture. I graduated from the University of Colorado with a degree in architecture in 2009. Held a few jobs in industrial design and structural design. Then I joined MB BIM them about three years ago, where I've been using-- don't know-- just Revit and in general BIM software. Been using it Revit for probably about 12 years now.

ERICH BRETZ: So how about you guys? Maybe can I get a show of hands, how many of you guys here work for precasters or maybe consultants to precasters? OK. So a handful. Good. How about general contractors? OK. More of those. And then structural engineers? OK. Great.

And other is the rest. So, good. That's really a pretty good mix. Nice. That's the wrong way to go.

So let me give you a really quick rundown of what it is that we do at MB BIM. I think it will help sort of frame a lot of the things that we're going to talk about here. So at MB BIM, we do a lot of fabrication modeling. So we do, OK, Rebar shop drawing, structural steel shop drawings. We do cold form, create concrete line drawings or lift drawings, and of course precast concrete shop drawings.

We also just kind of use BIM to solve general contractor-type problems, do a lot of Navisworks coordination and 3D coordination. Custom modeling, just a lot of weird stuff. Custom modeling kind of becomes like a constructability study type stuff, where we kind of model things in a very detailed manner and then use that to solve construction problems.

We're getting into VR. We've done a handful of VR simulations. And also we do virtual mockups and constructability studies. There's concrete lift drawings. We also use 3D Studio Max to do concrete. I'm sorry, for construction sequences. Then we do a handful of unique things. This is actually a head station inside of the world's largest gold mine, which also happens to be the world's largest copper mine.

We kind of use our engineering backgrounds to come up with an erection sequence for how do you build this little thing which ends up being roughly the height of, well, just this part of it isn't. But the thing is about the height of a 10-story building two miles deep into the ground.

So, generally speaking, our contractor clients kind of look at us as good problem solvers. And so we BIM to solve problems.

I'm guessing you guys have read the summary and learning objectives. And so on in terms of what we're going to talk about here about precast. It's kind of funny. I think my plans for this presentation changed a bit after having heard the opening keynote, which I hope you guys all had a chance to attend. Plus there was another presentation yesterday where someone from Europe was talking about automating the precast concrete layout process.

And I guess there's a couple of key messages here. First off, that's not how things are done here in the US for various reasons. And we can go into detail about what that all means and why. In a sense, really what we're doing-- and sort of the state of the art, so to speak here in the US-- is simply using 3D tools to create 2D CAD shop drawings that used to be under the old system.

So, for better or for worse, frankly, that's kind of what it is. And that's really a lot of what we do. And so a lot of the things that we're going to talk about I think that you might be asking yourself hey, this is great. But have I really improved the process here very much? And, unfortunately, I don't know. It just isn't the case.

On the positive side, I think there's hope. I'm guessing that a lot of you guys are familiar with the Revit 2018 tools and how they're sort of changing the way that the Revit does precast. And they're automating a lot of those tasks in particular with the piece drawings and the layout of walls and slabs.

And that stuff works great over in Europe because in Europe they do a lot of wall and slab precast concrete. But over here in the States, we're using columns, double T's, beams, those sorts of things. And just doesn't work over here.

So the process that we're going to present is going to appear to be as-- it's different than what the 2018 tools use. On the positive side, the 2018 tools are just the start of the development of that sort of mentality. And I think in the end the sort of ideal system is going to be sort of a hybrid of what it is that we're doing and what the 2018 tools are doing.

STEVEN RICHTER: Yeah. I think a good way to think of it is the 2018 tools from Revit kind of utilize walls and slabs and the segmentation of those walls and slabs into precast parts. This system is primarily focused around structural framing elements. So these are individual elements that exist unique to every element around them. And then those become your precast pieces.

ERICH BRETZ: So on the positive side, so you might be sitting there saying, hey, this is great. So what are we doing here? We have heard from our precaster clients that there is some very distinct advantages to using BIM processes. This is the first of them there is a reduction in NCRs. We find that producing precast pieces in Revit actually truly is an enhancement in quality.

Better communication tools to the guys in the shop that are actually fabbing the pieces, as well. The next two are probably the biggest advantages, I think. The enhanced coordination with the designers and other subcontractors. Using Revit and BIM processes actually just makes that sort of stuff possible. You know, pre-Revit, if you've got architects that are designing in Revit and then you're building your precast model in CAD, it's just really difficult to coordinate those two and to make those two things come together.

Along those same lines, we're actually seeing a lot of-- in the Denver area at least-- a lot of buildings in which the precast model is being populated at the same time as the architectural model. So in that regard, the precaster Castro kind of has to look at themselves as though they're a design partner.

So that functionality is unlocked. And to be quite honest, from a precaster standpoint, these guys are selling speed, right? So it just basically makes that possible now.

STEVEN RICHTER: I think something kind of along those lines that wasn't up there is it's important to keep in mind that the model is essentially a universal language. So it's a precast model. But it's a language that perhaps a steel erector can relate to.

And we actually saw that because we had a client who needed an erector and didn't have a precast erector available so they have to go to a steel erector. Well, you show them a model, they get a much better idea of what they're actually getting into.

ERICH BRETZ: And the next one, the piece to piece coordination. This one, we've heard that this is dramatically improved. When there's two pieces that come together that maybe they're supported by each other. With typical 2D CAD tools it's a lot harder to make sure that the-- because of the fact that your plans, your erection drawings, are detached from your piece drawings, it makes it easy for things to get jammed up. For maybe a T to get a little bit longer. For a spandrel beam, for example, to be a little bit long and then shipped out to the site. And then, holy cow, you realize that it just doesn't fit.

So that piece to piece coordination is dramatically improved. And then we've also heard that the material tracking is a lot better, too. So the ability to generate quantities of materials and to get material ordered faster is substantially better.

In terms of keys to success, I hate to say it. We're back to the 2D CAD world again since we haven't fundamentally changed the process. So a lot of these are going to be really similar to what you see in the 2D CAD world. I'll try to give you a couple pieces of information about how we accomplish these things in the Revit world. In terms of communicate, we always try to over communicate.

I don't know how many times, it just seems like that's the most common problem that you hear about, is that we just didn't communicate enough. So in terms of communication, we're providing our model and making it available to everyone in the project, providing a series of PDF erection drawings to the entire precast crews early on in projects. Making sure that they're looking at them as much as possible.

Want to reduce the amount of rework, minimize changes. For that, we've kind of got some very specific ideas about how does it that you ask for information, get information from the relevant people, in the right time frame. It's a little bit different than in Revit. And we'll talk about the warping problem in Revit and how that changes the amount of rework that you kind of have to do. And we'll talk about that in a bit.

On the positive side, using datas and schedules to drive content and quality control, this is something you definitely have to do. We use schedules to, in a sense, check the model without actually looking at the model elements. So when we're doing sort of quality control in our end, we're looking at the pieces, of course, in sort of an overall sense. But then also looking at the data that's in the model to make sure that things make sense.

And looking for relationships between numbers of plates. You know, if you have a specific plate where maybe this plate here, where every time that the plate that's cast into this wall column, every time that you have one of those, you should have the other one that's cast into the beam, making sure that those quantities match up one to one.

Just really simple things. But since you've got an overall global view of the project, you can do those sort of things.

Standardize, this is obvious. Everybody in the precaster world needs to standardize. For this, we're just using tons of legends, schedules. Trying to make that stuff, make data as repeatable as possible. And then automate, looking for tools that will actually perform some automation. We talked about the Revit 2018 tools, which are moving along. There's also a handful of other third party tools that are out that we'll talk about in a bit.

And then quality control. Something that we've noticed with our precasters jobs is that that quality control has to be done really on an extremely regular basis. And the primary reason for that is that we've found that models that are developed properly and are maintained throughout always end up performing much better when it comes to the piece drawing process than those that are not, that we don't perform quality control on. So you really set yourself up for success if you do a great deal of quality control.

OK. Laying out precaster framing. This is sort of the start of any one of our precaster jobs. So we're kind of tracing through this little process down here in the corner. The first thing, this is obvious. You kind of have to establish the grids, and the floor to floor heights, and the building envelope. Trying to think about in the Revit sense, what does that mean? This means passing your model on to the architects and engineers.

In an ideal world, I think the precaster model would essentially eliminate the need for the engineer's model. But the message here is simply that, as a precaster, you really need to make yourself a part of the design team. As the precaster, you're the person that's there with the most information about the precaster model. So use your model in that regard as though you're just another designer.

Typical details, typically for jobs, you can handle 90% of the details on a job or 90% of the connections in a job, with a handful of details you're going to use over and over and over. So we typically present our precaster client with a series of typical details that they then will sort of run up and down the chain to make sure that everybody's on the same board between for what's going to be used.

And these details go a long way towards sort of defining the typical plates that are going to used, the relationships between pieces, elevations, how pieces kind of connect together. So that's the goal there.

Erection sequence and crane location. This is a critical one, obviously. But, again, we haven't really improved a lot on the 2D CAD world here, with the exception of a Dynamo script. There's a Dynamo script that's available by Dieter. If you guys know who-- maybe you've seen the name Dieter. He's over in Europe and he developed a Dynamo script that takes as its input the weights of all the pieces. And then also it runs a generic algorithm on the location of the crane.

And it basically makes sure that every piece can be erected using the crane's load table from its position wherever it may sit on the site. And then it'll move the crane and it'll see, OK, how many pieces can I not erect here? And then it'll move it again. And it'll find sort of the optimal location for the crane.

The thing that we're trying to avoid here and the rework that we're trying to avoid is simply the relaying out of framing and then also connections. Precast connections are a little funny in Revit. Precast connections do have a relationship to an element. They're hosted to an element. But since they're connecting two pieces together, they don't maintain a relationship between both pieces.

So if one of the pieces moves, you're not guaranteed to get the connection to move with it. And remodeling connections and the framing that goes along with them is very problematic for a precast model.

Slopes and drainage. This is along those same lines. It's really complicated. It's really challenging for us to try to Rework. We'll talk in a second about the warping problem and how that's handled in Revit, but the real key there is just that when you warp framing, you have to not only modify the T's that are warped, but also the connections that support those T's. And rewarping that is an extremely time consuming process that we only do want to do once.

To accommodate that, we typically send through a series of plans to the architect that simply has highs and low points of garages. And them to approve that before incorporating those things.

And then panelization of vertical elements, this kind of goes back to the crane location problem, just making sure that those things are ironed out.

To accommodate all these things, we're just constantly keeping drawings in front of everyone, the erection crew, the production people, trying to make sure that everybody's on the same page in terms of what it is that they're going to get from a precast standpoint.

STEVEN RICHTER: I think with panelization, a couple of key things you want to keep in mind, just make sure that you're on the same page with the precast, as far as what their shipping requirements are, their weight requirements, their clearance, and their accessibility, as well as their erection department, with whether or not they'll be able to even access the plates from where they're at. Changes late in the game to the layout of your panelization are really going to cut into your budget.

Top in form, bottom in form, we send through plans like this to the architect and to the production people so that everybody's on the same page in terms of which face is going to be top in form. You can kind of see there's little red dots all over the place. There are notes that say this face is going to be top in form. The architect at that point knows what finish they're going to get on that face.

I guess I should've said to the surface finishes. So they know what they're going to get on each of those faces. They know that those faces are going to look just a little bit different than those that are on the form side of things.

Critical dimensions affecting primarily elements, things like building envelope. The thing that we're watching out for here is the fact that if you do change your edges, slab details, it's really common that that affects all the framing in that bay. And then oftentimes the framing in the next bay over, too. So really just want to minimize the amount of time that we're spending doing that.

And then cambers. You know, I think Steven I consider and talk for a day on cambers. I think the key here is just to keep those things in mind that you will have camber and that's going to eventually affect your model. And you need to consider that when laying out the plates and detailing it.

And finally, formliners. Because now you have all this data in the model, formliners-- I don't know if you guys know a whole lot about formliners, but these things are super expensive. They're heavy duty, and they're super expensive, and they're reusable.

So it kind of takes some strategy and planning out when it is that pieces are going to be poured. How many pours you can get out of one formliners. And you want to basically use as few of the formliners as you possibly can.

ERICH BRETZ: OK. So the warping problem How many of you guys here are familiar with the warping, or what warping is and why it's a problem? Only one? OK. OK. Well, we got a handful. That's good. But big picture, the problem here is simply that a double T in terms of its lifetime spends the first part of its life as a flat element When it's installed, and then in its final condition it's warped. So it's cast flat. It's stored flat. Shipped to the site flat. But then as it's installed it actually warps to the form of the supports that are supporting it.

So you see here what our implementation of this is. We use the Edge for Revit tools and what their implementation of it is. So in a sense because it does have sort of two lifetimes, since we're using the Revit model to create our piece drawings, what we have to do is we need both of those geometries. We need the flat geometry for when we're making our piece drawings. But we need the work geometry for when we're sort of laying out the embeds along a wall.

You need the warped geometry for when you kick your model over to a general contractor so that they can adequately perform their 3D clash detection. So we need both pieces of geometry. The Edge implementation allows you to control both of those geometries with one family. With a single family. It's essentially a container family. And that family drives the overall dimensions of the double T. It drives the length, the widths the layout of the legs. That sort of thing.

Since your flat geometry is going to be used for the installation of the pieces for your piece drawings, the embeds are shown on the flat geometry. So you can kind of see here, this is the flat one up top. The higher one. And then the lower one down here and down here, that's the warped one.

So the plates you can see in red here, the plates are all inside of the flat geometry. And then the warped geometry is going to drive the supporting elements, if that makes sense. So, in a funny way, there's an issue that kind of happens here. And it is that this plate here that's cast into the T, this plate here that's cast into the wall, actually are going to be attached to each other using a piece of erection material which is sitting on top of that.

And so, in a sense, as you're just sort of sitting here eyeballing it, it's a little hard to see, oh, yeah. I got that. I got that right. So just this whole process sort of makes this thing harder to QC. I think it's interesting to note that Tekla has an implementation of this, as well. So Tekla does a similar thing to what Revit can do in terms of creating fabrication models and drawings.

And their implementation of it actually involves unwinding the T from it's sort of installed condition. There's no doubt that this is a Band-Aid. And I think this kind of goes back to our original thought that, hey, that the tools are simply in development right now. Unfortunately, this is as good as it gets for what we're doing right now. But this is too clunky. It's inefficient and we're hopeful that the tools will get better.

STEVEN RICHTER: So kind of once you've gone through the panelizing, you have your layout coordinated, you can decide it's probably time to start putting in connections. Now you don't want to put all of your connections in at once. There's a lot of connections that you don't need. Essentially, if it doesn't show up on your erection drawings, you're going to be kind of spending time you don't need to spend putting in all these connections representative of the very end of the project.

So we found that, ideally, it's best to put these connections in once all the framing is complete, reviewed by production and direction, and you kind of have the level of coordination there. And the architect says, look, things that are in here are not going to change.

Now some connections you might want to model earlier are connections that impact the architecture. This might be stairs, or clearance head height issues, various code issues, or connections that you might put in that affect other trades. So, commonly, that's your steel fabricator, some of the fenestration or facade embeds that you might have, or some of the MEP penetrations.

A lot of times we see MEP penetrations coming down along columns. So if you have connections around columns, you're going to want to put those in, so that when they review the model, if they have a pipe coming through a connection that's right along your column, you're not going to have to be changing the entire piece just to accommodate that penetration.

ERICH BRETZ: Sorry. Well, maybe one other point that I guess you can say as part of the connections here is that we would recommend taking care of the connections, modeling the connections, designing the connections, and incorporating those into the erection drawings all at the same time. So you kind of get one guy that's stuck on solving this one connection problem. And he goes throughout the entire model and takes care of all that. I think it does a really good job of sort of making sure that you've touched every one of those kinds of connections. You've rooted out the ones that are, hey, maybe just a little bit different.

You've drawn the detail for it. And you've kind of got that mastered. And you can kind of move on to the next thing.

STEVEN RICHTER: So some strategies for modeling the connections. The connections can essentially-- the strategies can essentially be broken into two concepts. You either model the whole connection, which is a series of components, of the family, or you're going to model the single plate. I think when you're deciding what kind of strategy you want to use, you want to look down the line and see what repetition you're going to have with that connection.

If you have a lot of repetition maybe you build that connection as an individual element, you know, like a combination of those individual elements so that you can just one click, and you have three plates in. If you only use it once, you might just want to use the individual plates to make that connection.

Some of the techniques for this are implementing functional visibility into the actual pieces. So when you have a connection that's a combination of plates, you want to make sure that you can turn off any one of those plates at any given time. Because you might not actually want to use it in your model.

ERICH BRETZ: And then in terms of modeling whole connections versus single plates, I think the key here for us is simply that looking at like this one connection here, we model that as a whole connection. So looking at this connection, you've got a total of one, two, three, four, four pieces that you would have to manage at one time. And then this beam has two of them on it. And presumably, there's two on the other side of it as well. So you've got 4 times 4 times-- yeah.

So you've got a total of 16 little pieces there that you would be forced to manage. So every time one of these little things bounces around or gets moved or whatnot, then you've got to move every one of those pieces. It just becomes a real pain in the neck. So the goal here is to minimize that and simply make this into one entire family.

Functionally, in terms of how it's modeled in Revit is that when you make those families, you typically know which piece is going to have the erection material or the void. And set up the connection so that that is the face that the whole connection is hosted to. We could sit here and talk again for hours about that. If you guys would like to see some of that and kind of how that works, we've kind of got a couple of models here and we're happy to share those with you, as well. You can see how it works.

In terms of creation of erection drawings, at this point, just taking a really quick step back in terms of where we're at along the whole process. I don't want you to think that we're sort of going through this linear. Again, it's just like it was done when you're creating 2D shop drawings.

When you're modeling the connections, since we're going through and we're cutting all the sections around all of our erection drawings, presumably by the time you get to the erection drawing phase, and when you're ready to submit your erection drawings, the amount of work that should be done on your plans and elevations is really pretty minimal. At that point, you're just applying dimensions and tags and that sort of thing. There's really not a whole lot of work to be done there.

I think the thing to keep in mind is that you need to remember who it is that you're speaking to at that point The engineers and architects are obviously going to be reviewing your erection drawings. But then also the erectors are going to need those.

So it's always a good idea to go through your drawings to make sure that you've done a really good job of communicating to the erector what it is that they need to know as much as possible as well.

In terms of reuse of content, we've talked about this before, but we follow the general rule that if you have to type out a piece of text more than three times, that we make a legend for it. So it does apply to obvious things, like plan notes. But then any time that we have typical notes that appear on a plan, maybe this is some typical note next to a core, or something like that, we would turn that into a legend. And then just drop it on the sheets.

We find that the standardization of that kind of text has a really good-- it has a really nice effect of making erectors really efficient because they're not seeing the same information presented in a handful of different ways.

Details. Details are always a funny thing in Revit, even for designers. I think, at some point you'll be faced with the idea of, well, what I do about 2D details? Do I go ahead, with the trade being the possibility of using my details or just 2D line work? We made a decision that the 2D line work is a far better way to go with the obvious downside that you don't have any live data.

The issue is just that so much of this stuff is so reusable and so standard. And you're going to spend so much time cleaning up live details that we just don't really see a whole lot of value in using live details. Don't get me wrong. It's not that there's no value there. It's just that it's so much more efficient to use 2D dead details rather than live

We've also found that after you've created so many details, there's not a whole lot new under the sun as far as precast goes. Whenever you need a new detail, the most efficient way to make a new detail is to start with a detail that you already have. And then simply trace, you know, take a live section, trace the live section, and then create the new detail from that.

Common detail components. These are probably things you guys have in CAD, as well. But all of your standard plates, top view, bottom view, side view, front view, those sorts of things. And then, obviously, cross sections of all your precast components are really good things to have, as well. Yeah.

STEVEN RICHTER: We would hope as they start building off of the model more, some of the details, some of these kind of complicated notes and things can go away and be represented more in model form.

ERICH BRETZ: Yeah.

STEVEN RICHTER: So kind of once you have everything done for your erection drawings, you know, your details, your plates that are represented in the erection drawings, you can kind of start moving on to what you need to put in your piece to make your piece drawing.

There's a couple of strategies that you can do to model this. And similar to modeling the connections as a series of plates, you can actually model the pieces as a series of connections and/or plates. When deciding whether or not you're going to nest these plates into your piece, you really need to look at again the repetition of that connection, whether it's even going to be viable or you're going to nest all of these plates into this piece but only use them once. You know, those are decisions you have to make.

Things that we commonly see like on double T's or these T to T connections, they're on every single piece. That might as well just be nested in a family. They're always in the same place. They're on every piece. Put them in the family.

The same thing with the end bearing plates, or various other erection connections.

ERICH BRETZ: So the visibility of all these things then is controlled with these check-boxes over here. So this stuff obviously won't make a lot of sense to you. But these are standard plates for one of our precast clients. And so when these get checked, then it drives the geometry of that over there.

The data that remains good so when these pieces get wrapped up into assemblies, the data gets pushed through to them, and they're still able to be scheduled despite the fact that they're nested inside of other families.

STEVEN RICHTER: And that's just kind of a good strategy to follow as far as the functionality that you're building in your piece. Like really utilize the visibility parameters of these things. It's an easy thing to program into your piece. And it gives you a lot more flexibility when putting in pieces that have nested families in them.

ERICH BRETZ: So maybe something else is worth mentioning here too is that the obvious downside then-- so these little-- we call these things winglangs, those little winglangs, obviously, are going to be connected to another double T that's right adjacent to them. The downside then is that you've lost that intelligence. You've lost the ability for the connections to know that-- it now becomes a one-sided connection.

So if there's no element on the other side, it obviously isn't going to be connected to that element. Doing something like this though enhances efficiency so much that it's totally worth doing that. Again, thinking about if we had to go through and model each and every one of those little things on the side for 300 double T's, it just becomes something that's so hard to manage that it's just not worthwhile doing.

STEVEN RICHTER: So what kind of doesn't work well is modeling voids into your families. Just because the void is going to have to cut its host family. So you're not going to be able to move that connection to the family it's not hosted to and cut the void. It's just going to make you delete the connection. That's the typical error message that you might get.

So it's just important to consider where you're going to want to be putting or hosting this family as far as the elements that it cuts. So posted to the element that it's cutting.

ERICH BRETZ: And the thing that Steven is talking about here, you can kind of see, it's a little hard to see. But there is a void that's cut out of this wall family here. And to connect this wall piece to the wall piece below it, there's a void that's cut out of the upper wall family and then it's connected between the two with a little piece of erection material. The void basically provides a pocket for that erection material.

There's no great way to turn that void on or off. The void must cut some geometry. So if it's in there, if the family with the void is in your wall family, it must cut the wall. So you may not want the plate to go along with it. You can turn those off. But you can't turn the void off.

STEVEN RICHTER: Kind of another thing along those lines, is you'll see to where those two wall panels are kind of joining and the connection just adjacent to the one that Erich was talking about. We have a void cutting out where one of those plates sit. But the other plate doesn't have a void cutting out that family. So essentially it's just overlapping geometry.

You could build a void in there. But, again, you'd get this warning, especially when trying to build in the whole connection. So things that you probably don't want to model. I mean, there's kind of a simple rule here. If you're not going to quantify it, if you're precaster's not quantifying these things, there's not really a very good reason to model it.

I mean, that's pretty client specific. We know that there's some precasters out there that quantify strands. So maybe you'd want to model strand. But a lot of times we don't see it quantified. So there's no reason to model it. The only thing they really care about with the strand is the pattern. And we represent out with a legend that can just be copied and pasted onto whatever double T types it needs to be. Mesh is another common one. Grout too, although that's built into a lot of families. So it's just client specific.

AUDIENCE: Wouldn't you want to model strand just so six years down the road, when someone decides to drill a hole in a wall panel or [INAUDIBLE] they have that model to look at and, hey, it looks like I might be cutting some--

ERICH BRETZ: Yeah. Go ahead.

STEVEN RICHTER: I was just going to say, it's kind of an interesting question. I think we see that question as far as post tensioning goes too with slabs. In the fact that you could have it in your model, but the model doesn't really verify that they've built it that way, or the exact location, the final location, of where the strand is. I mean, it might move a little bit during the pour, or I mean there could be any number of reasons that the strand isn't exactly where they think it is.

AUDIENCE: Strand is [INAUDIBLE]

ERICH BRETZ: Yeah.

AUDIENCE: --as in [INAUDIBLE] usually [INAUDIBLE] 20 years down the line.

ERICH BRETZ: Yeah.

AUDIENCE: It's not moving.

ERICH BRETZ: I agree with you. I think there's plenty of compelling reasons that you actually would model the strand, you know? I mean, we see-- another reason that you would model it too is we kind of see strand getting jammed up with reinforcing end pieces, as well. So I think there's plenty of very compelling reasons to do it. Unfortunately, we just don't.

We control that. We can show you afterwards, too. We control that with a series of legends that sort drive the strand layout. And that information is used earlier by the production people when they're planning out the beds and what strands need to be jumped through the bulkheads and whatnot. But I'm not disagreeing with you. I agree that there are very compelling reasons to model that. Yeah.

STEVEN RICHTER: OK. So when you're modeling your reinforcing, there's kind of a couple of strategies that you can do. One is model groups. If you group your typical reinforcing together, it just allows you to click the group, copy it to the next piece. commonly seen in like spandrels, things like that, where you have like lots of end reinforcing. I think it's actually in the next slide, but-- yeah. No. No. There we go. So on the spandrels where you see you have lots of reinforcing that's very similar, create it into a group, and then just copy it around all over the place.

As far as the lifting is concerned, one thing that we employ is this Dynamo script to calculate the center of gravity for it. It's a really easy script. I mean, it takes 20 seconds to put it together on Dynamo. But it puts a little centroid where the center of gravity is on that piece. We then overlay a template. And the template's just a detail item that you can flash to the edge of the piece, they'll break the piece up into tenths or fifths or whatever your engineer kind of specifies.

And then you can align your lifting elements with that template. So that's kind of how we spaced out the lifting elements. On some of our standard just rectangular pieces, we nest the lifting elements into the piece. I mean, you know where they're going to be every single time, if there's no cut out. If the center of gravity is in the middle, the lifting doesn't change.

ERICH BRETZ: I think too it's worth mentioning that in order to make this-- since we're modeling every single piece and part that gets cast into the precast piece, well, with the exception of strand, like you said. You've got tens of thousands of little pieces that you're trying to manage at the same time. To remodel each one of these things over and over and over is a disaster.

So you've got to find creative ways to copy and repetition throughout your model. So when we're talking about using these so-called super families where we've got all of the pieces already embedded into the piece so we don't have to bother modeling those as one offs every time, or copying and modifying. That's really a huge key to success.

I mean, you can blow through a lot of time just trying to get all that stuff in your models. So piece drawings, I'm assuming that most of you guys know what piece drawings are. Piece drawings are really, just in a big picture sense, piece drawings are just the individual fabrication details that tell the guys at the bed how it is that the piece should come together.

The Revit assembly functionality is exactly what this is made for. So it allows you to annotate and to operate on a piece independent of the whole, any one element independent of the whole model. So our assemblies are-- so basically the way that we think about it is that you take a piece, you wrap it up in assembly, you dump all of the plates that are embedded into that piece into the assembly. You assign data. And then you use schedules, legends, and parameters to sort of represent and to change how it is that that data is shown.

This is a quick example. Our precaster wants us to show the plates, the straight bar, and the erection material shown separately. These are all from-- this is from a piece drawing, which you'll see over here too. So they want to see this information shown separately. So we control all that stuff in the assembly views. These are just assembly views using parameters and schedules.

Maybe this is a decent time to mention that we use, again, that we use the Edge for Revit system for this, which is OK. It really is in a sense sort of precast piece drawings out of a box. But it provides sort of the structure for this whole thing to happen. It provides the plate templates, the data management tools, with schedules, parameters, that sort of thing, for all of us to be able to happen.

Make extensive use of legends. Anything that we're going to do over and over we're going to put it into a legend. You know, I think the next generation of this, thinking too at least in terms of our office, is going to be in lieu of putting this-- this is a finished legend which appears on every single piece. In lieu of putting it into a legend, an even better way to do it would be to put it-- just embed it into the title block and control it with a check-box in the title block.

Because that way you're sure that it gets put on every sheet. Also able to use schedules to make sure that it is applied to every sheet when it needs to be. Whereas with a legend, you don't have the ability to control that to know if it is on a sheet or not. Yeah.

In terms of other piece drawing strategies, taking a really quick step back. In terms of piece drawings, we found that the erection drawing process takes about 60% of our time on a total project. And the piece sharing process is about-- I'm sorry. Pieces is 60%, erection drawing's 40%. So the piece writing process is super long. It takes a ton of time.

What we're seeing over here is sort of our automation of similar stuff at the ends of pieces. So this is a beam piece drawing over here. In reality, nearly every one of your beams is going to have the same data over at each end. And so it doesn't really make a lot of sense to recreate that data over and over and over.