R.S. issues - linking, decks and more

[dbaldacchino]

Hi all,

We've been tinkering with RS and found a few things that we had questions about.

a) First of all, in slabs, you can define a system family (deck) and you can also create a standalone deck, such as a 1 1/2" metal deck. It took us a while to find this feature, but once you see it, it's great. Now, if we're pushing a model to an analysis application such as ETABS or ROBOT, we'd like to have the gauge specified in Revit parameters. But this doesn't exist. Is there a reason why not? We understand that it can be added as a project parameter, but typically, if it's not defined in the system family, 3rd party developers cannot make a connection to it.

b) We're linking a RB model into RS and noticed that we cannot use the pick walls on the linked model to create slabs. Why is that?

c) The copy/monitor only works on grids, levels and columns. Will this be expanded to walls, roofs and floors in the future? I think these are very important elements to monitor. If the Architect moves a wall in the linked model (especially walls designated as Structural), the engineer needs to know that. He (or she )will have a wall drawn in the RS model and by accepting the change, that wall would move accordingly, thus ensuring coordination.

d) In case of roofs, in early stages of DD, the Architect typically leaves the roofs as flat. It would be nice if framing members could be "connected" to the bottom of a roof so that when the slope changes to 1/4" : 12" later on, the members would follow that slope. How are you guys/gals handling this?

I understand that both sides needs to talk, and there's no substitute for that. But if we're both modeling, I expect some automatic coordination in return. If we worked off the same model with worksets (we have in-house structural), then perhaps some of the above issues wouldn't exist. I'm not sure if that's a correct statement.

Any views/opinions? Thanks!

[rmcelvain.103137]

Hey David,

I'll try to give my view/opinion (limited as they are, I'm just adding to your questions basically ) to each one of your points...

a) I don't know what the reasoning is. My thought on it has to do with adding the guage parameter in the analysis program. Can you add it in ETABS and then have it be imported back into Revit?

b) no clue, have not tried it?

c) I agree with your thoughts on the importance of those elements, again I have not tried this feature but at this early stage I would guess there is a lot that will happen with the software. I would recommend posting an expanded version in the wishlist forum.

d) This one is something that seems really difficult. Take for instance the roof I attached HERE the lower portion creates a roof plane that when modeled is warped. When I sloped the joists to create this roof the top of the joist projected beyond the plane of the roof defined by the elevations. If I were to try and attach the joist to the plane of the roof .... well, it just isn't possible. I like the idea of controlling the top of member at each end (and if we can get it to work, any secondary framing would follow along with main framing slopes).

Thanks for the thread,

[Tom Weir]

Hi David,

c. The categories to be monitored will be expanded in the future.

d. Warped roofs are nearly impossible presently, the developers know it, and I think we can expect dramatic improvements in RS3 and 4.

<In case of roofs, in early stages of DD, the Architect typically leaves the roofs as flat> That is they currently work that way in 2D CAD drafting. If you have worked with architects using Revit Building it would not be like that. I think that the workflow has to change with modeling.
What I have been doing is to force (ask polietely) the architect to determine the ridges and valleys on the roof early in DD. That way I can set up the sloping roof portions. By making the bottom of the roof deck the current workplane the beams can then be attached to it, and move with it as the slopes are better defined and adjusted later in the design process
As more architects move to modeling we should see a better undertanding from them that lots happens in the early stages of the model and that we need that important info.

Have a great day....

Tom Weir
Los Angeles

[dbaldacchino]

Thanks for your reply. I know that Revit developers look through these forums from time to time and hope we might get some notes from them for why/how etc. or what to expect in the future. And I also like to hear what other user's experiences are.

About the gauge parameter, I've talked to someone at Robot and they said they could add such functionality through the API. For instace you would add a project parameter for gauge and assign it to decks, and the developers of Robot would add functionality to recognize that parameter and assign it to shear thickness, for example. But it would help if RS had that built-in to the slab family, or the stand-alone deck. That's just one minor issue that stood out right away as we started experimenting. I'm sure we'll have more as we learn to work with it better.

We're trying to see how far we can go with linking models before we resort to working off the same model. It has been suggested in previous posts by Revit Developers that, although it's possible to workset a model and have both structural and architectural working on one model, that it makes better sense to use linking for various reasons.

[dbaldacchino]

Thanks a lot for your feedback Tom.

I'm actually in Architectural and I was explaining how we typically attack a project. In later stages of DD we would have determined drainage, slopes etc. but in early DD, we just try to think in massing terms (big picture items).

So if a roof is flat and the engineer models the framing, would it be too hard to slope it after the fact? What if the roof slopes, and then later it moves up or down for some reason; is that also difficult? We're trying to anticipate potential problems to find the best way to collaborate together. I understand that you can pick a slab face and set it that as the workplane (as long as it's a flat surface), but don't know how you can "remake"/update the workplane if a roof changes so that those members update accordingly.

[rmcelvain.103137]

would it be too hard to slope it after the fact? What if the roof slopes, and then later it moves up or down for some reason; is that also difficult?

It depends on the situation, I would suspect the 1st situation (flat to sloped) would just require a redo of the framing. The 2nd doesn't seem to be too bad of a situation; mere elevation changes happen pretty cleanly.

The developers are aware of this issue (and others) and I have full confidence that they are striving to make it easier.

It's nice to hear you're keeping us in your thoughts regarding work process, makes me feel all warm and fuzzy. Which is something considering the weather rolling through here today; have you ever heard of thunder and lightning during a snow/freezing rain storm???

[Tom Weir]

Hi David,
The developers usually don't want to discuss future additions so as not to get people anticipoating stuff that will not be there. But they definately are listening, and perhaps we can move this thread to the wishlist forum.
As for sloping obviously if the beam member is going through a ridge or valley it won't work. You could do a flat roof and make the bottom of deck the workplane. Then alter the slope angle at any time and the framing will follow, or move the whole roof up or down. That's not a problem.
What do you do in DD when you are modeling in Revit Building though? Still draw just a flat roof? I think of massing as a more schemtaic stage.

Tom

[dbaldacchino]

Tom,

I'm not thinking that the developers would come and tell us exactly what to expect, perhaps just a teaser such as "we'll consider it for the future". Keeping the flame going you know

Well, early DD and SD overlap. There's no exact transition. It's a bit muddy, and we're really on our first true significant RB and RS project, so we're figuring a lot of things out as we go. We've already spent some time in both applications learning the tools etc. and experimenting before the start of the actual project.

As for roofs in early DD, I've been putting them in as flat at their appropriate level, but am fully aware that they'll be sloped. Typically you're figuring out a framing scheme first, before knowing how the roof will ultimately be broken down and where it slopes. Granted that you'll be thinking about that as you lay out the grid and structure too. Just to give you an idea, we do educational facilities and this project is a 2 story Middle School, about 173,000SF.

We've done some tests today where we selected a roof in the RB model as the workplane, and drew some framing. When the roof in RB was moved up or down, the framing moved in RS when the linked model was reloaded. So far, so good. But, when we rotated a roof, it gave an error and detached those framing members. It would only let us associate a plane that had the same slope as the beam. To gain control over the beam slope, we created a ref. plane at the beam slope, associated that as the workplane of the beams, then rotated the ref. plane. We then thought....what if we draw ref. planes to follow a roof (whether sloped or not) and then align them to the linked model? Well, Revit won't allow you to align to a linked model (I don't know why not). So then, you'd have to rotate the beam to the exact roof slope and re-associate with the roof once again. It seems like a lot of steps and lots of maintenance to me.

So we're kinda stumped. If we draw flat roofs and later we rotate them to the appropriate slope once the framing scheme is laid out, we lose the connection and have to go through some hoops to associate them back to the roof. If the slope changes for any reason (sloped to different slope, etc.) you lose the association and I think it should be easier to re-establish the connection. Or not lose the connection at all because planes can move up or down, but they can also rotate.

[Tom Weir]

Hi,
I am currently doing a high school campus, and two middle school campuses, so we seem to be in similar situations. Roof systems are one of the areas I have focused on with the developers and I think you can get a handle on the situation, though presently it is a bit tough.

For your scenario of starting with a flat roof plate when you add the beams, why not go to a section view and make the bottom of the deck the workplane, even though it is flat. Then if you change the slope angle the beams should follow the deck. Rotating and mirroring can be a pain, but you might try grouping the elements before you rotate. That sometimes help.

I have included 2 DWF files of one of the middle school campuses. The multi-purpose building over parking has multiple roof portions sloping all over the place. I had minimal information but set slopes for all of them.

Keep me posted. Very interesting thread.

Tom Weir
Los Angeles

[dbaldacchino]

Thanks again Tom!

I haven't messed much with dwf files, but I'm extremely impressed...90kb files with all that info. on each element?! Just awesome.

Back to the beam discussion.....if you attach the beams to the slab as modeled in RS, then yes, the beams rotate. The problem we found is when you use the RB slab (linked rvt model) to attach the beam to. It draws the beams fine, and if the slab in the linked model moves up or down, it keeps the association and the beams coordinate accordingly. But not if you change it's slope; you get the an error message (see attached jpg). I don't think the verbiage makes a lot of sense because there still is a plane. It's just rotated from what it was before.

Getting the beams re-associated with the linked file's slab is a real pain and I think we have to look at using the linked model more as a background and not attach structure to it. So using the linked RB model as a background, draw the slabs in RS first, use that as the plane and then make sure that when the RB roofs change slope, go back and update the slabs in RS manually. The beams will follow the slope of the slab in RS.

Another interesting thing I noticed was that in RS, you can only define 1 slope arrow on the roof slab. In RB, you're not limited to one. Although if you put a lot, the roof still doesn't deform as you expect it, without getting any errors. I found that to be kinda strange.