Curtain Wall Sealant Gap @ Jamb/Sills

[harrisbrett]

How is everyone dealing with the shim space/sealant at the jambs and sills of your Curtain Walls?

I've nested Detail Components within my mullions based on a Kawneer system which works fine. Are you drawing a 1/2" (or whatever gap) larger mullion for these conditions so you can then detail that space with sealant and shims at smaller scales? I'd like to find a solution that automates this as much as possible.

Another solution that came to mind would be adding another grid close to the bottom/jambs and then setting the bottom panel to be empty which you could then detail inside of.

How are people approaching this?

(Hope I was clear in explaining this)

[twiceroadsfool]

Add it in to the mullion sizes for the ends, then have it as a DC as you already mentioned.

Its not worth losing the efficiency of automatic embedding, etc, to actually have a "gap" in there.

Plus you want the "wall opening" to be the even dimension, if you think about it...

[robert.manna]

I think "we" (that would be a collective we for the firm) have tried just about every approach (including what you outline and what Aaron had to say). Pretty much everything has its drawbacks. My favored approach is to leave the curtain wall mullion alone, assume that the rough opening being created is the rough opening that the contractor is responsible for filling, and dimension accordingly. We don't account for the 1/2" gap when dimensioning where grid lines should be (did you/we in CAD?), and half the time grids/mullions line up with other stuff, or you can work off of a mullion corner, or something. Lastly, we just "fix" it in the detail views where we need to show a detailed mullion, sealant, flashing etc. I don't like embedding DC's into mullions because you can never really turn off the mullion geometry so either way you have to do a bunch of legwork, and I find the legwork on top of the "basic" mullion is easier then dealing with the DC in the mullion.

[Scott Womack]

I actually agree with Robert's approach completely. This works for 97 percent of all projects just fine. Just make sure there is a "General" curtainwall/storefront note to cover how you are dimensioning thses items/openings.

Personnally if FORCED into accounting for the 1/2" gap, I would create a left and right "end" mullion family, and offset it so is is "in" from the edge by 1/2" Thus, you could then by tabbing dimension to the ends of the curtainwall centerline for the "rough"" opening or masonry opening; or to the outside edges of the mullions, to get the actual dimension.

[robert.manna]

I'm actually not sure the 1/2 offset on the outside would work without having "mullion" geometry there. I think the wall that curtain wall is embedded into will shrink to fit (but I'm not 100% sure). We do have left/top right/bottom mullions as some people like to at least have the glazing pocket, however none of the mullions have the embedded detail components. I can see where not modeling the 1/2" would be a problem in design build, or firms that do it "all", however, I've yet to see a case where as the arch of record we are expected to dimension the curtain wall exactly down to the 1/2" in terms of accounting for these gaps and such.

[twiceroadsfool]

I'm actually not sure the 1/2 offset on the outside would work without having "mullion" geometry there. I think the wall that curtain wall is embedded into will shrink to fit (but I'm not 100% sure). We do have left/top right/bottom mullions as some people like to at least have the glazing pocket, however none of the mullions have the embedded detail components. I can see where not modeling the 1/2" would be a problem in design build, or firms that do it "all", however, I've yet to see a case where as the arch of record we are expected to dimension the curtain wall exactly down to the 1/2" in terms of accounting for these gaps and such.

Ive done a lot of projects where that half inch became an issue, or became significant when something wasnt meshing on site, so we account for it. Did we in CAD? Absolutely.

Having the mullion geometry there at the detail level isnt all that problematic. Youve got a couple of lines to linework tool, which is roughly the same amount of work as placing the DC seperately anyway, in whatever detail youre using to cover that shim space.

But i also find that a lot of people (NOT talking about you folks, you obviously know better) who are modeling this stuff dont really UNDERSTAND the built condition, and theyre not thinking about it. Then a detail gets drawn, and theres no shim space. Having it there, ahd having a conversation with them about WHY they have to use a different mullion at the ends (most of which are rule based systems anyway) is paramount.

Theres really no harm in doing it either way. You have to show it SOMEWHERE (imho...).

[eric.piotrowicz]

Aaron, Thank you. We had a similar situation arise where one of the PM's was insisting on getting the shim space accounted for, not that I disagreed with his point but nobody had asked for it before. In any case I came up with the same solution of creating an accurate profile and thus modeling the true built condition. It has been working great, especially after nesting the detail component. I was starting to question my methods while reading this thread based on thinking I might have over modeled. I'm glad there is someone else that is doing the same.:beer:

[twiceroadsfool]

Aaron, Thank you. We had a similar situation arise where one of the PM's was insisting on getting the shim space accounted for, not that I disagreed with his point but nobody had asked for it before. In any case I came up with the same solution of creating an accurate profile and thus modeling the true built condition. It has been working great, especially after nesting the detail component. I was starting to question my methods while reading this thread based on thinking I might have over modeled. I'm glad there is someone else that is doing the same.:beer:

With respect to overmodeling.... It doesnt get "modeled" for us any more than its a mullion thats a little wider. The DC takes care of the rest. I dont model fillets for sealant, and i dont model any backer rods and shims, LOL.

I have a tendancy to over detail things, in ARCHITECTURE. I dont care about software, or performance, or whatever. if the ARCHITECTURE isnt getting articulated, im not doing my job.

No one details something, until they learn that they have to. Weve had problems with it, so we do it that way.

Dont get me started on actual versus nominal, LOL...

[eric.piotrowicz]

RIght there with ya on the actual/nominal debate. Makes me irritated just thinking about it.
I did go a step further on the modeling since I added a 1/2"x1/2" return were the sealant goes since somebody demanded that it be shown and able to be dimensioned. It didn't seem to cause a perforance hit so I went ahead with it.

[twiceroadsfool]

To be honest, i worry less and less about the "over modeling" debate, unless its stuff that is entirely non value adding.

Other circumstances (project sizes, memory limitations, Operating systems, etc) have pushed us all to a new level of hardware that appears to be so above and beyond, im not longer really worried about the model being demanding, as long as its PERFORMING to serve the ARCHITECTURE.

So im with you. If youre going to show it and dimension it, id model it.

[t1.shep]

I am also trying to find the right balance between efficiency, usability, accuracy, etc. in regards to modeling and detailing curtain wall/storefronts. In my ideal world the curtain wall would cut the wall including the shim space/R.O. so that detailing would be quicker (less linework, cut profile editing, masking regions, adding detail components). This assumes that the mullion Detail Component is included in the mullion (like Kawneer's). However, I want the curtain wall to dimension to the mullion dimension and not the rough opening dimension. As others have said, when you make the mullion cut the wall including the rough opening, there is not a way (that I can find) to dimension to the actual mullion, it includes the entire R.O. Putting additional ref. planes in the mullion profile does not do anything.
As you can see in the attached image, I want to be able to dimension to the 2'-0" mullion to mullion dimension, but I want the CW to cut the wall down to the R.O. as other elements in the project will relate to the R.O. In the attached file I added a filled region to the Detail component that is nested into the mullion profile so it looks correct at the mullion, but you can see it doesn't cut the entire wall.
I don't want to have to add a calculated value to report the curtain wall dimensions to exclude the R.O. as that might change depending on the detail.
We also elevate our individual CW and so the model would need to look correct in elevation in that we only need to dimension the Mullion/frame outline and not the R.O. lines.
So, essentially, I need the model to look correct (cuts R.O.) but I need to only dimension to the edges of the mullions. Sounds much more simple when I put it that way.

[jamesgchambers]

This may be an "over-modelling" approach, but it's how we're doing it on our projects...

Since we need to dimension the MO/RO in overall drawings and then produce separate elevations of each curtainwall, we'll actually add a curtain grid just inside the outside of our curtainwall and replace the panels around the edge w/ a system panel type called "soft joint." it sounds like a lot of work, but it's all upfront work and future modifications are pretty simple and require much less clean up work. Then we set up our curtain wall elevations to exclude the "soft joint" panel type via filter.

We had tried the edge mullion approach before, but folks weren't satisfied with how the mullions didn't clean up (the verticals would project beyond the base of the horizontals), and therefore were having staff clean up the intersections with detailing.

Hate to say it, but it seems a simple add for the factory - shim space parameters for left, right, horizontal, and vertical edges. Granted it gets hairy when dealing with non-rectangular curtainwalls.

[Dave Jones]

This may be an "over-modelling" approach, but it's how we're doing it on our projects...

Since we need to dimension the MO/RO in overall drawings and then produce separate elevations of each curtainwall, we'll actually add a curtain grid just inside the outside of our curtainwall and replace the panels around the edge w/ a system panel type called "soft joint." it sounds like a lot of work, but it's all upfront work and future modifications are pretty simple and require much less clean up work. Then we set up our curtain wall elevations to exclude the "soft joint" panel type via filter.

We had tried the edge mullion approach before, but folks weren't satisfied with how the mullions didn't clean up (the verticals would project beyond the base of the horizontals), and therefore were having staff clean up the intersections with detailing.

Hate to say it, but it seems a simple add for the factory - shim space parameters for left, right, horizontal, and vertical edges. Granted it gets hairy when dealing with non-rectangular curtainwalls.

from my perspective as a specialty curtain wall detailer it doesn't matter if the project docs account for the caulking joint gap or not. What's important to me is that the documents indicate where the dimensions on plans, sections, elevations, or frame schedules are to and from. On almost every project I do this is my first RFI question because rarely is this information on the docs. Trying to figure out what a dimension extension line is relating to on a 1/4" = 1-0" frame elevation is no fun...(see t1-shep's 2'-0" dim above for instance)

and, as an aside, a caulking joint is not an aesthetic feature whose size (width) can be dictated by what you want to see. The width is determined by the design criteria of the building (story drift, LL and DL deflections, etc). I can't begin to count the number of "fights" that I've had with Architects over the years because they wanted to "see" a 3/8" joint when a 3/4" joint was required. OMG, you'd think I was pulling their arm out of the socket or something :p

[t1.shep]

from my perspective as a specialty curtain wall detailer it doesn't matter if the project docs account for the caulking joint gap or not. What's important to me is that the documents indicate where the dimensions on plans, sections, elevations, or frame schedules are to and from. On almost every project I do this is my first RFI question because rarely is this information on the docs. Trying to figure out what a dimension extension line is relating to on a 1/4" = 1-0" frame elevation is no fun...(see t1-shep's 2'-0" dim above for instance)

and, as an aside, a caulking joint is not an aesthetic feature whose size (width) can be dictated by what you want to see. The width is determined by the design criteria of the building (story drift, LL and DL deflections, etc). I can't begin to count the number of "fights" that I've had with Architects over the years because they wanted to "see" a 3/8" joint when a 3/4" joint was required. OMG, you'd think I was pulling their arm out of the socket or something :p

This is what I'm trying to get at...we only want to dimension the mullion to mullion dimensions, knowing that the R.O. and such will not be typically defined by us. But in the case of Kawneer's details you can see that their sill mullion has additional pieces and such to create the sill mullion. However, in their typical details they show that they dimension to the mullion as if it didn't have the extra stuff on it. So, i.e. a 2" mullion is what you are typically dimensioning to and the shim/caulking, and additional extrusions are the responsibility of the contractor and mfgr. The trouble with Revit is that to get all your views showing correctly and without doing a ton of extra detailing, linework, filled regions, etc. you need to account for the true R.O. size. However, this has impacts on how the elevations look (you now have "extra" lines on the elevation as you pickup the R.O. line and the Mullion line) or you have an elevation that looks as if the mullions are larger than they are because you need to include the shim space to get the sections and details to look correct.
Like someone previously mentioned, it'd be nice to have a R.O. offset so that your CW or window, or door, cuts the wall slightly more than the actual geometry. This can be done with windows and doors, but not so easily with CW.

[Dave Jones]

Like someone previously mentioned, it'd be nice to have a R.O. offset so that your CW or window, or door, cuts the wall slightly more than the actual geometry. This can be done with windows and doors, but not so easily with CW.

Revit curtain walls are an unruly beast that need a lot of work IMO. Hopefully in the future they can more closely be created with some semblance of reality. Like scheduleable corners and the ability to have curtain panels (glass lites) span vertical mullion splice locations...

[jkendall.200868]

As a curtainwall guy, I've seen the caulk joint become an issue many times. One of the common ones is when the caulk joint is not accounted for around a door with no sidelites. For example, you've got a standard 3'-0" wide door opening (which is usually minimum to meet ADA) with a 2" storefront frame on either side. If you don't account for the caulk joint, the rough opening will be made 3'-4". If you need 1/2" caulk joint on each side then your frame dimension becomes 3'-3" and your door opening becomes 2'-11" which may or may not meet ADA requirements, not to mention it makes one more unique door size on the project. Easy to fix in the design stage, not so easy to fix when the pre-cast panel guys have already made their surrounding panels.

It's a similar issue when sub-framing is not accounted for with a door in curtainwall. It's sometimes possible to do doors in CW without subframes, but it limits your door hardware choices. Not a big deal if there are sidelites to adjust.

I don't care if you dimension to RO or mullion frame dimension as long as you tell me which it is and what caulk joint you are assuming. I'd prefer RO with references to center or top of intermediate mullions so I can determine caulk joint requirements on my own and not have to "please verify" note to death on the shops when my DLO dimensions don't match up exactly with the architectural CW elevation sheet.

[mike.koutsoulias]

This post is probably too late but it could be useful for those of you that would like to generate accurate working drawing mullions sealant gap and all.
The easiest way to include the gap is to add the extra width to your perimeter mullions. Combined with a detail component and a profile using invisible lines in the mullion family it makes for a convincing detail in section and provides an actual modelled edge to dimension both the rough opening and the frame. See the attached pdf for an illustrated explanation.

[Dave Jones]

This post is probably too late but it could be useful for those of you that would like to generate accurate working drawing mullions sealant gap and all.
The easiest way to include the gap is to add the extra width to your perimeter mullions. Combined with a detail component and a profile using invisible lines in the mullion family it makes for a convincing detail in section and provides an actual modelled edge to dimension both the rough opening and the frame. See the attached pdf for an illustrated explanation.

all it needs is a parameter to flex the width of the caulking gap and you're there

[twiceroadsfool]

all it needs is a parameter to flex the width of the caulking gap and you're there

Thats what ours has, replete with a masing region so it can get detailed as necessary to conform with the adjacent consutrction. Plus, the profile "bumps" out slightly at the masing region, giving it the correct linework in elevation and detail.

[Dave Jones]

Thats what ours has, replete with a masing region so it can get detailed as necessary to conform with the adjacent consutrction. Plus, the profile "bumps" out slightly at the masing region, giving it the correct linework in elevation and detail.

I'm thinking of putting flip controls on my profiles to eliminate the lefts and rights, and tops and bottoms. I think that I could get this down to 3 profiles for each system. Panel 1 side for the perimeter, panel two sides for the intermediate, and panel no sides for specialty stuff like door jambs at walls. Think I'll work on that in my spare time :lol:

[twiceroadsfool]

It wont work. The "Profile" isnt what you see, when its loaded in to the Mullion.

But i dont do Up/Down or Left/Right in Revit. I do End/OppEnd, because of the Mirror Tools, and draw direction. :)

We end up with 4 for each system. End (Closed/adjacent to glass), Opp End (adjacent to glass/Closed), Center (glass/glass), End (Closed/Closed).

[Dave Jones]

It wont work. The "Profile" isnt what you see, when its loaded in to the Mullion.

But i dont do Up/Down or Left/Right in Revit. I do End/OppEnd, because of the Mirror Tools, and draw direction. :)

We end up with 4 for each system. End (Closed/adjacent to glass), Opp End (adjacent to glass/Closed), Center (glass/glass), End (Closed/Closed).

for me the profile is what I see when it's loaded into a mullion as I don't add details at the family level. If I did it that way I'd have to have 100 different profiles for each system to cover all of the varying details. For instance, the curtain wall project that I'm currently working on has a 34' high frame that the left jamb contains 9 different conditions from bottom to top. The other jamb has 7 conditions, 4 of which are different than the left jamb. Hence I'm a Revit modeler but an A******d detailer. You guys don't have to worry about this level of detail because you leave it to me ;)

[Dave Jones]

Thats what ours has, replete with a masing region so it can get detailed as necessary to conform with the adjacent consutrction. Plus, the profile "bumps" out slightly at the masing region, giving it the correct linework in elevation and detail.

Aaron, how do you get a massing region into a Profile-Mullion.rft based curtain wall profile family? Or, do you use a different template?

[Alfredo Medina]

Aaron, how do you get a massing region into a Profile-Mullion.rft based curtain wall profile family? Or, do you use a different template?

I think it was just a typo; I think Aaron meant to type "masking region", instead, which can be used indeed to obscure the additional width of end mullions in the project, (not in the family) and add some detail linework on top of that masking region.

[Alfredo Medina]

.... I don't like embedding DC's into mullions because you can never really turn off the mullion geometry so either way you have to do a bunch of legwork, and I find the legwork on top of the "basic" mullion is easier then dealing with the DC in the mullion.

There is a way to turn off the model mullion and show only the detail component. Check this solution: http://forums.augi.com/showthread.php?p=1100482#post1100482

[twiceroadsfool]

There is a way to turn off the model mullion and show only the detail component. Check this solution: http://forums.augi.com/showthread.php?p=1100482#post1100482

No offense meant, but thats really not a good solution. First: "White" doesnt mean the line isnt there, it means the line is white. That only works for hiding the line, assuming youre printing on white paper, in color. If you set your colors to plot to black, that doesnt work. If you export to other softwares, that doesnt work. if there are objects behind or below the mullion, it doesnt work.

Plus, ive never found the need. It took us two passes at setting up our Mullions. The Profile and geometry is accurate, and the detail covers the rest. No need to hide anything.

The nested Detail Components work great.

[Alfredo Medina]

... That only works for hiding the line, assuming youre printing on white paper, in color.

Well, most of the times I print on white paper, in color, even if there is no color in the drawing, so my white lines do disappear if I need, so it does work in most situations, unless there is another solution.

[twiceroadsfool]

Well as i said, i generally try to model them correctly so i dont need to hide the geometry. And i wasnt being snarky with the white paper comment, i apologize if it sounded that way. The reason i said white PAPER, was because of CAD exports, DWF, PDF, etc.

Plus, its great if you keep an offices standard plot styles set to plot color, but you need to mention that with the solution, because its a huge deal breaker. A lot of offices set drawings to plot black, so their logos dont come out lousy from gradients, etc.

Mullions are finicky, but- barring rediculous amounts of detail- i still fail to see how the exterior profile of the mullion is so bad that the modeled geometry needs to be turned off in a detail view.

[Dave Jones]

I think it was just a typo; I think Aaron meant to type "masking region", instead, which can be used indeed to obscure the additional width of end mullions in the project, (not in the family) and add some detail linework on top of that masking region.

I knew he meant masking region...I did, really :?

OK, so that's how I'm doing it now, creating curtain walls to outside frame dimensions then adding a masking region to indicate the caulking joint width and using the outside of the MR for a dimension point. I don't use any linework other than the masking region.

It would be nice to be able to create them in the profile family but I haven't figured out how to do it that doesn't create havoc at some turn or the other

[harrisbrett]

As the thread starter, I thought I'd pipe in finally. I do the nesting DC's within the larger mullion profile and as Aaron said, then let the DC take care of any detailing that I need with Masking regions, linework, etc.

I also do the multiple configs simliar to Aarons way with respect to jambs/sills/heads.

So far it works out perfectly. No complaints.

[TheMonk]

I do not believe flip controls are available for Revit family profiles. If you have one I would like to see it.

[htlmv]

105446

my profile for YKK left jamb