How do construct walls like "FRANK GEHRY"

[william.112158]

those of you who know his work, understand the concave sturcture, angled
walls and various forms that are used to make a "WALL".

In the Revit model, Im having difficult time constructing such "Walls". Can anyone offer a drirection/ a family or profile for which I can construct such a wall, would be most appreciated.

Im in the dark and I need help.

Lunch is one me for Great Tips.
thanks

[.chad]

Doesnt he use Form-Z?

[randolph.fritz102047]

Only solution I am aware of is to build a mass with the appropriate surfaces (for very complex forms, you may have to import a model), and then use wall-to-face or curtain system to face. Be warned that a lot of Revit's wall-surface automation doesn't operate on a curved, sloping wall--connecting wall profiles will have to be manually edited, line textures will have to be manually rendered, and so on.

[randolph.fritz102047]

Chad, Gehry actually builds (or at least built) physical models; his office uses aircraft CAD technology to digitize them. It's probably possible to get such forms into Revit by import.

[DanielleAnderson]

those of you who know his work, understand the concave sturcture, angled
walls and various forms that are used to make a "WALL".

In the Revit model, Im having difficult time constructing such "Walls". Can anyone offer a drirection/ a family or profile for which I can construct such a wall, would be most appreciated.

Im in the dark and I need help.

Lunch is one me for Great Tips.
thanks

There was a great presentation about this at AU last year, maybe you can track down the course materials. The best way to model these kind of shapes is to use the massing tool and then apply curtain walls to the forms. The building used in that presentation to demonstrate this technique was Frank Gehry's new building at MIT and some of Gaudi's work - so it is possible in Revit, just maybe not as easy as using CATIA (although I've never used CATIA so I don't really know). Play around with massing forms thinking about how they might form your walls, try out the blend tool to create some of these forms, and you'll be pretty surprised how crazy you can really get with Revit.
Hope that helps.

[randolph.fritz102047]

I actually took a stab at constructing some parabolic masonry forms in Revit, a few months back when I was first studying. I found that Revit's massing tools cannot create parabolas, let alone catenary arcs. So if one is serious about modeling a funicular structure, probably the way to do it would be to do what the engineers do, and build the model in Mathematica, or perhaps imitate Gaudi--use graphic statics for 2-D forms or build an actual funicular model and do a takeoff.

[david_peterson]

those of you who know his work, understand the concave sturcture, angled
walls and various forms that are used to make a "WALL".

In the Revit model, Im having difficult time constructing such "Walls". Can anyone offer a drirection/ a family or profile for which I can construct such a wall, would be most appreciated.

Im in the dark and I need help.

Lunch is one me for Great Tips.
thanksIf I remember correctly I think I saw that on an episode of the Simpsons. He just puts up the steel frame and then beats it into place with a wrecking ball until its "Perfect"

[.chad]

Chad, Gehry actually builds (or at least built) physical models; his office uses aircraft CAD technology to digitize them. It's probably possible to get such forms into Revit by import.oh. i guess it was eisenman who used form-z. learn something new every day :lol:

[Justin Marchiel]

If I remember correctly I think I saw that on an episode of the Simpsons. He just puts up the steel frame and then beats it into place with a wrecking ball until its "Perfect"

yeah was that the one where the convicts where escaping by sliding down the roof?

Good episode! LOL

Justin

[Fred Blome]

Google comes up with links like this with more pics from the Simpson episode: http://blog.davidteoh.com/archives/2005/11/frank_gehry_in.html

[AP23]

There was a great presentation about this at AU last year, maybe you can track down the course materials. The best way to model these kind of shapes is to use the massing tool and then apply curtain walls to the forms. The building used in that presentation to demonstrate this technique was Frank Gehry's new building at MIT and some of Gaudi's work - so it is possible in Revit, just maybe not as easy as using CATIA (although I've never used CATIA so I don't really know). Play around with massing forms thinking about how they might form your walls, try out the blend tool to create some of these forms, and you'll be pretty surprised how crazy you can really get with Revit.
Hope that helps.

Are you sure a curtain wall is applied? You'll get the "Wall type curtain Wall' is incompatible with picked face. Only basic wall types are accepted" message.

By the way, was that the BD22-1 : Wild Shapes Get Real with Autodesk® Revit® Building: Design Buildable Free-Form Shapes class by Matthew Jezyk. Unfortunately the session handout was never published on the website, so if anyone care to share it with us, would be great.

[WolffG]

I can't help but ask: why would you want to???

[randolph.fritz102047]

Curtain systems, not curtain walls, can be attached to masses.

[DanielleAnderson]

Are you sure a curtain wall is applied? You'll get the "Wall type curtain Wall' is incompatible with picked face. Only basic wall types are accepted" message.

By the way, was that the BD22-1 : Wild Shapes Get Real with Autodesk® Revit® Building: Design Buildable Free-Form Shapes class by Matthew Jezyk. Unfortunately the session handout was never published on the website, so if anyone care to share it with us, would be great.

Yah, that was the one. I'll see if I can find a few minutes to scan it and add it to this thread.

[AP23]

Curtain systems, not curtain walls, can be attached to masses.

yeah but curtain panels of the curtain system can't transform into the shape of the mass like basic wall do. Unfortunately, you can't replace these panels with basic wall either like you can with curtain walls. So the panels don't bend and you get these weird segmented shapes.

[ron.sanpedro]

yeah but curtain panels of the curtain system can't transform into the shape of the mass like basic wall do. Unfortunately, you can't replace these panels with basic wall either like you can with curtain walls. So the panels don't bend and you get these weird segmented shapes.

But a curtain wall panel can't bend in the real world either? To me this is a pro in Revit, it makes it hard to model things that are hard to build. Makes for some kinda gut level pricing feedback.

Gordon

[narlee]

"How do you construct walls like FRANK GEHRY"

Ok, first you find someone who will pay you FRANK GEHRY fees, then you....:).

[tatlin]

By the way, was that the BD22-1 : Wild Shapes Get Real with Autodesk® Revit® Building: Design Buildable Free-Form Shapes class by Matthew Jezyk. Unfortunately the session handout was never published on the website, so if anyone care to share it with us, would be great.Hmm. for some reason the handout and lecture notes were never published to the AU site. Let me look into that with other colleagues at adsk.

Please send me a PM if you'd like a copy.

thanks,

[L Wood]

Can’t beat around the bush with this issue. I love Revit… but it is not the opportune program for making Gehry-brand forms. The programs visual and form limitations affect the brand of Revit and of its users (moi). It’s a hurdle when marketing to students who are scanning multiple building types and using multiple building programs to produce project visuals. Revit can leave a bad taste in your mouth when put to a modeling taste test with Maya, or 3d Max.

Why Revit? I see the light at the end of the tunnel in the project pipeline efficiencies Revit provides… that is its trump card, and it pencils out in a huge way. Hence, I love/live Revit.

Looking for help for a similar problem, the attached Revit model came up in the forums. I don’t remember who shared it, but it’s been my “precious” for a few weeks. Making forms in Revit is difficult and inefficient but doable through massing and formulas...

(if it was easy, wouldn’t it be boring?)

[WolffG]

I love Revit… but it is not the opportune program for making Gehry-brand forms.

Kinda begs the question......why would you want to? I think there have been enough Airstream wrecks to go around.

[rupertv]

to me it is a case of have the creative freedom to "do anything" that my creative mind comes up with...if its buildable or not (or should that be cheap or not) i think is something that the designer should be in control of, not the software you use...i have found that doing anything moderately "off-centre" in terms of design sometimes needs a very frustrating work-around in Revit...

so the point is....let me (and my client) decide what we want to do and build...not the software i'm using....

[steedlei]

form.Z is the answer!! just make any shape in form.z then import to revit as generic model. Only disadvantage is you cant apply revit material like hatch pattern on section to imported model.

[ajayholland]

Looking for help for a similar problem, the attached Revit model came up in the forums. I don’t remember who shared it, but it’s been my “precious” for a few weeks. Making forms in Revit is difficult and inefficient but doable through massing and formulas...
(if it was easy, wouldn’t it be boring?)

It is a Phil Read construction, for sure. Search his posts (http://forums.augi.com/member.php?u=53389) for some other interesting models.

Phil’s imaginative work with Revit produces results that are ordered by the underlying mathematics that drive the families. The implication is that if it can be built in Revit, it probably can be constructed in reality.

When I visit architecture schools, as I did SCI-Arc last week, I am always amazed by the creativity and innocence of some of the design solutions on exhibit. Who is to say what can or cannot be built?

It is usually up to those who do not recognize the boundaries to ‘push the envelope’, and this in turn becomes the impetus for further technological innovation.

~AJH

[AP23]

There is no doubt that Autodesk doesn’t have the knowledge to develop such tools that are available in other 2d and 3d packages. The reason why, to this day, we don't have any of these tools is, because the majority of Revit users don't need them. These users prefer customizable elevation tags and an advanced text editor than nurbs.

People (and students) that do need to construct 'gerhyisk' buildings simple don't use Revit, but at the same token, the majority of these people use other Autodesk products. Financial wise, it wouldn’t matter for Autodesk.

[Joef]

I'm not sure which program actually has a full set of modeling tools plus the ability to produce a coordinated set of construction documents. Perhaps someone knowledgeable in these things could give us a list. Should be a fairly short list. Then perhaps we can get some users of these programs to tell us about them. I'm curious why a fully functional BIM program with NURBs and all that stuff isn't what we are all using instead of Revit. Did we all miss something here?

[knurrebusk]

This is why I feel Revit is on an intentional hold on!
Everything related to a more integrated BIM etc.

DWG is a dead horse!, it´s a useless labyrint of errors.
Still this stuff is the only export option from Revit.

Lightscape was a brilliant start in Radiocity renders, indoor lightning.
Only leftovers made the way to Max, slowly.


I could go on and on, just tragic greed combinded with a blind sheep heerd.
Sorry to say I´m one of the sheeps, since Revit make you addicted.

[mmodernc]

http://www.gehrytechnologies.com/

[knurrebusk]

This is not gonna help us Revit addicts!

Revit need cleening up from Autodesk, not out.

[Joef]

http://www.gehrytechnologies.com/

Doesn't do construction documents. From a case study:

"At the same time, 2D and 3D plans extracted from the CATIA model were detailed in AutoCAD to create contract drawings and specifications."

[knurrebusk]

I get depressed by this.
We should be further in front, the reason I´m very dissapointed with Autodesk.

Everybody told me to stay away from the bad wolf, but a wolf need a leader.
If Autodesk do not find a new Leader of the pack, time out.

Will be crushed like nothing very fast, this is not oil/gas.

[knurrebusk]

ap23

I worry if he is right, and if all my work for Revit is wasted.
I had a brother in law that told me to stop Drawing, and start selling.

If all this come together Autodesk is dead in BIM.

[twiceroadsfool]

Doesn't do construction documents. From a case study:

"At the same time, 2D and 3D plans extracted from the CATIA model were detailed in AutoCAD to create contract drawings and specifications."

It certainly does produce Construction Documents...

But let me tell you, if people complain about workarounds with Revit, then they dont even want to bother trying to document a building in DP. The third release made a lot of headway supposedly, but i tell you... Unless youre getting Gehryesque fees, theres no way youre turning a profit using that software.

Its an animal to document with...

[mmodernc]

Revit won't go away. But I wonder why complex shapes are not as forthcoming as in a NURBS software. Is it because it is a sold modeller? Autocad solids were a bit limited. Also I have found a lot of people are set in the way their documents look. So you need a lot of tools and flexibilty to turn a revit model into autocaddy drawings. Which are happening judging by last release -

[clog boy]

Let's be honest. The traditional way of constructing is also the most cost-efficient way: straight, vertical walls, rectangular floors, etc.
But let's say one would model a building in Maya, wouldn't it be possible to import it as a mass and do 'wall by face' etc? It would be great if you'd be able to model the Sydney Opera House from scratch using only Revit, but the truth is that's a one-off design with a very big budget.
We'll be stuck with drafting shoeboxes for the moment, since 90% of all buildings are fairly rectangular. Great designs require a great budget, and I think one would be able to design 90% of them also using Revit, but only if you're a great CAD-drafter.

If it would be easy to design innovative and yet cost-efficient structures, everyone would do it. Right? ;)

[AP23]

If it would be easy to design innovative and yet cost-efficient structures, everyone would do it. Right? ;)

They do. Just drive 50 km to the west and you’ll see that even little power substations and small bridge houses have a non-orthogonal shape, mostly designed by unknown young boutique firms. Or you can buy a copy of A10 magazine (a magazine cluttered with very complex buildings designed mostly by small to medium sized unknown firms).

So it is not only Mr Gehry or Miss Hadid with an unlimited budged doing gehryiks buildings. It’s also the thousands and thousand of boutiques firms and millions of architectural student’s world wide.

This “90% of buildings can be done in Revit” is really getting old and does not justify the absence of advance modelling tools. And while the idea to have every single “rhino-like” modelling tool in Revit would be to good to be true, it would be at least useful to have a few tools to be able recreate an idea in Revit that have been initially created in Rhino or FormZ.

[knurrebusk]

Agree!
English is not my native language, so I can´t top this.
Like someone ripped my instinct out of my head, and made words.

[twiceroadsfool]

This “90% of buildings can be done in Revit” is really getting old and does not justify the absence of advance modelling tools. And while the idea to have every single “rhino-like” modelling tool in Revit would be to good to be true, it would be at least useful to have a few tools to be able recreate an idea in Revit that have been initially created in Rhino or FormZ.


I dont even disagree with you, but our current options go something like this:

1. Revit, which can document most things, but struggles in modeling complex forms.

2. Digital Project, which can model just about anything, but is an absolute and utter nightmare to document with.

3. BS modelers, that dont document anything at all.

Im all for building a glorious 3D model, but you know what? Making a pretty picture and not being able to have it built doesnt interest me a whole lot...

[Steve_Stafford]

I think I finally understand, only dull dim-witted architects can possibly get by with Revit now. The "clever" architects out there doing "real" architecture have to continue using some other "clever" software until somebody at Autodesk wakes up and realizes they aren't hanging around with "clever" architects enough.

No offense to the dull dim-witted architects who like using Revit now, you just aren't cool enough to realize your plight. I just realized that makes me dull and dim-witted too... :sad: Well at least now I'm enlightened.

[gwnelson]

Steve, you do the dull, dim-witted client a real disservice with your lack of perception.

[knurrebusk]

Yes!
In many ways you just shoot your own foot.

Still Revit is my base, perhaps to much.
All of my best work has originated from Revit, because I not worry.
Never made a single huge mistake with Revit.

Perhaps we need a suisidal Revit (like extreme alfa code)

It´s like Sentryglas!
If we could bend it Whow!

Perhaps it´s possible, never read the white papers.
Dupont supriced me before.

This is where Revit has stagnated, perhaps all the effort should be put into Inventor.
Pro engineer is my nightmare, Inventor is opposite.

[mmodernc]

But I wonder why complex shapes are not as forthcoming as in a NURBS software. Is it because it is a sold modeller? Autocad solids were a bit limited. -

Here I go quoting moir again.
Could we have an Atuodesk answer to this one please?

[clog boy]

They do. Just drive 50 km to the west and you’ll see that even little power substations and small bridge houses have a non-orthogonal shape, mostly designed by unknown young boutique firms. Or you can buy a copy of A10 magazine (a magazine cluttered with very complex buildings designed mostly by small to medium sized unknown firms).

So it is not only Mr Gehry or Miss Hadid with an unlimited budged doing gehryiks buildings. It’s also the thousands and thousand of boutiques firms and millions of architectural student’s world wide.

This “90% of buildings can be done in Revit” is really getting old and does not justify the absence of advance modelling tools. And while the idea to have every single “rhino-like” modelling tool in Revit would be to good to be true, it would be at least useful to have a few tools to be able recreate an idea in Revit that have been initially created in Rhino or FormZ.

Just needed to hear someone say it out loud. Thanks.

Yet I still think complex shapes alone don't make for good architecture. It's shape and function. If someone would extend a wall diagonally two ways and cover it with bronze, while the perpendicular wall is placed four feet to the back and made from glass, that's architecture in my text book without using too many complex shapes.
It's the spice that makes a meal taste good, doesn't matter how (in)expensive the main ingredients are..

I predict more complex designs will become possible as they've been over the past few years. They just don't need to compromize the BIM concept.

[john.s.algeo]

Frank creates models, then digitizes them and brings the digitized model into CATIA. CATIA is extremely powerful and not very intuitive. I worked with it on the Walt Disney Concert Hall for the exterior metal panels.
I have modelled some multi-curved surfaces with the Massing tools and then applying walls to the surfaces. I would like to be able to find the rule (the straight line in a curved surface) but, unless you offset a curtain wall from the Mass face, Revit will not automatically find the rule.

[knurrebusk]

I wonder if a 2D sketch on a curved surface is possible inside the current Revit code,
or nurbs is the only way?

Modo 301 is a very strong polygon modeller, but is not delivering magic like
Vue Infinite/Xstreme do with replicateing nature.

Terragen 2 is getting old in fact!
Still beautyful, but a toy.

[clog boy]

Terragen 2 is getting old in fact!
Still beautyful, but a toy.

Word. But it still makes for pretty landscapes ^^

I miss the Edgesurf functionality we had in AutoCAD, if it would be implemented in Revit I'd like it not to be limited to 4 baselines. In fact, hexagonal cells would make for smoother surfaces.

Edgesurf surfaces can be calculated with, theoretically, but they would be a huge processor hog if you rely on them too much.

[tomnewsom]

Just a quick note - I sliced a cone in the family editor, and was able to use the resulting section as a work plane and picked the cut edges to create an extrusion. So parabolas are possible to create in revit. Just not very flexible!