Railings High Level

From my previous post its clear it will be difficult to hone in on any single aspect of a complex feature before hitting some high level items.  In comments on the previous post I listed some high level issues with the current railing feature:

• Indirect UI - Designers do not want to work in a spreadsheet.  Choosing a rail, clicking "edit",  then seeing all the geometry reduce to a single pink line is almost funny but I will refrain from laughing.  The baluster and rail dialogs need serious work or should be replaced by direct editing and simple properties.  Making a railing from scratch is the most difficult thing in Revit to do.  I think I am comfortable saying this.
• Missing objects - While generalized can be good the current tool is too general.  Handrails may take many forms but most railings have 0-2 and they have a relationship to the railing or host plane (grab distance).  Panels have been described as a "nightmare" or "a living nightmare" mainly because there is no concept of panel!  Rails end with hardware such as a rosette or even a simple radius cap.  Others are supported by brackets.  You might want to specify this or even count them.
  
• Too rigid - 13 types are often required to model a rail where one or two should do.  Making small tweaks needs to be supported once a general pattern is established.
• Bad geometry - rails are usually continuous so there is a need to get it right or provide a means to shape the rail path.  It also needs to consider the path is made of components.  Yikes it keep getting more complicated.
• Inputs - railings have codes and we need to ensure all the proper distances can be referenced and dimensioned and critical spacing can be input without too much math.  This means dimensioning baluster spacing in plan, extensions, height ect...
• Hosts - not enough of them.  Rails can not just go on stairs but tops of walls, topo surface, roofs, between columns ect...

Any more?

Next: Scenarios are extremely valuable.  (See this post for more background)  Scenarios provide the goals which can help inform system behavior.  I got a nice one yesterday:

When laying out a handrail you may start with support brackets at the ends and then every n'.  Now you realize there is an obstruction or support in the wall that requires relocating a bracket.  So you move the offending bracket 1' to the left.  What of the remaining brackets?  A design goal is usually to distribute the remaining brackets so they look considered and harmonious.

The nearest precedent in Revit are the soft division snaps you get when placing curtain-wall grids.  The system tells you where the 1/2 and 1/3 divisions are in each segment.  Thank you Revit - pat pat.

The key is understanding the interplay between the reality of varied building conditions and the designer who likes to impose pattern, symmetry and organization.

Please share any real stories if they come to mind.  A past project, a specific condition you solved.  The story will contain a wealth of information that will make predicting deficiencies in the existing tools easy and communicating them unnecessary.

BTW this blog gets a lot of long comments.  I'm always impressed by the time people contribute.  Thank you.

-erik

Comments

[Mathew Miler]

I was just trying to understand the current system better and was looking at the AU presentation Phil Read did on stairs and railings list last year. He talked about using curtain walls instead of a railing system at times. This gives us the graphical interface designers need.He also talked about that in version 1 that to create a railing system it was done with the curtain wall. That's my 2 cents.

[Erik Egbertson]

Curtain wall comes up a lot.  In a lot of ways its superior but in others it falls a bit short (curves, stairs, corner conditions ect..)
I also think that while you can override curtain wall type properties there is a fair amount of monkey business involved.

Simply offsetting a mullion might mean select mullion, unpin, delete, add grid, click mullion tool, choose partial segment, click on desired segment.  I think this could be streamlined.

Nesting is powerful and certainly cwall is much better when panels are involved.  Your 2 cents are valuable.

[Adam Sheather]

I think the new adaptive components will allow for these more complex railing and post systems especially using the conceptual tools as an outline.

I thing I still don't understand with all the new conceptual modelling tools and even with inplace modelling you still can't create inplace stairs or railings in Revit??

Why can't we just be allowed to create any object and give it whatever catagory we want?? I could make parametric inplace families for railings and stairs like I have done with walls!!!!

I hope 2011 has allowed this for the new adaptive objects and massing upgrades. IFC'ing stairs is currently the only way to get them in the right catagory and anyone who has used IFC knows the geometric translation is pretty hideous.

Even if we could create curtain systems and give them another catagory instead of just walls would open up alow for scheduling they are so so powerful and so underused. Suspended ceilings work great for panel scheduling, railings, stud work, precast (on occasion) special feature wall constructions I could go on.

It's already there it just needs someone to allow these objects to be catagorised correctly. Let us deal with making the document looking correct later but having it model and schedule alone wouldbe worth it.

Fingers crossed for 2011!!

[Michael Ruehr]

I agree with the Above
I use Curtain Walls for What they where made for and for Windows, Doors, Toilet&Shower Partitions, Kitchens, Interior Bathroom Tiling.....
Why
Curtain-walls are a great reference Object i can transfer and move complex assemblies with one object.
Eric what you mentioned as downsides looks minor to me, and could be solved as with Curtain walls with rule and non-rule based Types.
With Railings most of the time is spent on Thinking and Head Scratching.I would think With Adaptive Objects it should be possible to make a hosted Assembly Object (Curtain Wall)that can be assigned to different Categories and have some of the current restrictions solved (Non rectangular Panels and Rails that will not be cut by vertical Grid-lines...Boy this would make me happy

[Erik Egbertson]

The new tool certainly offer a lot of promise and approaching these similar problems with a generic solution that can place categorized content is compelling. There is more movement in 2011 and I expect this will continue.  The family editor did not change for years and it will not be replaced overnight.  The effort there is quite large hen you start breaking it down.  Im curious to know what percentage of the Revit users in your firms are using the new modeling tools.  In class we train studetns to use them and the learning curve is not too bad once they understand the new rules.

[Erik Egbertson]

Good to know what would make you happy.  Adaptive components are really cool.  High level a skeleton that can drive this type of content would offer a lot of power.  Can you elaborate more on the Thinking and Head Scratching?  Is it deciding the style?  Solving contextual conditions.  I want to read your mind.

[Michael Ruehr]

Eric
Thinking...What style of stair..
What is the best style railing going with the overall design and stair-style...
Giving options to the design team (review)....
How can i fit the stair in a always to small space...What material should be used...Estetics...Codes

Head Scratching....
How do i fix, support the Railings Posts...
Side Fixing, Bottom Plate,Grouted.....
How do i deal with the Rail on Winders
where do i stop with a Baluster Panel
How best to divide the Baluster-panels to have the same Size...
what do i do with the end Conditions...Space left..
Do i have better Alternating Patterns or do i want a makeup panel in the middle of the pattern or the end or both ends....i want an easy way to divide..see the result change quick...
Transitions and end conditions are prone to lots of problems, options... automation will most of the time fail...
Fixing is often random or does not follow the baluster pattern...Support of independent patterns would be great..Supports or block-outs are shown in plan-sets that do not show the railing so nested Symbols@Details are must...
Opacity-maps for Panels would be useful for very ornamental Patterns to reduce Model overload.

just a idea.. hosted Splines that can be used for Adaptive Components that can generate Child or independent Splines can then be grouped together as a Railing.The Splines could generate Hosted Point arrays for easy division(Rule Driven and/or Manual)...

[Erik Egbertson]

I like the way you think.  Yes I think a skeleton of sorts that can organize adaptive components is compelling.

for panels 2011 will support textures in non rendered views and I think you can make custom frit patterns using a generic material shader.  Ill try this out.  I agree fine detail can be handled in this way.

[JoeF]

Adaptive Components have been mentioned a few times in this thread. Anyone care to share a Revit example of this?

[Adam Sheather]

Adaptive Components are a 2011 feature so we have to wait a little longer to get our hands on them.