Rail Connections - Limitations

[rkuffel]

The Help menu reads as follows under Rail Connections,

"If Revit cannot create mitered joins when connections are made between railing segments, you can select one of the following options:

• Trim. Segments are cut with a vertical plane.
• Weld. Segments are joined in a manner as close to a miter as possible. Welded connections work best for circular rail profiles.

Can someone help me understand what is meant by "If Revit cannot create mitered joins?" Also can someone explain what is meant by "...close to a miter as possible?" I keep running into mitering and connection issues and it would help to know what the known limitations are instead of trying to figure it out by trial and error. It seems so simple to just extrude a railing along a path, but obviously that is not what Revit does. I could do these railing in about 5 minutes in Sketchup.

The impetus to this question is shown in the attached images:

[parveen.revit]

Revit has very limited capabilties in railings. It has been reported severa ltimes.

Sometimes, it won't be feasible to get the desired results.

SketchUp works on very different principal and programming platform. It's not fair to compare it wtih Revit.

[rkuffel]

I understand that Revit has railing limitations. My question is, what are those limitations? In the many years I have been using Revit, I have never seen any detailed explanation as to what those limitations are. The hints/help are vague at best. The problem is if a user doesn't have any frame of reference (known limitations), other than days and days of trial and error, than one doesn't know when a limitation has been hit or if there is just a parameter off or something slightly incorrect, e.g. a parameter setting.

I challenge anyone to quantify the limitations of railing connections. My guess is the only people that know what those limitations are, are the people at the factory.

P.S. I am a huge advocate of Revit, but I can't keep telling the many Sketchup users in my office who are floored by the complexity and limitations of certain tasks in Revit, that it isn't fair to compare the two products - that's just not a compelling argument. Frankly, I think they are right; a freeware that has better, more intuitive modeling capabilities...really? 

[parveen.revit]

Revit has limitations in dealing with railing but with experience, people has developed different workaround to overcome these hurdles. Autodesk provides a revit model wtih different compositions of railings and fences. my team has also developed two document in powerpoint and MS Word format to resovle these issues.

As far as SU is concerned, it is not a true BIM and you can toe concept skecthes easily. Comparision between Revit and SU is not fair.

[rosskirby]

Challenge accepted.  In my experience, the limitations of railings have typically been:

• with the joins (as in your example), especially in cases where you're truncating the pattern and the railing has to make some tight turns
• the inability to have a single railing extend over multiple levels and wrap back over itself (can't have overlapping lines in sketch mode); this is also true for stairs
• the inability to accurately or easily create complex railings, especially with respect to baluster placement and infilling patterns
• baluster panel families tend to "break" if the length of your railing isn't an exact multiple of the width of the baluster panel (i.e. using a 4' baluster panel in a 10' long railing means that you get a 1' empty space at each end because the baluster panel family won't truncate itself to fit)
• the inability to host a rail to anything other than a floor, stair or ramp (i.e. parapet railing)

That being said, the railing join problem in your post has been significantly improved in the 2013 release with the use of component railings, which are also much more easily customizable for complex situations and work better in multi-story applications.  So that's 3 of the issues.  I haven't had a chance to test the baluster panel issue in 2013, so I can't speak to that, but I'm pretty sure walls still can't host railings.

While your SketchUp users may be enamored with simplicity and ease of the modeling capabilities of SU, what they should be aware of is that SU is not an architectural modeling program, at least in the sense that is not geared specifically towards the architecture industry.  It may be easier for them to create the geometry in SU, but in terms of producing a full set of construction documents, and thus a building, from SU vs. Revit, Revit is going to win hands down every single time.  

Quick, make a door schedule in SU.  How about an occupant load or life safety or minimum plumbing facility schedule?  Wall sections with materials, details and keynotes that are all automatic?  No? How about something simple?  Insert a door into a wall.  Model a true curve, or a spline.  How about alternates?

That is why it's not a fair comparison.  They're two completely different programs with two completely different purposes.  Can they be complementary?  Absolutely.  Are they interchangable?  Absolutely not.  

Just my two cents.

[rkuffel]

Thank you very much for the reply.

[fg]

Yup, seven years later - not a lot of progress even though railing have been "revamped" several times.  I think a lot of the real limitations for railings stem from Revit's limitations with sweeps in general, and a lot of these stem from some very deep and hard coded limitations Revit has with precision and rounding. Revit will not in general sweep a profile where the curve returns into itself; and  because of "rounding errors" that occur with angles and curves, even if your geometry should theoretically clear itself - Revit fails to make the shape.  Unfortunately there is zero documentation or even acknowledgement by AutoDesk of these limitations.  The painful workarounds that experienced users have unearthed via trial and error typically involves extending paths, tweaking fillet radii, adding to offsets, etc in amounts just enough for Revit to finally resolve the geometry, but not too much to destroy the design intent.  Or they finally give up and build it in Rhino