Roof by Extrusion grows down?

[dbaldacchino]

I didn't realize, but roofs by extrusion grown downwards (like a floor) instead of up, like a roof by extrusion. This seems inconsistent to me and I think it whould grow upwards. What's your take?

[luigi]

There is a reason why the footprint grows upwards...and it took a while, and now I agree with it growing upward (I can now use the Top of Steel level for a roof)...and yes, I was a bit frustrated when I noticed the extrusion behaves opposite...also a bit ashamed when I was showing somebody how to do it by extrusion and told him that it would grow upward....but it didn't it grew downward :s

Got my vote...it's too inconsistent

I didn't realize, but roofs by extrusion grown downwards (like a floor) instead of up, like a roof by extrusion. This seems inconsistent to me and I think it whould grow upwards. What's your take?

[chodosh]

Jinkies! Good catch. I'm with you on this needing to be fixed and behave consistently one way or the other for roofs in general, but gravity usually wins, I'd like my structure to stay put and the assembly thickness to change as it would in the field more likely than not, beams and rafters below the roof, not within the assembly.

-LC

[dbaldacchino]

haha Luigi, that's when I find stuff out too, just when I'm trying to show something to someone and then look like a goober. This time I was working on modeling our house in my spare time (found some under the cushions on the couch) and was doing a round dormer with roof by extrusion and was expecting it to grow out and then figured I'll just offset the sketch. I kinda remembered that this is how it acts but at work, we seldom use the roof tool and we mainly default to the floor tool. Sometimes inconsistencies just jump out at ya :)

[sbrown]

The reason it grows down is because you are drawing the top edge of your roof, you can't draw the bottom because then how would you change the roof end conditions. And typically with roof by extrusion you are drawing in elevations so you would be sketching the ridge to align with another ridge or something like that.

[dbaldacchino]

Well true, but as long as you know the roof thickness, you can always offset the sketched line. I guess this is a topic that can be argued both ways. In my case, I knew what the inside dimension is (measured the dormer on the interior) so I wanted the roof to behave like the roof by footprint and grow up. It just seems inconsistent to me although I understand your argument.

[sfaust]

I don't see how drawing the top or bottom has an effect on the end conditions...

[patricks]

I rarely use roof by extrusion, but sure enough, EVERY time I have used it, I ALWAYS end up sketching the bottom surface (because I know where I want that to be), and then I have to offset by the roof thickness. It always happens for barrel roofs, as I know what the bottom of the roof is sitting on and its radius. Then I always have to offset upwards, and its annoying as heck if anything changes.

[sbrown]

Sounds like we would like the option to choose a location line for roofs, just like we'd like a location line for ceilings and floors. ie core boundary that doesn't move when we swap types.

[dbaldacchino]

I concur, that would be nice. Basically anything built with layers could benefit from this. In the case of a roof or a ceiling, having flip arrows or something similar would just change the offset rather than mirror the object.

[chodosh]

That'd be excellent to have a location line in the assembly. May resolve this issue... the quirky thing about the different behavior between extruded and by footprint is that it appears that in some analysis software, the walls that are attached to a roof by extrusion do not behave correctly, possibly because the assembly geometry is touching the bottom of the roof that is not the "gravitational base," so to speak. Has anyone else seen differences between the two roof behviours in analysis tools like Ecotect or IES, etc.?

-LC

[dbaldacchino]

Hmmm haven't tried that in VE-Ware. That would be a good experiment, although the shape-edited roof by footprint is not without it's share of quirks :)

[chodosh]

Yeah, no kidding, ...and we all now know you gotta have a roof at the top, or else, so getting this geopara-gravitation roof assembly thing workin is, well... kinda important so we don't keep these lists of "how to tweak your Revit model for exporting to _________" (fill in blank with any of the following: ecotect, equest, IES, gbxml, etc...)....

So, the question is still out there for someone at A'desk: why do they behave differently and is there a way to adapt this to a logic that is shared between both roof by footprint and by extrusion to satisfy everyone's needs and still have a consistent behavior for roof elements that we can count on? Is the location line the solution? Wish list item?

-LC

[dbaldacchino]

Go ahead Laura, articulate it and post to the wishlist ;)

[chodosh]

Done! Thanks for getting this discussion going, D.
Now stop picking on Steve.
-LC

[chodosh]

I tried to cover this with a twist relating it to similar behaviours of other systems in Revit and it made it in the list this cycle, you can comment HERE (http://www.forums.augi.com/showthread.php?t=91730) on whether or not I got it right... Believe me, no hard feelings if you don't agree with me.
:)
-LC

[dbaldacchino]

Sounds good! You have my vote ;) If there's any verbiage you want to change/adjust let me know. We can still edit the wishes before becoming part of the ballot.

[ws]

FWIW I use roofs by extrusion a lot - and while I find it more logical to set the top of the roof and have it offset down, I can never remember when I'm in the middle of an edit which way it will go after finishing.
More helpful would be to keep a ghost outline of the roof thickness while editing so that you could see the relationship of line to roof thickness.

Something similar occurs in Nemetschek's Allplan whereby roof thicknesses offset upwards from the roofplane but rafters offset downwards - logical I guess but again hard to remember.

[aaronrumple]

Why do we have to chose? Walls have justification. Shouldn't roofs and ceilings also ahve the same justifcation options?

[cliff collins]

Justification sounds like a great idea.

[chodosh]

There, that's the missing verbage, Dave: justification. And, that's what I was trying to get at. Walls can be justified by their assemblies. Roof have assemblies, too, so why do they behave more like simple booleans without structure? You guys nailed it.

[dbaldacchino]

Cool, I'll msg you with the proposed changes and we'll sort it out.

[mlgatzke]

Revit has never created residential roofs "correctly". I have ALWAYS needed to use a section view to move my roofs to their correct height once they've been placed. Truss manufacturers need 3 pieces of information from a roof: 1. pitch/slope 2. overhang 3. heel height. This 3rd item is where Revit has always failed to deliver, but I don't understand why. The "heel height" of a roof/truss is the height along the exterior wall structure (usually the sheathing) from the top of the wall's top-plate to the top of the truss. In my area, because of the cold winters, we have a minimum heel requirement of 8 inches. This allows for adequate insulation to be maintained along the entire surface of the ceiling or rafters (depending on your chosen method of insulation and insulation material).

I've never understood why Revit has never supplied the heel-height as a part of the parameters of each roof line. I just create my roof, cut a section through it, and move it to the "correct" height required. Not very Revit-istic, but it's the only way to get it to work.