MASSING - Face direction and orientation

[Anonymous]

It is very confusing to figure out how Revit massing determines the direction of a face normal and a face orientation. While Revit does have tools to flip surfaces, they do not always work and it is best to just model properly. However, there seems to be no official guide to modelling and making sure you get surfaces that are in the right orientation and normal direction.

Did I miss this guide/documentation? How can I tell users how to model in a way that will give predictable outcomes?

[barthbradley]

Funny, I was asking kind of  the same question a number of years ago.  Did you know that Ref Planes can be drawn upside-down?  Serious. 

 http://bimextension.com/revit-families-hosted-to-reference-planes-are-upside-down/

[Anonymous]

Yes, I did know that. That's actually pretty simple compared to massing, reference lines, etc.

Here's a tricky questions. Let's say you want to draw a square face, so you draw two reference lines. One higher than the other. Then you select them and make a form (face). How the heck are you supposed to figure out which way is front and up?

Am I right to think this is totally undocumented, but a very complex topic. I feel like 3 programmers at Autodesk know how it works and they never told anyone. And since the Revit massing is so broken and unpopular, no one has bothered to make a fuss out of it yet.

What a mess.

[Anonymous]

We discovered today that i you split a mass face into a number of segments, the normals are not affected. If you divide it by named reference planes however, the normals are flipped....

I have an employee trying to figure out all the hidden undocumented rules that will allow us to do massing without normal nightmares and she is finding that it reacts in a very illogical way. Doing the same actions results in different behaviours in different models.

This is both extremely frustrating and disappointing for a 'professional' architectural tool like Revit. I hope Revit Autodesk will give massing the attention it deserves so that it can become a functional tool.

Here is a link to vote for Autodesk to finish the massing tools properly.

https://forums.autodesk.com/t5/revit-ideas/revisit-massing-and-panel-tools-give-them-some-love/idi-p...

[Anonymous]

Here is some additional documentation on the issue of the normal flipping when dividing by named reference plans:

---

Notice we are using the Revit Normal diagnostic tool from Zach Kron, you can see it flip around for no reason.

http://buildz.blogspot.com/2011/03/that-aint-normal-diagnosing-point.html

[Anonymous]

Long story short: Autodesk needs to correct all of coding errors inside the Massing tools. There are some unforgivable miscalculations that happen with orientation, family instantiation, and other inconsistencies between family and project. Autodesk, can you please fix this by the next release so that these tools can be used effectively? There's really no excuse. It's almost really awesome, and it would be great to see some genuine improvements with Revit.

[adrian.esdaile]

Except... if they fix the coding errors in the Massing tools, guess what will happen to your nice complex model when you upgrade it to the 'new' version? It will break.... or not even be able to upgrade at all!

I've just done an 8,000 panel multi-curved facade out of Adaptives and the rules you have to learn to get it all to work are just nuts. If you draw lines anticlockwise, normals face the expected direction except if you're on a curved surface in which case the normal will be the exact opposite, and the normal at one corner will be pointing off that way instead of like the rest, except if it's a Tuesday afternoon between 3.14pm and 4.47pm in which case the normals are positive on clock-wise generated sketches, except if they have an odd number of segments in which case the normals are correct except the 4th normal which is reversed (but only if the date is divisible by 3) and you have your computer's time format set to Ethiopia (dawn is 00:00, midday is 06:00). And all that stuff that worked yesterday? Well, that's yesterday's rules; today's rules? Different. Seriously, it's that bad.

If Autodesk were to suddenly make Adpatives behave in a consistent and reliable manner my nice detailed facade is going to get trashed because I've had to rely on the daft, unpredictable, illogical and imprecise Adaptive system as it currently exists.

Remember when they upgraded the way Spot Elevations worked - and it broke all your spot elevations? I remember it, it was a fun night until 4.30am! Lessons learned: don't upgrade if you're on a deadline; expect problems when you do upgrade. (around 2006-2007)

Remember when they changed the way Window families worked? And it broke all your window families in your library? That was a fun couple of weeks repairing the mess. (2005 irrc)

Remember when they changed the way stairs worked? And half the stairs in your model went haywire? That was fun too! (2009 I think)

The more complex it gets and the more bugs that remain unfixed; and thus 'burnt in' to the core code (and our projects); the more Revit looks like a house of cards.

[Anonymous]

As far as I've noticed, nothing has been evolving in adaptive panels and massing since the last 5 years. I would love to have improved massing and adaptive panels even if it breaks everything.

[izo_cfmoller]

I couldn't`t agree more.
this issue has bothered me for a lot of time, going back and forth. 
through trial and error, i figured out a rule of thump within revit massing when it comes to face normals. 
Briefly, it depends on the rotation in which you initially draw the lines which will create the form of the face. 
if it is clockwise, normals face inwards.
if it is anticlockwise, normals face outwards. 
i hope this helps slightly. its a huge issue, and it is due of an update, Autodesk should address it asap. 

[syman2000]

This is why I always built the panel on center so if the panel goes haywire, it always stick in the middle. However what really bothers me is the fact that when upload model to Autodesk viewer, this is where your massing start to show thru the model. Screenshot below highlight issues within Revit and I don't think this was solved with the latest version.

Exporting FBX shows the surface correctly without issue.