WHY ARE #REVIT FAMILIES SO DYSFUNCTIONAL?

This has been an issue in the use of Revit for quite some time.  I and others have questions like:

• Why is there a category for Nurse Call Devices but not for numerous other needs?
• Why can we not control what categories or instances we need to be cuttable?
• Why do even cuttable categories work differently in various categories?
• Why so much variation between the family categories in their ability to display?
• Why does sharing a family change the rules of how it displays?
• How are we to even know all the different results possible from each category?
• Why does it need to be impossible to utilize the categories given to successfully complete a complex project?

This all comes together to ask the biggest and most crucial question:

If we are using BIM to create a virtual building to better the process and understanding of the components of the building, then why can we not trust the representation depicted in the families based on the preconceived determination of how different families are shown. Should they not all be a true section of the components wherever you are cutting them?  If not, then the process is faulty and prone to error.

So, lets take this issue to an example of work I am doing for a current project. We have terracotta and other materials that we want to panelize on the building but we want to see the actual cut through the material. If I have a thicker piece of terracotta than the others, as shown in the example below, then in most cases that piece will mask the piece being cut. These are all nested pieces of terracotta placed into the panel family of the same category so that they can be scheduled as their pieces as well as the whole panel.

So with this piece, lets look at some various categories that could be used for this.

Click Image to Open PDF

Here is a closer look at the two categories that are the closest to the true need of being truely cuttable.

The results are that the only category that can be trusted (other than some structural categories) is columns as long as you uncheck the show precut in plan option.  Generic model works until you share the nested components and a non-cuttable family like furniture will almost work when you share the nested components as it will show a representation of the actual nested component but will not actually cut it. So, because of how Revit is designed it leaves you between a rock and a hard place with the decision of what odd category should be used to represent what you need.

Anyone else have any thoughts or suggestions?

Christmas Gift - Stair Planning Family

Ever get frustrated creating a stair in Revit when you know that it will change a dozen or more times before the end of schematic design?  This family will give one the ability to create a schematic stair with the knowledge that it will occupy the correct amount of space with the most minimal time to produce multiple options.  No more fiddling with the stair tools within Revit before one is ready to.  I am giving this to all who want it as a Christmas Gift.  Go to the download section of the site or click here.

This was easy to produce after the effort that went into Fuzzy Math for Autodesk University and make sure to thank the BIM Troublemaker for making the comment during the session about how a planning family could be very beneficial.

Revit Stair Planning Family from Jason Grant on Vimeo.

"Super Revit Families" - Are they Good?

David Fano from Case and Design Reform brought up a good question on my last post about user research, "Have you found "super" families to be lighter than separate files?"

The simple answer is yes but I to had my reservations about them also when I first started creating super families.  Will they create a heavier model or slow down regeneration?  So to figure out the answer since testing this type of thing is difficult to create an accurate base line, I gave a shout out to some of the guys at Autodesk.  This is part of the response I recieved:

In families I would break down performance in to two different areas; graphic regeneration and model regeneration. The file size is only indication of the performance impact, it is not actual impact.  As an example a highly complex model might be large in file size but the amount of faces is what actually effects the graphic regeneration.  So if those faces are not visible in the view the performance hit is minimal. 

My question to them had included whether or not 1 super door family that was 2MB would be more or less impact than 10 doors that collectively added up to 5MB.  Based on the response above, I would assume that since none of the faces of the other nested elements are shown then the performance should not be affected.  I like to think also that the overall size does make some type of a difference but who knows.  That is just one of the mysteries of Revit.

The other comment I received was from Troels Olsen and it covers the other side of "Super Families" that I would like to discuss.  His comment was (in a shortened version), "One thing to consider here, is to create families that aren't overly complex. Partly because they may become heavy and prone to break, but perhaps more important, that they become so complex that it takes an expert to handle them."

The door had been my first super family and I have since become pretty efficient at providing enough flexibility to allow for user control without getting into the nested elements and I am also doing stability/testing to ensure that they do not break.  Usually I create the family two or three times before release to make sure that the path for it to be effective is the right one.  Do you use check boxes, family labels or dimensions to control the switching of elements?  But another question remains as to why create something that takes a lot of time and planning?

I believe that if created effectively, you give the users full control over adapting the element while at the same time ensuring that they all schedule and look correct.  Once done correct, it saves me a lot of time as a BIM Manager to change an aspect of the door or to add a parameter for scheduling. The user also, for example, has the ability to have different swings, different door panels and mix and match anything they choose.

One of my recent accomplishments of "super" families was for a university project that had an addition to a 1900's building where the existing windows needed to be modeled.  After reviewing the project, I realized that there were 14 types of windows and these could not be static families since each floor had different window dimensions.  Therefore, I had to decide if I was to create 14 parametric windows or one super window.  I decided on the super option and a day later had the finished product.  As you can see below, I have kept the parameters that the user adjusts to a minimum.  These include: Number of windows wide, Window size (Width + Height), Transom Size (Height Only), Transom (yes/no), Double Transom (yes/no) and a few random others.  I hid all the complex parts within the nested family.  Now, if the team comes back with different field dimensions, I only need to change one or two profiles in one family and not in 14 separate families.  A little food for thought....

What the user needs to change to use.

The Parameters hidden within.

The different options of the window.