Monday, June 8, 2009

Top Ten Architecture Modeling Mistakes

We've all heard it before. "The architects models are worthless and unusable." So here's my first crack at a top ten list of things architects need to avoid when modeling that inhibit further progress when dealing with contractors and consultants (i.e. structural, mechanical, electrical, plumbing, GIS and so on)

Top ten modeling mistakes architects make:

1.) Alignment - This is uber important. Especially when using the model for IES, CFD or other analytical tools. Floors connect to walls, parapets extend past roof lines, footings align with slabs or are integral to them. Oh yeah and windows and doors need to be modeled as constructed. Unfortunately a 500 lb aluminum storefront window can't bear on only the substrate and exterior finishes.

2.) No generic modeling. Or at least don't share the models until they have material and intelligence otherwise the models are distributing maybes and kind ofs which don't help any body and in reality can really confuse the team.

3.) Model as constructed. Walls are not 150 ft tall. Floors are not 1700 ft long without joints. Pre-cast panels are just that, panels. Also when concrete floors and deck pans are modeled they need to be modeled to their exact dimensions. Also ceilings have wire supports, which can easily be worked around pipes for the most part, however pipe hangers, duct hangers and other supporting appurtenances need to be modeled. Use best judgement.

4.) Don't use design options. A lot of tools will not recognize the difference between design option 1 or 2 when exported and lump everything together. Separate files work here and in reality work better at an early stage.

5.) Align and orient all structure to grid lines and exact locked and constrained dimensions. This saves on time (moving a grid line as opposed to 100 columns) and makes the modeling inherently more accurate.

6.) If a floor replicates exactly a number of times use groups and copy. You can always ungroup if the program becomes more complex.

7.) Don't model the structure if you have a structural engineer on the team who is doing BIM. The engineer's model is what should be built and tested, not the architects. Link it in, save time and if you need additional structure create a separate workset that you can send to the engineer to update his model.

8.) Limit 2D info. BIM is nice in the fact you can click on a component and it will tell the user what it is and that you can model as needed.

9.) Control your levels and families. Family creation needs to be created as accurately as possible and with flexible sizing options. This is the only time I encourage an architect to use a generic block and then go back later and change it to a worthwhile family object. Also learn how to model families correctly, these are often the biggest propagator for "blank" clashes. Levels need to be limited. When there are too many different objects hosted to too many differing levels it will get very confusing during the analysis phase unless there is some system put in place that makes sense to the team.

10.) Don't send junk. if the model isn't finished, if it's not usable and if it's not coherent it won't be to anyone else.

Thanks to Jeff Woodward from Midwest CAD who came up with this idea and I hope this helps begin a dialog.

Feel free to add on if you can think of anymore, these were just the "biggies" off the top of my head.


Erik said...

And fix the WARNINGS in the project. They bog down performance.

Great list by the way.

Vincetastic said...

Hey Brad, this top ten list is fantastic. I am not too familiar with the modeling techniques of architects, but my aunt is a partner with FLAD and Associates and has discussed some of the issues they have had with it. You can post this to our site and link back to your site. We are trying to create a directory for top ten lists where people can find your site. The coolest feature is you can let other people vote on the rankings of your list.

Jeff Woodward said...

Thanks for the effort and kind words. We already miss you in KC.

Anonymous said...

Part of the problem with this list is that we are assuming that the architect is being paid to model in BIM (and if they are to what extent?). Working on the design side of construction, there is no reason to model eveything little thing if it is not going to be utilized by us or the owner. I do agree that what is modeled should be done correctly but also efficiently. We normally do not do cost estimation or quantity take-offs, that falls on the contractor. So we can give our model up and let the contractor split the walls, floor pours, add cost info. and any other info the contractor needs to complete their portion of the work. Just not on the dime of the architects or engineers.

Anonymous said...

I am the closest thing our architecture firm has to a BIM manager, so i feel obliged to put in my 2 cents worth. You should preface this discussion by explaining the concept of "Design Intent Model" vs. "Construction Model". At our office, we're only doing design intent for now, as a direct (and improved)substitute for Autocad. Modeling to the level described here is for "Construction" level, which as you can imagine takes a LOT more time and expertise than the usual contract for traditional architectural services. It sounds like your complaints involve some of trying to do things with a "design intent model" that it was never meant for. in this context, these aren't "mistakes", just mis-applied efforts. Long story short, make sure the architect you are working with has been contracted to provide that level of modelling.

architect11 said...

Great comments! I think there are a couple of words I would like to add to clarify.

For starters, I am not recommending a design model at all, rather I am questioing the entire purpose of such a model. Aside from a schematic level of information and layout sharing with the owner, it frankly provides little information useful to the contractor and even the engineering teams.

The real question is when is the water the right temperature to have help jump in? Engineers really don't like reworking entire system designs by jumping in too early and watching their fees dwindle away while the design is constantly altered. Contractors dislike coordinating anything related to "design intent" only because it = liability.

And by all means architects can be more profitable by billing for more overall through the use of BIM (I'll post a few examples later) and more effiecient design processes by engaging MEPS and FP consultants after a high degree of model development. Educated owners realize what they are paying for. I recently had a University client tell me that he willingly pays 15% more fee for Construction level models produced by the Architect/Contractor team than design level documentation.

Good thoughts but something to consider is what real value is an architect bringing to the table with "design intent" documentation?

Corey Nardone said...

Great list. I've had personal experience with at least 6 of 10.