Desgning House in Revit with Complex Geometry + Angles: First Project with all Revit Consultant Support

[Limbatus]

Hello,
I am drawing a house in Southern California that will have complex geometry and angles. This will be the first undertaking of this scale for our office. We have had experience fabricating interiors and installations with irregular angles and parts, for about 15 years, but this is the first ground up construction project. We are currently assembling a team of consultants, and it has been agreed that the entire construction document set will be produced from ( or in conjunction with ) Revit. I am very excited about this project. Over the past 10 years we have been producing our Architectural deliverables using Revit, but we have never issued Revit Models to consultants as backgrounds. Are there any good discussion related to All Revit Project Management and Consultant Coordination? Where can I learn about best practices etc?
We are beginning to move into schematic design after about 2 years of concept work. We have some very developed rhinoceros models, as well as some physical models. Now it is time to begin the BIM process. Any advice or recommended reading would be greatly appreciated.
Thanks, Limbatus

9484794846

[Craig_L]

I have worked on several multi-discipline projects using revit now, I would have to say the best method I have seen (as far as clash detection, and resolution goes) has been the times when it was used in conjunction with Navisworks. Navisworks outputs the clash or problem with a screen shot, the element ID (so that in revit you can find the element by this number) and also it allows for much easier navigation of the model with all disciplines. As you can imagine a large or complex project with many disciplines modeled can get very slow in revit, navisworks seems to handle many models at once better (for review purposes).
I would suggest to you have a read of the National BIM Guidelines (and case studies) that have been developed and published in Australia.
These were written by people using the software and analysing their issues on projects, so they are practical and although not all of it may be relevant to you or your area, it will give you a good idea of what to expect and some tools for overcoming issues that arise.

http://www.construction-innovation.info/images/pdfs/BIM_CaseStudies_Book_191109_lores.pdf

Otherwise, looks like an interesting project...complex angles...let the wringing of hands and palming of faces commence.

[Limbatus]

Thanks Craig. I found that PDF pretty interesting. here's a few things i picked up:

Library Review - It seems obvious now, that all of our Structural Families should be Pre-Approved by the structural engineer. I don't have tons of experience modeling structural intersections and connections in Revit, but I have experienced limitations trying to join more that 2 structural framing families at once. I anticipate certain structural connections in this project requiring 4+ beams / columns coming together at a single point. we definitely need to make sure these are all joinable.

4D Sequencing - We've never really done anything like this through Revit. Always sequenced by the user with the help of phasing. Where can I learn more about other 4D features and Aspects of the software?

Clash Detection - We also have no clash detection experience. I read that NavisWorks is superior to Revit's built in features. Any experiences / stories or recommendations about this would be great.

[Craig_L]

I'll try to tackle these points from my experience for you, some of it I can't really answer. (sorry in advance for the wall of text)

1. Structural Families - Generally the OOB families will suffice, not in all cases. If you are going with the "seperate models" workflow (which I definately recommend for you) then this generally isnt too much of a problem, except for your initial objects that you will control.

What "we" tend to do is determine a handover date before we commence the project. I would suggest after Design Development (DD), ie after your client has approved the initial layout and your floor template and building shape is generally not going to change greatly.

What does this actually mean? When we recieve the architects model, we import it, set up our project coordinates etc to match, and then we proceed to copy/monitor levels, grids, slabs, walls, floors, columns, beams (all structural elements) Generally the architect doesn't do things like footings etc, we handle those. Now once we get to DD and the building isn't changing to the clients tastes all the time, we want to control the structural elements. So we ask that the architect delete all structural elements (easily done with some filters - walls, columns, beams, floors etc) and now they copy/monitor OUR elements in their place. This means in the early phases its the architect who controls these elements, and will dictate changes. Later on we (structural) want to keep track of these.

For coordination this works well, and Revit will give you warnings and a list of what has changed between the two models. When the structural team takes control of the elements they simply need to unmonitor the elements, because now it needs to be monitored in the arch. model. The architect keeps control of levels, grids and anything else thats not structural. We actually list these elements in a spreadsheet, and the handover dates (or expected handover dates).

Keep in mind also that huge changes for your flow on disciplines might be too hard to incorporate into a model for them 2 or 3 days before an issue is due, so set up dates in advance of deadlines to allow sufficient time for these changes - this needs to be fairly maleable as it just depends on the amount of change and what exactly has changed - your subbies will be able to tell you a rough idea of how long they need in most cases - and its a good idea to schedule regular meetings with the subbies so you can discuss upcoming changes and pre-approve etc.

The main thing I would warn against, is walls that built as composites. Structural just dont care about insulation, ply, and anything thats not bearing load, so in some cases instead of modelling your wall with composite layers of brick/gap/block/insulation/ply, break this up so it excludes the block, has a gap large enough to place the block, and place the block wall as a seperate piece, because later you will hand this over to structures and you dont want it all to be part of your composite, because this causes issues in documentation for the structures team (and will cause you problems for wall based families) and annoyingly yes this means placing openings twice...but again these get handed over with the walls later.

2. 4d - have hardly ever used it, check out phasing though which is done within revit and this can then be done also within navisworks although again, Im not that familiar with the 4D process we've never gone to that level of detail, although from a project management perspective I imagine its quite useful.

3. Clash detection, I've only ever been on the recieving end (thats to say I dont run the clash detection I just supply an out model for comments). I know that it is done in navisworks however, and what gets sent to me is an excel spreadsheets (which I believe is output by navisworks). Its quite simple in navis you can put in a parameter to say for example we want 5cm clearance around all columns - and it makes sure nothing impedes that. It also will check any element running into another, and flag it, you can accept it as is, or flag it and it then outputs all clashes in a spreadsheet with a screenshot, an element ID, which discipline owns it etc etc. I've seen this handled well, and not so well, depending on who manages the clash detection. Honestly I cant tell you how they do it, I just know it works well when I get a list of my disciplines clashes only what element number it is and a picture of it. I can also open the navis model after the clash detection, filter the disciplines I want to see and go look at that clash which as you can imagine makes it a lot easier to locate in a complex model environment. I believe also that revit will give you an excel based spreadsheet to work through with element ID and other identifying remarks as far as the copy monitor tool goes. I would recommend have a good read about navisworks and its capabilities, it might even be worth using a project management team that is well versed with BIM and Navis to handle your clash detection (we have found most architects outsource this now to specialists, Navis is expensive for the full blown version and it might just end up cheaper to outsource it to a specialist for you guys, and also you need someone versed in the program) and the coordination generally goes very nicely.

Hope that helps, I could probably go on for a long time, there are a lot of nuances and things to consider and its not evident until you dive into a project. Definately consider seperate models, this has other implications like it requires all of you to be using the same "build" version of revit, at least it used to Im not sure if this is something they phased out of revit but its worth checking into before you all start on a project and find you cant link anything or all of your copy/monitor elements get broken half way thru which would be a nightmare.

[Limbatus]

I'm interested in learning about Copy / Monitor + Clash Detection. Any good spots where I can read up on this? sample files would be great too. also, I want to know about the added benefits of Navisworks. Our firm currently doesnt pay for it, but it seems like it might be a worthwhile investment.

[david_peterson]

Navisworks can be a huge help depending on your construction firm. I've see projects where we get our models to zero clashes only to have one of the trades throw out the MEP drawings because they technically only show "design intent". This makes the whole process a waste of time IMHO. Copy monitor can be a good tool if you're sharing parts of the project. Things like Grids, levels and floors seem to be the most common. Some of this also depends on the type of families you use (Face Based vs Wall hosted vs Floor or ceiling hosted). Copy monitor works just like it says. You copy it from a linked file and set the monitor. When that element moves you'll get a warning. If it's get deleted, you should get a warning. If they add a grid down late in the project you'll need to be notified by the consultant. It can't monitor things that aren't in the model. And things that get deleted, and replace causes confusion for revit.
There's several books out there that have good material in them. The Mastering Series by Paul Aubin is decent. The one Tom Wier & David Harrington wrote for revit structure is great.
I believe there were also several AU classes that you should be able to pull from the Archives.
Hope this helps.

[Craig_L]

Copy monitor is fairly straight forward but as David mentioned there are nuances that revit likes and doesnt like in procedures.
The main thing to remember is if an object needs to resized or relocated, you dont delete it you leave the modelled item in place and alter it to suit. Deleting it in the host file will essentially "break" the link. What actually happens is it tells the 2ndary model that the item was deleted (but no one knows what replaced it) so deleting items should only be used when the item actually no longer exists.

Revit used to export a spreadsheet with copy monitor items on it and any changes, its fairly straight forward although it used to be a little too "verbose" and often I found myself wading thru pages of items that didnt need attention. The best way I found was to set up "coordination 3D views" where I color coded linked walls red, and my walls blue (for example). Also I filtered JUST walls and floors so I can focus on those elements, and this let me see visually changes that needed attention, of course this wasn't the only thing used in that process but was certainly helpful.

I would suggest get a Navisworks reseller (usually your CAD/Revit supplier) to give you a Navis demo.
We did this with our reseller, he organised a meeting, we had all of our disciplines attend, he had a real life navis works manager come in and give us a real life demo of how it worked and then Q&A at the end, it was well worth it, but in the end management decided it was too expensive for us, but we are also not often responsible for the clash detection (only the clash resolution for us) the detection was the architects responsibility more so than us, so for them it was useful, we only needed the navis viewer which is essentially free as far as I recall.

You can find a few tutorials, and blogs on the subject, even just watching the youtube tutorials will give you a nice idea of how it functions and some of the troubleshooting you might need.

http://bim.wikispaces.com/Revit+to+Navisworks+for+Clash+Detection
http://www.youtube.com/watch?v=o_9X3DM9t60

[jagostinho]

The main thing I would warn against, is walls that built as composites. Structural just dont care about insulation, ply, and anything thats not bearing load, so in some cases instead of modelling your wall with composite layers of brick/gap/block/insulation/ply, break this up so it excludes the block, has a gap large enough to place the block, and place the block wall as a seperate piece, because later you will hand this over to structures and you dont want it all to be part of your composite, because this causes issues in documentation for the structures team (and will cause you problems for wall based families) and annoyingly yes this means placing openings twice...but again these get handed over with the walls later.

What do you do suggest in these cases?
Should the architect substitute the block layer inside a wall/slab/roof with a mockup material?

[Craig_L]

No don't use mockup material.
Copy/Monitor the structural walls into the model

[jagostinho]

No don't use mockup material.
Copy/Monitor the structural walls into the model

And then use 3 wall types to represent a single wall?

Wall 1 - Exterior Finish
Wall 2 - Copy/Monitor Structural wall
Wall 3 - Interior Finish

sounds like difficult wall type management.

[Craig_L]

No, use 2 wall types.

Wall 1 - Exterior finish / Gap big enough to place block / Interior finish
Wall 2 - Structural Wall

[jagostinho]

I see.
So, I still need to create a "gap" material to accommodate the structural wall.

I see that the advantage on this is obviously the coordination.

But will bring more trouble while creating the architectural drawing series.
For instance, for automatic Wall Type Legend.
Or the case where one could use colored Plan Views by applying filters. For instance, for Fire Rating or Acoustic drawings.

Do not get me wrong. I am in favour of streamlined coordination (who isn't?!).
I am just trying to be aware of the traps of using such approach. So that we can be able to tackle them appropriately.

J.

[Craig_L]

I think you should find - these days with being able to tag and schedule from a linked model, that you can still get all of this to output for you. Legend is not that important, you can just put a hatch region for the structural zone, or create your legend with the copy monitored wall already in it.
There are traps, no doubt, I guess the question to consider is, are the traps worse or better than the documentation limitations?
In my experience, for coordination, and documentation the 2 piece method has been superior to the single wall piece method. The only real problem I encountered with it was that there are 2 openings placed because the opening is based on its host, and therefore needs to be created/hosted on both objects. There was a problem with the coordination of these openings, all be it small, and this came down to communication more than anything else - had we realised this implication we would have copy/monitored wall openings too, or at least tracked changes to them better.