Menu
In the interest of writing it all down to share…
I’m discovering what I’ve sorta always known. I’m a task-oriented problem solver that’s dedicated to sitting down and cranking out the work without really deviating – unless I have a competing task or deadline.
Over the last 10 days or so, I was granted an opportunity to develop the ‘existing conditions’ model to help out my last firm and jumpstart project development. I can only assume that current staffing levels preclude anyone sitting around with time to dedicate to this particular duty.
I happily took on this opportunity, partly because it was good to get back into an exercise to create something from scratch, and frankly – the income from it will keep a roof over my head for an extra couple of months and Christmas is around the corner.
The project started with an abbreviated look at a PDF of the architectural set of drawings from 1998 and a couple of questions.
“Can you provide an existing model of this? How long would it take?”
Yeah that wouldn’t be a problem.
(I know my personal level of detail that I model to, and it exceeds what most people think is adequate… I do have it in me to produce half-hearted work. I get yelled at for it – but I believe that if you’re going to put your name on something – it should reflect quality.)
It would probably take me a couple weeks on the outside, but that’s a conservative estimate. I can probably get it done in less time if you need it sooner.
(As an aggressive start-up, the answer is always YES, within reason…It’s better to undersell and over-serve than it is to over-promise and under-deliver. The first leads to a positive reputation. The latter will sink you. Fast.)
“Can you do it for less than $____?”
I can certainly beat that fee. It’s just me here, I have no overhead really. Working as an independent I immediately have to save half of it for taxes, but I can do it for less than that fee.
Then comes the quick check on the financials and implications of what I just agreed to. Quick proposal email written with deliverables: Fully operational Revit model of existing conditions using the firm’s template with reflected ceilings, generic lighting locations, existing power, worksets, etc. + the established fee.
I should mention here that even if I didn’t get the project officially, that I told them I’d probably just work on the model anyway just to have something to do. I’m competitive by nature, and I can’t tell whether its a strength or a weakness.
Whole thing was submitted to the client the next morning, accepted, and we were off to the races. I had a paying gig. Christmas can happen.
Now – to the project itself.
Without an official ‘go’ from my new client on the proposal I was now a part of, I started wading into the drawings. I rarely, if ever, wait for green light on projects. I’ve always tended to play with ideas and gather information well before anything is official. I’m not a fan of big surprises on projects, and knowledge is power. I understand that I assume risk here – but at this point – there’s not much to lose but a few days worth of time doing things I would otherwise be doing.
The building was adequately documented, but it’s been 23 years since it was ‘new’ and that’s a lot of time for minor remodeling and revisions to happen on a property. Personally – I’ve designed buildings that have changed ownership 2-3x in that timeframe and aren’t remotely the same anymore.
That’s the long way of saying that the drawings are fine as a starter – but there are going to be things that fall outside of my ability to observe and coordinate them. So I flagged them as things to look out for.
The plans were done with CAD, the elevations, sections and details done on the drafting table. Both were scanned prints from an original archived drawing set. You may or may not know – the large format roll-feed scanners can, and do stretch drawings – so graphically they’re not accurate. In this case it was combined with fudged dimensions, and plans that don’t align when you stack them back up.
I’m fairly certain that the elevator should be the same on all levels… If not – that would’ve made the papers somewhere.
I even broke apart the enlarged drawings into ‘way too large for common sense’ sized photoshop file to see if I could produce a scalable background… but that’s how I figured out that they didn’t align to each other. Even the overalls were stretched in a similar fashion, and graphically it was going to take a lot to undo that – for the limited usefulness it would provide.
I also couldn’t find a single column intersection that used BOTH the orthogonal and angled building grids, and no exterior architectural dimensions to rectify that relationship.
Then there’s the added wrinkle of dealing with drawing errors. In this case, the structural building grid and the architectural didn’t exactly match. Searching web images of the existing property show certain things that exist in the built environment that were handled generically, at best, in the drawings. I flagged all of that for the team that will be moving this project forward, and will remain a resource for the team. I tend to get fairly well versed with projects as part of the this process.
Certainly I’ve done more with less – so although it was an issue, it wasn’t an insurmountable one.
* PRO TIP: I know that the industry has tightened up over time with BIM and better practices, and doing a quality check on projects is largely a pain in the ass. There never seems to be enough time to get to it – but it’s also VITAL to the guy 30 years from now hacking together a model from your original working drawings. Archivally speaking – accuracy of the end result will save you time during construction, if not someone else’s time along the road.
Mistakes happen. If you aren’t making them, you’re not trying hard enough. When they do, own it. Fix it and try not to do it again. Pretty simple.
Always start with structure. I built out the structural grid(s) – orthogonal and then angled, and locked their respective directions together by locking dimension strings.
Then I grouped them which allowed me to move them around in unison without disrupting their relationships.
Once I found the magic columns located by intersecting gridlines – I pinned everything in place. But that wasn’t until much later in the modeling process.
I also learned thru developing this that there are sub-grids in some of the plan details, and the details didn’t necessarily match the overall architectural or structural plan locations.
Yeah, not clean – but close enough for jazz, and it was a way to start with some functionality in case things needed to adjust on the fly.
*PRO TIP:
The primary components necessary for existing conditions are:
Elements you’re not going to be moving much (if at all).
Much as the title suggests, and with some conflicting information, the only way to truly settle things out would be to model from the top, down and from the inside, outward.
In my experience, and it’s not especially glorious or even that much of an enlightened observation… The bulk of the organization of any stacked building type starts with the bathrooms and their associated mechanical chases. So I started there.
Building out and grouping all the existing bathroom types with left and right configurations, tubs, showers, accessible and alternate unit pairings… the entire matrix of possible outcomes.
Now, if you’ve done this as many times, for as many years as I have… coming across placeholder Revit families existing conditions documents isn’t difficult. Minor modifications were made for the shower unit to include the door configuration and the shower seat – but the rest is fairly simple.
Simple is a relative term though. There’s 12 bathroom types there.
Then a version of each of those with a narrower vanity because they pair with an accessible bathroom so the list above x2.
I started in the upper left. Completed one.
The process is similar for the rest of them. Copy, make a new group type – edit the group to reflect the dimensioned plans (& whatever photos you can find).
You now have the building blocks to start to develop room types. If Revit doesn’t freak out later in the project – you can also edit every one of these placed groups at the same time.
With a placeholder group developed for bathrooms, you can start to add in corridor and demising walls to start the general rooms. Now – interspersed in this field in the final model are the structural grid – which has 2 different bay sizes… (24′-8″ and 25′-0″), but that distance gets made up in the room entry hall. Bathrooms remain consistent.
*Sidenote – make sure that all of the groups have the same insertion point. Once you populate a floor plate, you can change group types into the appropriate bathroom type for that room, and it should be a clean replacement that doesn’t blow up the model. It might still freak out the model, but it’s easy to deal with if the placement point is in the same place.
I usually build this portion off to the side, so that you’ve got geometry in place to test drive families that you’re building in the model without impacting anything else you’re working on. If it blows something up, it’ll just take out some ‘sketches’ off to the side and not the whole project.
Once you get a field of rooms together, find the clear corridor dimension, drop a couple of reference lines and mirror. Voila, you’ve got rooms started. In this case, I then reviewed all of this against the mysterious grid lines, and what dimensional information I could find in the architectural drawings and slab plans to locate a floor plate edge.
The floor plate itself is also symmetrical across the east-west line, so I only had to start with the north half of the level to start.
Generate the room field, then the inside corners to locate the elevator core, then the end suites and stair and elevator cores… slowly nudging grids, refining locations and adding the structure for that floor plate. The target here is developing geometry, based on the dimensions given in the drawings.
In this case the plans were dimensioned with a 1″ tolerance and generic width walls – so with a little institutional knowledge I managed to get a floor plate developed.
Then it’s just a matter of grouping that floor plate and all its associated parts and copying to levels with a solidified column grid at the upper levels. Levels 4-7 established.
I also managed to locate a column in the inside corner rooms that helped to start to establish the organization on the lower floors.
Once all that is established, translate downward. Then you can work outward on additional components, site elements, etc.
The rest of the development of the model was fraught with additional challenges:
Traditionally – you probably won’t need the existing room mix. BUT, as an ‘extra’ I populated the entire model with existing rooms, matching the existing architectural. A couple days ago, I was sent a room matrix from a survey crew for the project. Since I did what I did – it was easy enough to compare and tweak anything needed, because I already had it scheduled. Had I not done that – there’d be an extra few hours worth of sorting all that out for confirmation. Like I said at the beginning – it’s better to include things that you may not need, than to get into a situation where it doesn’t exist and suddenly needs to.
I started doing this from the beginning, but I’ve always got a section cropped 3D working view open when I’m working on something. Its habit. I design and model in 3D, not just draft in plan and hope that it all works out.
Additional to this, I cross checked all my model groups with Enscape3D. When it was mentioned that all power and lighting needed to be located – I would add the lights to the room group, and update the Enscape model night scene to make sure that I didn’t have any outliers that didn’t have the lights on. Obviously just for my own entertainment – but hey, why not?
Copyright © 2023
