Real Models

[Anonymous]

I have just got a job recording a precinct of historic buildings. The drawings and computer models have to be very accurate showing all materials and construction.

I am considering building correct models to achieve accuracy. This would involve placing bricks, making windows and doors, making roof tiles with Solid Operations, etc.

I was wondering if anybody has tried this before and what the pitfalls and pluses might be?

Taking a brick by brick approach would eliminate non distinct or generalized wall fill sections and allow exact placement of openings, sills, etc. Making windows and doors would allow for beads, reveals, trims, correct recessing of glazing, sash ropes, etc. Services could be shown accurately and in correct locations.

Thanks in advance.

[Dwight]

Real models contain many polygons. A buzzillion. Slows the representation down.

I once rendered a building made of actual modeled bricks, but the guy forgot to put in the mortar so at exactly eye level you could see through it. Idiot.

ArchiCAD tries to make a compromise between real and symbolic representation. To keep the redraw speed up, you will carefully manage layers and only view what you really need to see at any given time.

Of course, your detailed model will take a long time. Forever, perhaps, so manage repeatability of elements well.

[Anonymous]

Thanks Dwight. I hope you werent too hard on the modeller who missed the mortar

I would build the wall of mortar and place the bricks within the wall I think.

I take your point about the complexity of shapes within the model. However I am being attracted to the idea of taking the model to a more accurate level.

Cheers,

[Anonymous]

Graeme,

I recommend trying your idea out on a small building and then duplicating it up to the approximate size/complexity of the real project. As fast as the machines are, the world is a complex place and there is certainly a point where the level of detail will overwhelm the machine.

I did a fully detailed model once with all the bricks, roof tiles, door panels, etc (I stopped short of putting screws in the hinges) and it was definitely too slow to work with. Admittedly the computer was about two orders of magnitude slower than what we have now, but the building was also about the size of a guard house.

[TomWaltz]

I have to add my voice to the concerns about performance. There is definitely a point where the number of polygons will absolutely drag down any 3D views. You also have to consider what the plans will look like with all the bricks and morter on them, which is a little unusual.

I love the idea of higher accuracy, but at some point you have to be able to justify it from a profitability standpoint. Will the extra time spent pay off somehow?

[Aussie John]

I wonder a little, what might be the point of the level of detail you are talking about. Are you thinking of making the builing out of square and out of plumb too (as all buildings are, not least historical ones).
If the number of bricks is important you should be able to model that with vectorial hatching with a brick fill suitably sized and spaced.
Maybe you could concentrate of the special features to model in 3D rather than the "nuts and bolts"

[Anonymous]

Thanks for your replies.

John, The point of the drawings is to document a set of historical buildings. I think it would be useful to have the detail where in the future say a certain number of bricks, or a certain window has to be replaced, then it would be possible to identify the existing materials specifically and then document the new materials. An accurate record of events could be maintained. In my state, in Australia (NSW), we have to prepare accurate drawings of any historical buildings to be altered under our heritage laws.

ArchiCAD is a good program but it has its limitations if used in an accurate modelling sense. The bricks are never quite in the right place, the windows and doors are never quite the same as those designed or being documented, etc, etc.

It is possible to adjust fills as you say but to have a correct model of the building is important, in this instance, and I am hesitant to fiddle with fills, windows, etc.

I am being involved more in historical work and I am seeing more accurate models in a commercial sense also. If the level of modelling can be raised to a better standard it may further develop my business. In the future I can see a real need for fully documented buildings for building maintenance, alterations and additions, space planning, town planning, historical recording. All municiple councils will want these records in my opinion and I hope to be able to provide some of these services.

However as many have said now the hardware may not be capable of building more accurate models.

Thanks again.

[Eduardo Rolon]

This might sound weird... but have you considered using SketchUp to create the bricks. It will be more flexible during modeling (which I think is the bigger problem) and you can export the object to AC. I agree that keeping the model manageable will be a issue no matter what you do so you should establish a strategy to handle the project. Instead of building a huge detailed model, build a rough/simplified model and then build the blow-ups with higher detail as different files (ex front facade, east facade etc.).

[Anonymous]

ejrolon wrote:
This might sound weird... but have you considered using SketchUp to create the bricks.

I think Graeme is looking for more intelligence in the model, not less. SketchUp is great for certain things but there is a point where adding detail becomes a dead end. SketchUp doesn't know a brick from a block from a broomstick.

Graeme,

It seems what you need is one of my proposals for future features in ArchiCAD. Get rid of the the demo/exist/new workaround with layers by putting that information into the elements themselves. Each element would have the option for an install date and a removal/demolition date (remove for reinstallation may be a bit more complex). That way you could maintain the history of the building and all subsequent renovations, and even take snapshots of it at any time.

Obviously working out the interface for managing this sort of information is, as the programmers like to say, not trivial. Not to mention how this might relate to subsequent sets of design and construction documents (and then there are all the design proposals - yikes!)

In the meantime (10 years?) good luck working out a near term best practice.

Remember also in the historical context that professional archivists laugh when they hear computer people talk about "archiving." To programmers twenty-five years seems like forever. To an archivist it is tomorrow.

[TomWaltz]

Graeme wrote:
However as many have said now the hardware may not be capable of building more accurate models.

I'm wondering if some kind of arrayed 3D brick / block object would help with this, so you could have 2D control and 3D control separated and not have hundreds of elements making up every wall. Something like what the Wall Framer Accessory object does.