Best practice for existing building shown on site survey?

[patricks]

I'm working on a large existing building renovation. The entire existing building has been modeled using existing building drawings, with demo items already demolished in the model.

This week I received a site survey, which shows the outline of the building. As usual, none of the building outline edges are parallel or at right angles, and the outline doesn't line up with my Revit model at all. For example, one exterior wall measures 81.5' according to the existing drawings and the Revit model, but that same wall measures 83 feet on the survey's building outline. I really don't think the actual building is off THAT much from the original construction drawings.

What do most people do in this situation? Just kinda-sorta get the survey file in place based on the building outline and your existing building Revit model?

If I tried to make the existing building's walls align to the survey's outline, it would be major hell trying to dimension stuff, and I mean dang near impossible. I know that buildings are never truly built square and plumb, but for dimensioning and other documentation purposes, you really have to draw it as if it were square and plumb.

[mthurnauer]

Just having completed a project where we modeled the existing conditions from the original building drawings, I can understand your plight. I agree that having the building all out of square would be very difficult to work from and if it is close I would leave it square and make sure your dimensioning method accounts properly for what your intent is and does not put you at risk of precisely dimensioning existing conditions which you may no are not true. Fo example, we were sure to dimension new walls to be centered on grid lines as such and not to give hard dimensions. Other walls are then parallel to these and offset specific distances, but dimension strings are not closed and the break in the strings are located where we can afford for there to be a deviation from the design. If however, you were to find a bust of 1'-6" between the survey and existing conditions, I would have a surveyor or field rep verify it. It very well could be off by that much. I had two projects in recent years that had the floor elevations off by over 10 inches. A change like that is not necessarily an error, but may have been a change that was not reflected on your drawings.

[andrewharle]

I too am trying to model and existing building where the survey drawings do not have parallel lines for walls.

I have downloaded a window (http://revitelemental.blogspot.com/2010/05/components-for-existing-building.html) from a blog from Chris Senior which is stretchable, and have produced my own stretchable doors. Chris shows a useful trick to make heights and widths instances not types.

However modelling irregular, non-parallel walls is the tricky bit. I am trying to do these as in-place families, but windows tend to jump around, and sometimes come up with errors.

ArchiCAD has a polygonal wall tool, that allows you to draw more or less any polygon shape as a wall and then place windows and doors within it, and polygonal wall joins up with other walls and behaves as a regular wall. It makes modelling existing buildings easy - I once modelled a Victorian church from a survey.

I would really like to see Revit come with a better set of tools for modelling existing buildings.

Meanwhile if anyone has got more experience of how to do it please add you thoughts.

[ws]

I'm not saying that it's particularly clever but I've done the attached in Revit and many similar buildings with varying wall thicknesses and tapers etc and been pleasantly surprised that Revit doesn't fall over ;)

Allplan, like Archicad mentioned, has a polygonal wall tool but you can also convert virtually any shape into a 'wall' and have windows and doors cut through automatically - and the windows and doors are stretchable...

Revit lacks that flexibility but for me, makes up for it in the ease of displaying 3D in 2D 'drawings'.

I take a sensible average for a really bendy wall and draw large variations as dotted lines and/or use the wonderful Cut Profile tool to amend the the apparent line of the wall in plan or section - as in the attached zoomed-in bit of 1:50 section drawing.

Digital surveys by specialist surveyors can be a nuisance as noted above - I think you just have to take a view on the importance of certain angles. Usually I do my best to align to plan angles, much as I do when I measure up myself.

[andrewharle]

William,

Thanks for your response and your tip on the 'cut profile tool'.

As an update on creating in-place wall components, the stretchable window with reveals that I referred to before does not work well in non-parallel walls. It can jump around all over the place, and when putting several in a wall, the wall fill pattern disappeared. It looks like a window or door in an irregular in-place family wall probably needs to have a single void forming it.

I have therefore created a simple stretchable window from the standard 'Sgl Plain' window that seems to behave more reliably. The only issue I have is that windows seem to stretch or shrink around their mid-point no matter what reference planes are set to define origin.

I have also aligned the tops of walls to the level lines in section and elevation, making their height adjust with the storey level.

[trombe]

I'm working on a large existing building renovation. The entire existing building has been modeled using existing building drawings, with demo items already demolished in the model.

This week I received a site survey, which shows the outline of the building. As usual, none of the building outline edges are parallel or at right angles, and the outline doesn't line up with my Revit model at all. For example, one exterior wall measures 81.5' according to the existing drawings and the Revit model, but that same wall measures 83 feet on the survey's building outline. I really don't think the actual building is off THAT much from the original construction drawings.......(abridged)


Hi Patricks,
I have trusted my site measure over the survey if its out , however its usually been only one wall or a corner.
Where I am relying upon the survey and only have that, I feel you have to trust the survey and run with it however, I have asked for confirmation where there is doubt as to accuracy.
This gives pause for consideration, provides you with more surety and leaves a critical paper trail from raising your concerns. I agree, a wall that much off is likely an error on the part of the surveyor (not enough coffee that morning ?)
regards
trombe

[bryanm]

Our office runs across this situation quite often as we do a lot of renovation and remodeling projects. In my experience the only thing that I trust 100% is my own tape measure. If there is a discrepancy between the drawn or modeled files then the best thing is probably to take the time to go field measure the building or area in question, or else you'll always be guessing which document is correct.

By the same token, if you are doing renovations on only one side of the building and a wall on the other side where there isn't any new work going on is off by some dimension then it may not be worth the time or effort to verify.

[gwnelson]

I have yet to find a survey where the walls are at right angles. Is this a legacy of sloppy carpentry or surveying? I trust the builder first, always.

[andrewharle]

Some thoughts on this...

The age of the building is important in determining whether to trust the design intent or the survey.

On older buildings like the one I am working on at present I am assuming that the survey is not accurate, but equally that walls in places vary in thickness and wall lines may not be parrallel.

On a project I worked on once modelling an modern existing building (in ArchiCAD) we buit our model from the original architects drawings, but then modified our model to suit an as-built survey the client insisted was required. It turned out that the drawings were more accurate than the survey, and we had several setting out problems on site because we had followed the survey.

As mthurnauer notes its best to check any major discrepancies.

[ws]

There's only one land/building surveyor I trust locally, and it's interesting that everyone else seems to trust their own measurements as well... based on experience I expect ;)

Even on my own surveys I never entirely trust those laser measure devices (i.e. Distos and the like) despite using them since the early days and always take several check measurements and/or use the min/max function into corners etc.

I still have tape measures in my bag, and pencil and paper are hard to beat when surveying old buildings.

I've tried the Bluetooth devices for transmitting measurements to a handheld but found it very hard work and went back to pen, pencil and clipboard :)

[trombe]

Ws,
where would we be without the Disto ! I love that thing !! - except that outside any decent distance its useless even with the cool little glasses that came with it !.
Then even out of whack walls matter not.
trombe

[ws]

Yes, they are great gadgets - started with a Mark II (the Mark I you needed 2 hands to hold) and have bought nearly every model since (I should start a museum).

Always remember the day I had to measure up an old stone barn, 2 ft deep in cow muck (yes, that's 2 ft /600 mm deep ) and was able to stand in the doorways and take measurements without standing within ;)

Currently my favourite is the D5 which has the colour tv camera in it and hi-res screen - a big improvement on the old lcd screens.

[andrewharle]

Further to my previous posts...

Am continuing to model the existing building using a mixture of Revit walls of approximate thickness, where survey wall lines are more or less parallel, and in-place family walls where they are not.

The problem I have is that whereas Revit walls can be cut by voids in subsequent phases, in-place walls can't, which means having to keep re-editing walls that are to be modified in the new construction.

Revit generally has very good modelling tools, so its a shame that it lacks a few simple tools for modelling existing buildings.

My Wish List items are therefore:

Tapering Walls - from the wall tool
Polygonal walls (sketch walls) - from the wall tool

p.s. an example of what's possible with ArchiCAD's wall tool
http://archicad-talk.graphisoft.com/files/thumbs/t_polygonal_join_111.jpg

[ws]

When things start to get complicated I often give up on using Phases and just duplicate the project to make an 'Existing' and 'Proposed' version.

Not ideal but it avoids a lot of issues with Phases and modelling fiddly alterations.

[STHRevit]

I don't think I have ever seen a survey drawing with building outlines at right angles. Not surprising I guess if you think about what might be on site, rocks,garden beds,plants etc etc. The surveyor can only do their best to get accurate levels and would not be too concerned about walls being straight.
We tend to put faith in the builder and assume tha the as con drawings are more accurate. Place a big bold note on you drawings stating that all dimensions are required to be checked and confirmed on site prior to construction, or something along those lines.

[benmay]

The other item to consider with existing builds is whether you actually need the accuracy of a few mm/inches here and there. These things look obvious when you have the survey turned on but once your documentation is complete you will mostly not notice the slight inaccuracy to a survey that may or may not be more accurate in the first place

From my experience on a few large projects with large areas of existing structure, one item to avoid is following the survey too closely, particularly for walls as you will end up with a lot of warnings in your project which can really start to affect performance, and you main workflow you should avoid is using pick lines on surveys. And a workflow I would suggest is modeling walls first by picking end points of the survey and straightening if there are slight misalignments and then modeling things like floors by selecting the modeled walls and not the survey

Where there are obvious areas of discrepancys the best approach I have found is to make note of any areas that look unusual on the survey and goto site and investigate, taking plenty of photos and measurements where required