In AutoCAD, I draw my model so that 1 unit equals 1 foot. As I understand it, architects commonly draw a model where 1 unit equals 1 inch.
I've been asked to do stakeout by an architect, to whom I had previously sent a topo drawing. He took my topo drawing, scaled it by a factor of 12, and translated and rotated my linework for the purposes of drawing the proposed building. Trouble is, when I ask him to specify what he wants staked out, he's sending me his drawing which is scaled and rotated, and which is not consistent with my base drawing and points, and I'm finding myself now trying to figure out what he's done, and figure out where the points are supposed to be staked out. I don't like it. Opportunity for error on both ends.
There must be a better way for architects and engineers to communicate. I suspect it is properly handled by AutoCAD setting(s). Can anyone clue me in to what would be the best way to improve this process?
That is a long war you've gotten yourself in.
I usually tell them that "I don't want MY modified [sarcasm](screwed up)[/sarcasm] drawing back, just send me YOUR drawing elements and tell me where you want it on the property".
> There must be a better way for architects and engineers to communicate.
We use Civil 3D and when we send our files to the designer they can only XREF the survey data into their drawing. They may rotate it or whatever, but at least they can't change the original drawing. Not sure what version you have, but that process is much safer.
What do they teach these guys in school?
There should be a common project coordinate system.
They should be able to provide something other than a digital drawing.
How is the contractor supposed to build the project, if there are no dimensions or coordinates provided? A properly annotated set of plans should have enough information so everything can be calculated from the hard copy plans.
However I suspect your situation is probably common place now days.
I'm not on Civil 3D, and don't anticipate going there. I'm using Carlson 2014.
Really? There's not a good way to handle this? This must be one of the most fundamental questions pertaining to CAD software design....
Microstation has the answer. You can specify secondary units allowing you to see the length of any given line in feet, inches, vars, meters or light years. This would be a great feature in AutoCAD or its knock-offs.
Bring it into your drawing, and scale and rotate it back. If he included your information, then you should have things to relate it to in both systems.
My Architectural Drafting professor (practicing Architect) beat into us that site plans are Civil plans (even when drawn by an architect) and therefore should be plotted at an engineering scale with CAD units of 1 unit = 1 foot. Architectural plans are plotted to architectural scales with CAD units of 1 unit = 1 inch. In actual practice I've never found that to be the case.
Typically when I receive a file, I open it, take a look under the hood to get a feel for the level of carnage I'm dealing with, then I close it. Next, I'll either insert the drawing as a block or an Xref, then scale 1/12 (always the fraction not the decimal equiv.), then translate & rotate as necessary to get the boundary back where it belongs. If they've committed no additional CAD Crimes their version of your drawing should perfectly overlay your original drawing. At this point I calculate the building perimeter, column lines, etc from the paper or pdf plans supplied and create a block of the calculated building. I then overlay my block of the building over their building shown on the xref'ed site plan and compare. This is also the point where I turn off the xref, leaving my calculated building visible, and check for conflicts with easements, setbacks, existing utilities and anything else that was shown in my original drawing. Quite often I discover that the latest revisions to the design are not reflected on the site plan or that conflicts have been overlooked because they've turned off half of the important info in my drawing so that their site plan looks "clean".
Request revised site plan, lather, rinse, and repeat.
> My Architectural Drafting professor (practicing Architect) beat into us that site plans are Civil plans (even when drawn by an architect) and therefore should be plotted at an engineering scale with CAD units of 1 unit = 1 foot. Architectural plans are plotted to architectural scales with CAD units of 1 unit = 1 inch. In actual practice I've never found that to be the case.
>
> Typically when I receive a file, I open it, take a look under the hood to get a feel for the level of carnage I'm dealing with, then I close it. Next, I'll either insert the drawing as a block or an Xref, then scale 1/12 (always the fraction not the decimal equiv.), then translate & rotate as necessary to get the boundary back where it belongs. If they've committed no additional CAD Crimes their version of your drawing should perfectly overlay your original drawing. At this point I calculate the building perimeter, column lines, etc from the paper or pdf plans supplied and create a block of the calculated building. I then overlay my block of the building over their building shown on the xref'ed site plan and compare. This is also the point where I turn off the xref, leaving my calculated building visible, and check for conflicts with easements, setbacks, existing utilities and anything else that was shown in my original drawing. Quite often I discover that the latest revisions to the design are not reflected on the site plan or that conflicts have been overlooked because they've turned off half of the important info in my drawing so that their site plan looks "clean".
>
> Request revised site plan, lather, rinse, and repeat.
Ditto, I have found that I am a better overseer of the scale rotate translate process from their Lunatic scaled world back to mine than they are.
Archyteks are always right
ALIGN command is handy. You just need 2 known common points in both drawings and you can move/scale/rorate either way in one command.
I use it a lot.
I second the xref option, to keep the design data as separate and unmodified. Inserting the Architects plan and scaling it will likely mess up any dimensions they've used.
On a related matter, if you're software will run lisp routines, find chkdim.lsp and it will identify all the dimensions that are modified and it will help show where the linework/symbols might not be held.
On a side note BIM coordination seems to be set up to be used from the design phase on, but rarely (never in my experience) is it based on the original survey plan.
> In AutoCAD, I draw my model so that 1 unit equals 1 foot. As I understand it, architects commonly draw a model where 1 unit equals 1 inch.
>
> I've been asked to do stakeout by an architect, to whom I had previously sent a topo drawing. He took my topo drawing, scaled it by a factor of 12, and translated and rotated my linework for the purposes of drawing the proposed building. Trouble is, when I ask him to specify what he wants staked out, he's sending me his drawing which is scaled and rotated, and which is not consistent with my base drawing and points, and I'm finding myself now trying to figure out what he's done, and figure out where the points are supposed to be staked out. I don't like it. Opportunity for error on both ends.
>
> There must be a better way for architects and engineers to communicate. I suspect it is properly handled by AutoCAD setting(s). Can anyone clue me in to what would be the best way to improve this process?
If the architect is a sub, as I most often our case, we outline the parameters on how we need to communicate and how the data will be used. Typically, we will let them use architect units but we always insist that they keep the orientation 100% as-is. This way all we have to do is specify the units in the xref file and it will come in correct.
When we're not the prime and were merely on a team with architects then things get a little trickier. As many have pointed out above, scale rotate and translate and double check all their data. As we know architects are notorious for putting 15'-10" on the plan but the actual Cad file will be 15.75' or something other than what is called out. This is the part that really frustrates me.
Mapman...
"... they can only XREF the survey data into their drawing.."
How do you control this?
"This must be one of the most fundamental questions pertaining to CAD software design.."
Its not a software issue...
:good: :good:
Mapman...
The servers will only allow read only access to anyone outside of the survey department. So essentially the file becomes a read-only file. Also, if we make an adjustment to the survey file it will automatically update the XREF'ed file and send notice that the file was altered.
This requires quite a bit of IT involvement. Since they determine who, within the department, can make adjustments and who is outside. AutoCad in most versions I have heard of have some capability to do this. But, I'm not IT and do not know exactly how this is done.
I had a plan yesterday that shows two north arrows. One for grid north (which is what we supplied the original info on) and the other for "project north". I just hope that they are smart enough to do all of the sun calcs etc using grid north!
Architects here use mm so our stuff often gets scaled by 1000. Sometimes around (0,0) which is easy to fix, sometimes around a random point, sometimes it gets rotated as well.
A mate did a detail survey for an architect to extend a building. After the design was done they then set out some grid lines for the builder. Accusations about the surveyor's competence flew when it was discovered that the new building didn't fit the existing. After lots of checks, the surveyor was fine but the architect had decided to lengthen the end wall of the existing building. Too bad his sterling efforts of lengthening a line in CAD didn't actually make the wall any longer in the real world.
If I have done the original detail, I will always overlay the design onto the original to check that things haven't been adjusted in some way.
Architects and engineers will screw with your data and there is not much you can do about it. On the very rare occasion I get a design back that fits perfectly and hasn't been scaled, rotated, etc I now make a point of sending a quick thank you email explaining how much easier (and therefore cheaper for the client) that it is. I like to think that in a few years this approach may pay off sometime.
Mathematical Closure
I once met a PhD Architectural Professor from MIT. I asked him to be sure to teach all of his students mathematical closure. His face turned pale white as he asked how many architects' plans do not mathematically close.
Better yet, I was working with one who had designed a building 50 SF over the maximum lot coverage. His building was 50' deep, so he decided to take 1' out of the middle. Then he thought, well, why not take 18" out of the middle? I replied, why not take out 16" and remove a joist? His response was, oh the Structural Engineer worries about those details.
Do you know how much ever loving waste there is in a McMansion, then you don't even try to reduce the number of cuts by holding a few cardinal 48", 96" dimensions? Every cut means another scrap for the dumpster...
Good luck with the war against the architects. They do one thing better than we do: they know how to charge for their services. There is a local design build firm with a $100k clause. If you do not use them to build the design plans (mind you, preliminaries with permits are about $100k and construction set is about $450k) then you get to shell out $100k to have a different builder. I believe their construction starts at about $500 a square foot. They would not be so bad if they did not beat up their subcontractors.