Extra decimals in dimensions

[dfi]

We have been encountering a strange phenomenon. Our dimensions are set to millimeters with no decimal places. When dimensioning between objects al the dimensions seem to be clean but when we place an overall dimension, the dimension is no longer clean. For example, we could have three dimensions of 200mm but when you place an overall dimension across the three dimensions of 200mm, you get a total of 601mm. If I add a few decimal places, you can see that the three dimensions of 200mm have extra decimals. I can only attribute this to sloppy modeling. In reality, it should not mean too much but this inaccuracy just seems to be compounding. Is there a way to reset all the positions of objects placed in a model to full millimeters? In the old AutoCAD days, we had a lisp that would clean imperfections like this up. Does anyone have any secrets to do the same in Revit?

[MikeJarosz]

You cannot edit Revit dimensions to make them read the way you want them to read. Revit dimensions merely report the size of the selected object, using the dimension format settings. If the object is not accurate, the dimension will reflect that fact. Revit went to great lengths to prevent fake dimensions. A fake dimension will compromise the integrity of the model. Think about it. Derived properties of an object like area, weight, center of gravity etc. are calculated using the actual sizes of an object. Revit has no way of reading your mind as to what you intended the dimension to be, only what you modeled. Imagine if engineering programs allowed such fakery. Acad created a whole generation of architects who tolerate this deceptive practice and accept it in their drawings.

The topic of overall dimensions not adding up to the sum of the parts has been covered many times in this forum. Because a dimension reads 200mm does not mean it is modeled as 200mm, because dimensions are rounded. If you create a ceiling tile grid that should be 200 x 200, but actually make it a fraction larger, the dimension will read 200, but over the length of a room the fraction overage will accumulate enough to throw the overall dimension over. An overal dimension is not the sum of the parts AS DISPLAYED. It is the difference between the start point and the end point selected. This is a basic principle of rounding off numbers. You can reproduce the same error in Excel, which I did here in this forum some years ago. I had two sums side by side in Excel. (1+1=2) and (1+1=3). The first sum used exact numbers. The second used (1.49+1.49=2.98) but I formated them to round to the nearest whole number like you want your millimeter dimensions. Excel dutifully displayed (1+1=3)!

You must think of your model as determining the dimensions, not the dimensions determining the model. I suggest you make a special dimension format called exact. Set it to millimeters, but use the custom format and set it to 4 or 5 decimal places. You can then take a dimension and toggle it between the exact format and your usual format. Most likely you will find that your 200mm has a trivial decimal inaccuracy that disappears when rounded, but reappears when you make the overall. You can also use the measure tool, which does not round off. I always remind my students that there is a difference between measuring and dimensioning. (As a BIM manager, I teach Revit classes).

Let's not get into the hidden reality that all this is being done in binary inside the Intel chip and has to get converted to decimal, itself a messy procedure. People get doctorates in this stuff!!!!

EDIT: AUGI coverted 2.8 and a parenthesis as a smiley face 1.49+1.49=2.98 is what I meant!

[patricks]

We use Imperial and keep all dimensions set to the finest accuracy of 1/256" for this reason!

I would change your units to display a few decimal places on your MM dimensions, then take some time to go around and clean up stuff that shows with garbage dimensions. If you prefer to not have decimal places showing on your dimensions, you can set your units to "Suppress Trailing Zeroes" which will only show decimal places where there are unclean dimensions.