Revit Accuracy – Must step back to AutoCAD??

[steve.clark]

We received an architectural model to use as a background that contained radial and circumferential grids, and made our structural grid system from the this model. In the course of working with the model, someone decided to verify that radial grids all met at one point, and sure enough, they didn’t. When we zoomed in, there were at least two workpoints. I say at least, because as we continued to zoom in, the grid lines disappeared one at a time, as if we zoomed past them. When we exported the model to dwg and checked, we discovered there was no workpoint, just a bunch of lines all intersecting each other. Admittedly, the intersections were very close together.

Can it be that Revit’s accuracy is limited to something on the order of 1/256 inches? Must we step back to AutoCAD to check these types of things?? Somebody, help restore my faith! Thanks.

[david_peterson]

We received an architectural model to use as a background that contained radial and circumferential grids, and made our structural grid system from the this model. In the course of working with the model, someone decided to verify that radial grids all met at one point, and sure enough, they didn’t. When we zoomed in, there were at least two workpoints. I say at least, because as we continued to zoom in, the grid lines disappeared one at a time, as if we zoomed past them. When we exported the model to dwg and checked, we discovered there was no workpoint, just a bunch of lines all intersecting each other. Admittedly, the intersections were very close together.

Can it be that Revit’s accuracy is limited to something on the order of 1/256 inches? Must we step back to AutoCAD to check these types of things?? Somebody, help restore my faith! Thanks.

That's why I NEVER copy off an arch drawing. I've never received one yet that hasn't had something geometrically wrong with it. I was given a floor plan once, drew the building off the dimensions given on the plan. The building was over 2 feet short of having closed exterior. Good Luck.

[.chad]

That's why I NEVER copy off an arch drawing. I've never received one yet that hasn't had something geometrically wrong with it. I was given a floor plan once, drew the building off the dimensions given on the plan. The building was over 2 feet short of having closed exterior. Good Luck.

um. how did it end up 2 feet short? my roommate actually told me about a drawing he had recently (he is a mech contractor) and the dimensions and actual construction were about a foot off.

it not really that hard to make dimensions in cad add up and be correct...

[Andre Baros]

I'm not sure about Revit's accuracy, but I have heard about it having issues with loosing fractions. i.e. 1/3+1/3+1/3=.9999 instead of 1. If your drawing details up a masonry wall and array that brick by 2-2/3 inches you can loose 1/2 an inch by the top of your wall. So if your grid lines got moved by someone a few times, you could end up with a problem.

The only time I've heard of buildings being a foot or two off was when someone at the local AOR messed up a metric to imperial conversion on drawings from a foreign design architect.

[steve.clark]

The shocking thing here is not that the architectural model was off, as we can all attest. It’s that Revit couldn’t tell us it was off. We had to use AutoCAD. I was hoping someone knew of a better way. This feels like a small step back for my project, but a giant leap back for Revit.

[cadkiller]

Can you post some files for us to look at?
I have an FTP site that you can use if the files are too big.
I would like to see exactly what you're talking about.

[david_peterson]

um. how did it end up 2 feet short? my roommate actually told me about a drawing he had recently (he is a mech contractor) and the dimensions and actual construction were about a foot off.

it not really that hard to make dimensions in cad add up and be correct...

It's real easy to screw them up when you override the dim text and not change the wall location.

[hand471037]

Can it be that Revit’s accuracy is limited to something on the order of 1/256 inches? Must we step back to AutoCAD to check these types of things?? Somebody, help restore my faith! Thanks.

You know, it's a funny thing. I generate models in Revit that I export to a CNC routing table. I've noticed some strange things too. I'll dig into it more and report back.

One reassuring thing I can say is that when I take elevations, plans, and sections of my models, and export them to DXF/DWG for the toolpathing, they always come out perfectly accurate. The cuts are dead-on. So the representations of the model itself are 'accurate' as far as I can tell. However with linework, detail lines, and graphic elements within Revit I have noticed some strange behavior, esp. in regards to curves and things.

One thing I would do is you say that you based your grid off of someone else's grid. Are you 100% certain that their grid is correct? If it's a radial grid, and their math is just a touch off, it could very well lead to a tiny inaccuracy like this. I'd take the dimensions, and break out the Trig and pencil and double check everything if it was really important. Garbage in, garbage out and all that...

[.chad]

It's real easy to screw them up when you override the dim text and not change the wall location.

i guess. i had correct dimensioning hammered into me my first few weeks here - my boss set my dim style to 1/128 and then told me that the only acceptable fraction was 1/2". i learned how to fix them after a few days of going home with some *big* headaches.

[david_peterson]

i guess. i had correct dimensioning hammered into me my first few weeks here - my boss set my dim style to 1/128 and then told me that the only acceptable fraction was 1/2". i learned how to fix them after a few days of going home with some *big* headaches.

I hope you were using nominal CMU