Hey all !
So I’ve been using Rhino for quite a few years and usually chose between large and small objects when it cones to templates intuitively. I know it is about the tolerances, but, when is an object large enough to be considered a “large object” (is a care a large or small object? is large compared to a ring, but small compared to a 5 story building…), and does it even matter? I mean, if an object is large but made of intricate small objects, how do you choose? I’m trying to get a better understanding of this…
Thank you for your input!
1 Like
The templates shipped and installed with Rhino are for convenience and are nothing special. The templates labeled “Large” have an absolute tolerance of 0.01 and 500 grid lines. The templates labeled “Small” have an absolute tolerance of 0.001 and 200 grid lines.
The absolute tolerance should be at least one order of magnitude smaller than the minimum size of details being modeled and the smallest manufacturing accuracy. An absolute tolerance tighter than needed is likely to result in more control points than needed with various operations.
The grid settings can be what ever the user prefers. The grid is just a modeling aid and has no effect on the geometry, other than providing a method to locate points if Grid Snap is on. Usualy it is preferable to model with Grid Snap off unless there is a specific need to snap to the grid. Many users like to model without a visible grid.
Any file can be saved as a template file using SaveAsTemplate. It will then be available for selection when a new file is started.
There are a lot of ways to look at tolerance. The smallest should be an order of magnitude smaller then you need to fabricate. That means a ring will be quite small, but a larger bridge or city block might be quite a bit further.
A conservative, lazy rule of thumb is to only count on 8 significant digits. So if you need 0.1 mm accuracy, then the tolerance might be set to .01, which means the maximum size of the object should not exceed 999,999.99 mm. You can actually exceed this a bit, but know it starts to get challenging out far.
Where we see this the most is a building set hundreds of kilometers from the origin because it is set at civil coordinates but then trying to calculate the intersection railing post with a handrail or something like that. We run out of digits to represent the problem.
Another challenge is working with other software that has hard coded tolerances. There is not always an easy way to set a larger or smaller tolerance later. Something like working with Revit, where it’s tolerance is 1 mm. That can be a challenge with small intricate parts and fabrication level models. Or solidworks which practical tolerance is around 0.0004 unit or so.
Not sure that helps, but this is my pragmatic approach to it.
I guess it is probably also important to note they surface fitting trimming and joins will use the current tolerance. While the majority of calculations will land within tolerance, some calculations may get close to the size of tolernace. And this tolernace is then saved in that trim/join.
So, if the tolernace is changed mid-modeling all the existing trims and joins are not re-calculated to the new tolerance. A process or refitting and re-trimming would need to occur, which can be challenging in certain cases.
Set the tolerance at the beginning of a project as tight as may be needed. Relaxing the tolerance and keeping the larger tolerance during modeling should not cause problems. Tightening the tolerance during modeling, even back to a previously tolerance, may cause problems.