So I have this symmetrical press that I am making and the faces of the lofts on the model have the same surface area. When I offset the surface it ends up being no longer symmetrical it's only a difference of about 0.0047mm but when I do an extrude down to the surface to make the female half it gets amplified to about 0.007mm and it causes the board to be slightly warped. Is this due to the way the offset is calculated, weird geometry, me being an idiot, a combination or something else?
@TimelesslyTiredYouth wrote:
The issue is likely due to how the offset surface calculation handles curved or complex geometries. CAD software often approximates offsets based on the normal vectors of the surface, which can introduce slight asymmetries, especially for lofted surfaces with varying curvature or slope. These minor deviations can then amplify when additional operations, such as extrusions, are performed. This is not unusual and isn't necessarily user error. To mitigate this, you could use manual adjustments, finer resolution settings, or re-evaluate the loft's geometry to ensure it is as mathematically symmetric as possible before applying offsets.
Kind regards
Rickty
CAD software does not usually approximate offsets.
Doing so would go completely against the purpose of CAD software, which is to provide precise models for the manufacture of physical artifacts.
CAD software, as opposed to mesh modeling software, represents 3D geometry with precise mathematical descriptions in the form of analytical and NURBS surfaces, often stitched into a BRep.
@muller.anthonyza please share the model. Export as a .f3d (or .f3z) file and attach it to a post.
It is much more likely that the measure tool makes approximations in its calculation of the surface area of a curved surface!
Even if you design the object in 2 different CAD software, and the objects are identical, and you measure things, you'll get differences.
We discussed this a few years ago, but it's very hard to track down such threads on the forum.
I am 60 years old and have worked in and around manufacturing for almost 40 years. I have over 30 years of professional CAD experience. I started working in 3D CAD in 1998 with SolidWorks and use a variety of 3D CAD software packages. For non-CAD modeling, I often use Blender (which started in the early 2000s) and occasionally SideFx Houdini.
What's your experience?
However there is a difference which is noticeable but I still cannot tell if it's my eyes deceiving me or if it is just skewed.
Hi Anthonyza,
It is possible the two faces were not symmetric to begin with. Please share the file here. I would like take a look to understand the behavior better.
Many thanks!
I suspect it is a bit of an optical illusion (exaggerated by the crosshatching), and also a possible misunderstanding of what offset does. Offset is measured along the surface normal direction at every point. I've added some arrows to your image to illustrate. Just visually, this looks like a uniform offset to me. I am reasonably confident that it is mathematically uniform, as well.
@TimelesslyTiredYouth wrote:
...Do you not question why i know this at my age, I would like to think you remember from a past conversation.
No, I don't question that. I don't think what you possess can be called "knowledge" yet.
You lack the ability to apply in practice what you write about. This is the second time, or third time you provide feedback to another user that sounds knowledgeable to to an inexperienced and unassuming user, but does not really apply to the problem at hand.
Also, if the model really would have such problems, you would unlikely be able to pinpoint the root cause and provide a solid fix for the problem.
You need to stop googling and reading things that might make theoretical sense. You need to start doing things. Download a model, with problem, analyze it. Play around with it. See what you learn from it. Listen to what other more experienced users have to say. Go back and experiment a bit more with the model. Maybe you can fix the problem and provide another user with real help other then general feedback.
Then you will also be able to determine if the user can improve their modeling skills or if the problem is really that the software isn't perfect.
Once you do that across a good spectrum of problems you can consider yourself somewhat knowledgeable.
That is how I would characterize myself "somewhat knowledgeable" 😉
Just for the record. I learned none of this in College! Computers like we have today did not exist. The internet did not exist! If you think you don't have time google stuff, real AND practice you have completely lost me!
You clearly are an intelligent person. Now, go practice and put what you read to practical good use.
@muller.anthonyza wrote:
here is the f3d file, beware I am a self taught 19 year old so some things may be a bit poorly done
No, the model is not done poorly!
But it does have room for improvement.
The first sketch is unnecessarily complicated. I would strongly advise against mirroring fit-point splines (works fine with control point splines). Editing a mirrored fit point spline can brake the symmetry. I would avoid mirroring entire sketches completely!
Mirror 3D geometry.
Similarly, don't mirror an entire sketch half. Extrude and then mirror geometry.
It creates more robust and much easier to debug sketches. It also creates computationally aster models.
Also, fully constrain and dimension sketches!
This fillet I would not sketch. I'd sketch a sharp corner and then apply the fillet as a solid modeling feature. Again, easier to sketch etc. Also, because the blue sketch objects are unconstrained, did you want for the fillet to terminate at the vertical line?
Her is what such a workflow would look like:
Edit: You will notice that I use mirroring of geometry and features a lot. That will absolutely guarantee that the geometry is perfectly symmetric, down to the internal precision of the geometric modeling kernel.
