It would be good to post the .3dm file with the curves. From the control points and the isocurves, looks like the curve on the left is a true arc, but the one on the right might be a polyline? Small breaks in tangency can cause filleting problems like your image.
Seems ok this way. Why it s that way, cannot understand, really, as before it was jut a larger number of edges I believe.
Actually I did in both versions import the curve as dxf from Vectric Aspire, my favourite tool to do simple 2D drawings which I work with since I believe 20 years.
Thank you all for your help!
That is the case. Dense polylines that approximate curves like that make for poor geometry. As per above, if you replace the dense polyline with an arc - in this case 47.75 radius makes a pretty good fit - your fillets will work and the geometry will be much cleaner.
Note that in any case the fillets on your extruded object will fail at a radius value of 3mm or above, because the arc in the left corner is 3mm and you canât wrap a larger radius fillet around a smaller one.
Well, personally I would learn to draw this stuff in Rhino, it makes excellent 2D geometry, as good as or better than many other programs - especially if you start working with free-form curves.
Actually I am not really good in Rhino. I know maybe 2% of it and especially I miss to know how to introduce simple geometries or change there sizes by numbers, like rectangles by entering dimensions and change those or at least get the dimension by clicking somehwere, center objects and the like. Most of the time I import 2D curves from Vectric and do create 3D objects like these:
Your large âarcâ is made of countless line segments. The radius is not constant. I would encourage you to lay off the Join and Boolean and FilletEdge, and instead use Rhinoâs practical 2D drafting and 3D NURBS surface modelling tools as mentioned by @Helvetosaur above. After all, that is what Rhino was developed for. With simple curves/surfaces, you will have no problem to realise your designs, and FilletSurf works well every time.
Will have to try to use curves that are made from a curve that is made from an exponential or hyperbolic poit list calculated by an excel table. As you can see in the last image I posted I am developing and manufacturing Speaker horns in different sizes and for different use in Audio systems.
Every item has a different curvature and needs to be designed a way it can be manufactured while keeping Audio Design correct and have some kind of attractive look to it.
I will keep on learning 2 Design in Rhino and hopefully get things in my head strong enough soI get the results easily and better as before. Thanks again for your help!
In Excel create a .csv or ..txt file with the point coordinates. The file needs to have the x coordinates in the first column, the y coordinates in the second column, and the z coordiantes in the third column. Import that file into Rhino.
Use InterpCrv to interpolate curve through the imported points.
Rebuild the curve through the points as a degree 3 (or degree 5) curve. Optional - reduce the number of points in the curve and check the deviation using PointDeviation. Check for smoothness using CurvatureGraph.
Like @davidcockey says, but as you need only a 2D curve to make a 3D surface of revolution (for our NC-lathe, however, Iâd only need 2D), just create x/y values with your audio-compliant formulas (I used f(x)=3*x^3 over the interval 0,2 > 6), export as suggested (make sure to use . for decimals, not the German , because Rhino wants dots) and once imported, you can draw a simple degree 5 curve without much noodling around, which has only a single span, and no jiggly curvature graph as an interpolated one.