# Conic Surface Input

• posted

I need to input a conic surface (hyperbola) into SW. I have been searching for a way to do this by inputting the vertex radii and the conic constant, but there dosent seem to be a direct way to enter this into SW.

I know about the xyz point table file which is an indirect method I might use if nothing else works.

Has anyone found a way to input a conic surface by entering the vertex radii and conic constant?

The conic constant is an optics term which is equal to the negative of the eccentricity squared - FYI

Michael

Loading thread data ...
• posted

I have seen posts byt others trying to make hyperbolae. The best analytic solution I have seen was to make a cone surface and intersect it with a plane. Good luck dialing in the particulars, though.

• posted

Is this supposed to be 100% accurate or just a representation? In other words is it to be CNC machined or made another way?

I can give a method to construct a hyperbola.

T> I need to input a conic surface (hyperbola) into SW. I have been

• posted

It sounds like you have an equation to work from, input the equation into Excel and generate a series of points. Then use the Curve Through XYZ Points to insert the point file. Last time I did this (SW 2005) I was not able to use the curve directly, I had to Convert it into another sketch and then create the feature. I was creating demo geometry for a Calculus class so I had the equation and limits to start and stop the curve. With the current functionality in the 3D Sketch you might be able to utilize those tools as well.

• posted

Sorry, just caught the XYZ reference in you post. My bad.

• posted

One other thought, have you looked at the Sketch Tool, Parabola and mirroring it across a center line?

• posted

Thanks to all for the answers.

Yes is has to be accurate it will be single point diamond turned. I did the xyz output table in the xy plane with z=3D0 and this worked to get my curve in SW. It would be best if SW had a way to input vertex radii and then a conic constant and then one could draw any of the conics - sphere, ellipse, hyperbola, and parabola. Michael

• posted

How accurate? I'd guess you will see approximation errors at about four decimal places even with the cone / plane intersection curve.

I don't know, but am wondering if for graphic purposes use what's convenient and for manufacture furnish the parameters wouldn't be the "way to go".

• posted

That's why using the curve-thru-points or a spline will be flawed. SW does not allow or control of degree. With the extra degree, a spline weaves in and out of the conic curve between defning points, even with tangent and curvature control.

• posted

If you're interested in trying a cone / plane intersection curve ...

formatting link
(I also think my original 'accuracy' estimate should be revised. With respect to the intersection curve itself, unless you're creating really 'peaky' curves, i.e. rho >= .9, you can probably expect six to eight decimal digits of accuray. Check if it's critical, though. It's easy enough to do.)

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.