Elipse

I know it's Sunday but could someone please check the area and perimeter
values of an ellipse in Sw.
I'm not getting the correct properties when using an ellipse in SW.
Until now I've always assumed SW gives me the correct values but I've just
been caught out on a job.
I checked the values manually, they were correct but they did not match the
results in Sw. The SW settings are set to max accuracy
What's more confusing is the area values given when you do a section query
in the sketch is different from the area values of the elliptical face of
the extrusion. I've checked the area/perimeter/volume values of spheres,
cubes etc and they work ok


Regards
Alex

 

Alex,

Since there is no exact formula for the perimeter of an ellipse, it could be
that your calculation is less accurate than SolidWorks. This site has a
discussion and various series and approximations that are much better than
some others I found:

http://home.att.net/~numericana/answer/ellipse.htm#elliptic

I tried three ellipses, with eccentricities of 1.111, 2.5 and 10 and the
perimeter values were all within a few hundredths of a percent of
Ramanujan's first formula, which is probably within the error of the
formula. The areas though were off by up to a third of a percent. This is
very interesting, because the area equation is exact. I guess SolidWorks is
calculating the area based on the curve values. Either the curve locations
are off a bit or there area calculation is not terribly accurate.

Seth,

On my test ellipses, the sketch area values were always closer than the face
values. Since the perimeter values are the same in the sketch and in the
model, why are the areas different? Did different people use different
methods to calculate the area in the two cases?


Jerry Steiger
Tripod Data Systems
"take the garbage out, dear"

 

On my test ellipses, the sketch area values were always closer than the face

Maybe different alogrithms or; surface / solid analysis usually reference a
tessellated version of the model. Area properties may use the same methods but
at different resolutions. Just a bit of idle speculation...