Replies

1. 
   BigNat | Apr 30, 2007 12:58am | #1

   I've built a number of elliptical tables and I have yet to find a hard-and-fast rule on leg placement. The best method is to draw it out to scale and decide what looks best. (the method I usually use is to divide the table's length by five and place the legs one fifth in from the ends, but this is simply for lack of a better idea and I only use it as a starting point). 57 degrees will work for the table he made, but since your ellipse may not be the same this may not be the right angle for you.

   Will the table have eliptical aprons? if it does, then one very important aesthetical consideration is getting the legs in the right plane. drawing a line that radiates from the centre of the table and placing the legs square to the line will make the legs look off. As will placing them square to the axes of the ellipse. The same applies to joint lines in a built-up edge. What I find looks best is to select the locations of the legs (or edge joints as the case may be), draw a normal to the ellipse at those locations and place your legs square to that line. In the case of edge joints, the normal itself is the joint line.

   Hope this helps,

   Nat

1. 
   user-156168 | Apr 30, 2007 09:44am | #3

   I'm sure that a lot of things come into play such as height etc. but usually I have pretty good luck with ;   1/3 of the length each way from the center or, 1/6 of the length from each end. 

   I bet there is a formula for proper placement if anyone knows what it is.

   Mighty Oak

2. 
   JiminFla | Apr 30, 2007 11:23pm | #5

   Nat:

   What method do you use to make the elliptical aprons?  Laminate bending?  Coopering?  Steam bending?

    

   Jim

1. 
   BigNat | May 03, 2007 09:52am | #6

   Jim,Yes, Yes and Yes. (Depending on the size and the client's wishes)
   I prefer laminate bending though. Just my own opinion, but I find it easier and less involved than steam bending and coopering (or in this case bricklaying) just doesn't look right (to my eye), unless you veneer the faces. If you're veneering the faces anyways, why not just laminate bend them thus eliminating the veneering step? Nat

1. 
   JiminFla | May 04, 2007 04:27pm | #7

   Thanks, Nat.

2. 
   byhammerandhand | Apr 30, 2007 12:59am | #2

   I think it would depend upon the proportions of major and minor axes. A coffee table that looks like a surfboard will need different placement than a conference table that's nearly round.

3. 
   [email protected] | Apr 30, 2007 11:18pm | #4

   I have a gut feeling, but am not quite sure how to explain it clearly.  You are trying to support the weight equally on all four legs.  And, if you draw a square from leg to leg the area outside the square is the same on all four sides.  I think that you want the same on an ellipse. 

   An ellipse is a circle seen from somewhere other than directly above,  and if you view an ellipse from an orientation outside the vertical to it's plane along it's major axis at the same angle it becomes a circle. 

   Start with a round table, seen from above.  If you draw lines out to the circle from the center to the 45-degree points, on the circle the legs are equidistant on the circle, and where they usually go on a round table.  Now if you tilt the perspective along one major axis, the legs would still be where they were but not equidistant from the axis.  However the area outside a rectangle drawn between the legs would be equal between each pair of adjacent legs, and the area supported by each leg would also be equal. 

   My instinct says the legs still belong there on the ellipse.  And I could draw it in cad, but can't remember the math to figure it out by hand.  I think it is at 90-degrees to the major axis, coming out from the centroids.  There is a name for this point, but geometry was 35 years ago, and I sure can't remember what the name is. 

4. 
   pillais | Jul 10, 2020 08:28pm | #8

   This provides a way to fins the center of gravity of an irregular shape https://www.youtube.com/watch?v=HaaL2wucyRE

   Then the center can be used to create a equilateral triangle with the maximum distance possible to the sides.