radius dish made with thickness sander

[Stray Feathers]

I've been researching how to make radius dishes with a router sled, and was about ready to do that, when I read about another method. Not sure if I am supposed to put a link to another forum, so essentially it is two layers of MDF, screwed together around the perimeter of the dished portion, with a spacer of appropriate thickness between the two layers in the centre. Then it is run through a thickness sander until the sanding marks get out to the edges. When it is taken apart, the result is a dished surface. The original description doesn't explain why you don't end up with a radius on both pieces, perhaps each with half the depth of the desired outcome? Has anyone used a method like this? I have a sander, and don't have a router sled, so I'm wondering . . .

[MaineGeezer]

If I'm understanding the method correctly, it sounds a bit hit-or-miss. I don't think you are guaranteed a uniform radius, or even a radius -- it could be a parabola or something. But I may be missing the concept. Also...I don't understand why both pieces don't deform, as you suggest. Thinking about it a little more, I think you're guaranteed NOT to have a uniform radius, because the edges are screwed together and essentially clamped flat where they are screwed together.

Assuming I'm envisioning the process correctly, it sounds like an ineffective way to make a radius dish to me.

[tippie53]

it may get you close but it is hit or miss

[bunny]

First problem is that you need to calculate the spacer diameter and thickness to get the desired concave dish radius which depends on the mdf disk felxibility. It's a 'reversed engineering' method and may be more difficult that the traditional approaches to making a radius dish.

Another thing is both mdf disks will obvously bend when screwed together so the base (bottom one) needs to be twice as thick or have some kind of reinforcement to remain flat. I have two 15' dishes that both arrived deformed resulting in a smaller working radius and a convex back. By screwing them together back to back I got a perfect 15' on both faces.

[tippie53]

To add to this discussion
As the top disk is sanded the pull of the dish will change as material is removed make the radius far from even.

[MaineGeezer]

Won't it clamp up sort of like this?

[tippie53]

Something but maybe not as much as you have it. What will happen is as the material is removed off the plate it becomes weaker and the bend changes and you can't control that so you may have a dish it a varying radius it will not be a true radius . As the stiffness changes so will the flex and you can understand that easy enough. What you think will be a 15 radius may end up being 15 at one point then 12 at another and maybe 10 at another.
See as the top looses the material it won't be uniform but the bottom will be so as the top is weakened the lower plate will deform the top more as it goes through the sander. Also you are assuming the the plates travel into the sander without the other applied force of the downward springs and this can add 2 flexing forces .
Not way that you can do this with any accuracy.

[MaineGeezer]

My drawing is clearly exaggerated -- I just wanted to 1) be sure I understood the idea and 2) illustrate that the result will not be a uniform radius.

[tippie53]

it is something like that.
You are in the right area .

[Stray Feathers]

These comments all make sense. MaineGeezer, you have described the process well, except I think it would be possible to screw the two pieces together without flattening them. On the other forum (of which I am not a member - the thread came up in a web search) several people thought it was quite clever, and compared it to a method of curving braces, by bending them and then sanding them flat, so when they are relaxed they have a curve on one edge. That sort of made sense to me, but I think there are way too many variables with the dish. Back to the original plan - thanks for input.