Hijack!

[Mark J]

I believe that the radius of the curve is being magnified on the convex curves and minified on the concave curves by a factor determined by the radius of the base.  

Think of a marching band turning a corner.  The person on the outside of the curve has a longer path describing an arc with a larger radius of curvature than the guy in the center of the line.  

[Coop]

14 minutes ago, Chestnut said:

Is your flush trim bit square to your router table? Is the guide bearing clean? Is the bit worn out?

I don't know if the bit being worn out matters but mine is getting pretty bad and i feel like it's not working as well as it should. I want to get one of those fancy big daddy bits but the cost is standing in the way.

Everything is smooth and goes back together well, it's just that the laminated strips don't match the template. It's not a pita, just a head scratcher.

[Coop]

6 minutes ago, Mark J said:

I believe that the radius of the curve is being magnified on the convex curves and minified on the concave curves by a factor determined by the radius of the base.  

Think of a marching band turning a corner.  The person on the outside of the curve has a longer path describing an arc with a larger radius of curvature than the guy in the center of the line.  

I was thinking it might have something to do with something that I had absolutely no knowledge of . But, by the same token, I can't imagine the 3.25" distance from the pattern to the bit, affecting it anymore than the thickness of a pencil lead when tracing around it. But, I will accept this answer and continue not loosing sleep over it!

[treeslayer]

32 minutes ago, K Cooper said:

I believe that the radius of the curve is being magnified on the convex curves and minified on the concave curves by a factor determined by the radius of the base.  

Math, i knew it!!

[Chestnut]

1 hour ago, Mark J said:

I believe that the radius of the curve is being magnified on the convex curves and minified on the concave curves by a factor determined by the radius of the base.  

Think of a marching band turning a corner.  The person on the outside of the curve has a longer path describing an arc with a larger radius of curvature than the guy in the center of the line.  

 

1 hour ago, K Cooper said:

I was thinking it might have something to do with something that I had absolutely no knowledge of . But, by the same token, I can't imagine the 3.25" distance from the pattern to the bit, affecting it anymore than the thickness of a pencil lead when tracing around it. But, I will accept this answer and continue not loosing sleep over it!

There doesn't have to be any math if you don't want there to be. As long as the router doesn't drift off the template and there is only 1 pass made the difference radiuss* don't matter. The diameter of the router bit and the path "automatically" account for the difference in radius.  You could do this free hand if you were good enough with a router.

Make sure the pieces you are inserting are the exact same thickness as the diameter of the bit that you are making the initial cut with. any variation in thickness will cause a gap.

* I don't know how to spell the plural form of radius.

 

To the technical the radius on a convex is increased by exactly that of the diameter of what ever is following the template. There is no factor. If the inside radius is 10" and the router base is 3.5" the new radius is 13.5" at the center. if your using a half inch bit the inside radius is 13.25" and the outside is 13.75". The same goes for concave except in subtraction. so it'd be 7.5" center with the outsides being 7.25" and7.75".

No if you want to talk spirals i'm going to throw my hands up and yell ARGH!!! ask a railroad engineer!

 

[Immortan D]

Ken maybe your router bit is not really at the center of your round base and you're getting isochronous curves.

[wdwerker]

Bit off center in base or base slightly out of round is my best guess. The base on my old huge PC 3 hp router that predates the 7518 is offset exactly 1/32 different from one side to the other. The instructions said to set up your straight edge or template exactly to the smaller side then make the first cut using The bigger side. That lets you trim 1/32 off cleanly to reach the proper size. So it could be something like that .

[Pondhockey]

7 hours ago, Immortan D said:

Ken maybe your router bit is not really at the center of your round base and you're getting isochronous curves.

This,

if you didn't center the router base.

 

[Coop]

I have a centering pin to be used with a guide bushing but it fits my PC and not the Bosch that I used. I just measured from the cutting edge of the bit to the outside of the base at severalpoints with a cheap caliper and was spot on. Y’all could be nice and say “yeah, it happens to me all the time, live with it”  

5 hours ago, wdwerker said:

 The instructions said to set up your straight edge or template exactly to the smaller side then make the first cut using The bigger side. That lets you trim 1/32 off cleanly to reach the proper size. So it could be something like that .

 I did something similar. I cut my template just a little long so that it overlaps each end of the cutting board for stability. But I marked the template at each end of the cutting board so I could place it in the same place each time. 

[wtnhighlander]

Ken, I find accuracy improved by working from a single center mark, rather than two edge marks, in situations like this. 

Also, clamping a guide rail or two on the overhang of the template will register it squarely against the work piece, if it must be removed and replaced.

[Coop]

Thanks guys for your suggestions and help. Having said all of this, everything clamps and glues up nicely when it’s all said and done. I was/am just bewildered as to why, after glueup, that the template doesn’t match. Que sera, sera. 

[Chestnut]

9 hours ago, K Cooper said:

Thanks guys for your suggestions and help. Having said all of this, everything clamps and glues up nicely when it’s all said and done. I was/am just bewildered as to why, after glueup, that the template doesn’t match. Que sera, sera. 

It's not supposed to. The curves are different because of the offset from the router base.

[Coop]

Nut, definitely not arguing but simply wondering. If I could have used a guide bushing in my router plate, would the template and laminated strip design be the same?  

[treeslayer]

Dang it Coop, now i've got to build one just to answer some questions i have about this. i'm beginning to think no mistake was made but if there was that's how i learn things, from other peoples mistakes but mostly from my own errors, if you could see my scrap bin you would understand.

[Mark J]

@K Cooper it all depends on how close the cuttig surface is to your template surface.  

If you're using a router bit that is say 1/2" diameter and it has a 1/2" guide bearing then the cut surface created by the router bit will follow the template exactly.  

If you use a 1/2" guide bushing on the router base you will get the same result.  

If your guid bushing is 1" then your cut surface will be subtly different than the template.  Your bushing is 1/2" larger than the bit.  So the bit will be 1/4" away from the template.  Each curve of the cut surface will have a radius of curvature increased or decreased by 1/4".  Small difference you probably won't notice to look at it.

But if the router base stands the bit 3" away from the template the +/- 3" change to the radiuses of curvature is going to stand out.

Here's another way to think of it.  Imagine you take a piece of paper and draw an arc with a compass.  Now take a second piece of paper and draw a second arc with a radius 3" greater than the first circle.  The curves are not going to match up, but they followed the same path so to speak.

[Coop]

1 hour ago, Mark J said:

Here's another way to think of it.  Imagine you take a piece of paper and draw an arc with a compass.  Now take a second piece of paper and draw a second arc with a radius 3" greater than the first circle.  The curves are not going to match up, but they followed the same path so to speak.

I hate to belabor this topic but since I started it and there are enough of you fine guys that are willing to help, then I’ll hang in there until I understand or give up. . Mark, in your example, if I take a compass and make a 180* arc (semicircle) starting on a line and finishing on the same line, then extend my compass by 3”, my arcs are equadistance from each other. I know, you’d think I was trying to wrap my arms around one of Newton’s law or something. Sorry for being so dense. 

[Chestnut]

2 hours ago, K Cooper said:

Nut, definitely not arguing but simply wondering. If I could have used a guide bushing in my router plate, would the template and laminated strip design be the same?  

Theoretically no but to your eyes probably. How big of a gap can you see these days 1/16th"?

It's all relative if the curve of the arc is large and the offset is small you won't be able to tell the difference if the curve of the arc is tiny and the offset is the same the difference will become noticeable..

It's like if you have $10 and i take a penny, that's hardly noticeable but if you have $1 and i take $0.50 you can't buy the Saturday paper any more.

[Spanky]

Coop, can see with his new glasses a 5 micron gap! 

[wtnhighlander]

36 minutes ago, K Cooper said:

I hate to belabor this topic but since I started it and there are enough of you fine guys that are willing to help, then I’ll hang in there until I understand or give up. . Mark, in your example, if I take a compass and make a 180* arc (semicircle) starting on a line and finishing on the same line, then extend my compass by 3”, my arcs are equadistance from each other. I know, you’d think I was trying to wrap my arms around one of Newton’s law or something. Sorry for being so dense. 

You're on the way, Coop. Now draw an opposing 180* arc beside each of tge originals, to form a pair of 'sine' waves, having different radii. Now cut along both waves, remove the waste between, and see how they don't fit together. 

This will be greatly exaggerated, compared to the curve in your board, since the arcs in the board were not  180*.

[Mark J]

1 hour ago, K Cooper said:

I hate to belabor this topic but since I started it and there are enough of you fine guys that are willing to help, then I’ll hang in there until I understand or give up. . Mark, in your example, if I take a compass and make a 180* arc (semicircle) starting on a line and finishing on the same line, then extend my compass by 3”, my arcs are equadistance from each other. I know, you’d think I was trying to wrap my arms around one of Newton’s law or something. Sorry for being so dense. 

Coop, you are correct when you note that the distance between the two curves is a constant 3 inches.  But the two curves will not have the same radius of curvature.  I think that last concept, radius of curvature, may be what's giving you pause.  And that strikes me as little hard to explain.  But here goes.  

This time instead of using the same center point to draw these two 180* arcs we will let the center points be different, but we'll make each arc start at the same point.  Begin by drawing the first arc alone.  Now draw a line from the center point of that arc to the point where the arc begins.  Let's call  the point the "start point" and the line the "base line".  Now add 3 inches to your compass.  We're going to keep the center point for the second arc on the "baseline",  but we are going to place the pencil point at the "start point".  Now when you draw the second arc it will not be equidistant to the first.  Furthermore the larger arc will extend farther away from the baseline.  The two curves have very different radiuses of curvature.  The larger arc is a "gentler" curve.  

[Coop]

NOW I’ve got it! You know damn well I’m lieing. Hey guys, Nut, Steve, Mark, and others that have tried to help, appreciate all of you. I’m just going to accept it as is and as long as they glue back together as they are suppose to, I will question it no further.   Coop

[JohnG]

Mark J got it. The fact that the template is a compound curve (not just a single curve, like a D shape) makes the difference especially obvious at the transition between the curves. Here's another visualization of it that matches your picture.

Imagine the white curve is the template. The dots are offset equally from the curve, like the bit in the center of the router base. At first glance, it looks like the shape matches.

 

The red(ish) line is if you moved the template over after making the cut. Now it is clear that the curve is not the same. The curve is close for parts of top curve and parts of the bottom curve, but there is a clear gap between the two.

 

[Immortan D]

About JohnG's visualizations,  looks like the curve could be somewhat fixed by running a second pass in the opposite direction.

[wtnhighlander]

Coop, did I miss a step in your process? Usually, these curves are started by following the template with a router to plow a shallow groove. Then cut through the middle of the groove with a band or jigsaw to seperate the pieces. Last, use a bearing-guided pattern bit to flush trim the cut edges to match the original groove. Regardless of how the original groove changed from the pattern, with this method the two halves should match, with the filler between them. If you just press the two curves together, they will be mismatched by the different radii resulting from the diameter of the original groove.

Is it possible the filler strip was slightly off from the original groove?

[Coop]

Yeah, I think @Mark J nailed it and @JohnG provided a good visual of the situation. And from what I understand, or not, the closer I am from the bit to the template, the less noticeable the difference will be. So using a guide bushing in my router plate is better than running the router plate against the template. Doing so means that I need a longer than 1” cutting edge bit that I have . Usually I seat the bit all the way into the collet, then retract it a tad. How far can I safely extend the bit from the collet to get extra length?