arc problem

[nestly]

I still think that flat spot is there

 

I've switched back to my original accessment' date=' the taper does extend all the way to the 72R arc. I think the red circle contributes to the illusion.

 

--> Proof

 

Eyetest, Which line is the extension of C. A or B?

[Jack_O'neill]

Everything is flat in that drawing. Even your C has flat spots on it!

[SLW210]

TTR will place the ARCs in question, once construction lines are drawn and then the tapered lines. The 26 and 18 are where the tapered lines meet the arc end points.

[Jack_O'neill]

I was kidding about still thinking the flat spots were there. I think it really was an illusion created by the dimension lines and the resolution of the jpg. It makes no sense for a flat to be there. Just me getting old and not seeing so good.

[nestly]

The method you used demonstrates what we were talking about, the drawing is insufficiently dimensioned for production. Even in your drawing, neither the R72 nor the R35 is actually tangent to the taper. ( ie both arcs can be extended/trimmed to intersect the tapered line at two locations)

[SLW210]

Yes they are tangent to the lines and the lines make up the end points of the taper. And if I hand that drawing to a machinist he had better be able to make it.

[nestly]

Yes they are tangent to the lines and the lines make up the end points of the taper. And if I hand that drawing to a machinist he had better be able to make it.

 

Nope, they are not. In fact if you zoom into the R35, you'll see that you trimmed it at the 2nd intersection as indicated by the overlap in geometry. The same condition exists at the R72,(ie the line can be extended to intersect the arc at a 2nd point that's more than 4 units away)

[eldon]

I suppose that I should post a drawing of my efforts so that it can be picked to pieces

ArcMkE.dwg

[nestly]

I suppose that I should post a drawing of my efforts so that it can be picked to pieces

 

It's not my intention to disparage, only to demonstrate that the drawing in question is not adequately dimensioned, and is apparently problematic even for experience AutoCAD users to replicate precisely. I expect you'd concede that you also "guessed" and/or settled for "close enough", right?

 

There is a solution, but at this point, my math isn't up snuff, and I'm not even sure how to get both ends perfect using AutoCAD without the use of constraints.

[Jack_O'neill]

I zoomed up on Hulk's drawing, drew a line from the endpoint of the angled portion off somewhere just to mark the endpoint. Then I turned on the grips, and moved the end to the tangent osnap of the 35 radius and asked for a distance. Before that, however, I reset units to 4 decimal places and viewres to 20000. You're both right...the end was indeed tangent to the arc. As for decimals there the distance from the end of the orginal line to my line was 0.0000. There is some extra circle there that didn't get trimmed off.

 

A CNC programmer would have no problem making this part dimensioned as is. Tool paths are generated off the model, you don't even really need dims to make it that way except to know how big a piece of metal to grab.

 

A conventional machinist is going to have to drag out his calculator and do some "figgerin" to be able to make this part. A conventional machinist, on a mill with a rotary table will need to know the angle of that taper so that when he cranks the rotary around, he knows where to stop and start cutting the straight portion. He'll also need to know approximately where the straight portion stops (the tangent point) so that he doesn't run past it. He'll also need to know the tangent points in degrees around the D35 end, so that he doesn't cut into the that R35 fillet. Same at the other end for the holes.

 

Long story short, making this on a CNC, it's good as is. On a conventional mill, you ain't got near enough info yet.

 

edit--now that I've sat here and looked at it some more, the conventional machinist will need to know where the center point of every radius is, because he's going to have to drill and ream a hole for the rotary table center pin at every one of them. Well, the little groove down the middle could be accomplished with the right size cutter, but the rest, gonna have to know where they are.

Edited August 1, 2011 by Jack_O'neill

[nestly]

You're both right...the end was indeed tangent to the arc. As for decimals there the distance from the end of the orginal line to my line was 0.0000. There is some extra circle there that didn't get trimmed off.

 

I don't think so. There is only one tapered line' date=' and there's only one arc at each end of the line. The endpoint of the R72 arc and the endpoint of the line are coincident, but they ARE NOT tangent unless I missed the memo announcing the redefinition of the word Tangent. The endpoint of the R35 arc and the endpoint of the line are NEITHER coincident, NOR tangent. The line appears to be snapped to the quadrants of each arc and then trimmed off. By definition, unless the line is horizontal (0 degrees), it cannot be tangent to a 270 degree quadrant.

 

Try it like this.... http://screencast.com/t/mfqpgWuKN

[Jack_O'neill]

I don't think so. There is only one tapered line, and there's only one arc at each end of the line. The endpoint of the R72 arc and the endpoint of the line are coincident, but they ARE NOT tangent unless I missed the memo announcing the redefinition of the word Tangent. The endpoint of the R35 arc and the endpoint of the line are NEITHER coincident, NOR tangent. The line appears to be snapped to the quadrants of each arc and then trimmed off. By definition, unless the line is horizontal (0 degrees), it cannot be tangent to a 270 degree quadrant.

 

Try it like this.... http://screencast.com/t/mfqpgWuKN

 

All I can tell you is what I did. Take it or leave it. I did it 3 times, with the same result all three times. If you turn the grips on that angled line, grab the one on the end, and move it tangent to the arc, that's where it puts it, right, wrong or otherwise.

 

In any case, I've burnt all the time on this I care to, got new work to do.

[nestly]

That's pretty interesting because when I do it your way, the discrepancy is very obvious. (more than 1 & 2 whole units respectively)

 

http://screencast.com/t/1Odj7VapYhW