Snail cam manufacture ?

[picclock]

I have a requirement coming up for a 'snail' shaped cam. basically this will raise a rod then drop it, rather like the water wheel powered hammers but only once per revolution.

Has anyone got any ideas on how to fabricate something like this. With CNC mill its an easy solution, but I don't have the funds for that kind of upgrade. The only other option that springs to mind is to calculate the distances from the centre/ verses the angle and mill accordingly.

Any better ideas very welcome

best regards

picclock

 

[Ken I]

Picclock,
Nothing wrong with your idea - that's how I would do it.

It will leave a cobbled surface but that hand finishes nicely with a file.

I would simply draw it on autocad (less tedious than doing the calculations) with all the cutter positiones arrayed around the cam, place the cam centre on 0,0 and print out the coordinate data.

You can do it as either an angular array (if you intend to use the RT) or as simple cartesian coordinates. If its a constant rate of rise then angular should be dead easy.

Post the cam drawing and I'll do it for you.

I'm guessing Marv's going to pull one of his programs out of his hat for just this.

Ken

 

[mklotz]

...
I'm guessing Marv's going to pull one of his programs out of his hat for just this.

Sorry, I don't have anything explicit for that because there are just so many spirals.

You might want to check to see if it satisfies one of these equations...

Archimedian spiral: r = a * theta

Hyperbolic spiral: r * theta = a

Logarithmic spiral: log r = a * theta

Parabolic spiral: (r - a)² = 4 * a * k * theta

If it does then using the equation to generate your cutting profile will be simpler than picking points off some cad drawing.

 

[Maryak]

Just thinking out loud

IF your lathe has a taper turning attachment ???

A tapered thread is a spiral, (I think) :-\

Over to you. ;D

Best Regards
Bob

 

[tel]

Somewhere here I have an article (hard copy) on turning the scroll for a three jaw chuck, which might get you on the right track. From memory it involved gearing the cross slide to the leadscrew? Anyone else recall that piece?

 

[Ken I]

Don't forget to allow for the cutter diameter - this will screw up any theoretical calculation.

See photo - the cam goes from 0° to 360° but the cutter only goes from 16° to 360°

Ken

 

[picclock]

@ Ken I

Thats a very good point i would likely have overlooked. And the drawing shows exactly what I need. I was considering using the rotary table on its side so that the surface would be cut by the end of the mill, with the mill diameter large compared to the cam width. This would give a slight concave effect to the cam surface if I leave the tool position unchanged, which may be a good thing.

@ MKlotz

Thanks for the formulae. I figured I'd just take the rise hight and divide by 360 to give the incremental change. I will probably print out an angle vers depth of cut table just so there is less chance of me messing up.

Many thanks for the ideas, I was just hoping that someone smarter than I had a better way. But at least now I know that this approach is the best one, short of getting CNC (Drool icon should go here ;D)

Best Regards

picclock

 

[Noitoen]

If the distance of the "step" is greater than the cutter's diameter, you could go from 0 to 360º. On the inside the cutter will go a little past the straight edge and eliminate the radius, leaving a "flat" surface related to the drop. "sorry about my English".

 

[Ken I]

I was considering using the rotary table on its side so that the surface would be cut by the end of the mill, with the mill diameter large compared to the cam width. This would give a slight concave effect to the cam surface if I leave the tool position unchanged, which may be a good thing.

That's better still - mark out and "hack" off the bulk then start high and work down so your cutter ends up in the sharp corner. Again the cutter can "interfere".
This will give you a very slight miniscus but nothing much.

Work everything on 4° (1 turn on th RT) steps - the polygon corners will deviate from true by very little - a quick dress with a file will sort it out.

Alternatively work from the tangent on centreline - no miniscus - but you are going to have to work out the last bits to get into the corner.

Ken