Mechanism idea needed for varying throw on a crankshaft

[picclock]

Hi

I'm (hypothetically - still sketching and thinking) using a shaft with an eccentric and ball race to power a pump piston.

However I'd like to vary the distance the pump piston moves, or the amount it pumps. I've considered several ideas that would work but they seem overly complex or a bit dodgy mechanically. I've come up with three ways so far, but I'm hopeful someone will know a better way.

Using a cam within a cam such that when both lobes are opposite the eccentric isnt :-), and when the lobes are aligned maximum throw is achieved. As this works on the relative position of the cams to each other I would have to figure out a way of rotating them when they are turning. Not impossible but quite complex. Best I can come up with is a gear on either side each connecting to the cams. One connects directly to a central gear the other to an idler then to the central gear. That way the central gear won't turn but if turned the relative position of the cams will change.

Another idea I had is to use a connecting lever to link the pump and eccentric shaft. When the lever is nominally at right angles it would give maximum movement. however when the connecting lever is near vertical it would just cause the pump shaft to move left and right a little.

I also considered a crank with a different offset at each end, and a rod end at right angles to it. Sliding the end along the crank would change the throw. Maybe a rod end with self aligning bearings ?

Does anyone know of a better or alternate way to do this ?

At the moment I'm favouring the last idea as it seems to create less problems.

Any help much appreciated as usual.

picclock

[kwackers]

Isn't this essentially what a steam loco's valve gear is? Simplest is probably a Stephensons valve gear - you'd only need one eccentric though since there'd be no need to reverse it.

[picclock]

Hi kwackers

Thanks for that. I've never heard of a Stephensons linkage before.

I know next to nothing about trains I'm sorry to say, and am a relative newcomer to engineering of any sort. 

The wiki link also led to some other linkage types very similar to this in principle.

Picclock

[RichardShute]

Another possibility is to leave the mechanical stroke fixed, but vary the effective or useful stroke as in some diesel injector pumps. The cylinder is actually a sleeve which can be rotated somewhat (say 60 deg or so) to open a spill port at varying heights due to a helical groove around it. Thre are various other similar schemes.

Purely for amusement, I like your idea of the geared concentric eccentrics. You could alternatively contrive the phase variation by a helical groove in conjuncion with an axial adjustment of a sleeve.

Richard

[Boucher]

The first thing that comes to my mind is the concept used in the variable volume piston type hydraulic pumps that use a wobble plate to drive the pistons when the plate is flat there is no piston motion as the plate is angled the pistons stroke increases accordingly.

[picclock]

Hi Richard

Using a helix to change the relative positions is a neat and simple idea i hadn't thought of  .

I would like to be able to adjust the throw with as little force as possible, hence the initial differential gearing type approach.

Still havn't really made a decision. I may make some models and see which turns out best.

Sadly work seems to have stolen my time at the moment, but if I come up with anything better or weirder that may be of interest I'll post it.

Best Regards

picclock

 

[ken572]

picclock,

Goto your local auto-truck junk yard and find some old swash plate,
AC cylindrical compressers, they made them in 3,5,7,9, piston units.
They shorten and or lengthen the stroke to change the cubic inch's
on the fly depending on the load needed. Vicker's also made hydraulic
swash plate pumps with a hand crank on the front to vary the speed
using the same as above.

Ken.

[BillTodd]

Using a cam within a cam such that when both lobes are opposite the eccentric isnt :-), and when the lobes are aligned maximum throw is achieved. As this works on the relative position of the cams to each other I would have to figure out a way of rotating them when they are turning. Not impossible but quite complex. Best I can come up with is a gear on either side each connecting to the cams. One connects directly to a central gear the other to an idler then to the central gear. That way the central gear won't turn but if turned the relative position of the cams will change.

Like this?...

 I have the same problem as you in that I also need a variable stroke mechanism that can be easily varied while running. This is the crank part of a provisional slotting head design I've been working on; The position of the crank pin can be moved from the centre of the crank (i.e. zero stroke) to the outside (max stroke) by a worm screw. The worm is suppose to be turned by pawls catching on the star head as the crank rotates past. I'm not yet convinced it will be up to the job, so I'm open to new ideas.

(the screw is supposed to be rotating smoothly but I haven't had time to remove the aliasing)

Bill

[picclock]

Hi Bill

I'm finding it hard to understand how your idea works. I did post some sketches that were lost when the site went down which I've attached using rod ends. They both rely on an axial movememt of a sleeve on the shaft. For me that works out quite well as it can be linked to an engine speed governor or similar device.

Best regards

picclock

[crabsign69]

  very clever there bill  i like it.     bet it would be hard to make but i like it.