How to suddenly stop a gear? (or how to suddenly stop a moving piston)

[pretend engine]

Hey, I'm trying to design a part for a prosthetic leg.

I need to stop a piston moving within a cylinder at any point in its motion. (this will happen when the patient puts weight on his foot)

I was thinking about having the piston turn a gear, and then when it's time to stop it, another gear-like piece will jam into the gear, stopping the system.

The gear won't be spinning fast, but will have a pretty strong force.

Does anyone have any thoughts on how this might be done? Thanks.

 

[etudiant]

If for use within a prosthetic, perhaps the piston could be hydraulically powered, with the patient's weight cutting off the fluid flow or the pump. Closing a valve is pretty gently compared to jamming a gear.

 

[Baluncore]

Gears do not like to stop suddenly. Very few teeth are in contact and must take all the punishment.

One advantage of a hydraulic piston is that a cross flow relief valve can provide a limit to the force when subjected to an impulse. This should protect the patient from further injury.

Why is there a piston in a cylinder, hydraulic or pneumatic?

 

[pretend engine]

Thanks to you both.

I like the hydraulic cut off idea. Is it possible to do that in a way where the fluid will not be restricting the movement of the piston when the valve is open? (not slowing the piston down while forcing fluid out of or sucking fluid into the cylinder)

Baluncore, I'm having trouble finding a site where I can learn about a cross flow relief valve. Can you explain it further?

The cylinder is actually just a guide channel for a rod that will be moving up and down. We need to be able to stop that rod at will. The sudden jarring of the rod will not adversely affect the user.

 

[Baluncore]

Hydraulics are unnecessarily complex and not needed for a rod in a guide.

The best way to stop a rod is with a tapered collet. Collets are effectively a circular wedge, they can be purchased in many sizes for low prices. Hold it open with a gentle axial spring pressure until it is triggered to close by reversing the spring direction. Two opposed collets can lock or release a rod that slides in both directions. Collet taper angle decides how easily it will lock or release.
http://en.wikipedia.org/wiki/Collet
http://en.wikipedia.org/wiki/Collet#ER_collets

 

[pretend engine]

Hydraulics are unnecessarily complex and not needed for a rod in a guide.

The best way to stop a rod is with a tapered collet. Collets are effectively a circular wedge, they can be purchased in many sizes for low prices. Hold it open with a gentle axial spring pressure until it is triggered to close by reversing the spring direction. Two opposed collets can lock or release a rod that slides in both directions. Collet taper angle decides how easily it will lock or release.
http://en.wikipedia.org/wiki/Collet
http://en.wikipedia.org/wiki/Collet#ER_collets

Great idea! I think so at least, it's hard for me to picture.

So friction stops the rod? Do you think that can hold up a persons weight?

Can you recommend an example that I can search for? The only context I can find is drill chucks.

 

[Baluncore]

Supporting a human with a collet is easy. Collets for holding a tool or work in a lathe or mill are optimised to be clamped tight with a screw cap. You would be better allowing the collet to slide on the rod, while seating when needed, in a tapered part of the guide. No screw clamp is needed.

Because the collet is tapered and you can select the taper, you decide the effective slope of the wedge. A gentle wedge or taper will clamp tight and be very hard to release. A steeper taper will be easier to release, but needs more axial force to hold the rod tight. If arranged correctly, you have the weight of the patient to assist in locking the collet to the rod. The pressure on the foot can operate the axial mechanism that causes the collet to clamp the rod.

 

[pretend engine]

Does this sketch look like what you're picturing?

http://imagizer.imageshack.us/v2/150x100q90/9/vn49.jpg

 

[Baluncore]

Sort of, but your wedge looks a bit steep. I would guess at a 30° included angle.

Avoid reactive material in sliding contact. A brass collet on a stainless steel rod might be OK.
You will need to calculate friction coefficients between inner and outer collet surface contacts.

In your diagram I have no idea what is connected to what, or what your symbols are.
This is a physics forum, mind reading is not an acceptable topic or reliable tool.

 

[pretend engine]

Sort of, but your wedge looks a bit steep. I would guess at a 30° included angle.

Avoid reactive material in sliding contact. A brass collet on a stainless steel rod might be OK.
You will need to calculate friction coefficients between inner and outer collet surface contacts.

In your diagram I have no idea what is connected to what, or what your symbols are.
This is a physics forum, mind reading is not an acceptable topic or reliable tool.

OK, thanks. F is the force being applied and V is the velocity vector of the rod.

By reactive materials, what do you mean?

Do you think the collet will unclamp easily, to allow the shaft to move freely again, or would you spring load it?

Again, thanks for the help. This seems like it will be perfect.

 

[Baluncore]

By reactive I mean chemically reactive between each other or in the warm wet environment.

The angle decides the locking and ease of release. A “morse taper” is a gentle taper and so is described as self locking. A collet must be steeper so it will come apart without a hammer.

I do not know your system states or mechanical arrangement. I guess a spring would hold the collet sufficiently to lock it, then a force opposing the spring on the collet would release the collet in the taper when load was released.Beside friction, as used with continuously variable collets or wedges, there are a couple of discrete ways.

Firstly, a gear rack (expensive) could be used with a short length of the same size rack as the “grip” rather than a pinion. See http://en.wikipedia.org/wiki/Rack_and_pinion The short rack section could be moved towards or away from the rack to lock or release.

Secondly, a threaded rod can be gripped with half nuts. Rotation is not needed. This is a cheap and available equivalent to the rack and rack. http://en.wikipedia.org/wiki/Split_nut

 

[Baluncore]

You need to avoid a self locking taper by using a more open angle and allowing the weight of the patient to clamp the taper. If your angle is too open there will be insufficient clamping force.

From page 2 of; http://www.practicalmachinist.com/vb/general-archive/ot-theory-self-holding-taper-79627/

for a taper to be self holding, the taper angle with respect to the centerline has to have its tangent less than the static friction coefficient for the two materials involved. That's how you can tell the break point between holding and non-holding tapers.

Interestingly, I guess this applies to wedges also. When will a job spit out a wedge? Answer, when the coefficient of friction between the wedge and the material is less than the tangent of the wedge angle.

 

[pretend engine]

Thanks, Baluncore. I need to digest all the information you've given me here. I really appreciate your interest.

We are a group of senior mechanical engineering students, and this is our first real engineering project. We are designing a knee with specific attributes to conform to the athletic demands of a specific patient.

Split nuts and collets were never mentioned in school, so I need to do some studying.

 

[pretend engine]

Firstly, a gear rack (expensive) could be used with a short length of the same size rack as the “grip” rather than a pinion. See http://en.wikipedia.org/wiki/Rack_and_pinion The short rack section could be moved towards or away from the rack to lock or release.

Secondly, a threaded rod can be gripped with half nuts. Rotation is not needed. This is a cheap and available equivalent to the rack and rack. http://en.wikipedia.org/wiki/Split_nut

Thanks for all the info. The taper discussion was very helpful and interesting.

I want to see if I'm understanding your ideas correctly, so here's another quick sketch. Hopefully it's decipherable. The top is the same collet sketch. bottom left: rack and rack, bottom right: split nut.

http://img543.imageshack.us/img543/4720/w81a.jpg

The black things in the collet sketch are an idea for how to lock the collet in either the clamped or unclamped position, based on the rotation of the hinge to the right, which corresponds to where in the gait the user is. The hinge might have a cam that pushes stops in place above or below the collet. Perhaps the hinge cam could also break the collet loose when necessary. That's not really important right now though, my main goal is to work out how to stop the rod.

From my view, here are the pros and cons of each method.

Collet:

Pros: Simple, durable, less expensive
Cons: possibly difficult to select the correct taper angle

Rack and Rack:

Pros: Fairly straightforward, easy to implement
Cons: Involves meshing of gears which could break, expensive

Split Nut:

Pros: Inexpensive, less math involved when compared to the collet
Cons: Seems vulnerable to wear

If you can, please let me know if I've correctly understood your ideas. Also let me know if you can think of other pros and cons for each method.

I look forward to showing you pictures of a real knee in 5 months! Thank you.

 

[Baluncore]

Yes, you seem to have the right ideas there. The holding springs in those positions may be a problem though.

Because the gear rack will have a flat faced tooth, (probably at about 20°), a more positive lock will be required to prevent the axial load forcing the rack segment radially out against the spring. I would consider using an over-centre latch, a rotating cam or a wedge moved axially to hold the segment against the rack.

Likewise, the half nuts on the screw thread have a thread included angle of about 60°, so it will also be pushed out. But in this case, the opposite sides of the same half of the nut will do a partial job of opposing the forces since they almost face each other. It will definitely need a positive lock though, as it will be modified later.
Because the half nuts are 180° parts, they will have to be pulled further apart to clear the screw thread than is needed for a rack segment. It is normal to remove the thread where the half nuts meet at the split so as to gain the clearance needed for only a short movement.

I would rewrite some of your Pros: and Cons:

Collet:
Pros: Infinitely adjustable. Off the shelf in many sizes. Uses a less expensive rod.
Cons: Lubrication, (body fluids or dust), could disable it. Must use correct taper angle. May wear since slip is possible.

Rack and Rack:
Pros: Gear teeth have trapezoidal teeth with flat tips so they will not wear. Uses same rack for both parts.
Cons: Discrete steps. Involves meshing of racks which may require movement to align before mesh. More accurate segment alignment control is required. Racks are more expensive. Asymmetric so structurally unbalanced hence may bend.

Split Nut:
Pros: Inexpensive. Will align and lock more easily than a gear rack because tips are tighter radius. Will probably not wear if the screw does not turn.
Cons: Discrete steps, but not as bad as a rack. Needs more radial movement to engage, unless the nut thread profile is reduced where the half nuts meet.

You can experiment with half nuts by getting a thread and nut. Then cut the nut in half but slightly off-centre and file the larger part to exactly half. Experiment with the prototype nut on the thread before you begin to modify the thread in the half nut.
As you remove the thread from the nut, (where you see the thread profile in section), you will find it takes less movement to lock.

 

[kevindeschamp]

vehicle engines stop pistons all the time. I vote for crank, connecting rod, piston

 

[pretend engine]

But the piston is part of a vehicle engine.. how would we stop the crank?

 

[Baluncore]

I need to stop a piston moving within a cylinder at any point in its motion. (this will happen when the patient puts weight on his foot)

vehicle engines stop pistons all the time. I vote for crank, connecting rod, piston

The piston must be able to stop at any arbitrary point in it's movement, with very little warning. There is no crank, let alone one that has both instantly variable throw and prophetic phase.