Need to increase efficiency during flywheel ramp-up

[RichLeimbach]

I'm currently using a small (35mm can) motor to ramp a 2" dia x .5" wide flywheel from 0 to 15K RPM in ~ .5s. Because of the inefficiency of the motor at the low-end range, I am exploring both mechanical and electrical improvements to the current design. I belive that you can link to the motor discussion here:

http://www.eng-tips.com/gviewthread.cfm/lev2/11/lev3/47/pid/237/qid/35300

I'm making this thread to see if anyone out there has a mechanical solution to the problem. What I'd like is a compact continuously varibale transmission design or a auto-shifting gear set-up. The main point is to increase efficiency in the low-end range of the ramp-up. I'm open to anything.
Any ideas?

 

[ve7brz]

Glad to see you got here Rich. I'll try to define the constant mesh transmission better and how it would work in a four or six speed implementation.

Talk at you later,

Bruce

 

[GregLocock]

Do you mean efficiency? If so then this is way out of my league.

However, if you just want a faster ramp-up for a given motor power could you use a stored air supply blowing on the rotor to get it going? Cheers

Greg Locock

 

[RichLeimbach]

Greg,
No dice. I am looking for efficiency. Also, packaging concerns keep me from being able to have ompressed air storage handy. Thanks for the suggestion.

 

[stressriser]

What about a centrifugal variable pulley? I believe I saw one advertised, we considered one on a design.

That way the motor starts out with high torque, the pulley changes diameter with speed, and transfers from high torque to high speed.

 

[RichLeimbach]

stressriser ... Thought about this, but never got it down to a reliable method in my head. Can you give me any info on the part you looked at (manufacturer, etc.). Any help is greatly appreciated.

 

[ve7brz]

Rich, streesrisers has a good idea. Take a look at the transmissions used in snowmobiles and auto-trans ATVs. If you could minaturize one of those you'd be a big step forward.

Bruce

 

[ve7brz]

Me again,

I decided I may as well describe the transmission to you if you're not familiar with them. Stressriser might be interested as well.

These tramsmissions are, by necessity, belt driven. Both driving and driven sheaves are split types. The driving sheave is held open by a spring and the driven is held closed by a spring. The driven pulley also acts as a clutch, being forced open wider than the belt.

Weights incorporated into the pulley force the sheave halves together at a given motor RPM. As the motor speed increases, the sheave continues to close further, walking the belt out toward the rim.

The driven pulley responds to the increasing tension on the belt by opening against the pressure of the integral spring. Thus a very wide range of ratios is provided over a comparatively small range of engine RPM.

A later development of these transmissions was load-sensing capability. In this development only one half of the driven sheave is connected to the output shaft. The other half of the sheave is free to rotate (to a limited extent) about the shaft. The two halves interact through a ramp and follower arrangement.

As tension on the drive belt increases there is a small amount of belt slippage on the 'fixed' half of the sheave, but since the 'free' half is not under similar tension it does not slip to the same degree, but instead rotates slightly relative to the 'fixed' half such that the two halves are forced closer together by the ramp and follower system. This forces the belt further out in the sheave which in turn forces the sheave of the driven pulley to open wider, and an increase in the (reduction) ratio results.

Hope that explains how they work. Where to find a miniature version is beyond me.

Bruce

 

[RichLeimbach]

Bruce,
The idea I was working on previously was close to the tension style you have suggested, but I couldn't figure out a way to get this to work correctly without a tensioner. I'll think a little more about the ramp and follower idea. maybe I could get this to fit cleanly inside the design.
I'd prefer to stay away from the swinging weights, as I think it could cause balance issues if placed on the motor and not dynamically balanced and I am concerned about bushing and brush wear if this happens.
Anyhow, thanks for the knowledge. Keep it coming. We'll get this thing solved before too long.

 

[Roach]

What would a variable frequency drive get you?
Does this compromise low end torque?

 

[ve7brz]

Hi Rich,

It seems as soon as I leave the computer another idea pops into my head. If you can measure the RPM of the motor (which is incredibly simple with a 3-phase motor), you can use a servo (as in model aircraft) to force the driving sheave halves to any position you desire.

This adds electronics (trivial) and the servo, while eliminating the swinging weights.

Bruce

 

[stressriser]

Rich,
Variable pitch pulley, start with this I guess. Should get you going

http://www.kartworld.com/TORQUE_CONVERTER_EXPLANATION.htm

 

[ve7brz]

I got so exited about the variable speed transmission I forgot what I came here for. Spiral spring. One revolution, controlled by escapement. Motor winds through gearbox between cycles.

Bruce

 

[RichLeimbach]

Bruce,
You and I think a lot alike. I checked into the spiral spring idea a couple of weeks ago. I need about 6 turns to get a good benefit in the system (from the spreadsheet), but with the torque that it would need to take up (~ 4 in-lb max), the spring ended up to be too wide to fit into my pinion (about .75&quot.
We already have electronics caplable of this in the system. I'll check into the servo idea. Think I could make it work.

stressriser (I like that name by the way),
Thanks for the info on the torque converters. I'll probably pick one of these up and see if I could shrink it into my system. It would be exactly what I'm looking for if I could get it to fit.

 

[RichLeimbach]

Roach,
By variable frequency drive, do you mean an AC motor with a special controller? Not sure exactly what you are referring to here.

 

[Roach]

I mean an AC inverter that converts an AC power feed to DC, and rectifies it back to a different frequency to decrease RPM of the motor. We use them all over our facility. We use them to ramp up ( control acceleration) of our motors.
They have all kinds of inputs that allow for different configurations.
Problem:

In doing this, the motor maintains a constant torque rating from 60hz (standard freq.) down (approx.),and loses torque above 60Hz. This may not be what you want considering you would have to size the motor large enough to supply a start-up torque for the flywheel, when you are really looking to size the motor for the torque required to maintain rotation of the flywheel at 15K RPM.

Just a thought.
Not the answer for efficiency I don't think, but I learned something in the process.

 

[dvd]

I recently saw a magnetic "shear pin" coupling from Rexnord. The principle this works under would allow you to get your motor rpm up quickly while transmitting a lower torque to your flywheel, which would allow it to catch up to the motor speed. The Rexnord website describes their product as follows:

"The Rex MagneLink coupling works on a
fundamental principle of nature involving
permanent magnets. The magnetic
coupling consists of two separate
components that have no physical
contact. Precision-machined, aluminum
rotors containing high-energy, permanent
magnets are mounted on one shaft, and a
hub conductor consisting of a steel
housing with copper ring mounts on the
other shaft. The coupling's ability to
transmit torque is created by the relative
motion between the copper conductor and
the magnets. This motion creates a
magnetic field in the copper that interacts
with the magnets, thus transmitting the
torque through the air."

They don't make a coupling as small as you need but you could possibly apply the same principles to a smaller coupling. You may end up with more inertia for the coupling than the flywheel, however. But that can be said about a number of the mechanical solutions.

 

[RX]

A couple of suggestions that may help reduce current draw of your motor during startup:

I used to race model cars in my younger days (now race real ones!). My experience was with electric motor cars which used small DC motors ('05' size) running on DC (7.2 V) batteries and controlled with a MOSFET electronic speed controller. Anyway, to reduce the wasted battery energy lost to heat the options were:
1) Electronically limit the current to the motor through the speed controller
2) 2 or even 3 speed automatically shifting geared transmissions (check model racing car suppliers - you may be able to adapt an existing suitably sized unit from a 1/8 th scale IC racing car)
3) Slipper clutch type arrangements - friction and/or fluid types were/are used on 1/10 electric racing cars.

With luck you may be able to find something off the shelf that will do the job, or at least can be further modified.

 

[RichLeimbach]

DVD ... Interesting idea. It would get rid of the heat losses that I would get with a slip clutch, but would have the same positive effect. I'll check into it. Thanks for the suggestion.

RX ... Current limiting doesn't help us in this instance because we are using an all inertial load (current limiting just increases the time in the low-RPM range. It does help our batteries however, so we may do a little of it in the end.

I'm already fitting a 2 speed RC car tranny to our system. My simulations show it should get me about a 30% performance increase. I'll post my results when I get the test run (today or tomorrow I think).

 

[Philrock]

Epilogics has a very ingenious (and somewhat complex) design for an infinitely variable transmission. Unlike most other designs, theirs is NOT friction-based. Here's a link to a very brief description of the device:

http://www.epilogics.com/corp/index.html

There isn't much info on the web site - you'll have to call them to get more.

pdr@baymachinedesign.com