"Memory Wire", "Muscle Wire" aka Flexinol, Nitinol...

Yes I have seen a similar product before.  Some say it works well and some claim it does not.  I believe it is a bit costly.

It's something I would like to experiment with one day for sure.

Memory Wire

There was a series of

There was a series of articles in MRR in the early 90s (I think) where a guy was using memory wire for all sorts of animation. It was pretty cool. It was supposed to be a multi-part series but I think only a few articles ran. The guy who was doing it was in Italy.

I think it was this one, March 1993.

Good luck!

Chris

Series of articles...

That's right Chris.  I have the article, written by Jacques Le Plat, in front of me.  In his article he uses memory wire to activate a Bachmann Ho crossing gate.

Mr. Le Plat confims what I have read, that there are two types of memory wire.  One-way, and reversing.  He writes, "One-way wire returns to it's original shape when heated, and any different shape will be forgotten."  "Reversing wire oscillates between two distinct shapes based on temperature."

Mr. Le Plat describes a simple electrical control circuit, but goes on to say that more presise control will require  more complex electronics in conjuction with the wire.

Mr. Le Plat concludes by saying he believes the possibilities for animation with memory wire are nearly endless.  I wonder if he has surfaced with another article?

I never saw one

I can't recall a follow up even though a thing in the month before where his layout was featured said there would be a series of articles.

I am also thinking maybe it wasnt Italy, but maybe France? It was definitely a Western European country where the layout was located.

Chris

Switch motors

This link has an intriguing display showing possible applications. The one showing a bell crank shaped object that is then replaced by a Flexinol device should work well to activate switches. http://www.dynalloy.com/index.html#Scene_1 The cost appears to be just pennies per turnout if we can figure out how to build the thing.

In searching...

I can only find three articles authored by Mr. Le Plat.  One on his N-scale mirror of John Allen's Gorre & Daphetid,(SP?)

one on concealing uncoupler magnets, and the "Magic Wire" article.  I wonder how he ever made out with other animations using the wire?

I had another link to a turnout actuator that used memory wire, but the site has changed servers and now the article is no longer online it seems.  I copied and saved the text and photos from the site for my own information, but I can't very well put them online without permission.  Dang copywrite laws anyway.

Another instance I recall was a memory wire uncoupler concept...but similarly the link is dead now.

Once someone figures this out ...

And once someone figures it out, send us an article - and include a video clip of the operation.

Should be a doozie - a perfect example of the kind of "pushing the envelope" article we want MRH to be known for!

Switch motors...

I see the wire is priced at $3.05 per foot.

The wire is able to "move" 4% - 5% of it's length according to the text on the page you have linked.  The math will take me awhile but assuming a wire length of 1/2" - 3/4" how much movement would 5% of that length be?

The pictures I copied of the turnout actuator show what looks like two pieces of wire being used to actuate the turnout.  Overall I would guess the modeler used 8 inches of wire.

The people who have critisized the idea in the past have mentioned the length of the wire needed to get enough movment as one of the issues to overcome.

Perhaps by using some mechanical advantage in the design this can be solved?

4-5%

You can figure on 1/2" of change for each 10" of wire at 5%.  I guess I'd better take a look at my crossing gates (installed but unpowered) and see if I can determine how much "stroke" there is between up and down.

I guess you could wind the wire up into a loop to get enough length in a limited space... somehow.

Switch motors

As I understand it, some wire "grows" and some changes shape and then returns to the original shape as in:

4-5%...

Thanks Jeff, I should have looked at it that way it would have been easier than trying to figure out 5% of 1/2 an inch.  DOH!

I think using levers, cams, pivots etc. it may be possible to get the mechanical advantage to handle larger movements with smaller lengths of memory wire.  Up to an 8% change is possible but apparently the wire will fail quickly.  At 4-5% the wire is reported to last through many cycles.  In the case of turnouts the wire need only actuate the turnout in one direction as a return spring could be employed to handle the reciprocal movement.

Jacques Le Plat used a memory wire 3.5 inches long that he states produced a .120 inch movement when voltage was applied.  Connecting it to a pivot arm he achieved .160 of travel on the actuating rod connected to the crossing gate.  He located the actuating rods attachment point as close to the gates pivot point as he could to reduce the linear travel and ultimately the length of the memory wire.

Look at RRCIRKITS web site

Look at RRCIRKITS web site for an article on muscle wire used for uncoupler

Switch motors...

Hey great video!  Yes I agree we can move a set of points with the wire.

This is how the author of the MRR article describes the two types of wire.

"There are two types of memory wire.  One-way, and reversing.  One-way wire returns to it's original shape when heated, and any different shape will be forgotten."  "Reversing wire oscillates between two distinct shapes based on temperature."

I have no doubt that a bunch opf savvy model railroaders can come up with all kinds of ways to animate things with memory wire.

RRCircuits

Yes I was going to post the link earlier but when I went to check it out the links to the pages with pictures would not load.  They are working fine now.

RRCircuits uncoupler

Other Memory Wire Info

+

Ruff diagram of possible  muscle wire turnout motor.

(E) is the turnout.  (A) is 2 muscle wires.  (B) is the + voltage for left or right pull.  (C) is the return or - side.  (D) is a metal throw bar actuator ala Tortoise. It's lower part is pivot and the return negative side of circuit.  So current flow from B to A to D to C.  Muscle wire can ot be soldered or glued, It is looped upon itself and clipped. The clip could be dropped on to the throw bar actuator and be part of the circuit. Other end of muscle wire is lopped and clipped at B.

Another idea is to have the muscle wire simply looped around the throw bar actuator and adjust the assembly so the actuator is to one side when muscle wire is at rest and move to other side when current is applied.   You could probably run this from a stationnary decoder or  a small Digitrax TL1. Turn on a light function and you get the turnout to move. 

Now to order the experimentor kit. 

Marc

You could probably also do it

You could probably also do it with one wire, using a spring for the return.

I think you could still use DCC control, just use one wire to go +/-. The other side tied to ground. Then when the controller flips polarity current will run when it is opposite your "common" polarity. Hope I am making sense.

Chris

Switch motors

Excellent. Your idea about using one piece of wire next to the turnout that pulls the points one way when in its natural state and the other when current is applied seems to me to be the most simple and inexpensive. There is a video at the website given by the first link that shows a lock being activated by what appears to be a single wire that changes shape under current.

Many of the products transition from one state to the other at 70 degrees or so Centegrade which is like 158 degrees Fahrenheit. Although it's not enough to start a fire, we would need to have some type of enclosure to prevent burns if touched with bare skin. Maybe some kind of plastic box that's readily available and preferably free (pill bottle?).
 

Been there, done that, gave up.

Hi,

We (Fast Tracks) spent a considerable amount of time working on a memory wire switch machine a couple years back and came to the conclusion that there is a good reason there are none on the market.  It can't be done (reasonably).

The biggest issue is the wire will stretch over time, especially if it is over stressed, which is quite easy to do if not careful, and impossible to monitor.

The other show stopper is the amount of wire needed to get enough travel and power to move the points.  Since there is such a small amount of shrinkage in the wire, quite a bit is needed to work properly.  A switch machine would end up quite large.  Longer wire can not be simply wrapped around pulleys as it contracts along its entire length and will only squeeze onto the pulleys.  A suitable force has to also be used to retract the wire in the form of a spring.

The Dynalloy site talks of hundreds of applications, but really, there are none.  I spoke at length to one of the engineers there and found out they could not come up with many uses other than their butterflies, which they make more money selling then they do with the wire!

There was a small linear actuator they were offering at the time that I thought would be ideal for a switch machine, but that turned out to be nothing more than vapourware that was used to generate a massive amount of investment in the company when they did an IPO.  A few months later it all seemed to disappear.  Speaking with the engineer he said the actuator was never actually built and all the images of it were renderings.  The only time they made a product using the wire was for a doll where the eyes moved, and after the manufacturer purchased a few thousand units they dropped it in favour of an easier to use motorized version.

We were hoping to develop a miniature switch machine that could actually be mounted on the head ties of the turnout and could be triggered with a laser pointer, but quickly found out that was physically impossible.  We thought of  enclosing the machine in a long straw like tube that could be easily inserted into a hole in the layout, but again, that would not work.

I still think the laser pointer would work to trigger a switch machine though, no need for a control panel, just a sensor near the turnout that can be hit with the beam, but that's another project.

We eventually abandoned the switch machine project.

This isn't to say it can't be done, but after spending many months working on the project we realized the benefits of the wire did not outweigh all its limitations, and in the end a cheap switch machine could not be built with it.

Tim Warris

Fast Tracks

Letter to Dynalloy

I have sent the following email to Dynalloy and hopefully we will receive a quick response.

Gentlemen,
There is an interesting converstion going on at Model-Railroad-Hobbyist.com regarding your product.

The gist of it is the need for a cheap, simple motor to move the switch points on a model railroad switch. Currently, this is done using small electric motors, selonoids or compressed air. Flexinol appears to be an excellant product for this application however we are in need of guidance as to the best way to proceed.

There is considerable interest and if the final product is inexpensive and reliable, potential sales are enormous.

I would encourage you to go to the Model-Railroad-Hobbyist.com website and read the thread titled: Switch motors. I would appreciate it if you would respond on the forum so that others may benefit from your input.

Using the muscle wire with

Your right about the single wire and spring. Take it further and use the throw bar actuator wire as the spring. Just have it offset so  the turnout is 'thrown' under actuator tension. Muscle wire brings it back to 'close' or vice-versa.  But I think it is best to  have the strain on the muscle wire at bare minimum hence 2 muscle wires.

Using the muscle wire with long turnout point rails from the frog on,  that are not hinged, would be troublesome I assume.  I've read that even Tortoise motors can encounter issues with this type of turnout.  The wire will eventualy fatigue from the strain.  But  switches such as  Shinohara or Atlas that do not have any counter-force to deal  with should be less of a strain on the muscle wire. 

From what I have read, cooling would be affected by enclosing the wire,  so the time required to relax the muscle wire is going to go up.   it would make for a SLOW acting switch. Not a big deal. 

The idea of having it under a throw bar is a problem in waiting. Having the device accessible from above or below  the bench work is simpler in case of maintenance.

Again this is speculation on my part, not having played with it. 

Marc 

Wire strength

Hi,

We were developing a commercial product an therefore the machine had to be powerful enough for all types of turnouts, commercial, handlaid or built with our tools.  Any turnout, once in place and ballasted can become very stiff, so a good strong machine is necessary.  That is the biggest advantage of the Tortoise machines, by increasing the wire size to .032" or larger they get very powerful and can overcome just about any situation.  The biggest flaw with Tortoise machines is the wimpy .022" wire they include, it just isn't powerful enough.

With memory wire, to get more power you need the thicker wire.  The thicker the wire, the longer the piece needed for the travel, it starts to become a catch-22.  If you mechanically multiply the power so a short travel and good power translates into longer travel and less power you have to develop a pretty small mechanism and reliability starts to become an issue, and the cost of the final product goes up.

Developing a machine around memory wire for commercial use that is inexpensive would be very challenging, especially for the MR market, which is fairly small.  We struggled to keep the cost similar to a Tortoise, and it usually ended up more.

Really, the biggest issue we found, and the show stopper is failure of the wire after a few thousand cycles.  We just could not replicate the durability claims Dynalloy made about the wire, even after carefully calibrating the loads on the wire.  If the wire in the machine could not cycle 10,000 times it would not be suitable for a commercial product.

I suspect the response you will get back from Dynalloy to the inquiry about developing a switch machine will be to offer their engineering services.  When ever we spoke to them it was clear that was what the are marketing.  They will not develop a product themselves, at least they weren't interested 3 years ago when we spoke to them.  The MR industry would be far to small, any less than a million units would not likely be worth their time.  For a fee, they will do the engineering work, but I suspect once the project got underway details of the limitations of memory wire would start to show up.
 

Tim

Wire Strength

Hi Tim,

First let me say how great it is that you are sharing Fast Tracks experiences with memory wire here at MRH.

In your experiements were you limiting the change in the wire to 3-5% maximum ?  I have read that going beyond this maximum will cause the wire to fail quickly and stretching will undoubtably occur.

I have also read that another way to increase the strength of pull is to multiply the number of wires, keeping the diameter and lengths smaller but increasing the force produced with increased numbers of wires.  Is this something the folks at Fast Tracks looked at as well?

What you have reported so far would certainly provide a good reason why the wire is not being used extensively.  I thought turnout actuators would be a natural fit for the wire but I was also anticipating other uses for it on a layout where animation is desireable.  Perhaps the wire is more suited for applications requiring less precision than turnout motors?  Any thoughts on that?

Thanks again for joining our discussion!

Switch Motor

Tim,

As to your testing of this concept...did you guys also try using the  wire that returns to a previous shape as opposed to the type that merely stretches?

Mike

Not to get off topic, but...

I see Tony's Train exchange, (link in the sponsors area on the front page) has a 12 pack of Tortise machines for $159.95.  That's just over $13 each!  Pretty good price I thought.