Material & Concept thoughts on this simple spring Mechanism

[see3p0]

Heres my problem.

I need to design a small spring type device that has minimal moving parts for a cleanroom environment.

____________________________

A sketch of the spring lever mechanism

http://files.engineering.com/getfil...a6ed-46c1-b2fa-0a7dc9ffa946&file=untitled.bmp

Part one is the lever arm
Part two is the spring material that is fixed to the lever arm
-The general shape of the stainless steel spring

http://files.engineering.com/getfile.aspx?folder=a3251a41-b37b-4c5e-9c30-2f7a5f4ec572&file=B_1.jpg

parts three is a cover piece
part four is a main body that the latch will contact with
______________________________________



The latch releases when a small downward force is applied to the top of the lever (part 1) as shown in my fairly rough sketch.

A force of 15 - 50 grams would suffice. I dont need a lot of force in the spring to keep the clip in the closed position

I would like to use a spring material (part 2) that will transfer this downward force and release the latch, part one will be of a ridgid shape and would ideally not flex when the downward force is applied.

The spring material (part two) would do all the flexing.

I was looking at a stainless spring steel something like, Stainless Type 302, ASTM A313.

Im not getting great results with the force to deflection of the spring (part 2).

If I use a small sized spring part it fails long before i am getting my desired results.

So my two questions are..

Is there a more suitable spring material that would deflect under small loads and not fail like the spring steel.

or maybe the shape of the mechanism i am trying to design is totally wrong.

__________________

note: this mechanism must be made from metal and kept as simple as possible. I have ruled out hinges and small individual springs as they are not ideal for the clean environment that this mechanism will be required to operate in.

Any input is greatly appreciated

 

[TheTick]

If a single leaf is not strong enough (or too brittle when it's thick enough), try multiple thin leaves, like the leaf spring on a car.

If the stainless doesn't work for you, perhaps something else? If yoou need antimicrobial properties, you can plate with silver.

 

[asimpson]

I assume clean room rated latches wont do:

http://www.terrauniversal.com/desiccators-dry-boxes/latches.php

How and where is spring failing?

Getting flat springs to work is difficult. Flexing can be concentrated in one area which takes all the punishment and fails. The trick is to distribute the flexing over as much of the spring as possible. Can you get spring to roll over radius.

Take a look at these for ideas;

http://www.google.ie/images?hl=en&s...77&q=flexure+devices&gbv=2&aq=f&aqi=&aql=&oq=

http://www.google.ie/imgres?imgurl=...page=1&ndsp=33&ved=1t:429,r:8,s:0&tx=54&ty=51

 

[Eltron]

If The Tick's suggestion of lamination won't work, you could always try to make it a bit thicker (within your design constraints of course). Is there some reason this can't be plastic?

Dan

http://www.eltronresearch.com

Dan's Blog

 

[ornerynorsk]

Ditto what Eltron posted. Go down to your local DIY and look at the childproof cabinet and drawer latches. No sense in re-inventing the wheel.

If for some reason a stock latch won't fit your configuration, one piece plastic latch with a strike plate of stainless or plastic should do the trick.

 

[MintJulep]

Is "down" at the bottom of your sketch?

If yes, why bother with a spring at all? Just rearrange the geometry a bit and let gravity do its thing.

 

[unclesyd]

I understand the need for simplicity in a clean room environment.

If you have to go with a flat spring and as stated above it is
tricky to do. I would look at 17/7 PH SS in the annealed
condition and heat treat after forming. Again as posted above
use as large a radius as possible.
You don't have to make spring to your exact configuration as
you can clamp a candidate material in vise and see the force
required to bend. The exposed length should be as long as the
distance from the center of the bend to the end. Not absolute
but in the ballpark.

http://www.brownmetals.com/home.asp#

If you have any quantity you might want to look at someone like Four Slide.

http://www.fourslide.com/flat-springs.htm

 

[tr1ntx]

Maybe an "over-center" or "cam lock" type device that uses the material flexibility as the elastic movement and holding force? A common one would be the latch on a toolbox (a metal one).

Regarding your mechanism, I've got a memory rattling around about a flat spring mechanism I've seen in everyday use. I think it might have been on luggage or an attache case or something similar, back before plastics took over and metal was still used. Raincoats maybe . . . you had a flap that folded over and it had a spring action tab on it. You slid that tab under a "loop" (it was flattened, a rectangular profile) and the tab sprung up and its back shoulder held against the loop. To release, you pressed down on the tab and pushed it back under the loop. I remember these on something (or things) from when I was a kid.

 

[TVP]

Specific to the materials question, I agree with unclesyd, look into 17/7 PH instead of 302. It is called Type 630 in ASTM A 313 (wire) and A 666 (flat, strip spring materials). Sandvik and others make this. Somers Thin Strip (part of Olin Brass) also has extremely high strength versions of Type 301 stainless, which is the strip steel version of Type 302.

 

[zekeman]

Why don't you post the design thicknesses you have used in your design.

Based on your sketch,I roughed out the stresses using a force of 50 grams I got a bending stress of only about 33,000 psi for a .005 inch thickness times .57" wide spring strip. But .005 is probably impractical and making it thicker will increase the stress proportional, but may give to small a width for your purpose. For example doubling the thickness to .010: will result in a width of .05"( varies inversely as the thickness cubed)

Analysis:

I estimate that the moment is almost constant over the spring.
F= 50gms=0.1 lb
L=0.8 inches
M=0.08 lb in


From sketch the deflection angle looks like
20 degrees=Pi/9
the deflection formula for constant M
deflection angle=M*r*@/EI
r= radius of spring estimate(0.5")
@=Pi/2 this is the spring working length angle (this estimate may be slightly high)
So, evaluating
Pi/9=.1*.8*.5*Pi/2/EI
Solving for EI
EI=.1*.8*.5*9/2=0.18
I=0.18/E
For a 0.005 thick strip, the strip width is
w=12I/t^3=0.57"
c=t/2=.0025
the stress is
Mc/I=.08*.0025/I=.08*.01*E/.18=33,000 psi

Now, the .005" thickness may be too thin for structural reasons; if so, you could increase the active length of spring by using 3/4 of a clock spring and improve the rigidity.

 

[plasgears]

Give it a KISS and combine everything into one part made of Nylon, e.g. Prototyping can be modest in cost and flexible for modifications. Consultation with a mfr can yield some ideas.

 

[desertfox]

Hi see3p0

I have uploaded a file detailing how to calculate the deflection of bent or shaped springs using a Strain Energy method.
If your going down the road of doing your own design then I would determine the deflection needed for the spring first and then worry about force and spring stress.
There are other methods to calculate deflection of bent or shaped springs that you can use, my only advice here is not to approximate it from linear beam theory deflection as it will lead to errors.
I have based my calculation on your sketch ie a straight beam with a quadrant at one end and used the section that zekeman used for the "I" value, you will see that for a 50gram load it will deflect about 5mm but the section as just yielded reaching a value of 251N/mm^2.
If you need a large deflection you need a very low stiffness spring, ie a lot lower than the one I have calculated, one way to do this is increase the spring working length so that more material is available to absorb the energy.
If you provide further details of your spring we might be able to help further.

desertfox

 

[zekeman]

Desertfox,

Good analysis, but if you look at the OP sketch the moment "arm" is closer to 20mm vs your 7mm. This could explain that I get 10 times the flexibility and the 20 degrees of motion that I estimated from the sketch.

 

[desertfox]

Hi zekeman

Yes the straight portion is 7.3mm but I add the radius of 12.7mm
to make up the 20mm overall length, thanks for the compliment.

desertfox

 

[unclesyd]

The OP should give a force he can live with to operate the latch. The applied force needed will have to go up to get spring he can work with like attaching it to the block.

I have some 3/", 1/2", %/8" wide steel (Bainitic and some hardened) strapping that I've used for springs to hold detent balls and if I get time I will try to bend one to the configuration in the OP's sketch.

 

[see3p0]

Thank you all for the insightful responses.

Apologies for the delay in getting back to you

@TheTick (Mechanical);
Multiple leaves is a good idea but unfortunately wouldn’t be ideal in this cleanroom environment.

@asimpson (Mechanical);
No Clean room latches wouldn’t suffice but thank you for the suggestion.
The c-flex bearings are an interesting concept, hadn’t heard of them before.

“where was the spring failing” –
The spring is failing at the curved part. I am using a simplified FEA analysis, 0.25mm thick stainless steel and presuming that the force from above will approximate to a identical force at right angles to the bottom of the face that will make contact and move. The spring is 5mm thick and I have tried a few different thicknesses.


@Eltron (Mechanical) & @ornerynorsk (Industrial) & @plasgears (Mechanical)

I cannot use plastics due to the environment that they will operate in unfortunately

@MintJulep (Mechanical),

I have to hold the cover piece onto the main body, unfortunately gravity alone wont suffice, It would make everything a whole lot easier though!

@unclesyd (Materials)
Thank you for the design suggestions, material suggestions and links

@tr1ntx (Mechanical) & TVP (Materials) thank you

@zekeman (Mechanical) Thank you so much for the calculations and clock spring type idea to increase the flexibility. I wasn’t expecting anyone to do calculations, much appreciated.

I will look into this if I cannot get a simple bent leaf spring to work sufficiently.

@desertfox (Mechanical)
Thank you too for taking the time and effort to do those calculations and scan and upload them. Did you use an online reference for the formulae or a what literature would you mind me asking.

It would be ideal to have some hand calculations to check if that the FEA is giving good results.

I would ideally like the spring to have 16 degrees of movement which is approximately 6mm’s in the horizontal. This could approximate to a 20mm length and 5mm radius of the vertical part of the spring.

I will definitely need to increase the length for the part that flexes. Giving it a longer lever to flex over. And yes should increase the radius of the spring too. Great suggestions, much appreciated.

@ unclesyd (Materials) Thank you too.


I was also considering that the spring may have a greater freedom to flex If I do not fix the whole length of the horixontal face but give it maybe 15mms of contact and fix it behind this.

Looking at the few and the calculations that have been completed here, Im not so confident I will get a 6mm horizontal movement with my current design..

 

[unclesyd]

I think everyone appreciates the recognition and feedback.

When you get the problem it resolved if possible please let the Forum know the results.

 

[desertfox]

Hi See3Po,
Any decent Strength of Materials text book should have a chapter which covers curved or shaped beams. The method I used is known as the strain energy method. Another theory would be Castigliano's, again this can be found in any decent text book. However, if you post dimensions of your spring and where it is restrained we can perhaps run a hand calc to assist you.
Desertfox

 

[zekeman]

"The spring is failing at the curved part. I am using a simplified FEA analysis, 0.25mm thick stainless steel and presuming that the force from above will approximate to a identical force at right angles to the bottom of the face that will make contact and move. The spring is 5mm thick and I have tried a few different thicknesses."

see3p0,

I'm a bit confused.

Do you mean the spring is .25mm times 5mm wide? And when you say you have tried other thicknesses, do you mean widths?

If you say the .25mm x 5mm spring is failing, I would look for a metallurgy issue, since by my calculations the stress for 20 degrees of rotation is well under the infinite life cycle values.

(1) Pi/9/(.5*pi*r=0.44=M/EI
basically,from my earlier post
And the stress is
(2) sigma=Mc/I
c=.5mm/2=.125mm=.005"
Dividing both equations
sigma=E*.44*c=66,000 psi
which is a very good working stress for stainless.
I got 33,000 psi for the .005"thickness I used in my earlier post.

Stainless 302 data I have is 130,000 psi working stress for flat spring material. So your failures may be due to manufacturing technique.



From sheet data on stainless 302 I found the allowed tensile stress is 130,000 psi

 

[gerhardl]

I am weak at mechanical calculations, but I have got the feeling that if you could alter and increase the shape of the spring form and radius in such a way that the spring in stead of a 'rounded approximate 90 deg angle' have a form that includes the first (rounded)third of an S-curve, a large radius swinging a bit up, before turning a large (but somewhat smaller, but fairly larger than original) circle radius down to reqired straight strech, this could perhaps alter your parameters and avoid the stresses at sharper angeled curves?