Flat spring takes on permanent set

[gareth71]

This is a legacy design made from galvanised spring steel, they were always made to a material spec of BS5216, minimum tensile strength 2500 N/mm²

The supplier is now unable to get this material and we've had samples in different new materials but they all take on a compression set, see the attached pdf for how the spring fits into the assembly.

Any thoughts on a better material to use which will still close the shutter, but does not get overstressed in this application? I have tried a similar design as a torsion spring but the thin (0.25mm) wire diameter means that long arm buckles before bending and a thicker (0.35mm) wire just takes on a permanent set again. It's possible that an extra turn on the torsion spring's coil would help (the trial part had only one active coil) but there's still the problem of the end dropping out of the shutter's slot and jamming up.

I don't have much space to go with a redesign, as you can see the shutter goes almost all the way up giving little space for the spring.

Thanks for any help you can give, I've been staring at this for too long and I've got the Engineer's equivalent of writer's block!

 

[israelkk]

From my search, specifying BS5216 is not enough (I don't have this spec in hand). This spec is a general spec for different types of spring wire materials and drawn conditions, the wire definition should include some class or type in addition to BS5216.

Secondly, as far as I understand BS5216 is dealing with spring wires, I am not sure that it includes flat spring wire (strip). Can you post the exact wire data? A Certificate of Test (C.O.T) document of the specific wire will help. The spring manufacturer should have it, it usually comes with the wire when he orders the wire from the wire manufacturer or dealer. The C.O.T. will list the spec or the wire and the actual tensile test results of the wire the spring were made off.

This spring reminds the spring that was used in the 3.5" floppy disks.

Can you provide how much permanent set you encountered (in degrees) with reference to the 80 degrees in the drawing.

As I see it, this is a very large deflection case of a cantilever beam (neglecting the 2.75 radius) , what is the desired spring force at maximum deflection (when the shutter is up as seen in the lower picture)?

https://sites.google.com/site/israelkk

 

[gareth71]

I'll ask the material supplier for the datasheet, but I do know it was tested at 2500N/mm². The thickness comes in at 0.26 rather than the 0.25 in the spec, as its stiffness is proportional to the cube of the thickness that make about 12% difference on its own.

Stiffness at full deflection doesn't really matter, but I need about 1N at the preload position to ensure the shutter closes, that makes the angle 65 degrees.

Agreed that this is a very large deflection of a cantilever, my only amazement is that this thing ever worked at all!

 

[israelkk]

How much permanent set the spring faced? After loading and unloading what was the angle compared to the initial 80 degrees?

 

[gareth71]

The angle after fully compressing it goes down to about 55 or 60 degrees.

I can't limit the shutter travel unfortunately so it's really getting past its elastic limit.

 

[israelkk]

There may be a way to use same wire with same spring final dimensions. Do you have a large deflection analysis of the spring? Did you calculate the maximum bending stress when the spring fully loaded compared to the ultimate tensile strength of the spring wire? If you have the calculation results and the wire tensile strength there is a way to tell if the spring can be fixed.

 

[gareth71]

Unfortunately I don't have any software to do a large deflection analysis, as I said this is a legacy product and I've got no design notes on what was considered initially (about 12 years ago).

 

[israelkk]

Here is what I suggest from my experience and expert:

1. Ask the manufacturer to make 4 NEW springs for testing.
1st spring should have a 90 to 95 degrees instead of the 80 degrees.
2nd spring should have a 100 to 105 degrees instead of the 80 degrees.
3rd spring should have a 110 to 115 degrees instead of the 80 degrees.
2nd spring should have a 120 to 125 degrees instead of the 80 degrees.

In your spring there was no tolerance for the 80 degrees but it should have. This is why I used 5 degrees tolerance for each spring.

Those four springs must be NEW springs. It is FORBIDDEN to take the 80 degrees spring and "open" it to the suggested new degrees. This following suggested process demands that the desired degree (for example 100 to 105) must be manufactured such that during manufacture the bending operation must be toward decreasing the degree and never increasing it. Meaning, that bending the wire during manufacturing and/or during loading must be in the same direction as it is loaded during operation.

Now, install each spring and load it to the maximum deflection 12 times by moving the shutter and releasing the shutter 12 times. This must be done for each of the four springs. Take the springs out and check the degree of each spring after the permanent set. The spring that resulted (after permanent set) with the degree closes to 80 degrees but larger than 80 degrees is the correct spring you need.

Let us assume that the third spring that started with 110 to 115 degrees resulted at 80 to 85 degrees, then this the spring that you need to manufacture. Even if it will gone permanent set during operation it will always be above 80 degrees.

If you had calculation results, there may be an option to use less test springs. To my best experience this process will yield the desired results, however, since you didn't supply calculation results and wire tensile data, and due to the highly non linear nature of the spring, there may be a slim chance that the end result may be slightly less than 80 degrees after permanent set but, it should be much more than 55 to 60 degrees.

Good luck and let me know the results.

 

[TVP]

gareth,

What have the new suppliers provided in terms of chemical composition, wire processing, and final mechanical properties? Any chance that one of the suppliers was ITA from Italy (

http://www.steelgroup.com/the-group/ita-spa)?

I have worked with them previously on very high strength, small diameter wire, which is what you are looking at for UTS > 2500 MPa. I'm not sure if they make shaped wire (also called flat wire), but I would certainly inquire with them. Usually martensitic microstructures are more resistant to stress relaxation than pearlitic microstructures (BS 5216/EN 10270-1/ASTM A228/etc.), so perhaps the previously used material really wasn't in accordance with BS 5216. Any chance you have a sample that can be metallurgically evaluated?

 

[MikeHalloran]

Are you sure that the original parts were hot dip galvanized, and not electroplated with zinc,
or mechanically plated?

Mike Halloran
Pembroke Pines, FL, USA

 

[israelkk]

How do you make sure that the load point on the spring wire is always at the end of the 17.5 mm long arm and it is not moving to shorter length when the shutter moves up and loads the spring?

 

[gareth71]

israelkk - thanks for the idea about getting different angles made up, that sounds like a good plan to evaluate what's best as the mechanics is more complex than my simple beam calculations can handle. You're right that the load point moves as the shutter does; it travels from one end of the radius to the other, and finally when the shutter is completely open it presses down on the whole arm of the spring.

TVP, that wasn't our original supplier but it might be another avenue to explore about sourcing the base material, thank you.

Mike, I'm just going by the drawing note that says they were galvanized but all I have is a photocopied version of an old drawing, it's the kind of thing where there might have been a spec change in the past or even a concession in place that never expired. There's a lot more forensic investigation to Engineering than I was taught in university...