Compression spring failure - breaking tips off sqrd & ground ends 4

[Ian63]

I have a compression spring made from range 2 wire and stressed to about 620 MPa shear stress. The ends are squared and ground.

These springs have been in use in our products for 3-4 years and we see no failures in the field at all. However when I put our product into a cyclic test program, running at about 15 cycles per minute, I get fragments breaking off the ends of the spring where it has been ground. This happens within 10,000 cycles.

Both ends are well contained in machined cups that are well aligned, and the spring is nearly completely supported from buckling sideways.

I am out of ideas to explain the failures (I’ve had 6 failures in the last week.)

Thanks for any ideas anyone can offer.

 

[sbozy25]

Well the first thing that comes to mind is that you are grinding the end such that you have thin material conditions on the very tip of the wire. The it is possible that your test apparatus rubs this thin end and in turn that causes the pieces to break off.

Is your test apparatus an exact replica of what the spring sits in when in the field?

Are you pulling springs out of the field and looking at them to make sure there are not any chipped ends?

More than likely you are not seeing this in the field is because.... well you just can't see it. I mean it is not like the part is going to fail if the end breaks up, as long as the parts breaking off are small, the spring will more than likely function just fine. Then the broken chips probably just fall out and dissapear.

My adivce would be to look at how thin you are grinding, I bet if you backed that off a hair you would have less of these failures. You also might try chamfering the ends, because it is possible the the grind step is causing a burr on the wire that is catching and breaking it off. A good chamfer should eliminate this from happening....

 

[israelkk]

One other option is that the griding process creates micro cracks on the spring ends if incorrectly done. Recall that spring material has approximately 50RC hardness therefore, it is susceptible to cracks during improper grinding. you can do a NDT test for cracks at the grounded surfaces. See MIL-STD-866C at

http://assist.daps.dla.mil/docimages/0005/02/86/866.PD9.

 

[TVP]

As a rule of thumb, the thickness of the ground tip should be at least 25% of the original wire diameter or 0.5 mm minimum. Any less, and the ends will be quite susceptible to breakage.

 

[Ian63]

Thankyou sbozy25, israelkk & TVP for your replies.

Sbozy25, I am cyclic testing the spring within the product as used out in the field. We service the products in house, and I have grilled the service dept to find the prevalence of broken springs. The answer is none.

isrealKK & TVP, the wire dia is 1.25mm and it is getting ground down to approx 0.6mm at the ends. I am suspicious of the grinding process setting up cracking.

The broken pieces are slivers of wire where the wire has cracked along its length for about 10mm (the spring is 12.5mm OD, so it's about 1/4 of a turn)

The broken pieces are the shape of fingernail clippings, and about 0.4mm of the original 1.25mm dia section.

I really appreciate the help on this one. The breakages are too frequent in testing (even though it's not seen in the field), for me to ignore without explaining the cause.

Ian.

 

[unclesyd]

Can you post some pictures?

 

[MikeHalloran]

Hmm.. "...well contained in machined cups...".

If the cups' ID is a close fit to the spring OD as installed, then the ends of the springs are stressed in bending in the plane of the spring seat as the springs are compressed and try to expand their OD. I think I've seen failures like you describe under similar circumstances.

I can't explain why the field fleet hasn't experienced the same failure ... unless they just haven't yet reached 10,000 cycles, yet.





Mike Halloran
Pembroke Pines, FL, USA

 

[JsTyLz]

Here's a few possibilities...

1. Decrease the frequency on the accelerated testing by half and see if you get the same failure. Generally the heat caused be accelerated testing should not make a difference, but the smaller the part, the less heat dissipation and.. blah blah blah you get the point.

2. If you have a metrology lab have the samples of the failed parts checked for grain size, chemical composition, etc.

3. As MH stated check the test setup to insure that you are not inducing a failure due to improper loading. As well as the stroke, the amount of force should also be simulated correctly.

It would seem to me that your test setup is under suspicion.

 

[TVP]

JsTyLz meant to say "metallurgy" or "metallurgical" not metrology. If the cracking is indeed along the axis of the wire, this is very suspicious. You definitely need to have some parts properly analyzed for the material characteristics as well as potential manufacturing (grinding) defects. Do you have access to a good metallurgical lab?

 

[JsTyLz]

Oops even engineers make mistakes :< (I think they are measuring it correctly.)

 

[sbozy25]

Are the springs peened?

 

[Ian63]

Guys,

Sorry for naive question, but what is the easiest way to get some pictures on this thread? Do I need to upload them to second site and link to it?

sbozy25, springs are not peened.

Mike, the cups at each end are 10mm ID with the spring OD of 9.25. Even under full compression I have about 0.5mm clearance in diameter.

N.B I've broken another 3 springs since starting this thread.


Ian.

 

[israelkk]

Did you try to look at the springs using at least X10 magnification. You may be surprise what you can see.

 

[CoryPad]

Yes, you use another server to host the images. Take a look at this:

faq559-1100

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Ian63]

OK,

I think I am having a buckling/fatigue failure as the spring can buckle sideways a few mm as it extends. I have revised my test setup by adding a centre dowel to support the end of the spring where the failures are happening & so far so good......


Here (I hope) is a pic of the broken bits.

 

[Ian63]

OK,

Now I have the hang of this images thing! The link to the photo is below

Ian.


IMG[

http://img182.imageshack.us/img182/2114/brokenspringbitsoo2.jpg

]

 

[sbozy25]

Hmmm.... Well those pictures do not really indicate the failure is due to the grind. I mean it is possible that the grind step initiated some cracks, but I doubt that was the only reason for the failure.

You say this product has been in service for 3-4 years now with no failures. Do you have test data from 3-4 years ago to see if they passed testing initially?

I can see a couple of potential problems:

1) Your spring supplier has changed wire supply companies and are now recieving inferior wire. It coudl be full of pitts that will cause fatigue cracks to form early.

2) Your spring supplier is not properly stress relieving the parts after coiling. If that is the case you can end up with brittle or soft parts. Both of which will not be good.

3) The spring was designed incorrectly to begin with and you just haven't seen the failures. You really should check parts that have been in service for a while and see if they have failed. When I say you, I mean you.... Don't take any one elses word for it, see it with your own 2 eyes. I am thinking the part is probably over-stressed to begin with and at the rate you are testing, it is causing fatigue to occur early.

Just my thoughts...