100k cycle spring failure 3

[subsearobot]

Hi!
I've got a compression spring that is failing at around 100k cycles. I am not all that experienced with spring design nuances...

I've tested 16 springs, of them, 3 have failed. One was broken in two places near the center when removed from the housing and inspected. the other 2 appeared bent (no longer cylindrical). when I tugged on them, they broke near the middle. felt like a brittle failure.

material is 17-7, heat treated.
Operating between about 40% and 80% compression
~1HZ

the (design) numbers:
load 1:2.95lb at 1.195" (spring length, not deflection)
load 2:6lbs at .88"
wire diam:.028"
coil OD:.180"

Is this design pushing a boundary that I am oblivious to?
Please comment!

 

[desertfox]

Hi

I assume this is a tension spring, can you post a picture of the failure and also the number of spring coils and end type

 

[tbuelna]

subsearobot,

At an operating frequency of 1Hz, I doubt you have a dynamic issue. Unfortunately, you did not provide sufficient information about your compression spring to determine whether it is overstressed for the number of load cycles it is being subjected to. It would also be helpful if you provided a detailed description of how the spring is constrained. For example, is there any contact at the coil OD/ID throughout its range of motion?

 

[subsearobot]

Hi, thanks for the questions.
It is a compression spring, ends ground. It is housed in a cylindrical housing, diam .197". the housing bore is held to a 16rms finish. the spring buckles in the housing. initial spring compression (.31") is controlled with a set screw. Travel is 8mm from its initial compression (about .31")
material is 17-7, HT to ch900 after coiling.

the design software gives the following:
26.8 total coils
free length 1.5
spring will buckle
spring rate 9.6 (I measured it at about 11)
solid height .75

the spring is impact loaded, metal plate on metal spring compressor. (it provides a tactile feedback to the operator of the equipment, then travels 8mm, where there is a hard stop to said metal plate.

What other information would help to determine if it is over stressed? The spring design software indicates a medium service spring, and our manufacturer said it should likely be good for up to a million cycles...

I am having trouble uploading a photo, I will as soon as i can.

thanks!

 

[desertfox]

Hi

Have a look at this site particularly clause 23.10 and 23.11, your spring is far to slender and that's why it's buckling.
The buckling increases the stress on one side of the spring and decreases it on the other

https://d2t1xqejof9utc.cloudfront.net/files/12615/CH-23.PDF?1338278308

 

[MintJulep]

the design software gives the following:
26.8 total coils
free length 1.5
spring will buckle
spring rate 9.6 (I measured it at about 11)
solid height .75

Buckling is bad.

 

[subsearobot]

is buckling bad if it is constrained inside a smooth bore?

I am replacing a previous design (though it's a little shorter); the previous design buckles, but i don't believe any failurse have happened...

this could be the simple explanation that I'm looking for!

 

[tbuelna]

is buckling bad if it is constrained inside a smooth bore?

Did you check to see if your spring is over-constrained by the bore at max compression?

There is only a nominal .017" diametral clearance between the bore surface and spring O.D. in a free state. The coil O.D. increases as the spring is compressed. Some of your spring O.D. dimensions may also be greater than .180" in a free state. This could result in interference between some coil O.D.'s and the bore, creating a binding condition.

As others noted, buckling is not good. With a small diameter compression spring having a large number of active coils, you might consider using a pin to guide the I.D. of the spring coils (if possible).

Good luck to you.
Terry

 

[MintJulep]

is buckling bad if it is constrained inside a smooth bore

If you do it 100,000 times, apparently so.

Also, 16 microinches isn't very smooth. Sort of like a file. [COLOR=][/color]

 

[desertfox]

Hi

It appears that the bore of the component might be to small for the spring outer diameter as noted by others above.
This article suggests the bore should be around 0.24" for your .18" diameter spring.

http://in.misumi-ec.com/contents/tech/press/58.html

I am a little surprised that your spring manufacturer hasn't pointed out about the buckling issue because in that first article it states that springs having a free length of more than four times the mean diameter of the spring is liable to buckling.

If the spring wasn't buckling then I believe that the stresses in the spring are not excessive, according to my calculation the shear stress at 6lb load would be in the order of 105,000 psi and the tensile for that wire is around the 300000 psi mark.
So the spring shear stress at full load is less than half the tensile strength of the material.

On the springs that have failed are there any signs of rubbing on the outer diameter if you look at them under an eye glass or better still, look down the bore of the spring holder and look for any marks.

I think overall the failure is down to the buckling and possible constraint of the spring coils in the bore of the holder, finally it might be worth talking to the spring manufacturer who should be able to advise.

 

[israelkk]

ROUND WIRE COMPRESSION SPRING INPUT DATA-TABLE 1
************************************************
< SQUARED & GROUND / OPEN OR PLAIN GROUND / OPEN OR PLAIN
/ SQUARED ONLY > (Spring ends) <1/2/3/4>..... 1
<SINGLE SPRING/NEST OF 2/NEST OF 3> <1/2/3>....... 1
MINIMUM OUTSIDE DIAMETER [inch].......................... .18
MAXIMUM OUTSIDE DIAMETER [inch].......................... .18
OUTSIDE DIAMETER STEP [inch]............................. .1
DIAMETRAL CLEARANCE BETWEEN NESTED SPRINGS [inch]........
MINIMUM WIRE DIAMETER [inch]............................. .028
MAXIMUM WIRE DIAMETER [inch]............................. .028
WIRE DIAMETER STEP [inch]................................ .001
MINIMUM NUMBER OF ACTIVE COILS .......................... 2
ACTIVE COILS STEP ....................................... .1
SPRING MAXIMUM WORKING LENGTH [inch]..................... 1.5
SPRING FORCE AT WORKING LENGTH [lb] ..................... 2.95
SPRING FORCE PERMISSIBLE DEVIATION [lb] .................
SPRING RATE [lb/inch].................................... 9.6
SPRING RATE DEVIATION [lb/inch].......................... .1
16. 17-7 PH AS COLD DRAWN AND AGED (CH)
MODULUS OF RIGIDITY [ksi]................................ 10999
SPRING DISPLACEMENT DURING WORK [inch]................... .31
SOLID LENGTH FACTOR = FACT (recommended 1.1 minimum ).... 1.08

NP DOUT DWIRE COILS LSOLID LWORK LFREE RATE
--------------------------------------------------------------------------
1 .180E+00 .280E-01 .254E+02 .767E+00 .114E+01 .145E+01 .950E+01

NP TAUA TAUSO FA FMAX CYCLE MASS
----------------------------------------------------------------
1 .133E+03 .146E+03 .590E+01 .648E+01 .100E+04 .229E-02

Above are the results of the calculations for your spring as "normal manufactured spring". As you can see the maximum life cycle is under 1000 CYCLEs. This result assumes a spring wire with minimum tensile strength as specified in a wire spec such as ASTM, AMS, etc. The reason your spring lived 100000 cycles is probably because the actual spring wire you have is stronger than the minimum allowed by ASTM spec. However, you can not trust future batches will be stronger too and how much stronger. You should ask you supplier how he came to the calculated 1000000 life cycles.

Can you supply more information on the spring manufacturing process?

 

[Tmoose]

"is buckling bad if it is constrained inside a smooth bore? "

My old dirt bike's fork springs get flats worn on the OD after just a decade or so.

http://czmadness.net/techdata/73Manual/73manual41.jpg

Note They are guided presumably by a small diameter about 8 inches long about 30% of the way down from the top.
They may actually be buckling and also rubbing against the larger fork tube ID as as I recall the won flat is nearly full length, and they make a humorous screen door or bed spring sound when rocking the bike against the front brake at idle.

 

[subsearobot]

I measured the compressed diameter, and it is very close to its original diameter.

Looking under the microscope, I do see some wear to the patina on one side of the spring. seems consistent to the spring rubbing on one side due to the buckling, and being guided by the housing.

 

[desertfox]

Hi

Well if you can see signs of wear that May also contribute to the failure by reducing the wire diameter of the spring by wearing a flat on it.

 

[subsearobot]

this seems to work in concert with another problem that we are having with this re-design. we are seeing galling between the rod and the housing where the rod exits. (rod connects the spring to the moving plate) both parts are stainless (303) and it's a very small contact area. the galling produces fine debris. We are working on a solution to that problem, and I suspected that the debris may contribute to the spring problem as well.

Once we test the improved rod design, will see how spring performance is.

thanks all!

 

[tbuelna]

...I measured the compressed diameter, and it is very close to its original diameter.

I did a quick calculation of your spring OD when compressed to .88" (per "load 2" condition in your OP) and I came up with an OD of .226". This is a diametral growth of .046" when the spring is compressed from a free length length of 1.5" to a final length of .88". It appears your guide bore of .197" diameter is a bit too small. As desertfox noted, the guide bore should have a diameter of .240" (or more) to prevent binding.

I'm not very good at math so maybe someone can check my numbers.

http://springipedia.com/compression-general-design.asp

 

[jlnsol]

Nobody is reacting on the 80% compression (!) during operation. In my view this a way too much. This spring needs to work heavily. 100k cycles lifetime is a very good result!