Spring failure reasons 3

[hamsuplo]

I have some returned springs that look like picture 3 and 4. There are many ways that the spring user could have misused it so that it turns into picture 3 or 4. But I have to assume that the spring user used it as meant to be used and that the failure is 100% due to the procedures before it gets to user (ie, metal producting, spring manufacturing, transportation)

We've been supplied by the same spring factory for the past decade, but only in the past year did we start receiving lots of returns of failed springs. So, obviously they have changed something in their manufacturing process which lowers the quality of their springs.

If anyone can give me some ideas why the spring would fail as it does in picture 3 or 4 (only one loop fails), I would appreciate it. Thanks in advance.

pictures of:
1. spring
2. how a spring is expected to deform under lateral deflection
3. what the spring I have looks like
4. another spring I have

 

[hydroman247]

Irregular diameter of the material throughout the spring?

 

[israelkk]

Standard extension springs are designed for pure extension. It is also has maximum allowed extension length to avoid yielding. Picture 4 can be explained by over extension during assembling or work that caused yielding. The spring assembling process must be designed such to avoid over loading of the spring. All other pictures show improper use or assembling process of a standard extension spring. If one wants to use a spring for lateral use the formulations used for extension spring are no longer valid and FEA is the best approach. For lateral use the spring ends has to fixed or in 90 degrees to the loading plane. Picture 2 seems to be an attempt to extend and load laterally at the same time which is a very unusual use of an extension spring.

Can you provide more information how the spring is designed to work?

 

[hamsuplo]

Our product is trampolines, the spring is not meant to be loaded laterally but that doesn't mean that it wouldn't be.

The springs are inspected before being shipped out, so the spring failure is not overloaded/yielded during the manufacturing stage, but during use.

But I must assume that the cause of failure is not misuse by user, but lower quality springs due to manufacturing process. I would like to find out any possible reasons for the decrease in quality of the springs.

Thanks for the answers so far; any more possible reasons as to why only one loop has yielded such as in picture 4?

 

[israelkk]

If the spring is properly installed and the ends are free to rotate with respect to the holding pins/axes there should not be any lateral/bending load. One possible cause can be if the ends are not aligned (there is an angle between the end). This will cause a parasite lateral (or bending load of the spring).

 

[drawoh]

hamsuplo,

Spring steel is (supposed to be) heat treated is it not?

Obvious cost saving!

--
JHG

 

[israelkk]

One more option is that the person who is jumping on the trampoline is landing with his shoe/feet on the springs instead of the fabric. If one lands on the middle of the spring length the actual number of active coils is decreased and spring is ended too causing lateral load and yield.

 

[hamsuplo]

drawoh:
The springs are heat treated, but for 3-4 minutes (344 deg C) or 4-5 minutes (317 deg C), I don't know whether this is sufficient for steel-70.

israelkk:
The deformation during use is irrelevent, I have to assume that they are used as they were designed to. I need to know possible reasons for why the springs produced in the last year are worse than they were before. Thanks for the tips though.

 

[Tmoose]

Maybe replacing a few springs is acceptable? at least for a while?

Any way to determine time the bad units were delivered, and possible if all the "bad" springs are in one or a few batches?

I'd start with -
- hardness test of material in deformed coil in 4
- hardness test and 30X magnification visual inspection of break and wire surface near break on 3
- non destructive Stretch Acceptance test of a percentage of springs in new shipments, and springs in stock, based on spring harvested from units with a few years successful field use

 

[hamsuplo]

Tmoose:

Not way to determine the time bad units were delivered, but most of the bad springs were produced the past year (returns started within the past year).

Your test recommendations are good ideas; can you explain to me the Stretch Acceptance test, what would be the allowable Stretch and how would I interpret the results to understand quality of spring?

Thanks.

 

[hydtools]

hamsuplo, what does your inspection process inspect? Only dimensions and not performance? You should be able to detect low quality springs at inspection for complete conformity to specification.

Ted

 

[hamsuplo]

hydrtools:
Dimensions (alignment of the hooks), and that they are not broken/deformed before shipping.
We also do small sample fatigue tests (100,000 cycles), and 50N tensile tests (the test results far exceed the requirement).

These test results don't explain why the springs deform as they do in pictures 3 or 4 though

 

[dgallup]

a) The kids jumping on the trampolines are getting heavier.
b) The springs are made from lower tensile strength material or slightly smaller diameter wire.

Spring rate is very sensitive to wire diameter. A change on .001" will make a big difference in the rate, if the spring is softer it will stretch farther under a given load. Once this exceeds the yield strength the spring is permanently deformed.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[Tmoose]

I just dreamed up the "acceptance" test. I was thinking Either you or the spring supplier should have an idea what the maximum load would be to reach the yield strength of the material, and what the maximum in-service load is expected to be, and how long the spring would be at that load.

Something along the lines of:
- Hang that load on a spring, measure the length and compare it to expectations, remove the weight and see if the spring returns to original length, maybe repeat 3 or 4 times to see if the spring yielded, but was still able to return to all coils touching like number 1.

- Then determine a load that yields the spring a certain amount, then apply that load to a few springs and record results and start an informational database.

 

[drawoh]

Tmoose,

I was thinking that testing springs was a good idea too.

You would have to take fatigue into account wouldn't you? Maybe load the thing several hundred times, somehow. It might just pass the first test.


--
JHG

 

[DesignerGuy16]

Shot peen the ID, will greatly relieve any stress concentrations. Didn't add a whole lot of cost when I've used it in the past.

Either that, or someone was jumping and landed foot first on a spring. They're designed to expand, not twist in the center like #3 in the picture you posted.

James Spisich
Design Engineer, CSWP

 

[1gibson]

Did the trampoline design recently add a safety net? This would increase the chances of jumping directly onto the springs.

Sorry this doesn't help you at all (you said you have to ignore the most likely causes of failure, improper use.)

 

[Compositepro]

Shot peening doesn't make any sense unless the springs were fracturing. Obviously, the the failures are due to yielding which is caused by exceeding the yield strength of the spring. You should have some requirement for minimum yield strength of your springs. Testing to make sure that the springs have more than this minimum yield strength is simple and non- destructive. Just hang an appropriate weight and the spring must return to its original length (no gap between turns).

 

[zekeman]

It is probable that the spring mfr has compromised the spring in some way, but you should have a better handle on the loads that they are subjected to and then call for a new design.
You can easily do this by taking the worst case loading, using probably 300 lb x n subjects jumping from some height to get the loading. Adding fatigue data life expectancy to the equation should result in a good design.
You probably had a marginal design to begin with and it doesn't take much to cause failure.
I think that the spring mfr should help you redesign the spring.

 

[tbuelna]

The comments from israelkk and zekeman are spot on.

Extension springs like this should only be subject to axial loads applied at the end loops, with all other kinematic DoF's unconstrained. If this were the case, you would not see the types of elastic failures shown in the images. Based on the images, it appears as though there was some excess combination of axial and radial loads applied that produced the elastic failure. Thus you have two options. One is to redesign the trampoline to shield the springs from any condition that would impart unwanted forces to the springs. The other would be to redesign the spring to accommodate any conceivable combination of loads that may occur during its lifetime.

Good luck to you.
Terry