compression %

[subsearobot]

Hi there.
I am designing a spring for a hand-held application. All spring action will be driven by hand, so speeds, impacts will be low.
Since it's a hand operated tool, I need a spring to be as compact as possible.

Let me preface this by saying that I am at a dangerous place: I know enough to be dangerous, but haven't designed enough springs to have either good intuition, nor great knowledge of the driving calculations. I am using Advanced Spring Design.

The question:
What are the risks in designing the spring to be at, say, 95% compression for the largest load/displacement? I am not all that concerned with linearity; think of this a tactile limit stop for the user. They will feel the spring, but doubtfully will register the non-linear rate over the travel of about 3/16".
According to ASD, the fatigue prediction is comfortably >10E7, our goal is less than 1/2 million cycles.

material is 17-7 h900

thanks for your thoughts

 

[israelkk]

"I need a spring to be as compact as possible" As far as I know ASD you will hardly be able to find the most compact spring because it doesn't have this feature. You will probably need to use trial and error but never will be sure that you have found the most compact spring.

If you want an intelligent answer to your question More info is needed, including spring load, deflection, space allowed (Compressed length, OD, ID), etc.

Material designation should have CH900 heat treatment not H900. The wire comes in the "C" condition, after heat treatment it becomes CH900 condition.

Theoretically, you can compress the spring to solid as long as it correctly designed. However, a check point for the force should be at 90% to assure some accuracy check from QA perspective. Beyond that point the force may increase exponentially with a difficult to control. From your question I assume you have trouble to find a spring that at 90% deflection will give the desired force.

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

 

[subsearobot]

Thanks Israelkk,
Yes, I am iterating. Since I design about 4 springs every couple years, this is not an unacceptable solution (time-wise).

I am trying to become a bit more educated so I know the right questions to ask, and can understand, my vendor.

good catch, I am in the habit of thinking I am working with machined 17-4. CH for 17-7

good point regarding the 90%. I was expecting, but did not see, non-linearity when I ran a test with an off-the-shelf spring having similar dimensions to my design (same OD, wire size, similar length, forces. but music wire instead of 17-7). would 17-7 be expected to perform similarly (that's my guess)

finally, I am conflicted about listing my data; if someone gives me a solution, I am less likely to learn stuff. but, as a beginning-level-expertise-spring guy, I think I can benefit from anything that a more experienced engineer may say. so, if you have any insight into whether I can get to a shorter L2, please comment.

Here is where I have arrived via iteration:
17-7 ch900 (require good corrosion resistance, plating music wire is not adequate)
wire: .025"
load1: 1.5#
length1: .41" <<<--- L1= L2+.17"
Load2: 1.88#
Length2: .24 <<<--- trying to minimize
OD: .375"
no surge, thumb actuated (no resonance concern)
ends closed, ground
ends fixed & constrained
require ~100% to test (probably 10-15 units, but that is not yet decided) to >500,000 cycles

thanks for your thoughts!

 

[israelkk]

When you say L2 shorter do you allow L1 shorter too?
Why the ends are fixed & constrained? Are they allowed to rotate?
Is the 0.375" fixed or can be smaller/larger?

When you dictate a force at L1 and q force at L2 with 0.17 travel you dictate the spring rate. Therefore, to get lower L2 only you have to allow changes in the OD and wire diameter.

Quick check shows that you can make a spring with same wire diameter, OD, shorter L1 and shorter L2 while maintaining the forces and travel and keep life >1,000,000.

 

[subsearobot]

Ok. my assumption regarding fixed& constrained was incorrect.

the ends are in short countersunk pockets, which are laterally constrained (similar to a pen). No rotational constraint. (I assumed fixed meant "fixed to the axis". not FBD the definition of fixed). It's amazing the assumptions we can make without even realizing it. my command of terminology is getting rusty!

yes, reducing L2 would reduce L1. We need a .17" delta between L2& L1. 3/8" OD is not (very) negotiable. I could go with a smaller OD, thought that's not the direction I want to go.

when you say I "can make a spring with same wire diameter, OD, shorter L1 and shorter L2 while maintaining the forces and travel and keep life >1,000,000", do you mean because my current L2 is at 95% instead of 100%? I had set an arbitrary design constraint at 95%; looks like I can take L2 to .20" and have a 98% compression. that does not leave me much wiggle room though. does that seem like a bad idea to you?

my real question is: "is it a bad idea to design it to 95% or 98% even? What am I overlooking?"

thanks again
s

 

[subsearobot]

attached is a plot of an off the shelf mcmaster spring that I grabbed as a reality check. the nonlinearity of this spring (similar to mine) is fine for our application. can I expect similar results from my current design, or is there a good chance of very different non-linearity after 85%?

 

[israelkk]

I have done some more calculations and here are my conclusions. The free length of the spring is ~1.1" therefore, to reach L2=0.224" you actually deflect the spring 0.876". Therefore, to solid there should be extra travel of

(0.876/0.9)-0.876=0.0973"

This dictates that the solid length is 0.224"-0.0973"=0.127"
For a wire diameter of 0.025" the number of total coils is 0.127"/0.025"=5.08 coils giving 3.08 active coils.

I do not believe you can have a spring that can meet the force and deflection not to mention to last more than 500,000 without some changes.

I don't know how many springs are you going to manufacture, however, to be able to do QA on the springs you can not ask for load accuracy beyond 90% deflection. Even for the 90% deflection it is commercially sound to ask for +/-10% tolerance on the force. You asks for two points of force and for both you need to give +/-10% tolerance.

I believe you should define only one load point (1.88 lb at L2), the spring geometry (wire material, wire diameter, OD and number of coils) will dictate the force at L1.

To get a sound design spring (shorter and with longer life than 500,000), to my best expertise you will need each spring to go a "preset/set-remove" secondary operation as the "final" production process. The spring should exhibit all the desired geometry and force/deflection properties "after" the preset operation. For the preset operation to be stable, each spring must be compressed to solid 12 times. This probably will make the spring price higher.

If you can accept OD lower than 0.375" you can find many more springs even with L1 smaller than 0.41" and L2 smaller than 0.224" but, all springs will have to go the preset operation.

I suspect that in the design process of your "hand-held application" the design of the spring left to the end of the design just to find that the space left for the spring is too small. If this is the case, you are not alone. During my over 40 years of experience I encountered endless such cases. I have no doubt that if the spring design was an integral part of the design process iterations, you were in a better position.

 

[subsearobot]

Good info, Isrealkk. I appreciate the your time! I will take some time to process/ understand your points.

the original design had a much more "relaxed" spring design (it operated below 85% compression). It also had an allowance for a different spring with a longer travel motion. once all the pieces came together, the unit was longer than desired, so we are circling back to remove any possible length.