Feasibility of component/spring retaining scheme? 1

[Bomble70]

Hi,

I'm looking for some advice/guidance as the attached problem is outside my experience as I'm a Manufacturing Engineer. Currently, I've been asked to assess the assembly of design whereby a component is retained within a casing by means of a retaining spring (comprised of 3 tabs). The parts are designed and supplied by a 3rd party and supplied to us for assembly into our design.

Based on the operating environment, the component needs a retaining force of ~44N under a minimum spring deflection of 0.3mm so we need an spring stiffness of ~49N/mm (44/3*0.3). However, due to the tolerances, the deflection can be 1.1mm so the retaining force is ~162N.

My concern isn't the force applied on the component under the min/max assembly gap but whether the deflection on the spring is feasible without breaking the tab (the spring is manufactured from beryllium copper with a yield stress of 190MPa). My instinct tells the current design isn't feasible but I don't have the experience to prove it!!.

I've attached a copy of the spring tab geometry but don't know how to determine the spring stiffness for the design and hence stress in the tab - I searched the forum for examples and found some threads that cover similar problems where the spring rate is determined using strain energy??

Hopefully, I defined the assembly issue but let me know if you need any further details.

Thanks for any help & advice
B70

 

[desertfox]

Hi Bomble70

I've designed springs like the one you have posted, now if you can give the radius of the humped section of the spring I might be able to look. I think your right though if the spring is compressed by 1.1mm it will be over stressed and not return to its original state.

one other thought why can't you ask the other company who designed it whether or not they considered the tolerance variation in relation to the spring

 

[Bomble70]

Hi Desertfox,

Thanks for the potential offer of help. I'm not in the office at the moment but I'm pretty sure the radius of the humped section is only R3mm. We got some samples from the supplier to support upcoming fit checks and have requested additional info on the spring properties (but are still awaiting info.) The rest of the assembly isn't avaiable yet to complete the trial/fit check

Being a typical engineer, I still like to know the theory and the logic behind the design.

B70

 

[tbuelna]

Attached are some basic design equations for linear wave springs from the Smalley Co. website. You may be able to get an approximation of the stresses in your spring using them. However, since your spring has very small dimensions, you'll need to carefully consider the impact of things like manufacturing and assembly tolerances. The spring rate and spring force at a given installed height can vary greatly with springs having such small size. It would be a good idea to have each spring compression set before being used. You may also need to measure the characteristics of each spring and sort them into matched sets.

Hope that helps.
Terry

 

[3DDave]

Too bad the cover isn't relieved to increase the travel for the spring and make it less sensitive to variations.

Why three little pieces? An answer that pops to mind on this is a simple disk with a number of radial cuts with the resulting tabs twisted a sufficient amount. I can't really help with theory on twisted tabs, but it is a lower spring rate version of certain lock-washers.

 

[Bomble70]

Hi Terry,

Thanks for the details and based on the part geometry and tensile modulus, I get a spring stiffness of 21.5N/mm and an operating stress of 846MPa, compared to a yield of 190MPa!! Whilst it's an approximation, it illustrates the unacceptable stress levels and the impact of small geometry changes have on the deflection and stress levels. That said, I did expect the spring stiffness to be higher.

Thanks
B70

 

[desertfox]

Hi Bomble70

Well I've calculated a spring stiffness of 48.62N/mm for a single spring but unfortunately any deflection over 0.05mm will result in the spring yielding. I'll double check my figures and post later.
It looks like the springs won't do the original 0.3mm deflection without yielding.

 

[tbuelna]

Bomble70-

3DDave's suggestion is excellent. What you need to do with your spring design is increase the working length. This should reduce the spring's peak stress levels and provide a more consistent force over its working stroke. If you can install the springs in grooves machined into the component top and/or cover bottom surface, you can increase the height of the springs and this will give you more freedom for selecting the profile and thickness of your linear wave spring.

 

[Bomble70]

Too bad the cover isn't relieved to increase the travel for the spring and make it less sensitive to variations.

Why three little pieces? An answer that pops to mind on this is a simple disk with a number of radial cuts with the resulting tabs twisted a sufficient amount. I can't really help with theory on twisted tabs, but it is a lower spring rate version of certain lock-washers.

Thanks for the suggestion but unfortunately, there isn't sufficient space to give a longer lever length, as both the height and diameter of the overall assembly are fixed. However, I'll query if the cover can rebated (I suspect not, giving it's wall thickness).

It's probably my sketch skills but the 3 radial tabs are connected/formed from a flat central outer ring. In addition to the 3 radial tabs, which are meant to provide the component retention force, the retaining spring also serves a couple of other assembly purposes;

- provides axial location of the component within the Can​

- there are 3 off long axial tabs (~40mm) that also come off the central outer ring to locate the component axially to the Can (I've no concerns with the axial tabs)​

- poka-yoke feature (i.e. if the component under the retaining spring is fitted upside down, than the cover can't be fitted)​

 

[desertfox]

Hi Bomble70

I checked my calculations and found an error, however I used strain energy theory and my final stiffness of the spring was 25.1N/mm, which is close to the answer you have.
Either way the spring is over stressed and the best way to alter the spring stiffness is to reduce the spring thickness from 0.3mm downwards, this then alters the stiffness following a cubed law.
Once the stiffness is reduced though, you'll need more deflection of the new spring to achieve the required load.

Another idea would be to mount compression springs vertically recessed into the cover and when the cover is tightened down the springs compress giving the desired load.
In addition the compression springs would be easier to produce and designing them much simpler.

desertfox