Belleville washers to keep preload under thermal variation 2

[franck]

Hi all,

I have an application for which I have to keep my preload over a wide range of temperature. Basically, I am preloading a glass-reinforced shim with some Ti6Al4V titanium fasteners. This assembly is going to experience a temperature variation of 160C (from ambient to –140C). The thickness of my shim is approximately 1 inch, its CTE equals approximately 40 ppm/C and its compression modulus is 34.5 GPa (titanium cte=8.6ppm and E=114GPa).

I was told that I could use Belleville washer to keep fairly constant my preload. I have all the equations regarding the sizing (washer stress and deflection) of those washers but I am missing the key point: how to make that washer a constant-spring force. Really I do not have a clue since obviously as the deflection of the washer changes, my preload is going to change and on top of that the temperature characteristics of the washer varies with temperature as well!

Is there some document outer there that could help me to design that constant force spring?

Best regards,

Franck

 

[israelkk]

You have to keep the h/t of the bellevile spring equall 1.41.

This gives a range of fairly constant force when the bellevile spring is loaded close to flat. These type of springs are not standard (to my best knowledge). You will have to make a custom one.

http://www.webspawner.com/users/israelkk/

 

[MintJulep]

You said it in the first sentence of the second paragraph. "Fairly constant". Thats pretty much as good as it gets.

BTW, have you considered that the modulus of the shim may change of this big tempertaure change?

 

[EdStainless]

You can use a stack of Belvs to keep the shim loaded, but the load will not be constant. Since your shim will expand so much more than the bolt/springs you will need to factor this in.
Do you have CTE data for the actual temperature range? Many alloys have very non-linear expansion properties.

= = = = = = = = = = = = = = = = = = = =
Corrosion never sleeps, but it can be managed.

http://www.trenttube.com/Trent/tech_form.htm

 

[israelkk]

MintJulep

How this modulus change will affect the "constant force" of the belleville spring? The belleville spring will give the "constant force" as long as the "constant force" range of the belleville spring is larger than the thickness changes of the shims.

 

[MintJulep]

israel kk,

You're right. It won't affect the springs within the limits you state. It may affect the behaviour of the joint as a whole.

 

[unclesyd]

Here are two sights that cover Belleville washers.
Either company is good and did have a great technical support and design information.
Key Bellevilles now has a CD for design.

http://www.keybellevilles.com/

http://www.asbg.com/spec/belleville.html

http://www.solonmfg.com/frames_belleville_springs.html

The solution to your problem might be to stack washers in series to keep a constant rate.


Another approach might be the Clover Spring Washer. I would take a serious look at this one.

http://www.cloverspring.com/index.htm

 

[israelkk]

unclesyd

A stack of washers in series doesn't give a constant force it creates a lower spring rate and therefore the changes in the force will be lower than using one spring. However, bellelle spring can be designed to give constant force of a range of spring deflection or even create a snap (bi-stable) spring choosing the right h/t ration.

 

[franck]

Yes, EdStainless, you are right the CTE is going to change over the temperature range. It starts at 45 ppm/C at ambient and reduces to 33 ppm/C at -140C. Fortunately, my shim material (G10 cryogenic grade) has been fully characterised at low temperature.

Thanks Israelkk! by looking at the plot, I understand how this constant spring can be designed by taking a ratio close to 1.4. Now, I have to think about the implementation!

Thanks all for your responses. Much appreciated!

 

[unclesyd]

israelkk
Take a look at the representations of various spring configurations at the following site.

http://www.keybellevilles.com/engine.htm

 

[israelkk]

unclesyd

I have looked at that web page and still don't see a constant force spring. Mechanical springs design is my main speciallty for 25 years. I had "a few" chances to design and build custom made belleville springs for contant force and snap through applications. If you will look at a good spring design book (such as SPEC's) you will see that a bellevile spring can be designed as "linear spring" where h/t<<1 or "constant force" spring where h/t=1.41 or as a "negative rate spring" where h/t > 1.41 or a "snap through (bi-stable)" spring where h/t>2.81 and nore.

I am familiar with all the companies mentioned in you first post, they are good companies but their catalog products usually do not deal with constant force springs because such spring can not be stacked.

I am also the developer of the "ultimate spring designer" expert system for mechanical springs design which is described in

http://www.webspawner.com/users/israelkk/

 

[unclesyd]

israelkk,
I concur and defer to your post.
I was taking a couple of applications where we use stacked Bellevilles to approximate a constant rate spring by limiting the stack deflection to under 60% of the height to get a linear deflection over a desired range to compensate for thermal expansion.

Anecdotal:
This system was designed by the old P&W group.

 

[drawoh]

franck,

You need to be approximately constant, right?

Work out how much your system will move. Select a spring system whose nominal compression is an order of magnitude greater than that movement. Bingo! Approximately constant force.

I do not know how much your system moves. A stack of belleville washers in series might do it.

Take a look at die springs. The blue painted, medium strength die springs are normally used for stripper plates on punch presses. If you want something compact for a given nominal force and travel, they are better than Belleville washers. Lots of manufacturers make them, so Google is your friend.

JHG