Thermal expansion Stress Problem

[wayt]

I need to bond a thick wall brass cylinder, ID 0.500", over an aluminum cylinder, OD 0.497" using red locktite. Brass cylinder wall thickness is 0.11". There are no external loads on the assembly. I have calculated the differential expansion of both over the operating temp range and found there is a max diff of +/- 0.1 mils. The rate of change in temperature is a maximum of 10C/minute. Will the stess built up in the assembly weaken the bond joint over time?
Is a press fit without locktite a better solution?

wayt

 

[desertfox]

Hi wayt

What is the final service temperature of the assembly the reason I ask is that you will have stresses generated by the intereference alone and then added stresses due to the temperature rise.
Further if your joint can carry the service load adequately without the locktite I wouldn't use it, in fact why do you want use an intereference fit with locktite anyway?
Without more detail of what your trying to achieve its difficult to advise further except that the locktite might be extruded out when you press fit the parts.

desertfox

 

[desertfox]

Hi wayt

Sorry I misread your post I thought you had an intereference fit prior to bonding.
However you might be better to just press fit them together it depends on what your goal is, also the final temperature to which assembly will be subjected needs to be considered.

desertfox

 

[wayt]

Thanks Desertfox for your initial comments.
The working temp is -20 to +70C. I am concerned about a press fit in production due to stress buildup and potential binding in the next element covering the brass cylinder. Also the added cost for tight tolerancing. The assemble has a very light load - another brass cylinder which slides over the first holding an optical element. The second brass cylinder must be free to move without any wobble over the temp range. The second cylinder is 1.5" long and moves laterally 0.5+".
wayt

 

[racookpe1978]

Fabricate it with an interference fit (2 to 4 mils will be enough since these are light loads tending to slide the two parts. You'll probably find you need a margin +0/-2 mils for the inner, and -0/+2 margin for the outer. Test a few parts to find the best/cheapest combination of wider tolerances, needed fit.)

Heat the outer (in hot water or oil) to about 130 to 150 degrees. Cool the inner (the smaller part) in ice water or by blowing it with CO2. Then slide the two together, lrt them reach room temperature. Your operating range of temps won't be large enough to let the two separate.

---

I understand it is the red locite that is "permanent." 8<)

 

[desertfox]

Hi wayt

Thanks for the response if you use the loctite then you need to know its thermal coefficient of expansion and Modulus of elasticity also the thickness of the adhesive layer and calculate the stress generated in the adhesive due to the expansion/contraction of the joint, or the easy route ask the adhesive manufacturer will it be okay for your specific application.
Now I know the temperature of the joint I'll calculate some stress values for purely the interference of say 2 mils and then the extra stresses due to the temperatue rise and post later.

desertfox

 

[Zesti]

The method described by racookpe1978 sounds like the way to go to me.

I would not put my trust in some sticky stuff to last over time. A fitting that is made by thermal expansion in the first place will last for ever.

 

[desertfox]

Hi wayt

If you have an intereference fit of a minimum 2 mils your hoop stress in the brass will be 110.5N/mm^2 which depending on your grade of brass seems okay to me.
The stresses due to temperature on an increase from say room temperature to 70 deg C ie a rise of 50 deg C are extremely small and not worth worrying about.
On the otherside ie cooling there isn't enough contraction for the joint to come loose however the force required to part the joint will be reduced but only by a very small amount.

desertfox