|
West is one of the better epoxy technology companies.
Epoxy to aluminum: The strongest initial bond will result if the aluminum is roughed up, as the 80 grit mentioned. The bond strength will also be faster to degrade, due to the formation of aluminum oxide.
Epoxy to anodized aluminum will not be as strong initially, but the bond will be far more stable over time.
If temperature variations are involved, the epoxy expansion coefficient (TCE) must be in the realm of the metals. Stycast 2850 does this, tracon has some bipaks which are much cheaper, there are others, I have just used those two. To achieve the TCE close to aluminum requires the epoxy system have filler, typically of the form of aluminum oxide powder.
Drilling the hole caused two things..first, it increased the local temperature, which causes increased shear forces at the interface, this can start the joint failure. Second, as the bit cuts through the epoxy layer, vertical forces between the metals will also cause stresses. The combo caused the joint to fail by peeling forces.
Recommendations:
1. Select an epoxy with known adhesive strength to the metals.
2. Select an epoxy with the lowest TCE available.
3. Choose the lowest curing temp possible. this means, two part systems, and preferrably room temp cure. For electronics apps which require elevated temp operation, a cure and post cure must be at or above the operation temp maximum. The glass transition temp must also be above operation.
4. Mechanical grabs are always best. For the copper to aluminum plate example of the origional poster, I would tap the aluminum, put some 4-40 flat head screws in, drill mating holes in the copper, countersink them. Overdrill the countersink in the copper, so that when the 4-40 heads are flush to the non bonded side of the copper, there is clearance from the screw to the copper. That provides a mechanical grab/lock...pulling the two pieces apart puts some epoxy into compression, that is where the epoxies perform the best.
Oh, almost forgot. When you examine the failed bond surfaces, look for faint lines of failure, somewhat akin to tree rings. When a bond like this fails as a result of temperature cycling, the failure usually starts at a corner, and works it's way towards the center of the mating surfaces. Sometimes you can see the tree rings caused by the different cycles.
Cheers, John
|
