Truckedup wrote: ↑Sun Sep 08, 2019 10:17 am
One Factor is you don't know the actual aluminum alloy, does sand cast or die cast or forged or plate aluminum all expand the same? I believe it's a decision based on actual experience, some of which has been expressed here....
Since it's an engine case, one can safely assume it is not plate or forged aluminum. That leaves cast aluminum as the method of manufacture. Going one step further, it could be sand cast but that's not like that bike mfr to use. The case more than likely could be permanent mold IMO.
Knowing where the estimated press fit numbers end up after a simple calculation is very useful and is a fundamental engineering exercise. Since a lot of smart guys spent a lot of time measuring CTE, I have reasonable trust in the numbers. I've also done a little thermal/structural FE to doublecheck this sort of thing and it substantiates the basic thermal properties written in most books.
Matweb is a good source for these kinds of material data.
Up above, I listed the CTE of 356T6 aluminum to be 2.32E-5 inch/(inch•Rankine) or m/(m•K),
which is incorrect for BS units and correct for SI units.
That's my bad, that's a rookie error.
As before, the 1.600 diameter is an assumed number. The OP would need to use the actual dimensions for proper estimations.
CTE for cast 356T6 is
11.9E-6-12.9E-6 inch/inch•Rankine
CTE for A360 is 1
1.6E-6-12.7E-6 inch/inch•Rankine (typical engine case alloy with higher copper content)
CTE for A380T5 is
11.7E-6-12.6E-6 inch/inch•Rankine (typical engine case alloy with lower copper content - reduces white fuzz when exporting across the pond)
You could use a simple spreadsheet to calculate the values for the diameter change for each alloy and get the range of expansion. Average them and you have a good idea of how the case changes.
So, to correct my error above:
For the aluminum case using an average
CTE of 12.0E-6 inch/(inch•Rankine):
The expanded diameter of the bearing bore using a 300F final temp and 68F starting temp and assuming the initial bore is Ø1.600" is about
1.6045"
Adding another 50*F to the case heating will add about another .001" to the bore diameter.
The contracted diameter of the bearing, assuming the steel is the major component using a 68F starting point and -166F final temp with the OD the same as the bore: Ø1.600"→
1.599". Note that at -166F, the steel is well below its brittle transition temp and care is needed to prevent breakage if malletizing the bearing.
So, it looks like about
1.6045-1.599 = 0.0055" of space could be created without going exotic and using 300F and the average CTE of both metals (.0053" using minimum CTEs). The OP mentioned .0045" interference at room temp so it appears using regular methods is adequate. Looking at the hoop stress in the case from this press fit, it is pushing the elastic limits of the case but that can be reviewed in another thread.
If the actual case and bearing measurements after doing the math for heating and cooling still show a press fit is needed, I don't see much chance for success. It is also interesting to note that around 87% of the press fit is "absorbed" by the engine case due to the bore expanding. Squeezing the steel cylinder into the bore takes up the remaining press of 11% of the original .0045".
For the OP's original question on whether using liquid N2 is a viable method should be directed to a Cryo company such as Cryopro or 300 Below.
The heating (oven) and cooling (frozen CO2) I used with aircraft APU parts always resulted in the bearing dropping in and waiting for the housing to cool enough to properly retain the bearing. Ag parts require a bit less precision and copious amounts of heat due to large castings in order to create a successful assembly.