change in bearing fit due to change in temperature 2

[electricpete]

This is probably basic to you'all but I am just not able to get my brain around it at the moment.

Let's say bearing and shaft are of similar material (similar thermal expansion coefficient) and have 0.001" interference at room temperature. Does the interference change if I heat up both the shaft and bearing to assumed uniform temperature?

What about if shaft and housing have 0.001" clearance... same question.

Thx.


=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

Wow, as I said one of those days...

Last question should have been:
What about if bearing and housing have 0.001" clearance... same question.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[israelkk]

No, they will expand or shrink the same amount therefore the clearance or the fit will stay the same.

 

[electricpete]

Thanks israelkk. I guess I should have know that. It makes sense now.

Now I am trying to understand NTN "Ball and Roller Bearings" Publication 2202/E section 7.2 (Calculation of interference). It states:

"To prevent loosening of the inner ring on steel shafts
due to temperature increases (difference between
bearing temperature and ambient temperature) caused
by bearing rotation, an interference fit must be given.
The required amount of interference can be found by
formula (7.3).
DeltaDT= 0.0015 * d * DT (7.3)
where,
DeltaDT : Required effective interference (for temperature)
mm
DT : Difference between bearing temperature and
ambient temperature °C
d : Bearing bore diameter mm"

So, it seems there are two possibilities:
#1 - are they assuming a differential coefficient of expansion between shaft and bearing materials?
#2 - Or maybe they should have defined DT differently as difference between shaft and bearing shaft temperature? Just trying to make sense out of it.

If I'm remembering right, 0.0015 E-3 / C is the coefficient of expansion for most carbone steel... so seems more likely they are assuming #2 - this coefficient of expansion for both materials and assuming a temperature difference between shaft and bearing?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[diamondjim]

Dt assume that the bearing only heats up
and that the shaft is at ambient temperature.
The heat will eventually transfer to the
shaft but probably never get as high as the
bearing temperature.

 

[desertfox]

Hi electricpete

Diamondjim is correct on this one, it appears that they are
allowing for differental thermal expansion of the bearing on the shaft ie:- that you haven't got uniform heating.
Therefore you would have to know what the difference in temperature was to calculate the required intereference.
So if the temperature difference was say 4 degrees C
then the expansion of the housing bore d would be:-

DELTADT= .0015*10^-3 * d * 4

which is also in this case is the required minimum intereference.

regards desertfox

 

[unclesyd]

If thermal growth is a possible problem you can compensate for thermal growth both on the inner and outer bearing race with a Tolerance Ring.

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

 

[Tmoose]

A typical steady state condition for motorized spindle bearings the inner ring is thought to be about 10 degrees C warmer than the outer ring. After a full bore start On the way to steady state a high speed spindle can experience enough thermal preload increase in seconds to tear the bearings' heart out. During the autopsy tear down the inner ring will often have rotated on the shaft. Kind of makes sense, since the heat transfer to the shaft depends a lot on contact pressure, so the hotter it gets, the looser it gets, the hotter it gets. That situation is somewhat reversed for the outer ring, where the OD gets tighter in the bore, for better heat transfer. But even that is not much help short term, since the slow-to-warm housing simply constrains the poor outer race in the beginning. One way to reduce thermal preload is relieve the bore over one race several thousandths of an inch.

 

[strokersix]

This is likely irrelevant for practical purposes but the clearance or interference magnitude will scale in exact proportion to the component growth using the assumptions given.

 

[electricpete]

stroker - Yes I see what you mean.
Start with a clearance of 0.001"
increase clearance by a few hundredths of a percent.
Still 0.001"


=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[plasgears]

It will save much misery if you consider all thermal conditions including transients during startup and shut down. Giving adequate clearance will make the difference between success and failure. I can vouch from experience.