Coefficient of Friction

[AIREDALE]

Looking for a value for the coefficient of friction between a steel bushing (4130, heat treated to 125000 psi/Rockwell hardness of C26-C33, cadmium plated per SAE-AMS QQ-P-416) and magnesium casing (ZE41A-T5). Unable to find on-line or in available references. Request a close approximation and/or reference.

 

[rb1957]

what's the fit tolerance ? (close, loose, sliding, interference, ...)

 

[AIREDALE]

Interference

 

[rb1957]

try this ...

http://www.roymech.co.uk/Useful_Tables/Tribology/co_of_frict.htm

not sure it'll get what you want (which i'm guessing is the force to remove a bushing, and not to install it)

 

[tbuelna]

AIRDALE,

If you're doing an analysis, my experience is that it's common to use different mu values depending upon what you are analyzing for.

For example, if you want to show by analysis that your interference fit has sufficient prevailing friction at the metal/metal contact interface to prevent the bushing from moving, then a conservative mu value would be in order. In aerospace, a widely accepted mu value for clean metal/metal contacts that must transfer loads by friction is 0.15.

On the other hand, if you want to show by analysis how much force may be required to overcome the static friction present at your clean metal/metal interface, then a mu value that is conservative in the other direction would be in order. For this approach, a static mu of 0.50 (or higher) would probably be suitable.

Also, don't forget to take the high CTE mismatch between steel and mag into account if there is a change in temperature involved.

Hope that helps.
Terry

 

[AIREDALE]

Good information all around. Very much appreciate the input.
The problem we are dealing with is an interferrence fit flanged bushing installed in a magnesium gear box casing mount foot. We are seeing a number of these bushings slip as the mounting bolt torque specification is approached/reached (656-724 in-lbs). Note that this is not a new design and has been in use for many years without reported problems.

I've been using the Engineers Edge Press Fit calculator in an attempt to determine the maximum sustainable torque for the installation (reverse engineering from the OEM spec). Using the conservative figure for mu of 0.15 indicated a flawed design that will allow slippage with a bore in the upper tolerance range and the bushing dia. in the lower tolerance range.

If I raise the coefficient of friction to a mu of 0.60 (magnesium on magnesium), the design works. The problem is determining an accurate mu for the steel/magnesium interface. I thought someone might have an accurate figure. Will have to perform tests to determine it.

Of course, we are also looking into any variations in processes which may be introducing a failure mechanism.

 

[rb1957]

sounds odd to me ... why'd the bushings be slipping as you torque the bolt ? the interference fit is between the bushing and the body, right? is the bolt interference fit as well ?? part of the question is how well the Cd plating is holding up ... it's easy to scratch off ... i'd expect you're freeze fitting the bushing ...

 

[AIREDALE]

Yes, process calls for freezing the bushing and installing within 1 minute. The bushing outside diameter is brushed with epoxy primer and pressed into bore using an arbor press.

 

[rb1957]

ok, freeze fit, wet installed ... needing the press indicates that frozen it is still a colse fit, maybe as it's coming up to room temp it needs some "encouragement" ... surprising that such a bushing is spinning when you torque up the fastener ... 600+ in.lbs indicates a pretty darn big fastener > 1/2" dia ?

still is there interference between the fastener and the bush ?

 

[racookpe1978]

Please draw your problem - Your comments are not clear.

(Conventional sliding or moving Friction coefficients aren't really a function/relevant in interference fit problems since the resistance to motion is developed before motion occurs.)

 

[rb1957]

i think what the OP is saying is torque is being applied to the bush (how?) and is being reacted by friction between the bush and the body, and the friction force is due to the normal force (from interference) * mu.

in part i think he's answered his question ... with minimal interference (adverse tolerances) you get slippage. wouldn't the appropriate mu be for Cd and Mg ? slipping is only part of the problem, 'cause once it slips i reckon it'll destroy the Cd plating.

myself, i'm not sure why the bushing is reacting torque, unless the fastener has interference with the bush.