Mild Steel Rolling/Surface Fatigue

[Martensyte]

Hi Group -

I'm trying to predict rolling fatigue life for cast low carbon steel of about 165 BHN, and at relatively low cycles (~10^5). Most of the work I've found on the subject of rolling fatigue seems to be oriented towards bearing design, with a sprinkling of work on cams and gears, so all the test data and contact life models I've found are more relevent to surface conditions of 450 BHN+, and much higher cycles (10^8+). Some work has been done for railway & crane wheel/rail design - materials are closer, but my application (wheeled vertical spillway gate) is much higher loading & lower cycles.

I've done a fair bit of literature review but haven't come up with much (Tallian, Lundberg/Palmgren, a few handbooks on the subject, AGMA standards, etc.) Just wondering if anyone might have an idea or two about how to approach this...

David G.
Vancouver, BC

 

[TVP]

What are the values of contact stress for these components? I find it difficult to believe that this application can be successful with contact stresses that are much higher than the 2-5 GPa that are routinely seen with roller bearings, etc.

 

[Martensyte]

The contact stress is about 1.5 GPa. I did manage a life calculation based on data I found for AISI 1030 at 10^8 cycles, and adjusting via log-log relation for the contact stress of interest, but that isn't a very convincing calculation. Had to assume S-N curve slope, question about failure mode differences, etc. 10^3 is a reasonable life based on our limited experience with this application, but we need to do a bit better than that given that it is a very "high consequence" application.

 

[Tmoose]

In addition to hardness and exotic heat treatment (for great tensile and fatigue strength) It takes immaculate steel, precise geometry and great lubrication to make the rolling bearings work despite the traditionally miserable combination of similar materials. Often it is not so much subsurface "fatigue" as surface destruction that starts as micro surface distress. I picture any cast steel is as full of inclusions and junk to resemble a sponge.

Can you get dissimilar materials and some form of lube in there?

 

[mewhg]

Can the application be scaled down to accommodate a small scale test? Something that might at least give an indication of order of magnitude durability?

 

[Martensyte]

Tmoose/mewhq: thanks for your comments. More about what we're actually doing - it's essentially a like-for-like replacement of existing cast steel wheels, of similar hardness, which have lasted for over 60 years without any indication of fatigue failure (corrosion is another matter). Despite this strong evidence that the design contact stresses are not excessive, our client is requesting that we review the existing design in light of modern standards & design practice. I'm just about at the point of concluding that there really isn't any such thing, except for some very rough guidlines that are published by USACE and a few hydroelectric authorities. I my view this is a pretty futile exercise, but our client is a conservative bunch.

Tmoose, the wheel path is actually about 230+ BHN, so there seems to be a good difference between the two contact surfaces. So, further hardening of the wheels would make the two surfaces even more similar, which might not really be an improvement. Lubrication is problematic, mainly because the location is within a fish-bearing waterway.

 

[Tmoose]

Like materials" was meant as steel-on-steel. There are mighty few steel-on-steel bearings in your car's engine, and the few that exist generally have gorgeous lubrication.
As near as I can tell playing with hardness to get for longer life is betting dollars to win pennies compared to lubrication.

Some plastics think water is lube.

 

[unclesyd]

Not a very comprehensive article but this might get you started.

http://www.rollpin.com/Usage_Guides/Tension Bushings/Section8.html