Material suggestion for a billet shaft/gear 1

[buzzedge]

Hi guys,

We're looking improve a small oem shaft/gear combo that has seen failures in one specific spot. It's a small shaft 13mm in diameter which has a 90 degree reduction down to 8mm. It has a gear, ~38mm diameter, at the other end of the 13mm section. We're not quite sure what kind of loads the shaft sees, the failures are somewhat rare but still we want to make an improvement. The failures have been where the 8mm section meets 13mm section. We can't modify the shaft dimensions in any way, the 90 degree reduction must stay.

We want to have a one piece billet shaft made but unsure of what material to use so the shaft itself is stronger without compromising the gear.

Thank you

 

[EdStainless]

What is it now? material and strength?
You will have very limited luck until you increase the radius at the transition.
Every little bit helps.
Pictures?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[Tmoose]

There are tricks to play with geometry by which a radiused fillet can simulate a 90 degree corner.

https://i.imgur.com/yAEA3WD.png

My choice would be C,, followed up with shot peening. I figure that would improve fatigue life at least 10X.

Sometimes it is possible to shorten the length of Ø13 mm shoulder and including a proper radius . Then add a washer with a heavy bevel one side of the ID to clear the new shaft radius.

 

[EdStainless]

And how many are you going to make, 6 or 6,000,000?
If you are only going to make a few I would suggest buying some pre-heat treated (quenched and tempered) 4130 bar.
We frequently used detail "C" for shoulders when we couldn't allow a large root radius.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

Posters have so far given good suggestions about improving resistance to fatigue fracture (we presume it is that), but also consider what is driving the failure. Misalignment or anything else that imposes bending must be identified and mitigated.

"Everyone is entitled to their own opinions, but they are not entitled to their own facts."

 

[buzzedge]

Thank you for the suggestions guys.

We don't plan to make a lot of these, less than 100 pieces, unknown what alloy the original is made of. The original version has a slight dip, sorry don't know the technical term, on the 8mm section just before it meets the 13mm section. This little "dip" takes away 0.25mm from the overall diameter, not sure how much strength it removes but for sure it must be the weakest point.

 

[EdStainless]

Take some hardness tests, you can get an idea of strength from that.
Sounds like they put the radius in the wrong direction.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[buzzedge]

I don't have access to a proper hardness testing tool but I do have a set of hardness tester files.

With HRC50 file I think I can start seeing some fine scratches on the shaft. Using HRC55 file I can definitely scratch the shaft.

The HRC50 file has a range of 50-52 and HRC55 file 55-57 so it seems the shaft outer surface is around HRC 53ish

 

[Tmoose]

To improve/correct the design I'd try verry hard to identify issues unique to the actual parts that broke.
- Pre-existing mfg defects like cracks from Heat treating or finishing grinding.
- un-intended geometrical stress raisers.
- Confirmation of hardness
- Observe the fracture surfaces with bright light and magnification. Compare the fracture surfaces to failure analysis literature.

If that investigation came up empty, I'd think about the loading and do some simplified stress analysis. Frankly the low reported failure rate seems to suggest the design is OK.
As suggested by others, check deeper into the possibility of unusual loading from manufacturing or assembly problems.

If the design proves marginal, I'd optimize geometry before throwing fancier materials at it. Swell materials can not compensate for overstress from bad geometry.
Some Successful aircraft have been made from wood, and some from petty exotic metals and man-made composites.

 

[mfgenggear]

Start with sending a part to met lab to analyze , material and heat treat properties. Then do a failure analysis of the fail parts at a lab.
The material will depend on torque, rpm, wear resistance required. Calculation per AGMA and or Din specifications.
A good gear program will easily do this.
And including cycles and safety ratio required.