Highest fatigue strength options for hand tool with significant stress concentration. 1

[Nereth1]

Hi all,

Our production engineering has designed a hand tool that significantly improves a process speed, but by its design, it has an 0.8mm thick x 5mm long cantilevered section with a stress raiser on one side, placed under unidrectional bending. Sketch attached.

The stress raiser is on the compressive side, so I'm not sure if it's actually an issue. Arguably it is drawing stress away from the tensile side I suppose, and we should consider making it even worse.

In any case, we can't find a material suitable to make this work for more than a few weeks. Money is effectively no object due to the amount of time this tool saves, versus the frustration of remaking them every week as they fail with what we have attempted thus far:

1) K1045 at around 400MPa yield, visibly bends after a dozen uses.
2) 4140 at around 900MPa, lasts a good few hundred uses.
3) Bohler K245, (

http://www.buau.com.au/cold-work.php)

heat treated professionally to around 58 Rc. Lasted a few thousand uses. This was not heat treated in a vaccuum furnace.

Next, we have been given a handful of tool steel options from the local Bohler rep, and were going to send the lot interstate for a more precise, cleaner heat treatment in a vaccuum furnace. Hopefully one of those works. If they end up failing in a month, I want to be prepared to get a little more exotic, for which I was hoping for the help of the experts here. What are everyone's thoughts on:

1) Other, non tool-steels for this application. I was thinking a lot of these hard tool steels are not designed for this kind of fatigue, so I was looking at other UHS steels e.g. Ferrium M54, but I don't know how they would compare.
2) What do we need to watch out for in the heat treat - e.g. as we are going up to high 50 Rc, I assume we are quenching pretty aggressively and not doing a lot of tempering - but would such a thin section attached to such a thick section be subject to massive internal stresses that would be counter productive in that case? How do we control that?
3) I have been thinking about secondary processes like nitriding to put a compressive preload in the tensile surface, but with such a thin section, would there be issues with us just accidentally nitriding most of the section depth, and having the centre of it go too far tensile and thus being counterproductive?

Thanks.

 

[racookpe1978]

0.8 mm thick at 5mm radius? I can see almost no way to get something that thin that is that long to withstand a bending moment.

Or am I backwards, and that is the notch into the final part that is under stress?

 

[tbuelna]

Your localized stress seems to be a result of the rather abrupt transition in section at the 0.8mm radius fillet. Change the fillet to something like an elliptical or parabolic profile. And taper the section thickness along the cantilevered length to provide a more even stress distribution.

 

[Nereth1]

racookpe1978,

No, you have it the right way around. Obviously something so small doesn't resist a lot of bending moment, but it still follows the same principles - smaller stuff just works on smaller forces.

tbuelna,

I think there is some work yet with changing the fillet, but I think the end result will be material * design, as changing the design alone will not solve the problem, so I'm really hoping for some materials advice in this discussion.

The fillet geometry is actually a very interesting discussion regardless - A greater stress concentration on the compressive side should slightly reduce the stress through the rest of the section (which includes the tensile side), and since this is uniaxial bending, that might be beneficial.

 

[dik]

Can you use a copper beryllium alloy?

Dik

 

[Nereth1]

I have never used one before in a design (and neither has anyone else in my company), but I don't see why not.

A quick google search reveals impressive strengths, but not as impressive as higher end steels - and I imagine it has no endurance limit. Any particular reason for recommending them?

 

[dik]

Copper beryllium has almost limitless fatigue resistance.

Dik

 

[Tmoose]

Got some pictures of the broken ones?
Due to the lack of background info provided, I'm imagining fatigue cracks originating from the corners of a simple rectangular cantilever.

https://i69.photobucket.com/albums/bb366/Alby2298/2007 Toyota Tacoma/7E6396FE-orig.jpg

Do your calculated stress levels explain the breakage?


Is the stress level due to a required amount of force, or the result of a required displacement of the end of the cantilever?

If only the 0.8 mm displacement is required, then making the part thinner and better shaped will reduce the bending stress.

However If a certain amount of force is required, with the limited info provided, it seems like some subtle geometry changes and surface prep like Mil spec shot peening, or prestressing the leaf to yield would provide a useful boost in fatigue life, even if a gorgeous optimum material is chosen.

 

[EdStainless]

To start with you need a remelted steel that has a very clean microstructure.
M54 or Airmet 100 would be offer significant improvement over any conventional steels.
Both of these will have much higher fatigue life than any Cu based alloy.
You also need a polished surface, no machining marks are allowed, and maybe you should look into shot peening as well.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[dik]

Catch the article, page 35

https://www.pa.msu.edu/people/edmunds/LArTPC/Other_Documents/guide_to_be_cu.pdf

I used to have an Anschutz target rifle 40 years ago that had a copper beryllium firing pin and it was stated in the literature that this alloy had an indefinite fatigue life... that's what made me suggest the alloy. I also understand that pure copper has almost an infinite fatigue life.

I'm not into materials like EdStainless, and, his suggestions are likely far better, and, he seems to work with materials all the time...

Dik

 

[Nereth1]

Hi all,

Some additional context for everyone:

- The cantilever is under a given force, not a given deflection.
- The cantilever is 0.8mm thick. The deflection is much less than this (I was thinking peening would simply damage it due to how thin it is?).
- One of the materials we are looking at is Bohler K890,

http://www.bucanada.ca/k890_microclean_c.htm,

which is supposed to be a cold work powder metallurgy tool steel, to allow a clean microstructure. How would this compare to e.g. Ferrium?

 

[Nereth1]

One other critical note guys, there are circlips sliding over this, I believe they are around Rc 50, so presumably this needs to be a few points higher than that to avoid abrasion? Otherwise, it needs to be somehow plated.

 

[Tmoose]

So maybe the cantilever surfaces are damaged by the circlips ? And maybe the cracks originate from surface damage?
Or, maybe not.

Got some pictures of the broken ones? and pictures and drawings of the assembly

https://equilibregaia.files.wordpress.com/2013/05/bc7a4-blind_men_and_elephant2.jpg?w=552&h=404

8 down, 12 to go.

 

[btrueblood]

What Ed said, you need to find maraging steels. Airmet 100 gets near 300 ksi tensile strength, there are 350 ksi versions of these alloys also, but they are subject to export controls.

https://en.wikipedia.org/wiki/Maraging_steel

 

[imcjoek]

Money is effectively no object due to the amount of time this tool saves, versus the frustration of remaking them every week as they fail

I don't know what this thing does, so my suggestion may be completely off-base...

Can you redesign to make these easy to replace and stamp out a million from some "simple" alloy rather than going for the solid-gold approach?
Make it a twice-a-week preventive maintenance problem instead of trying to find unobtainium?

 

[dgallup]

How many cycles do you need? Have you looked into 17-7? It solved one of my fatigue problems.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[EdStainless]

If you went with a PH stainless use 13-8 supertough.
But I would go with Airmet 100, it is a lot stronger.
The peening is done with very fine diameter shot, so the surface is still very smooth.

DIK, I based my fatigue comment on past experience with spring wire. BeCu is great at stuff when there is corrosion involved. There is also BeNi which is almost impervious to thermal fatigue, I have seen molds for glass that have been hit with molten glass and then cooled with an air blast, over 100 million times without any cracks.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[Nereth1]

So the following options have come up thus far which I need to choose between:

Maraging Steels (C300?)
Ferrium M54
Aermet 100
Bohler K890

What are everyones thoughts on the comparisons between them?

In terms of fatigue strength boosting processes:

- Peening
- Nitriding

Both for compressive stresses - I'm not sure how to choose between them, but I would think for such a thin section, nitriding would give better control of depth and internal stresses? Nitriding would also provide the surface hardness required to resist scratching from the circlips that are at Rc 50ish?

 

[tbuelna]

A maraging steel like C350 (AMS 6515) likely has the highest tensile strength. The residual surface compressive stress from shot peening helps when tensile stress from bending cycles is a problem. Machine the part, stress relieve and heat treat, and use some method to mechanically work the surface areas where tensile stresses are a concern. Conventional shot peening is OK, but there are other techniques like laser peening that are more effective with hard surfaces. If fretting damage from the clips is a concern, a final ISF process on the part would help.

 

[Nereth1]

Hi Tbuelna,

I will look into where I can get C350 then - I am in Australia, hopefully it is reasonable to get some.

I was under the impression that nitriding also induces compressive stresses in the surface - is it not as effective as peening?

I will look into ISF - Does this improve just surface finish, or does it also improve material properties? I am thinking that we shouldn't have big issues with scratching regardless, if we operate above around 60Rc, although I am not sure at what point diminishing returns will kick in in terms of reducing surface roughness to enhance fatigue resistance.

Thanks.