Shear Pin Theory & Practice 1

[NZHEng]

Hi All
I am currently redesigning a shear pin for a hydro turbine. The nature of the fluctuating loads makes this a very difficult problem from a fatigue stand-point and the pin solutions are, as a result, novel. As the selected material options are quite ductile I completed bench testing and have received unexpected results.

The materials tested are 2205 duplex stainless steel and E470 medium carbon steel (similar to 1045). Drawings of the basic test pieces are attached.
Taking ultimate shear strength as 0.67 of UTS, I designed these to fail at 572 kN in double shear.

All samples broke significantly higher than predicted by theory. The SS pins broke 39% and 44% higher. The E470 pins broke 32% and 36% higher. A sample chart of each is attached and photo of test setup.
The pressure gauge is calibrated and I have also calibrated the press with a load cell. I have measured the press bore and it compares well with the measured data. As such I believe the data captured is true.
Sample test charts and photos of broken pieces attached.

2205 Material Cert Properties: Yield = 476 MPa UTS = 737 MPa Elongation = 40%
E470 Material Cert Properties: Yield = 506 MPa UTS = 658 MPa Elongation = 22%

I have previously completed testing on both 1045 and 4140T shear pins which broke within 5% of predicted. The only significant difference being the body of the pin was much larger (90 mm opposed to 50 mm). The test pieces I am reporting on here were noted to be 'squashed' oval by 1-2 mm so wonder if test piece rigidity is a factor.

The displacement before fracture is significant, so my current theory is the high elongation is causing the the material to be drawn in a tensile manner before failure, thus skewing the results to a higher breaking load with the failure occurring due to a combination of both shear and tensile failure.

I am preparing some test pieces in 4140U and 4340U to test this theory, as these material are more brittle but still resistant to fatigue.

In the mean time, I would greatly appreciate any thoughts or other experience on such applications and as to why these are breaking so high over the predicted as am not confident I am understanding the mechanism.

 

[hydtools]

Work hardening may have increased material properties leading to higher failure stresses.

Ted

 

[EdStainless]

And those grooves are a bit narrow of the depth.
We used to make these from Monel K-500 (annealed and aged).
We had to make test pieces for each lot.
After about then lots we could just make two test pieces.
As I recall we were working about 20% higher than predicted.
Use materials with low work hardening, no austenitic or duplex stainless.

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

 

[Compositepro]

Ed, feedback like that can be worth a million dollars! At least $1. I am serious.

 

[NZHEng]

Thanks, that is some very helpful comments. Particularly regarding Ed's experience with higher breaking loads.
I see the Monel K-500 in the annealed and aged condition has a fairly high elongation also.
I guess it stands to reason the greater a materials elongation the greater the capacity for work hardening.

I anticipate moving to 4140U / 4340U will address this issue. I will post our findings from the next lot of testing next week.

 

[EdStainless]

We never used steels because even just a little corrosion changes everything.
Don't confuse ductility with strain hardening. They often don't go hand in hand.
High Mn alloys (stainless or not) often have higher work hardening rates.

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

 

[saplanti]

I feel that there is a design problem in the system or selection of material. Ductile materials are certified for the minimum yield and ultimate tensile properties. They may have higher values then the nominated in the standards.

Photos show the fatigue failure in the pin may be created by the involvement of bending as well. But it is difficult to see if this is double shear or single shear failure from the photographs. If it was a double shear failure this makes the shear failure control more difficult because of the range of properties.

It is too difficult to say something without seeing the full picture (the system drawings, why the safety shear is used and where). I would say that the casting materials are more predictable for failure ranges, however they are brittle, so it may be problem under cycling loads. This option may be suitable for bending failure for safety against overload. Leaving the safety failure under shear to pin may be a mistake in the design for the range of ductile material properties. But this needs to be investigated.

These comments are just made under the available information. I hope they help.

 

[rb1957]

structural fuses are tricky to work out. There are several examples in my field. For us, a key consideration is QC, and ensuring that (as far as we can) each part is expected to fail within a narrow load band.

You're right, getting the balance between static failure and fatigue performance is difficult.

Is excessive deformation a sufficient failure mode ?

Using the most ductile material may be your best bet.



another day in paradise, or is paradise one day closer ?

 

[NZHEng]

Thanks all for your input.
We completed additional testing with 4140U and 4340U samples.
These broke within 2% to 5% of theoretical.
I have also located some past data on 4140T samples which also broke within 5% of theoretical.

Being a stronger more brittle material made the difference I believe. I also suspect the rigidity of the body of the pin (away from the shear area) has a part to play in keeping the failure mode predominantly in shear.

Hope this helps someone else in the future.

 

[saplanti]

Good to hear that you solved the problem. I wish you can provide some pictures with the new material and design that you applied. I was wondering the shear failure type, and the shear angle on the cross sections. Kind regards.