High fatigue in the counterbore holes in the FEA 2

[AlexSK]

Hello all,
I was reading eng-tips for serveral years and it helped me to find solutions for many problems. Now I have a problem with fatigue in the counterbore holes in the FEA, so I will be happy to hear your comments about it.
I am analysing 6061-t6 aluminium frame with the heavy loaded bolted connections (counterbore holes, variation of mean stress in hi-ten bolts is close to the limit of +/-5%). Fatigue life of the frame is about 1e8 cycles @ reversible bending loads. It is not fatigue tolerant structure, as no one will inspect it for the next 25 years.
At the moment I see three things, refer to the attached images:
1) the std. counterbore holes with small fillets (~0.2-0.4mm) fail in fatigue due to the high stress range in the fillets. The calculated Miner's sum is as high as 10+ for the given histogram of the alternating loads.
2) another observation reveals high tensile mean stress (~80% of Sy) in the fillets once hi-ten bolts are torqued up. So, once I apply mean stress correction, the Miner's sum in the fillet becomes even higher.
3) the counter-bore holes with larger fillets (5mm or more) provide adequate fatigue, showing Miner's sum of 0.3-0.5, which is acceptable for design. The disadvantage of this solution is that it requires extensive milling of the counter-bore holes in order to get nice & large fillet.
Has anyone experienced similar problems with counterbore holes? Any comments?
Thanks,
AlexK
Sr. Mechanical Engineer
VIM AUSTRALIA
MARINE ENGINEERING

 

[FastMouse]

I have seen fatigue problems in aluminium alloy with deep spotfaces, similar to counterbores. The spotface fillet rad was, I think, 1/16 inch which is ~1.6 mm. (It was a while ago and the details escape me.) The in-service cracks occurred much earlier than the intended design service life. The original design substantiation was based on test (with necessary assumptions about the loading) and did not benefit from a high-fidelity FEA.

Based on the info supplied, I think that your points 1) to 3) indicate the solution.

 

[AlexSK]

Thanks FastMouse! That makes sense now

 

[tbuelna]

AlexSK,

Go ahead and specify a larger diameter counterbore with the larger fillet radius to meet your fatigue stress requirements. Most CNC programmers now prefer to spiral mill counterbores rather than using a dedicated piloted counterbore form cutter, since it usually requires fewer tool changes. In fact, for features like counterbores or holes that are not small or deep, CNC programmers will mill them whenever practical.

 

[AlexSK]

Sure Tbuelna, will do now

 

[Tmoose]

If the problem is solved with slightly larger radiuses, Is it worth investigating alternate details like tapered counterbores, stress fooling cuts outside the counterbores, shot peened counterbores and radiuses, countoured T-shaped steel washers under the bolt heads, etc?

 

[AlexSK]

I think there might be some other solutions too. I will have a meeting with my manager & manufacturing engineers next week to discuss what we can do with that

 

[dhengr]

AlexSK:
Have you actually seen fatigue and failures in the real world, or are you just seeing these high stresses on some FEA output? You are almost always going to see some of those types of high stress areas on FEA output, and you have to know how to interpret them and what they really mean, and how to fix them. It would be good to see a few sketches, real views of this piece, including dimensions, loads, etc. so we could understand what it is and what you are actually trying to do. You would likely get more meaningful help this way.

 

[AlexSK]

Dhengr, I've never seen counterbore holes with sharp spotface fillets in the structural parts and that confuses me a bit. Before posting this I have analysed the structure and keen to trust the results, as different model configurations & setups show similar results. Our company does not have testing facilities to prove the calculations and FEA, but so far (20years) our marine structures engineered to DNV/Lloyds/BV rules never had fatigue issues, so we are lucky from that point of view.
This issue with the counterbore holes appeared when the part was already machined and assembled. The engineering report does not say anything about the peak stress in the spotface fillets, apparently they were not considered in the FEA at that time (i.e. the heavy stressed counter-bore holes were analysed with 90deg corners @ spotfaces).
Just for clarity, I attached a couple of images, so you can see what I am talking about
Regards,
AlexK

http://files.engineering.com/getfil...-4f88-af87-002560d96673&file=beam_profile.jpg

http://files.engineering.com/getfil...52a3-472c-aaf0-4c6fd32fe4a6&file=IMG_1170.JPG

 

[SwinnyGG]

Where are you getting your high loads from?

Your images show a plate with a centrally located bushing for what I'm going to assume is locating a section of marine shafting.

Your forces indicated on this images also appear to indicate that all forces are in plane with the part- so where are you getting bending load?

On top of that, your image indicates that this part is isolation mounted- is it hard mounted anywhere? If not, isolation mounting is going to greatly limit how much load the plate can attract.

In short, your assumptions about the loads might be unrealistically high.

 

[AlexSK]

That's right, the loads are in plane with the plate. This plate is actually a beam on two pivots, that takes high vertical bending loads (~700kN) and horizontal yaw loads (~200kN) generated by the large marine stabiliser. The elastic mounts are engineered to withstand all the loads. The stresses in the hot spots of the beam are definitely higher when the beam is resting on the elastic mounts, than rigid supports (more DOFs).

 

[SwinnyGG]

Ahh- a stabilizer. That explains the high loads and moments.

Thanks for the detail.