Aluminum Connecting Rods - do they last

[jackboot]

I am designing an aluminum connecting rod with FEA.

What I have is the following:
design life - 36,000,000 cycles
6061-t6 material with "fatigue endurance" 14,000 psi
I know aluminum doesn't have a true endurance limit but I would like to design for 36,000,000 cycles.

For fatigue concerns:
What type of stress analysis should be used:
von Mises?
Principal?
Axial loads only?

The rod is a complex shape that has areas of tensile loading, but the majority of the load is compression (naturally). Should I be concerned with the compressive loads or only the tensile stressed areas?

Please offer some help as this is a real headache.

jackboot

 

[kenvlach]

Fatigue strength for Al 6061-T6 at room temperature is typically given as 13800 psi based upon 500,000,000 cycles of completely reversed stress using the R. R. Moore type of machine and specimen. From

http://www.matweb.com/search/SpecificMaterial.asp?bassnum=MA6016

and aluminum handbooks.

For loading at stress ratios –1.0, -0.50, 0.0, 0.50 and ‘Run-out,’ see Figure 3.6.2.2.8. Best-fit S/N curves for unnotched 6061-T6 aluminum alloy, various wrought products, longitudinal direction, page 3-289 of MIL-HDBK-5J. (Available at

http://assist1.daps.dla.mil/quicksearch/

but note: this Handbook is a 79 MB download.)

I cannot paste this figure, but here is the summary info:
Correlative Information for Figure 3.6.2.2.8

Product Form: Drawn rod, 0.75 inch diameter
Rolled bar, 1 x 7.5 inch
Properties: TUS, ksi TYS, ksi Temp.,^o F
45 40 RT

Specimen Details: Unnotched
0.200 inch net diameter

Surface Condition: Not specified
Reference: 3.2.1.1.8(a)

Test Parameters:
Loading - Axial
Frequency - 2000 cpm
Temperature - RT
Environment - Air

No. of Heats/Lots: Not specified

Equivalent Stress Equation:
Log N_f = 20.68 - 9.84 log (S_eq)
S_eq = S_max (1-R)^0.63
Std. Error of Estimate, Log (Life) = 0.48
Standard Deviation, Log (Life) = 1.18
R² = 83%

Sample Size = 55

[Caution: The equivalent stress model may
provide unrealistic life predictions for stress
ratios beyond those represented above.]

Reference:
3.2.1.1.8(a) Howell, F. M., and Miller, J. L., “Axial Stress, Fatigue Strength of Structural Aluminum Alloys,” American Society for Testing Materials, Vol. 55 (1955) (MIL-HDBK-5 Item 62-17).

Note: I am not a mechanical engineer, but if the connecting rods are to operate a significant time at higher temperature, it seems reasonable to reduce the fatigue limit in proportion to the reduction in YS. E.g., on going from RT to 212^oF, YS decreases from 45 ksi at 75^oF to 42 ksi at 212^oF. This gives a ‘fatigue limit’ of 12880 psi based upon the RT value of 13800 psi.

I also note that the figure in MIL-HDBK-5J gives higher stress limits for loadings other than stress ratio –1.0, so the 13800 & 12880 psi limits may be conservative.

 

[jackboot]

Thanks for the document-it is very informative.

I have calculated about the same number "fatigue limit" based on the stress condition - 12,700 psi.

However, how do I apply it. If I have any zones above this I am sure of premature failure? von Mises naturally depicts a much higher stress - how should the stresses be analyzed?

 

[SVanden]

For analysis of alum. wheels, I've used an eqn' from ASME of:

S = (202.3/sqrt)+14 (ksi) to get an endurance limit for 6061-T6.

For FEA, I try to keep von-Mises less than 80% of this value (compression or tension).

Higher operating temps would reduce this, how hot will it get?

Shane

 

[jackboot]

It should never see more than 190 F.

For your analysis - the "n" would be the number of cycles - correct? But the equation will always return a result greater than the endurance limit. At any rate - the 80% value is a factor of 1.25 (1/.8) which is in line with my "endurance limit adjustment".

However, my FEA depicts small areas well above this and if I understand your statement - I should have no areas in this "zone." Right?

Do you know what specification the ASME formula is listed?

I appreciate the input.

jackboot

 

[SVanden]

1) 190°F is fine for 6061. It starts artificially aging around 350°, so 190 will cause no problems.

2) "n" is number of cycles, and the ASME eqn I found used the standard fatigue endurance limit of 14 ksi for 1,000,000 cycles (thats the +14 in the eqn). Wheel testing max's out @ 500k, so I use it.

3) Can you refine the mesh throughout the model, or even where you are finding the high limits? Is it just a couple of nodes? (If so, I would probably throw them out, IF you can't refine the mesh in that area). If FEA shows OK, go ahead and cut some test parts (Al is cheap, and FEA is perfect world only). Take out any sharp inside corners that you can, and put at least R0.015" on any outside corners to help out w/fatigue crack propogation, polish if possible.

4) I found that eqn in a web search for a s-n curve of 6061. It was in an online thesis. If I find it I'll post it to you.

BTW - Upstream design software?
What FEA software are you using?
Con-rod designed for...?

 

[jackboot]

Thanks for the information:

BTW - Upstream design software?--Solidworks
What FEA software are you using? COSMOS
Con-rod designed for...? Piston Pump

jackboot