Mech Engineering question

Discussion in 'Community Discussion' started by dukebound85, Nov 9, 2009.

1. dukebound85 macrumors P6

Joined:
Jul 17, 2005
Location:
5045 feet above sea level
#1
OK so I have graduated and am trying to stay sharp on my skills

Sadly lots have been forgotten lol

I am working on review questions from the FE test and one is:

Strain:
1) .001
2) .0018
3) .003
4) .0009
5) .0011
6) .0017

Mean Strain:
1) 0
2) 0
3) 0
4) .02
5) .02
6) .02

Cycles to failure:
1) 500000
2) 12000
3) 2200
4) 480000
5) 45000
6) 7000

Assume Youngs Mod = 25e6psi

Question
1) Plot the design fatigue curve
2) show effect of mean stress

Any help would be awesome as my textbooks arent helping me

What I have done
1) converted the strain to stress by using youngs mod after reaarangin cycles from least to max
2) plotted stess vs cycles

I believe that is the right way yet i can not use excels function to get the fatigue curve. How do I get the fatigure curve in excel?

How do I incorporate mean effects

lol thanks at this odd request. just buggin me

2. paolo- macrumors 6502a

Joined:
Aug 24, 2008
#2
Hey, I'm currently taking a materials course, second year mech engineering student. I guess I could try to help you out, but I'm taking the course in french and don't know or haven't seen yet what is mean stress is. Wiki ain't of much help either.

3. Cannedkoala macrumors member

Joined:
Jun 8, 2009
#3
Mmm I do aerospace but we don't use S-N Curves.

Anywho, I think you're right about plotting stress vs. cycles. (SN Curve) That should give you a smooth-ish curve showing how many cycles to failure under a repeated sinusoidal application of a particular stress.

I think you should make two curves though. Under a sinusoidal loading the mean strain is going to be zero. If you shift the curve up a bit however, the mean strain is going to be higher. eg.

Max strain | Min strain | Mean strain | Change in strain
+100 -100 0 200
+200 0 100 200

The change in strain is the only thing that is relevant for fatigue. The actual maximum stress isn't as much of a factor.

It's worth noting that the mean strain is significantly more than the testing strain. I don't quite understand how your peak 'strain' can be lower than the mean. I'm going to take a guess and say that 'strain' is actually the 'change in strain' about the mean. Whether it is the full range of strain change or just the difference (amplitude) from the mean I can't say.

Having a mean strain of zero though makes it a bit trickier. From what I've learned, tension is the driver of fatigue, and compressive stresses don't have as much effect. So in my example above, the effective change in strain with regard to fatigue will be 100 and 200 respectively, because you can effectively ignore the negative stresses. I don't think you will need to worry about this though because it could show in your results.

As I said at the start, I think you should be making two curves, one for each mean strain. Then you plot two curves and compare the effect of a difference in mean strain. I'm going to guess that the zero mean strain will have more cycles to failure for a given stress because its effective stress amplitude is going to be half that of the mean strain of 0.02.

I could be wrong =D so don't rely on all this.