Design of Experiment- Cyclic Loading

[mhajinaw]

G'day guys,

I need to design an experiment to investigate the behavior of steel under cyclic loading. I am particularly interested in understanding the phenomenon of Bauschinger effect after reverse loading. I noticed that one of the common way to perform the test is through cyclic tension-compression test. But the main drawback for this test is that the specimen has the tendency to buckle as it goes through compression load.

I have this idea to conduct a test using cruciform shape specimen. Since the specimen has a shape of "t", what I will do first is to apply tension on vertical arm. I will just let the horizontal arm be free. This also means that I just conduct the test on a Uniaxial test machine. Based on my FEA simulation, it looks that the region at the centre of the specimen will also experience compression because I do not constraint the horizontal arm.

In order to reverse the load, I will take out the specimen and place the horizontal arm to the fixture of the test machine. I will apply tension on this horizontal arm which means that I shall obtain tension and at the same time the previous elongated arm (vertical arm) will return back to its original position. So, it seems to me that I can conduct tension-compression test by repeating this cycle.

But, maybe I miss a few important practical aspects. Hence, I would like to hear your thoughts on this idea.

Cheers

 

[MintJulep]

Seems like you are missing a few important practical aspects.

But it's hard to tell because your description is not very clear.

Maybe a sketch.

 

[mhajinaw]

Here, I've attached the 1/4 a quarter of the specimen sketch. I hope this will give you a good idea on the shape of specimen that I wrote in previous post.

As you can see, if I apply tension to the vertical arm and let the horizontal arm to be free, it looks to me that the horizontal arm will be shorten. And if I'm correct, this is due to the Poisson's effect. And if I now switch to the horizontal arm to be connected to the machine fixture and I applied tension, I will be able to get compression on the initial vertical arm. So, by repeating this cycle, I can somehow conduct cyclic testing. Am I right?

If something is not clear, just let me know again. I'll try to explain it.

 

[MintJulep]

it looks that the region at the centre of the specimen will also experience compression because I do not constraint the horizontal arm.

I guess that the boundary conditions in your FEA simulation are not consistent with the emphasized assumption.

 

[mhajinaw]

Well I actually apply boundary conditions that only allow the horizontal arm to move freely in horizontal direction (x-direction). That is what I meant by the previous quote. I think this is the right representation when modelling in FEA.

 

[SwinnyGG]

How do you intend to actually constrain this part in test?

Your FEA simulation is not realistic unless what you're calling the 'horizontal arm' is actually constrained so that it cannot rotate.

The force shown in your FEA plot applies a moment to the shape. Think about where that force is relative to the centroid of the cross section.