Tensile test: stress distribution

[pedrontts]

Hi,

I have been trying to simulate a tensile stress on a specimen with cylindrical cross-section (ASTM A370) using inventor and the result of the simulation shows that the maximum stress levels is in the zone between the gauge area and the fillet section. Considering that in lab experiments the specimens sometimes fracture in this region, I did not think this simulation result was very unreasonable, but I did some Googling and found that some other people get the maximum stress in the middle of the gauge section, somewhat far from the fillet section, when they run a similar simulation.

So the questions are:
- Is this stress distribution reasonable or is it completely wrong?
- If it is wrong, would anyone kindly point out possible mistakes in my simulation?

   

[TheCADWhisperer]

---

@pedrontts wrote:

- If it is wrong, would anyone kindly point out possible mistakes in my simulation? 

---

You didn’t Attach your file here?

Did you leave the ends free to translate and set to Detect and Eliminate Rigid Body Modes?

Why are there two arrows on each end?

[Frederick_Law · Accepted solution]

---

@pedrontts wrote:

but I did some Googling and found that some other people get the maximum stress in the middle of the gauge section, somewhat far from the fillet section, when they run a similar simulation.

So the questions are:
- Is this stress distribution reasonable or is it completely wrong?
- If it is wrong, would anyone kindly point out possible mistakes in my simulation?

   

---

My guess, the different is linear and non-linear simulation.

What you did is linear.  It do a simple analysis without deforming the part.

In real life test, the part will stretch.  Middle section get longer and smaller in cross section.

[pedrontts]

@Frederick_Law , you are right. By simply reducing the load to keep stresses below the yield strength (linear behaviour) I got a homogeneous stress distribution in the gauge section of the specimen.
Thanks.