Stress Singularities is a major concern when analyzing results as they considerably distort results. They are also a main cause for non-convergence of results. So the first question is -what is stress singularity? This can be best explained by the following example

The above bracket has a localized high stress around the force applied on a point. This Stress can be considerably higher than the operational stress and by applying a more dense mesh around this simply lead to a much higher stress. This phenomenon is known as stress singularity where the stress become infinite as illustrated by the following formula

**STRESS (infinite) = FORCE / AREA OF POINT (almost = 0)**

Therefore to avoid Stress Singularities when applying loads it is recommended not to apply loads at points and small edges.

Stress Singularities can also occur by applying constraints on points and small edges–even faces with sharp corners as illustrated below.

In the above example stress singularities resulted from using automatic convergence whereas the image on next page of the same model is showing the same stress in the area of interest by using the default mesh and no automatic convergence. So interpret results with care

Finally another cause of stress singularity is over simplification of components. Let’s look at the following example.

In this example the fillets were removed to simply the analysis, however when using the automatic convergence the maximum stress value would not converge as all the stress is concentrated around the edge as shown. In this scenario it would advisable to un suppress the fillets (or in cases when fillets are not modeled –is to use fillets to distribute loads).

**So in summary to avoid stress singularities within models is to:**

**Avoid applying loads on edges and small edges**
**Avoid restraining faces with sharp corners, including points and small edges.**
**Apply fillets and chamfers to evenly distribute loads**

** **I hope this will help you gain more confidence in your stress results…..

### Like this:

Like Loading...

*Related*