General Mechanical

Stress singularity question.

• Rameez_ul_Haq
Subscriber

If I am applying a concentrated load on a single node, I will expect a stress singularity near that node on the elements. However, if I apply a distributed load on a face, i won't expect the same thing. Why is that? Since as far as I know, if I choose a face and apply a load there, the load is transferred to each of the nodes those are a part of that face. That essentially means that each node has an external load applied to it. Why is this case different than if I just choose a single node and apply a load on it? Why would the former won't cause singularity but the latter will?

Moreover, if I think about it, each node of the model of FEA will actually have an internal load associated with it. So how is that not causing a singularity but applying a concentrated external load on a node is?

• Erik Kostson
Ansys Employee
The most important thing is that we should use a distributed force on 3 D elements rather than a nodal point force which is not realistic (since a force or pressure acts over an area - node does not ot have an area).

In any case as the effect of the loads and any singularities caused by them, vanish and do not influence much the stresses at the area of interest which should be located away from the load area (St Venant's principle).

So that is the most important thing here, distribute your force as over a region (as it is in reality), and if you use a single nodal point force (which is not recommended say on 2D and 3D elements) makes sure it is far away from the area of interest.

Erik
• Rameez_ul_Haq
Subscriber
thankyou for replying. Actually, I am already aware of what you have said. But my question was WHY? Why would the stress singularity be caused if an external load is applied on a single node but if I distribute the load on a face, ultimately that load is going to be divided to all the nodes (kind of) like a concentrated load. Why won't this cause a singularity?
• Erik Kostson
Ansys Employee

I do not have an exact explanation of this and so I can not answer you (in the end of the day there are somethings in FEA that we can not fully explain, we just need to appreciate them, spot them, and then try to resolve them or avoid them or accept them in our models).
One way perhaps to think is that when we use a face pressure or force (a total load say of 1000 N) is distributed and shared by say many of nodes. In a single nodal force one node is taking up all that force, and hence when we refine, we get that effect. While when we refine on a pressure load/ or force acting on an area/face, more and more nodes share the load as we refine, and hence the effect is not visible and even gets very small as many nodes share a fraction of the total load (or is not there due to the consistent load approach perhaps).
See also if this helps:

if you want to investigate further :
There are different type of loads - lumped loads, consistent etc - see for instance books on the FEM on the exact formulation and treatment of these, and perhaps this will help you.
That is all from me - all the best and take care.
Erik.
Viewing 3 reply threads
• You must be logged in to reply to this topic.