BarElement ConcentratedLoad returns incorrect results for moments

[SerGlushko]

Describe the bug Greetings! I've tried to apply moment as a ConcentratedLoad to a BarElement and I was getting varying incorrect support reactions depending on the IsoPoint. When the moment is applied to IsoPoint -1.0 or 1.0, however, the results are correct.

To Reproduce This is the code We've used and caused wrong result/ runtime error

var m1 = new Model();
m1.Nodes.Add(new Node(0, 0, 0) { Label = "n1" });
m1.Nodes.Add(new Node(0, 0, 6) { Label = "n2" });
m1.Nodes["n1"].Constraints = new Constraint(dx: DofConstraint.Fixed, dy: DofConstraint.Fixed, dz: DofConstraint.Fixed, rx: DofConstraint.Fixed, ry: DofConstraint.Fixed, rz: DofConstraint.Fixed);
m1.Elements.Add(new BarElement(m1.Nodes["n1"], m1.Nodes["n2"]) { Label = "r1" });
(m1.Elements["r1"] as BarElement).Behavior = BarElementBehaviours.FullFrame;
(m1.Elements["r1"] as BarElement).Section = new BriefFiniteElementNet.Sections.UniformParametric1DSection(16.43 / 10000, 541.22 / 100000000, 44.92 / 100000000, 541.22 / 100000000 + 44.92 / 100000000);
(m1.Elements["r1"] as BarElement).Material = BriefFiniteElementNet.Materials.UniformIsotropicMaterial.CreateFromYoungPoisson(210 * Math.Pow(10, 9), 0.3);
m1.Elements["r1"].Loads.Add(new BriefFiniteElementNet.Loads.ConcentratedLoad(new Force(0, 0, 0, 0, -2000, 0), new IsoPoint(0.0), CoordinationSystem.Global)); //The moment is applied to the center point of the element
m1.Solve();
var r11 = m1.Nodes[0].GetSupportReaction();
var r12 = m1.Nodes[1].GetSupportReaction();
Console.WriteLine("n0 reaction: {0}", r11);
Console.WriteLine("n1 reaction: {0}", r12);
Console.ReadKey();

Expected behavior Although near-zero Fx result for n1 is exected, My must be equal to -2000

Additional context The illustration:

Current results are as follows:

[epsi1on]

If user applies moment about Z instead of Y, then result is right.

probably there is problem with EulerBernoulliBeamHelper.GetLocalEquivalentNodalLoads() (pre-process phase) or EulerBernoulliBeamHelper.GetLocalInternalForceAt() (post-process phase)

Each one should be tested separately, unit tests for these two methods also should be revised for any probable calculation errors :D

[SerGlushko]

I've got a question about legacy components. Since FrameElement2Node is now legacy, are there any plans and maybe a time frame on ConcentratedLoad support for this type of element too? Thanks in advance.

[epsi1on]

This issue is in progress. Currently no plan to develop legacy code. Thanks

[epsi1on]

Problem fixed, it should work fine now,

However I'm afraid to say fixing this bug may create another bug, Could you please also test the correctness of internal force of this beam? (I mean BarElement.GetExactInternalForce() method)?

Thanks

[SerGlushko]

It seems the results are correct for both support reactions and internal forces in this example.

Trying to read exactly at the point where the moment is applied throws an exception, but I guess there's no way around it. Thanks!

[epsi1on]

Trying to read exactly at the point where the moment is applied throws an exception, but I guess there's no way around it.

Yes because the moment is not a single value at the point that moment is applied. Take a look at this example image:

you can calculate exact internal force a little before (like 1e-9) and a little after this location and draw or do the rest of calculation.

Take a look at file BarElement.cs method

public Force GetExactInternalForceAt(double xi, LoadCase loadCase)

to see how these points are calculated.

keyword: InvalidInternalForceLocationException