BriefFiniteElementNet Issue Tracker Rss Feedhttps://brieffiniteelementnet.codeplex.com/workitem/list/basicBriefFiniteElementNet Issue Tracker Rss DescriptionClosed Unassigned: Build Errors [7]http://brieffiniteelementnet.codeplex.com/workitem/7When trying to build the source from commit 5bec714f81cf4908108bfbbe59f09482e706146f or aecdc448dc542e310bfe04bbbec667ec59309516, the current latest commits on the CodePlex and GitHub versions of BFE, I get the following error:<br /><br />>Error	13	The tag 'ModelVisualizerControl' does not exist in XML namespace 'clr-namespace:BriefFiniteElementNet.Controls;assembly=BriefFiniteElementNet.Controls'. Line 13 Position 10.	X:...BFE.Net\BriefFiniteElementNet.SimpleUI\MainWindow.xaml	13	10	BriefFiniteElementNet.SimpleUI<br /><br />When I try to install using the VisualStudio NuGet Package Manager, I get:<br /><br />>Operation Failed: The element package\metadata\references\group must contain at least one <reference> child element.<br /><br />Could you please provide more detailed instructions on how to build BFE from source, or provide a stable release in the form of .dll's that can be referenced, or provide a stable NuGet package?<br /><br />Thanks!<br />Auri<br />Comments: seems the problem is fixed.epsi1onThu, 02 Mar 2017 05:50:50 GMTClosed Unassigned: Build Errors [7] 20170302055050ACommented Unassigned: Problem with Plate Bending Stress on triangle [8]http://brieffiniteelementnet.codeplex.com/workitem/8I created some TriangleFlatShell Elements with the behaviour FlatShellBehaviour.DrillingDof. The deformation is very simular in comparsion to Ansys. But I got very different bending stresses. The stress was of course not evaluated on nodal loads and supports. <br />The evaluation was done by this code :<br />```<br /> var globTensor = new Func<TriangleFlatShell, MembraneStressTensor>(elm =><br /> {<br /> <br /><br /> MembraneStressTensor xx = new MembraneStressTensor();<br /> <br /> BendingStressTensor xy = elm.GetInternalForce(0, 0, LoadCombination.DefaultLoadCombination).BendingTensor;<br /> xx.Sx = xy.M11;<br /> xx.Sy = xy.M22;<br /> xx.Sz = xy.M33;<br /><br /><br /> var rt = elm.RotateTensor(xx, Plane.XZPlane);<br /><br /> return rt;<br /><br /> });<br /><br /> var bfeElms = new[] { model.Elements.IndexOf(targetElement) }.Select(i => model.Elements[i]).Cast<TriangleFlatShell>().ToArray();<br /><br /> var bsT = globTensor(bfeElms[0]);<br /><br />```<br /><br />Does anyone has an idea why this happened ?<br />Comments: Thank you for reporting the problem. I think you are right and it is possible that there be some bugs in BFE codes. Finding and fixing these errors can needs significant amount of time. The BriefFiniteElementNet.Elements.TriangleFlatShell is going to be obsolete in near future and BriefFiniteElementNet.Elements.TriangleElement will be its successor, So rather to do not expense time on TriangleFlatShell and concentrate on TriangleElement. It would be great if you can have a brief look at code of method you are using, and see if you can find out something. By the way do you know any reference (book, paper, etc.) for transforming bending tensor from local to global system on shell element? Thanksepsi1onThu, 02 Mar 2017 05:46:04 GMTCommented Unassigned: Problem with Plate Bending Stress on triangle [8] 20170302054604ACommented Unassigned: Problem with Plate Bending Stress on triangle [8]http://brieffiniteelementnet.codeplex.com/workitem/8I created some TriangleFlatShell Elements with the behaviour FlatShellBehaviour.DrillingDof. The deformation is very simular in comparsion to Ansys. But I got very different bending stresses. The stress was of course not evaluated on nodal loads and supports. <br />The evaluation was done by this code :<br />```<br /> var globTensor = new Func<TriangleFlatShell, MembraneStressTensor>(elm =><br /> {<br /> <br /><br /> MembraneStressTensor xx = new MembraneStressTensor();<br /> <br /> BendingStressTensor xy = elm.GetInternalForce(0, 0, LoadCombination.DefaultLoadCombination).BendingTensor;<br /> xx.Sx = xy.M11;<br /> xx.Sy = xy.M22;<br /> xx.Sz = xy.M33;<br /><br /><br /> var rt = elm.RotateTensor(xx, Plane.XZPlane);<br /><br /> return rt;<br /><br /> });<br /><br /> var bfeElms = new[] { model.Elements.IndexOf(targetElement) }.Select(i => model.Elements[i]).Cast<TriangleFlatShell>().ToArray();<br /><br /> var bsT = globTensor(bfeElms[0]);<br /><br />```<br /><br />Does anyone has an idea why this happened ?<br />Comments: original topic here: https://brieffiniteelementnet.codeplex.com/discussions/661139epsi1onThu, 02 Mar 2017 04:58:24 GMTCommented Unassigned: Problem with Plate Bending Stress on triangle [8] 20170302045824ACreated Unassigned: Problem with Plate Bending Stress on triangle [8]http://brieffiniteelementnet.codeplex.com/workitem/8I created some TriangleFlatShell Elements with the behaviour FlatShellBehaviour.DrillingDof. The deformation is very simular in comparsion to Ansys. But I got very different bending stresses. The stress was of course not evaluated on nodal loads and supports. <br />The evaluation was done by this code :<br />```<br /> var globTensor = new Func<TriangleFlatShell, MembraneStressTensor>(elm =><br /> {<br /> <br /><br /> MembraneStressTensor xx = new MembraneStressTensor();<br /> <br /> BendingStressTensor xy = elm.GetInternalForce(0, 0, LoadCombination.DefaultLoadCombination).BendingTensor;<br /> xx.Sx = xy.M11;<br /> xx.Sy = xy.M22;<br /> xx.Sz = xy.M33;<br /><br /><br /> var rt = elm.RotateTensor(xx, Plane.XZPlane);<br /><br /> return rt;<br /><br /> });<br /><br /> var bfeElms = new[] { model.Elements.IndexOf(targetElement) }.Select(i => model.Elements[i]).Cast<TriangleFlatShell>().ToArray();<br /><br /> var bsT = globTensor(bfeElms[0]);<br /><br />```<br /><br />Does anyone has an idea why this happened ?<br />epsi1onThu, 02 Mar 2017 04:57:02 GMTCreated Unassigned: Problem with Plate Bending Stress on triangle [8] 20170302045702ACommented Unassigned: Build Errors [7]http://brieffiniteelementnet.codeplex.com/workitem/7When trying to build the source from commit 5bec714f81cf4908108bfbbe59f09482e706146f or aecdc448dc542e310bfe04bbbec667ec59309516, the current latest commits on the CodePlex and GitHub versions of BFE, I get the following error:<br /><br />>Error	13	The tag 'ModelVisualizerControl' does not exist in XML namespace 'clr-namespace:BriefFiniteElementNet.Controls;assembly=BriefFiniteElementNet.Controls'. Line 13 Position 10.	X:...BFE.Net\BriefFiniteElementNet.SimpleUI\MainWindow.xaml	13	10	BriefFiniteElementNet.SimpleUI<br /><br />When I try to install using the VisualStudio NuGet Package Manager, I get:<br /><br />>Operation Failed: The element package\metadata\references\group must contain at least one <reference> child element.<br /><br />Could you please provide more detailed instructions on how to build BFE from source, or provide a stable release in the form of .dll's that can be referenced, or provide a stable NuGet package?<br /><br />Thanks!<br />Auri<br />Comments: Fixed when compiling using Visual Studio 2015! (Also needed to add in a few missing references)aurimasmbWed, 01 Mar 2017 21:48:56 GMTCommented Unassigned: Build Errors [7] 20170301094856PCreated Unassigned: Build Errors [7]http://brieffiniteelementnet.codeplex.com/workitem/7When trying to build the source from commit 5bec714f81cf4908108bfbbe59f09482e706146f or aecdc448dc542e310bfe04bbbec667ec59309516, the current latest commits on the CodePlex and GitHub versions of BFE, I get the following error:<br /><br />>Error	13	The tag 'ModelVisualizerControl' does not exist in XML namespace 'clr-namespace:BriefFiniteElementNet.Controls;assembly=BriefFiniteElementNet.Controls'. Line 13 Position 10.	X:...BFE.Net\BriefFiniteElementNet.SimpleUI\MainWindow.xaml	13	10	BriefFiniteElementNet.SimpleUI<br /><br />When I try to install using the VisualStudio NuGet Package Manager, I get:<br /><br />>Operation Failed: The element package\metadata\references\group must contain at least one <reference> child element.<br /><br />Could you please provide more detailed instructions on how to build BFE from source, or provide a stable release in the form of .dll's that can be referenced, or provide a stable NuGet package?<br /><br />Thanks!<br />Auri<br />aurimasmbFri, 24 Feb 2017 08:35:54 GMTCreated Unassigned: Build Errors [7] 20170224083554ACommented Unassigned: Frame Element with both ends fixed [6]http://brieffiniteelementnet.codeplex.com/workitem/6why I cannot solve a beam with both ends fixed?<br /><br />----<br />Ability to solve models with no free DoF.<br />Comments: anyways ConjugateGradient seems to work fine, as @mostafanouh said. Thanksepsi1onWed, 25 May 2016 09:26:04 GMTCommented Unassigned: Frame Element with both ends fixed [6] 20160525092604ACommented Unassigned: Frame Element with both ends fixed [6]http://brieffiniteelementnet.codeplex.com/workitem/6why I cannot solve a beam with both ends fixed?<br /><br />----<br />Ability to solve models with no free DoF.<br />Comments: No problem. Problem is because stiffness matrix turns 0x0 and code does throws exception for that when trying to multiply 0xn matrix with nxm matrix. this is still in queueepsi1onWed, 25 May 2016 03:06:21 GMTCommented Unassigned: Frame Element with both ends fixed [6] 20160525030621ACommented Unassigned: Frame Element with both ends fixed [6]http://brieffiniteelementnet.codeplex.com/workitem/6why I cannot solve a beam with both ends fixed?<br /><br />----<br />Ability to solve models with no free DoF.<br />Comments: I'm sorry, my comment is wrong. btw, I can't delete it.mostafanouhThu, 19 May 2016 02:33:41 GMTCommented Unassigned: Frame Element with both ends fixed [6] 20160519023341ACommented Unassigned: Frame Element with both ends fixed [6]http://brieffiniteelementnet.codeplex.com/workitem/6why I cannot solve a beam with both ends fixed?<br /><br />----<br />Ability to solve models with no free DoF.<br />Comments: Of course you can, but try to use different type of solver (e.g. ConjugateGradient) because Cholesky solver is limited to positive definite matrices(all diagonal elements are positive) and the fixed-fixed frame element will result in nonpositive definite stiffness matrixmostafanouhThu, 12 May 2016 19:57:34 GMTCommented Unassigned: Frame Element with both ends fixed [6] 20160512075734PCommented Issue: Torsional constant for noncircular sections [3]http://brieffiniteelementnet.codeplex.com/workitem/3Sorry for late reply;<br />There's no explicit formula for any section with known set of coordinates. Integration over a finite element meshed section will find it.<br />Meanwhile, this [pdf ](http://www.dot.ca.gov/hq/esc/techpubs/manual/bridgemanuals/bridge-strudl-manual/page/appendix-c.pdf) contains a full explanation and precalculate torsional constant J for some sections.<br />This can be useful, something like, SectionGenerator.GetBoxSection(/* paramertes */) ..etc<br /><br />Comments: also look at line 341 to 371 from this file: https://github.com/rforsbach/Treu-Structure/blob/master/Sections/FemCrossSection.csepsi1onFri, 25 Mar 2016 09:01:58 GMTCommented Issue: Torsional constant for noncircular sections [3] 20160325090158ACreated Unassigned: Frame Element with both ends fixed [6]http://brieffiniteelementnet.codeplex.com/workitem/6why I cannot solve a beam with both ends fixed?<br /><br />----<br />Ability to solve models with no free DoF.<br />epsi1onWed, 02 Mar 2016 17:27:16 GMTCreated Unassigned: Frame Element with both ends fixed [6] 20160302052716PCreated Unassigned: Implementation of nonuniform beam [5]http://brieffiniteelementnet.codeplex.com/workitem/5Implement nonuniform beam (both stiffness and mass matrix) based on (one of) attached documents<br />epsi1onTue, 16 Feb 2016 07:04:17 GMTCreated Unassigned: Implementation of nonuniform beam [5] 20160216070417ACreated Unassigned: Maximum bending moment [4]http://brieffiniteelementnet.codeplex.com/workitem/4Add proper methods to loads and 1D elements to find out the maximum of specific component of internal force (like My or Mz) at element length directly.<br /><br />Original post from alitorabi ([link](https://brieffiniteelementnet.codeplex.com/discussions/650513#editor))<br />Actually I'm doing the same! but it's not always accurate and if you have >1000 elements it will take quite long time. <br />epsi1onTue, 26 Jan 2016 17:19:21 GMTCreated Unassigned: Maximum bending moment [4] 20160126051921PCreated Issue: Torsional constant for noncircular sections [3]http://brieffiniteelementnet.codeplex.com/workitem/3Sorry for late reply;<br />There's no explicit formula for any section with known set of coordinates. Integration over a finite element meshed section will find it.<br />Meanwhile, this [pdf ](http://www.dot.ca.gov/hq/esc/techpubs/manual/bridgemanuals/bridge-strudl-manual/page/appendix-c.pdf) contains a full explanation and precalculate torsional constant J for some sections.<br />This can be useful, something like, SectionGenerator.GetBoxSection(/* paramertes */) ..etc<br /><br />epsi1onSun, 17 Jan 2016 18:27:23 GMTCreated Issue: Torsional constant for noncircular sections [3] 20160117062723PClosed Feature: Implementing Rigid Element into library [2]http://brieffiniteelementnet.codeplex.com/workitem/2Implementing a rigid element into the library which using it reduces the stiffness matrix size. preferably decimal data type rather than double be used in its calculations for reducing the K matrix be used for getting better accuracy.<br />Comments: implemented by RigidElementepsi1onMon, 08 Dec 2014 05:37:07 GMTClosed Feature: Implementing Rigid Element into library [2] 20141208053707AClosed Feature: Support For Itterative Solver [1]http://brieffiniteelmentnet.codeplex.com/workitem/1Copied from here: https://brieffiniteelmentnet.codeplex.com/discussions/554813 originally written by wo80<br /><br />I'm not familiar with structural analysis using FEM, but I guess your matrices are always spd? You should consider using iterative solvers like cg and make the whole solving process more configurable to the user. I've used your benchmark application testing cholesky vs. cg:<br /><br />```<br /> Grid Size: 15x15x15<br /> Element Count: 9450<br /> Node Count: 3375<br /> DoF Count: 20250<br />```<br /><br />On my machine, cholesky takes ~20sec, while cg converges in ~2sec. Here's the code: [BriefFiniteElementNet-Solver.zip](http://wo80.bplaced.net/files/BriefFiniteElementNet-Solver.zip)<br /><br />I've added the Solver namespace, the only classes that changed are _Model_ and _StaticLinearAnalysisResult_. I haven't cared to make this accessible to the user. To change the solver type, open _StaticLinearAnalysisResult_ and find the lines<br /><br />```c#<br />//ISolver<double> solver = new PCG(new SSOR());<br />ISolver<double> solver = new DirectSolver();<br />```<br />Comments: Fixedepsi1onFri, 08 Aug 2014 13:19:13 GMTClosed Feature: Support For Itterative Solver [1] 20140808011913PReopened Feature: Support For Itterative Solver [1]http://brieffiniteelmentnet.codeplex.com/workitem/1Copied from here: https://brieffiniteelmentnet.codeplex.com/discussions/554813 originally written by wo80<br /><br />I'm not familiar with structural analysis using FEM, but I guess your matrices are always spd? You should consider using iterative solvers like cg and make the whole solving process more configurable to the user. I've used your benchmark application testing cholesky vs. cg:<br /><br />```<br /> Grid Size: 15x15x15<br /> Element Count: 9450<br /> Node Count: 3375<br /> DoF Count: 20250<br />```<br /><br />On my machine, cholesky takes ~20sec, while cg converges in ~2sec. Here's the code: [BriefFiniteElementNet-Solver.zip](http://wo80.bplaced.net/files/BriefFiniteElementNet-Solver.zip)<br /><br />I've added the Solver namespace, the only classes that changed are _Model_ and _StaticLinearAnalysisResult_. I haven't cared to make this accessible to the user. To change the solver type, open _StaticLinearAnalysisResult_ and find the lines<br /><br />```c#<br />//ISolver<double> solver = new PCG(new SSOR());<br />ISolver<double> solver = new DirectSolver();<br />```<br />epsi1onFri, 08 Aug 2014 13:18:51 GMTReopened Feature: Support For Itterative Solver [1] 20140808011851PClosed Feature: Support For Itterative Solver [1]http://brieffiniteelmentnet.codeplex.com/workitem/1Copied from here: https://brieffiniteelmentnet.codeplex.com/discussions/554813 originally written by wo80<br /><br />I'm not familiar with structural analysis using FEM, but I guess your matrices are always spd? You should consider using iterative solvers like cg and make the whole solving process more configurable to the user. I've used your benchmark application testing cholesky vs. cg:<br /><br />```<br /> Grid Size: 15x15x15<br /> Element Count: 9450<br /> Node Count: 3375<br /> DoF Count: 20250<br />```<br /><br />On my machine, cholesky takes ~20sec, while cg converges in ~2sec. Here's the code: [BriefFiniteElementNet-Solver.zip](http://wo80.bplaced.net/files/BriefFiniteElementNet-Solver.zip)<br /><br />I've added the Solver namespace, the only classes that changed are _Model_ and _StaticLinearAnalysisResult_. I haven't cared to make this accessible to the user. To change the solver type, open _StaticLinearAnalysisResult_ and find the lines<br /><br />```c#<br />//ISolver<double> solver = new PCG(new SSOR());<br />ISolver<double> solver = new DirectSolver();<br />```<br />epsi1onFri, 08 Aug 2014 13:18:44 GMTClosed Feature: Support For Itterative Solver [1] 20140808011844PCreated Feature: Implementing Rigid Element into library [2]http://brieffiniteelmentnet.codeplex.com/workitem/2Implementing a rigid element into the library which using it reduces the stiffness matrix size. preferably decimal data type rather than double be used in its calculations for reducing the K matrix be used for getting better accuracy.<br />epsi1onFri, 01 Aug 2014 16:43:44 GMTCreated Feature: Implementing Rigid Element into library [2] 20140801044344P