Numerical Simulation - Science method

@Christian M. Appendini @Nguyen Viet Thanh @Danial Ghaderi ； I'm sorry to bother you, but I really need the advice and suggestions from all of you experts

Thanks for your answer Emir Nezirić . I already tried this solution with the script below, but the computation time has exploded...

This script creates, for each iteration, a contact between a random number of nodes of the upper part and the surface of the lower part (with a random friction coefficient).

In this case, I considered between 15 and 50 nodes of the upper part for each contact created.

With this range, around 10 contacts are created.

import random

model = ExtAPI.DataModel.Project.Model

#Nodes in contact of the upper body

IDs = [i for i in range(9780,10167)]

x=0

limite=0

limiteNM1=0

count = 0

for i in IDs:

mu = random.uniform(0.8, 1)

nbNodes = random.randint(15, 50)

limite += nbNodes

newContact=model.Connections.AddContactRegion()

newContact.ContactType=ContactType.Frictional

newContact.FrictionCoefficient = mu

newContact.DynamicCoefficient = mu

newContact.Behavior = ContactBehavior.Asymmetric

newContact.ContactFormulation=ContactFormulation.NormalLagrange

if limite > len(IDs):

limite = len(IDs)

#Creation d'une named selection contenant des noeud

ns = Model.AddNamedSelection()

ns.ScopingMethod=GeometryDefineByType.Worksheet

Criteria = ns.GenerationCriteria

Criteria.Add()

Criteria[0].Action = SelectionActionType.Add

Criteria[0].EntityType = SelectionType.MeshNode

Criteria[0].Criterion = SelectionCriterionType.NodeNumber

Criteria[0].Operator = SelectionOperatorType.RangeInclude

Criteria[0].LowerBound = IDs[limiteNM1]

Criteria[0].UpperBound = IDs[limite-1]

ns.Generate()

#Verif contact

contact = ExtAPI.DataModel.GetObjectsByName(model.NamedSelections.Children[x+1].Name)

if contact:

newContact.SourceLocation = contact[0]

target = ExtAPI.DataModel.GetObjectsByName('bas')

if target:

newContact.TargetLocation = target[0]

x = x+1

for i in range(limite-limiteNM1):

ligne = [i, mu]

muVector.append(ligne)

limiteNM1 = limite

break

elif i == IDs[0]:

#Creation d'une named selection contenant des noeud

ns = Model.AddNamedSelection()

ns.ScopingMethod=GeometryDefineByType.Worksheet

Criteria = ns.GenerationCriteria

Criteria.Add()

Criteria[0].Action = SelectionActionType.Add

Criteria[0].EntityType = SelectionType.MeshNode

Criteria[0].Criterion = SelectionCriterionType.NodeNumber

Criteria[0].Operator = SelectionOperatorType.RangeInclude

Criteria[0].LowerBound = IDs[limiteNM1]

Criteria[0].UpperBound = IDs[limite-1]

Criteria.Add()

Criteria[1].Action = SelectionActionType.Add

Criteria[1].EntityType = SelectionType.MeshNode

Criteria[1].Criterion = SelectionCriterionType.NodeNumber

Criteria[1].Operator = SelectionOperatorType.Equal

Criteria[1].Value = 9392

ns.Generate()

ExtAPI.DataModel.Tree.Refresh()

#Verif contact

contact = ExtAPI.DataModel.GetObjectsByName(model.NamedSelections.Children[x+1].Name)

if contact:

newContact.SourceLocation = contact[0]

target = ExtAPI.DataModel.GetObjectsByName('bas')

if target:

newContact.TargetLocation = target[0]

x = x+1

for i in range(count, count+nbNodes):

ligne = [i, mu]

muVector.append(ligne)

limiteNM1 = limite

count = count + nbNodes

else:

#Creation d'une named selection contenant des noeud

ns = Model.AddNamedSelection()

ns.ScopingMethod=GeometryDefineByType.Worksheet

Criteria = ns.GenerationCriteria

Criteria.Add()

Criteria[0].Action = SelectionActionType.Add

Criteria[0].EntityType = SelectionType.MeshNode

Criteria[0].Criterion = SelectionCriterionType.NodeNumber

Criteria[0].Operator = SelectionOperatorType.RangeInclude

Criteria[0].LowerBound = IDs[limiteNM1]

Criteria[0].UpperBound = IDs[limite-1]

ns.Generate()

ExtAPI.DataModel.Tree.Refresh()

#Verif contact

contact = ExtAPI.DataModel.GetObjectsByName(model.NamedSelections.Children[x+1].Name)

if contact:

newContact.SourceLocation = contact[0]

target = ExtAPI.DataModel.GetObjectsByName('bas')

if target:

newContact.TargetLocation = target[0]

x = x+1

limiteNM1 = limite

count = count + nbNodes

Definitely yes! If you can smartly implement your numerical scheme you can almost get as accurate analysis as the continuous system provides. It is the same as implementing your mathematical analysis using machines. Now you need to take care of all basic needs related to numerical method, mesh, consistency, accuracy etc.

Hey there Rajeev Kumar Mandal! When it comes to numerical simulation software for polishing and finishing processes, one tool that comes to mind is COMSOL Multiphysics. It's a versatile platform that allows for the simulation of various physical phenomena, and its customization options make it suitable for engineering applications like polishing and finishing.

COMSOL Multiphysics enables you Rajeev Kumar Mandal to model and simulate a wide range of processes, taking into account factors such as material properties, surface interactions, and tool geometries. It's a robust solution for exploring the intricacies of polishing and finishing flat materials through numerical simulations.

Give it a shot, and feel free to dive into the specific details of your process. COMSOL should provide you Rajeev Kumar Mandal with a solid foundation for your numerical simulations. Let me know if you Rajeev Kumar Mandal need more suggestions or details!

First of all, the numerical stability does not depend on the round-off error. You can see the original paper of von Neumann, the discrete equation shows the stability property even if the exact precision is used, that is if you compute by hand.The round off error is a supplementary source to this framework and, depending on the spectra of the error distribution, can affect a result.

For this reason at least the double precision is mandatory in scientific commutation.

It seems you are encountering non-convergence issues with your mesoscopic simulation of a three-point bending test in ABAQUS and the load-CMOD (Crack Mouth Opening Displacement) curves are not reflecting the expected results.

Non-convergence in ABAQUS can occur due to a variety of reasons, and here are some general troubleshooting tips that might help you resolve the issue:

If after addressing these points you still face convergence issues, it may be beneficial to review the results of a converged step to determine if there are any physical reasons for the non-convergence, such as unrealistic stress concentrations or unexpected material behavior.

Hadi Basher Sir, Can you please the reference you used for the formula you mentioned so that it can be cited accordingly in paper.

Yes, COMSOL can efficiently handle the electroosmotic flow.

I know this question was long ago, but It is worth saying that I solved my problem using Ansys's Acoustic harmonic module, using a FEM-based solver, instead of CFX, which is a CFD-based solver. Ansys's Acoustic harmonic can accurately simulate harmonic loads and boundary conditions. Although this method is based on FEM, not CFD, It is a promising method to study FSI under harmonic loads. If you need to investigate a thermal or electrical field simultaneously with the acoustic field, use Ansys Coupled Field Harmonic.

Try with Flamelet generated manifold model either in Premixed or non premixed combustion model. Mark few cells at the burner exit and in the patch option (initialization) assign the value of progress variable as 1.

This should work

Nice work. Unsteady injection is indeed very interesting. I remember research on pulsed CO2 injection for recovery at Shell Research in early 1980's (L.de Jong ). The idea was that fast injection would promote fingering and because of this mixing.

Production would occur through the injection injection well after allowing time for mixing of the CO2 with the crude. Basically fast injection inducing fingering can if the injection is followed by a rest period to allow mixing can "stabilize" the contact front between C02 and the reservoir fluid. In case of injection in an aquifer, this will be less efficient than for oil, because the miscibility of CO2 and brine is limited. I therefore think that the CO2 sequestration problem is quite different from the enhanced oil recovery.

I. About simulation and experimental data ― Numerical models can be used to generate data ― numerical simulation ― which can be just as experimental as if generated by physical models. What seems of utmost importance is to precisely characterize the source of data; such as industrial data, pilot plant data, or laboratory data. What classifies data as experimental is not its source, but the compliance with the so-called experimental (scientific) method.

II. Application example ― The recursive least squares algorithm (RLS) allows for (real-time) dynamical application of least squares regression to time series. Years ago, while investigating adaptive control and energetic optimization of aerobic fermenters, I have applied the RLS algorithm with forgetting factor (RLS-FF) to estimate the parameters from the K_La correlation, used to predict the O₂ gas-liquid mass-transfer, while giving increased weight to most recent data. Estimates were improved by imposing sinusoidal disturbance to air flow and agitation speed (manipulated variables). The power dissipated by agitation was accessed by a torque meter (pilot plant). The proposed (adaptive) control algorithm compared favourably with PID. Simulations assessed the effect of numerically generated white Gaussian noise (2-sigma truncated) and of first order delay. This investigation was reported at (MSc Thesis):

Hello, according to your question about welding analysis, maybe this training will help you, according to the headings of this tutorial, I think you can get your answer.

here is that address :

Thanks for your response dear

Yes I've used this technique to get the results. But, as I mentioned i want to know why Floquet ports can't model lossy (as well as dispersive) half-space medium without using the Open add space boundary condition (BC).

Because, by considering a quarter wavelength space between the medium and Floquet port (via using Open add space BC ), we don't model a half-space anymore. Accordingly, we can obtain results approximately by considering sufficient thickness for the lossy medium.

Hi,

In my reasearch project, I have completed recently; a research for a typical turbocharger compressor optimisation. It has been studied air turbulent flow simulated the K-Omega SST turbulence model. My numerical simulations has been validated and I wrote some papers. Please have a look into my papers, hopefully it can help.

Regards,

Hi,

Could you please identify the type of help. Are you struggling in something , please clarify

Try this, you are welcome https://www.researchgate.net/publication/255766292_N-Like_rheograms_of_suspensions_of_magnetic_nanofibers

A symbolic computational software like MAPLE may be helpful.

Ziyi Wu, Thank you, I know what you said. But my question is about how to simulate dynamic contact angle instead of static contact angle.

Try File -> Export -> During Calculation -> Solution Data ->

choose export data every 1 time step, append file name with flow time and select the flow properties you want for the instantaneous snapshots.

if you export it as a HDF5 file, each snapshot can be read directly into python using h5py and you can perform whatever time series analysis you want.

Can you please, clarify how to couple PNP equations in COMSOL?

I don't get any convergence.

Thanks

I assume that the foundations are made of hard and compact rock which can be modeled as an elastic material. In this case, there are almost no differences in the response of the dam (static and/or seismic). Another reason could be related to easier modeling and less computational cost.

Personally, I prefer to model also the foundation layer.

Hi Faisal Qayyum,

Possible approach is to use Part>Create:3D-Shell-Sweep to draw longitudinal section of the tube (particularly, two paths for lofting), Shape>Wire>Sketch to draw circular edges, Shape>Solid>Loft to build the solid body without hole, Shape>Cut>Loft to make the hole, Tools>Geometry Edit…:Edge-Remove to remove unused part of longitudinal section.

I am trying to slove the standard rayleigh lamb frequency equation in isotropic plate using the newton Rapson method ( FORTRAN code ) but I am not able to capture all the root of the equation and modes, even not getting how to generate the data for plotting the dispersion curves... Kindly , please help.

Dear Dipaloke Majumder ,

So your input file looks like something like this ? (CLOAD statements for each node ?) :

*CLOAD

node number, direction, magnitude

*CLOAD

node number, direction, magnitude) or

*CLOAD

node number, direction, magnitude) or

...

Localization means discovering a region in the phase space, where the attractor is located. Usually, such a region lies in the basin of attraction. So, any point from the region tends to the attractor that allows to numerically compute it. Since for hidden attractors, their basins of attraction do not intersect with unstable manifolds of equilibria, one should develop an intuition for discovering such attractors. There are many developed approaches.

I recommend the following two surveys.

Strong Magnetic field inside the compact object plays the crucial role for understanding the variable x-ray burst. Change in internal structure due to the presence of superfluidity, superconductivity and external accretion also are responsible for this variable x-ray burst. What I understand that we have to consider these three first and then to compare with the simulation where the effect is changing, finally we will be able to understand what is actually going on.

Simulation can help us to know the magnitude of possible change and its location to find out inside the compact object.

With regard to Numerical Simulation, I have something to offer. But, what type of Differential Equations are included in the model ? is it ODE, stiff ODE or PDE and of what order?, what is the form of the DE? is it delay , fussy, Wave/Heat etc?. Let see the model maunwala@gmail.com

You can start watching Youtube videos when you learn how to simulate you can start the simulation yourself and when you have problems you can ask questions here or on this site https://forum.ansys.com/

Οf course you can ...e.g. solve the equation e^x+x^3-cos(x)=0 using the program for an interval -10<= x<=10

Reduce[e^x+x^3-cos(x)=0 && Abs[x]<20 ,x]

The same is done for any transcendental equation. But sometimes if it has many roots you will have to wait a long time and this depends on both the range of the interval and the number of roots within a given interval.

Follow

Dear Noorul Huda,

Currently, I am working on the development of Efficient Computational Reduction Methods for Hydrogen GT Combustion for flashback predictions. We are able to develop a methodology for predicting flashbacks using detailed kinetic simulations and progressing towards the development of fast computational processes for industrial applications, but the results are still not available in the public domain. You contact me directly, I can guide you in this regard.

Dear

Have you tried Ansys?

This is a good question.

This is a good question.

Hello dear;

I didn't research on Paleomagnetic studies, but i know 2 methods in order to behavioral study between two things. K-means clustering and Artificial Neural Network (ANN). you can read this combination method in this paper :

good luck

Thank you so much Dr. Smrity Dwivedi for the PDFs.

I have solved the problem without using higher-order basis functions!

Regards

Dear Y.C. Zhong

Thank you for the very thoughtful remarks. I know the literature mentioned in your note and I admire the style and depth of these books. In many ways, the book "Big Breasts and Wide Hips" connects in my perception with Marquez's "One Hundred Years of Solitude" due to a subtle and mysterious surrealism.

Regarding the Nobel prize examples from your comment, I would like to suggest an article in The Atlantic:

(I disagree with some of the opinions above but the paper is very interesting).

As far as my philosophy is concerned, I see a big role for the models of a continuum which, in the present mathematical form, were introduced by Euler in the 1750s. I also think that we can considerably improve our numerical weather prediction models, but I also share some of your skepticism about very long-term digital projections.

I think you should first fix the size of particles and also flow Reynolds number. And simply by doing an experiment on flat plate using particle image velocimetry technique your question will be answered.

There is some useful information in the link below:

Best

@nnaubo, the comsol software uses the Fourier equation to compute the thermal properties and you know there are different solutions of the Fourier equations. So in setting up your heat transfer model for the task at hand, you will have to decide which of the measurement methods you want to adopt I. E. Steady OR transient and which parameters of the solved Fourier equation you want to track I. E. Temperature and time dependent or other parameters. In computing the thermal properties, you will equally have to provide some of the other experimental variables like cross sectional area and the likes depending on your design

No experience with Hg code here. There are many paper and dissertations using ChaNGa to work on planetary/solar system questions.

Hi Zherui,

FS is a parameter that provides flexibility on correlating numerical with experimental results. I reproduced the formulation of plastic strain in the attached file, as one can see the FS value depends on the values of D_1 and D_2 and the negative hydrostatic pressure strength, T.

If one is confident with the remaining parameter values, then an acceptable strategy for deciding on FS value would be to minimise the error between experimental and numerical results based on multiple references. In the case of impact, the ballistic limit, projectile velocity-time profile and residual projectile velocity for instance.

Please consider [1] for a comprehensive analysis of the effects of FS.

For HDCs, FS=0.6 was previously used yielding satisfactory results for the ballistic limit and residual velocity and FS=1.5 for DoP.

In addition, FS=3.0 was used in [2] to validate Johnson and Holmquist [3] cases,

successfully reproducing the JH results.

Hope my answer helped.

Regards,

Theo

Rakesh Kumar from Imperial College London may help you out.

Vishal Singh, IIT Ropar, India may also be helpful to you.

Maybe you can alernatively consider the recursive least squares algorithm (RLS). RLS is the recursive application of the well-known least squares (LS) regression algorithm, so that each new data point is taken in account to modify (correct) a previous estimate of the parameters from some linear (or linearized) correlation thought to model the observed system. The method allows for the dynamical application of LS to time series acquired in real-time. As with LS, there may be several correlation equations with the corresponding set of dependent (observed) variables. For the recursive least squares algorithm with forgetting factor (RLS-FF), adquired data is weighted according to its age, with increased weight given to the most recent data.

Application example ― While investigating adaptive control and energetic optimization of aerobic fermenters, I have applied the RLS-FF algorithm to estimate the parameters from the K_La correlation, used to predict the O₂ gas-liquid mass-transfer, while giving increased weight to most recent data. Estimates were improved by imposing sinusoidal disturbance to air flow and agitation speed (manipulated variables). The proposed (adaptive) control algorithm compared favourably with PID. Simulations assessed the effect of numerically generated white Gaussian noise (2-sigma truncated) and of first order delay. This investigation was reported at (MSc Thesis):

plasticity induced crack closure (PICC) concept and three dimensional (3D) finite element method (FEM) were used to study the effect of compressive residual stress field on the fatigue crack growth from a hole. Furthermore, a new methodology on the basis of a correction factor was presented to increase the PICC precision. The result obtained was compared to two dimensional (2D) FEM, superposition method and Liu’s experimental data. To simulate the elasto-plastic behavior of the material, isotropic hardening was assumed and the Von-Mises yield criterion was implemented. A 3D mesh was built using eight-node hexahedral elements and one half of the specimen was modeled. The simulation results were fairly well correlated with experimental data. Furthermore, the 3D elasto-plastic FEM predicted a slightly smaller fatigue life than a 2D plane stress FEM. Applying the modified PICC method reduces the 3D FEM fatigue life prediction errors.

Dear Intekhab Alam ,

Hope you are well!

The optical properties of the HTL can change while the effect of these changes on the short circuit current remains small because the HTL is chosen such that its absorption coefficient is small and also its thickness is much smaller than the absorber thickness. So, the most incident radiation is absorbed by the absorber layer not by the either the HTL or the ETL.

Please follow the papers:

Best wishes

The answer depends on a scale of your problem. In some application there is a lot of emphasis on representation of the extremely fine details. Please see for example

In the end, it all depends on the capacity of your computer and the ability to digitize complex structures.

For a problem shown in your figure, you can represent the effect of a missing structure by appropriate parameterization of the surface roughness.

kindly check your mail

I modeled the FSW process using both ALE and CEL approaches. You need to specify a failure criterion, e.g., Johnson-Cook damage model. If you do so, you will notice a significant decrease in reaction force values - more realistic values. Also, you will achieve what you want.

Dear Hadi,

FEM using 2D or 3D stochastic model could be numerical simulation studying the effect of corrosion on stress concentration.

Refer site:

This might be useful for you. Just check this article.

Ashish

Dear Zmour,

It can be better in term of diffusion convection reaction. My opinion is little different, the least-squares method has better control of the streamline derivative than the SUPG.

Ashish

For JONSWAP, please check:

https://www.researchgate.net/publication/339272195_Development_of_submarine_cable_hydrodynamic_loads_calculation_program_using_MATLAB

For simple PM, please check pages 6 to 8:

Dear Yamina Soula and Marc R Roussel in our field of research, virtually all renowned journals ask for ca. 3-5 suggested reviewers during the online submission process. Ideally these should not come from the same institution or have been co-authors of previous publications. Many journals also offer the possibility to oppose certain reviewers, e.g,. colleagues who are direct competitors in the same area of research.

Interesting question. Fluent provides you platform for using different input boundary conditions. But it requires a proper meshing and validation of the experimental results. You somehow have to formulate a process of extracting the input boundary conditions from the IR image.

I would think a dynamical model (with nonlinearity) should be constructed to see if the oscillations are stable or unstable and if their amplitude (LCO) can exceed an allowed value

Sorry Ehasan, I have not written the code.

Here is one paper and some figures. I have more stuff, like mass transfer from nitrogen bubbles into water and cloud/droplet interactions. The mathematics are similar.

Hi, SLM-ed SS316L does not need damage model. Based on numerous experiences working with the compression of SS316L lattices no fracture will occur even up to densification. In addition, bulk SS316L (cylinder) sample does not even fracture up to 80 % compressive strain. A solely bilinear or multilinear plasticity model will work for SS316L.

If you are interested in damage model for a less ductile metal, such as Ti64, high strength steel etc, you can consider the Johnson-Cook damage model nontheless. If you are not planning to obtain them experimentally, you can reference them easily from various lattice papers.

Hi Lei! It would be better to share the material and damage model you are using in this study, instead of inp file. These are the two key points to suggest/comment.

Best Regards!

you can contact Mr Essaid El Kennassi to help in this question

You can use FNCorrosion software for analysis.

Muhammad Qasim Zafar , these problems are common and fundamental, so you can find them in welding theory books and molybdenum welding guidelines. Good luck!

Dear Ahmed Ramadhan Al-Obaidi , Danu Sulistiawan , Hamed Majidian , and Chennakesava Kadapa ,

Thank you very much for your reply and valuable contributions...🙏👍

Nice discussion...👍

Subrat Sahu , you want apply FDTD to what kind of equations (Maxwell, elasticity, Navier-Stokes, ...) ?

In my opinion, it depends on the model that you are going to develop. Turbulence is a 3D phenomenon caused by inlet jet mixing, plume entrainment, heat losses, etc. Using CFD, of course, you can model the turbulence. However, it is not the case for one-dimensional models since temperature gradient is only considered in axial direction, but turbulence should also be considered in radial and angular directions. Followings are papers that give some hints to solve this issue:

I hope this reply helps!

The answer depends on several factors. For numerical solutions that are relatively new, difficult, and are related to engineering problems, differences of up to 30% are sometimes still acceptable....

you can visit from my publication

Dear Hessam Keshavarz,

The numerical model for the simulation may require a time independent formulation (time scale of channel flows) which can be used to describe the gaseous flow in order to calculate heat source terms for a transient heat conduction equation for the solid. Input data required.

The two-dimensional flow field of the fluid can be solved through the model and input data are required.

(1) Inlet (velocity, temperature, density, species mass fractions)

(2) Wall conditions (axial temperature profile) the two-dimensional flow field of the fluid can be solved.

(3) The material properties (density ρ, heat capacity Cp and thermal conductivity λ) are functions of the local temperature and material (monolith, insulation and canning) and can also be specified as functions of the direction.

(4) Different adsorption sites on the metallic catalyst surface, as surface reactions are considered between NO, CO, and O2 only for the kinetic data of the mechanism may be considered.

(5) The parameters of the catalyst used, e.g., metal composition and loading, dispersion, viscosity etc.

(6) Thermal capacities and insulation, varying design parameters, as geometry changes, variations in engine out emissions etc.

Ashish