This is the best video I found in youtube..Which explaned very clearly about Implicit and Explicit. Thank you so much sir. I just subscribed ur channel. keep doing more and more.
Great Explanation with great simplicity Though i have solved many problems i learn today that computer applies implicit and explicit are only for dynamic events but for linear static we have direct methods Thank u very much sir
Always welcome Dear, I suggest to go with all videos of Abaqus playlist of my channel SimTech05..! Videos are not only Abaqus related..also every video have conceptual explanation. Which help you in any CAE software.
Thank you so much!!!! This video was very straightforward and probably the easiest to understand with respect to other cae videos! I will be sharing this with other cae beginners
Dear, For general static analysis mass property (density) not work. So mass matrix is null matrix. Same for damping. However for few static analysis we need mass properties and damping properties..! You mentioned more generalized form - for all static case acceleration will be zero. Velocity may zero or constant.
In the first video of this playlist, you said giving the reference of the book that when the loading frequency is less than 1/4 th of the natural frequency whereas in this video 1/3 rd. Although there should not be a fine demarcation between the two .
Dear, This concept of frequency just a thumb rule (some say 1/3rd some 1/4th). Every time for your model you need to check kinetic energy dominent or not...that is important.
While modelling a masonry wall by simplified micro modelling approach, there was no problem during surface based contact interaction using Explicit solution technique but I got problem during Implicit solution? What could be the reason and how can that be solved? Any idea?
Dear, There is no such direct command to converge solution. It always need to check warning/error message, need to observe interactions. Accordingly modification needed. However, as i mentioned in video Implicit solver may not able to handle complex contact conditions. But if you manipulate default settings in interactions you may solve such complex contact problems with implicit too.
In a separate video, you talk about quasi-static analyses using explicit. However, there is another in-built way of doing quasi-static analyses -- by using the quasi-static option in the Dynamic, Implicit method. Also, I have seen that this uses the static solver (static.exe), at least with the quasi-static option on. Do you know why? Thank you very much in advance.
Dear, Explicit is pure dynamic solver but if problem is sufficient slow we can artificially increase the speed of process hence less event time (for example event is for 10 sec but we can simulate the event for 0.1 sec)..it depends on you assumptions...as i explained in 2 videos...if you further reduce the time...it may possible system no longer remains in quasi-static condition...hence we need to take care of it. On other hand quasi static avilable with implicit technique is completely different.. whatever the boundary conditions...KE always remains zero... I'll explain this within next 3-4 videos. Keep in touch with Abaqus dynamic analysis playlist. th-cam.com/play/PLdXwpuOm8L9p1mPx4EaOCSRYT3Unj-1ql.html
Nice explanation. Nevertheless, at slide 75 the explicit governing equation is disregarding the stiffness matrix which seems wrong. You should have a stiffness matrix multiplying you {D}n vector (while is substracting the mass matrix of course).
Dear, The equation and explanation (page 75) adopted from the book - RD Cook Concepts & Applications of Finite Element Analysis. So i believe alll is good.
As always, thank you for the great video! I have a large model that contains 108 bolts and is a steel connection (grillage for transmission structures). The model has a static pressure applied on top of one of the members to check how the load is transferred through the entire structure. I am using static implicit but model diverges right at the beginning when I apply the pretension bolt forces. Do you think I can use explicit solver for this problem because the matrix doesn’t converge? If yes, how accurate my results will be? Thanks again and I look forward to hearing from you soon
Dear, @Adel Badiee If your load is static there is no meaning to run simulation by Explicit scheme. Error is in your modelling - most probably in interaction/contact definition. If you are using contact pair algorithm, I suggest to use - Slave adjustment - Adjust only to remove overclosure option. However u need to take care same element size in contact surfaces. Also take care geometrical errors (contact surfaces should not interact each other). I suggest to go with interaction videos of playlist - th-cam.com/video/HattYmp94Rk/w-d-xo.html th-cam.com/video/1yIbNUEOANw/w-d-xo.html Another video - Common Errors/mistakes in interaction ill upload soon.
This is the best video I`ve seen on explicit analysis!! But shouldnt there be a D_{n+1} term on the right hans side of the implisite scheme on the slide at 5:00? The implisit scheme looks to be depending on terms from timestep n only, not {n+1}. Wouldn`t this make the scheme explicit by definition?
Dear, 1. D_{n+1} is unknown which need to be find and depending on terms as mentioned. 2. Implicit depends non n+1, n, n-1 and subsequent terms. Look general equation carefully. (1st equation) However, I told a general criteria to distinguish implicit & explicit scheme. Mathematical things are not straight forward.
Dear sir first of all thanks for such detailed video. I have a question regarding the same. If i have two surfaces in contact and my model is subjected to quasi static cyclic loading ( specially in ADAS Dampers) which method will be more accurate, as it is having contact+ dynamic + time increment+ slow loading
When I run explicit solver of Abaqus it aborted and got this error "THIS KEYWORD IS NOT AVAILABLE IN Abaqus/Explicit" do you know what shall I do to solve this error
Dear, .msg file will show you which keyword not available/invalid for explicit step. Few common invalid input are - - 2nd order hex (C3D20), quad (S8) and other few elements not valid for explicit step. - XFEM/VCCT/Contour Integral not valid in explicit - Few contact definition not valid - Some material model also not available in Explicit etc. -
@@arsitech-ltd Ok but I modeled a concrete specimen with a hole using XFEM but the crack started from the precrack and is going till the hole but didn’t continue in the other side of the hole as a lot of cracks generated without any of them completely separated the specimen which doesn't make sense and I get a recommendation to use dynamic explicit solver. So, how can I model that in dynamic explicit solver as you mentioned that XFEM is not valid in dynamic explicit solver.
Dear, Whatsapp group is full (limited participants can join only)..so same telegram group is available..i suggest to join there. Also you can join our facebook community (link in discription). I shared few important notes on Facebook group..apart from that all i have - 2 FEA books and Abaqus documentation.
Thank you for this wonderful video. For a problem, where lateral load is applied on masonry wall with some precompression load, do we need to go for implicit ànalysis or explicit analysis?
Dear, If inertia is not dominant & nature of load is not function of time..u can use general static step aslo. Pre-compression load is not matter at all.
Dear, For the same, I suggest you to go through video - "quasi static analysis in FEA" Available at my channel SimTech05. th-cam.com/video/2P0cuAmjIc8/w-d-xo.html
Dear sir..kindly to clarify this..so in summary the Static: {f}=[K]{x}, Quasi Static:{f}=[C]{x'}+[K]{x}, Dynamic:{f}=[M]{x"}+[C]{x'}+[K]{x}..Am I right?
Great and very helpful video. Thankyou very much. Could you suggest a book related to such fem important things which can help to understand to simulate by writing code or improve simulation. Thankyou in advance. Please keep making such kind of helpful videos.
Dear, I suggest to read book - Concept and Application of FEA (Robert_D._Cook) for modarate and expert people in FEA. And Practical Finite Element Analysis - for moderate and beginners in FEA.
Thank you for that detailed explanation. Can you make video more on material used in dyna . most common interview question that I'm facing. Thank you again :)
I think claiming that the implicit schemes are more accurate than the explicit scheme is wrong. It is true that the force equilibrium is guarenteed when using explicit schemes, but this doesn't guarentee accuracy in dynamic problems. One could throw a ball, and calculate the force equilibrium in the ball at t = t_n, but the ball can still be at the wrong distance. This is because a Implicit Time Integration scheme still generates an error. There is a difference between the implicit system [K]{u} = {f} and implicit time integration scheme. Both contain the name implicit but it is not the same. [K]{u} = {f} can be solved exactly (or almost exact when using Newton raphson method if K depends on u). But implicit time integration schemes do not provide exact solutions in most cases. Please feel free to challange my claim. I'm a master student who just finished a few months of reasearch in the FEM field, so I am definetly no expert.
Dear, This is all about accuracy of operators used in scheme. Many researchers claimed that Modified beta Newmark operator more accurately predict the n+1 term if alpha and beta constant of scheme choosen wisely. However, for complex contact conditions and large deformation implicit may fail many time. But fail to converge solution is different thing and accuracy is different thing. On other hand Central difference operator is straight forward.. without iteration everytime direct solution of governing equation possible.. that's why for large analysis it required double precision...as single precision may have more truncation error. And Here I'm for knowledge sharing and knowledge gaining. Neither I'm accepting your challenge nore denying reply.
Thank you sir for making tutorial about such important topic. I liked it and understood it to some extent as I have a prior understanding of complicated mathematics. However, I would like to pen down the feedback that kindly make videos covering such important topics (like Implicit vs Explicit analysis) with the explanation of its physical significance . It will be helpful to those people who are new to FEA world or for the people who came to learn about implicit and explicit analysis for the first time. Otherwise It is very easy to copy and paste the equation from books and dictate it in video. Please, No offence! I found many difficulties in finding the helpful content during my initial learning of FEA and I still do so that's why making this suggestion🙏🙏
Dear, Always welcome and thank for appreciating work. Whatever i explained about implicit and explicit is only surface level.. I know most of users face problem in the same... so for me biggest challenge was - explain things in easiest way...not the mathematics behind it. Sometimes i feel i not even properly described implicit and explicit. I'll try to record one session of my advance batch and will update.
Great video, but please consider removing the background music for your future videos. It feels unnecessary and distracting from your explanation. Thanks.
Very helpful video
I dont know what can i say. Great contrubution to CAE world again. Thank you
Thanks for positive feedback again.
Keep knowledge sharing.
Great presentation!
Thanks for appreciating my work.
Such a great explanation! Thank you
Thank you very much for sharing. Keep them coming..
Thanks for appreciation.
This is the best video I found in youtube..Which explaned very clearly about Implicit and Explicit. Thank you so much sir. I just subscribed ur channel. keep doing more and more.
Dear,
Go through whole playlist of dynamic analysis and you will be amazed. 4 videos on Implicit explicit topic.
@@arsitech-ltd sure
Great Explanation with great simplicity
Though i have solved many problems i learn today that computer applies implicit and explicit are only for dynamic events but for linear static we have direct methods
Thank u very much sir
Always welcome Dear,
I suggest to go with all videos of Abaqus playlist of my channel SimTech05..! Videos are not only Abaqus related..also every video have conceptual explanation. Which help you in any CAE software.
excellent explanation, your videos are superb
Thanks for appreciating.
I suggest you go with other videos also.
Thanks for a clear presentation of the concepts. Request more such videos.
Dear,
I suggest to go with Abaqus playlist videos, most of them conceptual only.
Thank you so much!!!! This video was very straightforward and probably the easiest to understand with respect to other cae videos! I will be sharing this with other cae beginners
Thanks for appreciation dear.
Surely it should be helpful to others..!
Thanks, a lot Sir for sharing your experience. Keep it up, I am waiting for the next informative video.
New video on element quality check uploaded.
Have a look.
Always welcome.
At the time around 1:10', you said in static situation M=0 and C=0?
I think velocity and acceleration are zero, not those properties.
Dear,
For general static analysis mass property (density) not work. So mass matrix is null matrix. Same for damping.
However for few static analysis we need mass properties and damping properties..!
You mentioned more generalized form - for all static case acceleration will be zero. Velocity may zero or constant.
@@arsitech-ltd Thank you.
You can delete my comment.
Why dear,
You are also right in one aspect. Even if someone wrong...other people read that comment and will learn.
Have been looking for this.. Thanks alot bro
Suggest you to go through all videos of this playlist.
I had faced these questions related with implicit and explicit analysis. Thank you so much for this interesting thing. Sir🙏
Always welcome.
Keep knowledge sharing.
@@arsitech-ltd yes, Sir 👍
In the first video of this playlist, you said giving the reference of the book that when the loading frequency is less than 1/4 th of the natural frequency whereas in this video 1/3 rd. Although there should not be a fine demarcation between the two .
Dear,
This concept of frequency just a thumb rule (some say 1/3rd some 1/4th). Every time for your model you need to check kinetic energy dominent or not...that is important.
While modelling a masonry wall by simplified micro modelling approach, there was no problem during surface based contact interaction using Explicit solution technique but I got problem during Implicit solution? What could be the reason and how can that be solved? Any idea?
Dear,
There is no such direct command to converge solution. It always need to check warning/error message, need to observe interactions. Accordingly modification needed.
However, as i mentioned in video Implicit solver may not able to handle complex contact conditions. But if you manipulate default settings in interactions you may solve such complex contact problems with implicit too.
very helpful vedio
Very informative. Thank you sir : )
Most welcome!
In a separate video, you talk about quasi-static analyses using explicit. However, there is another in-built way of doing quasi-static analyses -- by using the quasi-static option in the Dynamic, Implicit method. Also, I have seen that this uses the static solver (static.exe), at least with the quasi-static option on. Do you know why? Thank you very much in advance.
Dear,
Explicit is pure dynamic solver but if problem is sufficient slow we can artificially increase the speed of process hence less event time (for example event is for 10 sec but we can simulate the event for 0.1 sec)..it depends on you assumptions...as i explained in 2 videos...if you further reduce the time...it may possible system no longer remains in quasi-static condition...hence we need to take care of it.
On other hand quasi static avilable with implicit technique is completely different.. whatever the boundary conditions...KE always remains zero... I'll explain this within next 3-4 videos.
Keep in touch with Abaqus dynamic analysis playlist.
th-cam.com/play/PLdXwpuOm8L9p1mPx4EaOCSRYT3Unj-1ql.html
@@arsitech-ltd Great, thank you very much. I look forward to your video.
@@arsitech-ltd But in that case, what changes if I change the step size T for a quasi-static dynamic, implicit job?
Huge difference.. just wait for next videos...as I'll get time I'll update. I just recovered from viral infection... still on rest.
@@arsitech-ltd Thank you. Get well soon.
Nice explanation. Nevertheless, at slide 75 the explicit governing equation is disregarding the stiffness matrix which seems wrong. You should have a stiffness matrix multiplying you {D}n vector (while is substracting the mass matrix of course).
Dear,
The equation and explanation (page 75) adopted from the book - RD Cook
Concepts & Applications of Finite Element Analysis. So i believe alll is good.
As always, thank you for the great video! I have a large model that contains 108 bolts and is a steel connection (grillage for transmission structures). The model has a static pressure applied on top of one of the members to check how the load is transferred through the entire structure. I am using static implicit but model diverges right at the beginning when I apply the pretension bolt forces. Do you think I can use explicit solver for this problem because the matrix doesn’t converge? If yes, how accurate my results will be?
Thanks again and I look forward to hearing from you soon
Dear, @Adel Badiee
If your load is static there is no meaning to run simulation by Explicit scheme. Error is in your modelling - most probably in interaction/contact definition.
If you are using contact pair algorithm, I suggest to use - Slave adjustment - Adjust only to remove overclosure option.
However u need to take care same element size in contact surfaces. Also take care geometrical errors (contact surfaces should not interact each other).
I suggest to go with interaction videos of playlist -
th-cam.com/video/HattYmp94Rk/w-d-xo.html
th-cam.com/video/1yIbNUEOANw/w-d-xo.html
Another video - Common Errors/mistakes in interaction ill upload soon.
This is the best video I`ve seen on explicit analysis!!
But shouldnt there be a D_{n+1} term on the right hans side of the implisite scheme on the slide at 5:00?
The implisit scheme looks to be depending on terms from timestep n only, not {n+1}. Wouldn`t this make the scheme explicit by definition?
Dear,
1. D_{n+1} is unknown which need to be find and depending on terms as mentioned.
2. Implicit depends non n+1, n, n-1 and subsequent terms. Look general equation carefully. (1st equation)
However, I told a general criteria to distinguish implicit & explicit scheme. Mathematical things are not straight forward.
@@arsitech-ltd Thanks for answer :)
Dear sir
first of all thanks for such detailed video. I have a question regarding the same. If i have two surfaces in contact and my model is subjected to quasi static cyclic loading ( specially in ADAS Dampers) which method will be more accurate, as it is having contact+ dynamic + time increment+ slow loading
Dear,
Go with all videos of playlist "Abaqus - dynamic analysis". You will get your answer. If still in doubt..let me know.
Thank you Sir for your Reply.
When I run explicit solver of Abaqus it aborted and got this error "THIS KEYWORD IS NOT AVAILABLE IN Abaqus/Explicit" do you know what shall I do to solve this error
Dear, Make sure all input you defined are valid for explicit step.
@@arsitech-ltd Ok but how can I check the valid inputs for explicit step.
Dear,
.msg file will show you which keyword not available/invalid for explicit step.
Few common invalid input are -
- 2nd order hex (C3D20), quad (S8) and other few elements not valid for explicit step.
- XFEM/VCCT/Contour Integral not valid in explicit
- Few contact definition not valid
- Some material model also not available in Explicit
etc.
-
@@arsitech-ltd Ok but I modeled a concrete specimen with a hole using XFEM but the crack started from the precrack and is going till the hole but didn’t continue in the other side of the hole as a lot of cracks generated without any of them completely separated the specimen which doesn't make sense and I get a recommendation to use dynamic explicit solver. So, how can I model that in dynamic explicit solver as you mentioned that XFEM is not valid in dynamic explicit solver.
any good book or journal I can use as a reference for this good information, sir?
Dear,
You can follow the book "Concept and Application of FEA (Robert_D._Cook)" & NPTEL lectures.
thank you sir
Hi
I am unable to join your whatsapp group.
I am keen to learn the background of FEA. Please share your notes and learnings also.
Dear,
Whatsapp group is full (limited participants can join only)..so same telegram group is available..i suggest to join there. Also you can join our facebook community (link in discription).
I shared few important notes on Facebook group..apart from that all i have - 2 FEA books and Abaqus documentation.
Thank you so much for the video. Can you please let me know if you posted a vedio to explain the difference between General Static and Riks Static?
Dear,
Till i not posted any such video. Anyways, Both are completely different approaches.
Thank you for this wonderful video. For a problem, where lateral load is applied on masonry wall with some precompression load, do we need to go for implicit ànalysis or explicit analysis?
Dear,
If inertia is not dominant & nature of load is not function of time..u can use general static step aslo.
Pre-compression load is not matter at all.
good explanation
Keep watching.
Sir can u pls put one video for defining excitation frequency
Dear,
For the same, I suggest you to go through video - "quasi static analysis in FEA"
Available at my channel SimTech05.
th-cam.com/video/2P0cuAmjIc8/w-d-xo.html
Sir waiting for your next video...plz
I was out of station of training program. Will upload next videos in this week.
Send the Telegram group???
Dear,
Plz find link in video description.
Dear sir..kindly to clarify this..so in summary the Static: {f}=[K]{x}, Quasi Static:{f}=[C]{x'}+[K]{x}, Dynamic:{f}=[M]{x"}+[C]{x'}+[K]{x}..Am I right?
Dear,
Quasi-static and dynamic analysis uses same governing equations. Inertia in Quasi-static is not dominant but we can not exclude mass completely.
@@arsitech-ltd Ok.. it's clear now..thanks again sir
Great and very helpful video.
Thankyou very much.
Could you suggest a book related to such fem important things which can help to understand to simulate by writing code or improve simulation.
Thankyou in advance.
Please keep making such kind of helpful videos.
Dear,
I suggest to read book - Concept and Application of FEA (Robert_D._Cook) for modarate and expert people in FEA.
And
Practical Finite Element Analysis - for moderate and beginners in FEA.
Thank you for that detailed explanation. Can you make video more on material used in dyna . most common interview question that I'm facing.
Thank you again :)
Dear,
Thanks for appreciation.
I'm just aware to LSdyna. Not expert. However, in future videos I'll talk about material models (common for all CAE).
I think claiming that the implicit schemes are more accurate than the explicit scheme is wrong. It is true that the force equilibrium is guarenteed when using explicit schemes, but this doesn't guarentee accuracy in dynamic problems. One could throw a ball, and calculate the force equilibrium in the ball at t = t_n, but the ball can still be at the wrong distance. This is because a Implicit Time Integration scheme still generates an error. There is a difference between the implicit system [K]{u} = {f} and implicit time integration scheme. Both contain the name implicit but it is not the same. [K]{u} = {f} can be solved exactly (or almost exact when using Newton raphson method if K depends on u). But implicit time integration schemes do not provide exact solutions in most cases.
Please feel free to challange my claim. I'm a master student who just finished a few months of reasearch in the FEM field, so I am definetly no expert.
Dear,
This is all about accuracy of operators used in scheme. Many researchers claimed that Modified beta Newmark operator more accurately predict the n+1 term if alpha and beta constant of scheme choosen wisely. However, for complex contact conditions and large deformation implicit may fail many time. But fail to converge solution is different thing and accuracy is different thing.
On other hand Central difference operator is straight forward.. without iteration everytime direct solution of governing equation possible.. that's why for large analysis it required double precision...as single precision may have more truncation error.
And
Here I'm for knowledge sharing and knowledge gaining. Neither I'm accepting your challenge nore denying reply.
thank you sir.
Thank you sir for making tutorial about such important topic. I liked it and understood it to some extent as I have a prior understanding of complicated mathematics. However, I would like to pen down the feedback that kindly make videos covering such important topics (like Implicit vs Explicit analysis) with the explanation of its physical significance . It will be helpful to those people who are new to FEA world or for the people who came to learn about implicit and explicit analysis for the first time. Otherwise It is very easy to copy and paste the equation from books and dictate it in video. Please, No offence! I found many difficulties in finding the helpful content during my initial learning of FEA and I still do so that's why making this suggestion🙏🙏
Dear,
Always welcome and thank for appreciating work.
Whatever i explained about implicit and explicit is only surface level.. I know most of users face problem in the same... so for me biggest challenge was - explain things in easiest way...not the mathematics behind it.
Sometimes i feel i not even properly described implicit and explicit.
I'll try to record one session of my advance batch and will update.
@@arsitech-ltd Thank you sir🙏
Always welcome dear.
Thanks
Welcome
Great video, but please consider removing the background music for your future videos. It feels unnecessary and distracting from your explanation. Thanks.
Dear,
I already stopped adding background music. However, few are old videos with background music. So plz adjust.
@@arsitech-ltd Thanks for letting me know.
Thank you sir