Видео

Hi Saher, you only need Python to perform Machine Learning. Abaqus is called within Python as well. So, I assume your issue is with how to use Machine Learning in Python. Once you figure that out, adding Abaqus is simple. Let me know if this is not clear and I'll try to come up with an example.

of course. You can pretty much model anything in ABAQUS ;)

Hi Saher. I suggest you comment the part of matplotlib. I've seen errors with that in the past that should be associated with the Python version you have installed in your PC conflicting with the Python version within Abaqus (This is just my theory, I never tested this).

@ Well, at first it showed me the errors in the Abaqus CAE list, now I can detect them through the command. I will try again can I have the python code you created please?

the final script is in my GitHub. You can find the link in the description of the video

It depends but the first thing I would suggest is to comment the part that runs the model. Then it should be fairly quick and you can run it manually to see why it is taking so long or if it is crashing

Well, 99% of the times you can remove any reaction. The solution method would be different but the end result would be the same. Now, in this problem, you need to remove one of the horizontal ones otherwise your structure becomes a mechanism and can't be solve (try to remove a vertical reaction here and you will see). Good luck ;)

Not really, if you are happy with the model, keep on going. More often than not, we do not really know what we want our model to do until we make the model and realize we need to change it. That is the point of this video, to show how to edit certain things without having to start from scratch. Good luck

That's very kind of you. Good luck with your studies :)

I'm happy it helped ;)

My pleasure ;)

Thanks Ed, coming up with analogies is the fun part. I'm glad you enjoyed it.

I'm happy to know this was helpful :)

My pleasure, good luck with your modelling :)

Hi. Please reach out to me on LinkedIn. We can schedule a call if needed. www.linkedin.com/in/luisantos090/

Awesome to hear :)

Thanks Alan, I'm glad to hear

Happy to hear that! Good luck with your studies :)

That's very kind Amer, I'm happy to know you enjoyed the videos.

Thank you, you're the legend! I'm happy it helped :)

I'm happy to hear that :) Good luck with your studies 💪

Thank you so much. You can connect with me on LinkedIn at www.linkedin.com/in/luisantos090/

Hi Mustafa. That's very kind of you to say, I'm very glad you enjoyed the videos

Muito obrigado José :) Fico muito feliz por saber que ajudou

Thank you for your kind comment. I never done Particle Swarm before but I'll have a look, I always wanted to learn it.

Thank you Mustafa. It should be a matter of days now :)

Hi Sylwester. As far as I know, load combinations are only available on the payed version of Ftool. Please confirm that on their website. Now, I guess you could do envelopes by exporting the moments to excel and then use a Max() and Min() functions

Hi Hasan, You can contact me on LinkedIn if you'd like: www.linkedin.com/in/luisantos090

​@@luisantos090 Hi Luis, Thank you. there is a problem that i could not send message for you. I want to get temperature, y-coordinate and pressure for some nodes. As usual when i get node number in the visualisation we see the node label. But the data for each node on their node list index number. what is your solution for this problem? actually node list index number and node label are different.

@Hasan H try to make the index equal to the node number minus 1. So, for example, if you want to know the output for node 2367, the index should be 2366. Give it a try and let me know if it doesn't work

Thank you Pankaj. I'm working now on the next episodes, focusing on optimisation using genetic algorithms :)

Thank you, that made my day. I'm really happy it helped :)

Hi Leonidas. At what minute is that triangle * rectangle? In any case the answer to your question is in this table. My students get a similar table on their notes. You can find it here. I hope it helps learnaboutstructures.com/Virtual-Work-for-Beams

@@luisantos090 Hola hermano , mira la verdad necesitamos ayuda en estos temas pq estamos estudiando para un examen. no c si podrías ayudarnos

Mohrs Integrals table :)

Hi Aziz. Thank you for your comment, that means a lot. There is indeed such command, I'll send it to you in 30 mins or so ;)

@@luisantos090 Thank you very much for your quick response, and for the effort. What I want is to save the maximum value of the Von Mises voltage for each run, either in an excel table or any other file, or just return those values to me via Abaqus command box.

@@azizseddiki6310 To save the data on an excel file, I use the following script (you will need to adapt it a bit but it should work: opFile = NewFolder+'/'+'DatainExcel.csv' try: opFileU = open(opFile,'w') opFileU.write("%10s,%10s "%('Force', 'U2') ) except IOError: PrintToScreen('cannot open opFILE') exit(0) for i in range(NrOfSteps): displacement = odb.steps['Step-1'].frames[i].fieldOutputs['U'].values[23].data[1] force = RefPointOutput[i][1] Displacements_midspan.append(-displacement) Applied_forces.append(-force/1000.0) opFileU.write("%10f,%10f " % (force, displacement)) opFileU.close()

@@azizseddiki6310 To get maximum values for displacements (you will need to dig to find the equivalent for von Mises voltage) I use the following code: Disps = odb.steps['Step-1'].frames[i].fieldOutputs['U'].getSubset(position = NODAL).bulkDataBlocks[0].data followed by: np.max(Disps,axis=0)[0] This will give me the maximum in the x-direction. You can do "[1]" or "[2]" for the y- and z-directions, respectively. Also, for this to work you need, at the top of your file, to "import numpy as np". I hope this helps :)

@@luisantos090 Thank you very much for your efforts to find a solution. I will try your instructions and give you feedback since I have tested it. I appreciate the time spent to solve my doubt. Thank you.

You're welcome Phap. Thanks for your message

Thank you Marwan, that's very kind. I can recommend other tutorials here: ruclips.net/video/yQRIOYYJFeA/видео.html ruclips.net/video/GE8-fSnwEUo/видео.html

Hi wehoo. Thanks for your kind comments. I am not planning to extend these topics much further but, for the future, which topics you would like to see covered? You can send me the list and I'll see what I can do :)

Thank you Nima, I hope it's useful to you

@@luisantos090 Thank you, it was. After this, I hope you make a video showing how to use neural networks and genetic algorithms.

@@nimapaslar8402 I will definetely show the genetic algorithms soon. The neural networks I'm still hesitant because I'm not sure how to make it work yet

You're welcome Amroune, that's very kind of you :)

Thank you Shervin. If you have any recommendations, let me know. I feel the post processing was the hardest part to demonstrate but I'm happy you enjoyed it

@@luisantos090 Definitely yes. But the procedure you mentioned in the video is really helpful. As some suggestions, it would be more than great if you extend the application of this procedure to some more expertized fields like the iteration in Eigen buckling analysis of a member under different shape factors or imperfections, XFEM, and propagation of cracks induced by low cycle fatigue recognized with python script. Anyway, thank you so much for distributing your knowledge for free!

@@shervinsafaeifaegh1540 thank you for that. I will show the buckling analysis and the imperfections in future videos. Regarding cracks and XFEM, it's not my field but I hope you can solve it nonetheless

Hi Daniel. Can you pinpoint in which second I made the mistake? There is a mistake in the video but it's then fixed. Let me know ;)

Hi Table, the 5 comes as the hypotenuse of a triangle. If you notice, bar AB has 4 m in the x-direction and then 3 m in the y-direction, which means the length of the bar AB is sqrt(4^2+3^2) = 5 m ... Then, the 4/5 is nothing more than the cosine of that angle. I hope that helps kakashi-sensei and feel free to ask anything else if you need

@@luisantos090 Oh i see very simple thank you very much, Kakashi appreciates it.