Пікірлер
@whcwcjecjecuecuw6654
@whcwcjecjecuecuw6654 10 күн бұрын
Afrer the topology optimization is completed it doesnt find the file? It suggests to combine results but that diesnt work either. What should I do?
@hr_3590
@hr_3590 11 күн бұрын
If executing in recent versions, replace : ax = fig.gca(projection='3d') by : ax = fig.add_subplot(111, projection='3d').
@ahmedaliyu430
@ahmedaliyu430 11 күн бұрын
Thanks for this quick tutorial. These are the sort of tutorials Altair need to have on their learning platform. Back to basics classic text book analysis.
@o.s.8068
@o.s.8068 19 күн бұрын
AI will impact FEA
@InnerVoiceofthepeople
@InnerVoiceofthepeople Ай бұрын
looking for more inspire simulation tutorials. so far everything is so great and also few simsolid examples
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
Glad you like it :)
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
For anybody wondering - I found the bug the next morning. Turns out i had the node id in the element.id variable (which is just a really really bad thing). It happened only for Hex elements as it has more than one line and i had the first cell of the line being the id - so for hex elements it took the node id from the 2nd line.
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
01:49:00 onwards is just a hopeless bug hunting session looking at the wrong part of the code. This definitely stresses the importance of tests - one or two lines of code would have caught that error days before this
@Anirban2050
@Anirban2050 Ай бұрын
(Are u developing in structural Analysis in finite element analysis) through python right!..
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
yes, adding to it. Not doing the fea in python but doing postprocessing of fea results.
@Anirban2050
@Anirban2050 Ай бұрын
@@ManuelRamsaier okay, this find of work i have explore yet . Thankyou very much doing live from Germany.
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
@@Anirban2050 sure. you're welcome
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
17:50 - short summary about the progress so far regarding the group detection algorithm 49:14 - How did we solve the problem of the many subgroups from the earlier livestream recap 01:00:00 - Pullrequesr for the graph-based detection algorithm 01:18:53 - Future Issues discussion 01:25:20 - Taking on the issue to add more data for the RBEs to the output (config, force, groups) 02:12:16 - Too many possitional agruments - making the linter happy
@abinalex714
@abinalex714 Ай бұрын
Defeaturing sometimes convert solid to shell or If sometimes I dont have solid but shell. i use volume tetramesh, supress many edges that are problematic then create 2d tria/quad mesh on the faces first and then volume mesh. To avoid tet collapse or some errors . This way I can control the surface mesh.
@abinalex714
@abinalex714 Ай бұрын
I really love this series.Thanks Manuel for doing this.
@abinalex714
@abinalex714 Ай бұрын
In the 156k model every slave node is disconnected, right because there are new nodes created in between. So the number of groups paths are like 2x slave nodes or something...
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
yes, but the connection detection over the whole graph (merge section of the connected comps) should take care of it. The 156k Model showed 2 groups - so that worked. What did not work was the forces.
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
Summary: 1:07:10
@DrSimulate
@DrSimulate Ай бұрын
This is a very nice video idea. Enjoyed watching!!
@abinalex714
@abinalex714 Ай бұрын
Hi Manuel, if you need a topic to cover. Can you cover about the stress linearization and by default how the stress are calculated for top and bottom section points for a shell and then how does one extract membrane stresses?
@anirbanmajumder4152
@anirbanmajumder4152 Ай бұрын
HI
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
Github: github.com/manuel1618/HyperworksTools
@gantovnik
@gantovnik Ай бұрын
Thank you for the stream. It is challenging to write a script in real-time. I am an engineer at Raytheon/Collins, but I also have a PhD (my thesis was about optimizing composites using genetic algorithms). As Ph.D. people, we always try to learn something new, right :) For me, it was tcl/tk in HyperMesh. It is LVL80 skill :) Only a few KZbinrs post examples of hypermesh coding: you, plus 3-4 other guys (from India and Vietnam). Of course, recently, Altair guys started to post interesting stuff on KZbin and Altair Exchange. Sometimes, we can read interesting posts in the Altair Forum. Thank you for your work. Yes, it is time-consuming and impossible to monetize it at all, but it is a great input to the development of aerospace and aviation and, eventually, human civilization. Keep producing great videos!
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
Thank you very much for the kind words. Yeah, I think you have to love it and be curious otherwise I would have given up long time ago. Speaking about tcl - this is really frustrating at times. The number of times I got help from tinh on the Forum I cannot count :-D But fun times...also here I think I will be going into tcl scripting agian as I have to make sure to use different properties for the comps...and that's still suboptimal. Maybe clustering, who knows. Thank you again and have a great time. If I can help in any way, you know where to find me :)
@abinalex714
@abinalex714 Ай бұрын
That bug was funny. The stream made me learn new things and was easy to follow.🎉
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
yeah :-D You never know what you find when coding.... Thanks for the compliment and thanks for being here and contributing.
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
Github Repo: github.com/manuel1618/HyperworksTools
@pradyumnchiwhane846
@pradyumnchiwhane846 Ай бұрын
Hi I want to improve the mid lines algorithm. I tried to make one tcl script but it going more critical for automation.
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
PR is appreciated. You can also detail your ideas here or by mail to: [email protected]
@noamaanshaikh5702
@noamaanshaikh5702 Ай бұрын
Hallo Herr Ramsaier, erstmal vielen Dank für ein tolles Video. Wenn ich richtig verstanden habe, gibts zwei Optionen mit wem man XFEM analysis machen kann, entweder Traction-separation cohesive behavior oder VCCT. Aber anhand der Abaqus Doku, VCCT kann keine Rissinitiierung von einer Oberfläche modellieren, die nicht bereits gerissen ist, daher muss ein Clearance zwischen zwei Oberflächen als Riss definiert werden. Können Sie vllt darüber einmal gehen und erläutern? VG
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
Hi, genau. XFEM ist ja eine andere Art der Analyse. Das VCCT Kritierium ist nur für das Risswachstum anzuwenden. Eine initale Schädigung muss vorliegen da diese (mit der dann vorhandenen Amplitude unter Last) gegen die Strain Energy Release Rate bilanziert wird und so direkt ermittelt wird, nach wie vielen Zyklen der Riss eine Elementlänge weiter gewachsen ist. Dann wird das wiederholt. Amplitude - SERR - Zyklen. Die xfem macht da mehr. Ich habe mich damit in dem Rahmen aber nicht beschäftigt. Der Prof. war kein Anhänger von xfem (eher das Gegenteil).
@shilpanayak3749
@shilpanayak3749 2 ай бұрын
Hi...im not able to put radius to a circle...it shows me in mm...how do i get the dimensions in radius or diameter for the circle???
@ManuelRamsaier
@ManuelRamsaier 2 ай бұрын
depending on the version you have - but check if there is a button called dimenions on the ribbon sketch. If you are in the sketch you can select the cirlce and define the diameter
@vs6x3
@vs6x3 3 ай бұрын
Thanks. The notebook I got was all wrong. Gave me wrong plot.
@WagaTouso
@WagaTouso 3 ай бұрын
Does this run locally on your pc or does it run in their servers? What are the specs of your pc?
@ManuelRamsaier
@ManuelRamsaier 3 ай бұрын
This runs locally on my pc. I have an AMD Ryzen CPU and decent amount of RAM (which is rather important for topology optimization): 128 GB But I see it run successfully on old laptops too. Speed is obviously different then.
@WagaTouso
@WagaTouso 3 ай бұрын
Thanks for replying. I have a ryzen 5 4600g with 32gb of ram, I think it should be enough for small models.
@ManuelRamsaier
@ManuelRamsaier 3 ай бұрын
​@@WagaTousodefinitely. You will be fine.
@TrueInsight
@TrueInsight 3 ай бұрын
Hey, there is also a way to visualize the reaction forces in the GUI of Inspire for the optistruct solver. You just need need to move the reaction slightly, like a remote load, you can do this in Inspire where after you have defined the support it gives a triad. After doing this you can click on the reaction force call out like you would for bolts or fasteners. If you are curious about this let me know, I can shoot you an image. I have gotten this question on support tickets a good bit. Thanks!
@ManuelRamsaier
@ManuelRamsaier 3 ай бұрын
oh absolutely! Could you send it please to [email protected] ? I will update that in the video then :)
@carlostomasmoralezama1218
@carlostomasmoralezama1218 2 ай бұрын
Good evening, could you explain in a video how you did it, thank you very much.
@wildphysicspianoboy97
@wildphysicspianoboy97 3 ай бұрын
This is gold, thank you very much
@ManuelRamsaier
@ManuelRamsaier 3 ай бұрын
@@wildphysicspianoboy97 you're welcome
@tnguyen3073
@tnguyen3073 3 ай бұрын
<3
@MOHITJOSHI-se1zf
@MOHITJOSHI-se1zf 4 ай бұрын
Would you explain what is different between Altair hyperworks and Altair inspire..
@ManuelRamsaier
@ManuelRamsaier 4 ай бұрын
Sure thing. In short, Inspire is targeted at the CAD engineer with stress calculation afinity. Its easier, well automated but limited in scope. Hyperworks has been renamed to Hypermesh and this is the full feature Preprocessor for the stress calc guy. Much more detailled, more functions etc. So, for a quick look at things, quick optimization: Inspire is a great tool. If you want to have detailled control and / or a more diffficult setup, then use Hypermesh.
@rabotokomoeketsi633
@rabotokomoeketsi633 4 ай бұрын
Thank you very much for this channel, by training I am a mechanical engineer working in the mining and mineral processing industry in South Africa. I have developed a great love for programming as well as thermodynamics and fluids after I worked on a project where the bearings on a ball mill were running hot and leading to failure. Having checked numerous drawings and machining tolerances we could not resolve the issue until the bearings were changed. No fault was detected in the manufacturing of those bearings though. I have thus decided to study in greater detail these two disciplines of fluids and heat transfer.
@ManuelRamsaier
@ManuelRamsaier 4 ай бұрын
Not the easiest topics to study but I bet it will pay off. If you have any questions, shoot me a message. All the best!
@OktoPutsch
@OktoPutsch 4 ай бұрын
very helpful advice, thank you very much ! it now sounds obvious but it never came to my mind how such little load changes may require more tweaks to cover a wider use of cases.
@ManuelRamsaier
@ManuelRamsaier 4 ай бұрын
Glad it helped! Thanks for the kind words!
@ramananj5079
@ramananj5079 4 ай бұрын
hmm.. RMS is a time duration of that particular g , so g * time
@maheshbk1975
@maheshbk1975 4 ай бұрын
Can you help me with a script to find list of node numbers in a model?
@ManuelRamsaier
@ManuelRamsaier 4 ай бұрын
Sure, can you be a bit more detailled in what you want to achieve?
@AbuIshaq1441
@AbuIshaq1441 4 ай бұрын
Mich würde interessieren, wie bist du fachlich zurechtgekommen. Also warst du als Hochschulabsolvent in der Lage sofort die Forschugn in Angriff zu nehmen oder musstest du nochmal einige Grundlagen aufholen ?
@ManuelRamsaier
@ManuelRamsaier 4 ай бұрын
Das war keine Schwierigkeit. Du musst den Willen und den Biss haben, dich in eine bislang unbekannte (oder zu wenig bekannte) Thematik einzuarbeiten und dranzubleiben. Ob du da an der FH oder an der Uni studiert hast, wird keinen großen Unterschied machen. Methodische Vorgehensweisen bezüglich der Forschung kann man sich leicht erarbeiten. Die Umsetzung ist da der deutlich größere Aufwand.
@AbuIshaq1441
@AbuIshaq1441 4 ай бұрын
@@ManuelRamsaier Danke für die informative Antwort. Was mich interessieren würde,war ob du auch Einarbtungszeite bekommen hast dich beispielsweise in Inhalte einzulesen oder auch nochmal Übungsaufgaben zu machen um nochmal alles zu vertiefen ich frage das alles, weil ich vor dem exakt selben Dilemma stehe und mir nicht sicher bin was auf mich zukommt.
@ManuelRamsaier
@ManuelRamsaier 4 ай бұрын
@@AbuIshaq1441 In meiner Erfahrung ist es so, dass du dir deine Zeit bei der Promotion frei einteilen kannst. Das Problem hier ist eher, dass die Promotion und das Projekt (von welchem du bezahlt wirst) nicht immer den gleichen Inhalt hat und du dann die Promotionsthemen quasi in deiner Freizeit machst, nachdem das Projekt bereits bedient wurde. Du kannst bei der Promotion dich so viel einarbeiten, wie du möchtest. Da gibt es in meiner Erfahrung keiner, der das dir verweigert. Wichtig ist eher: Es gibt nicht immer jemanden, der dich anspornt, das fertig zu machen. Die Promotion ist deine Sache und du musst das so vorantreiben, dass am Ende eine fertige Arbeit entsteht. Es ist ein Reifungsprozess auch gegenüber deinen Betreuern, der in der Abnabelung endet. Wenn du eine fixe Deadline für die Promotion hast (bspw. Industriepromotion) ist es nicht selten der Fall, dass man das nach der Zeit noch fertig macht (neben dem Beruf) und erst später abgibt. Zeit ist hier nicht das Problem. Dranzubleiben, den Biss zu haben, das ist die Schwierigkeit vielmehr.
@Rnrnr12367
@Rnrnr12367 2 ай бұрын
Ich bin jetzt seit 3 monaten dabei und kann dir garantieren, dass du dir sehr viel selbst beibringen werden musst. Du bist fachlich nach dem Studium sehr weit vom aktuellen Stand der Forschung entfernt. Die Grundlagen wirst du wiederholen müssen, aber sehr viel mehr Stoff aufbauend darauf verstehen müssen.
@sanjaip.s8429
@sanjaip.s8429 5 ай бұрын
Is there any solution for this problem than choosing non colinear indepedent nodes. My Indeoendent nodes are in X direction colinear. I tried giving dependent DOF 123 only. Still it showed the same error.
@AndreaScipioni-e3c
@AndreaScipioni-e3c 5 ай бұрын
sir do you give private lessons ?
@AccessDeniedYT
@AccessDeniedYT 5 ай бұрын
How come you're able to edit the properties of e.g. the curve you first created? Whenever I try to follow a tutorial that opens with the import of a .fem-file, all solver cards for properties, materials, as well as all additional 'things' I create in the model browser appear as 'ungrouped and are not editable. I cant seem to find a solution for this. I'm using HyperMesh 2023.1 with OptiStruct for the record
@AccessDeniedYT
@AccessDeniedYT 5 ай бұрын
Fixed it... Had to reload a different solver interface other than OptiStruct within HyperMesh and then reload the OptiStruct interface. Can't believe this took me 2 days to figure out.
@ManuelRamsaier
@ManuelRamsaier 5 ай бұрын
@@AccessDeniedYT We've all been there ;)
@VanHauTran5524
@VanHauTran5524 5 ай бұрын
Thanks very much. That so usefull
@ManuelRamsaier
@ManuelRamsaier 5 ай бұрын
Happy to help!
@kittiesletsgo
@kittiesletsgo 5 ай бұрын
Not related but: Can anyone help me what is wrong with this script? Tried to import it in Hyperworks, won't work. Heheh # Procedure to calculate the average element size proc calculateElementSize {} { set minSize [expr {1.0e20}] set maxSize [expr {-1.0e20}] set totalSize 0.0 set numElements 0 # Get all nodes in the model set allNodes [ hm_getmark nodes 1 ] # Loop through all nodes foreach node $allNodes { # Get connected elements for each node set connectedElements [ hm_getentityvalue nodes $node connectedelements ] # Loop through connected elements foreach element $connectedElements { # Get characteristic length of the element set size [ hm_getentityvalue elems $element characteristiclength ] # Update total size and number of elements set totalSize [ expr {$totalSize + $size} ] incr numElements # Update minimum and maximum sizes if {$size < $minSize} { set minSize $size } if {$size > $maxSize} { set maxSize $size } } } # Calculate average size set averageSize [ expr {$totalSize / $numElements} ] # Display results puts "Average Element Size: $averageSize" puts "Minimum Element Size: $minSize" puts "Maximum Element Size: $maxSize" } # Execute the procedure to calculate element size calculateElementSize
@Indonesia01ian
@Indonesia01ian 5 ай бұрын
Hi I am sorry for out of topic Did you know why the Hypermesh is so heavy application I have Hypermesh 2023 suite in my PC (Ryzen 5600X, 16Gb RAM, RTX 3060 12Gb). My computer lagging so much even if only I import geometry The geometry is a simple small 1Part (motorcycle shock absorber head) Actually there is nothing suspicious in the Windows Task manager (CPU 10~20%, RAM 50%) I don't know why this can possibly happen Is there anything that I can do to make the application lighter?
@ManuelRamsaier
@ManuelRamsaier 5 ай бұрын
Hey :) Can you describe a bit better what you mean with lagging? A delay when importing the geometry is completely normal. Even small geometric parts can be heavy to import as for example the number of faces which Hyperworks has to facet can be large even in small parts. If you want I can test it on my machine if you send me the geometry file to [email protected] If the lagging is related to turning of the model, make sure you have the latest gpu drivers installed. Your machine is suited well for the job. I have a similar one as you do.
@Indonesia01ian
@Indonesia01ian 5 ай бұрын
@@ManuelRamsaier Hi I made a video about it kzbin.info/www/bejne/o2jbhK17f8ubiKMsi=sdqG4riFy4-m12Oe I don't really know exactly what happen. I already uninstall then install it again but the problem still occur Interestingly I installed in my laptop i5 gen 9, RAM 16Gb, RTX1650 (all spec is lower than my PC) The software run smoothly
@kittiesletsgo
@kittiesletsgo 5 ай бұрын
Is there a script to find the average element size of your model in Hyperworks?
@ManuelRamsaier
@ManuelRamsaier 5 ай бұрын
Great idea. I don't know of any scripd that does it but it would be easy to code.
@kittiesletsgo
@kittiesletsgo 5 ай бұрын
@@ManuelRamsaier cool, waiting for it to be here. If it helps, I hope it can output like the criteria summary with it altogether. Hahahah
@kittiesletsgo
@kittiesletsgo 5 ай бұрын
@@ManuelRamsaier if you have time, may I perhaps ask for a script on this one? 😅
@ManuelRamsaier
@ManuelRamsaier 5 ай бұрын
@@kittiesletsgo unfortunately I am not really sure if I can do that in the current timeframe. I can give you a hint how I would approach it though: You have your input file (like a .fem file) and in there you have the elements, and the nodes. So I would go with a short python script which calculates the element size per element and then maybe the average per component ?! I mean its a pretty good defined scripting task and I would love to tackle it but right now I am busy with other things. But you can send me the input file to [email protected] and I may take a short look at it and see how complicated it is.
@ManuelRamsaier
@ManuelRamsaier 5 ай бұрын
Or the short version: Generate quads / trias (faces) and then use this command: 2021.help.altair.com/2021/hwdesktop/hwd/topics/reference/hm/hm_getaverageelemsize.htm
@soccergalsara
@soccergalsara 5 ай бұрын
I don't understand the point of 'b'
@ManuelRamsaier
@ManuelRamsaier 5 ай бұрын
It's just to make the equation easier to understand/write. It has no direct meaning but shortens the equation.
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
Get more details about the bending case: kzbin.info8KidL0ql9EU
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
Get Inspire Personal today: bit.ly/4aOKlfh
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
Get Inspire Personal today: bit.ly/4aOKlfh
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
Get Inspire Personal today: bit.ly/4aOKlfh
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
Get Inspire Personal today: bit.ly/4aOKlfh
@Anirban2050
@Anirban2050 Ай бұрын
Is it possible to do , structural Analysis Finite Element Analysis!
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
@@Anirban2050 absolutely!
@Anirban2050
@Anirban2050 Ай бұрын
@@ManuelRamsaier thank you sr
@ManuelRamsaier
@ManuelRamsaier Ай бұрын
@@Anirban2050 Sure, if I can help in any way, just leave me a comment :)
@Anirban2050
@Anirban2050 Ай бұрын
@@ManuelRamsaier sr i need software Altair hypermess.. student edition.. could not download it!!
@user-vn6rl4ep2z
@user-vn6rl4ep2z 6 ай бұрын
intresting short video👍👍👍. may i ask what is the software that you are using?
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
Sure, this is Inspire from Altair. You can get it from here for free: bit.ly/4aOKlfh
@omegadeepblue1407
@omegadeepblue1407 6 ай бұрын
Is only free for one year?
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
As far as I know you can reapply for another year. But I will check that once I talk to Altair next time and write it below.
@omegadeepblue1407
@omegadeepblue1407 6 ай бұрын
@@ManuelRamsaier Thanks Manuel
@dev4666
@dev4666 6 ай бұрын
Muss man für eine Promotion eine gewisse Begabung mitbringen oder reicht Interesse am Fach und gutes Durchhaltevermögen aus? Ich merke in meinem Bachelor Informatik Studium jeden Tag, dass ich hier nicht der Schlauste bin. Ich habe hier auch nicht die besten Noten. Trotzdem könnte ich mir vorstellen später mal zu promovieren.
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
Disziplin und Biss schlagen Intelligenz / Begabung auf lange Sicht. Es reicht dann evtl. nicht für eine bahnbrechende Promotion aber das ist halt alles sehr subjektiv und auch von anderen Umständen abhängig. Pauschal kann man das nicht beantworten. Aber man kann aussagen, dass der Mittelwert des IQs aller Promovierten kein "Genie-IQ" ist. Das haben auch durchschnittlich-Intelligente / Begabte mit Durchaltevermögen und Biss geschafft.
@kittiesletsgo
@kittiesletsgo 6 ай бұрын
is there a compilation of scripts or tpls that I can find? like even the simple one like naming and such?
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
There is an Altair Script Exchange which can be helpful for everything Hyperworks related. I have used it from time to time but not recently.
@kittiesletsgo
@kittiesletsgo 6 ай бұрын
@@ManuelRamsaier may I ask if where can I see it? are their tcls downloadable?
@ManuelRamsaier
@ManuelRamsaier 6 ай бұрын
@@kittiesletsgo Yes you can download them. You find this under: altair.service-now.com/community?id=altair_exchange (you have to login for this)