Hello
I have a case of stone that I can't process
Context:
4 parallel extruded profiles to form a floor that are considered to be welded together (addition of knots in SW along the length of the profile)
The floor makes in total: 2500x600 (thickness 50)
I was able to do quite easily the case of calculation where a load is evenly distributed on the surface of the profiles.
On the other hand, according to the applicable standard, I have to load the profiles with 300kg on a surface of 500mmx500mm centered on the profiles.
I can't model this case of calculation... Anyone have an idea?
I tried the mesh but it crashes all the time
Thank you in advance for your answers!
Hello
I don't have the model in front of me but you have to define a 500² area using a dividing line.
To do this, create a sketch with a 500mm square on each side (it's not the most difficult ;-) )
Create a dividing line
Insertion ==> Curve ==> dividing line.
Once this is done, you can use this bounded area to apply your static or dynamic load.
The advantage of the separation line is that its presence does not affect the simulation at all, unlike other techniques.
Here is a tutorial I made on the subject of dividing lines https://www.lynkoa.com/contenu/les-lignes-de-s%C3%A9paration-1
Second advantage if your load straddles two beams, it works properly. Be careful if your beams are separated to plan a large displacement with a distance limit. This allows depending on where you will move your sketch to have different deflections between two beams.
You should do several simulations by moving your separation sketch to safety, it will secure your results.
Kind regards
PS:
[Off-topic On]
I thought the standard was 500 kg per m² for metal frame flooring.
[Off-topic/Off]
1 Like
Good evening
I'll try to look at that tomorrow.
Kind regards
Hello
Thank you for your quick answer, The calculation works perfectly with the separation line system but only for mesh objects, when it comes to beams Simulation tells me that it is impossible to apply a localized force, whether it is a force or a mass at a distance.
At one point I succeeded by applying a force on points of my separation line, then dividing the applied force by the number of points, the results seemed consistent but after re-opening the Simulation model gives me another error when I didn't change anything:
Un ou plusieurs des chargements/déplacements imposés/contacts/entités associées au connecteur ne sont pas valides.
The mesh I don't even think about it, even if I mesh very coarsely during the calculation Simulation tells me that my PC is not powerful enough!
For the standard in fact it depends, there are 6 possible classes with different loads and each time 3 load cases:
-Evenly distributed load
-Load concentrated on an area of 500x500
-Concentrated load on an area of 200x200
The floor must have a deflection of less than 25mm under these stresses.
Hello
Is it possible for you to give indications on the sizes of the beams and especially their profile (IPN, UPN, etc...).
Please post us an image at least part of your assembly (attach a jpeg or in pictures format and not display on the forum because it is often illegible). You have to look at the appearance of the knots in your assembly (do they all have the same color)
I have an idea in the back of my head when I ask you this because for most edits, you should not use the beam mode but the volume mode.
In other words, you transform all your objects, beams, into volume and your life will be changed. To use the beam mode you need very strict conditions (no indentation on the edges, no light, frame butts and not mitres, etc...) and below the size of 200 x 200 the beam mode brings nothing, except PB.
Kind regards
Hello
I finally succeeded, it came from the dividing line, I had made a big rectangle of 500x500 that crossed the 4 profiles, then I applied my strength points.
My solution that works:
Make small squares over the entire area and distribute the force applied to these points (see Photo below:
https://ibb.co/hgncye
https://ibb.co/gHrRrz
In case you were wondering, there are nodes in the middle to simulate the fact that the profiles are welded together and therefore work together (I had some crazy results with the profiles considered as independent beams
I consider my problem solved, your answer was the right one, it was my application that was eroded!
Personal note => Do not run the dividing lines through the profiles!
For your information, there is another problem,
After saving the document, when it was reopened and after not changing anything, it was impossible to re-run the study!
...
I don't know what leads you to put knots in the middle, but for me if you accept this remark it's not the right way to proceed.
For the simulation, you have to get as close as possible to real life. And avoid the all-too-common mistake of imposing too many links on the model.
If I see the image before simulation which is the most interesting, it clearly shows that your profiles are not beams and should not be treated as such.
Because your profiles are welded together and not very worked, you should not use the shell mode either.
This means that everything must be switched to volume.
For the connection between the profiles, there are two ways, the good and the not so good.
SolidWorks allows you to put solder connectors, in this case ""solder bead""
The good one : If you put the weld seams, it also allows you to test if your seams are not too numerous, for example, or more annoying, insufficient.
The less good : You just have to make your four profiles together this will allow you to see the deflection to meet the standard. If you go through an office for the qualification, they will not retain the worst one, but will test the welds which can be the weak point if you are not careful.
I strongly suggest you adopt the right one because under heavy load, the welds sing (cf the Millau bridge with the welded caissons that sing when 35 tons pass over it.) The second reason is that if you ignore the soldering, then you might as well not connect the modules together.
Last point before you curse me definitively , in your case I would put the CG of the mass in the center of a cube of 500 on each side. This will allow you to take into account a possible spill or at least to check that there may not be any. I do remote charges at different heights to see the extreme case.
(I saw a machine go through a floor because the CG was very high while the mass of the machine was in line with that allowed by the floor)
So much for the friendly remarks and suggestions between professionals.
Kind regards
PS: for your non-repeatability problem it is prudent to always save the results of the simulation study and above all not to change the parts otherwise you cannot re-run the study without putting the right parameters back.
Thank you for this very complete answer, indeed I curse you, but it is above all my machine that will curse you when it has to calculate the deformation of the mesh!!
Indeed the raised CG I hadn't thought about it, it's quite coherent, especially since my two edge profiles don't have the same square moment as for the two central profiles.
In fact I made a shortcut by saying that the profiles are welded, they are in fact constrained between them in grooves, which makes them united in flexion.
=> In the beam mode this is the only solution I found to simulate this "link" between the profiles.
I'll have to work on my mesh to spare my machine, I'll keep you posted!
I'm thinking of trying to make 4 pieces with a constraint of non-penetration since there are no welds in reality @ follow ... In any case, thank you very much for your clear and complete answers!
1 Like