See also this link to understand the multitude of degrees of freedom via constraints on an assembly
and therefore minimizing it
http://forum.solidagora.com/techniques-assemblages-contraintes-t61.html
@+
2 Likes
So it's normal for the configuration... So the blocking bar is only used for viewing and opening, it's not very cool in the end! So I'm going to stay in the simplified one then^^
I have a PDM so my files locally. I don't use the toolbox. I use no assembly constraint, everything to a common origin. I only have external references on one part because I couldn't find anything else. It's quite heavy but the objective is to break everything at the end.
Surface repetitions I can't because I need the bodies for the mass.
Graphics Card: NVIDIA Quadro K2000
In any case, would making more lighter multibody guns be better than few multibody guns ?
1 Like
The fact is that
if you do + multibody and - assemblies
you will gain in constraints
so you will gain in speed of execution
The same goes for your constraints once the part, the body, the assembly is placed via the original constraints
you can remove them (the constraints) and fix these elements
Once the elements are fixed
No more constraints
so less calculation of degree of freedom
@+
I already have no constraints... Nowhere.
No, what I meant was: if I divide my pieces into several different rooms. That is to say, I make fewer bodies per piece. So it makes an assembly with more parts, but each part is less heavy than currently.
Does it struggle as much or would it gain in milling?
I already have no constraints... Nowhere.
Even if all your pieces are not constrained as you say, there are at least dimensions to position these pieces
otherwise explain to me how it stays in place
No, what I meant was: if I divide my pieces into several different rooms. That is to say, I make fewer bodies per piece. So it makes an assembly with more parts, but each part is less heavy than currently.
For me a room is a room with its own ref and material
so no multibody especially for large assemblies
Does it struggle as much or would it gain in milling?
try to do a multibody tone test against
an assembly with fixed parts where the constraints have been set
to obtain the correct position of the parts and remove them afterwards
@+
In fact you didn't understand my working method I think.
To put it simply: we are not allowed to stress in the assemblies and no more than 3 levels in the head assembly (so ASSEMBLY/Subassembly/Part).
So to avoid constraints, I place all my parts according to the same origin (via plans), and to avoid endless subassemblies I do multibodies. For example, a set of tubes welded on a plate: instead of making each tube and the plate individually to assemble them afterwards, I do everything in the same part in welded construction => I have multibody. I have the mass, the material, and the name of each body that appears in a nomenclature.
So from this method, my assemblies only have inserted parts fixed (so 0 constraints). And for the construction of the parts, there are sketches and functions in a classic creative tree with plans defined according to the general origin.
So I'm not going to change my working method, I can't. On the other hand, if I make parts with less body and therefore simpler parts, SW will necessarily calculate faster. I would have more parts in my assembly but they will all be easier to calculate, therefore faster. That's the theory I'm going to put into practice and I'll tell you if it works.
Thank you^^
If you place your parts via planes, you use a skeleton room in which your planes are stoked, right?
2 Likes
No, I'm not racking my brain. I recreate the same shots in each new room. I don't have 50 so it's fast. If we highlight them, it gives me a check in the assembly in case I messed up in the construction.
The skeleton part I find that it is not useful in my case. But I've used it to make mechanically welded structures (such as scaffolding, safety barriers), it's practical, everything follows when you change something, but it creates external references (that's forbidden for me).
thank you Ok so if I understood correctly
You position your plans via the general origin of your assembly
What level is this origin at in your assembly?
@+
Hello
Personally, I think that more pieces with less body will be faster.
I would also suggest that in your simplified configurations, for your parts, you merge as many bodies as possible, except for parts that would generate much more surfaces, for grating drawn blade by blade for example.
Having already traced this type of large assembly, good luck!! And patience...
Kind regards
3 Likes
For real large assemblies, there are several possible working methods...
You use the common origin technique (that's good)
Your need doesn't seem to be to make movements, but rather to "present things in space".
By seeing your 3D work, and what the needs will be in detail or overall MEP (drawing), it would allow you to further refine the choice of method in relation to your needs.
Note, given the number of pieces, the light mode is interesting, but when in sectional view, the cut files are resolved.
But otherwise, we can already talk about 2 methods that don't use "special functions" (speedpack, or other...)
The first would be to work in "Large Design Management" mode itself with two types of work:
- either by loading into memory only the areas worked
- or by updating the graphs of the desired areas.
Disadvantage in graphical mode, the sectional views will have the system color, and the dimension measurements will not be perfect (rounding problem).
(if necessary, I would use these methods)
The second would be to work only in useful areas.
Example: if it's a train with several wagons, we're going to work in the area of wagon 2, so the other areas won't be displayed or stored.
Example: if it's a circular set , we're going to work on the 0° to 30° slice, so the other areas won't be displayed or stored.
To do this, you need to create configs in the parent ASM, and set the affected parts in a deleted state:
- Default (=all)
- Zone1 (Work Zone 1)
- Zone 2 (Work Zone 2)
-etc...
If necessary, do the same in the assemblies of the level below.
Note on the PC, the graphics card is too small!! And watch out for the RAM.
Basically, for a CAD PC, unless you do processor calculation, you have to put good components in:
The graphics card (not a small one below Quadro 2000, nor a medium Quadro 2000, but rather a large Quadro 4000 style or higher)
The hard disk (SSD, and if possible work locally)
Windows in 64 bit
The RAM difrait 32 or even 64 GB (check the system once the ASM opens, how much is the RAM)
Then the processor (an average is usually enough, which allows you to put the price in the components higher)
The gain in performance of the graphics card will be beneficial every day, several times a day.
1 Like
there is also the working method of the sub-assemblies by "Model" (ASM saved in PRT, or passed in defeature which generates a PRT)
which still allows us to take into account the masses...!
1 Like
I don't think that transforming your assemblies into a part is a good solution because you will have a problem with your MEPs for the identification of parts unless you work with sub-assemblies for the final MEP and make a MEP plan for each sub-assembly and if you do this I advise you to look at my tutorial because it keeps the links between the created part and the assembly where it comes from
http://www.lynkoa.com/tutos/assemblages-complexes/cr%C3%A9er-une-pi%C3%A8ce-a-partir-dun-assemblage-pour-facilter-les-simulations
1 Like
I don't do MEP so it's one less problem.
As for the working method, I can't change it, it is imposed by the constraints of the BE. I already use simplified configs. I work in the parts, never in the assembly.
My problem is:
- the time to rebuild the multi-coprs parts (when I have to modify them) => sometimes 15-20 min, and sometimes the software crashes outright
- the time it takes to update/rebuild the assembly even if the blocking bar is activated in all parts
It's clear that my PC is not powerful enough, that's not a secret at the BE but there's nothing we can do about it, we won't change gear. So I try to optimize as best I can.
1 Like
As said aupavant avoid the multibody especially
if you have to make changes to it
then for the equipment we work with that we have................... can do otherwise
it gives you time to drink a coffee, smoke a cigarette ;-)
@+ ;-)
For the problem of reconstruction,
The "large design management (graphic/loading by area)" method saves time.
The "config zone" method does this too.
Otherwise, more generally, whatever the method,
you have to switch to the "rebuild/auto check" mode
to "rebuild/check disabled" and handle this manually (for this it changes options), in the PRTs, or the ASMs wanted only. And above all, forget the ctrl+Q, preferred Ctrl+B.
For my part, whatever the size of the ASM, I have gotten into the habit of working by managing the reconstructions, because it saves a lot of time, and to go faster. (and I find Solidworks more stable that way).
1 Like