Thread 993073

CAD thread? CAD thread.

>>993073 (OP)
which CAD software should I invest my time in?

>>993073 (OP)
Plasccity trial expired can't use it anymore, I feel sad

I actually did some stuff in FreeCAD with zero CAD knowledge. I hit a lot of bugs but it did work.

>>993112
How could you be so sure they were bugs if you entirely lacked any prior knowledge?

>>993073 (OP)
For me, it's Autodesk Inventor with 30% extra session crashing.

>>993237
>shit crashes
>shit explodes (constraint solver finding a solution that's not a solution)

>>993073 (OP)
Tried 123design it's shit

>>993102
Once you trial it that's it, developers made s tier security and it's uncrackable

Freecad is a weird combination of totally usable and frustratingly limiting. I am able to design most of what I need (fairly simple stuff though) but as soon as I try to do something in an order it doesn't like, the whole thing crashes. Will be nice in 5-10 years I think.

>solidworks simulation
>solution goes to 80% in 10 minutes
>sits there for the next 5 hours going back and forth between 3 and 17 gigs of ram
>"Study failed"
Every problem involving contact interactions has been a pain so far. I'm not even mad anymore.

cadurday edition

>>993073 (OP)
>>993287
nothing is uncrackable, he just paid crackers not to crack it

As an outsider I HATE 3d software.
Why do I need to install some arcane software whos GUI is build by fucking brain damaged people to do one fucking thing?
In this case engraved text on a curved surface.

Why can't these fuckwits make even importing models intuitive?
I made this in autocad btw and want text on the curved recess.
I'm attempting to use inventor but this shit is maddening.

>>993097
fucking none. learn a trade instead because shit is built by fucking twisted retards.
Autocad is ok

>want to insert a file into current document?
Hit OPEN. WTF fucking zoomer devs you total Cumbrians
>imported asset
>puts it in an asset tab with no visible way of importing into main workspace
I will murder on sight any 3D software dev I meet
Meltie over for now

I hate the antichrist

>>993587
I've managed to make a few things with freecad with it being the only cad software I've used. But it seema like I can never continue working on a saved project once it has been closed and reopened, nothing ever works again. I have to treat it like an 8bit videogame and complete it in one sitting.

It is incredibly satisfying when you actually get that cumbersome hunk of shit to make something slick though.

I feel bad for people who don't use Catia

>>993097
for me rhino, expensive as hell but easy to get around if you willing to renew the 3 month demo version with a different email each time lmao

>>993287
>>993745
I can crack it if you pay me.

>>994424
how much do you want?

>>993705
I deal with SW Simulation as a job, and I hate it with a passion, I tried asking management to get an Ansys license to no avail because it costs too much and there's not enough demand.
Going back to SW Simulation, first off the progress bars mean nothing, seriously, they barely qualify as vague suggestions. Secondly if your simulation runs for 5 hours and then fails you've got a stupid fine mesh while the analysis isn't properly modeled. Try with a VERY coarse mesh first, see if it runs, see if the results make sense; then, when you know it runs, use a finer mesh or, better, add mesh controls on the surfaces and edges you want to see in more detail.
If it fails, well I got a whole spreadsheet of errors and failures, full of possible causes and possible solutions, so I can't help you there without more informations.
To be clear, solve time doesn't mean the simulation is fucked, I had simulation run for 25 hours (it was an entire electric motor for an e-bike) and produce good results, but it all comes down to what you're analyzing and how you set up the simulation.

I'm going to use Blender to do CAD work again because learning another software is a pain

>>993792
Texturing with a displacement map would be way easier. You don't really see this type of process in real life machining because it's extremely complicated to perform.

>>994663
Not sure, it's an electron app LMAO.

>>994671
I need to do this to because converting files between programs is making me ultra-violent.
How hard is it to align objects? This is the most basic function I need speaking as the autoCAD psycho above.
It’s so unintuitive but I just want to zero in any shape I make and move by integer on an axis. That’s it.

Now every time I add a bolt connector in sw simulation the document units switch to meters. What the actual fuck. Every default I'm aware of is set to mm.

>>994667
In my experience coarse mesh fails the study more often than not, or at least yields questionable results. Fine mesh not so much aside from when solution takes so much time that I get tired of waiting and cancel it myself.
In that particular case the issue was probably parts slipping against each other under load (it was actually the point of study to see if they will), but instead of prompting about large displacement like usual, the solver decided to bug out for some reason.

>>995022
>every time I add a bolt connector in sw simulation the document units switch to meters
That's a new one.
>coarse mesh fails the study more often than not
Not to backpedal, but I meant "coarse" in a relative way, which is always dependant on the problem at hand. In any way, a failure after 5 hours is usually a symptom of bad modeling and mesh issues. Ideally you want that failure to happen much faster so you can troubleshoot your mesh (check for high aspect ratio elements, negative jacobian, shit base geometries, etc) more efficiently.
>parts slipping
Conctact analysis is a bitch in SW, getting it to behave requires esoteric knowledge not even Dassault has. Be mindful about the sleazy trick SW plays on us every time you mess with the model and then go back to the simulation: it may automatically switch the contact between faces you had previously defined to a self-contact, which usually doesn't work when you have two different bodies. I have wasted DAYS of machine time on this shit, fucking hell. Also, if you're studying slipping I assume you set a proper friction coefficient, but keep in mind SW doesn't like very high coefficients, and on the other side of the spectrum it won't go below 0,05.

Also, what's your usual node count? My workstation at the office is a somewhat recent Xeon with 128GB of ECC memory and a proper nvme drive, and I try staying below 3~4 million nodes at most (an analysis like that takes from 10 to 25 hours), and below 1 million if I can help it, with running times in the order of 4~8 hours; less than 500 thousand if I want something to shit out results within the day; if the problem allows it, I can use less than 200 thousand nodes if I want it to finish in a matter of minutes.

>>994663
$120, the code is compiled using bytenode, it's hard to decompile and the file seems to have an extra layer of obfuscation on the top.

>>995037
>a symptom of bad modeling and mesh issues
For one, that model had lots of unnecessary details that now that I think about it could be simplified or cut out altogether.

>it may automatically switch the contact between faces you had previously defined to a self-contact
Ah so that's what it was. I had a really cursed study a while ago where either some contacts would fall through each other for no obvious reason, or the study would fail due to "numerical difficulties". That's what probably was happening. Good to know.

>keep in mind SW doesn't like very high coefficients
The real coefficient would be at least 0.3, and the max 0.2 sw takes without complaints should've been just barely enough in theory (under the assumption of uniform contact pressure - which it isn't, hence the study) so the idea was that if 0.2 will work then the real thing will work even better.

>Also, what's your usual node count?
I don't really pay attention to that. I usually just tune mesh size by feel depending on how much time I'm willing to spend on it. Solution time can vary wildly depending on number of contacts and how well the model behaves after all. The study right now with 750k nodes, with every contact that have to watch out enabled just finished in 4.5 hours. With only essential contacts that move by design it takes 1h 40 min. With everything bonded it finishes in under 7 minutes. I'm running it on i5-12400, in a VM with 24 Gb of ram.
When I'm just fiddling with stuff I usually try to keep the solution time under half an hour, and only run longer studies when I want to verify the final design that I'm reasonably sure I won't have to fix and run again.

any fusion chads here? also i write websites in php and use an iphone. feels good to get paid and watch you nerds squeal over tools lmao

>>993705
>>994667
Have you tried Catia V5? I have an old R20 somewhere.
The V5 is the standard for Airbus and our aviation business overall.
SW is honorable but it's not as powerful.

>>995164
How do i contact you?

>>995885
>>994667 here. I work in a studio, we have all the softwares so whatever our client has, we can work with it too. I've seen CATIA, but never used it. UI is shit, and that's pretty much all I can say about it. As for the "powerful", I think it all comes down to handling large assemblies with millions of components and integration with a robust PDM that might set it apart from everything else. As for actually modeling in it, I don't know, but one of the good thing about SW is that being so widespread there's always something online that will help you learn or solve a situation, while the same can't be said about CATIA. In the end all CAD work pretty much the same (more or less). Just today I was reading SW2025 highlights and my impression is that DS is doing basically the absolute bare minimum to keep all these softwares afloat, because SW works like absolute ass, but the features and fixes are so few and far in between to be laughable. Also, unless you work in some massive corporation that does massive projects it's infinitely more likely for you to encounter SW than CATIA.

>>995886
What os?

Plasticity user here. How can I start learning to make AAA quality firearms? I know modeling in general fairly well but guns have been quite difficult for me in CAD software. Any good pirated courses? What guns should I create as a beginner?

Is there a CAD package that
>is cheapish (max $500/yr, inb4 pirate)
>has parameters that integrate with Google Sheets/Excel
>has text parameters that can be imported to a model/2d geometry
>has g-code export (not a need, but a nice to have)

I want to do a little sign business out of my garage. Ideally I would have products that are customizable without having to go in and change the model with each iteration.

I want to be able to put Basic Girl's(tm) last name/family motto/marriage date w/e into a spreadsheet and have a router table shit out

"(Basic Girl Name) and (Basic Guy Name) married (year) at (stupid town they live in)"

or w/e.

I use Inventor at work, string parameters are abhorrent. I have to do a work around to import text parameters from excel. It's also more than I need for mostly 2d signs.

Let's talk CAD!

for me, it's IronCAD. The tri-ball is GOAT

>>998446
alibre, probably.

>>998615
>>998446

Just mucked around with it a bit - they don't do the integration I need.

Currently I have to use an iPart in inventor, and then edit the iPart table in Excel since you cannot populate the regular parameter table with strings from Excel.

I am sure this functionality exists somewhere - how many part configs have different things engraved in them? Surely they don't expect engineers to manually change the engraving for each different part iteration.
wan00

>>994890
>How hard is it to align objects?
It's possible, the features are there, but it's very tedious to do in default Blender without addons

>>998636
yeah I think you have to script it in python
https://www.alibre.com/3d-cad-scripting/

>>998446
Inkscape is easy enough to script. The gcodetools plugin might suitable.

>>993783
The absolute worst take possible.
>>998446
Rhino has UserText on every object that you can export as CSV.
In fact I can recommend Rhino to everyone in this thread. One-time payment, no bs subscriptions, not gonna change, whatever you can't do in Rhino, you can do with the Grasshopper plugin, whatever you can't do with Grasshopper, you can do with someone else's extensions to it or write your own.

.t currently creating a factory control system using it, from 3d models to CAM auto-programming, part tracking, management and customer relations.

>>994379
I think this method doesn't work anymore.

>>1000881
>.t currently creating a factory control system using it, from 3d models to CAM auto-programming, part tracking, management and customer relations.
Very based. What kind of parts will be manufactured?

>>1000881
>>1001040
Buy an ad

>>1001059
>unironically thinking an entrepreneur is going to do his self-promotion on /3/ where 99% of posters here are learning to make 3D tits and anuses on Blender.
I think we are only two on this thread to use Rhino professionally and he seems clearly better than me so I'm curious what could be done with it.
inb4. phoneposting

There was a thread on /g/ a few weeks ago that said Freecad 1.0 was actually pretty good now. I just tested it. It's still shit.

>>993073 (OP)
does anyone have good tuts on manually building surface for G2 tolerance?

i tried to get into that shit half a decade ago but between a shit work environment and my own incompetence i bounced off it. thinking about giving it another try, there are a lot of jobs in my area that pay well if i can wrap my head around it

to be clear, not B side or engineering, but for organic sheet surface, like auto industry standard tolerance

also whats the industry standard software these days? catia? we were using alias, how well does that knowledge translate?

>>1001094
go on /g/, apple shills are everywhere. The fact that you don't believe that a small startup wouldn't shill their products when a massive marketing agency like apple does confirms that you're a shill.

>>1001103
>G2 tolerance
Tolerances are a specification on a technical drawing, not something you bake in your 3D model. They're an indication for manufacturing.
>industry standard software
All of them. If you have to learn from scratch I'd say start with Solidworks which is widely used, easy to learn, and has you have an infinite amount of online resources at your disposal to learn from and troubleshoot any issue you might have. Alternatively I can recommend Onshape (same company behind Creo), which is browser-based, free2play, and very very modern. In the end all parametric CAD softwares share the same core principles, just different UI/shortcuts/gestures.

>>996273
did you buy or pirate?

>>1001243
nice try, fed

freecad doesnt let you preview anything. you have to know the exact value ahead of time and put them in the left panel instead of a pop up box its also butt ugly. no smooth animation when switching angles. you have to dig through its terrible ux/ui just to change the grid values.
its not a decent alternative specially considering fusion 360 is free.
if you insist on using linyx exclusively then use ondhape atleast.
the only viable way to use freecad is by cutting primitives with the modifier.
dont get me wrong all cad software has clunky gimp tier ux ui but freecad is just the worst.
why cant i scale a segment or a shape in fusion 360 for example. why cant i fillet after using patch? why is the measuribg tool such shit when its so important.

>>998446
OpenSCAD ?

>>998446
theres a 3d printer plugin for fusion 360 not sure how good it is

I'm trying to make a quick thing and need to put a uniform indent all the way around on the selected face. Currently using Freecad, how would I do this? The closest thing I could find just clicking around would be the thickness tool, but it makes the indent too deep. Please be patient, am retard.

>>993073 (OP)
Is FreeCad usable now?

>indian YouTube tutorial
>Dude just doing an exercise from his school
>Don't use any shortcut
>Leave sketch unconstrained
>You can hear farm animals in the background

>>1002227
You may have already gotten your answer in the other FreeCAD thread, but here's an example of one way of doing it. You can define your profile in a sketch, then do a Revolution. In my image, I only did the top half of the profile, so I could mirror it later. You could define whatever crazy profile you need, or just a simple channel.

>>1002338
Thank you! This helped a lot.

>>993706
>grid on
amateur

>>1001630
It's a cool thing but gets really slow for more complicated designs. Also, making fillets is pain. I wish there was an OpenSCAD style software that did not have these limitations.

>>993097
OpenScad

is something like pic possible in freecad? Wondering if I should invest my time setting up Windows again if its not possible with freecad.

>>1001103
Do you mean G2 surface continuity? I learnt to surface by fucking around in Solidworks and Rhino and looking at online videos, tutorials and what used to be readily available presentations from Solidworks World. There was also Rhino stuff available but I haven't looked in ages.
I've used Solidworks for about 25 years so biased towards that - I have my own license - and it probably has the greatest userbase. DS are twats, though. Rhino is good but has a slightly different approach which might be familiar if you have used AutoCAD (old versions anyway). Surface building in Rhino is closer to alias than SW. Fusion is actually alright if you aren't doing anything really complex. Can't get on with Freecad but haven't tried 1.0 yet.
Solidworks is used a lot in manufacturing and product design (not always the best tool but ubiquitous) and Rhino in product design, furniture and architecture.

>>1002873
Sure why not?
>modeling threads
retarded

>>1002873
yes, but it would be painful, I found positioning and interfacing of parts in freecad to be abhorrent.

I've heard that CAD models are the best for 3D printers, but I'm looking to make something with a pretty organic shape and have been told Blender would work way better for that, and then I've heard Blender models aren't great for 3D printing.
Is this accurate?
Is there a specific software that would be ideal for something that needs to be printable but needs smooth curves that would be hard to model with splines or extra planes?

>>1003138
>I've heard that CAD models are the best for 3D printers, but I'm looking to make something with a pretty organic shape and have been told Blender would work way better for that, and then I've heard Blender models aren't great for 3D printing.

Blender is fine for 3d printing just make sure your mesh has enough detail. Organic shapes might be better off with a resin printer over a FDM style.

>>1003138
>mechanical part
CAD
>artsy piece of shit toy
Blender

Simple as.
>but but but blender can be used for engineering!!!!
No, shut up, it can't.

>>1003142
I'm doing both, so I'm guessing I should do the mechanical part in CAD and the artsy part in Blender and design them such that the artsy part can be connected with the mechanical parts with glued pins or snaps or screws or something? Can Blender do the precise measurements necessary to align fastening points?

I'm torn right now between learning Blender or trying to find Solidworks tutorials for more organic shapes. I suspect the time commitment is similar, which would give Solidworks the win because then I could test how the shell and the mechanisms articulate together in an assembly instead of printing both out and relying on trial and error.

>>1003159
You can import/export meshes into either and cut one part from the other, and cut screw holes from both. The inner part has to be a bit (say 0.4mm) smaller. Blender has an offset solid tool missing from freecad, though for a basic solid freecad's thickness tool can work too. Neither is very controlled so maybe you have to model the hole separately.

>>1003159
If you have parametric needs in your organic shapes, Power Surface is an add-on for Solidworks that expands its capabilities for free-form modeling. But I don't know how easy it is to pirate, unlike SW which is piss easy. You're probably better off doing as >>1003201 said though.

>>1003139
>and then I've heard Blender models aren't great for 3D printing
I think that's mostly because modelers turn on smooth shading, which does not work for 3D printing, you need geometry

>>1003201
>>1003202
Thanks for the tips! That information is immensely helpful. I had no idea about the Power Surface add-on or that you could transfer meshes between the two.

>>1003205
>smooth shading
Thanks for that. I didn't know it was a thing but now that I do I realize it was a staple in entire generations of video games.

>>993073 (OP)
I'm thinking about either doing Fusion, Freecad, or Plasdticity. I want to make game/animation assets but also 3D functional objects that work IRL. Whats the best one?

>>1003543
For assets go Blender, for 3D functional objects you have to be more specific, but any parametric CAD will essentially work.

>>1003106
>>modeling threads
>retarded
Good way to out yourself as a fucking amateur. I bet your repertoire includes dozens of reddit posts saying that modeling ANY threads is bad, without context.
Look at that image again: the tolerances on the thread diameter are so tight, and the parts so small, that a good engineer/draftsman would model them just to be sure.

>erm... ACKSHULLY I know every ANSI and ISO drill / thread diameter by heart
Fuck off nigger no you don't.

>>1003789
What the fuck are you talking about you retard those are completely normal threads, you write the size in the drawing and the machinist does them to spec and they just work.

Another great thing about not modeling threads is that usually the modelled hole is the correct size for predrill and the outside surface as well so you can easily add predrill size into the drawing as well.

>>1003807
Not other anon but modelling for rapid prototyping is common. Just did a load of pipe threads for MJF printed parts that worked very well. Did that to get all threads starting in the correct position and also wanted them working out of the bubblewrap without the user having to faff with taps and dies.
So no, can't always leave threads to the machinist.

>>1003830
3D printing is a whole different ball game.
And you actually end up having to manually model the threads which would be incredibly wasteful if it were a conventionally manufactured part.

>>1003861
Final part will be a moulding or a casting (with a few mods). The print represents almost exactly what the moulding will look like. The casting would be post-machined with threads added afterwards. There's a few arguments to be had over that.

>>1003862
Why not thread inserts?
If it large threads like a bottle cap threads then yeah sure mold 'em.

>>1003865
Internal threads yes of course where suitable (M5 mounting points) but not so much for external threads or watertight internal threads.

Can I use freecad to build something like a battleship?

>>1003889
What scale and level of detail?
Model / rendering purposes yes. Proper ship design uses specialist software or Catia, NX and lots of designers.

What new and interesting things have you discovered for 3d printing? For example I recently found 3dprintable.xyz. What else?

>>1003915
Buy an ad

>>1003543
Not freecad. Fusion is ok if you're willing to use cloudshit. I prefer plasticity though. And for functional 3d prints, I'd choose that over blender everyday.

>>993097
for 3D
Solidworks/Creo
for 2D
Autocad

But really, I dont know why would you just randomly learn CAD.

Just made my first pipe

>>1004011
Nice.

I got solidworks, and I feel like I can rule the world now.
Going to design a nuclear bomb after Im done with my steam engine

>>993073 (OP)
I want to be able to design stuff (or copy) to repair my headphones or other electronic appliances with 3d printing , is this the right place to ask? where should I start?

>>1004163
if only you knew how bad things really are.

>>1004171
Are you approaching this from scratch?
Short form might be:
1. Learn 3D CAD package (whichever you find easiest)
2. Learn about different types of 3D printer and materials (fdm or SLA if home machine, SLS etc if you send out to China or Shapeways etc)
3. Learn rules of designing for 3DP.
4. Take things apart (carefully), look at teardowns (ifixit) to see how things go together.
5. Go print some stuff and see what works, or doesn't. L
6. Iterate.

>>1004194
Should probably point out that learning CAD might take a bit of time to be proficient in if you are new to it.

>>1004194
yes I guess this should be what i should approach, thanks!
yes I think it would be from scratch, because I haven't found one single person who has some at least some experience on this field

>>1004200
any books, resources, threads in here , even plebbit that I should look into?

>>1004311
>>1004312
Will have a ponder. I have been involved in mech engineering and product design for 30+ years (formal instruction in both) with loads of learning on the job so not the best to suggest a starting point. However:
At a beginner's level, one of my kids is doing DT at school and there's some useful, simple design stuff in the exam revision guides you can buy from book stores.

CAD: FreeCAD and Solid Edge Community Ed are free (certain restrictions on SE). Also openSCAD, which I haven't used and Tinkercad
Fusion 360 and Onshape have or had free plans.
Solidworks has a makers edition which is apparently a PITA with embedded cloud shit. Shame as I use it at work and it is pretty good for what I do.

Some of the packages (Solidworks particularly) have okay built-in tutorials and there are loads of specific examples of functions on YouTube.
One good way of learning is to pick something you have lying around and just start modelling it. Begin with something simple and then increase difficulty. As a guide, blocky stuff like door hinges is easy, organic stuff like detergent bottles is hard. Moving on to take things apart for really in depth modelling will probably teach you more about how things go together than a bunch of books.

3DP: there used to be some best practice guides for FDM available from Stratasys, particularly "FDM for end use parts". Not sure if this has been wiped or stuck behind a registration form now. Simplify3d had a good guide to print settings.
Reddit's r/3dprinting "getting started" wiki has useful bits in it, though usual reddit caveats apply.
There's useful info at hubs dot com resources pages.

Phonefagging so formatting and speling may be a bit wonky.

Solidworks fluid simulator cost $14,000... Is there perhaps a cheaper way to do fluid dynamic simulations?

Just started using AutoCAD for the first time
Why does it seem like nothing has a hotkey? Do you genuinely have to type LAYER, press enter, then click a button in a GUI every time you want to make a new layer?

>>1004366
OpenFOAM. As the name implies, its free and open source.

holy shit, CAD is so based, like actually, really fucking based. I'm going to coom

Hello everyone, I am wondering how can I fill this in? I am wanting it to fill from surface circled in red upward, but keep the hole in the middle, the two diameters are slightly different in sizes, and wanted to I guess loft? from one to the other and fill inside the shape to help strengthen it. But I am not sure how to do this, this is in freecad 1.0.0.

>>1004782
If Z is the rotational symmetry, can you make the shape in the xy plane and extrude with a taper so it doesn't cross the other wall?

>>993782
>GUI is build by fucking brain damaged people
You've used Blender, I see.

>>1004373
no.

>>1004812
I actually ended up doing that 10 minutes after I made the comment, and felt dumb for asking.

>>1004373
I'm feeling like this too.
and how do I permanently change where the "base point" or "insertion point" of an object is?
I'm used to using turbocad where you just press D and click where you want it to be.
In autoCAD you can change where it is for one operation by pressing B and enter, but then it goes right back to where it was.
there is some CBPR command but you have to install it as a plug-in which is faggy and also it only works on blocks.
Seems like it is missing this very essential feature that I used every day at my old job.

>>1004819
i need to know what to do please. I want a one button hotkey like R to rotate not R-O-space then click a base point to start rotating.
turbocad for being really cheap did so much better of a job at this.
R45 and it's done. change the base point with D if you need to. man I miss that little cheapo software. it sucked at a lot of things though too.

Any resources for learning SolidWorks as a total beginner? I'm doing an internship at a furniture company and my supervisor told me that I should git gud at SW soon, or else he would break my wagie body.

>>1005357
SW' own tutorials are a good way to start, and then just look up on the jewtube for more tutorials, but beware of jeets. If you are stuck, try SW' Knowledge Base first (assuming those assholes at the company are actually paying for the license) and duckduckgo second: it's such a widespread software that chances are most of your issues have already been encountered and solved. As a last resort you can ask here, I'll answer when I can.
>internship
>they expect interns to just learn without tutorage, not even the basics
I'm sure they're the kind of company that only hires neo-graduates with 5 years of relevant experience, for minimum wage of course.

>>1005369
i miss when the SW forum was publicly viewable, found so many answers using it. probably why they locked it as i'm a filthy pirate. thankfully there's still the wayback machine ;)

>>1004782
Why not add a few break away supports? Tree type should do the trick.

>>1005369
I'd love to have a "no jeets" add-on for my browser where people can flag jeet tutorials and it hides the channel for everyone else with the add-on.

>>1005369
Thanks for the advice. I just started a couple days ago. There is a "200 SW exercises" book I found at anna's archive, just taking it easy with the first couple of models and fiddling around with the sketch and 3d tools.

I was going to watch the CAD/CAM by Mathlaban channel for guidance, but he's a jeet. Do you have any recs for non-jeet youtube channels or good exercise compilations? I think those in the book are okay, but idk how good that kind of practice translates into modelling components for furniture.

>>1005631
Any resource is as good as any if you're just starting out. I was half-joking about jeets: I hate how they speak, and that's on me, but the vast majority of their video tutorials won't really help you more than the stuff SW provides you with. That said, I came across good jeet channels, try watching a few videos and decide by yourself what you like best.
Keep in mind that once you grasp the core concepts of parametric 3D CAD (volume generation from 2D, volume addition and subtraction, the feature tree and its order of operation), it all boils down to knowing the advanced tools, keyboard shortcuts, and overall knowledge of the specific software peculiarities.

As for furniture, it depends: in principle you can do anything, but parametric software like SW aren't the best for flowy surfaces, like pic related. I mean: it can be done but it will be time consuming and painful. For basic shit like cabinets, shelves, and tables it's perfectly adequate.

>>1005638
Perfectly possible to do flowing, swoopy stuff. You do need to have a good grasp of modelling practices, when to use surfaces vs solids, when to model features directly rather than struggling to get something acceptable out of standard tools (eg fillets).
>>1004322 was me. Furniture has been about 10% of my work over the last 20 years.
Hard surfaces are relatively easy, squishy, irregular upholstery is a pain in the arse to model quickly, but I have learned to use Fusion (sculpt module) for those projects.

>>1005381
No that's just Dassault forcing all users on to their system. You will use 3DExperience and you will be happy.
Idiots alienated the most fanatical CAD user base out there.

>>993073 (OP)
Used to use Fusion360 but got caught designing a flame-thrower that connected off-the-shelf stock parts together and was banned and all my 250MB of files, includng my robotic hand and assault rifle, wiped.
Cunts and shit happens.
Anything FREE and like Fusion360 that isn't shit and amateurish crap since last I looked?

>>1005972
I need parametric that works, since I may need to alter the calibers, leade depth, shoulder angle on cartridges and other chamber specs PLUS I need shit that permits me to have master sketches and from those go to 3d modular design with ease like in Fusio360 but without american twats who will erase MY WORK AND MY DESIGNED as I need them for a patent application and the usa just ensured they won't be fucking getting the manufacturing package this millennia due to their cuntishness.
>I'd rather give it to China than sell it to filthy mericans
My rage is deep, I was raised to know forgiveness is for weaklings and I am not someone to fucking cross.

>>1003993
>But really, I dont know why would you just randomly learn CAD.
How else is an inventor supposed to try out whether their idea works in 3 dimensions then run the numbers on it in analysis?
Go get expensive jew-nigs to make one for $60k a pop on their 4-axis for shits and giggles?
Talk sense, kid.

During the lockdown I taught myself gunsmithing and studied the engineering history of fireams.
I have something new.
And no... it's not for the usa. You can spin on it. With any luck mericans will be being mowed down by it in the next war.
>got banned off Fusion360 a fews years ago and all my work ERASED
I have zero forgiveness... but I've moved on since then, refined the ideas and have something new so in a way it wasn't a total loss.
>only thing I lost was the flame-thrower I got banned for
>started as a simple brief I set myself...
>"it's rather hard to vandalize Confederate statues if you're on fire"
I just ran with the idea as an exercise. I can always re-do it. It was simple enough. Then the corporate pissflaps spotted it when I was doing stress analysis on the pump housing, somehow.
Got a permaban.
Shit on them.
THEN I found they erased all my designs.
You can imagine how I felt knowing they had simply torched MY WORK, MY DESIGNS, MY FUCKING WORK.
>pleased?
Oh, I'm pretty certain that wasn't it.
No smelly american touches my work.

>>1005976
Well... I'm back.
And I have something nice for thee.

>>1005631
holy shit. I legitimately screencapped this

>>1005631
How did you handle the corners?
I couldn't stay faithful to the dimensions without making them obtuse.
But I guess it's still correct because 90° wasn't specified in the drawing.

I see some extra faces on yours by the corner so I was just curious

i just had a dream that my cad part bumped up against some other guy's part during shipping and it may have changed the dimensions.
the other designer filled out some kind of incident form before I got a chance to, so then it was my duty to go into the part on my computer and fill out what's called an "updater schedule" to document any kind of deformations that could have happened and manufacturing considerations that may need to change due to it jostling up or whatever with another solid part.
or if the parts became fused how to get them apart.
can anyone with experience tell me if I'd really need to do this because of a shipping accident ?

>>1006519
Yeah, I don't think there's any issue with the degree of the corners. No idea of why it ended up looking like that. The sketch was done by mostly abusing the symmetry tool and kind of improving to fit the reference, and then extruding.

>>1005638
Thanks anon. I do see that SW isn't commonly used for furniture similar to your pic. Thankfully my company mostly does rectangular shite for banks and offices. Also I'm looking into Rhino 3d in case I need to do more curved stuff.

I do wonder where I can get some exercises for that. I got a pdf with some mock CSWA questions, but they are mostly focused towards mechanical parts, and not furniture components.

>>1007078
I got through the first 100 and found plenty of dimensions that were missing or just wrong.
I figure whoever made this was either being a little lazy or went out of their way to simulate the designer experience of being given poor information (which has happened to me many times in the professional world)

>>1004322
>3DP: there used to be some best practice guides for FDM available from Stratasys, particularly "FDM for end use parts".

This?
>https://www.stratasys.com/siteassets/3dlc/high-school/tag_fdm_enduseparts_en_1015_web.pdf

>>1007098
it's definitely on purpose to force you to look at the picture and reckon what the correct answer is yourself. you can't get wrong values when making such technical drawings out of a cad program unless you manually/deliberately override the text of the dimensions

>>993073 (OP)
I made my first complex model lads. What do you reckon?

>>1007513
Looks cool anon. I suppose you used an IRL helmet for the dimensions.
In which program did you model the helmet? Any good tutorials for the animation?

>>1007735
>animation
NTA, but many CAD softwares allow to configure an exploded model of your assembly, and to record an animation of course.

>>1007735
Thanks anon, didn’t use a helmet for sizing though did it from scratch. Took a bunch of unflattering pics of myself from various angles and then used them for proportion then measurements for sizing (plus 1-2 thin test prints). Fit turned out great.
I used fusion 360, which handily also has an explosion animator like >>1007743 mentioned, I think it’s called ‘animation environment’.
Result turned out pretty good too, pic rel

Which cad software is best for curved syrfaces like those of picrel? I want to try doing more organic forms now that I got some experience with the usual, rigid mechanical components in SW and inventor.

>>1008073
Plasticity, hands down. It's cheap too.

wait wait wait....i just downloaded F360 for the first time. are you telling me you cant actually customize the pan, orbit, and zoom hotkeys how you want them and instead just have to pick from one of 5 all equally janky fucking keysets bespoke to specific softwares?

why not just have customizable hotkeys?

i have to be missing something right? this program cant be that stupid 5 seconds into installing it

>>1008091
buy a 3d mouse
>shill
i bought a spacepilot second hand in 2013 and wouldn't want to use solidworks without it

>>1008120
>I'm a mech engineer/CAD monkey
>the studio in which I work doesn't have 3D mouse
>all the companies in which I end up doing consulting don't either
I know they're more expensive than the average piece of OEM crap they dare call a mouse and that you find everywhere, but it always strike me as odd now such a productivity enhancer is never implemented. I hate this place.

>>1008120
I dont need one, I just like to unify my controls between all the different softwares I use, nothing dived me nuts like every God damn 3d software having different controls and fucking up my muscle memory, but pretty much every other software I use has customizable hotkeys, I dont know why this one doesn't, it's such a basic feature, I mean ffs even Microsoft Word will let you customize your keys, it's not rocket science

>>1008160
I bet you can customise keys with a pro licence :^)

how is plasticity at handling large/complex files?

most of the tutorials i've seen have been pretty small or simple objects, but how well does this program work with a fully realized mechanical assembly with all sorts of parts, wires, springs, etc like a car or a machine?

does it run smooth or does it chug? my machine isnt a potato but it also isnt top of the line

>>1008331
>plasticity
>mechanical assembly

>>1008120
>>1008130
>unironically advocating for 3d mice
You suck at CAD.

A 3d mouse fucking sucks for anything beyond making a smooth pan/rotate to show your boss a cool model.
If your left hand EVER leaves your keyboard then you are slow and inefficient.
I don't care that your $400 accessory has a dozen bindable buttons.
My 5 dollar keyboard has 40+ buttons that can be reached by my left hand, multiplied by every ctrl,shift,alt modifier you can think of.
Every time you click a button on the GUI is time lost.

>>1007513
>jarvis look up how I can animate my camera smoothly and then tell me how to do it
anyway I do think the exploded model is cool

>>1008922
>anything beyond show your boss a cool model
You'd be surprised how important it is for non-technical people to see nice well presented things, especially when they do not understand them. I know because I learned it the hard way.
By the way you can /diy/ a space mouse for like $20, as long as you already have a 3D printer.

>>1008922
>I don't care that your $400 accessory has a dozen bindable buttons.
i spent the equivalent of USD50 on it 12 years ago, this isn't a matter of cost

Im not sure if this is the right place to ask because I haven't ever seen a request thread about 3d files, but i was wondering if any anon could help me modify this .stl file?
its a portal 2 floor button. I love this design /thing:3706013
but I wanted it to have a bottom covering it and 2 spaces for keyboard clickers like the model for this one /thing:5078900
Thank you to whoever helps!

>>1008130
No that's pretty standard, sadly. I freelance so bring a spacemouse if using my own kit but obviously not always possible.
I bought one of simple two-button ones in 2007 and it died about a year back, so great vfm for the use made of it. Now I have a wireless one that will hopefully see me to retirement.
The old (1990s) tennis ball mice were best to use - cost a fortune back then.
Not necessary but I notice the lack when faffing about in large assemblies or complex geom.

I'm still hanging out in the based cad thread. bump

Where can I find blueprints or exercises for SW sheet metal? I want to try modelling one of those old rigid metal cases for desktop computers, but I don't have any laying around for measuring.

>>993073 (OP)
Just made my first 3D print in TinkerCAD
:)

>>1009981
Google "sheet metal drawings for practice" and look through the image results. Plenty there.
Think I did as you were intending and modelled an old Dell Precision chassis for one of my first forays into sheet metal. It was built like a tank with some interesting clinch fastening (Tox?) going on that was a pain to sort in SW...

>>1010624
Slippery slope, anon. Your home will soon be festooned with plastic bits, some of which might be useful...

>>994379
Just pirate it bro

>>994379
imo Rhino's licensing model is (at least partially) holding back higher adoption. It's very surprising that McNeel hasn't yet offered a free or discounted personal/'maker' version of Rhino, whereas other companies (e.g., Autodesk) have such versions of their programs.
>figured that Plasticity answering the call from Blender's NURBS tools being dogshit for the past two decades would've been enough for old Bob to make his software more affordable, given the popularity that Plasticity generated

For fucks sake, is there really no way to duplicate a sketch in solidworks while keeping all its relations? I'm working out machine geometry and I need to have it sketched in multiple positions at the same time to figure out the rest of design so derived sketches are of no use. Blocks are useless either because they don't support external relations (to this day I have no idea what's their use case at all). Having to restore a dozen or so relations to master sketch dimensioned elements and hinge points each time just because goddamn frenchmen thought that under no circumstances you might have a reason to keep relations on copy-paste is such a huge pain in the ass.

>>1011603
copy pasting sketches keep the relations for me. the only thing you should lose is relations to the origin. unless i'm misunderstanding you?

>>1011619
to be clear, i'm hitting ctrl+c with a sketch feature selected, then ctrl+v with a target plane selected

>>1011603
I add construction line cross hairs, snap to external geom, that I dimension from. Copy sketch to new plane and then re-snap the crosshairs. A pain but doable...

>>1011603
You can use blocks as a quick and dirty way of copying geometry from edges converted from a ref model to another model without having to use external references. Example being a pcb hole pattern from an imported connector model, or a standard panel cut.
A other one is layout sketches. A pcb related example is placing boards in a product. Use blocks in sketches and mate pcb assemblies to block sketch elements. Easier and quicker to drag blocks in sketches than faff with mates...

>>1011625
>>1011624
>>1011619
Oh I missed a word, I mean external relations of course. My workflow right now is like this: I draw a master sketch where I dimension all the geometry (simplified using lines and triangles of course) and fix hinge points for movable parts in some eyeballed default position, then draw another sketch where I duplicate the moving parts and use external equal and coincident relations to the master sketch to dimension and fix them properly in hinges and then use some characteristic dimension (like suspension compression height in case of vehicle suspension) or just leave it underdefined to move it to a different position and see if it performs as expected and doesn't do anything weird. I make several of these sketches in different positions then I adjust the master model with movable sketches following the changes to dimensions and hinge geometry while keeping their respective different positions thanks to being related to the master sketch until I'm satisfied with it and can move on to drawing the actual physical parts. For something simple I actually draw all the relevant geometry within the same sketch several times in different positions and then with the right constraints SW automagically derives the proper geometry, but it becomes overcluttered really fast (pic related) and also there seems to be some sort of hardcoded calculation time or complexity limit where SW would just throw the towel and mark everything red as unsolvable even though you can still fiddle it in place manually (until the next tiny change or even just rebuild where it blows up again). So for more complex stuff several identical sketches all related to the main sketch, just with one or few dimensions adjusted later become a necessity.

>>1011652
>Comment too long
Derived sketches don't work for this because they're all identical to the original, you can't make individual changes to the copies; blocks are pretty much the same but even worse because they can't have external relations at all so the entire thing must be defined within the block and it's simply not feasible when it's not an already completed part you have to work around but a part of a larger model you're designing.

p.s. just had an idea as I was writing this, maybe you can use configurations to wrangle the derived sketches in several different positions at the same time? Will have to test this when I get to my machine.

>>1011652
>So for more complex stuff several identical sketches all related to the main sketch, just with one or few dimensions adjusted later become a necessity.
And forgot to add that all the external relations to the master sketch are lost on copy-paste so you have to restore them manually for each copy which makes it a great pain that has no reason to exist in the first place

>>1011653
>maybe you can use configurations to wrangle the derived sketches in several different positions at the same time
Ok tried it and it's a complete shitshow. 1. You can't have several sketches derived from different configurations of the same original sketch within the same part. They simply don't have a configure option. 2. You can't even do this in assembly. Every derived sketch follows the configuration the original part currently has. 3. You can do it in assembly if you copy-paste several copies of the same part that contains the master sketch, set the copies to different configurations, and derive sketches from each of them. They even properly follow changes to the master sketch since it's the same in every configuration. But it's a huge clusterfuck at this point and I'm not sure if it's usable in practice.

>>1011652
what are you actually trying to do?
i wouldn't call myself an expert in solidworks even though i've used it for many years (not formal training, just figuring it out as i go)
i have done a few things where i need different elements to line up in different positions, like a machine which can be operated at several angles, requiring pin holes to match up at each angle. initially i did this using a sketch, but now i do it using copies of the moving part(s) and setting them as envelopes. i can attach things like holes to said envelopes as external references without needing to draw messy sketches

>>1011692
-- also in general i've learned that solidworks doesn't particularly like "equal" and "tangent" relations, i've greatly reduced situation where large sketches suddenly go completely yellow by keeping those to a minimum

>>1011692
>what are you actually trying to do?
Not sure what to answer to that aside from what I've already said. I'm using sketches to derive the basic geometry of moving machinery, like vehicle suspension, a small excavator arm (on the pic above), or a laser beam positioning system I'm working on right now, and I have to make several copies of said sketched geometry in different positions to see that nothing weird is happening (like a laser beam missing a mirror in one of extreme positions) while I adjust the model to achieve desired results. I don't have professional SW training either so maybe there's a different way of doing that I'm just not aware of, but I'm yet to find evidence of any.

>>1011702
I get what you need, and I too use sketches that way sometimes, but Solidworks' Motion tab should allow you to calculate the various positions of a kinematic system, it might be straight up faster to just go back and forth between the two tabs than trying to wrangle SW into submission when complexity increases. Either that or use a different software.

>>993782
>>993792
CAD software is for engineers by engineers. I've gotten into it as software engineer and it hurts my brain since the approach is different, but same base concepts. [spoiler]Go for more geometrically simplistic designs and play with the software to get a better understanding of how to use it.[/spoiler]

>>1011707
I use motion too at later stages of the design process but it requires quite a bit of setup, and I sometimes don't even have a clear idea what the thing I'm making will end up looking like when I'm only starting with it. Throwing together a quick sketch in like 10 minutes is often easier and more productive. SW is the best tool I know for this, it's just this one thing with external relations that makes this unnecessarily tedious for no reason.

Anyone use progromatic CAD? I tried out CADQuery cause I'm already familiar with Python but then found out that it's basically been superseded by Build123D so I guess I should learn that instead? Or is there something else I should look into? Seems hard to find information about this niche.

is the fem workbench on freecad useless? Trying to compare some simple pressure load results and it sucks hard.

anyone happen to know how to get a 3d model out of a self-contained edrawings web viewer file? for example;
https://files.catbox.moe/5h39k9.html
closest i've gotten is there's a "modelContents" string which is 1.2MiB of base64, but the decoded file isn't in any format i recognise, and has no recognisable text in it either

Use astocad

>>1012864
>fiddy bucks for a UI
>of a FOSS
C'mon, for that price you can get Solidworks for Makers.

>>1012887
it's more of a donation system. looking at their faq, it's also foss, they just charge for the binaries. this is a perfectly legitimate way to monetise foss. you can build it yourself

>>1012889
>they just charge for the binaries
Not the first project that does this. Making builds for windows and mac is a PITA (as a programmer I want to kill myself ever time I have to touch windows or mac), so I'd approve of this model even if offering binary builds was the only thing they did.

>>1012890
yea, i was going to add that this isn't even that uncommon, but i figured that could be inferred. like for example you can in many cases find programs on the google play store or apple's equivalent where they charge for it even though they're completely foss. people are willing to pay for the convenience
that and of course building for windows is something i can understand wanting to be compensated for, and things like the apple store literally costs money to publish on

>>1012889
>you can build it yourself
I'm an engineer, not a /g/entooman, I have no idea how to do that. That said, from a purely ideological standpoint, if they keep it FOSS then I'm less against the asking price. Might give FreeCAD a try, I'm getting kinda pissed at OnShape's retarded way of handling way too many things (although some features are very neat), and if it's not completely trash I might even toss those guys my money for the privilege of being a mere user.

>>1012894
>/g/entooman
is it that obvious?
>if they keep it FOSS then I'm less against the asking price.
that is why i checked. i haven't heard of this one before.
>I have no idea how to do that
that's ok, that is expected as well, it's part of the incentive/convenience. if everyone knew how to do it to the extent that it didn't matter if they provided .exe installers or not, it wouldn't be much of an incentive.
keep in mind software freedom as far as the FSF is concerned is only about the availability and license of the source code, they aren't concerned with paywalling binaries (if you even provide them at all. many projects don't provide binaries whatsoever), as that's a service independent of the availability of the source code. not their business.

Fuck it, I'm just going to learn FreeCAD and try to stick with it. Not like I'm ever going to use CAD in my job anyway, so I mind as well learn it to 3D print shit

>>1012792
resorted to using renderdoc to extract the mesh data out of the browser itself, converting each mesh using a blender plugin to then export to something solidworks supports.
not very fun working with meshes in solidworks, but this is largely just for visual purposes and to trace around the cutout so it's not a big deal
very much thankfully the meshes show up in blender with their relative positions and even the correct scale, so it wasn't as much work as i had feared

Anyone who's interested in Free/AstoCad, this guy's channel is pretty based

https://www.youtube.com/@MangoJellySolutions/videos

Lol

I can not believe I got hired and fired in the span this thread has been up. bump

any autodesk fusion knowers here? i've got a display cutout in picrel that's fine as-is, but if i'd like to make the housing thicker i'd need to recalculate the depth. is there a good way to define a cut to go up to a certain distance from the opposite face?
can't figure out if i'm doing it wrong but it feels like fusion doesn't play very well with ref dimensions and it kinda seems like maybe an offset plane is my best bet?

>>1013085
8 months isn't that long or that short but the hell happened?

>>1013115
while not specific to fusion, perhaps as you hinted at, instead of cutting down from the side in your picture, make a reference plane offset from the other side and then cut the opposite direction

>>1002295
it's peak tutorial, honestly. you can see everything and know exactly how smart your co-workers will be.

>>1013125
yea figured,
guess i should stop being lazy since it's literally just one cut lol...

>>1013127
well unless fusion has an option to cut to an offset from a selected surface. i haven't used it so i don't know what options are available

I am working in Fusion. Some idiot 3d scanned a piece and ended up with a fucked up STL file with a messy mesh. Some portions like this rounded corner need to be cleaned and then rebuilt with some sketching and a loft.

How I do I erase these without damaging the surfaces nearby?

>>1013169
couldn't you just cut that corner off, re-extrude the corner in, and re-make the rounded corner. should be fine as long as you know the radius.

>>1013115
>8 months isn't that long or that short but the hell happened?

Not him but I had a similar situation happen to me recently. Got hired onto a contracting company and put on a contract with a large financial institution. Then the economic downturn happened and they decided not to renew the contract because of "budget constraints". My company kept me on the bench for 3 weeks, couldn't find a new contract to put me on, then let me go. Took me about a month to find a new job.

I find it hilarious that we're expected to give these fuckers 2 weeks notice but they can call us into a meeting with HR with 1 hour notice.

>>1013169
that looks really clean for something 3D-scanned, had 3d scanning technology really gotten this good?

>>993097
I've used onshape for free for a while with a throwaway email address, only tried Freecad (horrible, often can't calculate all constraints) and fusion360 (okay, didn't keep me hooked) and then recently onshape. Don't have to install anything either

FUCK

>>1001039
It still works

>>1013512
After all those months, I'm still using Blender
I have years of experience with Blender so it's easier than learning something new

>>1014169
blender isn't a cad program

>>1014173
blender is also a computer program aids your design process

>>1014260
you can use whatever software you like, but calling blender a cad program is like calling inkscape a cad program because it can draw lines and shapes. like yea, you *could* design something with it, eventually, but it's not meant for design. neither is blender, blender is a 3d modelling program, it's not intended to design real-world objects with

>>993073 (OP)
After reading through this thread and looking up the prices for each, as shit as it may be, I know the only possible option for me is freecad unless you make a shitload of money but if I start making money, I probably won't want to learn another tool from scratch

>>1014821
>learn another tool from scratch
there will be a lot of similarities, and depending on your future needs switching might be a good ROI.

>>1014821
You don't need to pay.

>>1014839
We will not take that route, chud.

>>1014260
blender maybe free but its not good at anything if you want a good modeling program use 3ds max is compatible with cat.

>>1014821
I used many CAD softwares and in the end they operate very similarly. Adapting to a new one is a matter of a couple of hours to get used to the new control scheme, you never truly start from scratch. If you don't want to have to deal with FreeCAD there's always OnShape.

I need to export a Plasticity file to Fusion, but STEP doesn't keep object named, is there a CAD format that keeps names when exporting?
Also the hierarchy in Fusion is a mess, why does every group have an origin, and if I delete an object in Fusion it deletes the whole group?

>>993097
Siemens NX

I am fond of version 10.0.3, last maintenance patch. Runs well in Qubes OS!

>>994033
>I have to treat it like an 8bit videogame and complete it in one sitting.
Not worth the download then. Sticking with NX...

>>994081
Doesn't CATIA use NX's Parasolid core?

Please help.
I'm trying to create an axial thrust bearing for a turntable.
I plan to print two races and the cage and use ordinary hard steel balls.
I want to create an assembly, not that it mattered rather to gain experience in the assembly workbench.
I'm looking to create the constraints as physically accurate as possible but maybe I'll have to accept some unnatural behaviour of the thing.
The first ring is grounded. I managed to constraint the first ball onto a vertex in the sketch I used for the groove. This is not right but works for now. The next constraint constraints the cage at one of its spherical holes to the ball. More balls can be constrained similarly to the cage. I have no idea how to get another ball to sit in the goove at another point without forcing it. I hope to find a way that will let me undo the unnatural constraint of the first ball to the sketches vertex. I'm okay with constraining to the groove face, groove edge, sketch.

>>1014980
Nope, CGM. Solidworks (also owned by Dassault) uses Parasolid, for which they pay a license, one of the reasons Dassault wants everyone on 3DExperience (CGM). Siemens (NX, Solid Edge, Parasolid) have also locked them out of the kernel's Synchronous tech.

>>1014978
Suemens NX works on linux?

Oh god solidworks is just bent on being a massive pain in the weirdest places
>running different part designs through simulation to find the best one
>try to use configurations to switch between them quickly
>turns out simulation studies are hard tied to the configuration that was active when they were created and the only way to run a study on a different configuration is to make a copy
>and if you need to make some changes to the study you have to either delete and make new copies or to manually edit every one of them making sure you won't miss anything then wonder why it went bonkers even if you have like a dozen
>ALL FOR YOUR CONVENIENCE OF COURSE
Can goddamn frenchies ever chill ffs

>>1016070
you can at least ctrl+click and drag all your connections/fixtures/etc into another study so there's no need to manually recreate them all

>>1016070
I've been dealing with SW simulation for work for quite some time now. You don't have to delete the copies, you just add a new configuration, do whatever you want as long as you don't delete previous functions (suppressing is fine tho), then copy your study using the new config, fix what broke, and now you can run the simulation. It's paramount you don't fuck with the other configurations or you lose the results of your previous simulations. It's bad, but could be worse. An alternative is making a Pack&Go which insulates your previous results by virtue of being a new different file. Just be aware you need to have a different name for everything, a prefix is sufficient, or opening both the original file and the copy will cause SW to shit itself and you run the risk of fucking the file you meant to not use.

>>1016077
You can just make a copy tied with another config with all that stuff already set, the problem is that it's an independent copy from now on and if you need to make some changes to the setup you have to edit every copy manually which is tedious and prone to error

>>1016081
In my case the forces/fixtures/connections are identical across all the configurations, it's only the stuff inbetween that's different. It's cool that I can have studies tied to configs, but holy shit just let me run the same study on multiple configs, how hard it can be. If shit breaks when I change between them that's my problem, no need to babysit me.

>2025
>can't open two different files with the same name simultaneously
Now this is just embarrassing

>>1016081
Oh and I don't care about saving results, I just write down what I need on a piece of paper after each run

>>1016099
>>1016100
The underlying issue is that Solidworks is an old CAD software that got things grafted to it over time. The simulation package itself is grafted on, the various capabilities of the simulator are grafted on. You can clearly see it by way of the UI: for instance there is a virtual wall contact boundary condition if you want to use one, but it's literally hidden and not in a place where one might think it should be. The solver itself shits the bed for the most stupid things, I could go on for hours. The sad part is that SW is probably the best one out there that integrates CAD with FEA in a single software. If you want an alternative, Inventor has its own FEA package, but I found it slightly lacking in some aspects compared to SW, but admittedly I haven't used it much.

>>993073 (OP)
Ah good ol' CAD. Good luck getting an AI to do this shit.

Is it possible to CAD something like this and then 3D print it into something wearable? I saw a guy CAD and 3D print a different suit of historical plate armor with very good results but it was substantially less detailed and somewhat industrial looking rather than handcrafted

>>1016482
you can certain draw up the pieces using cad software, though several pieces would need to be printed in sections and glued together. i'm not sure how well it would hold up unless you made the pieces quite thick. i'd be surprised if there wasn't some resources on how best to make large wearable 3d printed items

>>1016482
>3D print
Absolutely, but as >>1016483 said you definitely would have to divide the largest pieces, depending on the specific printer. You should look into propmaking techniques for 3D printing, I'm sure there's a wealth of knowledge out there on the subject.
>to CAD
That looks like pain. You definitely want to use a non-parametric software like Plasticity for objects like that, and you also should consider that whatever you design you factor in the 3D printing aspects like the minimun thickness of the object, printability (overhangs, plate contact area), and so on.
>substantially less detailed
For FDM printers layer thickness and nozzle size are what dictate the resolution of your print, but increasing resolution exponentially increase print time, so it's always a tradeoff. There are ways to get your cake and eat it too, but most likely is heavier post-processing on your prints than what you would normally have to do.

>>1016483
>>1016486
Thanks for suggestions. This is the other armor that I mentioned that inspired me and is completely made of PETG, sanded down, then painted over with a metallic finish, and held together with a combination of glue, leather, and rivets. It's cool although a little simple in design. So all the intricate detailing and fluting on the first armor I posted is what's really scaring me about the feasibility of CADing and printing it but you've guys have made me a bit more optimistic so I guess I'll try to work on my CAD skills and see what happens

>>1016490
this kind of model isn't exactly what i would call a beginner project in cad. like we're talking about reasonably complex surfaces and overlapping parts. not impossible by any means. i imagine if you're in the position to want to seriously try to make something like this however, you probably already have a good mental idea of how you'd draw this up.
certainly have a go, i've drawn some pretty complex machinery with no formal training either. i don't really touch surface modelling though. just don't expect to do it in a day

as a side note, if that photo is of 3d printed parts, that looks amazing, while it doesn't strike me as authentic antique armour, it doesn't look like printed plastic parts at all either

>>1016482
CADing this up in Solidworks or similar is possible but will take a while. Make sure you have some kind of reasonably accurate underlay / scanned mannequin / anthroprometric model to work over (not sure where you would find one).
>>1016490 looks easier but still complex if you haven't done any surfacing before. I have only played with Plasticity but think it could do the job. You might find Fusion's sculpt module better suited. I have done similar modelling in both Solidworks and Fusion and tweaking splines (lofts, boundary surfs, trims) can get a bit tedious. Sculpt bodies can be quicker to adjust but the quality of nurbs surfaces generated can be a bit variable, effecting downstream operations.

Have a go, let us know how you get on.

Looking for a free (preferably FOSS), offline-capable recommendation for GrapheneOS/Android. I love the UI and controls of the OnShape app, but it doesn't have an offline mode for ideation w/o internet. It's be nice if it could export all sketches, objects, and assemblies to OnShape and SolidWorks. Any ideas?

If I had followed the bandwagon and learned Grasshopper over a decade ago, I would've been able to more-easily rebuild a basic polyline into a decently bumpy and noisy coastline.
Instead, I resigned myself to my ignorance and gave myself carpal tunnel.

Is it hard to become a mechanical CAD modeler? I know Blender really well and have done mostly mechanical projects with it, but there's no jobs for that, even with Plasticity.
What's the best to get a job in that industry, Solidworks, Catia or Alias? And do you absolutely need to be an engineer to be good in that field?

>>1017919
Resident engineer working mostly as a CAD monkey here.
In no particular order:
>do you absolutely need to be an engineer to be good in that field?
No, being a CAD monkey is an entirely different job from being an engineer. Of course there's overlap and having some level of knowledge in the field helps, however most of it is codified, and even then each company has its own internal practices; in other words you'll learn with experience anyways. Most jobs as a CAD monkey end up being"take this existing machine and make it taller/shorter/larger/smaller/etc" and you will hardly calculate anything, so as long as you're not a drooling retard you'll be fine, but beware it's an alienating soul crushing job.

>the best
They're more or less all equivalent to each other and they operate very similarly, they just differ in UI and key bindings basically. In my experience the two main parametric CAD are Solidworks and Inventor, with probably Creo Parametric in third place, and then the likes of SolidEdge, Siemens NX, and CATIA, in this order. I mildly enjoy SolidEdge for its flexibility, I've recently warmed up to Inventor a lot, I'm in a long term hate/love relationship with Solidworks; I haven't tried the others yet.
If I had to suggest I'd say try OnShape because it is a pretty modern CAD (albeit rough around the edges) it's on cloud but free to use, and it's the easiest and fastest way to get your hands on a decent user friendly parametric CAD and practice. If you want to jump straight into a work-oriented CAD go for Solidworks.

>mechanical work
>in Blender or Plasticity
The sole idea gave me a stroke.
The main distinction between CAD like Solidworks and Plasticity is the parametric nature of the former: when creating shapes and assemblies you can specify relationships between objects and even tie each number and dimension to an equation. This allows you to tweak your 3D object by simply typing a different number and have everything update coherently (hopefully).

>>1013247
I am him and luckily I just got hired again.
looks like I'll be making missle housings in CAD.
boss rotate and then slam that shit into some poor bastards house, like an absolute boss.

>>1013250
aint no fuckin way that's the raw scan right? must've been cleaned up with meshmixer or something.

>>1005972
what the fuck how did they find out (assuming you aren't autistic enough to go around telling people). do they actively scan filenames or parameters?

>>1018168
There was something in the Ts&Cs a while back that said they reserved the right to delve through your data. I presumed that this was for AI training purposes but probably looking at what we're all up to. "Just in case".

Dassault are desperate to kick us all off SolidWorks and on to the cloud, doesn't take much of a leap to guess what they might be after...

>>1018181
fuck me backed up all my shit just in case, i gotta stop prefixing my variables with "nigger-" i guess.

>>1018181
jesus. once i finish up my shit i'm pirating an offline copy of solidworks/catia/nx

>>1018184
>>1018186
Consider the implications for those of us working on new tech. Spend years working on something to find it popping up in an AI generated result somewhere or be passed straight on to a competitor that has some relationship with the CAD vendor (ISTR rumours about the latter happening in aerospace 10-15 years back when cloud CAD started appearing).

>>1018188
i work in software and we'd never be caught dead having cloud vendor lock in like that. always thought professional CAD for those industries was always an on-premise company network kind of deal. any idea what the military uses?

>>1018189
Not sure. I think Catia/Enovia V6 /3DEXPERIENCE is available on-premises if you are someone the size of one of the big aerospace manufacturers, probably a prerequisite for defence industries. Otherwise you are using DS services.

Bruh wth blud

>>993073 (OP)
does REVIT count?

>>1005972
wtf
heh. I guess I'm not a paranoid schizo for not using cloud shit

Been using fusion 360 as babby's first CAD but i've seen way more quality of life stuff in solidworks. i hate the way fusion saves every fucking version (no idea why it can't be like git for software). can anyone comment on this? my main use case is modelling stuff for 3d mjf printing.

>>1018493
>MJF
>proprietary HP shit
The horror.

It all comes down to what kind of object you're modeling: if it's mechanical parts then Solidworks is pretty damn good, and it really shines when doing assemblies, but is really painful doing complex surfaces. Inventor is decent too, and it should be cheaper, but the assembly environment isn't as good as SW's; also if you shell out some extra money for Inventor you also get NASTRAN integration which as a solutor is as good as it gets; Solidworks too offers a FEA tool, but I hate it with a passion.

>>1018498
as opposed to the proprietary CAD software they're designing it on?

>>1013512
onshape gives me the shits because I use it to do the one thing I can't do in autocad and I forget how to do it every time I need it

Anyone know what type of mechanism I could use to keep these parts together?
I don't want to friction fit them.

>>1018512
What are they made of and what manufacturing technology is it going to be produced with? What are the functional requirements of the assembly?

Here's some options anyway:
>make a thread on the mating surfaces and screw them together if they're going to be made out of metal
>add a snap joint if they're made out of plastic
>add a bayonet mount

P.S. If you don't want a press fit you need at least a H7/g6 fit on that diameter.

>>1018513
Theyre made of plastic and an automatic manufactoring technology.

>>1018513
Cast silver
Maybe friction fit will be best
I also had this rough idea just now. leaves less room for error due to unpredictable cast shrinkage.

>>1018520
oh and basically I'd want them to snap in and never come out so that peg would have to be reshaped and to make sure it has enough flex to squeeze into the channel.

>>1018520
>>1018521
You made a snap joint right there. I'm not familiar with cast silver, but I'd advise against snap joints on metals as they usually can't elongate as much as polymers can, meaning the pegs would have to be very very long to flex without breaking. An interference fit would be the go to choice here, but Idk what level of precision silver casting has, meaning that to ensure the fit you may have to machine the mating surfaces. Look into shrink-fit too. All these solutions need to be engineered properly of course.

Another option is having some slight clearance and then drilling a hole and using a blind rivet to hold the two parts together.

>>998446
OpenSCAD. Free, pretty powerful, parametric, can take .json as input, and can be scripted by Python and other things.

>>1012357
I use OpenSCAD daily, it's great

>>1018368
>room not enclosed, area will not be calculated.
If those are generic walls im gonna fucking cry.

>>1018797
>>1018799
what can openscad do that fusion 360 with python api can't? every action that can be done within fusions ui can be written in python.

>>1018836
>what can openscad do that fusion 360 with python api can't?
Make you waste your time on typing walls of text to create what a few clicks will get you in a proper CAD made for human use.

>>1018902
yeah lmao honestly never saw the point. fusion 360 has integration with vscode+debugging so you can literally just record your gui clicks and translate them to code either way. maybe openscad can be used to generate models on the fly on headless servers.

>>1018905
It would've made great sense 40 years ago when computers didn't have enough power to handle real time 3d editing. Write the script in simple text editor then go take a break while it's rendering or does something else with the result. Sort of like LaTeX but for engineering. Too bad it came out 25 years too late to have any practical use and never managed to root itself in the industry or gather a religious nerd following like LaTeX did.

Has Plasticity been cracked yet? I often find myself needing to do some complex surfacing a few times a year sporadically so the license isn't really worth it for just my hobbies.

>>993073 (OP)
I want to learn CAD. Where can I learn and where can I download CAD?

>>993252
Kek

a
b

c

d

>>993097
>>998446
Just get FreeCAD. They've renovated the UI over the past year or two and it's actually bearable now. It even has OpenSCAD features. Just be aware of stuff like >>1003115.

Anyone else loves surfacing?
I feel like I'm ready to get a job at BYD

>>1019010
Only version 1.4.15. I check every once in a while.
Actually, there's a "perpetual trial" method on chinese sites, but they are locked behind payment. And it's still the trial, so the good stuff (xnurbs, square, etc) are not available.

If I had a little bit more talent, I would have purchased it.

>>1019583
yeah xnurbs is the shit i need. lmao imagine if BYD made all their stuff on pirated CAD.

DEATH TO PIRATES

(just kill me now senpai)

>>1019608
Plasticity is one of those software where pirating is just immoral, as their business model is not exploitative and the pricing is very affordable. Every other CAD has to die in a ditch of pain and misery though, pirating those software is unironically the moral thing to do.

>>1019621
I can respect not doing a subscription model, but $175 is steep to 3D print at home

>>1019621
I agree, however I'm a third world NEET with no hope of ever making this money back from the software.
I'll never financially recover from this purchase.

How do you guys make money from 3d?

>>1019623
Everything considered, I think it's a very reasonable price, especially because you can keep on using that software basically forever
>>1019624
>make money from 3D
I'm an engineer, but my 9 to 5 is being a good CAD monkey, I guess that counts. I hate 9 to 5 so much it's unreal.

>>1019624
>How do you guys make money from 3d?
I can't recommend it but I make around $200 a month from commissions

>>1019629
Maybe I should just make guns. How much are guncad bros making these days, I wonder?

>>1019636
>>1019637
become a plumber and make 15x more from what you make in le 3d lol

>>1019583
how much does xnurbs really contribute to great class a surfacing?

>>1020281
A lot. Not just xnurbs, but square too. It's basically impossible to get class A without the studio version. And even without this requirement, just modeling anything beyond the most basic lofts will take a very long time in the indie version compared to studio.

How do you accurately measure curves? I mostly do 3d printing and decided to print some radius gauges cause why not, while they have helped me when there's an actual radius curve, but theres times where its odd shape curve and seems to change radius size from beginning to end, I have tried get diameter right before the curve, and at the end and total height.

What methods have yo uused ot make it easier for calipers to do accurate measureents, or what other tools do you use ot help take measurements other than calipers?

Some stuff I have experimented with was take picture of the area I want to measure with a ruler both parallel as possible and in freeCAD use whatever I googled to help me change measurements to be equal to my ruler and traced the shape that way.

Another method I am trying is using air dry clay to press fit a part and hopefully make it easier to measure the shapes.

>>1020295
Doesnt fusion360 have free xnurbs and square in it already? Someone told me its useless to spend 300 dollars cus fusion free edition can do all of it. Is it worth buying studio license over going to fusion?

>>993097
Blender

>>1020409
> CAD
> floats as coordinates
Your plane gonna crash.

>>1020359
Sculpt, which was t-splines. Fusion is pretty good but its version of a feature tree is not easy to read if you have a complex model and it saves to the cloud.

>>1020465
Whats the difference between tsplines and xnurbs, is one exceeding the other in capabilities?

>>1020314
I ask myself if I really need to measure that curve. Sometimes you can get away with just matching a couple points.
If I have to match it perfectly, I'll print 2 layers to check.

>>1020490

Honestly its a cheap dollar general scrubber I bet that part of the handle size doesn't stay consistent. I'll just do a typical cone shape and be good enough it doesnt need to be an exact match.

>>1020481
>>1020507

>>1014169
I'm >>1013512
I've used blender for my 3d printing stuff for the first few months too, because I wasn't aware of good CAD and have quite a bit of experience with Blender. But with Onshape I can do things I can't (easily) do in blender, and it's just so quick to make something with fillets and dimension variables (so I can change some values to improve fitting, because my printer is crap and doesn't hit the dimensions well)
It's quite different, but also very fun. And when the browser/the website crashes, your progress is still saved online, so there's that

How major is implicit modelling and is it gonna overtake stuff like Breps for CAD in the future? i wanna mess around with nTopology but thats 30k per seat.

I've got this thing I want cut out of steel, but I made it in Blender so the file is all wrong. Anyone know where I can find a guide on how to properly convert this into a step file? I just tried dicking around in freecad to convert it, but the Chinese factory just told me to remake it in solidworks and I absolutely cannot do that.

>>1021152
Give up and remake it, the sooner the better, there's still no good way to convert STL to STEP
What's the thing?

>>1021153
Custom 0 iron golf club head. Never used a real CAD program before because I'm pretty good with the Blender and its fine for the junk I usually print, so I fear I may be out of my depth with something this complicated.

>>1021154
stl is a bit of a pain in the dick for any kind of cad or cam use, i suppose with the exception of 3d printing where all it's doing plotting layers and precision isn't really a big deal
issue is stl is just a bunch of triangles, there's no actual curves or other primitives, like you can't do even basic things like query the size of a hole since there's no concept of a hole in stl, it's just a bunch of triangles which may or may not be derived from a circle
solidworks has tools to help import stl's and convert elements into proper primitives, i've tried them before and it's often easier to just make the object again from scratch (maybe the few stl's i've tried were just particularly janky, i couldn't say)

>>1021157
-- or i just need to git gud, point is really that it's not a one-click process, at least outside of really simple examples

>>1021157
I think I have to decide whether I redo this from scratch or if I just get it sintered and try to make it not look like shit in the garage.

>>1021154
just try onshape with a throwaway email adress in an incognito tab and give it a shot

>>993097
Similar question. Is there a simple sorta-3D software for designing simple assemblies? I'm trying to figure out how my firearm action concept would operate, and it's really hard to do with layers in Paint.NET. Something very casual, it wouldn't have to do much.

>>1021161
Gonna give that a shot on the weekend, need to learn that anyway for a mechanical thing Im gonna make that probably won't work great if I blend it. Might settle for just getting it SLS printed from steel and buying a harbor freight polishing wheel because machining is expensive and this is one hell of a first project for a whole new class of software, like just look at that shit.

anyone here use ConjureSDF? haven't tried it but seems decent for hard surface stuff
I can't stand doing hard surface with regular tools so idk, just curious if it's worth it

>>1021391
I haven't and I wouldn't, it's locked to blender 3.3 and the next update is bring SDF features to geometry nodes so there's going to be a lot of SDF plugins popping up in a couple of months

rumors of pirated solidworks phoning home and "hobbyists" getting targeted legally by dassault lawyers how much is true? stories all over the internet about this should go full tinfoil and use pirated software on an airgapped machine only?

>>1021475
SW does phone home, if you let it go on the internet it will silently check your license and stop working. you should always keep it firewalled. i use simplewall on my SW VM, whitelisted so everything is blocked by default, no issues. at least, that's what i'd say if i pirated SW... of course.
as far as legal risks to actual hobbyists, idk if they'd consider it worth their time/money to go after you, but that is a risk you have to decide for yourself. if you do it in a business/commercial setting then... well don't do that.

>>1021476
thanks anon, i've seen people get legal action against them to pay the license fee just for making a few halloween decorations. in the case for digital hygiene what minimal lightweight VM do you recommend? my main use is FEA simulation and generative design (nTopology).

>>1021477
personally i use virt-manager/qemu on linux, with a passed-through second gpu. note that the vm is not related to isolating solidworks from the internet, simplewall is all you need for that. i just happen to not use windows as my main os, so it's really only there for solidworks, so the potential inconvenience of needing to allow programs to use the internet every time isn't an issue. there might be a way to deny internet access to everything in the SW program files folder without adding them all rather than making a whitelist (deny-by-default), but i haven't checked.

>>1021477
>>1021478
>simplewall is all you need for that.
well technically the built in windows firewall can do it as well, and i did that for a bit, but simplewall makes this easier

>>1021478
could i just deny the vm any network access in the first place and achieve the same thing?

>>1021480
sure, a completely offline VM or second computer is another way to do it, i don't strictly need that vm to access the internet, but it is convenient for me as i can then still use a web browser to access parts websites and the like, as well as my syncing software, but you can simply not give the vm a network adapter if you're paranoid.
i'm not personally familiar with gpu passthrough if your host (your main os) is windows though, so you'd need to look that up if you use windows

>>1021481
thanks again anon, my host is windows so gpu was trivial to set up.

>>1021475
so following this, can passing a file to a client, made with a pirated isolated solidwork get you in trouble easily?
what about the same but with an educational license?

>>1021558
Yes. D'assault is an very aggressive company, they WILL go after basement dwelling chuds and milk you dry of 40k+ dollars. On top of that, they have a specialized sector for that and an unknown, elite law firm by their side, whose only one client is D'assault themselves and no one else. They specialize in illegal usage of software and you will have near zero chances to win against them in court.

>>1021559
Proceeding with this, will you pirate and not dare to make a single little mistake like it catching your internet connection for a split second? You will pay dearly and get fucked up good, that law firm will come and fuck you hard.

Or better yet, stay away from D'assault and Autodesk.

Choice is yours, fellow /3/ chudlet.

>>1021560
i can even put a physical pc completely offline if necesary
in one of my previous jobs they did that for abaqus, was a pain in the ass to put simulations that were done in another computer. but in this case everything could be done on that pc
>>1021559
what i mean, would they know? i dunno why people as for solidwork in job postings, i actually expect talking to the clients to send them step files or whatever, as they are not industrial guys

>>1021560
can the legal team even reach places like russia or china?

>>1021558
i suppose there's no way to be certain, but i haven't needed to send out my sldasm/sldprt files, only technical drawings (pdf's) made with solidworks to manufacturers
files made with an educational license do carry that fact with them, you can't open educational files in the full version, so that's not as suited for professional work, on purpose of course

>>1021563
they operate everywhere AFAIK. third world redditors had problems with them.