cannot get radiused sheet metal part to flatten

[Mike Gera]

I'm having difficulty getting the attached part, 8575-09-02_CADForum.SLDPRT, to properly flatten.

The issue can be traced to the "flange fills - bottom" feature. If that feature is suppressed, the part will flatten.

Seeing my design intent in that feature, is there anything I could perhaps do differently?

Thank you for any replies.

[bnemec]

Model saved in 2024?

[Mike Gera]

Model saved in 2024?

Yes.

[AlexB]

The feature between your unfold and fold where you add material by the bend is causing the issue. I'm not sure what to suggest though.

[DLZ_SWX_User]

SO I just checked into this part and @AlexB has it correct. That added material is the issue. Try building the part by adding the 90° side flanges last & see if that works. I have had issues in the past arc's & flanges not playing well together. Might be something to talk to your VAR about as I think it should be able to flatten as you have it built.

I saved the part back to 2023 & built it as stated and was able to get it to flatten but when I bring it into 2024 it does not.

[XHawkeye]

FWIW: (Using 2024/2.0) it'll flatten if the left side of the sketch is changed to construction lines or the flange is flush/behind the bend.

[Mike Gera]

FWIW: (Using 2024/2.0) it'll flatten if the left side of the sketch is changed to construction lines or the flange is flush/behind the bend.

@XHawkeye

Is there any way you could send me a screenshot of that?

Thanks.

[Mike Gera]

This works, but it's not "correct". I wish there was a way to stop the bend at the point at which the flange ends.

ETA: There does not seem to be a way to edit the "Bend-Lines" sketch, which would, in theory, do the trick.

[tate876]

It looks like the flange fills feature might be causing issues when trying to flatten the part. One thing you could try is to adjust the geometry of the flange fills, maybe by breaking them into smaller sections or simplifying the design. Another option might be to check if any faces are over-defined or interfering with the flattening process.

[Dennis Bacon]

It looks like SW doesn't like two (or more) bend radii in the same zone. I would imagine that you could get by with adding a flat boss at the end. I am unable to use your file since it is a future version (I'm on 2020) for me but I get the gist from an eDrawings file of your model.

[HercalloY]

Yep as mentioned it seems its a situation SW can't handle. i tried to do it in every different way i could think of and no matter what, that same area errors out when it comes to generating the flat pattern. As soon as the Rolled section enters past the tangent location of the edge flange radius... Computer Says No.

Orange DIM Breaks Flat Pattern if Greater then 0

DIM Greater then 0 = Error Pattern Fails

DIM Less then 0 = Pattern Works

Works perfectly in NX. Modeled it exactly as you did (for curiosity), no problems (and no problems when done in a number of different modelling methods either).

[SPerman]

Works perfectly in NX. Modeled it exactly as you did (for curiosity), no problems (and no problems when done in a number of different modelling methods either).

I haven't done much sheet metal work in either program, but from my limited experience, NX has way fewer issues.

[Mike Gera]

I've come to the same conclusion: SolidWorks lacks the code depth to handle curved-to-straight sheet metal transitions.

@HercalloY and @SPerman

I've often wondered about NX. How do you like it so far? Pros, Cons? Expensive? Thanks.

[SPerman]

I haven't used it since I left my last job in 2017. Back then it was extremely robust. The idea of crashing while doing nothing of consequence was unheard of. Switching from NX to SW felt like a major step backwards. In fairness, some of my problems were from trying to use SW as if it were NX.


I would switch today, but the cost of the features I need is easily >4x the price of the same features in solidworks.

[Mike Gera]

I haven't used it since I left my last job in 2017. Back then it was extremely robust. The idea of crashing while doing nothing of consequence was unheard of. Switching from NX to SW felt like a major step backwards. In fairness, some of my problems were from trying to use SW as if it were NX.


I would switch today, but the cost of the features I need is easily >4x the price of the same features in solidworks.

That's the thing. None of us like to admit it, but SolidWorks is, at the end of the day, a cheap and dirty "for the masses" 3D modeling package. There's much better out there...for a price.

[jcapriotti]

That's the thing. None of us like to admit it, but SolidWorks is, at the end of the day, a cheap and dirty "for the masses" 3D modeling package. There's much better out there...for a price.

It's like a pair of Bose headphones....yeah they are better.....but $200-$300 is steep and most don't feel they need it or can afford it.

We do a lot of sheet metal with Solidworks for over 20 years now. But its all simple stuff....mostly 90° bends.....nothing really cylindrical or stamped. So NX would add little except cost in our situation.

[HercalloY]

You guys state the main limitation of price very well. I will note however that currently, if you don't need to push the feature/licensing boundary that NX can be had for cheaper (or around the same price for the license and my subs versus SW Prem are always cheaper for the a basic seat of NX). Yes you probably don't get the other "features" that SW throughs in in their basic forms that NX won't have (but not that many) and the base modeling functions ARE DAMN NICE. I've been extensively comparing the two for nigh on 20 years now with a more in-depth every day side by side usage in the last 10. And although there a small number of things SW does well that i wish NX had a direct capability to do... the no crashing and clean, consistent UI and the way it does things (aka professionally in most cases) has really made this SW lover since SW97 come to really like it.

If Siemens did say, 5 main things to bring it inline with SW functionality in the sense of "usability and modern functionality"... and made the price for higher end features less restrictive (coz lets face it, all they are doing is hurting their own sales by pricing themselves out of more purchases with their ridiculously high prices for extra features.. but that goes for all current software models IMO), no one with objective, rational thought would touch SW anymore with its annoying AF broken basic stuff all the time (i'm currently so sick of navigating to the working folder in SW no matter how many times i open a file or save-as to an external file and back to main file types etc.).

[Mike Gera]

I wonder what SpaceX uses.

[SPerman]

I have no idea about the source of this information

https://www.quora.com/What-CAD-software-does-SpaceX-use

SpaceX uses a variety of Computer Aided Design (CAD) software including CATIA V5, Siemens NX, and SolidWorks​​. Here's a more detailed breakdown:

CATIA (Computer Aided Three-dimensional Interactive Application):
Developed by Dassault Systèmes, CATIA is a powerhouse in the CAD world, especially favored in aerospace and automotive industries.SpaceX utilized CATIA for designing key components of their rockets like the Falcon 9 and Dragon spacecraft​.
It's great for 3D modeling, finite element analysis (FEA), and more, which are crucial for ensuring the viability and efficiency of the designs before they are brought to life.
Siemens NX:
Another major player in the CAD software scene.
It appears that Siemens' software suite was adopted by SpaceX especially after their initial work on the Falcon 1 rocket, which they initially started designing with a mid-range CAD program, but later switched to Siemens' more robust offerings for CAD, FEA, and product data management (PDM) software​.
Siemens NX also comes packed with tools for 3D modeling, finite element analysis, and more, akin to CATIA.
SolidWorks:
A 3D mechanical engineering design program enabling engineers to create and test virtual prototypes before manufacturing.
It’s mentioned that SpaceX designs its rockets and spacecraft using SolidWorks, which is adept for 3D mechanical engineering design and testing virtual prototypes before they hit the manufacturing floor​.
Autodesk and Matlab:
It's noted that SpaceX also uses a mix of software programs including Autodesk and Matlab alongside the above-mentioned CAD software, each having its own unique strengths that contribute to designing some of the most advanced rockets in the world​.
Each of these CAD software packages have their own unique strengths and capabilities, and IMO, the combination of these tools allows SpaceX to continually push the boundaries in aerospace design and engineering. From personal experience, having a robust CAD software suite is a game-changer in quickly iterating and refining designs, which in a field like aerospace, is absolutely critical. The detailed 3D models and simulations these tools provide help in identifying and mitigating potential issues early in the design process, saving tons of time and resources down the line .

[DLZ_SWX_User]

It looks like SW doesn't like two (or more) bend radii in the same zone. I would imagine that you could get by with adding a flat boss at the end. I am unable to use your file since it is a future version (I'm on 2020) for me but I get the gist from an eDrawings file of your model. image.pngimage.pngimage.pngimage.png

So @Dennis Bacon You are partly correct. If I change things up a bit & model it a bit differently I can get it to unfold, mostly correct. I still think that it is something that needs to be reported to Solid works.
First I made just an arc that stops at the point of the start of the Edge Flange bends. I then unfolded the part and made a cut on the end where the flange had been removed. (for a relieve notch). At that point I added the merged boss on one side and folded the part.

I then added the 2 90° edge flanges. I can now flatten the part but I only have the merged boss on one side. If I add the other side the part will not flatten!