Welcome! Log In Create A New Profile

Advanced

"CAD" files vs CAM files - a designer's plea

Posted by Nether10 
"CAD" files vs CAM files - a designer's plea
January 13, 2011 08:14AM
I'm a mechanical engineer that does design work. When I come across something cool like RepRap the first thing I want to do is download all the CAD files and have a look. Thus far, trying to do that has been frustrating at best, misleading at worst. And I think I know why.

The problem seems to be that most people will call just about any damned file generated by a computer a "CAD" file. Generally, when I see a "CAD" file link, it takes me to a bunch of STL files, which are only useful to stereolithography-type machines. They have no use to CAD users, in terms of "CAD" being defined as Computer Aided Design.

The key word there is "aided" - a human is still doing the design work, but using the computer as a tool. So CAD files should be files that are accessible and useful to humans, which I would classify as DXFs as the most basic, and going up through DWGs, IGES, STEP, Parasolid (X_T), and on up to native CAD package formats (SLDASM and SLDPRT for SolidWorks, for example).

So what would one call an STL file, or equivalent? I'd call it a CAM file - Computer Aided MANUFACTURING, which is really what it's all about.

Long shot here - any chance that we can agree that files useful to humans are CAD files, while files used by/useful to only computers are CAM files?
Re: "CAD" files vs CAM files - a designer's plea
January 13, 2011 08:59AM
You won't get any argument from me on this point although I have modified STL files but ONLY because the source files for the object be it AOI, OpenSCAD, IGES, STEP or whatever was not posted.


Bob Morrison
W├Ârth am Rhein, Germany
"Luke, use the source!"
BLOG - PHOTOS - Thingiverse
Re: "CAD" files vs CAM files - a designer's plea
January 13, 2011 09:36AM
Where are you seeing STL files labelled as CAD files?

There are STEP and IGES files for the Cartesian bot, AOI files for Adrian's extruder. Most of the more recent work is done in OpenScad and .scad files are available, but they are more suited to programmers than mechanical engineers.


[www.hydraraptor.blogspot.com]
Re: "CAD" files vs CAM files - a designer's plea
January 13, 2011 09:56AM
nophead Wrote:
-------------------------------------------------------
> Where are you seeing STL files labelled as CAD files?

Throughout the wiki, on the pages detailing other RepRap variants or RepStraps. There is a box on the right-hand-side denoting development status, what it's based on, author, etc., which includes "CAD files" as well. Just in general, nobody refers to CAM files, they call everything a CAD file. STLs and the like are a dime a dozen around here, but trying to find a decent set of solid models is BRUTAL. (Thankfully I've found a Mendel set, so I can get to work on that, should that be what I eventually select to build.)
Re: "CAD" files vs CAM files - a designer's plea
January 13, 2011 02:50PM
Long shot here - any chance that we can agree that files useful to humans are CAD files, while files used by/useful to only computers are CAM files?


This does not compute.


Random Precision
Re: "CAD" files vs CAM files - a designer's plea
January 13, 2011 03:19PM
Yes, I also agree that an STL file is not a CAD file. That's why I believe that a project which only releases their STL files should not qualify as open-source.
Re: "CAD" files vs CAM files - a designer's plea
January 13, 2011 03:44PM
Indeed, I would like to see STEP files published with any design. It is an accepted standard file format that almost any BREP modeler can handle.

I'm not sure what tool chain can make that happen right now. I found .scad files for Prusa Mendel, but I didn't find a way to export to STEP! OpenScad may be FOSS, but it's not suitable for design work. What can be done, other than model them over from scratch in your design tool of choice? Is there something more usable I should choose to use?
Re: "CAD" files vs CAM files - a designer's plea
January 15, 2011 06:50AM
An .stl IS a CAD file. Your design program might be advantageous because it recognises curved features, but you only need that facility if your design needs to be edited, right? The reprap developers have signed off on Mendel or Huxley designs and presented you with something that you can make and use. You're sniffing "It isn't a CAD file because I can't bash it up easily!" I use TurboCAD, PunchCAD and Alibre for work and play. Each of them can open and view .stl files, recognise them as manifold solids, consolidate coplanar triangulation as faces, and perform boolean ops on them. STEP objects are generally referred to as "dumb solids", and the advantage that they offer over .stls is curved surface definition. That isn't what makes something CAD or something else, it's just a lower bar for editing. The .dxf files you're willing to accept as CAD don't give curved-surface recognition or guarantee watertightness, what makes them more acceptable to you than .stl?

Edited 1 time(s). Last edit at 01/15/2011 08:23AM by murrayd.
Re: "CAD" files vs CAM files - a designer's plea
January 15, 2011 02:04PM
An .stl is not a CAD file, precisely because it doesn't allow the design to be edited.
Developers have not signed off on the designs. They encourage them to be altered and remade, and do not guarantee that the design is error-free. Doing so is exceptionally difficult when all you have to work with is an .stl file.
Re: "CAD" files vs CAM files - a designer's plea
January 15, 2011 04:04PM
Another problem with .stl is the loss of precision due to faceting curved faces. When the author only publishes the .stl file, part of the manufacturing process has already been determined, That is, at the time of export, the precision of the faceted approximation for curved faces is set in stone. Everybody else has to live with that, even if they would like to print with finer settings.

It's interesting that Alibre has some functions to interpret the .stl files and perform editing operations. I wasn't aware of that. But once it starts consolidating triangles into faces, you no longer have .stl compatible data. Until you export a new one, that is.

Hmm. Searching a bit, I found FreeCAD. An OpenCacade modeler capable of handling STEP files, and it can be scripted with Python for generating parameterized models. This seems so much more capable and easy to use. I wonder why it isn't more popular in the RepRap community.
Re: "CAD" files vs CAM files - a designer's plea
January 15, 2011 06:00PM
So although my applications can import .stls and edit them as easily as .stp or .igs files, I'm not working with CAD because you can't? The picture is a Mendel frame vertex imported as original, solidified, then facet consolidated and filleted.



Edited 2 time(s). Last edit at 01/15/2011 06:27PM by murrayd.
Re: "CAD" files vs CAM files - a designer's plea
January 15, 2011 06:45PM
Nonono. You're working with a CAD system that can import the CAM file and do useful things to it. Basically reverse engineer the faceted .stl file to create a BREP model. It doesn't save it as .stl though, unless you export. Right? On the other hand, Alibre's file structure is based on STEP!

I need to remember to look into this capability of Alibre. At the very least, I might be able to use it to back .stl files out to STEP. Then almost anything can edit it.

Edit: I see your filleted model still has faceted holes instead of round. This is part of the problem.

Edited 1 time(s). Last edit at 01/15/2011 06:47PM by Dale Dunn.
Re: "CAD" files vs CAM files - a designer's plea
January 15, 2011 07:33PM
Dale Dunn Wrote:
-------------------------------------------------------
> Hmm. Searching a bit, I found FreeCAD. An
> OpenCacade modeler capable of handling STEP files,
> and it can be scripted with Python for generating
> parameterized models. This seems so much more
> capable and easy to use. I wonder why it isn't
> more popular in the RepRap community.

People tend to use what they know. Programmers seem to prefer openscad, but modelers might prefer blender. My background is in Autodesk Inventor so this is the primary tool I use. To change programs requires learning a whole new set of menus and commands. That can be pretty hard sometimes.
Re: "CAD" files vs CAM files - a designer's plea
January 15, 2011 09:45PM
Having a look at FreeCAD this evening. It appears that it may have been passed over by developers because it is still in early development. Even so, it appears to be nearly as capable as OpenScad already wrt solid modeling. Very promising, but still a long way to go. It's too full of holes right now.
Re: "CAD" files vs CAM files - a designer's plea
January 15, 2011 11:17PM
The frame vertex wasn't "reverse engineered to create a BREP" in any way. .stls are manifold objects, like BREPs. The only considerable difference is curved surface definition. Alibre doesn't produce STEP objects unless you export them, and then your parametrisation is gone. STEP is then only advantageous if your receiving application is capable of feature recognition, and if your file has curved features. TurboCAD can toggle between mesh and solid, making vertex, CSG, facet and direct editing all available on the object at one session. Even FreeCAD has a function under the "part" menu: "create shape from mesh", select that result then "convert to solid". Select an .stl and apply that function, you've then got a solid that you can manipulate with CSG (or export as STEP, if that's your schtick), then convert back to .stl for production. "stl isn't CAD" is an arbitrary distinction, not a real one.




Edited 2 time(s). Last edit at 01/15/2011 11:32PM by murrayd.
Re: "CAD" files vs CAM files - a designer's plea
January 16, 2011 12:35AM
Yes, .stl files are manifold bodies, but that's about where the similarity with BREP models ends. Even if the only considerable difference is curved surface definition, that's a big difference!

I've been exploring these functions for converting between meshes and BREP. They still run afoul of the fact that the .stl mesh is an approximation of the original model. Curved face definition is lost, and faces that should be flat are still full of triangular facets. Even if I concede that .stl is a "CAD" format, I still contend that it is a poor format for interoperability and design work, unless you are working on a design that is well suited to mesh modeling.

Every model format has its strengths. BREP is good for precise, feature-based prismatic solids and NURBS surfaces. CGM, I'm not familiar with. I think it has fallen out of mainstream use in industry. Mesh modelers are good for less precise designs, and they excel at more "organic" shapes. Translating between BREP and mesh is very difficult to do well, and that's the problem I have with .stl files. I've been experimenting on Wade's extruder frame from the Prusa .stl files, and I can't get a model that is usable in a BREP environment. I'd like to see what TurboCAD can do with it, but I wouldn't have to if the models were published in a neutral BREP format. Perhaps I should learn Blender to work with .stl files? Is there a free tool that can turn an .stl file into a useful BREP?

What I was saying about Alibre's file format is this: Alibre doesn't store parts as .stp files. What it does do is represent the BREP model geometry in the STEP format within the Alibre part file. It was a selling point when Alibre came out, that the STEP interchange was a high quality because STEP was the natural storage format.

I think perhaps we should first settle what kind of models are most useful for the task at hand. Exchanging files for character animation or some other "artwork" is probably best done with a mesh model such as would be used by blender. Exchanging files for machine designs (like a 3D printer) is best done with a BREP format, such as used by all the mainstream CAD tools. Can we agree that these are the strengths these formats play to?
Re: "CAD" files vs CAM files - a designer's plea
January 16, 2011 02:17AM
All I can say to you is that standing about stamping your foot because you don't like the format isn't likely to change anything. The people who got there before you made it work.
Re: "CAD" files vs CAM files - a designer's plea
January 16, 2011 02:48AM
copy-and-paste from parallel discussion in reprap-dev (which you guys should be in, since you are devs grinning smiley ):



Oh, exactly.

We've got a whole bunch of sub-populations talking past each other,
each championing their favorite tool. Each tool is weakly or strongly
non-compatible with other file types.

The best way to deal with this is to take a "whatever gets you through
the night" approach: use your favorite tool, upload files, and be very
polite to people that try to get you to switch to their favorite
tool. It's a pure beauty contest. If we were to force a vote, we'd
just have the non-plurality go off in a huff, keep developing, and
stop uploading files.

As an example, we get an annual surge of mech-e students that have
just learned a proprietary CAD program and they want to make RepRap
better. If we tell them "openscad is mandatory" they get annoyed and
evaporate. If we tell them, "welcome aboard! (and don't mind the
chaos)" they'll have more fun and RepRap will keep getting better. grinning smiley

There are no real action items here, I'm afraid. It's the same with
weakly-printed (fast-rep) repraps versus strongly-printed (slow-rep
repraps). We're not going to reach consensus, therefore no action
items, therefore let's not worry about it and just keep reprapping. ^_^

Andrew, let me know if you have any trouble uploading. The mediawiki
gets a bit fussy when presented with "media", i.e. cad files. We're
working on it: [lists.reprap.org]

Cheers,
Sebastien
p.s. As much fun as it may be to discuss the philosophical
implications of all this, (and getting very meta), I'm personally much
more keen on discussion that results in actual dev of repraps taking
place ...

On January 15, 2011 01:52:23 pm David Johnson wrote:
> But can heekscad do anything with the native scad file and openscad do
> anything with the native heek file. That would make those formats pretty
> dead as well using that definition.
>
> On Sun, Jan 16, 2011 at 4:07 AM, Neil Underwood
wrote:
> > Andrew, one issue though is that an exported STEP is a simi dead format.
> > The native formats of the open source CAD programs are the most workable.
> > Even though Heekscad can import IGES and STEP, it's can't do much with
> > them, on the other hand Openscad and Heekscad can do anything with their
> > native Scad and Heek files.
> >
> >
> > On Sat, Jan 15, 2011 at 8:10 AM, Andrew Netherton <
> >
> > [email protected]> wrote:
> >> I'm working on taking a STEP assembly of the Mendel, fixing it up a
> >> bit (parts/components missing, out of place, etc.) in SolidWorks, and
> >> making the finished product (both with and without whatever tweaks I
> >> throw in) available in SolidWorks, STEP, and whatever other format is
> >> requested. Watch the wiki for the file, but likely not for at least a
> >> week or two.
> >>
> >> Andrew
> >>
> >> _______________________________________________
> >> reprap-dev mailing list
> >> reprap-dev@lists.reprap.org
> >> [reprap.org]
> >
> > _______________________________________________
> > reprap-dev mailing list
> > reprap-dev@lists.reprap.org
> > [reprap.org]


-Sebastien, RepRap.org library gnome.

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything! grinning smiley
Re: "CAD" files vs CAM files - a designer's plea
January 16, 2011 03:51PM
murrayd Wrote:
-------------------------------------------------------
> All I can say to you is that standing about
> stamping your foot because you don't like the
> format isn't likely to change anything. The
> people who got there before you made it work.


Well, maybe I need to examine my tone, but my intention is not to whine about my favorite CAD system. I don't expect people to invest in a seat of SolidWorks. What I'm trying to do is advocate for posting a neutral format that people can use with their favorite systems. More in a bit.
Re: "CAD" files vs CAM files - a designer's plea
January 16, 2011 04:52PM
SebastienBailard Wrote:
-------------------------------------------------------
> copy-and-paste from parallel discussion in
> reprap-dev (which you guys should be in, since you
> are devs grinning smiley ):

...

> > > Andrew, one issue though is that an exported
> STEP is a simi dead format.
> > > The native formats of the open source CAD
> programs are the most workable.
> > > Even though Heekscad can import IGES and STEP,
> it's can't do much with
> > > them, on the other hand Openscad and Heekscad
> can do anything with their
> > > native Scad and Heek files.
...


What I'm trying to advocate for (apparently not well, since I'm having to summarize now instead of earlier) is posting the best formats possible so that others can use whatever gets them through. I don't mean to ask for everyone to standardize on STEP, but, if possible, make it available also. The more formats we make our work available in, the more developers will be able to contribute.

When I joined this thread, I asked what the best tool is for working with these files. That got lost in the discussion of what "CAD files" are. So, if I want to look at or work on Prusa Mendel, my options for files seem to be .stl or .scad. Is OpenScad my only practical option other than rebuilding the models in my system of choice?

I work in a multi-CAD environment on a routine basis. The only way we've found to cope is to export to all of the relevant formats for every release. We don't all get to enjoy the benefits of working in the native format, but we do get to work with clean geometry. As I learn more about the open source CAD situation, I'm finding that this does not seem to be attainable yet. If the core developers are working on OpenScad, it's seems there is nothing to do about neutral BREPs except wait for a BREP import/export to be implemented, or remodel the parts in my system of choice.
Re: "CAD" files vs CAM files - a designer's plea
January 16, 2011 05:40PM
Dale Dunn Wrote:
-------------------------------------------------------
> SebastienBailard Wrote:
> --------------------------------------------------
> -----
> > copy-and-paste from parallel discussion in
> > reprap-dev (which you guys should be in, since
> you
> > are devs grinning smiley ):
>
> ...
>
> > > > Andrew, one issue though is that an
> exported
> > STEP is a simi dead format.
> > > > The native formats of the open source CAD
> > programs are the most workable.
> > > > Even though Heekscad can import IGES and
> STEP,
> > it's can't do much with
> > > > them, on the other hand Openscad and
> Heekscad
> > can do anything with their
> > > > native Scad and Heek files.
> ...
>
>
> What I'm trying to advocate for (apparently not
> well, since I'm having to summarize now instead of
> earlier) is posting the best formats possible so
> that others can use whatever gets them through. I
> don't mean to ask for everyone to standardize on
> STEP, but, if possible, make it available also.
> The more formats we make our work available in,
> the more developers will be able to contribute.
>
> When I joined this thread, I asked what the best
> tool is for working with these files. That got
> lost in the discussion of what "CAD files" are.
> So, if I want to look at or work on Prusa Mendel,
> my options for files seem to be .stl or .scad. Is
> OpenScad my only practical option other than
> rebuilding the models in my system of choice?

For Prusa yes. It is designed in openscad so you are limited to what openscad can export.

>
> I work in a multi-CAD environment on a routine
> basis. The only way we've found to cope is to
> export to all of the relevant formats for every
> release. We don't all get to enjoy the benefits of
> working in the native format, but we do get to
> work with clean geometry. As I learn more about
> the open source CAD situation, I'm finding that
> this does not seem to be attainable yet. If the
> core developers are working on OpenScad, it's
> seems there is nothing to do about neutral BREPs
> except wait for a BREP import/export to be
> implemented, or remodel the parts in my system of
> choice.

Openscad is a completely different way of working. It allows much more collaboration than a visual cad system because you can use normal software version control to show differences and merge changes. When combined with git it fits the reprap project perfectly and I think that is the way the it will go. The only problem is designing hardware becomes writing software, which is great for programmers, but not mechanical engineers. It is very easy to use though, I picked it up this weekend. It is very limited in what it can do but reprap parts are very simple shapes. It's like AOI without the GUI and the boolean bugs.

You want everybody to use STEP, but very few open source programs support it. None of the ones commonly used by the reprap do.

The tool I use is CoCreate PE, which is fantastic free tool, but it can only save in its own closed format and STL. So anything I contribute to the community is STL. I put the source on Thingiverse as well but it is only useful to other CoCreate users.


[www.hydraraptor.blogspot.com]
Re: "CAD" files vs CAM files - a designer's plea
January 16, 2011 07:34PM
nophead Wrote:
...
> You want everybody to use STEP, but very few open
> source programs support it. None of the ones
> commonly used by the reprap do.
...

That's what I'm finding out. I can see that if I want to contribute design work to the community, I'll pretty much need to work in these formats. That's the way it works in other mechanical design realms too. You use what's compatible with the others in the project. Whether we like it or not, a de facto standard emerges. I don't mind if everybody doesn't use STEP, but it sure would be nice if everybody could make it available in order to expand compatibility. That's just not an option with the present state of tools being used by the community.

Your comments about using software version control are interesting. Mechanical design has both traditional and software-based revision control processes. I've often wondered how different they are from software version control, and whether one could be adapted to the other. Perhaps this is an opportunity to explore that.
STEP is more accessible than it used to be
January 16, 2011 07:54PM
We really are spoiled for choice. I was thinking on it last night, and I realised that when I started using CAD, around twenty years ago, .stl was the only universal model format for manifold objects, IGES was at version 4, and STEP didn't exist...
But STEP has made inroads into free and open-source CAD, mainly because of the solids kernel that Matra Datavision abandoned and open-sourced, Open Cascade.
HeeksCAD, FreeCAD and NaroCAD all use it, and I try to keep up with each of them, because they're being pretty actively developed. HeeksCAD is geared towards CAM/CNC, NaroCAD towards direct editing, and FreeCAD towards parametrisation and part history.
I did spend some time playing with FreeCAD last night, and it's at a pretty mature level. There's interplay between graphical and model objects (things SNAP to one another! Woohoo!), solids boolean reliably, and the only real problem I could find (although its more likely that I just haven't read that part of the manuals and tutes yet) is that there isn't a way of highlighting selected edges for blends or chamfers. It has parametrisation, a history tree, IGES, STEP, .stl and .dxf exchange, in addition to SVG import and drawing output with model views.
Of these three modelers, the only one that really allows conic surface (cubic patch) production is NaroCAD, through curve sweeps, but if you've a program that does IGES surfaces, say Marcus Bole's PolyCAD (which is free but not open source), FreeCAD can thicken them to allow applying them to prismatic solids through booleans. Ten years ago, capabilities like those started at the best part of a thousand dollars and headed on upwards.

Edited 3 time(s). Last edit at 01/16/2011 08:07PM by murrayd.
Re: "CAD" files vs CAM files - a designer's plea
January 16, 2011 09:34PM
When I joined this thread, I asked what the best tool is for working with these files. That got lost in the discussion of what "CAD files" are. So, if I want to look at or work on Prusa Mendel, my options for files seem to be .stl or .scad. Is OpenScad my only practical option other than rebuilding the models in my system of choice?

Nope, happily in a week or two:

> > [email protected]> wrote:
> >> I'm working on taking a STEP assembly of the Mendel, fixing it up a
> >> bit (parts/components missing, out of place, etc.) in SolidWorks, and
> >> making the finished product (both with and without whatever tweaks I
> >> throw in) available in SolidWorks, STEP, and whatever other format is
> >> requested. Watch the wiki for the file, but likely not for at least a
> >> week or two.

Yes, I also agree that an STL file is not a CAD file. That's why I believe that a project which only releases their STL files should not qualify as open-source.

Yesssss ...a bit of nuance here Jacob. grinning smiley The best thing to do in such cases is to gently ask the developer in question to upload the useful source (i.e. CAD or CAD-like) file to the wiki. The worst thing to do is to angrily invoke the GPL and get all confrontational ...

That's what I'm finding out. I can see that if I want to contribute design work to the community, I'll pretty much need to work in these formats. That's the way it works in other mechanical design realms too. You use what's compatible with the others in the project. Whether we like it or not, a de facto standard emerges. I don't mind if everybody doesn't use STEP, but it sure would be nice if everybody could make it available in order to expand compatibility. That's just not an option with the present state of tools being used by the community.

The best thing to do is use your favorite tool and put up the most "universal" files that tool can export. I don't think anything is going to bridge the divide between the subpopulations of tools anytime soon. When RepRap finds a fork in the road, it tends to take both branches. smiling bouncing smiley


-Sebastien, RepRap.org library gnome.

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything! grinning smiley
Re: "CAD" files vs CAM files - a designer's plea
January 16, 2011 10:54PM
Yes, Sebastien, I fully agree. My aim with that statement was not to bring a hammer down on people who have to-date uploaded only STLs, but to make them aware that such files are good for printing but essentially useless for making changes to. I think it's possible that many people simply aren't aware of how much of a dead-end STL files are, since they often do upload them with a license that permits changes.

STLs are to CAD models as a PDF full of photographs of words is to a text file. You can send them both to a printer just fine, but it's a big hassle to fix a typo.

Anyway, so basically you're right, we should encourage people to upload the most universal files that their tools allow. There's two independent criteria for universality though.

The first is lossiness. If file A contains more information about the model than file B, such as feature history, curved surface definition, or geometric relations, then most programs that can open file A will be able to use it to generate file B. No program that can open file B will be able to use it to generate file A though. The translation from A to B is lossy and irreversable, and that means that file A is a more "universal" file type than B.

The second is compatibility. If file A can only be opened by a proprietary program that costs a thousand dollars, and file B can be opened by every program, then by this measurement file B is more "universal" than file type A.

Maybe whenever there's a situation like this, the best thing Alice can do is to just spam the uploader with as many filetypes as she can export. Then Bob can choose the most lossless file that his particular CAD program can import. Alternatively we can try to agree on a community-wide standard format, but I don't think we'll find any single format that is as lossless as a native file and as interchangeable as an STL file. So the former method is less of a compromise.

The downside is that it makes revision history a bit tricky, because the different file formats corresponding to a single model will not necessarily stay up-to-date with each other, unless changes are only made to the master "root" model and that one is used to update the lossy copies in other formats. But then only people who can edit the root format will be able to make changes, so it might not be very accessible.

That, or we can just let the revision tree become somewhat messy, where people modify the lowest branch on the tree that they can, and hope that someone else with better software can copy the changes down to the root. In this case, the most up-to-date file would always be the one that is hardest to modify. For this reason, as well as because of the repeat labour involved in copying changes up to higher levels, maintaining a model would require a lot more work. But it would also make the file the most accessible.

Edited 2 time(s). Last edit at 01/16/2011 11:13PM by jbayless.
Re: "CAD" files vs CAM files - a designer's plea
January 17, 2011 10:04AM
Anybody tried something like this? [www.solveering.com]

Even if you take openscad, convert to STL and then to STEP and then edit it in a GUI, you can't go back to openscad again, so you can't collaborate on Prusa Mendel, only fork it.


[www.hydraraptor.blogspot.com]
Re: "CAD" files vs CAM files - a designer's plea
January 17, 2011 10:21AM
nophead Wrote:
-------------------------------------------------------
> Anybody tried something like this?
> [www.solveering.com]
> ep.html
>
> Even if you take openscad, convert to STL and then
> to STEP and then edit it in a GUI, you can't go
> back to openscad again, so you can't collaborate
> on Prusa Mendel, only fork it.

Murrayd pointed out that TurboCAD has a similar function (though not free). The examples given for TurboCAD and Solveering both had errors in the planar faces and round features that are still faceted. In theory, it should be possible to to examine other features that appear to be primitive (perhaps with user intervention) and replace them with appropriate curved faces. There may still be losses, though.

Perhaps the best use for someone in the larval stage or participation, with suitable tools (such as myself), would be to devote some time to recreating Repraps from .stl files in a CAD system capable of exporting many formats. Where would someone ignorant of the version control tools used by RepRap go to learn how best to make such files available at the current revision level, and how to keep up wit changes?
Re: "CAD" files vs CAM files - a designer's plea
January 17, 2011 05:54PM
There's a function to eliminate edges, consolidating faces, in FreeCAD. Convert mesh to object, then object to solid, then It's the blue "up arrow" on selected faces. Not as to-the-point as the function in TurboCAD. Trying it out, it's a bit shaky, ie, sometimes doesn't like to attempt too many faces conjoined at once, or the weather, or the way I'm holding my tongue, or the way I'm looking at it. Other methods are mostly much more direct.
Circular features, no. The reason being that the facet arrangement of cylindrical faces isn't regular, for .stls it's predicated on how the vertices meet adjoining faces, because the rules governing that are part of the .stl format definition. The only way to do it, even semiautomatically is to add a hole filling, then subtract a *cylindrical* cylinder from the solid. Accurate location can't be done off the faceted hole representation, though.
But as has been pointed out, losses from reverse-engineering .stls are almost inevitable. Just in the available Mendel frame vertex .stl, two faces have been shattered into triangles that are no longer coplanar, and minutely out of register with the remainder of the geometry. But given that Mendel was originally built in Solid Edge, I wonder if the .stl flaw is from Solid Edge's conversion? Doesn't matter in the real world, anyhoo.
Re: "CAD" files vs CAM files - a designer's plea
January 17, 2011 08:02PM
Perhaps the best use for someone in the larval stage or participation, with suitable tools (such as myself), would be to devote some time to recreating Repraps from .stl files in a CAD system capable of exporting many formats.

> > [email protected]> wrote:
> >> I'm working on taking a STEP assembly of the Mendel, fixing it up a
> >> bit (parts/components missing, out of place, etc.) in SolidWorks, and
> >> making the finished product (both with and without whatever tweaks I
> >> throw in) available in SolidWorks, STEP, and whatever other format is
> >> requested. Watch the wiki for the file, but likely not for at least a
> >> week or two.

grinning smiley

Where would someone ignorant of the version control tools used by RepRap go to learn how best to make such files available at the current revision level, and how to keep up with changes?

irc://chat.freenode.net/#reprap
http://reprap.org/pipermail/reprap-dev/


-Sebastien, RepRap.org library gnome.

Remember, you're all RepRap developers (once you've joined the super-secret developer mailing list), and the wiki, RepRap.org, [reprap.org] is for everyone and everything! grinning smiley
Re: "CAD" files vs CAM files - a designer's plea
January 17, 2011 10:17PM
Signed up!

OpenScad is an interesting little platform. I can see why programmers take to it. Painful from my perspective.

At first I thought I would just patch up the .stl files, but the .scad files contain formulas that document what we in the SolidWorks community call "design intent". I'll translate that too, but it's going to take a while to catch up to this moving target. A long while if MakerGear delivers my Prusa soon. Hopefully I can capture enough from these formulas to be able to quickly generate different sized Prusas.
Sorry, only registered users may post in this forum.

Click here to login