Parametric Prusa Mendel?

Hello all.

I am starting a build of an SAE Prusa Mendel. I would like to contribute and make improvements to the design but can not find parametric files (i.e. .iges, .stp, .sldprt). I work primarily in SolidWorks which does not understand openscad files. As such, I am going to create a solidworks model of the prusa mendel and put it up. Before I do that, however, I have a few questions:

Is there a non-openscad parametric set of files for the prusa mendel available?

Are there .dxfs or dimensioned drawings of the prusa mendel parts? (I will put these together as well if not)

What parts are identical from the original Mendel to the Prusa version? It seems silly to remodel parts that have already been modeled.

Would anyone be interested in these models? Thanks.

The Prusa source files (.scad) and .stl files are hosted here on GitHub. There are no parametric models other than the .scad files. The design is still in flux, and there are still issues to be dealt with. In other words, it's a moving target.

There is a regular Mendel that was modeled in SW, though I can't remember where it is right now. It is complete (everything, including screws and nuts), but I don't think it was done with significant modification in mind. It looks to be more of a modeling exercise.

I've started work similar to what you have in mind (in SW 2011) by importing the .stl files and modeling over them as a reference while interpreting the design intent from the OpenSCAD code. I've been working from the metric files, and the Prusa design has progressed somewhat since I worked on it last. I think I downloaded the files January 17th. I had most of the printed components done, but got distracted while working on the Wade's gears (Arduino Inventor's Kit projects. Fun!). The big hangup was that I didn't find a parametric gear model available in SW that I thought was suitable for the project. I have that partly done, and was going to upload it to 3dcc. I may not finish soon, since I expect to receive my MakerGear Prusa before I can get back to it. I need to work out how to upload what I did to GitHub.

Phew! That was not trivial for a git noob. Much still to learn. My SolidWorks files are available here.

Hi
I am one of the people who created the Solidworks files for a Mendel, located
[objects.reprap.org]
I am almost through with my Mendel (group effort) and we hope to start a Prusa style in the next couple of weeks (group effort means we all eventually want to have our own)
Thanks for turning Prusa into solidworks. I am going to have to wait till I get my Solidworks upgraded from 2010 to 2011. I was redesigning the Mendel to go Bearingless, but Prusa beat me to it, with a better design. Gotta love the open source universe.
Anyway, I am sure I will be in contact with you soon to talk about your design. I notice they are all part files. Have you tried doing any assemblies yet?
Thanks again

1

No, I haven't had time to do any assemblies yet. With my Prusa due to arrive tomorrow, that's not likely to change soon. After I get it up and running, I'll probably start thinking about mods, then pick the project up again. Even when I do, I'll be working on a fork of the MakerGear fork, so I'll be going in a bit of a different direction. Keep in mind, some of the work I've done is already outdated. Prusajr is still actively developing the design.

I have not had a chance to look at your files but I am wondering how you created the frame vertices. I am generating them now but the dxf file the model is based on has ill-defined dimensions. I attached a picture (units are inches) to show what dimensions I pulled off. It seems as though the shorter sides near the holes for horizontal rods were meant to be the same size as one another. As were the two inner, medium sidess, and the two outer, medium sides. Does anyone know the correct dimensions for these sides?

Here is the same base drawing dimensioned how it seems it was meant to be dimensioned. I am worried about changing these dimensions slightly and what that may do to the overall design.

I found the same thing. My theory is that the vertices are based on published .dxf files from the original Mendel. I think the only change for Prusa was to add feet to some of the vertices. How the original Mendel vertices (if I'm right) got distorted, I have no idea. The asymmetry is probably not enough prevent someone from using the vertices in either orientation. I found myself wondering how much this contributes to people having difficulty getting their axes running smoothly. Probably not much.

Anyhow, my plan was to eventually build an assembly with a frame shape that followed a layout sketch (using in-context relations). Using that, you could rapidly generate various size and shape derivatives. Toward that end, I reworked the vertex model a bit until it was symmetrical about an axis that bisects the frame rails. The changes were actually dimensionally small, less than 0.5mm IIRC (I was working from the metric Prusa). The differences are probably small enough that parts printed from my model would be mechanically interchangeable with other Prusa or Mendel vertices. The accuracy of the printers isn't much smaller.

My Prusa arrived today, so this is officially back burner for me.

harrison Wrote:
-------------------------------------------------------
> Here is the same base drawing dimensioned how it
> seems it was meant to be dimensioned. I am worried
> about changing these dimensions slightly and what
> that may do to the overall design.
>
> [i.imgur.com]

Wasn't the original Mendel design in mm, then converted for inch?

Edited 1 time(s). Last edit at 02/28/2011 05:36PM by Dale Dunn.

Yes, the original Mendel was designed in mm then converted. I was just trying to show even numbers. I have put together the Prusa frame in solidworks. The frame vertices are done. I also modeled the majority of the simple parts. Time to do the more complicated bits. Thanks for your replies. Were there any other strange bits you ran into?

The only other thing I noticed was that the rod clamps were asymmetrical. Functionally a non-issue, and a trivial fix in OpenSCAD. I submitted that to the project on GitHub, and I think it had already been changed by the time you started work on this. Amusingly, MakerGear is or was still using the old version. I checked for this when I was checking in my Prusa kit last night. Rick's design forked from the Prusa sometime late last year, so this is no surprise. Still it's always interesting to me to see something I worked on (or looked at) in CAD implemented in the physical world later.

Interesting. I'm looking forward to being a contributor to the project, though it has come to be my understanding that there is little interest for a solidworks model as this would likely create another fork in the Prusa development. I took a look at the mendel frame vertex.dxf file on the original mendel github and noticed that it is not a symmetric profile (units are mm):



This isn't likely to be a big issue as the largest difference between sides that should be symmetric is about .25 mm or .01". This obviously is not an issue as many people have printed these and built successful machines. I wonder if it is even worth updating and generating a new .dxf file.

Thanks again for your help.

Harrison - I'd be really interested in the SolidWorks version of the Prusa once you've finished with it. Or, if you'd be willing, it'd be great to have a copy of what you have now. We're looking at getting some parts printed and our supplier wants SolidWorks files. Others might be in a similar position.

I noticed that Dale's uploads to Git, at least for me, show up blank in Rhino. I am kind of a noob with this stuff, so the fault could be on my end. Any advice would be appreciated.

Glad someone's doing this!
Justin

Dale Dunn Wrote:
-------------------------------------------------------
> The Prusa source files (.scad) and .stl files are
> hosted here on GitHub. There are no parametric
> models other than the .scad files. The design is
> still in flux, and there are still issues to be
> dealt with. In other words, it's a moving target.
>
> There is a regular Mendel that was modeled in SW,
> though I can't remember where it is right now. It
> is complete (everything, including screws and
> nuts), but I don't think it was done with
> significant modification in mind. It looks to be
> more of a modeling exercise.
>
> I've started work similar to what you have in mind
> (in SW 2011) by importing the .stl files and
> modeling over them as a reference while
> interpreting the design intent from the OpenSCAD
> code. I've been working from the metric files, and
> the Prusa design has progressed somewhat since I
> worked on it last. I think I downloaded the files
> January 17th. I had most of the printed
> components done, but got distracted while working
> on the Wade's gears (Arduino Inventor's Kit
> projects. Fun!). The big hangup was that I didn't
> find a parametric gear model available in SW that
> I thought was suitable for the project. I have
> that partly done, and was going to upload it to
> 3dcc. I may not finish soon, since I expect to
> receive my MakerGear Prusa before I can get back
> to it. I need to work out how to upload what I did
> to GitHub.

Could it be that the SolidWorks import plug-in for Rhino isn't ready for SW 2011 yet? I'm not a Rhino user, so I'm just guessing here. I probably should have uploaded STEP and ParaSolids versions too, just for completeness.

Dale,

Is there anyway you could save these in 2010 format? I would really appreciate it. Thanks,

Alex

Parametric modelers, as a rule, do not save to previous versions. Sorry.

Hey Dale,

If you save your files as step files, I would be happy to convert them with the feature recognition in SolidWorks and save them as 2010 .sldprts. I am working on the SAE version but the physical construction of my Prusa Mendel has taken priority.

My hope is to generate a table from which these parts will be generated so that you could load the parts as either SAE or Metric variations. Though this would really just be a modeling exercise with the openSCAD versions already generating both files.

I looks like I uploaded a folder full of STEP to github a couple weeks ago.

It's too bad about the version compatibility issues. If you use FeatureWorks, you'll lose the parameterization I built into the models to mirror the OpenSCAD code for each part.

is there a way to make a larger scale prusa mendel?

There area few larger designs described in the wiki. For modest increases in size, you can probably use all the same printed parts with longer rods and belts. You'll need your own heated platform. I'm not sure how large you can make it before you'll need to start altering the basic configuration. Have a look at the designs in the wiki, and see what you think.