Resize a STL File For Me

[MarkF]

I just finished my first model where I imported a photo to use as a reference.  I wanted to try using this capability for a while but didn't have a need.  It was remarkably easy and my first print was spot on size.

I wanted to print road crossing ramps for the track on my model railroad layout.  To start, I laid a piece of paper over the rails and trace them.  From there, I took a photo with a ruler in the view for scaling in FreeCAD.  (You need to take the photo as perpendicular to the object as possible.)  After importing the photo into FreeCAD, you need to scale it for use.  Here, I right clicked on the photo in the tree and selected the ends of the ruler and gave FreeCAD the length.  After scaling, it was just a matter of drawing the shape.

A nice bit of kit for the toolbox.

[bostonman]

It was remarkably easy and my first print was spot on size.

You're just showing off now.

Not sure if you have a scanner, but could you have used a scanner so the picture came out more perpendicular (ignoring that you got the picture spot on)? I've scanned things before by leaving the cover open and covering it with something to avoid external light getting in the scan.

Have you ever considered a 3D scanner? I've done searches for them and found many, but I don't have much of a need for one except bragging rights.

Also, I'm sure you know this, but if you need multi color prints, they can always be painted.

I haven't tinkered with the dimensions of the bench foot you made but plan to. Currently the cold and recent snow is limiting many hobbies. Although very simple and basic compared to what you (and others) created, I pieced together a plate for my Arbor Press.

My press is small, got it off CL for $35 a few years ago and used it several months ago to replace bearings in a drill press. The issue is that it didn't come with a plate or anything else. Long story short, I printed a cone shape to assist with removing some of the bearings and remained shocked how tough the print was. Not sure how strong this plate will be (I'm planning to print at 80% or maybe more infill), but it may be enough for any future projects. The slot and hole sizes are somewhat random as I didn't find any standards online. The slot on the press is about 1-1/8", so I made one slot that width.

[CatalinaWOW]

A flatbed scanner works very well for collecting geometry.  And I usually don't have to put a shroud over the object to get good enough contrast.  You do run into problems with anything that isn't flat against the scanner surface.  The problems are fairly obvious in the image and you can usually apply some Kentucky windage to work it out.  But it is fabulous for making gaskets and the like.  The newest versions of FreeCAD have improved image import dramatically.  It is really easy now.

If you have any serious work to do on your press I suspect your 3D printed parts won't be up to the task.  I have tried a couple of times to do things like this.  An example a two inch thick shim to allow use of a tool from a 6" swing atlas lathe on my 10" swing lathe.  It looked and felt seriously strong, but the first time I attempted a real cut it failed completely.

I was gifted one of the moderate cost 3D point scanners a couple of years ago.  It is pretty impressive for non-precision work.  Scanning a persons face allows making a print that looks like the person.  But it is fiddly to use, the software is temperamental, the workflow isn't totally obvious and the accuracy isn't much beyond the James Watt fit specification for his steam engines (can't put a five farthing piece between the piston and cylinder wall).  It is possible that they have improved in the last couple of years, but I would recommend saving your money.

[MarkF]

I do have a flatbed scanner, but you have to be able to get the object onto it. 
A 7x10 ft railroad doesn't fit very well.   
I don't think I would find much use for a 3D scanner.  I just don't do that kind of work.

Using a photo has a niche place in CAD modelling. 
This is the only case I have been able to use this capability.
It was a savior in this unique case to get the complex geometry into FreeCAD.

As for strength, I use mostly PLA and found it's not as strong as I would think.
I doubt a base plate for your press would survive your first press.
Even carbon fiber filament would be hard put for tough tasks (if you could even print it).

[bostonman]

MarkF, you never mentioned how you created the angle for the table saw foot. You mentioned it being beyond my level, but I'm always up for learning existing methods so maybe I can apply them in the future.

Maybe I won't remember how to do it, but at least I know what exists so I can try searching for it. Things are much harder when you don't know what to look for or terminology on what to look for.

The angle iron is only 1-3/4" of an each on each side, and approximately 0.07" thick (slightly over 1/16th" of an inch thick). It's not a very big foot, but the angle was the only part preventing me from creating one for some time.

Not sure if I'll print it using TPU or PLA. The TPU is more forgiving with sizing since it stretches and it's durable, but might grab the floor too much.

[MarkF]

If you expand the AdditivePipe in the tree, you will see two sketches.  One is the profile of the shape I want to create and the second is the path to sweep the profile over.  It's like doing a Pad except that it follows a path.  You will also note the two Pockets following.  Their sketches also have the angle needed to trim the ends of the Pipe to bend them in at the bottom of the foot.

The pink parallelogram is the profile and the yellow line is the path.



Another view with one of the Pocket sketches (green) to trim the ends:



You would change the path to match the width of the angle iron plus twice the thickness of the foot.

You would change the top of the parallelogram for the thickness of the angle iron plus twice the thickness of the foot.

I would print with PLA.

[CatalinaWOW]

Another option would be to add a solid plate to the bottom of the foot, print it in TPU to allow it to have a tight fit on the leg, and then bond a material with the properties you want onto the bottom.  Rubber, nylon, felt or whatever.  Or go the hardware store, pick a foot that has the floor interface that you want, and then make the appropriate shape so this foot sleeve mates to the foot purchased in the store.  An adapter so to speak.  One very appealing option is to make the adapter mate to one of the screw thread adjustable feet that you can buy.  Then you can level your saw even if the floor isn't level or flat.

The real beauty of 3D printing is that there are very few constraints on the shape.  The sky is the limit.

[bostonman]

The real beauty of 3D printing is that there are very few constraints on the shape.  The sky is the limit.

I fully agree and why I think this stuff is so cool. Even just spinning a shape on the screen looks cool.

Unfortunately, as mentioned, my brain lacks some creative ability (I've always been more of the person who disassembled stuff, figured out how it worked, repaired it, etc...) and I also tend to draw stuff in 2D models.

When I went for my BS, I took a drafting course and was taught to draw showing the front, side, and top view (I forget the term for this) and that's how I show most drawings when I'm explaining something to someone. Others whip out nice 3D hand drawings when they explain stuff.

Having said this, I think these are reasons I struggle with creating 3D shapes along with stuck in the mental mode of a simple 2D model that gets padded (or a revolution for a shape).

The table saw itself (a freebie off CL years ago) didn't have a fence. Recently I thought of using that aluminum stuff that can be screwed together (I don't know the name of it - it has lines in it along with holes - but have some long pieces). They are straight which may be a perfect fence. I came up with the idea to model the handle part along with the hook and maybe threaded rod to reach the other end of the table.

I have the image in my head based off Internet pictures, but struggle to come up with an actual working design. This all relates to what I stated above regarding lack of creative ability.

[MarkF]

DIY table saw fence:

 

[bostonman]

Thanks for the links. I may have seen that video before when i searched for DIY fences.

People come up with some elaborate, and simple ways of making stuff. Some though make it look easier than it is.

[bostonman]

I'm back to struggling with exporting to a STL.

As told to me, the 'tree' on the left is a cake recipe and I select the last step to export. For whatever reason the holes are not exporting.

In some cases I get half the drawing or part of the drawing.

Although many ways exist to do 3D drawings, I drew the top half and mirrored it. After I created a standoffs and mirrored those (although I think a way exists to do an array).

If anyone digs into the drawing and/or steps, I think they'll have a good laugh because I probably took the hard road including many construction lines that probably didn't need to be used.

[MarkF]

Sorry.  But, there are so many errors that I don't even know where to start.
You have a serious phobia with mirrors and you are not consistent with them.  Take the corner standoffs to start.  You mirror them I suppose so their all the same.  However, one is not like the other three.

My suggestion is the start over with a new model and place the center of the coordinate system in the center of the model.  Then, model only one quadrant with just the solid base, with one corner section and one standoff. 
Then mirror right to left.  Then mirror those top to bottom.  Afterward pocket all the holes.  And finally, do the fillets.

You MUST systematically divide the model into pieces and build it up.  I can't isolate all the pieces and neither can FreeCAD.  PAD002 should look like a mirror of PAD003 but some how it has the lower half attached.

You can't export the model because the geometry is broken in places.

[rteodor]

I would do it a bit differently using only 'Part Design' and 'Spreadsheet' workbenches. I was a bit lazy: all dimensions would have to go into the spreadsheet...
By exploiting symmetry as much as possible starting from the sketch, there is no need for any mirroring.

To be noted in the first sketch, I may have simplified it and use arches for the sides. That I think was your intention but I might be wrong.

[MarkF]

I would do it a bit differently using only 'Part Design' and 'Spreadsheet' workbenches. I was a bit lazy: all dimensions would have to go into the spreadsheet...
By exploiting symmetry as much as possible starting from the sketch, there is no need for any mirroring.

To be noted in the first sketch, I may have simplified it and use arches for the sides. That I think was your intention but I might be wrong.

That's probably the best way.  He is not really saving anything in the long run from doing the mirroring.  In the end trying to put the holes and fillets in afterward is a lot more difficult and involved then just drawing each piece outright.  In fact it is down right a struggle to do.  A bit of déjà vu for me.  You end up with something that is NOT part of a body and you can not operate upon it.  There are some tricks that I keep forgetting to make it work.

[MarkF]

@Bostonman

Here is the Display Box cleaned up.

I deleted everything except the three basic sketches:  the bottom, the top sides, and one standoff.
Then I mirrored and positioned the sketches into their correct location.  Padding each as I went.
After which, I pocketed the three holes.  Followed by the fillets.

Just mirroring and padding the sketches in a more logical order allowed me to export the last item in the tree (i.e. the fillet).  I could have just clicked on 'Body' for the export.

You really need to plan a more systematic approach as you build-up the model.
I also tweaked a few of the sketches trying to add constraints and keep the dimensions the same as you had.

[rteodor]

I would do it a bit differently using only 'Part Design' and 'Spreadsheet' workbenches. I was a bit lazy: all dimensions would have to go into the spreadsheet...
By exploiting symmetry as much as possible starting from the sketch, there is no need for any mirroring.

To be noted in the first sketch, I may have simplified it and use arches for the sides. That I think was your intention but I might be wrong.

That's probably the best way.  He is not really saving anything in the long run from doing the mirroring.  In the end trying to put the holes and fillets in afterward is a lot more difficult and involved then just drawing each piece outright.  In fact it is down right a struggle to do.  A bit of déjà vu for me.  You end up with something that is NOT part of a body and you can not operate upon it.  There are some tricks that I keep forgetting to make it work.

I feel that something is either missing from FreeCAD sketcher, or I missed it (or that is only on newer versions): like for those stand-offs, drawing one corner should be all that's needed. Then using a tool like 'Symmetry' one corner should be mirrored over the axes. But by using 'Symmetry' no constraints are transferred to the other side so the constraints must be added manually. I would expect some sketcher tool like 'Strict Symmetry' to do that so I would not have to.

[bostonman]

Sorry.  But, there are so many errors that I don't even know where to start.
You have a serious phobia with mirrors and you are not consistent with them.

MarkF, I knew my approach/design would cause a chuckle and I laughed at your reply because it had so much truth. The entire time I worked on this (let's ignore the times I started over) I kept saying: I must be doing something wrong, it shouldn't be this confusing, etc...

My suggestion is the start over with a new model and place the center of the coordinate system in the center of the model.

I believe the steps/approach was similar. I began with the base drawing only half at the X/Y lines in quadrant one and padded it. Drawing and padding rectangles I've mastered at this point - if only this was all required to be a whiz at 3D modeling.

The struggle and complete hack job was the curved part for the mounting hole. Trying to get the constraints to comply was an uphill battle. Each time I added a constraint, it would cause something else to go haywire. If I attempted to widen a curve, it would cause another line to stretch.

My initial goal was to have it look similar to this but I gave up and just made it "any old thing":

https://www.polycase.com/media/catalog/product/cache/4dcd68b41454ebf75c0e6c85988f4416/l/p/lp-31fmbt_27.jpg

After "finishing" the base, I clicked on the face and added a sketch, then drew the the walls and padded. Now the top half was complete and thought adding standoffs after was the best approach. Usually in the past, if I mirror a section that includes a hole, I get a mirrored image, but the hole is the sketch. When I attempt adding a groove to make the hole, I get an error that (I'm loosely quoting) the sketch isn't mapped to the face; so I do get frustrated with mirror because I never know what to expect.

When I mirrored the top half, everything seemed good (although my fear of mirroring is that the mirror sometimes doesn't end up on the correct axis or orientated correctly and I end up yelling at my computer as to why it's out in left field).

Finally I made the standoff and mirrored them. I did the top left first, mirrored, crossed my fingers on where the mirror would land (ironically it landed in the correct spot - bottom left), tried mirroring the top left again to make the top right, but it placed the standoff out in left field, so I manually positioned it, and believe mirrored that one to make the bottom right.

In my opinion, I completed all the steps perfectly to make the worst, most confusing, tree structure in FreeCAD history and end up with a model I can't export to STL.

Update: To add even more annoyance at my end, without thinking, I made the walls of the box exactly 2" x 3" to match the dimensions of the PCB I just made (and ordered). I should have made the box slightly larger, but knew trying to change dimensions in my drawing wouldn't work. All I was hoping for was that my printer is slightly out of cal and would print it 2.00001" x 3.00001". In other words, I was going to print and pray.

[MarkF]

The mounting tabs on the outside of the box can be just simple square corners cutoff to 45^o or any pleasing angle.  You could also fillet the points to round them if you wanted it smoother.



It is possible to constrain the dimensions of the box so that you can modify the inside of the box to accommodate the PCB.  Here's an example where you can change the sketch for the 'Pad_walls' to allow the PCB to fit.

[bostonman]

Thanks.

I wasn't trying to get too fancy, but I also took the opportunity to place a little style to it in hopes of increasing my modeling skills.

[MarkF]

I don't know what you use for creating PCBs.
I use DipTrace for making all of mine and I export them as a STEP file.
With a STEP file of the PCB, you can import it into FreeCAD and build a case with mounting standoffs.
I have done a few cases now and it's of real benefit to be able to insure everything lines up and you have enough clearance for the PCB.

This is a case I did for my railroad hand throttle:



And the finished case with knob inspired by this thread:

[bostonman]

The program I'd like to use for PCBs is Altium because it's used in industry.

Unfortunately it's rare I lay out PCBs, but I use a registered program called PCAD 2006. Years ago the company I worked for was upgrading and didn't have a need for PCAD anymore. At the time I didn't know anything about PCB layout except overhearing terminology (and working on circuits), but wanted to learn.

They gave the full version to me along with their libraries including having my co-worker/buddy to help with any questions.

Sadly it doesn't have a STEP export feature. I've only made about five PCBs since around 2008, so I haven't had a desire to learn another; especially if it's one not used in industry.

As with FreeCAD, Ideally I like to learn software used in industry so when I walk into a company, I can have experience with common software and be an asset.

[CatalinaWOW]

The user interface varies among the various CAD/ECAD programs. The fundamental concepts don't.  If you have experience using the concepts in any one of the programs you are far closer to being able to use one of the others than you were as a newby.  And even if you are using a "pro" version it only helps if the place you walk into also uses it.

[bostonman]

The fundamental concepts don't.  If you have experience using the concepts in any one of the programs you are far closer to being able to use one of the others than you were as a newby.

I fully agree and somewhat misspoke. My brain was focused on PCB software, and, since I've only made about five PCBs, it hasn't been worth my time to learn other software.

My statement was meant more for: if I'm going to learn new PCB layout software, I'd prefer to use an industry standard, but agree, what I know in PCAD is transferable to others.

I have a........... "full copy" of Altium, I think from 2014, but, when I attempted to start using it, I realized adapting to it would take long and wanted to get my board laid out quickly.

With the exception of not being able to export to a STEP file, most of what I do is transferable.

MarkF, I wanted to mention something. Where you make various stuff for your train(s), do you ever incorporate text? I've tinkered with text that extrudes out, but haven't figured out how to make it embedded.

In any case, I had an idea that when I learn how to embed text in a design, to fill it with something to make it stand out. I thought maybe if you want to have text to show "ON", "OFF", "FWD", "REV", etc... that maybe it would give your designs a nice touch. I'm uncertain what to fill it with, but I thought of some watery epoxy that would fill the embedded text so it becomes even with the surface.

[MarkF]

I actually have done two models with lettering for the layout.
In both models I extruded the letters. I painted the letters on the 'railroad crossing sign'.
I didn't paint the letters on the 'turnout motor' model because I didn't want them to stand out. 
The lettering is there to identify the turnout to the operator when they want to switch it.
You can see both of these in my previous photo.



I can't think of a model where I embedded the letters.  Embedded ones could be painted as well.
In order to embed the letters, I believe you would first need to extrude the letters and then
from the 'Part" workbench cut the extrusion from the main body.

Also in that previous photo most everything that is gray in the photo I 3D printed.

• Tunnel portals and blocks for the tunnel walls.
  
• Extended platform for freight station.
  
• Traffic Jersey barriers.
  
• Concrete curbs for parking lots.
  

[MarkF]

Nice quick video on padding and pocketing text onto a shape: