Cat hut for winter

I’m in the process of designing and building a small hut/kennel for my two cats
They live outside and need a warm place to rest, so I’m planning on a fairly large well-insulated house

Design

The build will use 9mm 3-plys OSB panels, assembled with finger joints, glued and screwed for reinforcement
Note: Different textures in the 3d model are only used to facilitate viewing the fingers

In-between walls, I’ll add some 45mm insulating foam, hence the double-walled structure

The whole structure is divided into 2 separate “modules” I can screw together, with a single common “roof” above

”Sleep module”

“Terrace module”

This allows me to have smaller parts , mainly for cost optimization (smaller stock needed)
This will also allow me to change some parts of the build if need be (eg: a larger terrace on the right side so that I can leave some food too)

Modeling

The hardest part of this project is probably to design the fingers and corresponding holes.

I failed miserably at designing them at first, wasting 2 days with manually defined fingers that would break after every change in dimensions…

Then I discovered the “BoxJoint” plugin :slight_smile:

BoxJoint Plugin

This plugin is a true holy grail !

It’s quite simple to use:

  • First just model your two boards with a simple “butt joint”

  • Click on the add-in

  • Select the two faces you want to join

  • Click OK… Boom! It’s done….
  • Need to change some parameters? No problem, double-click on the feature in the timeline
    image

And you know what’s the best part of this? It’s free! :smiley:

Dogbones Plugin

Now we have finger joints, but as you may already know, they’re not “machineable” as-is
The bit radius will keep them from sitting flush
We need to add “dogbones” to the holes and fingers to accomodate for the bit radius

I tried two plugins for this:

- Nifty Dogbones - Nifty Dogbone for Autodesk® Fusion® | Fusion | Autodesk App Store

Both plugin do the exact samle thing, and have the exact same options

Thay also have the same flaw;
it’s only a script that generates dogbones as a one-time operation, and you can’t modify the parameters later. You’ll have to delete the feature group and re-do the process.

The first one is free and open-source, and worked fine at first.
But it totally failed later in the project and I was unable to update or even re-model the dogbones, Fusion360 would just crash every time…

The second one is paid (20$) and I suspect it’s heavily based on the first one…
It’s a lot faster and more stable though, so I may end-up buying it anyway…

Here’s how it goes:

Note: As mentioned, the resulting feature will be a group of operation without any mean to edit or modify the parameters. I’d be quite ok with that for the free plugin, but I think it’s a bit of a shame for the $20 one

Layout

As usual, I use the trusty “MapsBoard Pro” plugin to create the layout for machining

Next steps ?

This is just an “ongoing thread” to document this project
I just wanted to share some activity and cool findings with thos plugins

Now I need to buy the OSB sheets, break them into manageable dimensions, dust off the LR3, and cut all of this :smiley:

I’ll keep you posted :slight_smile:

13 Likes

This is quality content! I would love to see more of this from you and from other users.

2 Likes

I’m very basic in fusion. I usually export a dxf of the sketch I’m using and then import that into ESTLcam. Does the boxjoint plugin modify the sketch or the body? If the body, then do you just export that into your cam and program it like a 3d object even though its only 2d?

same. If you keep it all in fusion I’d love to see how you do all the setups in the CAM environment,

It modifies the body directly.

I didn’t really care about this because I use Fusion360 as my CAM program directly, I find the process a lot easier and error-free

If you want to use ESTLCam though, you can set mapsboard pro output to “DXF” and work from here

Note: I didn’t mention it for clarity’s sake, but BoxJoint plugin has a dogbone feature. It will generate “3D dogbones” that are a lot nicer visually, but also a real pain to machine. That’s why I use a separate “dogbone” plugin to generate them, so that I get a simpler / easier to machine geometry, and this one will be export-able to DXF without any problem

Never tried ESTLCam so I can’t really compare, but I always found Fusion360CAM to be pretty straightforward and easy

Create your setup, add clearing and contour operations, export you NC program, done…

Rinse and repeat for every stock sheet…

Also, If I happen to modify my model, the CAM program would also update automatically, which is a big plus for me… no export/import versionning error in this process…

2 Likes

looks like I’m buying Mapboard Pro! That export Dxf would let me use Estlcam the same way as I do now. I’ll give Fusion a try as well. Though my models are often a mash of other models that require some manual tracing in Estlcam to get usuable gcode lol Need to up my skills in cad.

I love this hut. Design question… Is there a way to periodically clean it out without tearing it apart?

Yeah you really need to design and model with CAM in mind
Making the model as a “block” and then trying to split it into sheets in the end won’t work

But that’s part of the process, while I’m desiging, I’m also thinking the whole time about how I will cut and assemble the elements
All parts need to be “flat”, “extrude” is the one and (almost) only tool you should use while modeling those :slight_smile:

Keeping the model very “parametric” also helps a lot, as I can play around with the sizes in the end to make it all fit in the stock I have available, or adapt to other constraints

Some details about how I designed the parts

Initial sketches

Start with two sketches that will capture the dimensions
Think of these as your “Industrial drawings - front and side”

Note: Everything should be black, any “blue” line is not properly constrained and must be fixed

Begining the box

Then extrude the “sheets”

Notice I extruded each wall as a separate body rather than extruding a solid box, I’m modeling the sheets, not the whole object

Also note that in all extrude/cut operation I never referenced any “real” dimension in inch/cm, and always refered to projected points, so that the dimensions are all driven by the two initial sketches

I’m extruding to the max height, and then use a “cut” operation to remove the slanted part as a way to minimize operations, but you could also extrude 4 sheets from the side sketch

Finger joints are added

And a third sketch and cut for the flap

Adding some features

At this point, I already have my main features set and done, so I’ll create a group out of these

image

It really helps when you look back and make some modifications, and it also forces you to “think in features”, not mixing things up

As an example, I’ll then add the “cat head” cutout and plexiglass window, and this will also be a single group

If you select the group in the timeline, you can see the affected zone is well defined
And if I ever decide to delete it, things should keep in place :slight_smile:

Keeping some strict rules like that will greatly help you produce a truly maintain-able and machine-able model
There will absolutely be some errors and break when changing parameters (lost projections, various bugs and constraints that weren’t right, …) either because you didn’t define the “good” constraint, or because Fusion has some weir sheningans sometimes, but it will be easier to fix if the model is clean in the first place

I plan on having the roof removable
Any other part should be easily accessible from there, walls aren’t closed on top for easy replacement of the foam if need be

The whole structure will be outside, but slightly protected by another structure, so everything should be pretty safe

1 Like

While I’m showing off some tools, Let’s try something new… :slight_smile:

I’m trying to optimize the stock I need for all those pieces
My LR3 has a work area of =~ 820x1800, and I need to be able to carry the panels in my car (ideally inside the car, or on the roof) over 50km

Normally, I would look toroughly at the sizes, grab a pencil and paper, and think hard about this…

But now we have AI :smiley:

I want to optimize the CNC cutting of wooden pieces in 2500x1250mm panels.

The dimensions of the pieces to be cut are as follows:
- 782x578
- 398x728
- 318x728
- 398x578
- 316x368
- 206x190
- 398x578
- 398x578
- 398x578
- 398x578
- 398x578
- 368x578
- 325x618
- 618x391

My machine has a maximum cutting area of ​​800x1500

How many panels do I need?
How do I cut them to make them easier to transport (with a minimum number of cuts per panel) ?

Here’s the response I got:

To optimize the CNC cutting of your wooden parts in 2500x1250 mm panels, while respecting your machine’s maximum cutting area of 800x1500 mm, here’s a structured approach:

1. Constraints to Respect

  • Panel size: 2500x1250 mm.

  • Maximum cutting area: 800x1500 mm.

  • Goal: Minimize the number of panels used and the number of cuts per panel to facilitate transport.

2. Parts Analysis

You have 13 parts with the following dimensions (in mm):

  • 782x578 (1x)

  • 398x728 (1x)

  • 318x728 (1x)

  • 398x578 (5x)

  • 316x368 (1x)

  • 206x190 (1x)

  • 368x578 (1x)

  • 325x618 (1x)

  • 618x391 (1x)

3. Cutting Strategy

Since your machine can only cut a maximum area of 800x1500 mm at a time, each 2500x1250 mm panel must be divided into smaller sections that fit within this limit.

Step 1: Virtual Panel Division

A 2500x1250 mm panel can be divided into:

  • 3 sections of 800x1250 mm (since 2500/800 ≈ 3.125).

Step 2: Optimizing Layout

Group parts by size to minimize waste.

Example Layout for One Panel (2500x1250 mm):
  • Section 1 (800x1250 mm):

    • 2x (398x578)

    • 1x (398x728)

    • 1x (318x728)

    • 1x (316x368)

    • Remaining space: Can be used for other parts or will be waste.

  • Section 2 (800x1250 mm):

    • 2x (398x578)

    • 1x (368x578)

    • 1x (325x618)

    • Remaining space: Can be used for other parts or will be waste.

  • Section 3 (900x1250 mm, since 2500 - 2x800 = 900):

    • 1x (782x578)

    • 1x (618x391)

    • 1x (206x190)

    • Remaining space: Can be used for other parts or will be waste.

Estimated Result:
  • Panel 1:

    • Section 1: 2x (398x578), 1x (398x728), 1x (318x728), 1x (316x368)

    • Section 2: 2x (398x578), 1x (368x578), 1x (325x618)

    • Section 3: 1x (782x578), 1x (618x391), 1x (206x190)

  • Panel 2:

    • Section 1: 1x (398x578) (only one left after Panel 1)

    • Remaining space: Can be used for other parts if needed.

4. Number of Panels Needed

With this approach, 2 panels should be sufficient to cut all the parts, with efficient space usage.

5. Simplified Visualization

Here’s an example layout for Panel 1 (adjust as needed):

Section

Parts (LxH)

Remaining Space (LxH)

Section 1

2x(398x578), 1x(398x728), 1x(318x728), 1x(316x368)

Minimal waste

Section 2

2x(398x578), 1x(368x578), 1x(325x618)

Minimal waste

Section 3

1x(782x578), 1x(618x391), 1x(206x190)

Minimal waste

6. Transport Recommendations

  • Group parts by size for easier handling.

  • Cut panels into 3 sections (800x1250 mm) to fit the machine’s limit.

  • Use spacers or separators to prevent movement during transport.

7. Verification and Adjustment

For precise optimization, use nesting software (such as CutList Optimizer, DeepNest, or OptiCut). These tools automatically visualize the layout and minimize waste.

How cool is that? :smiley:

It even generated a visual representation when I kindly asked… although it doesn’t make any sense!!! :smiley:

I managed to make it output a simplified cutlist though, and it doesn’t seem too bad

OK, let’s plug that into MapsBoard Pro

Mmhhh not quite, but we’re almost getting there….

Manually adding a 800x1250 section solves the problem and we still have some left…

Now I only need to find a way to cut those sheets on the shop’s parking lot :stuck_out_tongue:
Hint: I made a small purchase but shhhh… don’t tell my wife :stuck_out_tongue:

If it was the Einhell she won’t notice, if you also bought the Festool she will because there isn’t any money for food left. :sweat_smile:

1 Like

So… that’s 04h42 machining time at least… whooo…
@Tokoloshe can you drop in real quick and lend me yourt fast LR4 ? :smiley:

And nope, it’s “only” the Einhell (it was a steal… 50€…), the festool one I bought long ago (with tracks and an MFT3)
No regrets, great tools, and they’ve seen some use :smiley:

I guess it isn’t faster than yours, I am just stupid enough to run it faster. :stuck_out_tongue:

1 Like

Put a few 2x4’s to prop up the sheet and cut with one of the Kreg edge guides. Should fit on your nice new tool. Sure people look at you but awkwardness / shame at breaching social norms are personal feelings, you can learn to ignore them with enough practice.

Concrete curbs work too, or so I’ve heard. :wink:

1 Like

Not a big fan of the Kreg system, It’s quite pricey and bulky I think

I already have this Parkside guide rail/clamp that works pretty well.
It’s very light and portable, and at 8€ it won’t break the bank

Only thing I’d change would be some kind of way to see where the blade will land (a bit like a track saw rail)

I’m just not quite sure it’s long enough for the sheets I want to buy, I’ll need to check
If not, I think I’ll just build a quick and dirty plywood tracksaw as a makeshift…

EDIT: Max length is around 90cm… I need 125cm… plywood track it is then…. or maybe a DIY system similar to the Kreg one for my one-time use… I’ll see…

I use a bit of 3mm MDF with a straight bit of 16mm mdf glued to it! :rofl:

Exactly my thought… and use the saw itself to cut the excess so that it “auto-calibrates” itself…

But do I have a 150mm long (and straight) mdf strip left? that’s the whole point :stuck_out_tongue:

1 Like