I’ve had several questions asking what my design process was for creating the 3D models I print. Stuart and I agreed that it would be a good topic, but until now I really didn’t have a good example project that was fresh in my mind. Then I was notified about an Instagram post by @jeremy.joseph:
The video post shows where he used wooden cleats to make a Milwaukee Packout hanging system. I was asked by another person if it was possible to 3D print something similar.
Advertisement
A small caveat before I start: as with many things, this is the type of skill that takes practice and experience to develop and hone, and I am in no way a master. There are probably easier and better ways to do things, but that doesn’t stop me from trying nor should it stop you.
Design Goals
In the past I’ve hesitated to design a tool box hanger, because I wasn’t sure that a 3D printed part would be strong enough to withstand the amount of force a loaded Milwaukee Packout toolbox or organizer could put on a hanging cleat. I didn’t think fused thermoplastic would stand a chance, but when I saw the Packout cleats made from thin wood, I realized that maybe I could have been wrong.
I didn’t make a formal list at the time, but as I set about designing the part I kept these criteria in mind.
The part should:
- Be as strong as possible
- Maximize contact with the slot in the bottom of a Packout box
- Require minimal material and print time
- Be easy to print with no supports
The “strong as possible” criterion is pretty self explanatory – it just means that I needed to make sure the areas that would come under stress were as thick as possible. I needed to maximize contact with the slot in the bottom to spread the force over a wider area, because the wider the area the force is spread over, the less likely there would be a point of failure under normal use.
It’s always a good idea to minimize the amount of material used, as the less material used, the lower the part will cost to make. Also, using less material can – but not always – reduce print time. Not having to print and remove supports also reduces the amount of material used, and the amount of time required to make the object.
Advertisement
Design Programs
I primarily use two different programs to design models: openSCAD and Fusion360.
openSCAD is basically a programming language, where you create designs by coding different shapes and then combining and manipulating them. The power of this program is that you can create a model that is easily to manipulate by just changing a few variables. Models written using this code can be modified by users on Thingiverse.com to create customized versions of your item.
OpenSCAD is completely open source, with no cost, and you are completely free to commercialize your designs any way you want. It works on Mac, Linux, and Windows computers, and doesn’t require a huge amount of processing power to run — except when you actually preview or render the model.
Fusion360, on the other hand, is more like a conventional CAD program. The basis for Fusion360 models are sketches, which are similar to the 2D CAD drawings you are probably used to seeing. Once you have your basic two-dimensional sketch, you can extrude the model into three dimensions, manipulate it, and even add more sketches to define the profile of the model from different angles.
Fusion360 runs on Mac or Windows computers and requires a 64 bit processor. It is not completely free to use in all situations, but there is a free startup license for companies under a certain income limit, or hobbyists. This means that if you want to make money off your models, you are somewhat limited unless you purchase a license.
Why do I use openSCAD sometimes and Fusion360 other times? I tend to use openSCAD for simpler models that I want people to be able to customize. It’s also much harder, but not impossible to create fillets, concave surfaces, and chamfers in openSCAD. When I need to design something more complex, that’s when I will use Fusion360.
In this case, there’s really no reason to customize the parts, and I also planned on making use of fillets to make the cases slide onto the cleats easier, and so I used Fusion360.
Cleat Revisions
To hold and retain the Packout cases, there are two separate parts: the latch catch, and the cleat. The latch catch isn’t always necessary and isn’t the critical component — that is if I couldn’t make a usable cleat, the latch would be irrelevant, so I concentrated on the cleat first.
My first revision was based on the wooden cleats that @jeremy.joseph made. After modeling the simple cleat, I found it didn’t fully engage the slots in the bottom of the Packout case. It turns out that he actually mislabeled one of the dimensions, but by that time I had already started measuring to refine the part.
After printing my first version of the cleat, one thing I noticed when I looked at the Packout cases was that the front and back slots weren’t flat, they had this ridge that was preventing my cleat from fully engaging the slot. I didn’t want to make the cleat thinner because then it would just rest on that ridge, and a thinner part would also be more prone to failure.
My second revision incorporated the notches to avoid the ridges in the case slots, and I also reduced the size of the overhanging tabs to remove material. When I slid the cleat into the bottom of the Packout case, the cleats still weren’t fully engaging the slot. I noticed that the end of the slots were sloped, so I filed away the front of the tabs to make them engage even further.
For the third version of the cleat, I further removed material by rounding the back part, and I incorporated an angle into the front part of the cleat.
At that point, I was ready to actually try mounting the Packout cleats to see how well they worked. For this, I switched filament types, from PLA to a stronger type: PET-G — it’s basically the same material that water bottles are made from. I also decided that I would add a slope to the outside of the tabs to assist in aligning the Packout case.
When I actually mounted two of these cleats to a board and tried sliding a Packout case onto them, I still had problems getting the case to align with the cleats. At this point I decided to redesign the cleats to make aligning the Packout boxes easier.
I cut off the back half of the cleat because it wasn’t serving any purpose, and extended the front to catch in the valley between the slots. This made it much easier to align the Packout case, as it meant I needed to be close, but not exact. As I pulled the case forward onto the cleats, they would snap into perfect alignment.
In my last revision I just tweaked the height. You can see in the above photo that the base of the Packout wasn’t resting on the board, so I just lowered the total height.
To mount the cleats, I inserted them into the bottom of a Packout case and covered them with carpet tape (double sided sticky tape), aligned them on the board exactly how I wanted, pressed down hard to get the tape to stick, and removed the case. Then I screwed the cleats in place.
Latch Catch Revisions
With the cleats finished, I started work on the latch catch. I needed to modify @jeremy.joseph’s design, because I couldn’t figure out how to correctly model a countersunk screw head on a ramp. It seems like an easy thing to do, but for some reason I couldn’t get the bore and the countersink to align just right.
Instead, I started out with a flat spot on the top of the catch to screw it in. I also added tapered sides ending in wings to direct the latch into the right position.
After printing the first design, I found that the latch didn’t behave like I thought it should. The ramp and flat spot ended up being too long, it interfered with the cleats catching the bottom of the box correctly. To fix this I made a steeper ramp, but this made it more difficult to push the box over the catch.
In my final design, I widened the entire catch and placed the screw holes on the sides. This allowed me to decrease the length of the ramp while still having a slope shallow enough for the box to easily slide over.
The black latch catch is actually the same version as the orange catch. I purchased some black PET-G to give the end product a bit more professional finish.
Finished Product
After 6 cleat versions and 3 latch catch versions, I had a working system. Above is a demo of how the Packout cleats work. I’m only showing a half-width box, but it should also work with a full-width box just the same.
I’m still not convinced that these cleats would be strong enough to hold a fully loaded Packout case in place inside a moving van, but I think they should hold up to light use in a garage or workshop if you don’t overload the cases or try to mount one of the large boxes.
I’m thinking about modifying the design, so that I can cut the cleats out of aluminum on a mill, but that’s another project.
I’ve uploaded the STL files to Thingiverse.com (see link below) and I’ve included the Fusion 360 archive files if you want to see how they were drawn.
Download Files (via Thingiverse)
Warning: This is a 3D-printed part, and as such it is NOT suited for high strength or heavy load applications.
Parting Thoughts
Hopefully I’ve been able to give you some insight into how I design parts for 3D printing. Not every project requires the same process, but in general I find that an iterative process where I focus on only a few parameters at a time works best for me.
I find that making a crude model first allows me to see how the part will interact with the real world and helps me solve problems early in the design, before making major changes would be painful.
Still, you shouldn’t be afraid to redesign a part or throw away a design completely. While it might seem like a waste, the insights you gain from your previous iterations give you valuable knowledge about the problem you are trying to solve.
Brian A
Sweet!
I dont know what the holdup is on them releasing the packout wall mount plates and the damn packout flat dolly I am dying for. Was Q3 2018, then Q4, now supposedly Q2 2019. Any inside info?
Bill
I honestly don’t understand the design of the packout system from Milwaukee. If the idea is to be able to stack holders onto a hand-truck and carry all of your stuff, congratulations. However, why include a handle like a suit case? Why not make foam inserts available. I am sure that I just missed all of that, but now I am guessing not if people are trying to make their own inserts.
Doresoom
Milwaukee has come out with their own foam inserts for the Packout cases recently, once they realized everyone was using Kaizen Foam with them. I’ve used the Kaizen and it’s great, but haven’t used the Milwaukee foam yet.
Brian A
Its identical, I got foam from 2 sources and the genuine Milwaukee packout tool case which comes with foam. Cant tell the difference. Sorry Milwaukee for saying this lol
Bill
Thanks. That is good to know. Didn’t realize that had done that. I just knew that I was being left out of something. 🙂
James C
I don’t yet have the cart. I have three small organizers and one low profile small organizer. I like the briefcase handle because it’s easy to lock together a few organizers and carry it as one unit.
Chris
Check out your local woodcraft or rockler.
They had the best prices I could find on the foam you would need.
This is my DIY job.
https://imgur.com/2BpksV0
PHILIP JOHN
I modified my pacout to fit under my truck cover. Had to modify the handle. I did it so I can keep the handle in its slots.
Jim
that’s my only complaint about the packout. its too tall with the handle to fit under the bed cover. What did you do ?
PETE
I believe they have/are updating the rolling box to be able to remove the handle so it’ll fit under a bed cover. I believe i saw it on instagram?
Benjamen
Maybe you saw this post from @toolaholic:
https://www.instagram.com/p/BvrQu-fnNxM/
He’s visiting Kaizen Inserts and Brian is showing him two of the new products his company is coming out with. One of which is an aftermarket way to remove the handle.
Corey
Man, DeWalt’s had that equipment forever lol just some ribbing, but for real my tough system dolly in the back of my truck is 5 years old now after 3 companies, countless industrial jobs, 3 moves, and it still works perfect. I’m quoting a bulk order of the wall racks for my job trailer soon, too. These kinds of storage options really do make life and work just a little easier.
Frank D
Great post!
I don’t have any particular mounting needs, other than perhaps creating a four wheel dolly, but will have to check out Fusion 360.
TonyT
An important note on strength: FFF (filament) printed parts are anisotropic: the parts are weaker along the Z axis than X or Y, so how you orient the parts on the printer bed impacts strength as well as need for supports.
As you’ve shown, it is very possible to modify Fusion 360 designs, but you need to run Fusion 360. Part of the original idea behind parametric CAD was to be able to automatically create families of parts (using a table or spreadsheet to drive the key parameters).
So far I’ve mostly used DesignSpark Mechanical for my 3D designs; it’s kind of like SketchUp for mechanical design (you can do a lot of pushing and pulling), but there’s definitely still a learning curve for doing more complex parts. It’s free but not open source.
I’m going to start learning Fusion 360 because it has CAM support , and I can’t think of any other free or affordable program that does.
Benjamen
Yes, and some filaments are more prone weakness in the Z axis than others. The worst I’ve seen is “wet” nylon (nylon that’s absorbed humidity), it peels apart layer by layer with your fingers. The reason I chose PET-G is that the inter layer fusion is pretty darn good, better than I’m able to get with ABS.
It also really depends on printer settings, Stuart and I discussed a section on printer settings, but it was really beyond the scope of this post.
That said I’ll rotate a print to prevent the z-axis from getting stressed, but the cleat would have needed too much support and probably taken 2x as long to print. And even then I wouldn’t subject it to a fully loaded Packout box
Speaking of CNC, Fusion360 is also one of the only free programs that does engraving/v-carving too. And simulation…that’s next on my list of things to learn.
TonyT
Nozzle size and layer thickness also have an impact – bigger nozzle & larger layer size means better adhesion between layers and stronger Z axis.
The engraving & carving is likely tied to the CAM support. I’m trying hard not to look too much at the Sienci LongMill CNC router Kickstarter ($1300 CAD for 30″x30″, but I don’t really have time or room for something like that yet….)
Drew M
Interesting.
“The LongMill competes against several machines that exist in the midsize hobby desktop or benchtop CNC routers. You might have seen by now that the market is dominated by the X-Carve and the Shapeoko 3, with machines like the Stepcraft following along.”
Having seen and talked to people who “used” an X-Carve, I would say that they’re setting the bar pretty low to try to develop something better.
I get that people want an inexpensive but good cnc router however, I honestly believe there are compromises on rigidity and control hardware that can’t be made and have the final product still be useable.
Using angle aluminum for the ways is an interesting idea IF there is a tube of some kind to give it better rigidity in torsion. I would also be concerned about vibration on that design.
IMHO, if a cnc machine can’t handle a full spindle load cut in all directions and positions without significant tool deflection then the designer failed.
Chris
You could look into something more DIY…
https://openbuildspartstore.com/
TonyT
That can be more expensive (but more educational) if my time is worth something.
Chris Ball
Just to quickly add to the debate, with careful design and a well tuned machine, severe use parts can absolutely be printed. Right now my skid steer has a nylon bushing I printed for the door track that has been in use for about three years (300 something hours), replacing a 175 dollar assembly. My truck has a couple ABS parts that I printed that are probably 7 or eight years old now. Those I fully expected to have to replace every year or two and except for one or two I have lost over the years I have had no issues (they hold down roll doors that otherwise vibrate loose and the shock cord tying the two parts of the assembly together definitely is only good for a year or so).
In my experience, at least for ABS and Nylon, if you stress test a couple samples and are confident in them, their properties are pretty stable over time, they won’t be as strong as the same part injection moulded but since the design is up to you, most of the time you should be able to make what you need. Clearance and other constraints may or may not make any particular project impossible for 3d printing but you can say the same for injection molding or aluminum for that matter.
FixItChuck
Yup, basically what mechanical engineers do day in and day out… except we kind of have to hit the mark on the first try. With million $ + hardware and reputation and etc on the line, making sure things fit and work as intended on the first try is pretty important.
Zachary Drumm
Not sure how you guys make that happen. I mean, it is so easy to overlook a little piece like that lip he had to redesign 3 times (woulda taken me more tries than that!).
How do you catch those little things on the first try? 3d imaging? Clay modeling?
Kent
>> This means that if you want to make money off your models, you are somewhat limited unless you purchase a license.
Last I looked, a paid license for Fusion 360 is required at $100,000 of income from your company. At that point, it’s $50/month or $500/year. You could sell 20,000 of those brackets at $5 each before you needed to drop $500 on a license.
For a product from a for profit company, this is simply an amazing deal, and in no way does it keep people from working or tinkering. Also part of the deal is really good CAM software (Autodesk bought HSM Cam a few years back), which is usually another few thousand dollars.
Sure it doesn’t run on Linux, but if you Linux, you’re used to that by now.
Except for the very small percent of people in the world who think *everything* should be free, there’s not many valid complaints about F360 or the licensing agreement.
Benjamen
It maybe wasn’t the best way to phrase it, but the intent was to warn people that if you were using it for commercial gain, there’s possibly a price. I’ll think about rewording it.
The license terms are based on total company sales of $100K. If you have one employee, there’s a good chance you are over that limit. If you are selling 3D designs on Etsy as a side business, you’re probably not.
https://knowledge.autodesk.com/support/fusion-360/troubleshooting/caas/sfdcarticles/sfdcarticles/Do-I-qualify-for-free-use-of-Fusion-360.html
https://www.autodesk.com/company/legal-notices-trademarks/terms-of-service-autodesk360-web-services/autodesk-web-services-entitlements
I’m not arguing that Fusion360 isn’t a good deal, and I’m all for paying for good software, just remember you need to be aware of the licensing requirements and that they can change. With software like openSCAD, you don’t have to worry about that.
I’ve been scouring the forums and Autodesk says that as a hobbiest you can use it for one year free then you can apply to extend your subscription. They are very carefully wording their agreement to reserve the right to charge you after you’ve used it a year. But they haven’t yet.
As far as I can tell the Hobbyist and Startup groups get the “startup” license. I’ve reapplied for a subscription to the “startup” license a few times and they ask you some questions that I can’t remember.
Kent
I’ve been using it for at least 5 years, and the questions have never been more difficult that “are you making more than $100,000?”.
I don’t want to be argumentative with you, but if you’re selling a product and bringing in over 100k, you should be able to drop $500 on software.
This is perhaps the least limiting professional software license I’ve ever seen.
TonyT
For me, the great part of F360 are the advanced features like CAM and simulation. I do wonder what Autodesk will do if everyone switches over the F360 – it’s A LOT cheaper than their other software such as Autodesk, Inventor, etc.
If you don’t need CAM, etc, there are a number of good alternatives, especially on Windows. For Linux users, the pickings are pretty slim; off the top of my head, I’d say the only affordable alternatives are VariCAD (which supposedly isn’t the most intuitive) and BrisCAD (which can do 3D, but it’s not the traditional SolidWorks style).
If you have a iPad Pro, there’s Shapr3D.
Discalimer: my experience is limited to a bit of DesignSpark Mechanical, Alibre, and ZW3D, in the near future I’ll be adding F360 and maybe openSCAD
The What?
I hate to piss on your 3d printing parade, but why didn’t you just make a shelf to put it on? Why would you opt to not have a shelf to conveniently place the box wherever you want standing up or laying flat? What do the cleats have to offer over the convenience of a shelf. Even better question is why does it have to lock into place when you set it down? What’s wrong with just setting it down? Having to line up the box with the cleats every single time you go to reach for the box would get old pretty fast. I could see a use for something that in a work van where you have a rail to mount things on or a rail on a wall so it’s not in the way. I get that you own a 3d printer and you wanted to attempt to emulate what someone else did,but I would certainly trust a piece of wood to hold the box up on a flat surface without falling off as opposed to a piece of plastic. I don’t think that was a piece of thin wood that the guy used to make his cleats either. It looked like he used a piece of 2x and cut it to size and routed the sides to fit the box onto and left enough meat in the back to offer enough rigidity to where he could drill and countersink the screws without it compromising its ability to hold itself up while placing a load on the front of it vertically. The rigidity along with ductility wood has is superior to plastic except for something like luon or mdf. You don’t see plastic being used to bear the tensile or shear strength of a load on any sort of structure.
Benjamen
It’s obviously something people want. Milwaukee is eventually coming out with a wall hanger. Kaizen Inserts just introduced a wall hanger. My Thing on Thingiverse has 600 downloads.
It doesn’t have to lock, you can use the cleats without the lock. Some people may want it to lock
I explained that I redesigned the cleat to make it easier to align. You can actually come at the cleats from different angles and offsets and it will pull the box into alignment.
It’s not about the total thickness of the wood, it’s about where it’s thinnest: the tabs that hold the box. They are under 1/4″ of an inch thick (about .215″) If the wood grain is aligned the long way, you could easily break a tab.