I’ve been slowly working on the layout for my custom build. I’m really stuck on the idea of having some sheet metal parts machined and folded for the build. Has anyone else in the community done this? Is there a large cost (I know this is highly dependent on the part geometry but did you feel it was outrageous )? Any suggestions/comments are welcome it is always nice to have other points of view.
Few things:
- The actual build will have a longer linear stage, right? The aspect ratios shown here will give you a very small working area.
- It might be fine, but I think you’re going to see undesired flex between the belt and arm joints. Because the carriage joints are offset from the line of action on the belts, there’s going to be quite a bit of torque on your short bearings and bearing carriers. A longer bearing and longer carriage could be a good idea.
- Are you going to add an additional constraint to keep the end-effector parallel to the rods, or are you going to place the nozzle coaxial with the end joints on both arms like the Morgan does?
I think +Peter van der Walt has some experience with this…
Custom sheet metal and machined part pricing is based on a number if factors, mostly ones that are stuck as is for what you want. Material price is obvious, 6061 or 1060 aluminum for machined parts and 5052 aluminum for bent parts? See McMaster or onlinemerals for rough pricing.
Material choice also affects machining costs. Some materials are annoying to machine. Aluminum is easier to machine (though gummy when cutting and can’t be abrasion wheel cut) but harder to weld.
Then finish, anodized, painted, powder coat, it as machined? My advice, just spray paint yourself.
Quality and tighter tolerances increase price too.
Part quantity also affects price greatly! 1 of an item often costs as much as 4 of the same item due to set up costs.
I made my own metal parts, but my tolerances were not great due to my tools. I have priced out professionally machined parts and it was costly.
It may be wise if you don’t have machining experience to look for local machinists for help, they are everywhere. If they have material laying around they may be willing to give you fair pricing, but for a machine like a 3D printer, the amount of parts you need, you may be looking at $200-$500 of machining at a rough guess. Avoid welding and it’ll save a lot of money, especially in aluminum. Try to machine less even at the cost of material, machining time costs more than eaw material. Avoid post processing steps when possible. And try to minimize set up steps. So, machine a part all from one side, flipping it over gets more costly, as does changing tools.
If all your sheet parts are laser cut and bent from one side it will be cheaper. Tapped holes will increase cost, as well complex bends.
Finally, I’ve found the overlap between sheet metal houses and machinists is small. You may need two different people. The sheet metal house I use claims to do machining, but anything beyond basics they outsource and that costs more than you going direct.
@Ryan_Carlyle Yes, the plan for the final build would be to have a ~600 mm rod carriage in both the x and y direction and a ~300 mm arm length. Which should mean I could get about 300 x 300-(effector height) x 600 - (y carriage thickness) build area.
I’ve certainly thought about the torque, but also was trying to minimize the total carriage thickness…if you can see a better solution I’d be up for suggestions. I’m not entirely pleased with the design of the x carriages yet either. I’m not enjoying my options for securing the belt to the carriage.
I’ve thought of it, but what are the issues associated with no cooling fan (both for the hot end and extruded filament)? Are there any? Also, I’m not sure there is an easy way to do this sideways.
Yes. I’ve had a printer frame of my own design laser cut and bent from 1/8" aluminum 5052. @Clockwerk_Mech arranged it and the parts came out great. Hint: get the holes cut to the tap diameter but do the tapping yourself. Some combo drill-tap bits from Harbor Freight are great for this. 5 sets of parts is usually a better deal than 1 so if you want multiple printers or can get a few people to buy in and split the cost with you, that’s an interesting way to handle the finances.
Can also spray paint or powder coat yourself.
Oh, if you can do the design without bending, get some 6061 aluminum and cnc it yourself.
@raykholo best place in the USA to get 12x12 1/8 plates of 6061 alum?
@Ariel_Yahni_UniKpty I’ve gotten 3/16" and 1/4" on eBay cheap. 1/8" tend to be a little more expensive. There’s a seller called metal remnants or something like that.
Might be a bit warped. Onlinemetals if you need something close to perfect.
Great advice here. My advice is to print the design first- if you have a printer. On a new design, you WILL find problems and need to change things s few times. Wait until you know it works well first. Then I’d make the machine in metal yourself with aluminum plate… Hack it out any way you can. A drill press and a jigsaw go a long way. This may turn up more changes. Setting yourself up to build in metal will be so valuable if you are going to be making machines. Then when everything is sorted, you can get it made somewhere. My $.02
Brook
Some more advice… Laser cut steel is very cheap (often cheaper than buying the material only at a big box store). Laser cut aluminium is harder to cut and thus thickness limited (3/8" max usually). Bending is where it gets more complicated/expensive.
Like @Brook_Drumm says you can try printing it first but any large scale parts will be difficult to make accurately.
To be honest, my “making/creation time” is very limited so even with the extra cost I’ll get parts laser cut every time, it is always accurate and saves way too many hours to count. (even if it takes multiple iterations)
Some data points- 400x400mm sheet of 1/16" (16ga) steel laser cut is only about 20-25$. A 400mm 2" x2" 16ga steel tube laser cut on all 4 sides is around 50$.
Sounds great everyone. My first two years at work were as a mechanical designer, but the tolerances and materials we use are not really comparable to this type of small scale project. So far the design entirely avoids tapped holes. I do plan on prototyping with my current printer just to make sure assembly and tool clearances are where I want them.
@raykholo my professional experience has thus far taught me not to tap into bare aluminum if you want to have a long lifespan item which may have to be disassembled multiple times (this is due to the hardness mismatch between aluminum and the steel your fasteners are most likely made from). Of course thread degradation will still take a number of cycles…but theoretically bolt preload, vibration, and possible human over- torquing will lead to stripped threads.
@Neal_Grieb true. And heli-coils and pem nuts are expensive. Nutserts aren’t so bad I guess. But for 3d printers yeah, bolts and nuts make more sense for parts that get serviced often.
Also, make sure you cut the holes too small and bring them up to the right diameter and concentricity after the bending is done. Unless you get it bent on a jig(expensive), you are never going to get it concentric and parallel with just bending. We bend a bracket for extruder stepper and the two holes, base and shaft, need to be concentric. It increases cost by quite a bit. So we post process to align the holes and then anodise.
There are quite a lot of plans out there to build a sheet metal brake so you can do the bending yourself, but you will need a mill to do the holes properly.
After doing the math, it seems my original intuition was correct. A quick bit of Python scripting shows that using a 200 mm arm length, 7.5 mm stepper pulley diameter, and assuming that both steppers move outward at the same time, I only get about 48 mm of sub 100 micron resolution (if the steppers have a 1.8 degree resolution). I could augment this by having only a single arm move, but that seems very complicated.
I’ll probably scrap this idea and focus on something simpler. Thoughts?