PDA

View Full Version : Newbie with a few questions about 3d printing. Please help!



brad91
12-18-2016, 07:35 AM
Hi everyone,

Im new to the whole 3d printer world and looking at building one to start a business. I am a design draftsman currently so know my way around autocad and inventor.

Im wanting to build parts for cars. Replacement parts that you can no longer buy and also new things. Anyway a few questions I have:

Has anybody ever tried to print a complete bodykit for a car? (fiberglass sucks so was hoping I could build something big enough and print entire bodykits?) Ive attached a picture of what I am looking at printing. I feel SLS would work better but I dare say id need some serious amounts of material to fill the bed haha.

If so, what plastic would I need? It would need to have a little bit of flex to it.

What style of printer is best? From my research the only affordable and way I can see building something as big as I want is to use FDM but if anyone thinks I could build a SLS printer im all ears as I beleive the finish will be a ton better?

What sort of finish am I going to be able to get? If im doing parts for cars to paint and finish how smooth am I going to be able to get the finish? I know it will need finishing but if I can get it good enough I can throw some gel coat/hi build primer over that would be fine.

Is there anything else I need to know? A mate of mine has built a cnc router and is giving me his contacts for the extrusions and ballscrews I will need to build the frame easily. I can also cut all the parts I need his router and can get parts machined up.

Also to anyone who runs a business. Is there much of a market for 3d parts? Especially on the larger scale?

Thanks heaps guys,

Brad

Jim Dawson
12-18-2016, 12:48 PM
What you propose sounds like an ambitious project. There are some auto manufactures that are experimenting with 3D printed body parts, but I don't know how successful they have been thus far. A couple of things that they have going for them is almost unlimited funding and the ability to hire the top talent in the field as well as having the supporting infrastructure.

Having said that, I guess it's possible for an individual to tackle such a project. First you might build a smaller printer and make some 1/10 scale parts to get the technique down. This would give you an opportunity to learn about the available materials and how they react under various conditions.

As far as market, I'm sure that if you can provide a quality product then people will be happy to buy your parts.

dharmic
12-18-2016, 05:19 PM
Ambitious, indeed.

Most 3D printers don't get anywhere near the build volume, so you'd be looking at a custom and quite large printer to begin with. And they're not built that big for a couple of reasons:
1. If you start to factor in your time, it's just not going to be worth it. A 2 litre part can take 24 hours to print, a bumper is going to be more like 2 weeks to print. And you cannot leave them unsupervised - if anything goes wrong and you're not there you're looking at some real fire hazards.
2. Physics. The plastic cooling after it's been extruded creates some serious internal stresses in the part. Often, on a smaller bed, they're enough to buckle the part enough that it pops off the print bed. Now you're talking about something an order of magnitude larger than the usual printer and I'd be thinking the buckling would be inevitable.
3. Forget production finish. Even the finest print at 0.05mm layer height (which will be even slower again) leaves a pretty rough finish by comparison to a moulded product.

Materials - I use ABS or Nylon for smaller structural parts. Both have strengths and weaknesses and quirks which add complications to the printing process.

I think you'd be much better off for stuff on this scale looking at a router to make moulds and either lay glass or look at vac-forming ABS or HDPE sheet.

underthetire
12-18-2016, 10:50 PM
Look at this, it may give you an idea. They use a glass/carbon infused filament. This machine is huge.
Shelby Cobra (http://web.ornl.gov/sci/manufacturing/shelby/)

Sent from my A3-A20FHD using Tapatalk

brad91
12-19-2016, 09:26 PM
Ambitious, indeed.

Most 3D printers don't get anywhere near the build volume, so you'd be looking at a custom and quite large printer to begin with. And they're not built that big for a couple of reasons:
1. If you start to factor in your time, it's just not going to be worth it. A 2 litre part can take 24 hours to print, a bumper is going to be more like 2 weeks to print. And you cannot leave them unsupervised - if anything goes wrong and you're not there you're looking at some real fire hazards.
2. Physics. The plastic cooling after it's been extruded creates some serious internal stresses in the part. Often, on a smaller bed, they're enough to buckle the part enough that it pops off the print bed. Now you're talking about something an order of magnitude larger than the usual printer and I'd be thinking the buckling would be inevitable.
3. Forget production finish. Even the finest print at 0.05mm layer height (which will be even slower again) leaves a pretty rough finish by comparison to a moulded product.

Materials - I use ABS or Nylon for smaller structural parts. Both have strengths and weaknesses and quirks which add complications to the printing process.

I think you'd be much better off for stuff on this scale looking at a router to make moulds and either lay glass or look at vac-forming ABS or HDPE sheet.


The space of keeping the moulds would kill me though. I can have a shed with one printer in it. Or a shed with a couple moulds haha. Im looking at doing it for the drift/race car community so alot of the stuff would be skyline and silvia/180sx kits. But each car can have 10 different kits and if you can only offer one your market is going to be very little. If I could just print off whatever that would be awesome.

I wasnt aware of the unsupervised bit though... I was hoping I could just start it up and still go to a day job and come home to printed parts.... How flammable is the stuff?


I saw the shelby that was printed and that was sort of my inspiration. If I could get it smooth enough that hi build primer could fill the imperfections that would be great but as you said it would take weeks which for a $300 sell price isnt economically viable.

Being able to print reproduction panels for classic cars would be great though as well. There is a huge market in australia for classic holden panels. Steel guards are around $800ea. fiberglass is around $800 a pair and they need a fair amount of work to look good before going on a show car.

The advantage I can see to plastic bodykits is that they can have a nudge before cracking like fiberglass does. Standard bumpers can take a decent hit wheras fiberglass kits like I run only need a slight bump and you are spending a weekend repairing it.

Another advantage I could see to the size from a business perspective is clients. Anyone can buy a desktop 3d printer. If I had something large enough I dare say it could possibly be the largest in Aus. Which surely would have to help my business prospects.

So you really cant leave your printer alone? Is it like you cant walk away? Or just check on it every hour or so? Is that the case with all types of printing? Like could I get away with leaving it with one type of printer over another?

Really sucks with how long it takes to print. Im guessing that cobra would have taken a month or so solid to print? Which in round figures if you charged $100/hr for your printer and assuming it ran 24 hours a day would be around $72000. You can buy a complete one here for that.

I guess I could go for a smaller one but is there likely to be clients?

Sorry if this seems disjointed and doesnt make sense. Ive had to stop and go back to work a few times whilst writing this haha.

Thanks
Brad

dharmic
12-19-2016, 09:35 PM
Completely understand where you're at.

To the big printers - yeah, there are commercial and custom made printers that size but you're looking at huge bucks. I don't know how they get around the issues of lifting and shrinkage etc. I do know how much of a battle it is going from 100mm diameter prints to 300mm diameter prints though, and can only surmise those difficulties will continue to scale up as you scale the work to 1.4m or more on a bumper.

And leaving printers alone? Well, there's a few things that can go wrong. The worst is when the temperature sensor in the hot end fails (which happens a lot more often than you may think). So the controller thinks the heater is at room temperature and keeps pumping power into the heater element. At some point the plastic in the nozzle can catch fire and you come home from a day at the office to no more house. There's ways around this with automatic fire extinguishers and heat triggered e-stops etc but it all gets a bit nuts. More mundane issues include things like a fault in the printing process turning the job to crap, so you get home to find you've blown a great big pile of plastic and need to spend three days chipping it away from where it's concreted itself in a big pile in your machine's guts.

I treat mine pretty much like I treat my router, mill, anything in the shop that moves: I will get it running and walk away, but I won't leave the site and I won't stray out of earshot of the job for more than a few minutes.

Jim Dawson
12-19-2016, 10:58 PM
It is possible to 3D print stuff that does not use a thermal process. There is one company that is 3D printing concrete houses. I suppose it might be possible to use something like a 2 part polyester resin and some kind of filler. Maybe even to get it to self level. I have no idea how to go about this, but once you had the basic machine, then it's a matter of finding the right materials and application method. The print head would change depending on material and method, but the X, Y, Z machine would remain substantially the same.

brad91
12-19-2016, 10:59 PM
I would definately be going down the path of custom made.

We turned an old magic eye cutter at work into CNC with a basic $400 ebay kit, computer from the dump and some wiring and a few gears machined on our lathe.

I was planning a similiar approach but obviously have the build my own frame. I design mechanical stuff for a living and am competent on lathes/welding/machining etc so I wasnt concerned about the build. I am however concerned about the time factor of printing.

Anything stopping you running 4 or so print heads at once each one slightly higher than the next? Effectively cutting the print time down to a quarter of usual?

One guy at work had a good idea. Print the base layer with tabs on it. Have the machine pause/just carefully do it whilst printing and screw the bottom layer down in all 4 corners. This would stop the part lifting. But on something that big im also imagining you would need a heater head that re-heats the plastic as your coming back to the other side as the material wouldnt bond together otherwise.

Alternatively what about an SLS printer? I havnt read anything on those parts warping?

I dont have a mega budget but 'if' this is something that could be feasible and as I would be doing most of the work myself and with the frame material cost around 5k (my mate said his 3m long router material was around this) and the motors etc being under $1000 I would estimate $10,000 should see something decent? (no idea how much more SLS is?)

I was really hoping it could work as I was seriously looking at trying to make some decent money from this so I can become my own boss one day