What's new
What's new

Complete newbie looking for direction - auto parts

Raccin

Plastic
Joined
Aug 21, 2016
Location
Akron, OH, USA
Greetings all! I've embarked on a journey to manufacture automotive parts as a side gig, which may one day become my main job (not yet though). My primary focus right now is lighting, whether it's taillights, parking lights, etc. I've got a handle on the electronic side no problem, but due to the nature of the beast, I also need to manufacture the rest of the unit also.

I'm seriously lacking direction on this one. My original thought was to make them out of various plastics by using a CNC machine to drill blocks of metal, which in turn would be placed together and I'd pour plastic into it for some kind of molding process. I started going down this path, only to realize that I had no clue what software to get, and the more I researched, the more I questioned its feasibility long-term.

I am perfectly capable of doing the 3D designs myself, but my question becomes how to translate that into reality. Is CNC machining even the right approach? I really would like to do all this in-house. As a hobby, I don't have millions to spend on it like most auto manufacturers do, but I understand that manufacturing is an expensive business to begin with (or so it seems), so I can work with that.

ANY guidance, pointers, experience, etc. is highly valued. Been searching the forums for a beginner's guide (heck, even Google), but nothing found yet.

EDIT:
My location doesn't seem to want to save. Set it twice already, but doesn't show up. Should be "Akron, Ohio, USA" for anyone wondering.
 
Milling or turning the housings is perfectly acceptable for proving your concept and checking for figment but will really fall apart the second you have any quantities to produce at a marketable price point. Just like most of your competitors you will eventually need to look into injection moulding if you will be doing plastic housings.

If you don't have an experience in machining you're most definitely underestimating the difficulty of what you're hoping to accomplish, never mind the capital investment. Working from a budget figure $10-15k for a 90s VMC rigged to your door, another $2-5k for some tool holders and workholding etc just to get you started and then start at $2-15k for software if you don't need CAD software and you're going to have a lot of money tied up in things you don't know how to use.

You say you're capable of doing the design work yourself but what exactly does that mean? There's a big difference between can work Google sketchup and a practicing engineer.

You're going to be better off farming out the manufacturing to people like us who have experience in manufacturing. If you can afford the learning experience I'd keep doing whatever you're doing that lets you afford it, there's a reason opening a hotdog cart seems like the answer some days...

Not to say it can't be done but you're in for one hell of a learning curve.
 
Calibrate Your business plan.

Whats Your target cost ? Qty / yr ?

If you are making replicas for museum autos, or aftermarket stuff, and can charge 500$ each, its a goer.
Get er done.

Manufacturing molds and molding:
If You need to work, You cannot afford to even start manufacturing these.
To make the molds, figure 30-60k in stuff.
A basic, old, small molding machine, figure 15-20k.
Add 1000-2000 work hours. Yes, that many.

Anyone with a shop has way over 100k in stuff to make these efficiently.
Find out cost of mold, about 5000$ (china import) to 4x that.

And then, out pop say 10 parts/cycle, in say 4 secs (could be 4 times/sec).
And can make 1000+ parts per day at less than 1$/part.

The people doing these types of parts can make them for half the cost You could, with old obsolete equipment.


All sorts of silicone molds, etc. exist for small quantity manufacturing.
At 1/50 below mentioned startup costs.
But the parts may or may not be appropriate for the use.
 
Last edited:
Seems like maybe I'm in the "design and assembly" business then. Okay, that's actually really useful information. The electronics side wound up being way too expensive to do the assembly work myself, the cheapest machine I found that could assemble a whole board was in the ballpark of the machines you guys are quoting, something like $5,000 to $10,000. Big investment, and one that I'm not likely to recouperate given maintenance costs.

So here's the next question: I've been designing the electronics myself, going elsewhere for the actual production of the boards and the assembly of the components. Is this an approach I can take with the manufacturing side of things also? As in, if I make the 3D CAD model, can it be manufactured?

I fully understand that whatever 3D model I provide may require tweaking, hopefully I can find a shop who is willing to work with me to teach me how to make better models as this progresses.

Side note: that $500 number, was that for a single taillight? I'm in the market to make at least 50 of each side and sell all of them, may be more like 100 when it's said and done. Does that change the per-unit price at all? For my first project, a NOS setup is I believe around $350, so it would be killer to meet or beat that price on any newly manufacturered units, even without electronic costs added onto that.
 
Seems like maybe I'm in the "design and assembly" business then. Okay, that's actually really useful information. The electronics side wound up being way too expensive to do the assembly work myself, the cheapest machine I found that could assemble a whole board was in the ballpark of the machines you guys are quoting, something like $5,000 to $10,000. Big investment, and one that I'm not likely to recouperate given maintenance costs.

If swinging $5k to $10k for electronics assembly machines is out of your budget, then CNC mill or lathe and support equipment will be as well. I have both cheap / old SMT and wave assembly lines as well as cheap / old CNC lathe and mill. The upkeep and tooling cost on the electronics stuff is considerably less than the machine shop machines.
 
Side note: that $500 number, was that for a single taillight? I'm in the market to make at least 50 of each side and sell all of them, may be more like 100 when it's said and done.

i make parts for a vintage gm vehicle. it's nowhere as lucrative as most people think. i have eliminated many items over the years as i can no longer do other work to subsidize items that sit on the shelves for years.
 
Just want to repeat my previous question as it seems it may have gotten missed: If I design it, can someone manufacture it? Fixing any flaws in the design is welcomed.

If swinging $5k to $10k for electronics assembly machines is out of your budget, then CNC mill or lathe and support equipment will be as well. I have both cheap / old SMT and wave assembly lines as well as cheap / old CNC lathe and mill. The upkeep and tooling cost on the electronics stuff is considerably less than the machine shop machines.
That seems to be the consensus. It's mostly low-quantity parts, and being able to profit five figures requires over a decade of this to accomplish I'd imagine, or some serious financial breakthrough that I can't bet on.

i make parts for a vintage gm vehicle. it's nowhere as lucrative as most people think. i have eliminated many items over the years as i can no longer do other work to subsidize items that sit on the shelves for years.
My strategy is not one of maximizing profits, it's creating products that may be a bit expensive to most, but are still sellable in some small quantity. That's why I ask about the pricing, if it's $500 just to produce one, that's well within the target mark for this kind of thing, especially if costs go down if I batch order, say, 50 of them. I do have a shop that sees hundreds come through every year express serious interest in my project, even guided me quite a bit on what they'd like to see, so it's possible the 50 number may actually be low.

I scoured for the better part of a year trying to find someone who was willing to even just convert the current colored lenses to clear ones so I could retrofit LED boards to them, and there isn't anyone that would pick up the task that I could find. If I'm wrong, and such a place exists, I'd love to know who they are and get in touch with them. Haven't found it yet though, so I'm under the impression such a place does not currently exist.
 
If you want clear lenses, for LEDs, then could they be 3D printing in clear material?
 
If you want clear lenses, for LEDs, then could they be 3D printing in clear material?
Haven't found a way to do it without gaps, but now that you mention it, I've been using purely filament-based printers so far. I know powder-based printers do a much better job, but I haven't found one cheap enough for persomal use. For a side gig, that's a different story!

Thanks man, even if I ran one giant circle that started with 3D printing, I think your metaphorical knock upside the head was what I needed to get me to go back. I'm way over-complicating this.

Thanks a ton! :-)
 
I understand you caution about revealing more about your target market but I would like to point out a few things that may be redundant if you already know of them but worth saying just in case.

There are an enormous mount of LED lights for automotive, motorcycle, and marine use. Most standard bulbs are also available in LED versions and there are numerous 12V modules also available that can retrofit existing light housings such as stop lights.

Since most dedicated purpose auto lamps have colored lenses I'm going to guess that you want clear so you can use different colored LEDs in the same housing for something like red tail light/stop light which turns yellow when flashing as a directional. Because many LEDs can cause eye damage they are a regulated item and you will not be allowed to use them on public roads until you can prove they are safe for other drivers. That will mean certification and testing.
 
You need to read up on manufacturing methods.
Study small scale stuff, and hobby stuff, like re-enactors, craft work, cosplay etc.

In qty small CNC machine stuff may be very competitive if the underlying material is not expensive.

Clear plastics can be machined, and many shops machine plastics routinely.

I suspect a single piece of under 10" max diameter, might cost downto 10-20$, for machining services.
In qty 50, done in less than a day (might be 2 hours for 50, depending).

Its the kind of stuff many here do.

The important part is understanding the manufacturing vs $$ aspect.

The molded lenses cost, say, 1$, qty 5000.
The machined lenses, say 20-40$ material and 20$ machining, qty 20.

So 40-60$ cost each, this aspect (potentially down to 30$).

So, if you are selling them at 200-500$, you have a viable business, where the total max qty is limited to a tiny subset of total users (67M new cars/year), 1B cars-in-use.

Find out who makes similar parts in plastics.
Hint: marine industry.
 
If I design it, can someone manufacture it? Fixing any flaws in the design is welcomed.

Of course if you can design it someone can manufacture it but there will be some limitations. Look into the limitations of molding. Every side needs a draft angle about a parting line. If you have zero or negative draft you will not be able to pull the part out of the mold unless the part is really flexible like rubber. Also when you start getting into complex shapes with parting lines that are not a flat plane it will really stretch your CAD skills as well as your CAD program abilities depending what you have. Some CAD programs have tools to split molds otherwise you will have to draft the parting lines manually which can be really time consuming. The molds are not just a negative of the part they have to have the gating and venting designed into them. You may also need ejectors - places that an ejector can press the part out of the mold. Depending on the design you could have issues with warping, uneven shrinkage, thin cross sections not filling, bubbles, or surface finish defects. I would imagine you could find a manufacturing that will take your CAD model of the final part and they could work with you to make needed revisions to make it manufacturerable but this could get very expensive. As the saying goes if you have to ask how much you can't afford it.

Most plastics are made with very expensive injection molding machines that work with pressures on the order of tens of thousands psi of pressure. There are some hobby grade bench top hand powered injection molding machines. I know a guy that has one. Its made by Emco not to be confused with Enco. Its limited to a 4oz shot IIRC. There is cold casting which is usually a two part plastic resins that you can pour into molds but I don't know how well suited the material would be for your application. You could even cast epoxy. If you do a cold cast resin like this you will probably need to put it in a vacuum chamber to get all the air bubble out. One other option would be to deform sheets of plastic to a mold similar to blow molding or vacuum molding. A clear acrylic plastic (plexiglass is a trade name for a popular clear acrylic sheeting) can be deformed to the shape of a mold when it is heated. You could experiment carving molds out of wood by hand and heating a sheet of plexiglass with a heat gun and deforming it over the mold.

If you are the mechanically inclined type that has a lot of perseverance you could probably accomplish one of these methods out of your garage but only after spending at least twice the amount of time you think it is going to take and blowing whatever budget you had in mind. If the part you are trying to make has a flat bottom, flat walls and simple radius on the corners you don't need a CNC. An old worn out knee mill or even benchtop mill in the corner of your garage will get you there. If the design is more complicated you could still do it to a point without a CNC but its going to take a lot of thinking. Using a manual mill is like drawing with an etch a sketch. Straight lines are easy but curves are not. A rotary table and ball mills and taper mills give you some more options but only to an extent.

It is hard to find out where to start I know. Just keep digging. Look up videos of machining and molding processes, search forums, look up books in your local library, look for local machinist clubs. NYC CNC has a youtube channel that you might find helpful.

FYI some terminology to help you out. When machining with a mill unless specifically doing a drilling operation the cutting operation is called milling not to be confused with drilling. When using a lathe to machine parts it is usually refereed to turning.
 
Regarding the LED part, the problem isn't with sourcing a good bulb, it's tbe design of the OEM unit. By it's very nature, it emits a poor light distribution, and suffers severely from "hot spot" problems. The community behind the car at large has been searching for a solution, experimenting with aftermarket units, mostly unsuccessfully. It doesn't help that the bulb bases are quite rare, and the circuitboard powering the bulbs themselves were seemingly made of cardboard. The whole design is flawed, and needs a modern revamp to address. Most guys spend at least $300 on a good setup, and that's before they touch the lenses and such, which can run $350 per side.

Anyways, injection molding is something I will research, but I will be primarily focusing in 3D print methods for now. If I can 3D print the part I need, already have some experience with the process, and simply trade cost for time and effort, that's something I can bank on.
 
I don't see your hobby grade Fused Deposition Modeling (FDM) printers that are all the rage right now are not going to get you there. I don't see any way you are going to get close to the optical properties you are looking for. They smoo a bunch of lines together that are not perfectly bonded with a less than ideal surface smoothness. They are mediocre at best for making prototype mach ups, trust me I have one. One of the technologies like SLS or one of the photo resin technologies might get you there. In my opinion you would be better off sticking with traditional manufacturing techniques.
 
Look into protolabs.com. They make it their business to produce short run injection molded parts using aluminum molds starting with your CAD model. Mold cost will be much less than full production tools. Part cost a little higher, part quality a little lower but may be good enough for your purpose. Happy customer, not connected otherwise.
 
I don't see your hobby grade Fused Deposition Modeling (FDM) printers that are all the rage right now are not going to get you there. I don't see any way you are going to get close to the optical properties you are looking for. They smoo a bunch of lines together that are not perfectly bonded with a less than ideal surface smoothness. They are mediocre at best for making prototype mach ups, trust me I have one. One of the technologies like SLS or one of the photo resin technologies might get you there. In my opinion you would be better off sticking with traditional manufacturing techniques.
That's what I said in my prior post:
Haven't found a way to [3D Print] without gaps, but now that you mention it, I've been using purely filament-based printers so far. I know powder-based printers do a much better job, but I haven't found one cheap enough for persomal use. For a side gig, that's a different story!
I have three filament-based printers and zero powder-based ones. FDM is what I'm calling filament-based (i.e. you give it a spool of filament and it prints it out line by line), SLS and photo-resin is what I'm calling powder-based (you pour in a powder and it uses a laser or light etc. to bond it together). I couldn't remember the technical terms offhand, my apologies if I was unclear there. Powder-based printers have always been more expensive than the filament-based ones, but for my use-case, it's perfect. This is, of course, assuming I can find a 3D printer that can print a clear material with no gaps in it, which may be a challenge. Samples first, of course, no sense spending $10,000 on a machine that is useless for my use case!


Look into protolabs.com. They make it their business to produce short run injection molded parts using aluminum molds starting with your CAD model. Mold cost will be much less than full production tools. Part cost a little higher, part quality a little lower but may be good enough for your purpose. Happy customer, not connected otherwise.
Thank you for the recommendation, I will look into it!
 
SLS and photo-resin is what I'm calling powder-based (you pour in a powder and it uses a laser or light etc. to bond it together).

There are printers that use a liquid resin that is cured by light. Its usually classified as Resin (DLP/SLA) technology. I think this would be your only hope although I personally think its a waste of time compared to traditional molding technology for what you want to do... Just saying in case you have not heard of these. Was not sure if we were talking about the same thing. Versus the powder based printers that use a plaster powder. ZCorp (now bought out by 3D systems) called it 3DP. It sucks too (I have an old 310 model).
 
Raccin:

My background is automotive engineering (but not lighting, though I've worked along side a few lighting egrs on some projects). If you "manufacture" lighting asm for automotive use and it is not tested to comply with DOT regulations (FMVSS 108 I think), then technically it is "for off road use only". If you don't put in those weasel words you could get in a bit of trouble - not that it seems to affect those driving around with crappy (clear lens LED, poor optics and non functional reflectors) aftermarket taillights. You might want to poke around here a bit (SEMA is the trade association for aftermarket parts, you could visit Vegas for the annual SEMA show (and possibly be overwhelmed by your competition)) Auto Regulations | Federal Regulation of Aftermarket Parts | SEMA
Here is some light reading for testing to FMVSS 109: http://www.google.com/url?sa=t&rct=...D_IXS4cIW1dNV_33deOM6g&bvm=bv.130731782,d.dmo
 
Short and sweet. There are no (as in NONE, ZERO, ZILCH, NADA) additive machining (3d prototyping) machines that will produce optically clear AND smooth parts suitable for usable automotive lighting assemblies. And even if there WERE (hell, maybe there is some technology that JUST came out that I am unaware of), doing short run production on them would be exceedingly expensive and slow.

Somebody mentioned Protolabs..... also check out Quickparts.... both will make prototype molds for you (quick turnaround) and are easy to deal with. Not cheap, but WAY cheaper than an Objet machine or a line of credit with Stratasys....
 
I looked at a vendor's lens products at the Charlotte, NC, Auto Fair a few years ago. He used original glass lenses as a pattern and made flexible silicone molds from them. For clear lenses, he pored a clear plastic into his mold. The finished plastic lenses were very good. He used a transparent colored plastic for the lenses that required color. I do not remember who the vendor is, and have no contact information.

Jim
 








 
Back
Top