Manufacturing Suitability (CNC/Extrude?/Cast) for Part

[snoop77]

Hi all,

I've a small part (15.7mm width × 35.0mm depth × 28.0mm height) which I intended to have CNC'd with a type 2 anodised black finish. Everything I read said anodised casting doesn't really work well and this part requires a good cosmetic finish. With the chip shortage still underway I'm restricted to 50pcs but need 1,000pcs longer term and the cost benefit of casting I do need to explore.

With this in mind I've put it out to a bunch of Chinese one of which in particular is willing to "guarantee" that casting the part and anodising it will be fine. Specifically he's said there are two methods and the second will give a finish like CNC but it's double the cost at $2 and the mould is higher. I asked him to explain it but he's sidelined the question saying to trust him, only saying that the anodisation for this method would give a deeper colour. I've been burned several times by Chinese quality and have no intention to do so.

Rough prices with anodisation and a laser engrave on the bottom are:

250pcs CNC: $5
50pcs CNC: $15
1,000pcs Cast: $1 each + $1,000 mould
1,000pcs Cast (or extrusion or something): $2 each + $1,600 mould.


It's worth going to cast if I can nail down the quality and hope to pick people's expertise and experience on which method and finish to go for.

View attachment spec.odt.zip
View attachment case.step.zip

 

[DDoug]

Are you still planning to send this work to China ?

 

[snoop77]

Not necessarily!

All being equal if there's someone in the UK that can do it and the additional cost isn't astronomical it counters the reduced risk.

 

[implmex]

Hi snoop77:
There are a whole lot of ambiguities about this part which make it difficult to advise you intelligently, but let me try to give you something from what I can see so far:

1) You have a really variable part thickness, and this will impact any cast part appearance massively, so if appearance is crucial, casting of any kind is not your friend.

2) As designed, obviously you cannot extrude this part, so that process is out if you want a near net part from your extrusion process.
It's an exceptionally poor candidate to extrude and then machine and assemble, because you gain very little for your cost, you increase the assembly complexity, and now you have to figure out how to fixture and mill the extrusion, how to fasten the end caps and, on top of it all, aluminum extrusion is hard to mill nicely (too soft and gummy).

3) If this is to be diecast or investment cast, the geometry will have to change to accommodate those processes...you will need draft, you have to plan a gate location, you have to plan for ejector pin marks. you have to plan for flash and how to trim it.

4) In my opinion, you don't have enough parts to justify all the up front cost and screwing around needed to invoke a high volume production method like diecasting or investment casting.
A thousand parts is nothing...a couple of days for a good CNC shop with modern equipment.

So IMO, machined from billet is how this part is shaping up.
Not only do you get design flexibility, you also get consistent material properties, a proven anodizing process, and a better part appearance, all for a lower cost than all the screwing around that alternative processes imply.


BTW, at 5 bucks a part and a thousand parts, you are talking about chicken feed dollars...this is a project a commercial customer would pay for out of petty cash.
If the financials of this kind of outlay are a strain for you, your business venture has other, more pressing problems to solve than what processing method gives the best outcome.

Not to sound harsh or anything, but for the money you're talking about, and the pathetic potential profit for me, I wouldn't even look at the job seriously; and I'm a small one-man shop.


Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

Oh yeah, as an afterthought; if you don't want to fuck around with anodizing, can you make the part out of black Delrin instead?

 

[Comatose]

In my experience, the quality of anodize, machined or cast, is a bit of a crap shoot whether you do it in China or domestically. Expect some fall-out either way. Purportedly, even Apple scraps a LOT of cases at the finishing step.

That's a really, really small bottom radius to CNC that deep. I'm not even sure how one would do it. 1mm radius, 25mm tall, from the picture. This is a part that should be cast if you can't open that up. If you can, then the CNC cost would go down.

Casting, the top arch would need to go to maintain even wall thickness, but this sounds like an electronics enclosure, as I assume that mates to a battery of some sort. That could be square without causing problems.

I've had plenty of molds made in China. You can get really, really good quality. Better than domestic quality, even. But you end up paying better than domestic prices for it. I had one part go from a $1200 mold to a $45,000 mold, for the same part, to actually get what I want. (In fairness, it was a silicone mold, and went from being a compression mold to an LSR mold, but also went from a commodity to a high end specialty molder.) Set your expectations accordingly.

At the prices you're expecting, with the quantities you're expecting, China or another low-lost-of-labor country is going to be your only option, unfortunately. Nobody on this forum is gonna make 50 of these for $15 each. Not even the guys in their garages.

 

[DDoug]

Not necessarily!

All being equal if there's someone in the UK that can do it and the additional cost isn't astronomical it counters the reduced risk.

So you need help from us, and what will you settle for ?
10% more at most ?

What doo you consider a "Fair Price" ?

 

[snoop77]

I didn't see it as being a problem with the 25mm depth as you say, would a 2mm bit have too much flex? I could make it 1.5mm radius and heighten the mount a little to compensate.

 

[Comatose]

I didn't see it as being a problem with the 25mm depth as you say, would a 2mm bit have too much flex? I could make it 1.5mm radius and heighten the mount a little to compensate.

Yes, a 2mm bit has way too much flex. There are ways to do it, but all those ways are slow or expensive. A 3mm bit would be marginal. A 4mm bit would make it a piece of cake. Can you handle a 2mm radius down there?

 

[snoop77]

I've no idea how many the Chinese supplier is expecting to throw away, perhaps for mould + annodisation it's 50%. I'd also like him to elaborate on what the second casting method he mentioned is exactly but haven't had the details so far. They are pushing for mould over CNC but I'd be silly to think that's for my benefit.

Digger -> I'm not that far down the conversation yet, more so it's to find out what people think is the better method for the quantities I'm talking about, and there's been some excellent advice so far - keep it coming guys.

 

[snoop77]

Yes, a 2mm bit has way too much flex. There are ways to do it, but all those ways are slow or expensive. A 3mm bit would be marginal. A 4mm bit would make it a piece of cake. Can you handle a 2mm radius down there?

I'll have a look, it's not ideal. Would 3.5mm work? This is for a zip tie it wouldn't have to be as absolute as the 1mm diameter front holes.

I also have the inner lower two corners which I thought would be pointed out as an issue. They're also 1mm diameter!

 

[snoop77]

I've adjusted the zip-tie hole and enlarged the inner lower corners to match the 2mm radius.

I imagine filleting the front will be difficult so used a 0.25mm fillet/burr.

View attachment case.step.zip

 

[Comatose]

The inner lower corners were the ones that I meant.

Generally speaking, the bigger tool you can use when machining, the better. Bigger tools can be pushed faster, take bigger cuts and leave a better finish, up to about a 12mm diameter. Past that, you get horsepower limited on a normal setup. With the following exception:

The length to diameter of the cutting tool is important. Up to about 4:1 length to diameter is fairly standard stuff. A 3mm diameter tool cutting 12mm deep or less, a 12mm tool down about 48mm, not a big deal. 6:1 causes some grief. 8:1 is flirting with the edge of possible, as well as the edge of finding tooling. 12:1, you're looking at less straightforward techniques, like having to tilt the part on a 5 axis machine and use a tool with a wider shank. That's harder, slower, and requires a more expensive machine. Best to keep it below 6:1 as much as you can.

 

[snoop77]

The inner lower corners were the ones that I meant.

Generally speaking, the bigger tool you can use when machining, the better. Bigger tools can be pushed faster, take bigger cuts and leave a better finish, up to about a 12mm diameter. Past that, you get horsepower limited on a normal setup. With the following exception:

The length to diameter of the cutting tool is important. Up to about 4:1 length to diameter is fairly standard stuff. A 3mm diameter tool cutting 12mm deep or less, a 12mm tool down about 48mm, not a big deal. 6:1 causes some grief. 8:1 is flirting with the edge of possible, as well as the edge of finding tooling. 12:1, you're looking at less straightforward techniques, like having to tilt the part on a 5 axis machine and use a tool with a wider shank. That's harder, slower, and requires a more expensive machine. Best to keep it below 6:1 as much as you can.

Useful to know for this and all my further designs, thanks!

I've amended and will 3D print.

I'm fairly sure the Chinese company I was conversing with don't really know what they're doing, they've asked me why I don't think extrusion will work, as in they haven't looked at the designs. Neither did they spot that inside radius that would be impractical to CNC prior to sending a quote. Doesn't instil confidence when they say they can anodise casting just fine!

 

[snoop77]

The detail is coming out from the Chinese company, the cost has jumped from $1 to $3 and they want to extrude and weld the face. They've also said if there's a lot of failure the cost could jump again.

I can't see how welding the face will be good enough - it's the primary cosmetic face.

 

[implmex]

Hi snoop77:
You wrote:
"I'm fairly sure the Chinese company I was conversing with don't really know what they're doing,"

Yeah you sure got that right...no argument there.
The processing plan they're trying to talk you into is one of the stupidest ones I've seen in forty plus years of doing this sort of product development. (Assuming you articulated your goals to them clearly beforehand).

You want a thousand or a few thousand parts.
You have features all over the part that cannot be extruded but must be milled either from a block, or an extruded blank (unless you cast the part), so the extrusion gains you almost nothing except extra cost and more downstream problems.
The primary cosmetic face is one they intend to weld around.
You have highly variable cross sections that will create cosmetic blemishes in a cast or molded or extruded part.

The part needs to be black and of good appearance.
So if there was enough volume to justify, I'd be injection molding this part from a plastic like ABS, and if it needed superior strength and toughness I'd choose something like Acetal or maybe glass filled polycarbonate.

With your limited volume, I'd explore having them milled from Delrin blocks rather than aluminum and save myself the anodizing hassle.
I'd put in a label bed and make a stick on label...screw the ano and laser engraving.

So from what I've heard so far, I advise you run away from this vendor and find a better one.

Cheers

Marcus
Implant Mechanix • Design & Innovation > HOME
Vancouver Wire EDM -- Wire EDM Machining

 

[snoop77]

It will look shit.
And black anodise casting....IME that will look shit too.
Extrusion can look good if processed and etched before anodise.
Machine from solid (6082-T6) and a note on the drawing stating cosmetic face required (and discussing it with a vendor) will be gravy.
You just have to find someone in the UK that will do it for nothing and supply the material free too....

Haha! I'm not after it being done for nothing. Going to the Chinese was an automatic default and not one I'm tied to. I suspect a British CNC'ing shop will be handling larger orders and won't be interested.

I should have mentioned aluminium is for heat-sink properties, plastic is a no-go.

It's being milled and certainly not by this company as I don't think they understand the requirements - I clearly defined the cosmetic properties for one.

Cheers,

Andrew