/diy/ - Do It Yourself

I've got a question, how precisely on 30 degrees does my topslide need to be for threading? it's really shitty so I can only reliably get within 1 or 2 degrees.

>>780747
feeding with the compound is done so only the leading edge of the threading tool does the cut. can't remember why. maybe it's not important.

figure as long as the tool's got the right grind on it you'll get a workable thread.

>>780747
+-10 degree at least, its not that important

>>780727

My brain on OP: What are all those nails doing. Wait, those are shavings. Wait, holy shit *that's* an endmill.

>>780844
Well, dicks. that means I'll have to touch off every damn time.

While I'm here, I've got a project that could use some feedback.

this is the part i want to make. it has to have a .150" hole through the entire length of it, and then there's a counterbore on that hole, which then gets threaded (the counterbore). There will also be a round profile cut after the counterbore, as pictured. lastly, there is some geometry on the outside that has to be done last, on the mill. I do not have a center drill.

currently, I intend to start with square stock, squaring it up on the mill so that it is perfect. then, while still in the mill, I will find the center of this square, and drill my .150" through hole. I will set it up to turn between centers using this hole. I will turn all of the outer profile, but leave square ends. I will then move this to the mill, and do the final milling stuff, like that slot. now I must put it back in the lathe, but this time in a 4-jaw chuck, because I will be doing that end work with the counterbore and the tapping and the spherical bit. If I have a shitty 4-jaw chuck, how well can I expect it to hold the part on-axis? how can i know that it's on axis?

I want to turn some acrylic on the lathe. How do I get it to be translucent afterwards? do I need some kind of polish? would washing it in diluted acid work? how about a heat gun?

>>781077
>How do I get it to be translucent afterwards

Translucent? It's translucent by default after any machining operation.

Transparent? Lot more difficult. About the only way to do it is with some purpose-made acrylic polish and a lot of time/elbow grease.

>>781073
turn up a bit of rod that's a tight fit to the bore and indicate off of that. (unless it's one of those 4-jaw scroll chucks in which case i'm sorry for your lots)
will it get yanked out of alignment when you mash a ball cutter into it? probably. glad it's not my problem lol

>>781077
polish, flame or acetone.
http://www.youtube.com/watch?v=7na8kQ78vkQ
unsurprisingly the most boring and tiresome method (hand polishing) gives the best results.

>>780727
ME undergrad here. my school has a fully stocked, state of the art machine shop in the basement of our ME building. how do i go from dinking around in solidworks to actually making something?

>>781122
put both hands behind your back and slide them down towards your thighs, somewhere in the middle you'll find your own ass.

>>780747
Truthfully, not at all.

It is mainly to preserve tool life, but if you don't do too much threading you are good. Also, generally its 29.5 that I've heard, either way, I do most of my threading on a prototrak lathe that does all the work for me other than checking the parts.

>>781073

My suggestion is do all your mill work first, then switch to the lathe. If you have access to them, use collets. For your four jaw, I have little experience with them as we don't do much with them at my shop, but the basic idea is to use a dial indicator to bring it in to zero. If you get it centered and locked up, you'll be good to go.

Also, internal threads are best done with a tap. you might need to modify your tap to bottom out but it'll be a lot easier for you in the long run.

>>781150
what?

>>781150

I want to know what cheap knock off brand drill bit that was so I never go near them.

>>781122
see the OP. theres a link to go to for beginners.

>>781227
Well considering they used a filter on the photo. Who's to say they knew what they were doing to begin with

>>781122
>ME undergrad here

apparently your in school, audit a machining class or 10

>>780747
>>781073
same part?

why are you starting with square stock when finished part has no square sides?

how are you going to turn on centers without a center drill?

>>781383

This man is wise. I agree with his questioning. Here's how I'd approach the part:

>start with round stock an extra inch or two long
>put in normal 3jaw chuck
>face, center drill, set up with live center
>turn outside profile
>drill hole through center
>drill hole for threads
>turn that radius
>tap dat ass
>take out of chuck, cut slightly longer than finished length
>put back in chuck with rough end out
>face to finished length
>bore out the end
>take to milling machine for slots and friend

No fucking around with a 4jaw required

>>781387

thanks anon, id go a slightly different route


>start with round stock an extra inch or two long
>put in normal 3jaw chuck
>face, center drill, set up with live center

face, center drill, drill through hole, ream through hole to tolerance, can use the through hole later to dial in concentricity. set up with chuck and tailstock center

>turn outside profile
>
>drill hole for threads
>turn that radius
>tap dat ass

take part to mill, use square collet block to cut the 2 cross axis slots, 2crenelations on the end, and profile the recesses on the outside of the piece

back to lathe, part off extra stock, check concetricity, face the end, and bore .520, do this last because +.010 toleerance

>>781073
hexagons man.
use a hex bar stock. mounts easy in a vice and a 3 jaw chuck. best shape

>>781383
Same part. Starting with square stock so that after everything is turned, I can put it on the mill and cut the cross axis stuff, 90 degrees apart. and I'm going to drill a center hole in the mill, it will be the only center hole I need.

>>781387
>>781406

I don't have collet blocks, a center drill, or a rotary table, and the tapping has to be done on the lathe, as the entire purpose of the part is to demonstrate machine aptitude.

>>781619
I haven't put too much thought into this but it's an idea. Do all the turning from some round stock to begin with. Once all that's done mill one flat where one of the cross holes will be, lay that flat down to mill another flat 180 degrees opposite it. Then clamp it up by those two flats to mill another flat 90 degrees to those, then flip the part over for the final flat. Once all that's down it's a simple process to mill all the cross holes using the remaining flats to level the part.

>>781624
milling flats covers the rotational axis, but what about the axis of tilt when mounted in the vise? normally it would be done with parallels, but I don't have short parallels or gauge blocks.

>>781634
Just put it all the way down in the vice, use a narrow end mill, and be careful.

>>781634
Alright that's a stupid question. but it still requires a center drill, which i dont have.

>>781641
Then buy one, or use the shortest drill bit you have. I don't know why you'd rather do things the hard way and possible screw up instead of buying some cheap supplies that you'll probably need for future projects any way.

>>781645
Because it's not my machine tools, so I'm not sinking money into it. and you can't assume someone on DIY has money to do something, as half the time that's the reason theyre here anyway. doing this the hard way is what will get me a job to have money for the future.

>>781671
>Because it's not my machine tools, so I'm not sinking money into it
It's like a couple dollars. Where are you even working that has access to a mill and a lathe but not a single centre drill?

>and you can't assume someone on DIY has money to do something
I think it's reasonable to assume that anyone that has internet access can also spare a few dollars. I'm not telling you to go buy a 5 axis mill.

>doing this the hard way is what will get me a job to have money for the future.
If I was a potential employer I'd want people that do the work efficiently and know what equipment they need to get the job done. I'd rather put a tiny bit of money into the job to get it done quicker and correctly on the first go.

>>781073


do they not teach you how to use a dial indicator?

want to radius the corners on 2.5"x1/4" steel flat bar stock whats the best way to do this?

also, how to route metal economically.

>>781679
its 300 dollars, and a small town school with shithead professors and no budget. the only reason they have what they do is because I found it and fixed it up for them.

>>781762
>its 300 dollars

>center drill
>$300

What the fuck are you searching for? Center drills are among the cheapest bits of machine tooling that exist...

http://www.harborfreight.com/center-drill-countersink-set-5-pc-60381.html

http://www.amazon.com/Industrial-Center-Countersink-Degree-Included/dp/B00M54BPRQ/ref=sr_1_1?s=industrial&ie=UTF8&qid=1425870888&sr=1-1

>>781770
you ignorant slut
the fucking chuck, i dont have a chuck
jesus christ how dense are you

>>781795
>jesus christ how dense are you
Not as dense as you it would seem. For a start the majority of drill chucks are a lot cheaper than that, secondly you could just hold the damn drill bit in a collet.

If you can't figure this basic shit out you may as well quit now because even if you managed to get a machining job it wouldn't last long before they fire your incompetent arse.

I already said that I didn't have any collets. I already figured out this 'basic shit' with the original method i described, which no one has yet to put any holes in. you all just keep describing how you would do it, with better tools and more shit i have to buy. news flash, theres a always a much more accurate, easier way that involves spending a shittone of money. this isnt buy it yourself, its do it yourself, now if you can't tell me one goddamned thing about the method I've already described or provide any other useful information, you can fuck yourself all the way back to /b/ you thoughtless retard.

>>781762
i dont think you know what a center drill is. from your posts in the thread this whole idea seems out of your league. if you didnt want the advice in the thread you should not have asked for it.
a center drill is like $3. its has a 60* angle. a normal drill bit is double that. it doesn't seem like you understand how to use a 4 jaw chuck.
Milling your center hole on the mill, my fuck,

those slots are like, the least critical looking part. if you want an exercise in inefficient work flow and complicated jigging by all means.
if you already know everything, why are you even going to school?
>>781624
pretty much this. exactly this. or if you are a faggy baby, you can use a hexagon bar stock so you can easily use a 3jaw chuck and a vice on the milling machine.
also you cannot make this part, exactly to drawing, with the tools you have access to anyways
>>781073
>>780747

last but not least, lol 1/4-20 fuckin casual

>>781634

Grab piece of stock that is pretty square then.

It won't be as good as a parallel but it works nicely.

>>781073
>While I'm here, I've got a project that could use some feedback.
>this is the part i want to make. it has to have a .150" hole through the entire length of it,,,,
>I do not have a center drill.

I think what you want here is called a "coaxial indicator" (pic related)
it is a special dial device for centering a mill spindle over a round object or over an already-existing hole.
there is more expensive ones but the generic China one I have works to better than .001" easily.

the only hard part here is that you want both ends counterbored on exactly the same axis. with the coax indicator, that is easy to do.

what I would do is starting with the squared-off block (assuming you are doing manual machining and not CNC):
1) cut the sideways holes (the rounded-rectangle ones shown in white in OP pic)
2) drill a .110" hole all the way through, then a .147" hole and then ream it to .150" (to make sure its perfectly straight & smooth)
3) now put it on the lathe and use the .150" center hole as lathe centers to round off the outside
4) put it back on the mill and use the centering gauge to center on the .150" thru-hole to do the counterboring on both ends

you don't absolutely NEED the coax indicator, you can jerry-rig something using normal dials.. but the coax indicator is way, way faster and easier to use. if you have a mill, it's easily worth $70
the exact problem with trying to use normal dials is that they don't stay facing forward where you can see them.

>>781815
>4) put it back on the mill and use the centering gauge to center on the .150" thru-hole to do the counterboring on both ends
(correcting my own post)
gad dammit, not a centering gauge
here I meant to say "use the coaxial indicator to do the counterboring on both ends"

centering gauges are,,,, well, I don't know. I don't have one. I've never seen where such a thing would be useful.

>>781815
>>781817
this is useful information to me.
for what it's worth, I don't need the hole to be perfect (I don't think)--there's another part with a .150" stem that will be going up through this part, although the stem is non-load bearing (that is what the threaded part is for) so it just needs to be a clearance fit. If you read
>>781073
You'll see that I intend to do some millwork, which will give me a perfect rectangular prism shaped stock, and I'll be able to do my center hole exactly on-axis with a #24 bit, making it .152" . I feel like if I do ALL the millwork first, then the constant leaving and re-entering of the cutter in the material will create a lot of chatter, shitting up my precision and my finish. Also, I am nervous about clamping my part directly with a 3-jaw, as it may marr the surface while I do the end work.

I have considered making an attachment for my dial indicator, one that could measure the inside of a hole, something like the pic?

>>781834
>I have considered making an attachment for my dial indicator, one that could measure the inside of a hole, something like the pic?
that could be useful for a lathe, but its not going to help much on a mill
and it wouldn't even measure accurately on a lathe, but it might help you center stuff on a bore easier.

the reason it might help on a lathe is that on a lathe the part turns, so the dial can stay stationary (where you can read it)
on a mill, the feeler must rotate at least 270° around the part to be any help--so you're not going to be able to read the dial all the time

here is a video showing how its used:
https://www.youtube.com/watch?v=Ik2ePadnU8g

there is another video (the #1 google result) that a guy talks for 15 minutes about why coax indicators don't really "measure" well (cosine error) but that's not the point. the actual measurement isn't what you want.
it is -only- used for centering under the spindle. when you adjust the part to where the indicator needle doesn't move, that means the part is centered.

Has anyone got any more links to more resources? I know enco is a good place to buy stuff, but other than that i got nothing.

>>781860
>Has anyone got any more links to more resources? I know enco is a good place to buy stuff, but other than that i got nothing.
I get a lot from enco, most all of my consumable stuff.
MSC Direct is their 'other' site, but the main difference I see on MSC is that they offer less low-end stuff and more high-end stuff.

I check Grizzly occasionally. they're Chinese-made and not that cheap but sometimes they have nicer features on equipment they carry. their consumable stuff usually costs WAY too much

also there is a website named Shar's, but they sell a lot of their stuff a bit cheaper on ebay as discount_machine

all these places mostly sell Chinese-made stuff, do note.
"good" machinist stuff costs 5-10X what the Chinese stuff does. people like to say "don't buy that chinese crap" but they have no clue

>>781806
>I already said that I didn't have any collets.
Really? Must've missed that bit. If that's the case, how are you intending to hold the end mills in the mill?

I can only think of 3 options. Either you're going to use a drill chuck, if you think that's acceptable then you don't deserve a machining job (and that still wouldn't explain why you could chuck up a centre drill), you're using bits with a taper shaft that go straight into the head (can't remember the term for those) in which case they're likely way to large to work for your piece, or you simply don't have any idea what a collet is.

>I already figured out this 'basic shit' with the original method i described, which no one has yet to put any holes in.
No, you figured out a janky way to do it (not that that is particularly hard) that takes more time and effort and is more likely to result in screwing up and having to start all over again. We're not going to point out every possible way you could fuck up, there are far too many of them. We're trying to tell you the easy and correct way yo do this shit, the way you would be expected to do it if you were being paid for it. It's not our fault that you're lacking even basic supplies (that would cost you like $50 at most).

>>781880

JTS is alright

i get most of my material from enco

>>781806
jesus christ you dont even have collets, why are you even using the mill then?
. >I already figured out this 'basic shit'
You havent figured out shit. everyone has pointed out several holes.in your awful work flow.
Unless you are literally eating garbage, spending $3 for a fucking center drill (which i still dont think you know what that is), if you have such a problem with that than why even be a machinist? its not like it would be the last tool you buy.

There is zero reason to drill the center hole in the mill. you will be mounting it in the lathe anyways, you migh aswell do the center drill there, so its like uhhhhh center. relative to your rotational axis. youve even said the 4jaw is shitty and you dont know how to use it. you are going to spend more time mounting that, trying to get your part concentric to your center hole (if you understand what that means)
>>781834
>hey i want to do more overly complicated shit that i dont understand!

>perfect rectangular prism shaped stock
>no collets
thats just not going to happen. and even with a 4jaw chuck, you dont need a perfect rectangle, you just need to understand how to use a 4 jaw chuck.

>I am nervous about clamping my part directly with a 3-jaw, as it may marr the surface while I do the end work.

and bringing everything into perspective relative to your skill level.
Use shims. or tape. to dont smash it into your part like you are trying to fuse the 2 together,. or leave extra stock on your work piece to accommodate for it and then as last step you part it off with a parting tool.

>>781880
>>781860
I use http://kar.ca/

Dont bother with shit tools. despite budget or use. start buying real tools and it will show as your work becomes less shitty and you struggle less with better results. good machinst stuff should not cost 5-10xs. my Korean band taps and Bosnian reamers were way less than i thought they would be.

All my materials I buy local from metal suppliers.

>>782080
I have collets for my mill but not my lathe. but I don't have a tailstock drill chuck. I even have the fucking center drills but not the chuck. it's fucking expensive, and I'm not buying one.

there cannot be any drilling done in the lathe. period.

I can't just jam stuff inbetween the jaws and the stock because it won't be on center. as it is my 3-jaw has a couple thousandths runout, so anything that gets clamped gets clamped once. and even if I could use it twice, I can't leave extra stock on the end because the end are the part that needs all the complicated lathe work done.

If you didn't have your head so far up your own ass trying to tell me i need to buy a million and one things, you'd see that the plan I had is the only one that guarantees precision, which is the most important thing for a machinist, is to understand how to make a part with the cuts needed ad also the precision and indexibility.

if you have nothing useful to provide other than 'hurr durr buy things' then get the fuck out of the thread.

>>782155
But the tasks you want to perform with accuracy require specific tools in order for them to be done t tolerance.
Unless you can accept a wider tolerance then alternative methods of performing those machining operations are not going to produce the desired result.

That's just how it is with machining. What you can accomplish is going to be driven by what tools you have available and your ability to use them to produce the part as drawn.

The rest is a matter of compromise based on time, tools, material, and tolerances.

I don't have access to a lathe but I need to turn a part, so I am putting it in a collet in the mill and turning it down by holding a tool in the vise. The only problem is that I don't have any turning tools or tool blanks, nor do I want to spend the money for a tool I won't use again. I do have a dull drill bit which I can probably grind a cutting edge into just like a lathe bit. Will this suffice? Or am I much better off buying a tool blank?

>>782208
Look in your bin of broken carbide tools.

>>781691
Edge rounding tool on a mill or a regular wood router carbide roundover in a mill

>>782217
I'm a hobbyist, I don't even own carbide tooling. Only broken HSS endmills and dull HSS drillbits

>>782220
>>782220
technically yes, since it's all HSS one way or another... but hot damn, that's sketchy. dont forget to take a vid a bring a link back for us.

>>782220
>I'm a hobbyist, I don't even own carbide tooling. Only broken HSS endmills and dull HSS drillbits
If you are "turning" a part in your mill (I do this occasionally for small parts, if it is easier or faster than using the lathe),,,,
it would be a lot better for you to buy a couple HSS lathe bit blanks and use those instead--just for the reason that the end mills and drills are round, and you have no way to grind 4 flats on them correctly (perpendicular and perfectly flat, that is).

I have a few carbide mill bits around that I use occasionally.

I also have a set of the cheap brazed carbide lathe bits, but I hardly ever use them. They have flat faces and that doesn't work very well on the littler (11" x 26") lathe I have. ....All the lathe bits I normally use are ground from HSS blanks, and are very hooked (pic related). If I try to use any flat-faced lathe bit, it howls and leaves a shitty surface finish. A deeply hooked bit just hisses almost silently.... as it cuts off a bird's nest. But I usually can't avoid that.

Some of the stuff that all the books say to do on BIG lathes, simply does not work on small lathes. I've tried to use two different cutoff tools and neither works--they scream and chatter. I think they are too wide, the blade is 1/16" I guess. If they were 1/4th that thickness they might do better. And I've not bothered to try to make anything to do that.

>>782349
>I've tried to use two different cutoff tools and neither works--they scream and chatter. I think they are too wide, the blade is 1/16" I guess. If they were 1/4th that thickness they might do better. And I've not bothered to try to make anything to do that.


you have to be very very careful with how you profile the edge. i run it as slow as possible on a 9" lathe, flood the area in wd40 for aluminum, then feed it slowly by hand.

normally this guy parts at some high speed on steel with some expensive inserts but here he tries parting a circle ( even brazes his needed bits ) - doesn't work though

https://www.youtube.com/watch?feature=player_detailpage&v=tftydxyTvcM#t=1511

Guys, guys, I need to make a chisel texture (preferably kinda random looking) in wood. I have a cnc mill, but I don't have any of those fancy-shmancy texture making cams. Do you think it's possible and feasible to make if all I have is a drawing soft + 3d soft + cambam or should I ditch this idea?

>>781619

>I don't have collet blocks, a center drill, or a rotary table, and the tapping has to be done on the lathe, as the entire purpose of the part is to demonstrate machine aptitude.

Make some tooling for your aptitude demonstration.

A collet and some parallel clamps would cover much of the same set of tasks.

I imagine making a drawbar would be a real achievement, too. What are you using in the milling machine right now?

>>781371
>>781227
>>781150
How does a drill bit bend like that? All I've ever seen or used snap and have very little deflection in the first place. I don't even think you can get them hot enough from normal drilling to anneal them enough to allow bending.

I mean how cheap do you have to go to find something like that?

>>783113
Probably harbor freight.

>>783113
Could be a nail used as a bit. We used to get pallets that were nailed together with nails that spiraling on them

>>783113
masonry bit (soft shank, brazed carbide tip)
drill motor dropped off a ladder or scaffold

this is where the 'walking under a ladder is unlucky' superstition comes from

>>783130
I did some research, seems it happens in the UK a lot (pic related). Don't by, "Rolson," brand.

>>783154
It isn't, those have less pronounced fluting.

>>783200
The bit in >>781150 isn't a masonary bit though.

>>783202
>I did some research, seems it happens in the UK a lot (pic related). Don't by, "Rolson," brand.
in another thread from a few weeks back, anon was trying to drill a hole in a thin stainless-steel bowl. and the drill wouldn't go through, and he wanted to know what to try next.

,,,,and in that thread, I am one who said "stainless is especially tough to drill in,,,, and besides that fact: most department store drill bits (and taps) (sold in the USA at least) are absolute shit, they are way to soft and may not go through a thin sheet of stainless even once..."
and there was people who didn't believe me

if you need to drill in metal and you can only buy department-store drill bits, then do yourself a favor and buy the cobalt-steel drill bits.
do not buy ANY of the department store high-speed-steel bits, they're only really made for hanging pictures on walls--drilling in plaster and wood.

if you want HSS drill bits that are good, buy name-brand bits from a machinist supplier online. you can get them plain/uncoated, you can buy them individually and they'll still last at least 2-3X as long as the department-store titanium-coated ones.

>>783209
Most of those drill bits people have trouble with actually have on the packaging what metals/materials they can/can't drill through. People just need to read the labels and use the proper tool.

It is the equivalent of using a deadblow hammer to install nails and bitching about how they don't work right.

>>783202
>>783209
>>783215
Mabey next time buy a bit that is specially made for metal not a wood bit

>>782155
you sound like a joy to be around. I would certainly like to be your co-worker. Reacting angrily and slinging insults when your poorly-planned process gets met with reasonable criticism and constructive advice is the mark of a true professional. You have a bright future in the skilled trades, and I wish you all the best of luck on your journey.

>>783202
Drill bits like this from rolson, silverline etc are about £1 for 10, tops £2 for 10 - and that drill bit is clearly a woo drill bit, my guess would be they tried drilling in something metal.
The cheap drill bits are great in my eyes - if your on a piece of work you've put 50 hours into, use a decent expensive one, an old forgemasters etc, but if your knocking something out or it's first cut, using a drill bit like this with lots of lube should work (at least for a few holes - they go blunt fast)

are dremel brand dremels better than off brand ones?

>>784402
Dremels are ok, better than harbor freight or walmart black and decker rotary tools, but by no means are they truly great rotary tools like a Foredom rotary.
Its a good idea to have one around in case you need it.

>>783215
>Most of those drill bits people have trouble with actually have on the packaging what metals/materials they can/can't drill through.
at most (US) hardware stores tho, all they say is that the auger bits are for wood-only and the twist bits are "for wood or metal"
many of the twist bit sets only give the coating type in the description--black oxide or titanium. they don't tell you anything about the turd they put that coating *on*

some sets say if they are high-speed steel or only carbon steel--but I've seen the HSS ones turn to mush just the same when drilling in thicker plain carbon steel.

>>784402
>are dremel brand dremels better than off brand ones?
they are somewhat better. also the off-brand ones may not fit into the Dremel-brand accessories, if you might need any of them.

the Dremel-type bits and sawblades are also supiciously soft, I have found. the specs usually say "for wood, plastics and soft metals".... don't count on them working well in aluminum or steel. and forget about stainless.

the carbide bits work okay (for grinding). and the cutoff wheels work okay.

I mainly use my dremels with the cutoff wheels, or with wire brushes for cleaning/polishing smaller rusted metal parts.
if I have a tiny rusted metal screw I want to save, the little wire brushes on a dremel will get all the rust off but still not beat up the screw threading like using a bench grinder would

>>783261
>>783236
>>783209
https://broadside23.wordpress.com/2012/01/14/bend-o-matic-drill-bit/

>A 5 mil bit of MDF caused this.

So much for your metal theories.

How can you mill a hemisphere out of a block of aluminum? Apparently something like this requires CAD and CAM software but I don't even know where to start. OP links didn't really detail anything like that.

>>784470
>So much for your metal theories.
I have broken HSS drill bits before and they bend only a little before snapping off entirely. for any drill to bend like that pic looks like shit to me.

I can see how it might happen--if the drill didn't hold concentric, and you were pushing hard on it, especially when drilling horizontally in an odd or uncomfortable position. The weight of the drill kinda "hangs" on the bit, and starts wobbling as it spins.

>>784485
>How can you mill a hemisphere out of a block of aluminum? Apparently something like this requires CAD and CAM software but I don't even know where to start.
does not require a CNC.
in olden days of yore, you would do this on a manual lathe with a device called a "radius cutter" or "ball cutter"
one such video:
https://www.youtube.com/watch?v=UwGLL7cMdHk

>>784498

Thanks for that. Now I just have to find someplace to learn how to use a metal mill.