help with turning (or otherwise shaping) a small dome

bbaley

Registered
Registered
Joined
Nov 14, 2018
Messages
62
Hi,
I am just getting started researching how one might turn or form a small "dome".
By dome I mean not a full radius half-dome (ball end, etc)

but a convex shape for which the radius would be MUCH larger than the piece being machined.
The convex dome need not be a specific dimension with an exact tolerance.
I am imagining a piece of 1" round stock on which the end is shaped a dome where the center/highest point is maybe 2.5mm / 0.1"

My first thought was to make some sort of jig to hold a cordless drill that I could use on the belt sander - where the jig bottom plate had a bolt and a slot (cut with router and large circle jig) such that I could "turn" the part on this large sweep/radius... but I am figuring there is some lathe option possibly ?

I have seen lathe ball turning attachments and the like - but those seem a) to be for small radii , and also complicated to make (for me).

what would you do ?

dome_1.png
 
If the curve doesn't have to be geometrically accurate, it can be done free hand. You can rough out the curve by calculating points along it and cuttin with a combination of the compound movement and cross slide movement. Set the compound so movement is parallel to the spindle axis and cut a series of steps. Depending on your patience and the number of steps, you can get fairly close. Once the rough profile is cut, use a file to break the steps. As you get close to your desired shape, switch to emery cloth or sandpaper. I use WD40 as a lubricant with the sandpaper. Switch to finner grades of sandpaper until you reach your desired finish. It is possible to reach a mirror finish that way.
 
What kind of tolerance are we talking about and by what means are you to measure this? What is the mating part? What quantity? If moderatly low tolerance and high quantity I would say to hand grind a form tool then the profile would be cut in one operation.
 
Think I would grind a hand ground form tool. So I went and ground a tool. And formed this on alum on my 11 inch Logan. Might be a bit much for a smaller lathe.
But pic is of 1 in alum stock and tool I ground.
1543945580616.jpeg
 
Last edited:
What kind of tolerance are we talking about and by what means are you to measure this? What is the mating part? What quantity? If moderatly low tolerance and high quantity I would say to hand grind a form tool then the profile would be cut in one operation.

sooo... I'd have to grind a form tool that is at least slightly over 0.5" wide cutting edge for a 1" dia round stock...
But I think that is pushing it for my little lathe and the little AXA tool holders you think ?

Or am I misunderstanding.
 
If the curve doesn't have to be geometrically accurate, it can be done free hand. You can rough out the curve by calculating points along it and cuttin with a combination of the compound movement and cross slide movement. Set the compound so movement is parallel to the spindle axis and cut a series of steps. Depending on your patience and the number of steps, you can get fairly close. Once the rough profile is cut, use a file to break the steps. As you get close to your desired shape, switch to emery cloth or sandpaper. I use WD40 as a lubricant with the sandpaper. Switch to finner grades of sandpaper until you reach your desired finish. It is possible to reach a mirror finish that way.

What kind of tolerance are we talking about and by what means are you to measure this? What is the mating part? What quantity? If moderatly low tolerance and high quantity I would say to hand grind a form tool then the profile would be cut in one operation.

not a large quantity, but sufficient to warrant a repeatable method.
low tolerance definitely.

There is no requirement dimension - just a rough target. A half-dome has proven to be not a shallow/soft enough contour for the application, and a simple very shallow angle also does not work well, so the short dome (large radius) seems to work the best. I made a test on hardwood - by hand, which was close enough I think to prove the shape.
 
Forgot to ask, what is the material? A four way tool post is more ridged then the QCTP. There is no better way than to try.
 
Is the dome mostly cosmetic you do you need it to be repeatable with some precision? If it is just for appearance I've had good lock using a file to shape.

I have a radius cutter, and I don't see why it couldn't be set to cut a much larger radius. I don't think it is limited to cutting 1/2 ball ends.
 
Here is how I did the nose case on my model radial engine. The section shape had to be pretty accurate. If there is a way for you to generate a table of convenient X-Y step-over coordinates corresponding to your curve shape (ideally a CAD program which makes quick work of it). Then its basically a series of cuts, blue the surface & careful profiling with file, then abrasive paper until the blue is gone. Depending on your desired resolution & how the work is mounted, I would make one dimension nominal (like 0.020", 0.040", 0.060...) and then let the corresponding intercepts be whatever odd dimensions they come out as.

IMG_7158_edited-1.jpg

IMG_7159_edited-1.jpg

IMG_7162_edited-1.jpg

IMG_7173_edited-1.jpg
 
Back
Top