Turning (approximating) a Domed Surface

[Dr_GMJNParticipant @dr_gmjn]

All,

I need to turn a dome shape for a pair of cylinder caps. Similar to a typical smokebox door kind of thing.

I've heard of using co-ordinates on the cross and top slide, presumably in conjunction with turning the top-slide to different tangent angles. Can anyone tell me the procedure, and also what's the best way of smoothing the facets generated? I want them to be identical.

Material is cast iron. Radius of the dome is about 50mm, and the diameter of the domed disc is about 25mm

Thanks.

[Dr_GMJNParticipant @dr_gmjn]

[John HinkleyParticipant @johnhinkley26699]

That size of dome lends itself ideally to a radius turning attachment.

John

[John HaineParticipant @johnhaine32865]

…or CNC turning…

[Clive Brown 1Participant @clivebrown1]

I would do that type of turning with a hand-held graver. Mine is a length of 1/4" sq. hss mounted in a file handle. A hand-turning rest is required. Mine is usually a piece of square or rectangular BMS suitably positioned in the tool-post.

The graver is sharpened by grinding its end-face flat at about 45 deg.

Works very well on most materials.

[John HaineParticipant @johnhaine32865]

Or a Turnado – type attachment for the lathe? Approximate in facets them smooth by hand turning?

[JasonBModerator @jasonb]

I tend to rough them out with combined hand twiddling of the cross and top slide handwheels to get a shape that looks pleasing to the eye and then finish with a small handheld scraper. If you want them both identical then work out some co-ordinates in CAD and use those with a small round nosed tool then blend by hand

[Dr_GMJNParticipant @dr_gmjn]

Thanks all.

Jason – I’m not sure what the co-ordinated would do.Do you mean drive, say, a 0.8 radius tool to a series of locations, and blend the remaining peaks? I’d imagine I’d need dozens of positions to do that?

Or do you move the tool to subsequent locations by moving first in x, then in y to give steps, and blend the steps?

It should probably be obvious but I can’t visualise the process.

[John PParticipant @johnp77052]

Posted by Dr_GMJN 27/09/2021 16:30:39

All,

I need to turn a dome shape for a pair of cylinder caps. Similar to a typical
smokebox door kind of thing.

I've heard of using co-ordinates on the cross and top slide,
presumably in conjunction with turning the top-slide to different
tangent angles. Can anyone tell me the procedure, and also what's the best
way of smoothing the facets generated? I want them to be identical.

Material is cast iron. Radius of the dome is about 50mm, and
the diameter of the domed disc is about 25mm

Thanks.
—————————————————————-
The easy route to turn this type of radius
is to use fixed length rod the radius you need
with point on each end and one fixed place
on the lathe bed for one end of the
radius rod and one fixed place on the crosslide
for the other end see sketch.

In use some pressure is applied with the carriage
feed handle to keep the rod in place.
With the tool tip on centre the radius rod is parallel
with the lathe bed,as the crosslide is moved out the
pressure applied to the carrage feed moves the tool
in the required radius,depth of cut is applied via the topslide.

John

[Clive Brown 1Participant @clivebrown1]

I made this cover recently with the hand-graver I described. Pretty much free-hand and took only a few minutes.

A bit more care perhaps if 2 identical domes were wanted.

Brass in this case, but CI would be much the same.

[DC31kParticipant @dc31k]

It might be worth drawing or visualising what he wants to do before making too many more suggestions.

Using the 50mm radius and 25mm diameter mentioned in the first post, the curve pokes out about 1.6mm at its centre.

[John PParticipant @johnp77052]

Just found this photo re my last post showing

set up for turning small radius.

John

[bernard towersParticipant @bernardtowers37738]

John p , just beat me to I have videos but the site won’t let me upload them. Easy to make and set up and fascinating to watch especially if you have pcf. Dr_GMJN if you want videos message or mail me.

[JasonBModerator @jasonb]

Posted by Dr_GMJN on 27/09/2021 18:41:31:

 

Jason – I’m not sure what the co-ordinated would do.Do you mean drive, say, a 0.8 radius tool to a series of locations, and blend the remaining peaks? I’d imagine I’d need dozens of positions to do that?

If you draw a section through the cylinder end cover and then draw a line to follow the profile set away by the radius of the tool's tip you can then pick co-ordinates along that line.

For a relatively flat dome like you need the cross slide can be moved a convinient amount each time say 0.020" or 0.030" so draw a series of lines at your chosen spacing and then measure offsets for each one. I have taken the point of contact in the ctr to be 0,0 and you can see the further I bring the tool away from ctr with the cross slide the deeper each plunge cut is. eg at 6.5mm from the middle I'm plunging in 0.491mm.

You then just blend in the high spots to give you this

It works with any shape such as this decorative column capital

 

Edited By JasonB on 27/09/2021 19:18:58

[JasonBModerator @jasonb]

For the benefit of DC31K and anyone else who can't work out what is needed here is a section through the cylinder cover, it's the domed cream coloured surface that needs to be machined

[Anonymous]

An alternative to the rather neat rod method is to follow a template. Here's a smokebox door being machined using an automated (hydraulic) follower:

The template could just as easily be followed by hand manipulation of the slides. The method has the advantage of repeatability.

Andrew

[Ron LadenParticipant @ronladen17547]

I have used the guide rod method a few times, its simple, quick to do and works a treat.

A picture below of the convex face I turned on the class 22 buffers.

Ron

[Dr_GMJNParticipant @dr_gmjn]

Thanks everyone – there’s a lot of interesting info here. I like the guide rod idea, but having looked at Jason’s sketch I wonder if an elliptical arc would look better.

Ill sketch some geometry out and the decide between co-ordinates or guide rod.

In fact the guide rod looks too cool an idea to ignore – I’m going to try it anyway on some scrap.

Thanks again all.

[John ReeseParticipant @johnreese12848]

A comment on the radius rod (guide rod) method. The rod needs to be parallel with the bed when the cutter is on the spindle axis (centered on the work..

[Ron LadenParticipant @ronladen17547]

Thanks John, that is true if you are working to a given radius (I wasn't on the buffers) then the rod needs to be the length of that radius dimension and set parallel to the bed. Also with a 50mm radius you can see that the rod would be too short to run between the head and the cross slide so a fixed extension is needed as seen in the John P picture. Thinking about it the extension doesn't have to come off the head if it's easier then something clamped across the bed would also work.

Ron

Edited By Ron Laden on 28/09/2021 05:20:39

[Andy_GParticipant @andy_g]

Posted by Dr_GMJN on 27/09/2021 18:41:31:

Do you mean drive, say, a 0.8 radius tool to a series of locations, and blend the remaining peaks? I’d imagine I’d need dozens of positions to do that?

Or do you move the tool to subsequent locations by moving first in x, then in y to give steps, and blend the steps?

I stepped in Z by 0.2mm and turned to calculated X positions to generate the radiussed section at the right hand end of this part. Because my cross slide has a 1.25mm pitch, I converted my X coordinates to full turns & vernier readings to save mental arithmetic & inevitable mistakes during turning! You can see the crib sheet in the background.

Very little blending needed with a large radius tool.

I keep meaning to rig up a (manual) template follower, but haven’t got around to it yet.

[Howard LewisParticipant @howardlewis46836]

50 mm radius on a 25 mm bar sounds like a job for a radius turning tool

Either a special just for this job or ma more sophisticated one for other, later, jobs.

At a pinch, maybe a boring head mounted on the toolpost, so that it can rotate about its axis, might suffice.

(Making a tool, to make a tool, to do a job! )

More ways of skinning a cat, (Not that I am suggesting making a tool for skinning cats. mFirst objection would mbe from ours! )

Howard

[JasonBModerator @jasonb]

It's not on the end of a 25mm bar, look at the images posted.

Being quite a thin item if held in a chuck most ball turning tools will foul the chuck

[Howard LewisParticipant @howardlewis46836]

Misread the bit about 25 mm diameter.

Approaching the problem from another point, make up a concave form tool, run slow and feed very gently along the lathe axis?

If you really fancy your chances grind from a 1/2" HSS toolbit!

Put some gauge plate, off set in the 4 jaw and carefully turn a concave to make the form tool. Harden / temper as you see fit. Having no clearance the tool would need to be mounted at angle to provide a clearance angle..

For brass, with care, and only one or two to make, you might even get away as is.

Howard

[Dr_GMJNParticipant @dr_gmjn]

I didn't buy castings for these end caps, I got some cast iron bar, and will make them from that.

I'm going to go for an arc, not an ellipse, because I think the apex looks more defined like that (rather than a bit flat if I use an ellipse).

I'm also going to make the apex of the dome project about 1 mm from the bolt face rather than having it flush (as per instructions), again to make the form a bit more defined. If it looks wrong I'll just try again with different settings.

[Author]

Posts