General questions about machine dovetails

You see both a and b.

Hardinge Hlv the bed is an A.

Milling machines are normally B. On the saddle, the screw covers run in that top groove.

Full flat is only useful if it allows easier cleaning. Really not needed for bearing area.

Sent from my SM-S911U using Tapatalk
 
@jwmay On my lathe cross slide and compound for example there's no contact on top, and there's not supposed to be. The contact surfaces are like my drawing A.

IMG_2664.jpeg
 
Ok thanks for trying to enlighten me.

Anyhow, I’m going to bow out.

Maybe it would help if I knew where this question came from, or what use you have in the answer. It appears I was able to offer an example of what you’d never seen fairly easily. Not having seen something isn’t really proof there’s a reason it doesn’t exist. Or that there’s a reason the ones you’ve seen were different. It’s only evidence that you haven’t seen it.

But most manufacturing decisions come down to cost. So if I had to bet on why a person may have seen more of one style than another, I would say the decision was made in the accounting office.
 
@jwmay It's hard to tell from an image, but the image you posted is probably an example of my drawing A. Getting all three horizontal surfaces in good contact would be next to impossible and provide little or no benefit.

My question was prompted by frustration with my Chinese mini lathe, and as you say, it goes back to the cost of quality manufacturing. Budget Chinese manufacturers seem to do okay with dovetails (at least most of the time) but they seem to have big problems making a saddle that rides properly on prismatic ways. I wonder why Chinese mini lathes don't have a bed like my drawing B.
 
Lathe ways fascinate me. My understanding:

The primary design requirement for a machine way is usually to allow movement in only one of six degrees of freedom, preventing movement in all others. Secondary design goals include ease of manufacturing, ease of maintenance, longevity (durability to avoid wear), and ease of assembly among other things.

Let's ignore the compound and tailstock on a lathe and focus on the cross-slide for a moment.

For a lathe cross-slide, the six degrees of freedom are:
  1. Translation along X (the only desired cross-slide movement, toward and away from the operator)
  2. Translation along Y (floor-to-ceiling movement, never desirable except for milling operations on a lathe)
  3. Translation along Z (along the bed ways — only provided by carriage/saddle movement, undesirable for the cross-slide itself)
  4. Rotation around X (cross-slide roll: side to side tipping — undesirable)
  5. Rotation around Y (cross-slide yaw: like the back end of a car sliding on ice — undesirable)
  6. Rotation around Z (cross-slide pitch: fore and aft pitching like a car going up and over a hill — undesirable)
Regarding your second question, the saddle itself is only supposed to move linearly in Z (item 3 above).

Note that the cutting forces are almost always within the extremities of the bed ways (and the saddle + cross-slide + compound has a lot of mass, so gravity is working in your favor).

Some lathe beds (Atlas, for example) have box ways (vertical and horizontal surfaces only). If the sides of box ways wear, some sort of adjustment in the saddle is usually required to take up the slack (else you could get translation in X or rotation around Y). Gravity and cutting forces are constant, though, so even wear on the top of box ways isn't a concern (the saddle just sits lower).

Some lathes (e.g. Hardinge and clones) have a single dovetail way with a large top bearing surface and angled dovetails to prevent pitch around Z, and translation and rotation in X and Y. Note that unlike most cross-slides and compounds, the top of the male bed dovetail on a Hardinge HLV-H IS a primary bearing surface. Both the saddle and the tailstock ride on this massive hunk of metal (which is one reason these are such rigid lathes).

Most other lathes have at least one prismatic ("vee") way. Prismatic ways constrain yaw around Y or linear translation in X without the need for adjustment (the saddle just tends to sit lower if things wear evenly). Gravity and cutting forces generally eliminate pitch around Z, as well linear translation and rotation in Y (though clamps that ride under the ways are still there for crashes and the like).

Since the bed is the largest part and heaviest part of a lathe, all of these shapes are relatively easy to manufacture and re-grind/re-scrape. Even with dovetail beds, it's only the smaller saddle and cross-slide that have relatively inaccessible surfaces to grind/scrape.

Regarding your first question, a cross-slide (for example) is only supposed to translate along X (item 1 above).

The cross-slide is smaller and shorter, and the cutting tool almost always overhangs the bearing points, so gravity and inertia aren't working in your favor and lifting forces are much more of a concern. Thus dovetails are typically necessary: you need to prevent lifting and twisting during normal operation.

With cross-slides though, the top of the saddle is usually a male dovetail, and the cross-slide itself is female. The top of the male portion is typically NOT a bearing surface (as KKrum points out), instead it's the "feet" of the cross-slide that bear. Those surfaces on both the saddle and the cross-slide receive the most wear from cutting forces, but are fairly accessible and easy to grind/scrape. The left "foot" also helps to prevent roll along X as it is farther "outboard" than the top of the dovetail.

The internal angled dovetail portions in the cross slide are more of a pain to repair (usually involving handscraping rather than a power scraper). They see less uneven wear, though, as the cutting forces from turning and facing are typically more downward than toward the tailstock.
 
As a precursor to scraping dovetail surfaces, a hacksaw is used to seperate sutfaces. A cut is made where surfaces (at an angle) meet. As per Edward Connely's book that I went to the yrouble of photocopying many years ago. No wound up tossing it.

(Apologies if this was already mentioned).
 
It's hard to tell from an image, but the image you posted is probably an example of my drawing A. Getting all three horizontal surfaces in good contact would be next to impossible and provide little or no benefit.
My thinking exactly. Glad to know we’re on the same page with that. lol

My question was prompted by frustration with my Chinese mini lathe, and as you say, it goes back to the cost of quality manufacturing. Budget Chinese manufacturers seem to do okay with dovetails (at least most of the time) but they seem to have big problems making a saddle that rides properly on prismatic ways. I wonder why Chinese mini lathes don't have a bed like my drawing B.
Definitely cost is the reason for the poorly built products available. And these days, with giant groups of buyers happily buying low grade products and joyfully modifying them into decent machines, it’s unlikely to change. I’d go so far as to say it’s become trendy to buy budget built stuff and fettle it to perfection.

It’s a good feeling to “improve” a brand new purchase. Think aftermarket exhaust on your Mustang, or bigger tires on your pickup…

Prismatic ways on a mini lathe are gonna be a couple orders of magnitude cheaper and faster to produce than a bed and carriage with accurate dovetail slides.

But it’s finally starting to get through my thick skull. Thanks for your patience.
 
Thanks for that thorough explanation, @Rex Walters. And thanks especially for defining your axes. 3D game and CAD developers have holy wars about which way the axes go, and there are many systems.
 
I think with machine tools the general convention is that Z is the spindle axis, but X and Y might be debatable. A lathe is basically a vertical mill laying down on its back.
 
A cut is made where surfaces (at an angle) meet.
To clarify, it's impossible to scrape all the way into the interior corner of the dovetailed section. You need to relieve the inside something like this:

Screenshot 2024-02-03 at 8.35.40 AM.png

As per the Connolly book, you can use a hacksaw blade (or 2-3 bundled together) to abrade cast iron. I've done it: it works but takes a long time. A cut off wheel or similar is probably easier if you can maneuver it into place.
 
Back
Top