Thread: Should I scrape?

My second hand 12x36 has a quiet badly worn cross slide with some scoring, worn to the point that it affect the cross slide travel and I want to fix it but the only time I have ever scraped something was a 2" block at trade school with a el'cheapo scraper.

The limited tools I have at my disposal are:

Surface grinder

Limited use of a surface plate

height gauge

And I will probably have to grind my own scraper from a file.

Will it be possible to get the job done and parallel with what I have? and if so, what is the best approach, a step by step guide would be appreciated.

My plan was to first remove the cross slide and grind it flat then use it as the master to blue and scrape the saddle dovetails, providing it will still make contact with the saddle dovetails...

I don't have a straight edge but I could machine up some cast iron to fit in the dovetails and grind an edge if that would be good enough to get me out of trouble.

I'm still not sure how to get both surfaces running parallel to each other, and then the gib will need attention after all the material has been removed

I don't want to start something I can't finish, so.. what do you reckon?

I can't speak for you, and whether you should or should not attempt it, and I am by no means even a novice at this game, but my thoughts on your post anyway, hopefully will give you some things to think about: If you look in the project threads (Project Banarama), you will see I tackled the compound slide assembly and the flat ways of my cross slide with not much more tooling (no grinder either!). I didn't bother with the dovetails of the cross slide at this stage, as I mentioned in that thread it was more of a 'preliminary clean up' operation to get them flat after doing my compound. The gib does hang out the back a bit, but it's functional. In that thread somewhere you will see a small dovetail prism I made up to use for the dovetails on the compound - unfortunately not long enough for the cross slide dovetails.

Aligning of the two dovetails can be done with a few measuring jigs that are easily made up, however I took the lazy approach, and used two round pins and a micrometer to measure across them, scraped them in until they measured parallel. Of course, I had the advantage that they didn't have to align to anything other than each other, as the entire compound rotates, with the cross slide of course you will need to ensure that they are perpendicular to the bed ways, with any slight error tending to be closer to the headstock at the rear bedway. One of the reasons I haven't tackled mine yet - I need to go through and align the bed correctly, and then I would like to check out the underside of the saddle before getting too carried away on the top.

Nick Mueller on Youtube has some good videos that indicate the process and sequence of scraping in various parts.

While you can certainly grind the ways on the cross slide, beware that the top is very unlikely to be flat, and will warp the part when the magnet pulls it down, thus giving you a warped surface once the magnet is released. You can of course shim under the part to try and solve this, but a ground surface is not going to give you a great result with the blue. I'd blue the flat ways on the cross slide against the surface plate first, and then consider scraping that flat rather than grinding, but if you do grind scrape it afterwards. Check against the top surface before starting, to see where you need to work in order to keep the flat ways on the bottom parallel to the top surface. Remember the compound relies on this top face being in plane with the bed - imagine the path the tool takes when using the compound slide for a shallow taper if your cross slide is tilted 5 degrees down on the tailstock side. Then once you've scraped the flat ways flat against the surface plate, you can dust the top of it off on the surface grinder.

Once you have this part scraped and flat, and assuming that the cross slide is longer than the saddle, you then use this in order to blue and scrape the flat ways on the saddle to mate.

Only now is there any point in looking at the dovetails, and this is where you will require a dovetail prism/straight edge that has been scraped in against your surface plate, as well as a sled of some sort to measure their alignment. I think with my compound I did the base dovetails first, then used that as the master for the top piece, but if you watch the Nick Mueller videos the correct order will become apparent.

Gibs can be shimmed - see Stefan Gotteswinters Youtube video on rebuilding an import compound slide WITHOUT scraping for the general idea.

Scraping is fairly easy. Scraping to alignment, and fiddly little dovetails - somewhat more difficult. Keeping track of which surfaces need to be parallel/perpendicular/planar to other surfaces, now that is the real trick. And for advanced technique, Marcus will likely be along soon to point out how much 'curve' to scrape in to allow for later wear - way above my paygrade.

You need to fairly carefully consider how far down the rabbit hole you wish to go, as ideally you would start at the bottom up. Scrape the underside of the saddle, then scrape the flat ways on top planar to the bed ways. Then scrape the dovetails "perpendicular" to the bed ways (allowing for a very slight bias to facing concave). Then of course scraping in the top part of the cross slide, paying attention to the top being parallel to the flat ways on the bottom in order for the compound to be correct. Then finally do the compound. Of course, I did cart before horse, as after doing my compound the cross slide and saddle were that warped I figured I may as well make it at least flat, as that would need to happen anyway - and there was quite an improvement in feel and operation just from doing those flat ways. Is it right, no, but it's a hell of a lot better than it was.

However, I would suggest if you're not too confident about the process, consider starting with the compound slide assembly instead. It can be treated as a self contained unit, in that as long as the flat ways are parallel to the circular base, there is no other outside alignments to worry about. Marcus will comment that you'd ideally scrape it to account for wear in the cross slide, but if the plan is to do that later anyway, it shouldn't matter too much (unless you use the compound a lot). Don't underestimate the benefits of scraping in the compound - if you blue the top part on the surface plate, and find it as warped as mine, you'll gain quite a lot of rigidity by scraping it all in, and some valuable experience/confidence to tackle (or avoid!) the rest of the job.

And plenty of reading of PracticalMachinist is invaluable in getting a handle on the process. Marcus has a Facebook group going that would likely be of assistance if you're on Facebook (I'm not).

There are plenty more around with infinitely more knowledge than I have on the matter, and hopefully some of them will make some input, but in the meantime hopefully I've given you a few things from a beginners perspective to ponder that will ultimately help.

Having gotten carried away answering the actual question, and now taking a step back - is it even necessary to scrape it, or is there an easier way to solve the problem?

Do you have a picture of what you’re dealing with? Enough wear to bind up the cross slide on what I assume is a Chinese or Taiwanese lathe sounds rather unusual, although if it’s been used fairly heavily in an industrial setting or never oiled certainly possible. It’s definitely not something simpler, like the cross slide screw being worn, and someone having tightened up the split nut to reduce backlash? Or is there some big burrs and ridges causing the problem that could simply be stoned out or removed with a burr file? Perhaps some chips stuck in the wrong place? Just spitballing ideas here, without knowing how much you’ve looked into it, but maybe scraping it is not strictly required in order to reach your goal?

• I really wouldn't bother with grinding a file, making a clamp style scraper that takes a carbide insert will save you a lot of time and pain in the long run

• I wouldnt bother grinding the top, as its already ground but peppered in dents, scraping will take the average of the original plane.
• the standard height gauges are awful for scraping when working to tenths, make a three point stand if you can.
• a cheap and nasty slow speed diamond lap is quick to make and a necessity for working with carbide scrapers
  
• if you make a dovetail straight edge from cast iron the design is important, I love my L shaped dovetail straight edges but the uneven weight means they must be scraped and used as a master in a very specific way or they will induce an error.

I've just pasted this from a post I made in the facebook group I made for aussie scrapers hopefully this helps re the scraping of dovetails. should you have any questions its probably easier to give me a bell on the landline (07) 5522 1726, I don't scrape as a business I'm just a metal machinist that really enjoys scraping.

engine lathes fall into two distinct categories "short slide" style as seen on the green DSG and the more modern "long slide" style on the colchester. should you rebuild a lathe of the Modern long style you can scrape the flat ways using a surface plate but the fixed dovetail will require a dovetail straight edge.


Should your dovetail straight edge be too short (risking a concave error) or if you are afraid of causing an convex error from a straight edge of sufficient length it is a simple matter to scrape the side adjacent to the fixed dovetail so it can sit against the plate in a stable fashion, doing so allows one to accurately inspect the dovetail as long as the face in red (A) has two points of contact against the flat way and the green face (B) is either scraped or sits on three points for stability, once scraped the cross slide can be used as a master to scrape in the mating ways on the saddle


The for mentioned method however is not case for a short slide cross slide, on a short slide lathe the cross slide ways on the saddle must first be scraped and then are used as a master to spot the shorter cross slide. This ideally requires a dovetail straight edge of sufficient length but their are methods around it. The short slide also differs in that unlike its longer counterpart it is not scraped for true flat it is scraped with a heavy relief in the middle 30% of the flat ways, fixed dovetail and gib otherwise it will quickly lose its stability as it wears.

diagram 121.png118254643_645204839456602_533979952911406728_n.png

Thanks for the good info Jekyll and Hyde & marcuschrist.

I've attached some photos of what I'm dealing with, they aren't great pics but you can make out the wear tapering from front to back, it definitely had a hard life before me.

I've had it apart a couple of times, adjusted the nuts and backed off the gibbs, I was able to improve cross slide travel, but it's still not great, and I would like to address the wear, even if it's not the direct cause of my issues it's something I want to fix and work my way up from there.

marcuschrist; I'll have a look for your facebook page and contact you there.

It might be time to buy a dovetail straight edge, and/or a regular straight edge for the task, and any future tasks, I'll check all the usual places.

IMG_0468.jpgIMG_0466.jpgIMG_0471.jpgIMG_0467.jpg

Here is some light reading for you
https://metalworkforums.com/f65/t152...slide-scraping

I had a quick look, there are at least some of the pictures still there. (some may have gone missing due to a forum screw up)
I can likely find any missing ones if you need a look at something.

The flats have to be scraped before you can accurately measure the dovetail parallelism but its worth grabbing some dowel pins and taking a quick measurement too see some actual numbers as it's very possibly an issue with screw/nut alignment. the following will get a bit long winded and I apologize but it will likely save a few people some confusion later down the track

it should be said that most scrapers when doing fine finishing work will use indicators whose increments are in either 0.002mm (2 micron) or 0.0001" (a tenth). when working to a tenth there is a whole host of issues you can encounter that will induce error. one such error is the Reuleaux effect, it is pictured below in its pure exaggerated form. the cheap and nasty import dowel pins found at fastener shops in aus will often suffer from this in the realm of 0.0004"+ and this form of error cannot be detected with a micrometer as the diagram below demonstrates, however when you use such a dowelpin to measure a dovetail you can get drastically different measurements depending on how either dowel pin is orientated so you may measure an error, step scrape according and remeasure only to find a completely different set of numbers that make no sense. its one of those many chasing your tail moments you can find when scraping.
reuleaux_triangle.gif

The solution? I use two matched sets of Hoffman rollers for all dovetail work which is extreme overkill but for those looking to be accurate on the cheap the replacement individual gauge pins are dirt cheap from Meyer. if you want an interesting demonstration of the Reuleaux effect, take an awfully bell mouthed 3 jaw chuck or shim some roundbar at the rear of the jaws and take a fairly deep cut with no support from a center. you can mic the part and get fairly consistent measurements around the OD however if you set an indicator against the work you will see the trilobe. hopefully this has been helpful to some and solved a few mysteries.

Originally Posted by NedsHead

Thanks for the good info Jekyll and Hyde & marcuschrist.

I've attached some photos of what I'm dealing with, they aren't great pics but you can make out the wear tapering from front to back, it definitely had a hard life before me.

I've had it apart a couple of times, adjusted the nuts and backed off the gibbs, I was able to improve cross slide travel, but it's still not great, and I would like to address the wear, even if it's not the direct cause of my issues it's something I want to fix and work my way up from there.

marcuschrist; I'll have a look for your facebook page and contact you there.

It might be time to buy a dovetail straight edge, and/or a regular straight edge for the task, and any future tasks, I'll check all the usual places.

IMG_0468.jpgIMG_0466.jpgIMG_0471.jpgIMG_0467.jpg

Hi Ned, Guys,

Those slides don't look as if they have seen any lubrication for a long time !

It does beg the question Marcus, where does one buy gauge pins locally that are trustable? I am soon to embark on a scraping adventure also and am tooling up atm. My lathe is old school so no lovely flat sides/top etc. Not wishing to hijack, hopefully the answer adds to the topic here and helps the OP.

Originally Posted by StrayAlien

It does beg the question Marcus, where does one buy gauge pins locally that are trustable? I am soon to embark on a scraping adventure also and am tooling up atm. My lathe is old school so no lovely flat sides/top etc. Not wishing to hijack, hopefully the answer adds to the topic here and helps the OP.

I lucked out and got the two matched sets of Hoffman Rollers and one larger set of Hoffman rollers that were more or less unused, all originally came out of the toolroom of an munitions factory. I had them for quite some time before I actually used them and was getting mixed results from dowel pins previously without understanding the cause of problem. I've never had to purchase any gauge pins locally and I'm an eternal cheapskate that will wait weeks if need be for international shipping to buy quality NOS rather than buying import goods at a premium locally. You can of course simply mark what ever you happen to use if its roundness is questionable and orient both pins the same way each time you take a measurement but its always nicer to not have to bother. also my fault for derailing the thread a bit but I am curious to see the numbers to get an idea of roughly how bad the wear actually is on Neds cross slide ways

Originally Posted by marcuschrist

I lucked out and got the two matched sets of Hoffman Rollers and one larger set of Hoffman rollers that were more or less unused, all originally came out of the toolroom of an munitions factory. I had them for quite some time before I actually used them and was getting mixed results from dowel pins previously without understanding the cause of problem. I've never had to purchase any gauge pins locally and I'm an eternal cheapskate that will wait weeks if need be for international shipping to buy quality NOS rather than buying import goods at a premium locally. You can of course simply mark what ever you happen to use if its roundness is questionable and orient both pins the same way each time you take a measurement but its always nicer to not have to bother. also my fault for derailing the thread a bit but I am curious to see the numbers to get an idea of roughly how bad the wear actually is on Neds cross slide ways

Ever checked out endmill shanks for roundness? I may or may not have used two pieces I cut off a broken endmill (depending on your answer )

*edit* Actually, on reflection I don't think I had a broken endmill the right size, think what I actually used was a couple of pieces of 8mm ground rod for linear bearings. Surplus from my cheap Chinese 3d printer, so probably not too good! Nonetheless, worth knowing if endmill shanks are any good?

[QUOTE=StrayAlien;1977500where does one buy gauge pins locally that are trustable? [/QUOTE]
Well they aren't gauge pins, but they should be "pretty" round.
https://metalworkforums.com/f65/t167...-ground-dowels
Hounslow Engineering - page 3

IF I recall correctly, you can check for Reuleaux effect by having the dowel in a Vee block while measuring. Though there are other shapes, so if you really want to get into it you need different angle Vee blocks.