Page 1 of 3 123 LastLast
Results 1 to 15 of 40
  1. #1
    Join Date
    Aug 2010
    Location
    Toorloo Arm, VIC
    Age
    39
    Posts
    1,270

    Default Project Banarama (AKA the LSO based "My first scraping kit)

    So the inevitable happened, after my post a while ago on the ways of my AL335, which I am now referring to like many before me as a 'lathe shaped object'.

    Armed with a Sandvik scraper, a Carbatec surface plate and some Permatex Prussian Blue, I launched an assault on the compound, on the basis I couldn't exactly make it much worse without seriously trying.

    First picture shows the finished parts before reassembly. Is it perfect? Probably not, I'm sure the gurus will be rolling their eyes at the terrible job, but there is a laundry list of improvements that make it far better than what I started with.

    compound done.jpg

    The flat ways on the upper slide were touching on two diagonal corners only, and rocking by about 0.03mm. Scraped flat, whilst checking periodically the top face with a test indicator to try and keep it pretty much parallel to the ways. For most of it, that makes no difference, but there is a 'pad' that the toolpost itself sits on that I wanted parallel to the flat ways. Doesn't actually matter, but that's what I wanted.

    The circular base was convex by about 0.05mm across the 25mm odd distance next to the hold down studs, now ever so slightly concave. It must have been rocking to some extent there since it was mostly sitting on the 'ears' around the hold down studs. The keen observer will have noted I scraped a non sliding surface, which was sort of a mistake, but not for the reason expected. I wanted it flat, so that I could clamp it to the rotary table on the mill without it distorting (not having a lathe functional for some strange reason). The progress shot will show how much out it was.

    compound ref.jpg

    Originally, however, I'd intended to clamp the 123 blocks (needed due to a circular boss on the base) in the mill vice, skim the flat ways to take out a bit of bow and very slight height difference (maybe 0.02mm), and then pass over the top with a facemill - THEN do the base on the rotary table. Problem was, when I started indicating it in the mill, the difference in height between the flat ways was completely different to what I'd measured on the surface plate. Given that I could see on the surface plate the 123 blocks were sitting up in a V shape thanks to the convex base, I figured the vice must have been pulling on the 123 blocks and warping the base. So I scraped the top surface flat, and then ran the indicator over the circular base. 0.2mm difference from side to side. Bolted it to the rotary table, after checking that it stayed flat through it's rotation, and checked that I still had that 0.2mm variation, then milled the base. Checking it back on the surface plate, now shows that the flat ways had 0.2mm difference in height, as well as a touch of bow. Those 123 blocks must have been sitting pretty skewed, I thought. Checking the base showed, annoyingly, that it was still a touch convex... The quick witted among you may see where this is going...

    Clamp it back in the mill vice by the 123 blocks, start indicating it in... WHAT is going on here? Try as I might, I couldn't get it to sit down with less than about 0.35mm difference between the flat ways, despite the top sitting flat. Shimming with paper, loosening, readjusting.... With a sudden burst of intelligence, I thought 'I wonder if...' Marked a texta mark on the top face, swung the indicator over it hanging out to the left. Then moved the table over, swung the spindle around 180 degrees, and swept the same spot. 0.15mm. SON OF A..... Some weeks previously, I'd tilted the head on the Bridgeport to the side to put an angle on something I was doing. I'd then eyeballed it straight again to finish off what I was doing, as it was non critical anyway. I THOUGHT I had properly retrammed it afterwards... Turns out no, I never got round to it.

    So, in conclusion, measure six hundred and fourty three times, then measure once more for good luck. Then, maybe think about cutting. Turns out the top face was tilted at 0.2mm to all the critical features, which is where the mistake comes in. What got me in addition to the mill tram confusion though, was that I meaured the thickness of the flat ways when I got this 0.2mm figure in the first place, and sure enough, one side was 0.2mm thicker than the other... In the end, makes no real difference, just had to make a new gib, and reskim the bottom on the rotary table the tiniest amount to get rid of the convexity. But in any other part of the machine, that could have been a fatal flaw.

    compound base.jpg

    Moved on, scraped the flats of the base to the 'master' of the upper. Now, given that the upper part of my compound is the first thing I ever scraped, I'm not going to give any advice or instructions. Except this. Permatex Prussian Blue is terrible. End of story. Do not pass Go, do not collect $200. Seriously, don't even try using it. The red in the photo is permanent marker I was using as a highlighter, and you can see smears of Permatex in there, and it's still bloody difficult to see whats going on. Australia Post finally delivered me some pigment towards the end of the compound, and I mixed up some of Machtools secret sauce from the old scraping thread on here. Night and day difference, with the Permatex you'd have it thick on the plate, and still look at the part going "I think there's a bit of blue there...." Phils (Machtool) mix, even when the surface plate looks DRY, rub the part on it, flip it over, and yep. Deep blue spots there, there, there. You need very, very little. It's almost more of a paste compared to the Permatex, but that's what makes it work so well. I also have some "Stuarts Micrometer Blue" somewhere on the way, it will be interesting to see how it goes. Cheap and easy, think it was about $15 on eBay shipped from the UK, so might be a viable option for those disinclined to roll their own. HiSpot and canode seem to be almost unobtanium in Australia without some expense...

    compound permatex is terrible.jpg

    A short break, to take on my second ever scraping project - a dovetail prism (I think it's in the background of a shot somewhere). Which first had to be rescued from the old Alfetta flywheel it was hiding in. After tilting the head to cut the angled face, I wound it back up to 5 degrees off tram, and left it there. Stops so much oil leaking out, but very obvious that it needs retramming. Lesson learned.

    Prism scraped, and..... tiny lumpy dovetails are painful to scrape. Let us never speak of them again. The two parallel dovetail ways were not. By about 0.06mm. They are now, well within 0.005mm. Guessing by the variation (or lack thereof) in where the line lines up on the micrometer, I'm claiming it's probably close to 0.002mm. Maybe better, but given I'm interpolating a 0.01mm micrometer and measuring dowel pins, that would be very bold indeed. About this point my pigment arrived, so after mixing it up and seeing how much easier it made it, and how much more information it gave, I went back over everything and touched up a few spots. A pre-adjustment shot, taken with the surface plate basically dry as I mentioned earlier. This is as bright as Permatex gets when you slather it on, but with Machtool home brew, this is about as... not bright? as it gets. Difficult to take photos of light bluing, especially under fluros.

    compound nearly done.jpg

    Machined and scraped the gib to fit everything else. I had intended to put new oil feed holes in, with oil grooves, but the way the dovetails overlap it just wasn't going to work out. They only contact on about 50% of their height, just not enough room for an oil groove. It's kind of like the flat ways on the base were cut about 3mm too low (okay, maybe it's 3.2mm - on one side anyway). If the upper sat 3mm higher, the dovetails would have almost full overlap. Anyway, being that it's a compound, and I don't actually use it that often (this was actually as much, if not more, about improving rigidity of the toolpost connection to the cross slide as it was about the compounds accuracy and operation), I decided it wasn't worth the hassle of putting in ports and grooves just for the flat ways. So while I'd intended to put a zigzag oil groove in, and not do any flaking due to lack of real estate, I actually did the opposite. Careful inspection of the first photo of all the parts done will show some terrible flaking done in the 'never exposed' area of the upper, after a whopping 2 minutes of practice on a spare chunk of cast iron.

    Put the upper in the mill, skimmed the 'pad' that the toolpost sits on. Took out the almost 0.1mm of lumps in it, then had to scrape a little to get it square to the flat ways. Came off the mill about 0.01mm out, probably because the sides of the upper aren't quite square. If a part doesn't sit hard on both parallels in my vice, the back isn't square to the bottom - end of story. And the front parallel was just ever so slightly floating. It could also be wear in the Y way of the mill. A surface grinder and mag chuck would be nice...

    Assembled the whole shebang, and it's just effortless to operate, one finger on the handwheel and silky smooth. Now, before I bolt this back on the lathe, let's just see exactly how horrible other parts of this beginner scraping kit are....

    cross slide 1.jpg

    Now, I already knew the cross slide looked like this, having peered under it with a mirror previous, but I wanted to see just how bad it actually checked out, and whether I could do something to improve it. I'm not ready to go all out on this, I need to finally get onto building the stand for the lathe, so that I can actually level the damn thing without it lifting a foot off the ground at the tailstock end. I know the bed is twisted, but without a decent stand I haven't been able to sort it. Not a fan of bolting things to floors if I can help it, especially in my space. However, I wanted to see if there were any big lumps it was rocking on, and what was actually there.

    cross slide 2.jpg
    cross slide 3.jpg

    Interestingly, this is the ONLY surface in the whole cross slide assembly that has been "factory scraped". All the dovetails, and the saddle are straight mill finish (possibly grind on the saddle).
    Notice those bright shiny spots in the middle? Well, that could get interesting. Because I put a heavy layer of blue on the surface plate, and this is what I got...

    cross slide first blue.jpg

    No blue anywhere near those points. In fact, where is the blue? Hmmm. Put even MORE blue on the surface plate. We're now at the unquestionably far-too-much blue to get any sort of useful information from point (If we weren't before...)

    cross slide try again.jpg

    Errr..... Right.

    Looks about the same, just brighter. Well, since I'm here....

    cross slide progress.jpg

    This is where it's at after 8 passes. Seems like it's pretty straight by how quickly it started to come in. Which makes me suspicious that the saddle is Banarama part... whatever I'm up to. Not going much further than this at the moment, probably a couple more passes to get those two corners in a bit, then a quick blue of the saddle to see whats going on there. If it's got a big hump, I'll rough it down to more or less flat, then throw the whole thing back together. My logic of it is that if I know the cross slide is sitting more or less flat on the flat ways, and not rocking on random high spots as it moves through it's travel, it's going to make measuring and analysing a bit easier once I've built the stand etc. Just one less variable to deal with at that stages, and as an added bonus, it will be a bit nicer to use in the meantime.

    Oh yeah - and as an added laugh, can someone explain this, er, "design choice" to me? Cause I'm out of ideas.... But I did laugh when I took it out. Only thing you can do with this things, is laugh. They do make excellent learn-to-scrape kits though, you could do a absolutely terrible job and still end up with a better machine! Take even a bit of time, and the quality of that result goes up quite rapidly...

    please explain.jpg

  2. #2
    Join Date
    May 2011
    Location
    Murray Bridge S Aust.
    Age
    71
    Posts
    5,942

    Default

    Thanks for the interesting how you did it report and pics.
    I think the design choice was "we're out of counter sunk's but a cap screw will still do the same job".Makes you wonder about some people.
    Kryn
    To grow old is mandatory, growing up is optional.

  3. #3
    Join Date
    Sep 2010
    Location
    Port Sorell, Tasmania
    Posts
    74

    Default

    Nice write up J&H, thanks for taking the time to post it here. I have a 336 and pulled the compound and and cross slide off for a good clean when I got it. The ways on mine look about the same standard as yours and I expect the contact on the mating surfaces is similar. Would like to do the same thing with mine. Do you have any training in scraping or are you self taught? If a beginner can do what you have done it gives me confidence.

    Do you have a design in mind for the lathe stand. I also wound prefer not to have the lathe bolted to the floor. I am thinking along the lines of stand with a heavy "backbone" using 200 or 250mm SHS on legs with levelling pads.

  4. #4
    Join Date
    Mar 2011
    Location
    Southern Flinders Ranges
    Posts
    1,536

    Default

    Thank you for the good read over this mornings second coffee..

  5. #5
    Join Date
    Jul 2017
    Location
    Burleigh heads QLD
    Age
    29
    Posts
    114

    Default

    scraping everything into parallelism does make everything easier to inspect but the best practice method for a compound specifically is to scrape the base with a inclination upwards in its travel equal to the deflection with the gib set appropriately. The further the top slide traverses the more it deflects downwards off a true center line which causes the taper being cut to increase in diameter along its length but not in a linear fashion, in reality you really only really notice in a lathe whose compound is worn as well as saddle/cross slide wear which causes the True path of the cross slide to dip exceedingly low, Lower Tool height causes a larger diameter but the relationship is fairly minor. Measuring this deflection allows you to preempt it and compensate according which will lessen its effect later in the lathes life.

    a good example of this would be if you had a sample taper chucked up and dialed in on that taper with an indicator mounted to the top slide. under no tool load you could be perfectly dialed and assume all is good, however unknown to use at the time the saddle is worn convex, the cross slide has more wear towards the chuck and there is slop/wear in the compound. once you begin turning the compressive load would cause this stack of errors to drive the true path of the compound slide downwards diving deeper down the further it traverses outwards. this drop causes a larger diameter where you may have expected to profile in black you get the profile in red.


    Cross slide deviation diagram 1.PNG
    RE blue I generally either mix my own or take an off the shelf blue and add prussian blue oil and a small amount of diesel to help it go into solution and for highlighter I just use the red iron oxide pigment but only for finish scraping. one of the local scrapers (located in brisbane) has a 1 litre bottle of the canode and has happily handed portions in the past, I'm no huge fan on water based blues but your wife will stop cursing at you for all the blue stains. Outstanding work by the way, as a first project you've nailed it.

  6. #6
    Join Date
    Aug 2010
    Location
    Toorloo Arm, VIC
    Age
    39
    Posts
    1,270

    Default

    Quote Originally Posted by tony_A View Post
    Nice write up J&H, thanks for taking the time to post it here. I have a 336 and pulled the compound and and cross slide off for a good clean when I got it. The ways on mine look about the same standard as yours and I expect the contact on the mating surfaces is similar. Would like to do the same thing with mine. Do you have any training in scraping or are you self taught? If a beginner can do what you have done it gives me confidence.

    Do you have a design in mind for the lathe stand. I also wound prefer not to have the lathe bolted to the floor. I am thinking along the lines of stand with a heavy "backbone" using 200 or 250mm SHS on legs with levelling pads.
    Hi Tony - I wasn't joking when I said the flat ways on the compound upper were the first thing I've ever scraped. Never done a course, didn't even own a scraper until I bought one for this job. What I HAVE done, is read many threads on Practical Machinist, where you will find Richard King ever present in any scraping threads. Now, I'm not real sure what to make of the guy himself from his posts, I get the feeling he might have been an evangelical preacher in another life, but he has his "4 rules". If you keep them in mind as you go, you'll be well on the way. A little more on that in the post to follow, I saw your reply this morning, and took some photos as I experimented a bit today. Stefan Gotteswinter on Youtube has also done two compound slide videos, one on scraping a Schaublin compound, and a "no scrape" approach on a Chinese compound. I've used a little of both in my approach. Nick Mueller also has some helpful videos up on Youtube. Plenty of reading and viewing out there, including some old threads on this forum. Take note of advice from Machtool on this forum and also Practical Machinist, he ran a few classes here at one point, and knows what he's on about (having done it for many years professionally). Always plenty of sorting wheat from chaff when reading on the internet, as always it's up to the reader to apply their own logic tests to anything said.

    H&F sell a complete compound assembly for the AL335 for about $400. Now whether it fits an AL336, maybe, maybe not. But I would imagine they can supply one for the AL336 for a similar price if needed, so worst case if you somehow manage to completely destroy it, that's what it will cost you to throw the old one in the bin and pretend you never tried. Although, in my novice opinion, you'd have to make some serious errors to get to that point. As long as you think about which surfaces NEED to be aligned to each other, in which way, before you start, have a plan of progression to the end, and keep that goal in mind as you go, checking periodically, the process is fairly slow, and you should pick up any issues fairly early. If indeed you have two diagonally opposed high corners, or a big hump in the middle, you can always start (as I did) by knocking them relatively flat with the scraper, and then see how you're feeling at that point. You'll either feel up to refining it more, and really working on getting it flat, or you'll be hesitant and decide to get some other practice in. Either way, you'll have improved things by some amount.

    Brief example of the sort of 'check' I mean - first thing I did today was plonk the cross slide on the surface plate, and swept the top with the test indicator to see if I could get away with biasing my scraping to the way that wasn't making good contact. Turns out, 95% of the top surface was dead nuts, only the front left corner was high. The triangle formed by the tangent of the circular hole for the compound studs and the left and front sides of the cross slide is high by 0.02mm. No point scraping anything other than even in my opinion, as leaving it that way means the minimum of material needs to be removed from the top. Touching off on the highest point with a surface grinder, and then taking 0.03mm off the whole thing will clean the whole lot up nicely. Unfortunately, that does mean my compound for the forseeable future will be tilted by that 0.02mm. On the plus side, I've found an excuse to seriously look at getting a surface grinder.

    If, for example, it had been 0.02mm high in the front left, but fairly evenly tapering down to the back right corner, I could have chosen to step scrape and bring the flat ways at that point down more than the back right corner, thus needing less or no cleanup on the top. Or, if the top was also lumpy as hell anyway, I might have decided stuff it, and just taken more off the top at the end to clean it up. These are the sorts of decisions you need to make along the way, and when you get down to cross slides and saddles, you'll have other alignment constraints to deal with at times. With the compound however, you are basically free to treat it as a self contained unit. Stefan Gotteswinter and Nick Muellers videos will give you much more insight into checking things along the way, and the order of which surfaces to attack when.

    Re stand - I've come up with a design to utilise some of what I already have, and to suit my desires for storage (I can't afford to waste that space under the machine). Key to this design is getting rid of 4.3m of 6 inch channel I've had kicking around for ages, and is always just too short for whatever I think of to use it for. The intention is to make a rectangle out of it to bolt the lathe to via some thick plates with jacking screws incorporated, and build legs/shelves/drawers off the bottom of this channel section. The majority of the lower part will be probably 50x50x5 SHS, and I intend to probably stitch weld 3mm plate to the legs/frame in order to brace it and help it resist racking. It will be serious overkill, at one point I calculated it would weigh somewhere around 200kg on it's own, and I think it'll be a fair bit heavier than that now. I figure this lathe could do with a serious mass increase. I can upload a screenshot of the design if you like, but it probably won't help you that much.

    Quote Originally Posted by marcuschrist View Post
    scraping everything into parallelism does make everything easier to inspect but the best practice method for a compound specifically is to scrape the base with a inclination upwards in its travel equal to the deflection with the gib set appropriately. The further the top slide traverses the more it deflects downwards off a true center line which causes the taper being cut to increase in diameter along its length but not in a linear fashion, in reality you really only really notice in a lathe whose compound is worn as well as saddle/cross slide wear which causes the True path of the cross slide to dip exceedingly low, Lower Tool height causes a larger diameter but the relationship is fairly minor. Measuring this deflection allows you to preempt it and compensate according which will lessen its effect later in the lathes life.

    a good example of this would be if you had a sample taper chucked up and dialed in on that taper with an indicator mounted to the top slide. under no tool load you could be perfectly dialed and assume all is good, however unknown to use at the time the saddle is worn convex, the cross slide has more wear towards the chuck and there is slop/wear in the compound. once you begin turning the compressive load would cause this stack of errors to drive the true path of the compound slide downwards diving deeper down the further it traverses outwards. this drop causes a larger diameter where you may have expected to profile in black you get the profile in red.

    RE blue I generally either mix my own or take an off the shelf blue and add prussian blue oil and a small amount of diesel to help it go into solution and for highlighter I just use the red iron oxide pigment but only for finish scraping. one of the local scrapers (located in brisbane) has a 1 litre bottle of the canode and has happily handed portions in the past, I'm no huge fan on water based blues but your wife will stop cursing at you for all the blue stains. Outstanding work by the way, as a first project you've nailed it.
    First off, thanks very much for the kind words, much appreciated.

    No argument from me on your suggested alignment technique. I considered whether to go to the extent of predicting wear by scraping more sparsely in areas, or 'selective' alignment, and decided for the amount I use the compound (rarely, but just enough to not be bothered with a solid toolpost), it probably wasn't worth the head scratching at the time, especially given the cart is well and truly up the road while the horse is still in the stable. Everything under the compound is an unknown, and I want to address all of it at some point - so no gain in scraping the compound to what's already there. I figured I can always revisit it at a later stage if I desire, but parallel is a vast improvement over... whatever that was as delivered. When I come to doing the cross slide for real though, I'll certainly be reading up on this area and making some decisions on allowances for future wear and deflection - same with the saddle/carriage arrangement.

    I have much to research when I get to tackling that whole area proper, but having now got the compound to the shape I was trying to make it, I can now focus on the 'thinking' side of scraping, knowing that I have a least a basic handle on the actual "doing".

  7. #7
    Join Date
    Sep 2010
    Location
    Port Sorell, Tasmania
    Posts
    74

    Default

    Thanks for the detailed reply J&H. My learning process sounds similar to yours, reading on the internet and you tube. Encouraging to see that acceptable (to me) results can be obtained.
    Did you get your scraper from an OZ supplier or have to import one?
    I don't have a surface grinder and am expecting the gibs to need replacing or shimming (as Stephan did) and that seems to me to be the biggest/ trickiest part of the whole job.

    Your thoughts for a stand sound to be along similar lines to mine.

  8. #8
    Join Date
    Aug 2010
    Location
    Toorloo Arm, VIC
    Age
    39
    Posts
    1,270

    Default

    Following on from yesterday, I kept knocking back the cross slide. Took a few shots along the way, of how I as a complete beginner teaching myself am working, and of a few things I've noticed along the way that might help other beginners get a bit more confidence to rip in a bit harder earlier on.

    I have been going fairly slowly, checking with the blue often to see I'm heading in the right direction, but I've figured one or two things out (rightly or wrongly). By no means is this intended to be instructional, more demonstrative of the effects of your actions along the way, and how quickly you can change the appearance of your pattern. The importance of hinging the part, as explained and demonstrated well by others in many places, I have found to be completely correct. I feel it alone tells you as much, if not more than the blue - to the extent you can almost predict what the blue will look like when you take it off the plate. It also, of course, tells you when you might want to steer clear of areas of blue because they may not actually be relevant.

    One thing I do at times, is draw up "exclusion zones" with a texta around areas with no, or very light blueing on them. You can of course just avoid them anyway, but occasionally I like to have an easier to see boundary that doesn't get lost in all the little chips on the surface...

    exclusion.jpg

    So after a bit of time, it looked like this, with still some material to remove to get the ends down.

    working away.jpg

    So, I'm working away, more or less following Richard Kings "scrape individual marks" mantra. Blue it up, scrape the blue areas at 45 degrees to the axis of the ways in as neat a row of scrapes as I can manage, put another row above that and repeat until scraping air. Come back at 90 degrees to the previous direction, working the same areas, trying to aim to hit most of the darker spots remaining, but not worrying overly much about it. Clean, stone, blue, repeat. Eventually, I get to this - probably helped by the fact I realised my scraper had slowly acquired far too much negative rake. Fixing that up made for some far bigger chips than I'd previously been making, with less effort. Keeping your scraper sharp is a big help, requires far less effort to scrape. I was finding about 2 cycles was getting me to the point where I wanted to touch it up a bit, but I suspect I may be one of the only people in history who scrapes too deep as a beginner. But it should work out just fine when I get to the stage of really refining the surface whenever it all comes apart again for the real deal.

    dont panic.jpg

    Now, at this point, I would have previously been thinking "oh no - where'd all my points go!". (I think scrapers should come with "Don't Panic" inscribed on the shaft in large friendly letters.) Having played with this a little already earlier in the day, I took that photo, then I took a single pass, just one direction, and only took the highest spots off - the silvery bits with the dark blue ring around them. As shown below - you can see how many I DIDN'T decapitate peeking out of the chips.

    single high spot pass.jpg

    That gave me the result below. Not by any means perfect, but a noticeable improvement in count, and repeating the process two or three more times would bring it back no worries - providing you're up to that point, of course. Just an indicator of how, as long as you're keeping a consistent pattern to your scraping, and covering the general area you need to, it seems like everything stays pretty well under control (to me, anyway).

    after single high spot pass.jpg

    However, I wasn't actually at that point in my opinion, as I still had no noticeable contact points out on the end. Normally, though, I would have chased the high points around for a while until I had heaps of contact again, then got back into taking broader areas down. I decided it was too early for that. However, it was getting closer, so I started to change my approach somewhat. Still doing the up-and-back routine, but dropped the length of my scrapes to less than half, to really try and start bringing the checkerboard back out. Pattern shown below, just don't laugh at my drunken weaving.

    tightening up.jpg

    After a single cycle, this is what I had:

    after one tighten up.jpg

    I drew in a couple of exclusion zones, and hit the rest again for a single cycle with the same tight pattern, with the following result. As you can see, although the PPI is low, they are spread more or less evenly across the whole surface (with the exception of the tail, and a touch up the front), and in a semi regular pattern. Doing the same another time or two would likely really bring out a regular pattern

    after two tight passes.jpg

    At this point though, keeping in mind my goal at this point in time, I didn't see any value in going much further across the whole thing - remember, I'll be coming back to this at some stage in the (distant?) future, and some wear may well occur during that time. Or the casting might move a bit. The last little bit of the tail isn't in contact, but 98% of the time for what I do, that's hanging out in thin air anyway. The front corner does at least have some minimal contact. So I went through and did a single pass decapitating the high spots, same as earlier.

    high spot chip 2.jpg

    Looks kind of sparse, but at this point I was still using the blue from yesterday on the plate, I hadn't topped it up at all. This is part of the reason it looked so bad earlier when I mentioned "Don't Panic". I feel like the amount of blue I had on there would be well suited to chasing 40PPI or higher. So for a fairer comparison to earlier, I topped up the blue on the surface plate a little. Still probably less than I had at the start of the day, but plenty.

    increase blue.jpg

    Now that looks quite a bit better. Is it too much blue? Maybe a tiny bit, but I still have shiny high spots surrounded by dark blue, just there's a bit more definition around them, a bit easier to tell points that are generally high areas, as opposed to 'point' highs, and you can also see a few lower high points that would come into contact with a bit more refinement. Speaking of which, that's where I've decided to leave it. Again, with my personal current target in mind, this is good enough for the time being. It IS actually 20PPI even in the areas that look a bit sparse, and a couple of passes picking the highest spots off and splitting some of the bigger ones would make it far better, and when the time comes I'll spend some time taking it a bit further down to get those last little spots on the ends in contact, and refine it far more.

    For the moment though, it will work much better than it did, and it's suitable enough for a crude master. If I were actually fitting the saddle to it 100% at the moment, I would want it better - but this is not that. I may, once I've checked everything out in the future, need to lower one or other of the flat ways on the top of the saddle by some amount or other. This is simply getting it 'generally flat' which would need to happen anyway. So, blue on the cross slide, and check it out on the saddle.

    saddle 1.jpg

    Thats..... even worse than I expected. I see the hump on the right, that I fully expected from the shiny spots on the cross slide, but the left? Hang on, there is the faintest line of blue on this scratch that runs almost full length of the ways. Wipe clean, and some aggressive stoning of the scratch, and try again:

    saddle 2.jpg

    That looks more logical. NOW, here is where the trick comes in. Remember I mentioned the importance of hinging, and it telling you where to leave the blue alone? This is the same kind of deal, but viewed in a different way. The blue closest to the front, I'm going to leave well alone for the minute (ignore the scrape marks there, I may or may not have been briefly testing the sharpness of my scraper when I first shaped it ). The blue in the middle on the right hand side is correct. When I sit the flat side of my short dovetail prism on it, the prism spins on its centre, right in the middle of that blue. Same on the left, although not as definitively central, but generally still pivots around the middle. Also not visible in the shot, is a tiny band of blue on the very back edge of the left hand way. It may have been scraped off the cross slide, or it may be high.

    But with regards the blue at the front - this is where the 'trick' of scraping starts to come in to catch the unwary beginner. It would seem rather logical, if you hadn't done some serious reading, to just scrape all that blue off, and see whats what. However, imagine this. If the way is completely flat, and you put a feeler gauge near the back where that blue is, the cross slide is sitting on the front of the way, and the feeler gauge. These are our two blue points. Scraping both at this point might well be akin to taking the feeler guage out, and milling the thickness of the feeler guage of the front of the ways. If I had the saddle up on the surface plate, and I knew the saddle to bed ways were correct, I could meaure the saddle to cross slide ways with an indicator, and know for sure which is correct. This option I don't have at the moment. I have, however, ascertained by spinning the flat side of my dovetail prism that I'm pretty sure the blued portion towards the rear of both saddle-cross slide ways is a high, and thus will need to be removed regardless of whatever else needs to happen to correct any other misalignments.

    More thinking shall be applied. I'd like to get a basic level of 'flat' into it before reassembly, or it'll just chew a hollow in the cross slide over time. But given the lathe is against a wall, and I don't fancy doing left handed backwards scraping, this may not happen... Depends, I need to spend some more time assessing it to figure out what I may or may not be able to get away with without causing too big a headache later.

  9. #9
    Join Date
    Aug 2010
    Location
    Toorloo Arm, VIC
    Age
    39
    Posts
    1,270

    Default

    Quote Originally Posted by tony_A View Post
    Thanks for the detailed reply J&H. My learning process sounds similar to yours, reading on the internet and you tube. Encouraging to see that acceptable (to me) results can be obtained.
    Did you get your scraper from an OZ supplier or have to import one?
    I don't have a surface grinder and am expecting the gibs to need replacing or shimming (as Stephan did) and that seems to me to be the biggest/ trickiest part of the whole job.

    Your thoughts for a stand sound to be along similar lines to mine.
    I got my scraper from https://www.greenwood-tools.co.uk/sh...er-256840.html

    Think it cost about $120 posted for the 20mm wide one. Cheapest I could find, although Marcus might have some better leads.

    As far as the gib goes - I don't have a surface grinder either. I DO have a Bridgeport style mill though. Put chunk of flat cast iron in vice, with enough sticking up for the job required, and long enough side to side to give you the gib plus 50mm. Mill the front and back faces with an endmill, to the height required for the whole piece. Swivel vice on table, and use a dial indicator one of the two faces you just machined to set it at the right angle. In my case, from measuring the gib, and adding to it the play (checked as shown in Stefans video), I determined I needed it to gain 2.45mm over 145mm of travel. Machine back face at that angle, until thin end is the thickness of the old gib at the last point where it's still inside the compound. Tapered part done. Nod head down by (in my case) 28 degrees, install slitting saw. Lop top of gib off, to create appropriate angle. Raise knee by desired gib height plus thickness of slitting saw, lop gib off rest of stock. Scrape. Don't drop it. It bends. Don't drop it a second time. It's very annoying when you just about had it a perfect fit. Don't forget to retram head and vice.

    Is this the correct, textbook way? Probably not, but it gave me a usable result. Mine didn't warp all over the place after machining this way, but I did use an 30 year old car flywheel, so it's probably pretty stable anyway. Durabar or flocast or whatever may well behave completely different, never used them.

  10. #10
    Join Date
    Sep 2010
    Location
    Port Sorell, Tasmania
    Posts
    74

    Default

    Thanks again J&H, I haven't got any excuses not to give it a go. Wont happen this year though and
    if I had a reason to go to Brisbane I would be happy to pay for a day or two of tuition from Marcus.

  11. #11
    Join Date
    Jul 2017
    Location
    Burleigh heads QLD
    Age
    29
    Posts
    114

    Default

    Forget about the hinging for a moment you can happily setup a mag base on the saddle and bring the tip into contact on each corner of the cross slide to to check for rock

    generally if your cross slide is a 1:1 or greater like most modern engine lathes you can scrape the cross slide as a master first but in an ideal world before you scrape the saddle cross slide ways(assuming the Bed is ok/has been ground) you Bed the saddle in first, if the saddle has warped or Worn in the usual fashion it wont be supported correctly. its also relatively important to relieve the middle 30% of the saddle ways on the flat and V way. as the saddle is restrained to a single axis you cannot hinge it but after relieving the center you only have to worry about a twist which can be checked using an mag base on the bed with your indicator checking each corner for rock.

    once the saddle is scraped you must also worry about the weight of the apron, I know this is long winded but consider this you're currently scraping the cross slide ways on the saddle in its ideal state. the weight of the apron is inducing its affect in the same manner it will in use but when you remove the saddle to set it up to scrape the dovetails the same load wont be present. it will be minor but it will effect your print and your measurement of the dovetails later down the track.

    for anyone reading this later down the track a good example of someone in the scraping chat who worked on a similar size lathe, he scraped the cross slide and its mating ways but only on the flat as he did not want to tackle the dovetails yet as at the time I dont think he had a dovetail straight edge scraped in. in doing so he induced a similar error as above but the big problem he unknowingly caused was that the gib which was now maxed out in its travel was Bent. when in place a bent gib is forced somewhat straight however when removed it springs into its bent state. he epoxied a brass shim to the not working side of the gib which common but as he did not first straighten the gib it was now forcibly restrained in its bent state so although the flats on both slides where perfect he actually induced a massive error in forcing the gib convex.


    RE the cross slide if you have a modern style lathe like left you can scraped your cross slide as a master but ideally you do so in tandem with a dovetail straight edge, to add this this you scrape Both to true flat however if you have the older short slide style like the right its important to relieve the middle 30% of the cross slide in both the flats, fixed dovetail and gib. if you scrape for perfect flat on a short slide then even tenths of wear at the far ends will have it rocking on the flats and the dovetail. by relieving the center it will be always sit flat even as it wears.

    118254643_645204839456602_533979952911406728_n.png


    funnily enough my partner and I where supposed to be heading to tassy earlier this year and I had plans to bring my Biax, hand scrapers etc but the flights were cancelled due to Covid. There's only one Tasmanian on the scraping chat who I was going to meetup with to do a bit of scraping and I still may when restrictions lift. RE the scraper just make your own, my favorite rougher is the mushroom handle one that Donnie made me. only important thing is to use a facemill to cut the seat so that the radius of a scraper blade will sit it tight without rocking (see pic below and video)

    https://www.youtube.com/watch?v=9HkVGMi5SsY

    blade sharpening.pngCRJN0968.JPGIMG_9172.jpgIMG_9229.jpgIMG_9230.jpgIMG_9232.jpg

  12. #12
    Join Date
    Jul 2017
    Location
    Burleigh heads QLD
    Age
    29
    Posts
    114

    Default

    also RE trying to work with a Micrometer to measure dovetails, The measurements in scraping are all comparative nothing is an absolute measurement so you don't necessarily need one for every slide. Basically 3+2 point indicator carriages/sleds are your friends and will allow you to measure alignment in almost any slide regardless of Box way vs dovetail, interval vs external etc. the two pictures to the left is a my Go to carriage for box ways and flat bed lathes, and the picture on the right is from the connelly book demonstrating how crude they can be.

    if you go the connelly style they can be dodgy weldments with ball bearings welded to each pad and bearing races to take the average or just rounded feet. The problem with most as pictured to the right is that they don't do a very good job of taking the average of a scraped surface and as such small errors can be introduced. if possible a 3+2 carriage should be ground or scraped and as such it will give you much less erratic readings. I will also that for scraping you should get yourself 2 micron increment dial indicator or a imperial tenths indicator. I personally work in imperial but as mentioned previously all scraping measurements are comparative so either will work well. I personally find that 1 micron increment indicators will drive you up the wall for most scraping but one of the guys in the scraping chat who does all the precision granite orders works excursively in micron to flex on all of us.


    118765386_1678910432281028_7596098310794065762_n.jpg118774324_744624319712949_7004557488325701888_n.jpg3+2 .PNG

    sorry for all the rambling by the way, I realize my posts can often be long winded and confusing at times but hopefully you gain something from it.

  13. #13
    Join Date
    Nov 2017
    Location
    Geelong, Australia
    Age
    57
    Posts
    2,651

    Default

    Quote Originally Posted by marcuschrist View Post
    ....
    sorry for all the rambling by the way, I realize my posts can often be long winded and confusing at times but hopefully you gain something from it.
    I'm just chirping from the sideline as I have nothing scraping-wise to add - but no apology needed Marcus.
    Its great to have such detailed input, so from my perspective ramble away!
    I'm definitely finding your posts very useful. I've only done a little bit of scraping but I know there's more in my future, so being able to follow along with someone else's project while someone experienced brain dumps relevant info is awesome

    Steve

  14. #14
    Join Date
    Jul 2017
    Location
    Burleigh heads QLD
    Age
    29
    Posts
    114

    Default

    Quote Originally Posted by OxxAndBert View Post
    I'm just chirping from the sideline as I have nothing scraping-wise to add - but no apology needed Marcus.
    Its great to have such detailed input, so from my perspective ramble away!
    I'm definitely finding your posts very useful. I've only done a little bit of scraping but I know there's more in my future, so being able to follow along with someone else's project while someone experienced brain dumps relevant info is awesome

    Steve
    I really wish we had someone interested in film and editing to follow some of the rebuilds in the group, some of the guys have come up with awesome techniques not just to measure alignment but to measure Sag and compensate accordingly, Dan the guy that organizes the Granite group buys has been doing some wild setups on his surface plate for his grinder rebuild, in some cases he will scrape for flat and then position multiple elements of a the grinder as they are supported its usual running state to exactly quantify the nature of sag and then scrape accordingly. he also has no way of controlling the temperature in his shop but checks the temperature and will only do finish scraping at approx 21 degrees. I personally think its bordering on insanity and its several levels of overkill but he's doing it more as an exercise in learning how things behave at a micron level.

    One thing I really wish I had learned earlier was the counter intuitive nature of cast iron distortion from Peening and how huge parts can be straightened. Common example is Mill tables, the bow in most mill tables is actually a result of over tightening of T nuts, using under sized T nuts or the dreaded quick action rhombus T nuts. The peening effect in the area in red exerts a force outwards inducing a bow/doming in the table. Impact damage in the areas in blue also contribute. With this in mind if you skim the under side of the T slot and have the top ground the vast majority of bow will relax out of a table. this can be combined or replaced with Violently peening the underside of the Mill table to provide a counter acting force. a Needle scaler with tips ground bulbous or hammer hammer achieves this well.

    Good example of this is a recent job for a close friend, Pic on the left is the cross slide off a CNC he just bought mid rough scraping. he bought it with a tooling plate mounted for gang tooling but at some point in its life they had a turret mounted on one side so the top face copped an insane beating from years of Tools chips etc pounding into it. the degree peening from all abuse actually caused the entire cross slide to bow so the slides where only contacting at the far extremes which causes four pads of extreme wear and gauling. by rough scraping past the abuse on the top face the entire slide actually partially straightened itself out which reduced the amount I had to take out of the slides.

    With this knowledge in mind you can use peening or the relief of peened/damaged areas to make "course corrections" but without it you may end up scraping a slide first and then later address another area only to have the previously generated flat surface develop into a convex one from you unknowingly releasing tension in a surface. Same goes for scraping a part and the machining off the raw cast skin, the resulting change in internal stresses causes your once flat surface to distort. Hopefully this information helps someone from making the same silly mistakes I'd done many times until the penny dropped.

    118255485_365929801079593_994157967863501998_n.jpg4-38.jpg

  15. #15
    Join Date
    Aug 2010
    Location
    Toorloo Arm, VIC
    Age
    39
    Posts
    1,270

    Default

    Quote Originally Posted by OxxAndBert View Post
    I'm definitely finding your posts very useful. I've only done a little bit of scraping but I know there's more in my future, so being able to follow along with someone else's project while someone experienced brain dumps relevant info is awesome

    Steve
    Don't get too excited, this project is rapidly coming to a point where it goes on hold for many months, pending my gathering tools and finding motivation to build the stand (and managing to prise my wallet open for the steel to do it.)

    No pictures, I only took one today and it's blurry anyway. Been dealing with a headache all day, so just been doing rather than documenting. Current status is that the saddle-cross slide ways are now flat, although low on point count. Struggling with everything today, having difficulty sharpening the scraper, and getting into arguments with the blue not wanting to be the right thickness. Polar opposite of yesterday. However, I re-read Richard Kings 4 rules last night after mentioning them, and realised I hadn't really been spacing my scrapes quite far enough apart, so focused on that today, and seems to have gone well. Gave up on trying to get the blue to co-operate just before (I'd had enough anyway today), so I cleaned the faces, smeared some way oil on, and floated the cross slide. Silky smooth - until it displaces some oil, then it gets a bit sticky. Probably the low PPI on the saddle? Lifting it up off the saddle... takes great effort, but once you finally crack the oil film, and put it back down, glides beautifully again, until it displaces some of the oil. Meh. Tomorrows job is to bring the point count up on the saddle a bit, and then throw it back together - I've achieved what I wanted to at this time.

    Marcus, no problem with your ramblings here. I'm reading them, and filing them away for the stage when I actually do the job properly. In the meantime, I need to collect/make a number of tools. A sled is one, I was just too lazy to build one for the compound. A dovetail prism long enough for the cross slide is another, along with maybe a straightedge or two. An large assortment of indicator bases and arms is high on the list - everything I've tried to do recently (not just scraping) has often been an exercise in frustration getting the indicator set up. A 2 micron indicator? Not real convinced on that one, but I'll consider it - I have enough difficulty working to 0.01mm. The big thing I'm lacking though, is a surface plate bigger than a postage stamp (when funds allow). Even the cross slide slightly overhangs, and while it's worked well enough this time, for this goal, its very far from ideal. Thinking at this stage of 600x600 or 600x900, any idea what your guys were paying for those sizes?

    On making your own scraper - where do you get your carbide from? I couldn't find Sandvik blanks for anything even close to a reasonable price, it wasn't much more expensive to buy the whole scraper, with the advantage that I was starting from a known point, and any issues I ran into were me, not the scraper or carbide. I wouldn't mind making one or two different ones though.

    Just looking back through your posts trying to pick out the other things I wanted to respond to, and realised I should have read your suggestion for checking rock less cross eyed. I went on a whole different tangent based on what I THOUGHT you'd said, gave up, and just whacked the top of it in various places lightly with a plastic hammer and figured out what was rocking and hollow that way. Scrape relevant blue a bit and repeat. Agricultural, but got there. I'm happy I've taken minimum viable off to get to flat, leaving me some room to move when I come back to this at whatever time.

    On the topic of mill T-slots - something I have previously read. Of course, I've discovered now I have an OML vice which is very flat, that my mill (Supermax bridgeport copy) table has a slight bow in the front to back direction (centred on the middle slot), and the clamping rails I made up slightly change the tram of the vice as I tighten them down. What is weird to me about that though, is that it's the same all the way from the drain pocket for coolant at one end right to the drain pocket at the other end. Would the T-slot deformation normally bow the table front to back evenly all the way to the ends, or is it more likely in this case that the casting itself has relaxed over time? I sort of had the idea for some reason that damage to the T-slots would bow the table in the long axis, left to right (which it may well be, haven't looked, can't do much about it at this point)...

Page 1 of 3 123 LastLast

Similar Threads

  1. Micro Petrol Air Compressor "Project Loan Out"
    By gazza2009au in forum METALWORK PROJECTS
    Replies: 58
    Last Post: 28th Sep 2019, 09:20 PM
  2. First scraping project
    By OxxAndBert in forum METALWORK GENERAL
    Replies: 10
    Last Post: 8th Jan 2018, 02:30 PM
  3. Hand Scraping a Hand Plane to .0001" Accuracy
    By Ulma Doctor in forum METALWORK GENERAL
    Replies: 4
    Last Post: 28th Feb 2015, 06:38 PM
  4. New Scraping Project
    By scottyd in forum METALWORK GENERAL
    Replies: 11
    Last Post: 21st Oct 2014, 11:29 PM
  5. Swiss Scraping or scraping from the shoulder
    By Ueee in forum METALWORK GENERAL
    Replies: 3
    Last Post: 30th Jul 2012, 01:31 AM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •