How to determine which surface to scrape first on a bed ways.

I have been searching these threads and YouTube to try to find out the thought processes to determine the sequences for scraping in for example a lathe bed with two inverted V's and two flats. There is plenty of information and video on beds with flat ways and lots of cross slides dovetails etc

So far I have not found anything which describes the sequence once the 'mapping out' for wear has been completed on anything other than flat bed ways.
I fully appreciate that such information may be hidden in this forum somewhere - but Ive not come across it yet and Ive been reading it for a few years now.
A little surprised as its something every project has to go through at the start.

Any thoughts, links to videos etc much appreciated

Thanks
Mat

PS - am interested to see if I get any replies - please do let me know if you would like your replies included / excluded in any future work I undertake on YT - I will try and include as many as possible ;-)

PPS - Happy New Year to everyone :-)

As Moderator I have moved this post to a new thread, the old thread was closed.

Charles

More power to the OP if he wants to hand scrape this lathe bed, but I'm still curious if a professional scraper (which I am not) would ever scrape a lathe bed with two or more v's? To me seems to be enough scraping in the rest of a lathe re-build, why not start with a known good way grinder tolerance foundation?

L7

lucky7 said:

More power to the OP if he wants to hand scrape this lathe bed, but I'm still curious if a professional scraper (which I am not) would ever scrape a lathe bed with two or more v's? To me seems to be enough scraping in the rest of a lathe re-build, why not start with a known good way grinder tolerance foundation?

L7

Click to expand...

From his original post I dont think he does want to scrape the whole bed. But it might be the only way to get what he wants out of it. The real issue is mapping and determining how much to take off and where. This has been discussed many times as the real secret to scraping. The actual mechanical act of scraping isnt as important as knowing where and where not.

In most machines there are machined sections that can be used to reference from, not sure if the OP has that on his machine or not.

Charles

lucky7 said:

More power to the OP if he wants to hand scrape this lathe bed, but I'm still curious if a professional scraper (which I am not) would ever scrape a lathe bed with two or more v's? To me seems to be enough scraping in the rest of a lathe re-build, why not start with a known good way grinder tolerance foundation?

L7

Click to expand...

Couple of points about the original question / post & how it relates to my 'project'

1. I know hand scraping was used extensively in the manufacture of larger machines (think 60ft + bed length with 20" shears) which were at best planed to +/- 0.0005"/10" or better - I'm basing this on what Ive read and discussions Ive had over the last year with Chris (former Craven trained (1960's)and then machine rebuilder). These machines were far to large to move to a grinder or too large for what I understand as ways grinders - (im still reading about those). So if hand scraping was the final 'fitting operation' for such machines which were produced to a working tolerance of 0.0002"/10" (spec' the fitters had to work to in the machine bays) - there must be a consistent approach to achieve the result. That was what I hoped to gain from my time with Chris - thats a work in progress I am hoping to continue this year.

I figured, I would ask the question on this forum to try and gather the views from a wide field. In turn I thought it may help others looking at bed ways refurb' and (like me) not in a position to 'send it out' or get someone in. I was surprised I couldn't easily find a thread on the subject. I asked about bed ways as I see it should cover all machine types not just lathes - or maybe the approach does vary - so as Lucky& queries would the pro's scrape the full bed or is it just not done these days ?

2. If the bed condition on my Holbrook is typical for its age - then getting a decent original surface to work from is going to prove a challenge:- the V's are worn and have been stoned down removing the ridges to indicate the level of wear, the tops of the Vs 'cleaned up in a similar fashion' by some former owner - already the suggestions for head stock area, below rack among others throws up some options for datum s from which to measure wear and assist mapping out.

This is not a commercial exercise on my part, I cant foresee any situation where such a lathe would be rebuilt in a shop - simply just not viable the price paid for it reflects that - it is certainly an opportunity for me to learn further skills and techniques in machine reconditioning and when Im done hopefully I will have a useful machine tool in my shop.

I gained a huge amount of support from the thread started on my shaper rebuild a while back - this lathe is the next logical step for me, presenting as it does new challenges for alignment and tighter tolerances. The project has already raised a number of issues for long Straight Edges with no suitable master ref' to work to - I hope as I progress with the lathe, other topics can be discussed as they arise.

lucky7 said:

why not start with a known good way grinder tolerance foundation?

L7

Click to expand...

Well I am trying. I guess I am going to have to donate a few 44 gallon drums, some boards, rope and a sheet so Lurk can lash together a boat to bring it over.

I think cost has a lot to do with it. Way grinding to do it properly is quite slow and thus usually expensive on top of that there can be a sizeable freight bill. When I did my milling machine table and saddle the freight bill alone was ~$5500. Although I think I went arse about face, as I freighted the 10 tonne grinder to here not the mill to the grinder.

Full disclosure, I have no bed ways under my belt (yet), but I've done a fair amount of wear measuring and leveling.

IMO, plaining OR grinding is sought not for the precision they offer (although they do), but because of the amount of material they remove while maintaining consistent geometry. When ways are manufactured and finish scraped, they are not bringing down the material any further, they are just getting any remaining high spots and providing oil retention pockets over the whole surface.

Scraping the group of surfaces of a bed way as opposed to plaining or grinding them first to bring them down past wear and damage is more than possible and many have done it, you just have to maintain the existing geometry, or bring it back in some cases.

The flat surfaces between the ways practically never wear, but can get beat up. That said they are often your best datum points for measuring the ways. Understanding how the wear happened also helps. The bed left the factory square and true, but now the carriage ways close to the head-stock have wore down and the tail-stock ways on the far right wore down.

It's right along with the old "measure twice, cut once" philosophy. Just know where the metal is high and know how much you are taking off with each pass, and print/measure multiple times before you take another pass.

RC99 said:

Well I am trying. I guess I am going to have to donate a few 44 gallon drums, some boards, rope and a sheet so Lurk can lash together a boat to bring it over.

I think cost has a lot to do with it. Way grinding to do it properly is quite slow and thus usually expensive on top of that there can be a sizeable freight bill. When I did my milling machine table and saddle the freight bill alone was ~$5500. Although I think I went arse about face, as I freighted the 10 tonne grinder to here not the mill to the grinder.

Click to expand...

Reckon hed need a beer or 3 by the time he rocked up at yours!

Hope the drawing is something like right and helps rather than confuses

Just the usual non pro 2c. Looking at the bed, the TS V and flat ways run right under the headstock, surfaces 9/10 also run the full length of the bed. Pretty sure surface 10 will show little if any wear, also surface 9 should be virgin at the HS end with little to no wear at the TS end, could mic at the TS and HS ends for a confirmation. The vertical surfaces (unmarked) between 9/10 and 4/3 might well be able to be used for a horizontal location along with 1/11.
Im thinking after the usual setting up the bed and general measuring for wear with a half decent straight edge and feelers, id start straight down with the TS ways, with those roughed in itll make a nice reference to bring in the saddle ways with an indicator riding the TS base. Or summit like that, I feel rustier than that bed

Well thats a better diagram than my crayons can do for sure !

What you describe is what I am hoping. Time will tell when I get the lathe cleaned off and start measuring.

I will try a record all the stages of the bed assessment and what worked / didn't work in terms of alignment - it may prove useful for others along the way.

Mat

If 10 is unmolested, I would scrape 9 first, then move onto six. You use a micrometer to measure parallelism from 10 to 9.

You can transfer the parallelism of 9 to 6 with a level device.

What RC says makes good sense . With 9 and 6 parallel you can get an idea where 8 is pointing, providing theres confidence in surface B, using a Kingway type effort riding 6/9 and B you can easy indicate 8, looking for wear at the TS end. With 8 in, 7 is a doddle.

While its on my mind:-
No harm leaving a hole in a surface while you use the important part of it to bring in something else, gives you some wiggle room in the event of a mistake.

Print takes care of straightness (mostly), focus on the numbers at the ends of the bed (I wasted time here )
Tenths indicators on the shelf, till the beds roughed in

Im really looking forward to this getting started, Mats definitely got the minerals for the task imo, keen to see how he attacks things. Dont take too long tho Mat, its only a month before the rain starts!

Now that the "other stuff" is out of the way, I'm delighted to see this discussion shaping up the way it ought to have from the beginning. Just dunno why that couldn't have happened to start with. Much thanks, Demon and RC.

Demon73 said:

What RC says makes good sense . With 9 and 6 parallel you can get an idea where 8 is pointing, providing theres confidence in surface B, using a Kingway type effort riding 6/9 and B you can easy indicate 8, looking wear at the TS end. With 8 in, 7 is a doddle.

While its on my mind:-
No harm leaving a hole in a surface while you use the important part of it to bring in something else, gives you some wiggle room in the event of a mistake.
View attachment 246105
Print takes care of straightness (mostly), focus on the numbers at the ends of the bed (I wasted time here )
Tenths indicators on the shelf, till the beds roughed in

Im really looking forward to this getting started, Mats definitely got the minerals for the task imo, keen to see how he attacks things. Dont take too long tho Mat, its only a month before the rain starts!

Click to expand...

Did not have much to say up to this point. But you jiggled it.

One thing became pretty clear for me in bringing in the two surfaces (6 and 9) is when you start with the initial mapping out the bed, you have to make sure you level it within the limitations of the wear and make absolutely sure you do not accidentally move it afterward. It would be very easy to chase a twist and not concentrate on the "straightness of the bed". I would most likely just scrape 9 flat then try to find a common point between 6 and 9 and map the difference along the length +-. adjust the bed so the deviation of 6 from 9 is symmetrical, i.e you have about the same in + as in -. Re-scrape 9 for straight and scrape 6 for straight, re-level and repeat till both straight and level. once you established 6 and 9...the rest can be measured from those two surfaces. If all goes well it should not take many iterations.

dee
;-D

Lurk,

Do you have the Connely book? I do not have my copy in front of me, it may have the sequence of what you are looking for all mapped out.

dcsipo said:

Did not have much to say up to this point. But you jiggled it.

One thing became pretty clear for me in bringing in the two surfaces (6 and 9) is when you start with the initial mapping out the bed, you have to make sure you level it within the limitations of the wear and make absolutely sure you do not accidentally move it afterward. It would be very easy to chase a twist and not concentrate on the "straightness of the bed". I would most likely just scrape 9 flat then try to find a common point between 6 and 9 and map the difference along the length +-. adjust the bed so the deviation of 6 from 9 is symmetrical, i.e you have about the same in + as in -. Re-scrape 9 for straight and scrape 6 for straight, re-level and repeat till both straight and level. once you established 6 and 9...the rest can be measured from those two surfaces. If all goes well it should not take many iterations.

dee
;-D

Click to expand...

I be honest Dee, I might be a bit lost here .

When you say level, do you mean that the bed be free of twist as can best be measured from unworn surfaces? If so I totally agree, but tbh would be amazed if a bed like that has any note worthy twist, its no long skinny lite weight. Checking level at each end of the bed between the shears, or with a box level from surface 11 for anything horrible would only take a mo. If it was found to be bad then its up on tray and bases and torqued straight before the fun begins . The important thing for me with surface 9 is that the ends mike the same at 9/10, id be hoping that 10 just needs stoning cos it looks a right PITA to get an there and work it.

Reminds me of a conversation that came up with my bed re working it on three points or bases. Be interesting to hear how RC etal would approach the setup for grinding.

Rightly mentioned above, the order of operations and mapping depends. Turcite & hard ways have really changed the game but that doesn’t mean Machine Tool Reconditioning by Connelly is useless.

Connelly orders the lathe map-rebuild in roughly the reverse order of the construction sequence. Keep in mind this work assumes everything is done by hand in the rebuild. Mostly the only tools available being scrapers, surface plates, some pins & bars, squares and maybe dial indicators… So it starts by fixing the compound, cross slide, saddle ways & then working down to the bed. IMO, for scraping “straight down” to fix the bed geometry would be easier with a good master → Connelly has it!

For Lurks purpose here the carriage (saddle) would be "made right" first & then used as a sled for mapping the ways if following Connellys’ order list. That work happens on a bench and is checked (square to crosslide) on a surface plate.

All the above mentioned unused bed surfaces are valid for reference planes when it gets to the point of mapping the bed with the sled.

The real truth about soft way machines & slideways is the short member of the two (saddle here) is MILES worse than the long ways. Old timey, it wasn’t unusual for just the saddle & cross slide to be refreshed & the machine put back in service.

This is gonna be a lot of work for sure...

Good luck,
Matt

Given that the HS and TS areas of the saddle flat way are probably unworn, would it make any sense to use them to align when you scrape the T/S flat way first, and then use that as the consistent reference for the rest of it?

Seems like that would keep you parallel to all the accessory items, like the rack, the slot for the taper attachment, leadscrew brackets, and maybe the carriage gib ways as well. The reverse would work also, if the headstock was on the T/S ways and they go right through under it.

Maybe there is a good reason not to do that, in which case I'd like to know it.

The bed may be better than the carriage, but the area near the chuck is usually worn down and no good for reference. I'd expect to trust the T/S ways better than the carriage ways, but they could be spotted to see what the real truth is.

JST said:

Given that the HS and TS areas of the saddle flat way are probably unworn, would it make any sense to use them to align when you scrape the T/S flat way first, and then use that as the consistent reference for the rest of it?

The bed may be better than the carriage, but the area near the chuck is usually worn down and no good for reference. I'd expect to trust the T/S ways better than the carriage ways, but they could be spotted to see what the real truth is.

Click to expand...

Valid point, that’s why crap like this is so variable.

If the TS ways are found to be the best for a longer length and you want to map the bed that way, → then (following Connelly) you would refresh the tailstock base to the best part of the bed and use that for a sled.

I have an old Regal 17 that if you don’t know how to push the tailstock before locking you’re gonna have a miserable time with shaft work though...

Good luck,
Matt

I was not thinking of that as a reference all along, but as a "height" reference for each end, to "point" the scraping by checking at each end in order to keep the scraped way level with the original factory alignment. That can then be a reference to keep the rest in line.

I'm thinking that would minimize the amount of fiddling with the auxiliary stuff and gib ways.

Maybe a good plan, maybe impractical