Scraping and Struggling.

[Anorak Bob]

You blokes are going to come down of me like a ton of bricks....

I am trying to create a flat surface on a piece of cast iron measuring 130 x 80mm. It is clamped to the mill table because there are more milling operations involved and more importantly, there is 0.001 " fall in the table on the X axis over the 130mm length. I did not check before I started cutting.:doh:
This is not a job for the surface plate that have. It is cast iron, heavy and not something that lends itself to delicate positioning on the workpiece.

I started off using a lathe faceplate, then a Moore and Wright angle plate and finally a surface ground magnetic chuck as a portable surface plate. I can see the heads shaking.:o

I've lost the lid off the bearing blue tube and as a result there is blue from A to B.

Traversing an indicator along the X axis reveals nary a flutter of needle movement but there is 0.0005" fall towards the front edge in the Y. I have to keep going.

A couple of real basic questions. Is bearing blue the stuff I should be using and even when very thinly spread on my "surface plate", it smears somewhat on the workpiece when the plate is rubbed on the surface, is this what usually happens?

Bob the tyro.

[Bryan]

Bob, some bearing blues work better than others. I have some of that powerplus and another which I like better. I forget exactly why now. I will check the brand tomorrow (it's late here). Smearing will happen on a ground surface. Spotting masters always have a little texture, whether from scraping on an iron tool, or the the natural texture of the stone with granite.

[Anorak Bob]

Thank you Bryan.

[Stustoys]

Hi BT,
I'm confused. Why cant you take the part the surface plate?(I assume because you don't want to reset?)
The problem with what you are doing(if I understand you) is you will end up with a flat face but it wont be parallel to the bottom face(or are you going to come back and fix that up later?)
I found spotting off surface ground surfaces did smear easily. Are you getting consistent spotting with the magnetic chuck?

Certainly looks nice.

Stuart

[Anorak Bob]

Hello Stu,

I didn't want to remove the workpiece from the table because I have to bevel the long edges and then bore a center hole and cut a tee slot. Leaving it in place assures parallelism. Let's say my adjustment of the table on the X axis is a shabby as it is on the Y, the chances of repositioning the workpiece precisely are less than slim, if I removed it for more convenient access.

The bottom being out of whack is of no account. It will be machined true when I turn the dovetail spigot. The supposedly flat scrapped surface will be bolted to the lathe faceplate.

Given the slipperyness of the surface ground mag chuck, do you consider it better to use the flycut surface of the angle plate as my surface plate?

BT

[Michael G]

Bob, the amount that you are taking off by scraping is in the order of a thou or two. On that basis doing the machining operations and then removing the part to scrape should not be an issue.
Connelly sometimes advocates using one part of a machine as a master for another part that is being scraped in. On that basis, the best thing to use may be the mating part.

Michael

[Bryan]

Connelly sometimes advocates using one part of a machine as a master for another part that is being scraped in. On that basis, the best thing to use may be the mating part.

Michael

Yes, but after scraping the first part to a known reference I think.

Bob, the better bearing blue I have is Holt's. But I think that's the least of your worries. You could use your faceplate or angle plate, after first scraping them - in theory to reference standard - using your surface plate. But if your surface plate is cast iron and second hand it can't really be trusted unless you've checked it against a known reference. The other problem is what are you running your indicator on? Obviously the mill table. But how do you know how flat that is? I don't mean to be discouraging, but unless you start with something known to be flat (enough), you could waste a lot of time.

I don't really understand your problem with the mill setup but I'll leave that for others with more milling experience.

[.RC.]

B
Connelly sometimes advocates using one part of a machine as a master for another part that is being scraped in. On that basis, the best thing to use may be the mating part.

Michael

But that is only when the other part has been scraped first.. In fact you always use the mating part as a master when doing final work, like fitting a tailstock or saddle to the ways..

[Anorak Bob]

Bryan, Stu, Michael and Richard,

Thank you for the advice. Given that nothing I can use as a master can be guaranteed as flat I'm sorely tempted to mount up the 28mm slot drill in the horizontal spindle and remove yesterday's handiwork. I can't use the mating base of the dividing head as a master because of the projecting bevelled edges. Bit of a catch 22.

The universal table requires realignment if it is moved from zero on any axis. A fixed table would be a handy thing to have.

BT

[Michael G]

How about a trip to a glazier and getting a small piece of the thickest plate glass they have? That stuff is usually pretty flat and as a first approximation will give you a reference.

Michael

[Abratool]

Bob
I just happened to look up the forum & discover your current problem
I agree with Michael, its going to be difficult to scrape a surface flat to a ground surface however its a start. I just checked my hand scraped surface plates one measures 7" by 9" & the other 9"by 12" the smaller one would be ideal for your job.
I only wish you were a bit closer as I would drop over this afternoon.But its about a 5 day drive, I remember it last year, crossing the Nullabor.
I use the Power Plus bearing blue & its ok providing you only use a little on the tip of the finger, too much & it just gets confusing & a mess.
There must be an old Tooly in your area that could loan you a reference scraped surface to use. It would not take long to bring a surface that size to flatness.
The other thought I had, was I have a roll of half thou .0005" thick paper (very thin) about 1/2" wide left over from my Tooling days. After you got the surface flat by scraping, the underside could be shimmed up with the half thou paper to set the horizontal surface clocked in at zero, for subsequent machining of your dovetail.
I am not sure if this all will help but if it is of assistance I can post over some of the half thou paper to you.
Once you start chasing these half thous ,the whole game gets interesting, to say the least.
regards
Bruce:)
ps Now to get back to a wooden stair case, I have been building for the last 35 hours.
Looking forward to its completion, & getting back to metal stuff, & dial gauges.

[Stustoys]

It will be machined true when I turn the dovetail spigot.
I'd forgotten about that part :D. I should have gone back the the drawing.



But that is only when the other part has been scraped first.
If BT was making a way Yes. But I dont think the parts are going to move in use. So if the parts are going to be clamped together in the same orientation Do they need to be flat to a master?

BT,
Why not machine the Tee slot(+the relief/rebat what ever you want to call it) and bevelled edges now and use your dividing head as a master?


Stuart(likely missing something as normal)


p.s. you might want to spot your dividing head to the other table you have to be sure they fit well.

[RayG]

Hi BT,

I think the 2-4-6 blocks we did on the scraping course would do the trick as a reference surface, I think they should be about the right size for your job. Interested?

Regards
Ray

[Pete F]

Sorry Bob, I'm completely confused, normal I know :D

Why can't you finish the milling operations and then scrape that surface from the surface plate? You can still confirm parallelism by indicating off the surface plate. I'm sure I've missed something here, but can't quite see what it is.

Pete

[Anorak Bob]

What's causing me the most angst are the two bevels. They are critical because they provide the alignment of the head on the base. If I was to accurately machine them prior to scraping the flat face, any removal of material from that flat face is going to result in an increase in width between the bevels. The head will no longer fit. If I was to increase that width by say 0.0005" and then rely on scraping the bevel to ensure a correct fit I am all of a sudden out of my depth because I would have to scrape an edge about 10mm wide and 130mm long at 60 degress, uniformly.

Stu is correct. The dividing head is clamped on this base in one fixed position.

BT.

[Pete F]

If I was to increase that width by say 0.0005" and then rely on scraping the bevel to ensure a correct fit I am all of a sudden out of my depth because I would have to scrape an edge about 10mm wide and 130mm long at 60 degress, uniformly.



Bob, wouldn't you then simply be scraping to fit, using the dividing head as the master?

[Stustoys]

Hi BT,
The way I'm seeing it now, you dont have any alignment issues until you machine the face with the spigot on it. (other than the bevels to the face)

Give it a go, if you cant get all three faces as good as you want you wont lose much height machining enough off to start over. I think I'd start by machining the bevels a little deep to ensure the flat faces meant, then scrape the flat face until the bevels meet(?). Sure sounds easy enough ;)

On the bevels you could over scrape the middle 80mm so its touching on the ends.(the same way they have removed the center section of the dividing head and the other table.) You can also machine the middle of the flat face of course. That will make your life a little easier.

Stuart

[Abratool]

Bob
I would think about this job overnight.
There have been a few suggestions, however, if these are digested into the brain overnight, a simple solution often presents itself.
Thats been my experience :rolleyes:
The job is too good to mess up !
regards
Bruce

[Greg Q]

I skipped over a few replies since I have been up for...let's see...18 hours now. The order of scraping starts off with a known, quantified surface reference. You then transfer all that flat goodness to piece one by scraping it with reference to the master. Then you can use that newly flat component to transfer the quality flat to the other machine elements.

Bob, It is apparent that I need to ad some water soluble Canode spotting ink to the package soon to be enroute. Others may differ, but I still like it best for scraping.

Next month I shall have a nifty Czechoslovakian granite surface plate available for a special forum price. Kindly advise if you'll be coveting it. I shall also be in PER sometime in Feb for tuition. I scrape for coffee.

Greg

[Stustoys]

Hi Greg,

The order of scraping starts off with a known, quantified surface reference. You then transfer all that flat goodness to piece one by scraping it with reference to the master. Then you can use that newly flat component to transfer the quality flat to the other machine elements.
But is Bob really interested in it being flat? or does he just want it to match what he has?(with any luck at all its pretty flat anyway).
He could check the other table he has(the one the dividing head normally goes on) against his master, then check his dividing head to that table But this assumes he wants to scrape both of those, if he doesn't want to scrape them ATM isn't the best he can do to match the dividing head as it is, there is no movement after all?
If at some future date he chooses to scrape the table and the dividing head he will of course have to rescrape the new base.

Stuart

[Greg Q]

Hi Greg,

But is Bob really interested in it being flat? or does he just want it to match what he has?(with any luck at all its pretty flat anyway).
He could check the other table he has(the one the dividing head normally goes on) against his master, then check his dividing head to that table But this assumes he wants to scrape both of those, if he doesn't want to scrape them ATM isn't the best he can do to match the dividing head as it is, there is no movement after all?
If at some future date he chooses to scrape the table and the dividing head he will of course have to rescrape the new base.

Stuart

You may be right Stu, I'm too shattered to honestly digest it all, although I imagine its pretty straightforward. I wonder why I can still type though? At any rate; matching mating, fixed components to each other can be cheerfully done absent any third party flat reference as you so rightly point out. I do the same when scraping woodworking tool components into a more intimate relathionthip* with each other.

*Did I listhp just then?

Greg

[Bryan]

I think Greg needth thum coffee. Or maybe thleep.

I think you're right Stuart, I retract all my waffle about references. Bob I wish I had something helpful to say about the alignment problem but it's well beyond my experience. Where's the other Phil when you need him? The un-handlebared one.

[Steamwhisperer]

I heard that (insert smiley here)

[Michael G]

Bob, it would be interesting to see the fit of the bits you currently have (I think I've picked up that you have a mating pair and want to match up a 3rd part). Given the precision involved in getting three surfaces to match up exactly, I'm leaning towards the idea that only two of them are defining the mate (and they are probably the two wedge surfaces). After all, only two planes are required to define a line.
If you can blue up one part and see what transfers over, that may reduce your work load a bit - there's no point in scraping the large surface beautifully flat if it is really only in clearance. (That is, clearance enough that when the bolt is tightened the surfaces contact with minimal deflection)
If on the other hand all three surfaces are scraped in I can only ask the question "Why have those cunning chaps put a redundant surface in there?"

Michael

[Stustoys]

Hi Guys,
About a 50/50 chance I'd say :)
I would have thought your coffee could get you well past 18 hours, maybe its losing it effect?

As I'm seeing it there isn't an alignment issue until Bob starts on the second face.

The scraping on non moving parts doesn't need to be better than about 5(?) points per inch right?

The reasons I can think of that might be why there is a third surface are
Increased rigidity,
Would the bevels themselves be wide enough to ensure alignment(in roll) on their own?
and maybe it would remove any height changes from different torques on the hold down bolts if only the bevels were used.

Of course they could all be wrong.

Stuart

[Michael G]

I'm really going to throw a spanner in the works now -
Looking at the photo in post 9, the section that Bob is trying to replicate does not look to have been scraped. Possibly a filed or fine machined finish is adequate - the parts do not not move relative to each other after all. The only requirement is that the contact is solid and the two parts locate together properly.
Have a 3rd surface in there would certainly provide a more rigid mate, but I also wonder how to achieve that in practice. The other part of Bob's mating pair is the one with the scraping marks. It suggests that that is scraped into the section, rather than the other way around. If you consider that scraping typically removes tenth's of thous, chasing the perfect fit-up could be very time intensive. The slightest bit of wear and there goes the location.
Having thought about it a bit more, these are the two ways that I'd be approaching the problem. As Bruce suggests, it's the sort of thing that really requires a bit of hands on to consider (and I'm not making that trip either...).

192833

Michael

[.RC.]

Sorry Bob, I'm completely confused, normal I know :D



Pete

Same here, we were discussing flat, now scraping angles has come into the conversation.. What does the overall part look like, including it's mating part... How big are both parts?

[Michael G]

Whoops. Just realised a mistake in my previous post. Option two won't work as drawn. You have to start too loose (clearance on the sides) so that it is contacting on the top and then slowly machine the top of the plug down until it is a good fit.

Pete & RC, my understanding is that Bob wants to create a form the same as the far right photo in post 9, to mate with the part shown second from the left. He has started with a flat surface and then wants to use that as a datum to put bevels on the sides & cut T slots. while he started talking about scraping a flat, using the mating part was suggested as a form gauge and things have gone on from there. (Hence the need for angles)

Michael

[Abratool]

Michael
Thanks for that correction, as I was concerned & thought I had overlooked something.
This job of Bobs is one I cannot get off my mind.
I even woke up at about 3 am this morning trying to figure a solution.
Like you said it would be a lot easier if we were over at his place seeing it all in front of us.
However my latest thoughts are to machine the 2 angles so that say a feeler gauge of .002" or thinner would just fit between the 2 mating flat surfaces.
Then handscrape the 2 angles so that the 2 flat surfaces come closer together & get some blue showing, at which time all could be mated accurately, on all 3 mating areas.
regards
Bruce:)

[.RC.]

while he started talking about scraping a flat, using the mating part was suggested as a form gauge and things have gone on from there. (Hence the need for angles)

Michael

He is going to have to use the matiung part as a form tool...

All three surfaces are going to have be scraped in at the same time... To me it looks fairly straight forward.. Pointless scraping the big flat area first as that is the last bit that will touch the mating part when you start scraping the sides....

I would be machining/scraping the base section first, then using that side as a datum for the machining of the top...

[Anorak Bob]

I applied blue to the swivel saddle ( Schaublin's name for the base ) and fitted the dividing head. Pretty even distribution of blue on all 3 mating surfaces. Applied blue to the head and fitted it to the base, same result. The base is not scraped, the machining marks are linear, maybe planed.

The sketch shows a way of me being able to creep up on the fit of the bevels. The thing I need to be ever conscious of is that I need to bore a hole exactly in the center between the bevels.

Bob.

[Stustoys]

Hi BT,
I'm not sure if we are helping you bail or turning the water on harder?

I know I've asked this before some where but I cant find it.
I'm wondering again along the lines of RC.
How about you
rough the bevels
machine the tee slot
bolt it to the face plate and machine the spigot
make the base
scape the base parallel
scrape the bottom face of the table flat to the base
mount base and table on mill
fly cut one side of the top flat face.
rotate the table 180 degrees fly cut the other half
set up your bevel cutter
cut one side
(now for the fiddly part)
rotate 180 degree as near to exactly as you choose to get
cut the other bevel
repeat until you are happy with the size.

The 180 degree turns mean that you will be centered on the spigot and any errors in your table alignment will be halved(across the flat face) and at opposite ends. I think this would be easier to scrape in than if you machine it in one setting and make a wedge.

My dumbest idea ever?

Stuart

[Anorak Bob]

Stu,

Your sequence of events makes sense. The problem I have is that the fitting of my base to the Schaublin head is the last operation and the most risk ridden (for me). The scraping of the three surfaces, while at the same time maintaining concentricity, is a task I find daunting.

I would prefer to meet with disaster at this early stage rather than the end.

Bob.

[Michael G]

Next thought -
(I hope you run out of problems before we run out of thoughts)
If you make up the base - even if just a piece of metal with just the spigot hole and bolt that to the mill table, you can use a DTI to get that concentric with the spindle. Truing the base up (making it parallel to the mill axis) should not be difficult as you would use the same technique as you do for aligning a vice. Using a combination of the table screws and the DTI (or an indicator) you should then be able to determine whether or not the sides are equidistant from the spigot. Any cyclic error on the table screws should be minimised as you will be turning the screw the same number of turns (although in different directions) for each measurement.

Thanks for the photos in your 11:19am post. It clarifies things a bit but also shows that the whole top surface does not have to be flat - just the bevels and a strip along side. On your base it's probably worth machining that bit that does not have to be flat down so that you can concentrate your efforts on the important bits.

Michael

[Stustoys]

Hi BT,
I'm struggling to see how you will be able to drill the hole between the bevels close enough* to the ensure alignment of the spigot.

I dont think you'd be able to wreak it while scraping, it will take you a long time to remove 1mm(?)
Still you can try it your way. If that doesnt work you'll likely only have to machine 0.5mm off to try it my way.

Stuart

*As always what is "close enough"?

[Anorak Bob]

Stu, Michael. Happy New Year.

Thanks for thinking about this when you should be doing other things. I'm having trouble typing. Peroni, 2.99 clear skin cab merlot and red Johnny don't mix real well.

I borrowed a center microscope and a beautifully ground angle plate from another forum member. Blued up the plate and placed it on the scraped surface. The results were that bad that I will remachine the surface tomorrow. A slot drill cut surface will be better than what I have managed. The centrer scope will allow me to accurately pick up an edge.

BT

[Stustoys]

Morning BT,
Happy New Year.
Hows the head? lol
You might be surprised how bad your freshly machined surface looks when spotted.

I've had another thought, How much adjustment is there in the Tee bolt clamping system? It might be good to machine the Tee slot deep so if you need a couple of goes to get the top face right you wont run out of adjustment?

Stuart

[Abratool]

Bob
I have been impressed with the number of ideas feeding into your problem job.
There was an old saying in the tooling game, that sometimes difficult jobs can be "worried through"
I think your particular machining challenge falls into that category.
Even though I have been working on a timber staircase rebuilding job for the past 50 hours or so, at my home, your problem job is in the back of the mind.
This morning there was consideration of using precision steel rollers to measure the distance between the dovetails on your job, but as mentioned there is a wealth of knowlege, being sent to you.
All the best to you & all on the forum for 2012.
regards
Bruce

[Anorak Bob]

Good Morning Fellas.:U

The huge drink of water at 3am did the trick.

I do greatly appreciate the consideration and advice given to me by you all for what should be a simple little project. The pursuit of accuracy has proven to be quite a challenge. Attempting to match the accuracy of the original fixture is ambitious. Without the appropriate equipment, I imagine unattainable. I do need to acquire a surface plate, without one I'm wasting my time.

I will remove the workpiece and tram the table and spindle for starters. I'll also send GQ a message re his Czech special.

All the Best.
BT

[Pete F]

I do need to acquire a surface plate, without one I'm wasting my time.

Ah! In my suggestions I was assuming you had one Bob. Yes you need a surface plate. In a couple of months I expect to have one for sale :wink:

Pete

[Anorak Bob]

Ah! In my suggestions I was assuming you had one Bob. Yes you need a surface plate. In a couple of months I expect to have one for sale :wink:

Pete

I just hit the buy it now button for Gregory's Czech plate. I'm starting to see light on at the tunnel's end.

Thanks Pete.

BT

[Pete F]

I'm starting to see light on at the tunnel's end.

Thanks Pete.

BT

In my case, more often than not that light turns out to be a train! :oo:

[jhovel]

Bob, the Carbatec granit plate I purchased just before the scraping course has proved invaluable to me. It's 9"x12" and not too heavy - and surprisingly accurate.
I can recommend that. It was $59.

Joe

[Anorak Bob]

Thank you Joe,

Sounds like a useful companion for the 400 x 500 plate Greg has. I will check it out.

BT

[RayG]

Hi BT,

Been thinking about the problem of scraping three surfaces to come together at the same time, here is one approach, others with more scraping experience might have a better method, first, I wouldn't use the surface plate, since you are scraping to suit an existing part, then my thought is that, that should be your reference.

Using Stuart's idea of reversing using the spigot as the center of rotation should ensure that your bevelled sides are centered on the spigot. Which leave the problem of how to get close enough to scrape, I would take it slowly gradually increasing the bevel until you get down to a thou or so clearance at the top. I hope this sketch helps.

http://www.backsaw.net/pics/Misc/ScrapingViseBase.jpg

The central idea I want to convey is that it's better for stability that the three surfaces to be very slightly concave, than to have a high spot in the middle.

I see it as an exercise in patience.
You need to get into that Zen mind set for scraping. :)

Regards
Ray

[Abratool]

Hi BT,

Been thinking about the problem of scraping three surfaces to come together at the same time, here is one approach, others with more scraping experience might have a better method, first, I wouldn't use the surface plate, since you are scraping to suit an existing part, then my thought is that, that should be your reference.

Using Stuart's idea of reversing using the spigot as the center of rotation should ensure that your bevelled sides are centered on the spigot. Which leave the problem of how to get close enough to scrape, I would take it slowly gradually increasing the bevel until you get down to a thou or so clearance at the top. I hope this sketch helps.

http://www.backsaw.net/pics/Misc/ScrapingViseBase.jpg

The central idea I want to convey is that it's better for stability that the three surfaces to be very slightly concave, than to have a high spot in the middle.

I see it as an exercise in patience.
You need to get into that Zen mind set for scraping. :)

Regards
Ray
Ray
Same thinking that I had. Please see my post on 31st Dec @9 52 am.
I think if the bevels were scraped down & say a feeler gauge was used for the last 1.5 thou or 2 thou to ensure the larger flat surface was getting close to its mate, then that would be as you have mentioned, better than trying to get a result with a surface plate.
The master or reference piece in this job is the finished mating piece that Bob already has.We know that piece is as good as one could wish for.
Once that small gap closes up as the bevel scraping proceeds, blue is applied & finish scraping achieved.
I think also on these jobs you have got to hold the tongue the right way...Zen mindset or something similiar. :rolleyes:
regards
Bruce

[Pete F]

Ray, that's pretty much what I was suggesting above, I probably just didn't communicate it very well; scrape the flat surface, then the bevels to fit.

I would expect the bevels to come in to alignment together if carefully scraped. I don't see any reason they should wildly differ as they're both being scraped at the same time. However if concerned about keeping them central to the hole for the spigot, I think I'd turn up a plug with a flange on it to a close fit in that spigot hole, and mount an indicator to the plug. Sweep the dti to the bevels to ensure they're remaining central. Maybe I've misunderstood what is being made here.

Pete

Edit: I should add that's coming from somebody with very little scraping experience, and virtually none on scraping to alignment. So take the suggestion for what it's worth. It's just how I would approach the problem as I see it.

[Anorak Bob]

I'm getting real squirmy about this. The idea of scraping two bevels evenly to ensure concentricity while at the same time maintaining both parallelism and 60 degrees is pushing me well beyond my skill level. I know there are some who think this is a piece of but the only scraping I have done is the puerile example displayed earlier in this thread.

I spent considerable time yesterday realigning the mill table. I'm thinking of remachining the top flat section and cutting a 30mm wide set down as per the original base, then cutting one bevel and measuring by means of a pair of dowels, dovetail style, cut the second bevel. I can blue the thing up and see how bad it is. I have enough height in the block to repeat this a number of times.

Once again Boys, I appreciate your thoughts.

BT

[.RC.]

I'm getting real squirmy about this. The idea of scraping two bevels evenly to ensure concentricity while at the same time maintaining both parallelism and 60 degrees is pushing me well beyond my skill level.
BT

That is why you do it first, them machine everything else from there...

I actually can not see too much difficulty with the scraping you want to do...

Sounds to me you need to go to a scraping class.... Who knows what the future will bring there... :)

[Abratool]

Bob
There are some out here to use your words "Think its a piece of "
and it is, really, yes really.A damn good exercise in concentration.
Just narrow down the parameters, as mentioned in the various suggestions, & start "worrying it off".
It will all work, just take your time, & dont rush it, or get frustrated, as I frequently say there is no bus to catch.
regards
Bruce:)
ps.......I still like the idea of machining, then scraping the 2 bevels until the flat surface eliminates the light coming through & starts to see a bit of blue. The precision rollers to measure as you proceed would assist.

[Anorak Bob]

Stuart was correct about how bad the results would be if I machined off the scraped surface.

I'll refit and carefully tram the vertical head. Then I will have another go. I've removed about 0.5 mm, no big deal. That's a brand new, never used Clarkson slot drill. Obviously not intended for nice finishing.:no:

[Abratool]

Bob
Now ,Bob........ lets get serious,..... this job is not all that hard......Really. :rolleyes:
Maybe you need another drink of water at 3 am as per one of your previous messages a bit of water at that time of the morning does help!
May I suggest re reading Raygs comment of Jan2 @ 1.44 pm. For me as I read it..... it makes a lot of sense.
Here are my considered thought out suggestions, reconsidering raygs comments.....at this time of night around 10 .45 pm in the eastern state of NSW
Forget about trying to obtain flatness with a Clarkson cutter or hand scraping the flat surface to a surface plate, or any thing else about the flat surface..... Just forget it for the time being.
Machine the bevels or the 60 degree angles, so there is a thou & a half or 2 thou of clearance between the larger flat surfaces, use rollers or toolmakers buttons (on their sides) or dowels to have a predetermined dimension to work to , so that not too much is machined off.This can be calculated by using the same rollers as measuring tools off your finished Shaublin reference which is..... the only true reference.
All the rest are superfluous ,(Not needed)
You have the good equipment, centering microscopes, indicators etc after you achieve the above to follow through to completion.
Go ! :)
,You can do it :2tsup:
regards
Bruce


Forget about trying to machine the flat surface with Clarkson cutters or trying to scrape it flat to surface plates.

[Anorak Bob]

Bob



Forget about trying to machine the flat surface with Clarkson cutters or trying to scrape it flat to surface plates.

After the failure of the horizontally mounted slot drill in creating a flat surface, I remounted the vertical head and spent considerable time tramming it. Refitted the PRESTO:roll: 1 5/8" end mill and had another go. The results displayed an even criss crossing of cuts. Ignore the set down section on the block, that's the result from using a 30mm slot drill and a faster feed.

I cut the bevel to size on one side then, progressively measuring using a pair of dowels, cut the second bevel. Sadly 0.02mm undersize. It's surprising how sloppy a fit it was with the dividing head fitted.

I refitted the end mill and very, very carefully remachined the upper flat surface, 0.01mm each cut, knocking the high spots off with a hard Arkansas stone and test fitting the head between cuts. Blued up, there is contact on all surfaces. With a 1um indicator measuring play at the end of a bar projecting 100mm beyond the spindle nose and with my bulk pressing down on the dividing head, there is 4um movement ( 0.00016" ). I don't know if the play is from a sloppy fit or from the head riding up on the bevels when twisted. I'll machine the tee slot next and find out.

Given my lack of knowledge and skill in the art of scraping, I'm happy to leave the fit as machined for the time being.

The perception will be that I ignored most of the advice proffered, but in doing so I discovered a bit more about my mill and it's capabilities.

My next challenge will be finding the center and finding out how to align the Hensoldt centering microscope to achieve the former.

BT

[RayG]

Hi BT,

Nicely done.. 4 microns, wow, that's an impressive bit of work.. :2tsup:

Regards
Ray

[Anorak Bob]

Ray,

Luck would have been responsible for that result.

BT

[Stustoys]

Hi BT,
Looking great so far. I certainly wouldnt be losing sleep over 0.004mm just yet :).

Stuart

[Michael G]

I'd certainly leave it at that for the moment - you have no relative movement between the parts so I'd suggest that scraping is optional

Michael

[Anorak Bob]

I'd certainly leave it at that for the moment - you have no relative movement between the parts so I'd suggest that scraping is optional

Michael

Well I must have got something right. I machined the tee slot this morning and clamped the head in place. With a non Schaublin W20 collet and a 1/2" test bar, a transfer punch that had less run out than anything else that came to hand, I set up a 1um indicator. At the spindle nose there was no needle movement. 90mm from the nose there was 5um TIR (0.0002"). I rotated the test bar to the mid position between the extremes of the run out and traversed the table the length of the bar. There was no discernable needle movement. I checked this a number of times with the same result. Switched to a 0.0001" DTI with the same results. I traversed the indicator along both the vertical and horizontal centerlines of the bar. Removed and reclamped the head on the base. Same results again. No bad for a 50 year old machine.

Now comes another challenge. Finding center. A thought had occurred that the centerline of the dividing head may not align with that of the factory swivel base given that the head ( and tailstock ) are scraped to fit the base. I had thought that I could use the Hensoldt to align the mill spindle with a W20 dead center collet mounted in the dividing head spindle. Sadly, there is insufficient room to install the scope.

My current thinking is to install a 10 mm bar in the dividing head and the same size bar in a collet chuck mounted in the vertical head. Using a loupe, bring them into contact , then wind the spindle over 10mm.

Any other suggestions?

BT

[Stustoys]

Hi BT.


Using a loupe, bring them into contact , then wind the spindle over 10mm.
I'd forget the loupe and use feeler gauge or shim. Using a loupe so see light between two round parts will drive you nuts I think its going to be a point contact. You could try both and see if they agree, also you could test from both sides.

Back the the shed for me.

Stuart

[Abratool]

Bob
I just got home from being away up north, fishing, catching mud crabs, & body boarding in the Yamba Surf. Also enjoying grandkids.
The first thing I looked up was how you were progressing & it was very pleasing to read your report. The blue looks to be in the right places & the machining is exceptional as the indicator reading has shown.
I do not think scraping will be needed, it is really good as it is.
Well done! :2tsup:
Bruce

[Anorak Bob]

If I set up my Mahr 1um indicator square to a 10mm rod in the dividing head and zero it I can then gently retract the stylus with the cable release allowing me to move the table up to the position of the same 10mm rod held in a collet and chuck that I trust. I can then gently release the stylus and position the spindle to zero the indicator. More accurate than a loupe. I will need to make a flat faced stylus.

What do you reckon, pissing in the wind? The setting up square will be the tricky bit.

BT

[Stustoys]

Hi BT,
I dont see how the flat faced stylus is going to make things easier. You'd need to be perfectly** square in two plains. I think you'd be better off with the normal stylus. Though I do like that you end up on center and dont have to worry about removing the offset like my idea below.

How about this, one 1um indicator on the table set against the 10mm rod in the dividing head. Another 1um indicator on the machine set against the 10mm rod in the spindle. bring the rods together until you have movement on either indicator. Repeat as many times as you like on both sides of the bar until you are happy?(if you had flats ground on the rod until they where 5mm thick that would let you be on center)

Maybe while you have it set up, turn the dividing head 180deg and see if it all comes out the same?


Stuart

[Abratool]

Bob
Been thinking about your next step & feel a pointed round tip stylus would minimise error compared to a flat type.
regards
Bruce:)

[RayG]

Hi BT,

A cylindrical stylus would minimize the error from not being on-center with the indicator.

Regards
Ray

[Anorak Bob]

My centering efforts have been thwarted. The plan to use the Mahr's cable release to gently lift then lower the stylus onto the test rod won't work. It lacks uniform repeatablity, 0 to 5 um variation. I have to work out another plan. I need to devise some way of rigidly fixing the indicator back to the immobile body of the mill. There is too much movement in the Noga support.

I pulled my Bausch and Lomb microscope apart in a futile attempt to create a 43x loupe. I can't hold the thing still enough and clamping it to something isn't easy.

More head scratching:doh:

BT

[Stustoys]

Hi BT,

Have you tried a DTI? The lower pressure helps I think.
I've found when you get that fine that the normal old bar and clamp indicator holders work better as you can make the arms very short(but my Noga is a copy I have no idea how good the real ones are).


"A cylindrical stylus would minimize the error from not being on-center with the indicator"
But how do you get it square the both plans?

Stuart

p.s. Some may say 0.0002" was pretty close ;)

[Anorak Bob]

Stu,

Thanks for the response. My most accurate DTI is a 0.0001" Mitutoyo, coarse in comparison with the Millimess. I have a stout armed Eclipse mag base I will try with tomorrow.

The Israeli Noga is big, flexible and superb but I also have a couple of 19 dollar minature pretend hydraulic versions that have proven unbeatable in a number of setups. They get in where the Noga can't. 1um, 0.000039" is nudging FA. Doesn't take much to move that needle.

I will let you know how I get on.

BT

[Michael G]

I remember being told a story once by an ex toolmaker - when he was an apprentice he spent some time in his company's metrology lab. One day the supervisor announced to him that they were going to measure a micron.
I can't remember all the things they did, but the measurement was done at lunch time so the press line was not working, and I think they even stopped people walking & driving forklifts outside the lab. The lab was of course temperature controlled and the items being measured had been in that environment for some time to get a uniform temperature. They would have worn gloves and were careful where they breathed so that air currents or the heat of their breathe did not disturb things.
I also read a story where the grand old man of Mitutoyo came to their top production area to "show the troops how it was done". He blued up a surface plate, did a couple of movements to spot the item that was being scraped and then told people to leave it for 15 minutes because the heat from sliding it on the surface plate was enough to cause some distortion in the surface (we're taking a Master Master plate here).
What I'm trying to say here is that while you may have equipment capable of resolving to micron levels, unless you are being every careful, the uncertainty surrounding the measurement is probably such that you are not really accurately and repeatably getting that. In that case the gauge that reads to 1/10th of a thou is probably adequate for the work you are doing.

Michael

[Abratool]

I remember being told a story once by an ex toolmaker - when he was an apprentice he spent some time in his company's metrology lab. One day the supervisor announced to him that they were going to measure a micron.
I can't remember all the things they did, but the measurement was done at lunch time so the press line was not working, and I think they even stopped people walking & driving forklifts outside the lab. The lab was of course temperature controlled and the items being measured had been in that environment for some time to get a uniform temperature. They would have worn gloves and were careful where they breathed so that air currents or the heat of their breathe did not disturb things.
I also read a story where the grand old man of Mitutoyo came to their top production area to "show the troops how it was done". He blued up a surface plate, did a couple of movements to spot the item that was being scraped and then told people to leave it for 15 minutes because the heat from sliding it on the surface plate was enough to cause some distortion in the surface (we're taking a Master Master plate here).
What I'm trying to say here is that while you may have equipment capable of resolving to micron levels, unless you are being every careful, the uncertainty surrounding the measurement is probably such that you are not really accurately and repeatably getting that. In that case the gauge that reads to 1/10th of a thou is probably adequate for the work you are doing.

Michael
I read your post & all I could thing of was "Lets rebuild Australia"
Bring it back, please :)
We had plenty of know how, why did we let it all go ?
regards
Bruce

[Anorak Bob]

Michael,

I am fully aware of my limitations and and those of my equipment. I'm using the Millimess because I have it. The accuracy of the machined base was simply a fluke.

Bob.

[Abratool]

Michael,

I am fully aware of my limitations and and those of my equipment. I'm using the Millimess because I have it. The accuracy of the machined base was simply a fluke.

Bob.
I think it may have been skill,& know how.:2tsup:
regards
Bruce

[Michael G]

Bob, I apologise if you thought I was slighting your abilities and methods. One of the reasons that we are involved in this hobby is to try and improve and I agree with Bruce that skill and know how may be the reason things worked for you. Occasionally I encounter students how don't believe they are capable of getting things right and they attribute everything to luck too - it's a tough job convincing them that it is possible they are capable and can "own" some success.

What I was trying to say in a long winded way is that there is too much uncertainty in measuring to a micron in a home workshop. (Because of my background I tend to use specialist terms - in this case I'm meaning the likelihood of a measurement being correct rather than a false value due to measurement error, thermal conditions, technique, vibration, etc, etc,...). If that's the only gauge that you have then that's what you have to use, but with a gauge capable of that level of precision you may end up chasing your tail because you are reading values that are a result of the environment rather than the work being measured. I've done it myself - I have a very precise spirit level (0.005mm per metre from memory) and once when I was trying to level a mill table I was having all sorts of trouble because the slightest bit of grit or imperfection on the table was enough to throw things out. Switching to a less precise level let me get to good enough without erroneous readings throwing things out.

Michael

[Dave J]

Your doing some nice work as usual Bob.:2tsup:
The way you ended up doing the top last was the way I would have done it. I replied to the other thread about the base and doing it that way, but not sure what happened to it.
As to the micron accuracy, from what I remember of your shed it all sealed off, so you would have a better chance than most of us with open sheds chasing microns.

Dave

[Anorak Bob]

Michael,

When I purchased the Millimess indicator, it was not a purchase fueled by aspiration, pretention or illusion. I bought it because it is a requirement for the servicing of the horizontal spindle in the 13. Schaublin state in their service manual that a 0.001mm indicator is to be used when adjusting the spindle bearings. A task that I have yet to undertake.

My use of the indicator has been limited to simple positioning, levelling and the measurement of runout. I use the indicator for the final accurate realignment of the universal table. Repeatability has not been an issue in any of these functions which are really no more than observations of deflection.

At the age of 59, I still have the ability to comprehend and I do understand your comments and more importantly, why you made them.

Bob.

[Bryan]

Bob, sorry if this is a bit agricultural, but I would drill a starter hole as near as I could, leaving plenty of meat. Then bore, measure, adjust, repeat. Measure from the hole edge to the part edge with a mike.

[Anorak Bob]

Bryan,

What I have yet to ascertain is whether Schaublin when scraping the dividing head and tailstocks to fit the long swivel saddle, maintained vertical centerline alignment with the saddle. The head and tailstock are a matched set so even if their common centerline is out of whack with that of the saddle, there is no ill effect.

With my proposed rotatable swivel base, vertical centerline alignment becomes important. Hence my idea of establishing the center hole position through measurement and positioning of bars held in both the vertical and horizontal spindles.As Michael points out, not a simple task if pursuing toolroom accuracy.

I only have a ball anviled 0-1'' micrometer, the base is too wide and to complicate matters I would have to measure off the bevel. And this is all if the the head is centered on the base.

Then there is the issue of the suitability of the dovetail spigot and it's fit in the base.

I might just start painting the back of the house! :no:

An overwhelmed BT

[Stustoys]

Hi BT,

Have you tried turning the dividing head 180 and seeing if its close to the same center line or are you just going to mark one end of your new table and always set it up that way?
You could also remove and replace the dividing head and see how repeatable that is.
Of course you need repeatability in your measurments first. Did you try the other base yet?

Remember you aren't trying to measure a micron, you're trying to repeatably measure as close to nothing as you choose to get :)

Once you've made the spigot you can knock out as many bases as needed until you're happy(wont the bases be easy and cheap compared to the table?). Sure there may be better ways to fix the table to the base but not without using more height.

If you really get stuck you can paint my place as well :D

Stuart

[Anorak Bob]

Hello Stu,

I've been enduring the discomfort of the heat in the shed achieving not much other than confirming Michael's comments relating to repeatability outside of the controlled toolroom environment. Using a 0.01mm indicator reveals a lateral spindle displacement on the x axis of 0.01mm. I have been indicating off the more managable tailstock spindle. I will graduate to the headstock after lunch. I have it set up with a collet and endmill that show no discernable run out when the spindle is rotated. I have to see whether I can position the indicator such that it will allow end for end readings.

BT

[RayG]

Hi BT,

I can't offer any helpful suggestions ( I like Stuart's idea of reversing though) but I can offer some encouragement, often it's patience and determination that wins through, and you obviously have both of those in spades.

For what you are doing you don't need to do absolute measurements in the micron range, making relative measurements in the micron range is a different issue altogether, and much more achievable.

Making setup stuff, like vises, parallels, squares, angle plates, etc, etc, you really do like those things to be an order of magnitude more accurate than the job requires, simply because those errors can accumulate, sometimes if you are lucky they can cancel as well.

So, hang in there, the painting will wait. :D

Regards
Ray

[Anorak Bob]

OK Boys.

I've spent the last warm hour toing and froing the dividing head on my base and the lateral centerline displacement is no more than 0.0002" or 0.005mm. The Millimess is too sensitive to be of any practical use.

With a 0.0001" dial indicator set up fairly square to the centerline of the DH spindle, the repeatability is within 0.0001" or 0.0025mm. There is 0.0001" runout in the collet / endmill setup and I'm indicating for repeatability at the midpoint of the runout. Repeatability is based an removal and replacement of the dividing head on the base.

I think I should be able to position the vertical spindle fairly accurately off the tenth indicator simply by elevating and traversing the table then carefully zeroing the indicator with the same endmill held in a chuck. The table does not move on the Y axis. Sounds easy.:roll:

The lateral displacement is what you can see in the second photo showing the yellow 0.01mm indicator, the needle is slightly left of zero.

How's the beautiful tiny Federal tenth indicator. A generous gift from fellow forum member "C-47". Thank you Alan.

BT

[Abratool]

OK Boys.

I've spent the last warm hour toing and froing the dividing head on my base and the lateral centerline displacement is no more than 0.0002" or 0.005mm. The Millimess is too sensitive to be of any practical use.

With a 0.0001" dial indicator set up fairly square to the centerline of the DH spindle, the repeatability is within 0.0001" or 0.0025mm. There is 0.0001" runout in the collet / endmill setup and I'm indicating for repeatability at the midpoint of the runout. Repeatability is based an removal and replacement of the dividing head on the base.

I think I should be able to position the vertical spindle fairly accurately off the tenth indicator simply by elevating and traversing the table then carefully zeroing the indicator with the same endmill held in a chuck. The table does not move on the Y axis. Sounds easy.:roll:

The lateral displacement is what you can see in the second photo showing the yellow 0.01mm indicator, the needle is slightly left of zero.

How's the beautiful tiny Federal tenth indicator. A generous gift from fellow forum member "C-47". Thank you Alan.

BT
Bob
Persistance is paying off.
Looks like beautiful work being undertaken.
Indirectly you have contributed to my 12 yr old grandson who whilst visiting us has been looking over my shoulder, while I communicate on this Forum, particularly with you.
He has told me he is going back home to Grafton (8 hrs drive north from here) to set up a "neat" workshop, & wants to join the forum. He proposes to start off with wood projects. I built his father a solid workbench with vises etc fitted so he is off to a good start.
regards
Bruce
ps I only have 1/2 thou indicators,So I better come up to speed & get a 1/10th thou indicator.

[C-47]

Bob,
Glad the indicator has turned out to be of some use and keep up the good work. Alan.

[Michael G]

Go Bob!

[Stustoys]

"drill drill drill" or should that be "bore bore bore"?

How you have that all well in hand.
The dial gauge stand holding the 0.0001" dail gauge, whats going on inside the vertical post?(at the bottom of the screw slot)

Stuart

[Anorak Bob]

"drill drill drill" or should that be "bore bore bore"?

How you have that all well in hand.
The dial gauge stand holding the 0.0001" dail gauge, whats going on inside the vertical post?(at the bottom of the screw slot)

Stuart

Stu,

I just slipped up to the shed to look at the two small ridges in the screw slot. Dismantling the thing would lead to mutilation and tears. I have no idea why they are there. The shaft is screwed to the base securely, too securely for me to remove.

I'ts a really neat indicator stand, one of my earliest Ebay purchases. Eclipse made an even heavier version of that stand as did Baty and Mercer. I used to look for one but they are heavy suggesting potentially high shipping costs. Far more rigid than the Noga but lacking the flexibility.

BT

[Stustoys]

HI BT,
So it does look like a separate part then.....strange. still I'm sure they had their reasons.
I've used the Eclipse one that looks to be about the same size as yours, the fine adjustment is way better than any of the ones I have used that are out on the end near the gauge.

Stuart

[Anorak Bob]

Stu,

I have a few Starrett and Eclipse mag stands without adjusters and they a bit of a pain to use. I also have a Starrett with the snake arm and fine adjustment but it is a bit unweildy to use on a regular basis. The Noga is my go to stand. You need one.:U

BT

[Anorak Bob]

All this for one lousy hole!

I about with a few ideas for setting the dial indicator on the centerline of the horizontal spindle. Above or below the centerline and I may as well be using a wooden rule for the job.:doh:
Dawned on me after wasting an hour or so that I needed to be able to move the indicator vertically over the test bar ( 10mm slot drill ) to fairly accurately establish the centerline.The solution was a bit Heathish but worked. I set the indicator up so that it was zeroed at the midpoint of the 2 tenths test bar runout. The same test bar set up in a collet chuck indicated double the runout so I split the difference and moved the vertical spindle into position after traversing the table to determine the vertical centerline of the test bar. Started the hole with a slot drill and finished off with power feed on the slowest setting. The hole is 10.22mm in diameter, I was aiming for 10 but trying to set the boring head to cut to a finished size proved to be too large a challenge for me.:no:

The hole size doesn't matter, more of a concern was making sure the shaft of the boring bar didn't collide with the tee slot sides. I will clean up the ends of the base and then move over to the lathe.

I do apologize for this painful odyssey but if only one other person has learned something from my efforts then it has been worthwhile.

BT

[Stustoys]

Hi BT,
Nicely done.
Now onwards to the lathe.

Stuart

[Abratool]

Always interesting Bob, & one never stops learning, that is the enjoyment.
regards
Bruce