Thread: Scraping a No 5 0n youtube

Flat to 0.0002" - Stanley No.5 Jackplane hand scraped... - YouTube

you can file and lap to less than .0012 by a good bit (based on my experience), and it doesn't usually take that long.

but scraping like a machinist would fine surfaces is definitely several orders more accurate. Takes forever, though.

But the guy has a good point - if you want to practice on something, an older inexpensive plane is a great starting point. The castings are void free and good quality and they are very dry and compliant.

Filing them is an enormous mess, but it can be done quickly. I know I mentioned something on here about turning a bastard file to use the edges of the gradual curve - it may have been you. that works an absolute treat even on a dulling file, but it's not the way someone who checks their hair in the mirror twice every time they wee is going to finish a sole due to the stray marks left here and there after the whole thing is already flat.

To the guys credit, he mentioned in his intro, that it’s over kill, but it was still an enjoyable watch.

Cheers Matt.

I've been scraping planes for years using much the same equipment, with the additional use of filing for initial hogging work. They look pretty and are a delightfully predictable to use.

BILD0066-small.JPG
BILD0057-small.JPG
BILD0058.jpg

Plus, with the wipe of an oily rag you can do this party rick with two scraped planes


BILD0078-small.JPG

The one issue I had with the video was the comment he made regarding the older steel type scraper; he said you'd be going back to the grinder every few minutes.... my scrapers have never seen a grinder since their initial shaping and construction. I used to use standard oilstones but now I just give the tip a couple of licks on a well-worn 1000 diamond stone, followed by a rub on each flat to knock off the micro-bur.

Cast iron scrapes beautifully with a properly sharpened tool. I'd love to be able to copy the old-fashioned scroll patterns that vintage fitters used to be able to impart to their work but I don't have the time left to me to learn that degree of muscle memory!

And then you take that ultra finely tuned plane, this is to the tolerances of the space shuttle, and use it on wood that moves 0.5mm in a single day LOL. Back n forth, day in and day out... And! leave the crap blade in it LOL.

Originally Posted by The Spin Doctor

And then you take that ultra finely tuned plane, this is to the tolerances of the space shuttle, and use it on wood that moves 0.5mm in a single day LOL. Back n forth, day in and day out...

Hi,
Let me see if I can convince you otherwise. There is really no link between wood movement and what tolerance of sole flatness you're willing to accept. Yes wood moves. Most people who would go to the trouble of scraping a plane sole are not naive to the fact. One could also argue by your logic ( and it is a common misapprehension) the machines that are used to manufacture something metal, say an internal combustion engine that undergoes thermal expansion don't need particularly high tolerances because some of those parts are going to move by 0.5mm day in day out.

Let's separate out wood movement. You make something to a particular tolerance on the particular day you assemble it (so it goes together) and you use your knowledge of wood movement to design it in a way that it can be accommodated without blowing the structure apart. Leave it at that.

Now what does a plane sole do? Apart from supporting the wood right in front of the cut, the main role is to register the blade to a desired depth within the wood. When a plane sole is flat, the toe registers the blade to a depth and as the cut proceeds and the heel comes into contact, it keeps the registration of the blade exactly the same. The shaving removed is a consistent thickness once the marks have been removed.
If any length plane's sole is concave it makes the plane useless. The blade is lifted off the wood as the cut proceeds so a smoother can't work, and the jointer will plane the ends and in towards the centre as the ends are cut down and eventually you will have planed a convex curve which is useless for a matched glue joint.

A smoother's sole can be a little bit convex and still work. There is an argument that it's an advantage to have a bit of convexity on a smoother since it allows localised areas to be worked on a little more specifically than a flat smoother which will want to travel on higher ground further afield.

The role of a jointer is to plane edges straight so that matched glue joints can, well... match; so there is no/minimal gaps or a small gap towards the middle for a sprung joint. Now a plane, any plane, that has a flat sole will still plane a concave curve that is entirely due to the blade projection. The more the blade projection the tighter the curve and the worse the concavity will be. *So as a side note, the sharper the blade the thinner the shaving and the smaller the blade projection can be, which is a good thing*. Years ago I did all the circle maths to determine what happens with blade projection and convex soles. It's boring and arcane so let me sum up:
For a #7 jointer with a flat sole and a blade projection of 0.05mm the worst the plane can do is create a curve on the edge of 2.4m boards so that the matched joint has a gap of 2mm. Acceptable. If the blade projection was 0.1mm the gap could be no more than 4mm, a 0.2mm projection - 8mm - you get the idea. Now if the sole is convex, it simply resolves into an effective increase in the blade projection. So if the #7 has a 0.05mm blade extension but is convex by just 0.15mm then the matched joints you planed would have a gap of 8mm without you doing some partial passes leaving out the middle. People do this, it's part of the art of planing apparently, but it is fudgy and fussy. I myself prefer to plane in full passes with a sharp plane and flat sole and accept the 2mm gap with a #7. By the way, if you had a flat #4 with a sharp blade and a blade extension of 0.1mm and jointed the two boards the gap in the matched joint would be 24mm! Which is why you want to joint with the longest plane you have if doing it entirely by hand work. *However if you're just taking machine marks off after using a jointer, even a #4 with a flat sole and a small blade projection is acceptable for one or two passes.

Now scraping does not really aim to bring all the sole to a geometric plane, just the toe, around the mouth and at the heel. Have a look at my first photo and you will see it is not all scraped level. Indeed scraping is the only way you can selectively relieve inconsequential areas of the sole if you think it is an advantage, as some do. The texture of the scraped surface is also more slippery than a lapped sole. And lapping can so easily lead to convex soles as many have discovered.
Sorry for the long post!

mic-d . Can you spell out some of this please? I don't understand what your saying.

Originally Posted by mic-d

The role of a jointer is to plane edges straight so that matched glue joints can, well... match; so there is no/minimal gaps or a small gap towards the middle for a sprung joint. Now a plane, any plane, that has a flat sole will still plane a concave curve that is entirely due to the blade projection.

Why does a flat plane sole with blade projection plane a curve ? Are you saying if you enter and leave planing an edge from one end to the other a plane leaves a curve?
Does this have anything to do with how the plane is held?
You know some people plane joints using the front of the sole with two hands holding the plane forward of the mouth on the body. Almost like hanging 10 on a surfboard.
Edit. Holding it on the sides not the very front end.

Originally Posted by mic-d

The more the blade projection the tighter the curve and the worse the convexity will be.

Why have we gone from having a concave curve to now having convexity?

Rob

It'd be nice to come across a second hand surface plate cheaper than this new one. The size looks fine for jointer planes.
Any ideas of the quality of these?

Untitleddf.png



And what does this flatness measurement below mean exactly?

Untitleddfg.png
Rob.

probably was micrometers or something.

Can't say anything for that one, but the chinese tool room grade granite plates that were or are still sold for cheap - and in sizes up to 12x18 here - are actually very good and very flat.

I have no idea what they're called now, but MSC industrial in the states at one point was willing to ship a 9x12 or 12x18 (the latter is pretty heavy - it's much thicker) granite B grade plate for about $70 total.

It has to be at least twice that courtesy of at least the enormous cost in shipping here in the states.

tool room B on my older certificate said +/- .0001" or something, but I don't know what measurement that's over. My plate is too dirty now from being used as a platform to check it with an unabused starrett straight edge, but I would bet they're both a good match for each other (starrett was .0002" per foot on the rules I think? I have one that I use in the shop and one that just stays in the house to check things)

Actually, this video is really excellent. A long time ago I wanted to try scraping, just didn't want to do it on everything. George Wilson offered me a break on an unused biax, but that seemed like a beginner buying a turbo katana motorcycle as a first bike, so I put it off.

I have everything other than the scraper (but bought an older NOS anderson scraper off of ebay yesterday) and the reference paste, but that's an oil paste with pigment. I'll just make that by mulling pigment into linseed oil.

eons ago I bought the 12x18x3" b grade tool room plate in anticipation of using it for something - probably sandpaper for flattening at the time, and never used it for anything. I'm going to work over 1 of my 4s vs. another that I've flattened reasonably flat and see if there is any difference in feel.

I remember george at one point making a plane that had an elephant profile in the side, and I asked him if how flat it was and he responded that it was scraped and flaked (I think that's the right order for the operations).

The comments the guy in the video made ring true - I've noticed more friction, especially on end grain - of premium planes vs. hand lapped planes. I never really thought about it, but just felt it. The function of a stanley 4 vs. a LV la jack long ago and how much less effort it was to plane the same wood with the 4 on a long end grain panel was shocking.

Originally Posted by auscab

mic-d . Can you spell out some of this please? I don't understand what your saying.



Why does a flat plane sole with blade projection plane a curve ? Are you saying if you enter and leave planing an edge from one end to the other a plane leaves a curve?
Does this have anything to do with how the plane is held?
You know some people plane joints using the front of the sole with two hands holding the plane forward of the mouth on the body. Almost like hanging 10 on a surfboard.



Why have we gone from having a concave curve to now having convexity?

Rob

The second question first, that's my slip up, it should be concave curve planed on the wood, sorry for the confusion

First question is easy to show, hard to explain in words in a short way. But no nothing to do with the way the plane is held. Let me say first that the concave planing 'error' is a maximum. A flat plane will trend towards straight because of its stability in registering to the work, a convex plane will trend towards a maximum error (without user input) because of the instability registering to the work.

Look at an ideal case first where you've got a board freshly jointed off the jointer and straight and you're taking the machine marks off. You firmly register the toe of the flat plane on the flat wood edge. It registers positively because the sole is flat. You proceed with the pass and heel passes onto the wood. You could visualise there is now a gap behind the blade where the wood has been planed away, especially if you're still pressing down firmly on the toe (even using a hang ten method) because the sole is flat (doesn't step like an electric plane or a jointer). Then you bring weight and power more evenly front and back of the plane and the heel settles onto the wood so imperceptibly. This pivot decreases the depth of the blade ever so slightly, but I did the calculations and it is such a minute change as to be practically zero so it can be dismissed - it doesn't have any impact on the flatness of the planed edge. You finish the pass and exit the wood with most of the weight on the rear. The shaving is consistent along the length of the board, the board began straight and it is still straight. With a convex plane, what is registered on the toe at the start? It's a curve, but with weight on the toe the plane is probably rocked forward, making the depth of cut a little less at the start, then proceeds to a full depth shaving through the middle of the pass and the plane rocks back towards the end of the pass as weight shifts to the rear and the cut becomes shallower again. You've started a concave curve on the board which will only get worse with more passes. If you're just taking cutter marks off, probably inconsequential, but not if you're starting with RS timber.

Look at the following greatly exaggerated image of a plane sole, it shows the gist of how you may get to a curved edge with a flat plane but it requires a lot more passes, like what you need to dress RS timber. But for a flat #7 the curve depth is only 2mm over 2400mm and really only that maximum if the board is concave to begin with. If it's straight the first case above applies, and also for a convex board, if you manage some well made partial passes in the guts of the board to get an initial straight.
planecurve.jpg

With a convex plane there is no getting around the fact that the effective blade projection is the sum of the projection right at the mouth and the depth of the convex curve of the sole. This leads to a greater error (depth of curve) on the planed wood edge by simple geometric arrangement of the toe-blade-heel. And because the sole is curved and unstable, the registration is more than likely to lead to the maximum error. While again, labouring the point, the flat sole is just the blade projection at the mouth, and the sole registers very positively so these errors are maximums and you can do much better than that.

Hope this helps clarify, it takes a lot of thought to write this economically!

Cheers
M

Rob, I don't know what the flatness measure is, maybe a typo. My granite plate from Mc Jing was flat to 1.1uM. It is only a 400mm x 300mm plate so if your plate was 3.253uM/m that would be about right and more than enough for scraping. If you're thinking about doing it I can help with tool recommendations.

I agree that this is a well presented and informative You tube video and if DW gives it a tick , that is a big wrap. I am a big hand plane guy , have been through the long learning curve and have a few with the right pedigree, stanleys and record and a Turner that i have tweaked and fettled, have every sharpening device and can even freehand to satisfactory result.I have lapped the Soles but that may have been a wasted effort .lots testing and cross referencing shows the flattest surface in my workspace is a piece of 6mm glass , testing with a feeler gauge thats goes down to.04 mm and best straight edge the beam of an old large stanley square. my flattest sole is a made in canada no4 sweetheart and it also takes the most consistent wide thinnest shaving of my planes on a good day.I can get it that the 3 critical points of the sole should be coplaner and what mic-d says makes total sense and i want to go the next step and scrape a couple of planes ,no5s.Toquita and carbartec offer a granite surfacing plate, probably the same, 300x230x37 for under $100. McJing offering is out of stock. a 6inch Groz flat metalworkers scraper that is described as ideal for the task is under $20 online, the engineers dye and a printers roller also available at modest cost, i would scrape the planes in a woodworking bench vise Is this an adequate set of kit or am i doomed to failure here would appreciate your knowledgeable opinion , thanks Ross