Thread: Hand Plane fettlers should read this

Cool, I have a think about how to set it up while I finish grinding the plane blades.

FYI if you wanna know more about the interferometer you can check out RayG's post here
https://www.woodworkforums.com/f65/hp...ometer-160818/

but definitely check out the youtube vid super cool piece of kit IMO.

oh and BTW I'm Josh of RayG progeny in case you were wondering.

I started fettlijg hand planes about twenty years ago, using a chunk of 20mm glass. I started off using abrasive paste as recommended in some magazines of the day. Of course the glass gets hollowed despite the best attempts at using all of the plate. This ends up making a canoe-shaped plane body for obvious reasons.

Next I glued sandpaper strips to a new chunk of glass. This fixed the canoe problem, but I found that I would often end up with a plane that was flat diagonally across the sole from upper left to lower right when looking at the sole.

Since I started scraping I have been able to eliminate the above errors. Scraping though is some ways farther down the rabbit hole, and the values get more removed from real world practicality. When I scrape things I try to get them flat to within a couple of microns...a principle in metrology that references have to be ten times the accuracy of the workpeice. At this level hand warmth will quickly cause even a very robust casting to twist with differential heat, causing varying spotting patterns. Its vexing.

Hand planes will often be called upon to take 0.001" shavings...that's been my aim since I read it somewhere in an expert article. I mostly use my good planes for trimming fits and joints, hence the fine shaving fetish. Obviously a flat sole is better than a Pringles chip; scraping can give improvements in frog fit too, bedding more solidly and helping to eliminate chatter.

A solid granite flat reference with a thin spotting dye spread is the best, most stable shop reference for flatness you can get. Well, short of an interferometer (I have seen said laser rig and am deeply impressed)

Greg

Hi Greg, have you ever tested the flexibility of a cast iron plane by doing what I've done and applied a twisting force to it while spotting or have any comment about the plausibility of the results I've shown being due to either twisting the plane or just an error in spotting?

Cheers
Michael

Hi Michael...

Yeah, its easy to torque a bench plane out of flat...you can see it in spotting against a surface plate as you report, and this is I think the source of the error in lapping. This also raises the question of flexibiliy of the plane in use. A common bench plane of the Bailey pattern is limp linguine compared to an H model casting, or the heavier Bedrock pattern. Of course Infill planes with heavy bodies, soles and a tight-fitting dense hardwood infill probably deserve the high regard in which they are held.*

I have come to suspect that my occasional frustrations with Bailey pattern planes comes from exactly that ham-fisted uneven manipulation of the tool. When you are in a quiet zen state the shavings snick out of the throat even and consistent. When tense or rushing I think its easy for an unskilled (me) user to twist the plane enough to degrade the performance of the plane and turn the whole enterprise into a struggle against the tool instead of a calm partnership.

*my smooth planes, in order of delight, go from a gunmetal infill plane, to a L-N bronze four, to a Norris A51 to a 1910 Stanley four. I realise that thick blades and cap irons add no small amount of goodness so it is not purely a review of real or perceived rigidity.

If Josh wants to make a real thing of interferometer science fun, I'd be happy to volunteer all of the above tools. It might even become a definitive study.

Greg

It is an interesting thought that the ability of the plane body to resist flexing/twisting might actually have an impact not only on lapping the sole of a plane but the performance of the plane on the wood even if its sole is perfectly flat.

I wonder how much the plane would have to twist to make a difference. my guess it would be a function of thickness of the shaving one was trying to achieve? I'm talking smoothing planes here BTW.

Does it average or compound an error?

It would also be nice to know how far down the rabbit hole of flatness there is no point of going any farther since any manipulation by the user is twisting it way past that. There is definite value knowing when it is too far is too far.

Having had some more time to ruminate on this I will observe that in any planing session the plane is probably going to get a lot of even friction heat from doing its job...enough to probably offset any uneven body heat from the user.

I am serious about the 0.001" shavings...it seems pretty easy to set a well-tuned plane to that level, and leave it there. A few more strokes is better than messing with a sweet blade position. As you know it is pretty easy to distort a casting like a plane body that much, ruining the ability to take a full-width shaving. It all boils down to the fact that an open U-shaped object is crap at resisting twisting...by definition it has comparatively little rigidity against torsional stresses.

Greg

On edit: spelling, and emphasis added for clarity

I have only really ever spent a lot of time smoothing with my bronze LN4( <3's ). I have never owned or used another smoothing plane except in school *shudders*. So I really don't know how much twist you can get out of Stanley 4 or 5 1/2 and if it was noticeable but it seems you can.

I will surface grind one of RayG's #4 or #5's flat and square(I'm sure he will not mind :P) and see about a setup that will allow for repeatable results across different planes.

Does any one have a good methodology for precisely applying a constant torque repeatedly? ATM I'm thinking a bar with a weight.

interesting pics mic-d,

now i'm curious in what way wooden planes might be affected, not as much as we might think is my initial thoughts, also , without testing i tend to think cliftons are more rigid than bailey et al , i have some infill, LN, stanley, woodies as well but my impression has always been the cliftons to be amongst the most rigid , they are much thicker castings so weight/mass might have something to do with it, on the other hand the woodies have performed well to, i wonder if its possible that a (typical) wood plane doesnt twist the same as a steel bailey..i have all that specific machinist granite as well, might be interesting to try and twist some planes


cheers
chippy

Also I might mention ... the standard "old world wisdom" is that to lap a plane, leave in the frog and clamped-down plane iron because they "distort the plane" in working configuration.

I have wondered if that was true, and I think you didn't find that with a Clifton plane(?)

But perhaps they act not to distort the plane, but to reinforce it against deformation.
The converse being - that a plane may distort (a bit?) more without the frog in.

The plane you twisted and spotted - was it fully kitted?

Thanks,
Paul

Originally Posted by pmcgee

Also I might mention ... the standard "old world wisdom" is that to lap a plane, leave in the frog and clamped-down plane iron because they "distort the plane" in working configuration.

I have wondered if that was true, and I think you didn't find that with a Clifton plane(?)

But perhaps they act not to distort the plane, but to reinforce it against deformation.
The converse being - that a plane may distort (a bit?) more without the frog in.

The plane you twisted and spotted - was it fully kitted?

Thanks,
Paul

I found out the hard way that even adjusting the tension on a frog screw will change the spotting- So it is best to leave the frog in while scraping or as Greg says, you will be vexed when you put it back and respot the plane and for lapping perhaps it contributes to rigidity. In service however, if I have had to move the frog for a wider mouth (maybe I fitted a wider blade), I don't let that bother me, although I do tension the screw back to the same marks. I also did some trials with and without the lever cap and blade assembly clamped in - didn't make a difference, so I leave the blade out when scraping.
All my scraped planes are capable of taking a sub 1thou" shaving in service. I think some of the keys are very sharp blade, wax the sole so there isn't as much effort (and possible twisting) and a relaxed grip on the plane.

Originally Posted by Brobdingnagian

I have only really ever spent a lot of time smoothing with my bronze LN4( <3's ). I have never owned or used another smoothing plane except in school *shudders*. So I really don't know how much twist you can get out of Stanley 4 or 5 1/2 and if it was noticeable but it seems you can.

I will surface grind one of RayG's #4 or #5's flat and square(I'm sure he will not mind :P) and see about a setup that will allow for repeatable results across different planes.

Does any one have a good methodology for precisely applying a constant torque repeatedly? ATM I'm thinking a bar with a weight.

Perhaps I'm telling you how to suck eggs but people have had some problems with surface grinding planes. p'raps it is to do with the heat generated, but more likely, the clamping involved predistorts the sole and it is ground flat in that state and then springs back when released. Perhaps a scraped plane would be better. Also, what is the width of the laser beam and how will you ensure that with the application of force and the movement of the sole (either twisting or accidental displacement) will not cause the laser to hit a different point on the sole? Be ok if the sole in the immediate area of the laser is flat within a micron or so, but if not the laser might drop into a hole or hit a peak and you will get an artefact.

Originally Posted by mic-d

You bloody skeptic!!

Can't help it Michael - 40-odd years of having to double-check everything one does leaves scars........

And yeah, alright, I'll admit to being born with a sceptical nature to boot.

Ok, so does this twisting, and the uneven wear that it may cause over time become the "fingerprint" of the plane? Do we unconsciously compensate for this as we are working, in the way that we hold the plane, the places we apply pressure? Does the plane then wear into us a bit like a pair of shoes? Would this twist manifest itself differently if I were to use a plane that wasn't mine, but had been in use by another woodworker for many years? Is this all rather esoteric and off topic?

All very interesting!

Cheers,
Virg.

Originally Posted by mic-d

Perhaps I'm telling you how to suck eggs but people have had some problems with surface grinding planes. p'raps it is to do with the heat generated, but more likely, the clamping involved predistorts the sole and it is ground flat in that state and then springs back when released. Perhaps a scraped plane would be better. Also, what is the width of the laser beam and how will you ensure that with the application of force and the movement of the sole (either twisting or accidental displacement) will not cause the laser to hit a different point on the sole? Be ok if the sole in the immediate area of the laser is flat within a micron or so, but if not the laser might drop into a hole or hit a peak and you will get an artefact.

Not at all, I recon there would be a difference between a surface ground sole and a scraped one. I can add another possible reason; a scraped sole would have little wax lubrication pockets which would make using it much smoother/easier (ie the same as flaking does for machine tools with oil lubrication).

The laser beam width has very little to with the measurement but rather the relative distance the laser travels after it goes through a beam splitter. I would not be trying to measuring the straightness or flatness of the sole but rather the angular deflection perpendicular to the main axis of the sole with a fixed torque. The kind of error your talking about can be minimized if the beam splitter is place one the sole of the plane and then reset before the application of torque. But then again that is only one idea, there are quite a few different setup that could work well. I will try to be as scientific as possible.

I just tried on a couple of the planes behind my shoulder as I'm typing this to see how much force it takes to introduce a visible twist; the answer a lot less than I thought. :S even the relative stout LN#2 will move visibly with moderate sideways pressure (about 0.5kg), and RayG's infill is MUCH more rigid.