Thread: Setting a handsaw

Originally Posted by riverbuilder

How much set is too much set?
As far as I can remember, all the old man did was set a saw, cut across a board about an inch, and rattle the saw back and forth. Is this adequate? Is there any technical technique? Or measurement method?
What does everyone else do?

RB

Are you talking about large logging saws. Either way most people use a saw set of some description with the majority being the "Pliers" style.

Regards
Paul

I’m talking about standard, 26” long, 6 teeth per inch, panel saws, driven by arm power. I know how to set them and what to use, but the size of the kerf produced, is there a standard for it? For instance, is it 1 &1/2 times the thickness of the saw plate? Or double? It would vary a bit depending on the timber type cut for the test I would think, slightly anyway.

Originally Posted by riverbuilder

I’m talking about standard, 26” long, 6 teeth per inch, panel saws, driven by arm power. I know how to set them and what to use, but the size of the kerf produced, is there a standard for it? For instance, is it 1 &1/2 times the thickness of the saw plate? Or double? It would vary a bit depending on the timber type cut for the test I would think, slightly anyway.

RB

It varies: Green timber a lot of set, softwoods, partuclarly those that are fluffy, quite a bit and seasoned hardwoods the least of all.

I have to say I never measure the kerf thickness, but would start at half the saw plate again for hardwood and work up from there. if your 26" saw has a plate that is .036" thick, you might start at a kerf of a little over .050" and work up from there. I think that would be minimum on a taper ground blade. It is easier to increase set than reduce as there is a chance of breaking teeth. generally people aim for the narrowest kerf that is practical, as I am sure you appreciate, but you may have to go with a set that covers a range of timbers.

I am sure others will have more precise figures available.

Regards
Paul

I've read all sorts of figures ranging from 20% to 50%, and they could all be correct according to circumstances. As Paul said, green (wet) wood needs lots of set, and softwoods (more fibrous) generally need more than hardwoods at the same MC. The easiest way that I know for getting a reasonable approximation of how much set you have is to get out the feeler gauges & a good set of calipers, or micrometer if you have one. Cut as clean a kerf as you can in some hardish wood & measure its width as accurately as you can using the feeler gauge, subtract the thickness of your saw plate, and calculate the % of set from that. Backsaws have constant-thickness plates, but if you are checking a tapered saw, you'll need to measure close to the tooth line.

You could be obsessive and spend hours or days mucking about like this with a feeler gauge & a micrometer checking the width of kerfs, thickness of plates, & adjusting setting pliers and getting some precise % of set, where the saws are cutting perfectly, with no binding & minimal "rattle". OK, so you get them just so on the day & for the wood you used as a test material, then next week, after several humid days, you cut something a bit different & your saws bind, or after several very dry days, they rattle in the cut!

I mucked about a lot when I first started making backsaws & had several similar saws at one time that I could sharpen & set differently for comparisons. It was hardly a proper controlled trial, but I did find (for backsaws), that a set that gave a kerf somewhere between 15 & 20% wider than the thickness of the saw plate was the best compromise (for me) for cutting the typical hardwoods & most softwoods I would use in cabinetmaking. I aim for the maximum set I can tolerate to begin with, which gives me 2 or 3 re-sharpens before the teeth lose too much set & start binding.

The practical solution is to do what riverbuilder's old pot & most other blokes of my dad's generation did; find a sort of sweet spot for your saws by trial & error that works well enough in 99% of the situations it's used for, without having a clue what % the set is. In the days of framing with green hardwoods, the saws used for cutting the framing needed lots of set, so they'd have additional saws set more finely for cutting dry stuff like flooring.

The actual amount of set you "need" depends on personal preference & experience - a novice usually needs a bit more set because they have a tendency to force the saw, and also find it harder to keep the saw cutting straight, needing lots of correction. A finer set exacerbates the effects of both, so it's generally better to have a slightly "overset" saw when you first start using the arm-powered things. You may not get as clean a cut, but it will be much easier going....

Cheers,

I've played around with kerf measurements but mainly on band saws and chainsaws and my experience is that kerf sizes can be quite a bit wider than "blade width + 2x set"

For example the milling bandsaw at at the tree loppers yard uses a band that is 1.3mm thick and we normally use a set of around 0.5mm which should create a kerf of around 2.3mm instead it's more like around 3mm. Maybe bandsaws are different to handsaws

Measuring real kerf is quite tricky because cuts are rarely smooth enough to allow a feeler gauge or callipers to measure it accurately.
This diagram shows the potential problems.
kerfreal.jpg
Both feeler gauge and callipers will generally underestimate real kerf sizes with with callipers giving more variable results.
Measuring at the start (where saw first enters wood) or end of a cut where saw leaves the wood can have additional problems where the saw blade may flex a touch and widen the kerf.
What I found was a bit more reliable? was to make a test cut, and then cut through the wood perpendicular to the test cut such that it it exposed the cut pattern above and then measure that.
Whatever is used I found I had to accept that an absolute measurement was not really possible

These problems are more obvious in the case of chainsaw kerfs - see below especially where highly pointed chisel shaped cutters are used.
kerf.jpg
What I have settled on to measure kerf width is long thin tapered piece of hardwood (I call mine a "kerf gauge") marked out with the thickness (calibrated using a calliper at set points along the taper) of teh wood.

Lee Valley/Veritas sell these as taper gauges for silly money but they typically start at 1mm and only have 0.1mm resolution
Veritas Tapered Gauges - Lee Valley Tools.

I "tap" the gauge into the kerf by dropping a small wooden mallet from about the same height in the hope that this flattens some of the high and fibres protruding into the kerf and measure it several times along the cut.
It's a lot quick and easier to use in the field instead of fumbling with feeler gauges or risking damaging sets of digital callipers.
I use it as a relative rather absolute measure but recognise that even that does not eliminate all the issues.

For thinner kerf measurements as made by handsaws a finely tapered piece of metal would probably be better than wood, should be easy to make for someone with a metal milling machine.

Bob, indeed, as you point out, measuring kerf widths can be a fraught & inaccurate procedure! I was going to mention some of what you covered, but you did a much better job so I'm glad I waited.

Your mention of chainsaws is particularly apt. I've been slabbing up some radiata these last 2 weeks with my "ripper" slabbing rail. I tried measuring from the inner setting of the saw, as close as it gets to the rail, allowing what I measured as a typical kerf width between each slab or plank. But by the time I worked my way in & got to the last cut, the inner slab/plank was always thinner than I'd planned. So now I start the saw & make a cut, then measure the next plank off, allowing 3-4mm fudge-factor, move the saw over & make a good distinct cut there, then repeat til I get to the last cut. It's easy to re-set the saw in each short kerf as I work into the centre of the log, and I get no surprises when I get to the last cut.

My attempts at kerf-measuring were purely to establish some rough parameters so I had a place to start for setting backsaws. I hope I didn't give the impression I had the process sorted out for anyone other than myself. And don't anyone get the idea I always work with razor-sharp, perfectly set saws - I'm as guilty as anyone in putting off sharpening when trying to finish something, or just not feeling in the mood to set up & do it.....

Cheers,

Yep chainsaw are tricky to work with to get log straight cuts.
When I first started out after a few logs I saw the opposite effect to yours with the slab getting a few mm thicker as it got towards the end. This happened even when I tightened the mill verticals as hard as I dared (its only Al). It turned out that I had slightly uneven length cutters (the set facing one slide were different to those on onto other) which across the length cause the bar to dive into the cut. The locking mechanism for the mill verticals were cam locks but the vibe was enough to pull the vertical through the locking mechanism. I added some locking nuts to the all thread vertical bar adjusters and this caused the mill to jam part way through the cut until I identified the chain issue. These days I file the chain to constant raker angle so the length of the cutter doesn't matter so much.

Other issues with chains mills are parallel-ness of bar with mill and general flexibility. If the are not rigid enough the vibe can increase kerf width. Rocking the mill sideways in the cut can cause poor finish and widen the kerf on a irregular basis. I found its better if the mill is kept in the same position all the way thru the cut.

25% of blade thickness for seasoned timber was what we were told as apprentices. Seems to work just fine.