HOWTO: Kerf Measurement and Jig


#1

Understanding the interaction between your tools and materials is critical to making parts that fit well. A key metric to understand is kerf: the amount of material removed by a cut. Compensating for kerf is the difference between joints that are sloppy or that fit with just the right amount of force.

Even though a laser cutter removes only a fine amount of material, that slight difference in dimension between your plans and the physical material adds up. It’s worth noting that your kerf can vary between batches of the same material, power and speed settings, and over the lifetime of the machine. Because a laser beam is only in focus at a specific point, the kerf also varies throughout the thickness of a cut.

Here’s a jig you can use to measure kerf:
kerf%20tool

It works by comparing a reference length (A) to the length of a set of blocks (B) which have been cut from the same stock. That difference is the sum of the kerfs of each cut. It’s much easier to measure this compound kerf than to try to measure the kerf directly by cutting a square of known size.

Assemble the jig

Each block has OK or Flip engraved on it. This allows us to compensate for the vaguely trapezoidal shape of the blocks due to divergence of the laser beam. We’ll flip every other block left-over-right. This example is made with Proofgrade Thick Acrylic to show you how the jig should be assembled. You’ll notice that the Flip marks are backwards and on the bottom of the block when inserted into the jig.

Measure the reference

If we had an ideal cutter that could make perfectly planar cuts down at the molecular level, the reference portion of the jig and the blocks should have exactly the same length. We’ll start by measuring length (A), the top-most width of the jig. You should take several measurements and average them together.

Measure the blocks

The jig has a couple of cut-outs to make it easy to insert calipers to measure the width of the blocks. Again, you should take several measurements and average them out. If you’re feeling adventurous, you can measure the blocks outside of the jig.

Compute the kerf

In this case, my average (A) value was 109.64mm and (B) is 107.5mm. This gives us (109.64 - 107.5) / 10, or .214mm / 0.0084in. This is a starting value that I’ll use, adding an extra thousandth of an inch for joints, or subtracting one or two thousandths for clearances.

Additional reading

If you’re a Fusion 360 user, I highly recommend the plugin DXF for Laser, which allows you to export faces as DXF files and allows you to specify a kerf value to apply to the source drawing.


Laptop riser with cable management
#2

I think I’m missing the obvious here, but you have 11 blocks, and that is 12 kerfs ?
John :upside_down_face:
On second thought A has a kerf missing at each end - is that it ?
Or 22 half kerfs across B, and 2 half kerfs across A !


#3

It’s a “fencepost” problem: If you want to build a fence 10 yards long using 1-yard beams, how many posts do you need? Eleven.

This jig assumes that the reference edges and the edges of the two outermost blocks are co-linear. Since the same amount of material ought to have been removed, we don’t need to think about the outside kerfs. This leaves us with only the 10 cuts that separate the 11 blocks.


#4

Ah, but my explanation shows how confused I get.
Time for a cup of tea, I think.
:upside_down_face:


#5

Nice! Thanks for sharing the file. :grinning:


#6

Over the years I directed the day to day work of hundreds of engineers and programmers, and folks that had PhDs in Math or Physics. There was one Physics PhD that was brilliant but communication between us was almost impossible. He thought in abstract terms and I did not. Had to assign another engineer to be a translator between us.

For my simple mind… The first measurement already includes the two outside cuts so there are 10 additional kerfs to average.


#7

Very nice jig and write up.

Just a couple of thoughts from earlier conversations and personal experiences.

I love the DFX to Laser plug-in but I’ve come to prefer adjusting for kerf right in Fusion 360. This way I get total control over what compensates and what does not.

Also, while your jig most likely gives very good numbers ( I love the /10 thing) I find that differnt typs of joints make a difference so I tend to play with a sample joint to get things fitting just right.

Anyway, thanks for sharing this. It definitely adds to the conversation.


#8

i always do this. i try to figure kerf. then i make my box in boxes.py. then i snip out a section of both sides (maybe 2-3 fingers) and cut that out of some sample of the material and make sure the pieces fit. and if i’m off in my kerf, i go back to boxes.py and adjust the number slightly and try another sample joint. it’s tedious, but even when i think i have kerf right, my real world sample can be either slightly too tight or slightly too loose.


#9

I’ve also noticed this. Thought I had settings dialed in, made 20 copies of something, and they all fit together a little differently.


#10

that’s because of this:

and even the same sheet can vary in thickness from one end to the other.


#11

Well, thank you for this—looks very useful!