knife-sharpening-angle-guide
Use when sharpening kitchen, pocket, or workshop knives (whetstone, guided rod, or pull-through) — matches the correct edge angle to knife class (Japanese 12-15° vs German 20° vs cleaver 25°+), builds the grit-progression ladder from chip repair to polish, computes the physical sharpening-pass height to hold an angle by hand, and explains burr formation/checking so the edge actually gets sharp instead of just shiny.
Knife Sharpening Angle Guide
Overview
Sharpening fails for exactly three reasons: wrong angle for the knife class
(a 12° edge on a German chef's knife folds on the first pumpkin), skipped
grits (trying to finish on a 300-grit stone makes a toothy scraper, not an
edge), and never raising a burr (polishing an edge that was never formed).
scripts/edge_guide.py holds the lookup tables: per-class target angles,
asymmetric ratios for Japanese blades, grit ladders chosen from your
starting defect, pass-height math ("hold the spine X mm above the stone at
this blade height"), and burr-stage checks per side. It also decodes steel
types into sharpening-feel expectations (soft German steel vs powdered
metallurgy) and pick-a-stone advice from whatever stones you own.
When to Use
- Knife is dull, chipped, or "sharpened" repeatedly by a pull-through and now won't take an edge.
- First time with whetstones; you don't know what angle or grit to use.
- Mixed drawer: German chef's, Japanese gyuto/santoku, paring, bread, boning, pocket knives — one angle does not fit all.
- Deciding whether a knife is worth sharpening vs the steel is just bad.
- Restoring a thrift/flea-market find with chips and a heel swale.
Don't use for: serrated bread-knife sharpening (rod-and-ceramic per- scallop technique — different process), scissors (bevel geometry differs), or axe/machete geometry (much blunter, different domain).
How It Works
- Angle by knife class. Japanese double-bevel (gyuto, santoku): 12–15° per side; German/Western chef's: 20° (17 for a thin laser); paring 15–17°; boning/fillet flexibility 18–22°; pocket/EDC 20–25° (edge stability over sharpness); heavy deba/cleaver 25°+ with a hammer-spine left blunt.
- Asymmetric ratios. Many Japanese blades run 70/30 or 80/20 — the script tells you passes per side to maintain the factory asymmetry rather than accidentally centering it.
- Grit ladder from defect state.
- Chips or tip repair: 220–400 → 1000 → finisher.
- Very dull / first stone session: 400–800 → 1000 → 3000–6000 (or 8000 for polish lovers).
- Refresh a working edge: 1000 → 3000/6000.
- The rule the script enforces: never jump more than ~3× grit in one step; each stage must remove the previous stage's scratches.
- Hold-the-angle math. Height of the spine above the stone for a target angle = blade width × sin(angle). For a 45 mm German knife at 20°: raise ~15 mm. The script prints a per-knife height table at 15° and 20° so hand-sharpeners have a physical reference, plus the "matchstick/marker-trick" description.
- Burr discipline. Sharpen one side until a consistent burr forms the FULL length (feel by stropping thumb AWAY from edge on the opposite face); then mirror passes on side two; then one-lightest-pass alternating to shed the burr; then strop. The script narrates the stage checks.
- Steel decoder. Soft stainless (X50CrMoV15 etc.): sharpens fast, dulls fast, forgiving. Hard Japanese (VG-10, ~60 HRC): burrs are small and brittle — lighter pressure. Powdered steels (SG2/R2, S35VN): slow on stones, need diamond/CBN for the burr stage. The script prints expectations so you don't misread steel behavior as technique failure.
- Stropping + maintenance. 1–2 µm compound on leather or bare cardboard; weekly stropping pushes stone sessions to months apart. The honing-rod doctrine: rods realign, stones remove — a rod cannot sharpen a dull knife, only postpone the stone.
Using the Script
# angle + grit plan for a German chef's knife, currently very dull
python3 scripts/edge_guide.py --knife german-chefs --condition very-dull
# full drawer plan
python3 scripts/edge_guide.py --drawer
# pass-height table for hand-holding angles
python3 scripts/edge_guide.py --heights --blade-mm 45
# what stones you need if you own these
python3 scripts/edge_guide.py --stones-owned 1000 6000
# steel expectations
python3 scripts/edge_guide.py --steel vg10
Common Pitfalls
- One angle for everything. 15° on a boning knife folds on bone; 25° on a gyuto makes it a wedge. Class first, habit second.
- Chasing polish without a burr. If no burr ever formed, no new edge geometry exists — the knife will look great and cut nothing. Burr first, always, full length.
- Skipping grits. 400 → 8000 leaves 400-grit scratches under a shiny bevel: an edge that fails in a week. Max ~3× jumps.
- Pushing pressure. Weight of the knife (plus fingers) only; pressing bends the edge away and produces a wire edge that snaps off dull.
- Pull-through sharpeners on good knives. They rip carbide chunks out of hard steel and set a sloppy 20°+ micro-bevel. Fine for a $10 beater, destructive on VG-10+.
- Ignoring asymmetry. Centering a 70/30 Japanese blade changes its steering and food-release behavior; count passes per side.
- Testing sharpness wrong. Paper-slice test (phone-book paper, one smooth pull) and tomato-skin test, not the thumb-across-edge gamble.
Verification Checklist
- Angle chosen from knife class table (or factory bevel matched).
- Grit ladder has no >3× jumps from your starting condition.
- Burr raised full-length on side one BEFORE switching sides.
- Alternating lightest passes shed the burr; no wire edge remains.
- Stropped; paper test slices cleanly without snagging.
- Asymmetric blades: pass counts per side preserved (70/30 etc.).
- Downloads
- Security audit
- Pass
- Last updated
- 3w ago
- Current version
- v1.0.0
- License
- MIT-0
