What size steel for a carving knife?

A carving knife is long and narrow, and its height at the heel is what sets the bar. The preset is an 8 1/2 in blade, 1.15 in tall at the heel, with a 3/32 in (0.094 in) spine and a 5 in full tang. Add 1/8 in for clean-up and you need 1.275 in, which rounds up to 1 1/2 in. The 3/32 in spine sets the thickness, so the bar is 1 1/2 × 3/32 in. Forged or ground, it is the same bar.

The matching KraftyDesign blank (the "Draw this blank" link) is a long, narrow blade with a mostly flat edge and a sharp point set a little below the spine. The handle lines up with the spine, and the edge drops below the handle at the heel, which gives your knuckles a little room above the board.

Thin matters on this knife. America's Test Kitchen (ATK) tested slicing knives, the carving knife's round-tipped cousin, and liked blades that were thin but just stiff enough to control. Its winner was 1.5 mm (about 0.06 in) at the spine. At about 2.4 mm, 3/32 in bar leaves room to grind a taper toward the tip. The calculator starts on 1084, and AEB-L and 15N20 give the same 1 1/2 × 3/32 in bar. Larrin Thomas of Knife Steel Nerds says AEB-L excels where thin edges are needed, kitchen knives among them. 14C28N is sold in metric thicknesses, so the calculator flags 3/32 in as uncommon for it. Check what your supplier stocks.

The calculator starts on stock removal for this one. Stock removal takes the whole 13 1/2 in profile plus clean-up: 13 3/4 in of bar per blade. That is about 9 oz of bar for a blade that finishes around 5 oz. A 12 in bar is too short. The 18 and 24 in bars give one blade each, a 36 in bar gives two and a 48 in bar gives three at the lowest cost per blade.

Forging the bevels and the point stretches the blade, so about 9 1/2 in of the same bar (about 6 oz) is enough. The blade finishes near 4 oz. The 12 and 18 in bars give one each, 24 in gives two and 36 in gives three. A 48 in bar gives five at the lowest cost per blade. If you start thicker and forge down, set "Forged thickness" to "One size up". That starts you at 1 1/2 × 1/8 in, and it only applies in forged mode. The calculator won't forge AEB-L, CPM-154, 14C28N, MagnaCut or A2. Pick any of them and it works out the bar as stock removal, even with Forged selected.

Long, thin blades like this one can warp in the quench. In a BladeForums thread on thin W2 kitchen knives, one poster's blades were 0.086 to 0.098 in thick at the spine by the plunge, up to 9 1/2 in long. One of them came out warped from end to end and took three heats clamped to a length of angle stock to straighten. Another poster in that thread works from about 3/32 in bandsaw steel, and cuts the profile and heat treats before grinding anything. In another thread, on a thin 10 in slicer, one poster still got a little warp even with a plate quench.

What a carving knife is for

A carving knife is for slicing big cooked meats. Wikipedia lists poultry, roasts and hams, and gives the length as 8 to 15 in, so the 8 1/2 in preset sits at the short end. Wikipedia also says a carving knife is much thinner than a chef's knife, especially at the spine. That lets it cut thinner, more precise slices. It is often sold with a carving fork, a long two-pronged fork that holds the meat steady while you carve. Many carving and slicing knives have a row of hollows ground into the sides near the edge, sold as a Granton edge or a Kullenschliff. Wikipedia says Granton hollows are normally found on meat carving knives. On meat they seem to work. In ATK's slicing knife test, slices of brisket and turkey stayed in place instead of sticking to the blade, and all of its top-ranked knives had them. The review adds that the scallops make the blade thinner and lighter so it slides through food, while the spine stays stiff enough for control. William Grant of Sheffield got a US patent in 1931 for a blade with a row of hollows beside the edge. The patent itself says the edge is particularly suited to cutting tomatoes, fruit and the like.

The carving knife is older than the fork that goes with it. The Victoria and Albert Museum (V&A) in London says the carving fork first appeared in the late medieval period, for use with a carving knife that had evolved earlier. Sets of knife and fork became more common from the mid-17th century. In the 18th century the host or hostess carved the roast at the table, and the V&A says carving was seen as an important social skill for men and women. Then a change in how dinner was served moved the carving to the servants, away from the table. In the 19th century carving sets were usually sold boxed, in two, three or five pieces. The bigger sets added a smaller knife and fork for poultry and game birds to the usual knife, fork and steel. A Sheffield carving knife in the V&A, hallmarked 1884 to 1885, is 31.4 cm (about 12 1/4 in) long overall. Its gallery label says the blade has been shortened by repeated sharpening on the pointed steel. Today the sets are less common. A knife expert quoted by ATK says carving knives used to come in sets with a fork, and that the sets have lost favor over the last few decades. ATK links that to fewer families sitting down to a Sunday roast.

The Granton stories that follow come without sources, so read them as lore rather than history. Two knife retailers, Russums and House of Knives, say Granton knives are made by a Sheffield firm with roots going back to 1601. One poster on The Kitchen Knife Fora says Granton knives were designed in the 1920s for carveries, butchers' shops, hotels and restaurants. Even the date of the patent is muddled. Wikipedia and House of Knives say the Granton edge was patented in 1928, though Wikipedia flags that line as needing a citation. William Grant's US patent cites a British application from February 1929 instead.

Stock for a carving knife

In a BladeForums thread on a turkey carver, one poster suggests starting from bar about 0.100 in thick for a 10 in blade about 1 1/8 in wide. That poster would finish the spine at 0.080 in, tapering to 0.040 to 0.050 in at the tip, and grind the bevels after heat treat. For a 10 in slicer planned in 1/16 in A2, one poster in another BladeForums thread recommends 3/32 in stock instead, because 1/16 in would flex too much to cut well.

Widths assume blade height + 1/8 in rounded up to stock; thickness is the finished spine rounded up. Change the inputs above for your own pattern.

Steels that suit it

Questions