Is denser wood stronger?

Denser wood is harder to carve, and stronger, but nowhere near as much. The 2010 USDA Wood Handbook ties both to specific gravity, and hardness rises about twice as fast as bending strength. When a board fights you, my answer is that the problem is not the wood, the problem is the tool.

You bought the heavier board because heavy sounded strong. Then the first relief cuts crumbled at the edge, and the background tore instead of coming away clean. Being annoyed about that is fair.

Density really does change what happens under your gouge, and it changes it a great deal, but almost none of that change lands where most woodworking talk puts it. Below are the handbook numbers, and beside them what I watch happen in limewood and in oak.

Here is the whole page in six lines, before any arithmetic:

• Density gives a good index of strength, and never a promise.
• The 2010 Wood Handbook's table shows a doubling of specific gravity multiplying side hardness by about 4.26, and stiffness by only 1.62.
• One live oak row in the handbook's own species table is denser than pignut hickory, and bends worse.
• Hardness is one steel ball pressed into one flat face. It says nothing about a knock.
• In the handbook's spread figures, gravity varies about ten percent inside a species, while toughness varies thirty four.
• A board changes density inside itself, and I read it by colour.

How do you define wood strength?

Strength is four separate tests, and a board can win one test while losing another. Ask which test they mean. The handbook prints no column called strength.

It prints bending strength, stiffness, crushing resistance, hardness and several more, each measured on its own small block of clear straight grained wood. Those blocks are prepared to the small clear specimen standard, which exists partly to separate the influence of density from where the tree grew and how high up the trunk the board came from. Every figure you will read below started life as a clean laboratory stick with no knots in it. Your board has knots.

The bending test, or modulus of rupture, measures the load a beam carries before it snaps. Elastic modulus is stiffness, meaning how far that same beam sags on the way there. Crushing resistance is what a post does under a downward load. Toughness is the odd one out.

The 2010 handbook defines toughness as the energy needed to cause rapid complete failure in a centrally loaded bending specimen, which is about as close as a laboratory gets to a carved bracket being knocked off a shelf by a sleeve. Toughness decides whether your work survives the house it lives in. Almost nobody quotes it.

Four tests, four answers, and no way to collapse them into a single word. Each property has its own page here, and I would rather send you to those than rebuild them badly, so begin with the strongest wood question if ranking is what you came for.

How is wood density measured?

Density is written down as specific gravity, which is a ratio and not a weight. Cell wall material has a gravity of about 1.5 in every species. Oak and balsa are built from nearly the same stuff.

What differs is how much air sits in the cavities, so the ratio really counts how much wood substance fills a given space, and the Forest Products Laboratory chapter on mechanical properties states that plainly. My page on specific gravity walks the idea through slowly. A catch waits there that spoils half the chart comparisons on the internet.

The Wood Database counts at least five different ways to express it, which is why American beech gets published as 0.54 in one place and 0.73 in another. Same tree. Two honest numbers. Published dried weights are standardised at 12%, the level wood settles to near seventy degrees and about sixty five percent humidity, and freshly felled stock can weigh more than double its ovendry weight.

So a board feels heavier this month than it did last month without a single fibre changing. Wanting the plain definition with none of the arithmetic is reasonable, and I wrote what wood density means for exactly that reader. Companion pages cover high density species and the low density end, plus a short one answering whether dense wood floats.

What is wood hardness?

Hardness is one test with a steel ball, and it measures nothing else. Everything else you have heard about hard wood is inference.

The Wood Database describes the Janka hardness number as the force needed to embed a ball of 0.444 inches, or 11.28 millimetres, to half the ball's diameter in a sample dried to 12%. White oak lands near 1,350 pounds of force. Basswood sits around 410. Lignum vitae, at the far end of that scale, reads 4,390.

Not once have I quoted one of those numbers at the bench, and I am not about to start. They are useful, and they answer a question I do not have. A ball pressed slowly into a flat face models a chair leg meeting a floor very well, and a gouge slicing across the grain at an angle very badly. Should the scale itself interest you, a whole page covers what Janka hardness means, and another chases the densest wood ever recorded.

---

How much harder does denser wood actually get?

Doubling the density multiplies hardness by about four, and stiffness by roughly half again. Table 5-11a of the 2010 Wood Handbook gives each property as a power function of gravity.

For hardwoods dried to 12%, side hardness works out as 15,300 multiplied by G raised to 2.09, in newtons. Bending uses 171,300 and an exponent of 1.13. Stiffness gets 16,500, with the exponent dropping right down to 0.70. Maximum crushing resistance runs on 76,000 and an exponent of 0.89, which is the strange one.

Read those exponents and the argument falls out on its own. An exponent above two means hardness climbs faster than the density does. Below one, as with crushing at 0.89, doubling G buys you less than double the resistance to being squashed. Stiffness at 0.70 barely notices that the board got twice as heavy.

Put numbers on it. A wood at 0.40 gravity and a wood at 0.80 sit one doubling apart. The heavy one resists your edge four and a quarter times as hard, and carries just over twice the load before a stem breaks. You will feel the first difference straight away. You may never meet the second. Both facts come off one line of one table.

That table is the article. Density buys resistance under the edge far faster than it buys anything the finished carving will ever use. Four and a quarter times the trouble for two and a fifth times the strength is a poor trade, and you pay it with every stroke.

The 2021 edition, published as the current Wood Handbook, revised these constants, fitting hardwood side hardness at an exponent of 2.15 rather than 2.09, so quote one edition and stay inside it. The 2010 volume used here is catalogued at the Forest Service record for GTR-190, and the newer one at the record for GTR-282.

Is the densest wood always the strongest?

Live oak and pignut hickory settle that in one line of the handbook. Two hardwoods, one much heavier, and the heavy one loses.

Measured at twelve percent, the handbook records live oak at a gravity of 0.88 with a modulus of rupture of 127,000 kilopascals. Pignut hickory, considerably lighter at 0.75, is recorded at 139,000. So the denser wood is about nine percent weaker in a bending test, and it is less stiff into the bargain, 13,700 megapascals against 15,600.

Buried in that chapter is the quiet part. Gravity is called a good index of mechanical properties as long as the wood is clear, straight grained and free from defects, and the same page warns that those power functions will not accurately predict an individual species value or an individual specimen. They were fitted across sixty six hardwood species and forty three softwoods. An average is a wonderful thing to argue with and a poor thing to carve.

Then comes the spread. Averaged over about fifty species, the handbook's Table 5-6 puts the coefficient of variation for gravity at ten percent, and for toughness at thirty four. So the steadiest number in the whole set is the one everybody quotes, and the property deciding whether a carving survives a knock is the wobbliest of the lot. Bad pairing, if you were hoping one predicts the other. On the practical side of it, the woods that fight the gouge is the page you want.

---

Why does one board change hardness inside itself?

Density will not be one value for a board, and I read the change by colour before picking a gouge. On limewood you can see the variation with your eyes.

Even inside the same lime wood there are variations of softness, and the colour changes along with them. Lighter here, darker there. In my case the darker patch is the softer one, which is the opposite of the olive.

Same board, same tool. The middle tore out in lumps and the lobes came away smooth.

Basswood behaves the same way, and knowing it will stop you panicking. If you see a dark spot in basswood, the material there is a little mushy, which is absolutely fine. You carve it, you do not fight it. Olive runs the other way. Really stubborn, that wood, and that is the nature of olive. Some parts show the variation of colour, and on olive it is harder to carve on a darker wood and easier to carve on a lighter one. So density is different, and the colour code has flipped between two species.

Two woods, two opposite readings of the same clue. No chart carries that for you. You learn it board by board, with a gouge already in your hand.

Here the laboratory and the bench finally agree with each other. Curves get fitted to species averages, and then the chapter tells you the curve will not predict your specimen. I am standing over one board watching it change under one gouge. Both statements say the same thing in different accents.

Take that seriously and it changes how you lay out a drawing. Put the thinnest stem where the board carved cleanly, and let the coarse ground fall on the patch that tore. Nobody can plan that from a chart. You plan it from the first chips off the board.

This next one runs about a quarter of an hour, and in it I argue that basswood is the woodworker's MDF.

Does denser wood need a different tool?

A denser board changes the tool before it changes the effort. On the harder woods a long bevel is a problem, because the metal behind the edge becomes really weak at the point.

Oak does that. Cherry does that. Not every oak, mind you, since brown oak is a little softer than a white American oak. My full reasoning on edge geometry lives in the course about tools, and the short version is that a slender wedge of steel cannot survive being driven into something that will not part easily.

To carve oak you will most likely need a mallet, and the bevel has to suit how hard you intend to hit. Mine is not measured. I place my hand underneath the handle, take hold, and the knuckle should almost touch the piece of wood, which means the edge is already grabbing material. That is the best bevel for me, mallet or no mallet. Press me for a number and it comes to about, I would say, between 18 and 22 degrees, if you want to be really technical, but I am not measuring. Some of my own bevels are actually really extreme, a good deal lower than that. Whether you need a mallet at all is its own question.

Crunchy fibre changes the movement as well. Carving a circle in a dry apron, I could not simply place a gouge and rotate it around the centre, because I would snap the centres clean off, and those centres are the high points I have to keep. Little by little instead. Nothing clever about it at all.

That job is the reason I reach for the V-tool when the wood is really dry. A palm V-tool, to be exact, because I get a little more control with the palm one. You are not supposed to hit that with a mallet. I am barely touching it, and even so the mallet gives me much more control of the tool.

One depth rule goes with all of this, and it matters more the harder the wood gets. Do not go more than an eighth of an inch. Later you can always take it a little deeper, but not at this stage.

---

How do you carve a denser board without tearing it?

Work the order, not the muscle, and the dense board stops tearing. These six things change when the wood under me is harder than limewood, in the order I change them.

Step 1, read the board first

Look for the colour changes before a pencil goes anywhere near the surface. Darker patches in lime and basswood are usually the softer material, and a dark spot in basswood means mushy fibre that carves perfectly well. Put your most delicate detail somewhere else if the drawing allows it.

Torn fibre, not a clean sliced surface. This is what the wood does when it is denser than the stroke allows.

Step 2, shorten the bevel and sharpen it properly

A long bevel goes weak at the point, and hard material finds that weakness immediately. Grind it shorter and strop it. My sharpening course covers the whole sequence, and you will find a free lesson on sharpening a gouge if you would rather try before anything else.

Step 3, do not skip the relief cuts

With crunchy wood those relief cuts are not optional. Straight down, all the way round, before any waste leaves the board. In basswood you can skip a few and get away with it. Oak will charge you for every one you skipped, and it charges in torn fibre along a line you cannot put back.

Step 4, take smaller bites

Little by little is the whole method. A rotating gouge that would sweep a full curve in limewood snaps the high points out of dry crunchy fibre, so break the movement into short passes and leave those high points standing. The mallet does the work. Your wrist does the aiming.

Step 5, watch for fibre that presses instead of slicing

Walnut runs a little harder than lime and basswood. It is not a hard wood by any real measure, but it is harder, and pressing into it makes the fibres smush both ways. You can go deeper with a knife than you expect. Later, even after you clean the shape nicely, those pressed fibres pop back up and ruin the surface. All I want from that first pass is a line to place a gouge into.

Step 6, blame the tool before the board

A few species are almost impossible to carve, and I will grant you those. Walnut, cherry, basswood and lime are not among them. When my own work smushed and I complained to my teacher that my wood was no good, he gave me the same answer every time. There is no such a thing as a bad wood. Only dull, not sharp tools. He was right, and I say it now to anyone who complains about a board.

Can wood density change over time?

Wood substance does not change, but the number on the chart moves with the water. Weight and density are not the same claim.

Freshly felled stock of some species holds more water by weight than fibre, and nothing shrinks at all until the free water has gone. Below the fibre saturation point, which the chapter on moisture relations averages at about thirty percent, the cell walls begin giving up water and the board shrinks around its own substance.

The grain runs one way and the drawing runs the other.

Indoor furniture and trim are meant to be installed at about eight percent, according to the drying chapter of the same handbook, and six percent in the dry southwest. One more wrinkle sits underneath all this. A published gravity also reflects gums, resins and extractives, which add weight while contributing little to mechanical properties, so two boards at the same figure can behave unlike each other under your edge.

How much a board then moves is a thickness question, and I handled it in whether thicker wood carves better and in what shrinkage means. The tangential to radial ratio is the number that matters there, not the weight. Underneath the lot sits the fibre picture, where quartersawn stock checks and splits less than flatsawn, whatever the species weighs. And what makes a wood less dense answers the other direction.

Does density decide how long a carving lasts?

Moisture decides that, and density barely gets a vote. Air dried wood usually sits under twenty percent moisture content, which turns out to be the whole story.

The chapter on biodeterioration is blunt about the consequence, which is that wood kept air dry will not decay, and any decay already in it stops progressing. Keep it dry and it stays. Soak it and no amount of density saves it.

The proof of that sits in museums, carved in one of the softest woods on the list. Grinling Gibbons, who lived from 1648 to 1721, carved his lace cravat out of a single block of limewood weighing less than 150 grams, no more than an apple, and the V and A says it has inspired wonder for over three centuries. The piece itself is catalogued as London work of about 1690.

Then there is Tilman Riemenschneider's limewood Mary Salome and Zebedee, carved in Wurzburg around 1505. That one is still here too, and the museum record notes it was seriously worm infested at some point and the holes were plugged with wood. Five hundred years, in soft material, with an insect problem, and it survived. I wrote more about him in his unpainted limewood, and about survival generally in how long carvings last.

---

So which wood should you carve?

Basswood, limewood, linden, and I will not apologise for it. Probably the best material for wood carving there is.

You can do a lot more with limewood than you can with oak. Oak is more crunchy, and it has a lot of grain. None of that is a complaint about oak. It is a description of what oak asks from you before it gives you anything back.

Heart pine deserves a mention here, and it is not even a hardwood. Hard as a rock. A client of mine demolished nineteenth century multi storey factories along the East Coast and harvested the heart pine posts out of them. He slices those posts into veneer about an eighth of an inch thick and sells it on to flooring makers, who laminate it onto plywood and sell it really expensive. Then he asked me to carve a table for his office, and that was huge, about 12 feet. It took me a long time and I broke lots of gouges carving it by hand. Driving a nail into that stuff is almost impossible. Heart pine comes from the middle of a very big tree, close to the heart, which is also why those posts barely move. Density belongs to the piece on your bench, not to the species name on the invoice.

Walnut makes a fair middle road once your edge is right, and the Wood Database puts black walnut at 1,010 pounds of force with a bending rupture of 14,600 pounds per square inch. I have written up carving in walnut separately. A species by species read lives in the wood guide, with pages on basswood and red oak, and the whole cluster sits under understanding wood for carving. Beginners do better starting in soft material, which is what the beginner course assumes from the first lesson.

---

Bring a board, sharp tools and a stubborn streak. Shall we carve?