How do you measure how hard a balm or lip balm is?

Penetration depth in tenths of a millimetre, force to a fixed depth in grams, and the bend test for sticks. Published figures, the formula behind them, and home rigs.

Balm hardness is measured in one of two ways, and they run in opposite directions. A penetration test drops a loaded needle or cone onto the surface for a fixed time and reports how far it sank, in tenths of a millimetre: the bigger the number, the softer the balm. A texture analyser pushes a probe to a fixed depth and reports the peak force, in grams: the bigger the number, the harder the balm. Neat white beeswax reads about 22 tenths of a millimetre on a supplier's needle test, and patented solid lipsticks are reported to need at least 30 grams-force at 20 C. The fact that catches people out is that "harder" depends on the test. In one study, the lipstick that was easiest to bend was also the firmest under the needle, so a stick can be weak and firm at once.

• Neat beeswax: about 22 dmm
• Lipstick floor: 30 gf at 20 C
• Haighton: C = K W / p to the 1.6
• Sensory vs needle: R = -0.729
• Beeswax stick: 5,267 g to rupture

Two numbers called hardness, and which way each one runs

Every hardness figure you will meet for a balm, a wax or a lipstick comes from one of a handful of tests. They answer slightly different questions, and their numbers are not interchangeable.

The two penetration tests put a fixed load on the sample and measure distance. The instrumented tests drive the probe a fixed distance and measure force. So a "hardness of 20" means nothing without its unit: twenty tenths of a millimetre is a firm wax, twenty grams of force on a needle a soft balm.

None of these is the yield stress that a rheometer measures. They are empirical: each gives a reliable comparison inside one method, which is why the rheology and yield stress page keeps its pascals separate from penetration readings. The balm texture science page covers the words people use for what they feel; this page covers the numbers.

Needle penetration: the number on a wax data sheet

Needle penetration is the hardness figure printed on wax specifications. A weighted needle rests on a levelled sample at a controlled temperature, is released for a fixed time, and the depth it reaches is read off a dial in tenths of a millimetre, usually written dmm. The method most often named for it is ASTM D1321, written for petroleum waxes and borrowed for natural ones.

Kahl's technical data sheets, for one supplier's range, give these figures.

Two things follow. First, beeswax is the soft one. A sunflower wax that reads 2 to 6 dmm lets the needle in four to eleven times less far, so gram-for-gram swaps are a guess until measured; the wax substitution calculator works through the conversion. Second, the vegan substitute is sold on a range, 10 to 20 dmm, that sits just under beeswax's 22, which is what "substitute" means in practice: matched on hardness, not chemistry.

A neat wax figure tells you about the wax, not your balm. Hardness of the raw material and the ability to gel oil are different properties; the oleogels and wax networks page shows how a very hard wax can be a weak gelator. To know what a wax does at 10 or 15 percent in your oils, you have to make the balm and measure it, and the full set of wax properties is on waxes compared.

Cone penetration, and turning a depth into a yield value

For soft solids, a cone works better than a needle. Its contact area grows as it sinks, so the load spreads and the cone comes to rest even in soft material. The cone method is old and well worked out for fats. Writing for the Society of Cosmetic Chemists in 1966, P. Sherman of Unilever described it plainly: "A needle or cone (or sphere) penetrates the sample with a given force for a predetermined time, and the depth of penetration is measured."

The Haighton formula

A cone reading can be converted into a yield value in grams per square centimetre. Unilever's US patent 6,365,175 sets out the version used for fat spreads: a 40 degree cone is placed just touching the surface and "allowed to drop for 5 seconds", and

C = K × W / p^1.6

where C is the yield value in g/cm2, W the total weight of cone and sliding stem in grams, p the penetration in tenths of a millimetre, and K a factor for the cone angle, 5,840 for a 40 degree cone. The patent's cone weighed 80 g. The conversion lets two labs with different cone weights compare results.

Sherman added the classification, credited to Haighton, that gives the C-value its meaning, and a practical limit: "Reproducibility is most satisfactory when p is restricted to 7.5 - 20 mm."

Haighton's top band is "too hard" above 1,500 g/cm2, and the bottom "very soft to just pourable" below 50. For an 80 g cone, the edge of "too hard" falls at about 36 dmm and the boundary between "plastic and spreadable" and "hard, but satisfactory spreadable" at about 54. A firm stick balm will sit beyond the cone's useful range, which is where a heavier cone or a needle takes over.

Where the formula wobbles

The exponent is not a law of nature. The patent uses 1.6. Sherman wrote the same equation with an exponent that depends "on the properties of the sample being tested" and is "usually approximating to 2", with his own cone constant. A C-value from a paper is only comparable with yours if you use the same K and exponent. Haighton's bands are also a scale of spreadability, and the patent applies the method to fat spreads. They suit body butters and soft salves, which are fats meant to spread, and suit a lip balm less, since there "too hard to spread" is close to the design.

Texture analyser probes: force to a fixed depth

Cosmetic labs mostly use a texture analyser: a motorised arm that drives a probe at a set speed to a set depth while a load cell records the force. Three published lipstick studies give their settings in full, which makes them a usable template.

Esposito and colleagues give the benchmark: "Acceptable solid lipsticks for the consumers should achieve at least 30 g-force (gf) at 20 °C for 8.1 and 12.7 mm diameter stick, as reported for current patented lipsticks." That is about 0.3 newtons, a low bar. It is a minimum for a stick to function, not a target.

Notice that Kasparaviciene's lab turned the test around: they pushed until a fixed force was reached and reported depth, which makes it a penetration test again, with smaller numbers meaning harder. The same instrument can produce either kind of figure, so read the method before the result.

A texture analyser trace shows more than the peak. Huynh and colleagues note that "the needle penetration test can also indicate the presence of unwanted air bubbles or grainy texture, which would be seen as fluctuations in the force values during penetration". A jagged trace is the instrumental version of the fault on gritty balm from hard wax.

Why the bend test and the needle test disagree

Huynh and colleagues made four lipsticks at a fixed oil to wax ratio, changing only one wax, and tested each by bending and by needle. Their fourth formula, with 7 percent of an algal wax in place of candelilla, was "statistically significantly softer and easier to bend" than the control at 25 C. Under the needle, the same lipstick "was the firmest compared to all other lipsticks". Weakest by one test, firmest by the other.

Both are right, because they measure different failures. A needle measures local firmness, which Huynh ties to application: too firm and a lipstick "may feel waxy and dragging", too soft and pay-off runs too high. A bend test measures the force to bend a stick "until it broke off at the base", which Esposito reads as "an indication of the brittleness of the sample". The fourth lipstick was firm but the least stiff of the four: the authors found the algal wax "made the lipsticks more flexible".

The practical rule is to measure the failure you care about. For a balm in a tin, a penetration or needle-force test is enough. For a twist-up tube, the bend test matters as much as the needle, because a stick that snaps when extended is a failed product even if it feels right on the lips. That is the fault on lip balm won't twist up, and it is why lip balm ratios carry more wax for tubes than for tins.

Compression to rupture: the suppository tester

Pharmacopoeias already have a method for the strength of a moulded fat: the apparatus for testing suppositories, which loads a sample in steps until it collapses. Budai and colleagues at Semmelweis University used it, as described in the European Pharmacopoeia chapter 2.9.24, to measure "the hardness or ability to resist the rupture of sticks" made from yellow beeswax and plant oils, each stick 3.0 g.

Three findings matter for a balm maker. Oil softened the sticks without moving their melting behaviour much: at 60/40 beeswax and sunflower, softening point barely moved (64.0 to 63.5 C) while compression strength fell by 29 percent. Hardness and melting point are separate properties, so a drop point will not tell you whether a batch is firm. Second, at 40 to 60 percent oil "stick bases formulated with castor oil are the hardest", and at 70 percent the oils "do not result in different compression strength values". Third, adding carnauba or candelilla at 1 to 2 percent made the sticks harder, and "higher concentrations do not further alter the measured compression strengths". Beyond that, carnauba kept raising the softening point up to 3 percent and the drop point up to 4, while candelilla raised neither, which is the trade set out on beeswax vs carnauba wax.

The oil also moved the softening point once there was enough of it. With jojoba it fell from 61.3 C at 40 percent oil to 42.5 C at 80 percent; with castor it only fell from 62.5 to 57.3 C. Jojoba sticks at 70 to 80 percent oil were, in the authors' words, "too soft and resulted in physically uncomfortable stick application".

Does the number match what a finger feels?

An instrument is only worth the trouble if it agrees with people. Kasparaviciene and colleagues had sixty volunteers score thirteen lipsticks for hardness on a 0 to 10 scale and measured needle penetration depth on the same sticks. They found a "negative corelation" between sensory hardness and penetration depth, with a Spearman coefficient of R = -0.729: the harder people said a stick was, the less far the needle went. The amount of beeswax "had higher influence on hardness and penetration depth than amount of oils mixture". Beeswax content also tracked melting point closely (R = 0.912), while hardness and melting point were only moderately linked (R = 0.484).

A coefficient of -0.729 is strong but not perfect: a finger mixes firmness with glide, tack and warmth, and a needle measures one of those. Use the instrument to tell you whether a change happened, and a panel to tell you whether anyone will like it, as described on sensory testing.

What you must hold constant, or the number means nothing

The formula is often the smallest thing moving a hardness reading. Fix these, write them down next to every result, and a cheap rig becomes trustworthy.

1. Temperature. Every study above conditioned samples before testing, and every one found warm sticks softer. Huynh's lipsticks at 45 C were softer on both bend and needle "as expected". Test at one stated temperature, and if you sell into hot places, test there as well. Why a balm softens far below its melting point is set out on solid fat content.
2. Age. Huynh and Esposito both held their samples for 24 hours before testing. A balm keeps firming for days as its crystals settle into a more stable form, so a same day reading measures the pour, not the formula. Balm hardens over time covers the drift.
3. Cooling history. Two tins of one batch cooled differently are two different materials. Pour at a fixed temperature and cool every sample in the same place; crystallisation and cooling rate explains why.
4. Geometry. Test in the middle of the surface, well away from the wall, with at least 10 mm of balm under the deepest point the probe reaches. Use the same container size every time.
5. Replicates. Huynh tested five sticks of each formula and Esposito six. Three is the least that shows you the scatter; a single reading cannot tell a real change from noise.

Three home rigs that give repeatable numbers

None of these gives a figure you can compare with a published paper. Each gives a figure you can compare with your own last batch, which is what batch to batch inconsistency needs.

1. The weighted drop: a home penetrometer

This is the drop test: a needle or cone test made from a dowel. Take a smooth rod of fixed diameter, 3 to 4 mm, in a loose vertical guide such as a drinking straw or a drilled block clamped above the tin. Fix a weight to the top so the moving part weighs a known amount, 50 or 100 g. Lower the tip until it just touches the surface, release it for 5 seconds, then lift it out and measure the hole depth with the depth rod of a digital caliper. Record it in tenths of a millimetre so it reads in the same direction as a wax sheet: larger is softer. If you can turn or buy a 40 degree cone of known weight, you can go one step further and convert the depth with the Haighton formula above, keeping inside 7.5 to 20 mm.

2. The scale push: force to a fixed depth

This copies what a texture analyser does, in reverse. Stand the tin on a kitchen or jewellery scale, tare it, and push a 2 mm rod straight down by hand, slowly and steadily, to a depth mark on the rod, say 5 mm. Read the highest figure the scale shows. That peak in grams is a needle-force hardness: larger is harder. Push steadily to the mark; speed changes the reading. A 0.1 g scale is fine here, since the forces run to tens or hundreds of grams, and weighing and calibration covers checking the scale itself.

3. The stick bend: will it snap in use?

For tubes, wind the stick fully up, lay the tube horizontally in a clamp, and hang a small paper cup from a thread looped 3 mm below the tip, the same position Huynh's bending blade pressed. Add water a few grams at a time until the stick breaks, then weigh cup and water. That weight is the bend strength. Run it at room temperature and again after an hour somewhere warm.

Reading your results: what counts as a real change

Work out the spread before you trust a difference. Take your three to five readings on one batch. If the gap between two batches is smaller than the spread within one, nothing has changed that the rig can see. If it is two or three times the spread, it is real.

Then use the number to steer. If a balm reads softer than the target, add a point of wax at a time and test again; balm too hard or too soft gives the adjustments and the wax ratio calculator the starting ratios. Budai's results suggest the first point or two of a hard wax moves hardness most, so expect diminishing returns. And do not expect hardness to explain drag: Spearman's -0.729 leaves room for two balms of equal firmness that feel different, which is the subject of drag and poor glide.

Where these numbers stop applying

The wax penetrations are one supplier's sheet values, with the method not given on the page, and are typical figures rather than a guarantee for any lot. The Haighton conversion and bands come from fat spreads, and the exponent differs between sources, so C-values only compare when the formula is the same. The texture analyser settings and the 30 gf floor come from lipsticks, which carry more wax than most balms. The compression strengths are for 3 g beeswax sticks tested on a pharmacopoeia suppository tester, with glycerol and lanolin in most samples, and the sensory correlation comes from sixty untrained volunteers scoring thirteen lipsticks.

None of the studies tested a shea or cocoa butter balm in a tin. The methods carry over; the numbers do not. What does carry over is the shape of the answer: a hardness figure only means something with its probe, load, time, temperature and age attached, and two honest tests can rank the same products in opposite order.

Reviewed and updated 22 October 2026. Spotted an error? Tell us and we will fix and log it.