What is yield stress, and why does a balm hold its shape until you rub it?

Yield stress is the stress a balm resists before it flows. Measured values for soft wax gels run 2 to 18 Pa. How it is measured, and why stirring cuts it.

Yield stress is the stress a material can carry without flowing. Below it, a balm behaves as a soft solid: it holds its shape in the tin, keeps a pigment particle where it was poured, and stands up as a stick. Above it, the crystal network breaks and the balm flows, which is what happens under a fingertip. The numbers are smaller than most people expect. A rice bran oil gel with 3 percent beeswax, the least wax that makes it solid, measured a yield stress of 2.29 pascals and began to flow outright at 18.55 pascals. The fact that changes how you make balm is that a broken wax network does not rebuild itself at room temperature. In the same study the beeswax gel recovered under 2 percent of its stiffness after shearing, so a balm stirred while it sets stays weaker until it is melted again.

• 3% beeswax gel: 2.29 Pa yield, 18.55 Pa flow
• Carnauba gel: 16.32 Pa still, 0.43 Pa sheared
• Beeswax gel recovers 1.89%
• Petrolatum G' about 4,500 Pa at 37 C
• Wax lip balm G' down tenfold at 50 C

Four quantities that get mixed up

Most confusion about balm texture comes from four quantities that are easy to blur. A figure in pascals means something quite different from one in pascal seconds.

The distinction that matters is between viscosity and yield stress. Castor oil has a high viscosity and no yield stress: tip the jar and it flows, slowly. Tip a tin of balm and nothing moves, however long you wait. That is why the pigment sinking page argues that viscosity only slows a falling particle while a yield stress stops it, and why the glossary defines yield stress as the difference between a product you can scoop and one you have to scrape.

The yield stress is structural. As the melt cools, wax and hard fat crystallise into a continuous network of touching crystals with the liquid oil held in the spaces; the oleogels and wax networks page sets out how little wax this takes. The stress needed to break enough crystal contacts that the network stops spanning the sample is the yield stress.

What the measured numbers look like

Published rheology on finished balms is scarce. Wax oleogels, the skeleton of a balm without the butter, have been measured carefully, and so has petroleum jelly as a pharmaceutical base. The table gathers figures from four studies. They were not measured the same way, which the next section explains, but together they set the order of magnitude.

Three things stand out. A wax gel at its gelling minimum is extraordinarily weak: two or three pascals is about the pressure of three sheets of printer paper lying on a table, which is why a balm uses two to seven times the gelling minimum. The gelling minimum is not a measure of strength either: carnauba needed twice as much wax as beeswax to gel rice bran oil, 6 percent against 3, and still gave the weaker gel on both measures. And the way a gel was made changed its yield stress almost fortyfold.

No study we found gives a yield stress for a finished lip balm or salve in pascals with its method stated. Tsupko and colleagues measured two commercial lip balms at 25, 37 and 50 C in 2023 but reported the comparison as a ranking and in orders of magnitude; their results are in the temperature section.

Yield stress is not one number

Two figures for the same kind of product usually disagree because they were measured differently. The structure does not snap at one exact stress; it gives way progressively, and each method picks a different point on the way down.

The amplitude sweep: two points, eight times apart

The usual test is an oscillatory amplitude sweep: the sample sits between two plates, one of which twists back and forth at a fixed frequency, 1 Hz in Wang's study, while the strain is stepped up from 0.01 to 100 percent. At small strains the response is linear and nothing is damaged. Then the network starts to break.

Wang and colleagues took the yield stress as "the end of LVR, indicating the softening point of oleogels", where LVR is the linear viscoelastic region, and fixed it at the point where the storage modulus fell 5 percent below its plateau. They took a second number, the flow point, as "the crossing over of G′ and G″, indicating that gels underwent permanent deformation". For the 3 percent beeswax gel those two points were 2.29 and 18.55 Pa, a factor of about eight apart. For the carnauba gel they were 1.82 and 4.36 Pa, a factor of about 2.4. A report that says only "yield stress" can be out by nearly an order of magnitude.

The gap is itself informative. Beeswax gave way gradually over a wide stress range; carnauba went from softening to flowing in a narrow one. That matches what the balm texture science page describes: a beeswax balm softens forgivingly as you press, while carnauba makes a structure that gives way abruptly.

Stress ramps and flow curves

Simpler methods ramp the stress up until the sample moves, or fit a flow curve with the Herschel-Bulkley equation (stress equals a yield stress plus a consistency term times shear rate to a power), as Su and colleagues did for candelilla oleogels in 2022. A fitted yield stress depends on how low in shear rate the data go, so it is a model parameter rather than a direct measurement.

Large amplitude tests give higher numbers again

Gillece and colleagues at Ashland compared a standard stress sweep on skin creams with large amplitude oscillatory shear (LAOS), which analyses the full distorted signal. For a buttery cream, the large amplitude results "(394/293 Pa, respectively) suggest that the yield magnitudes from LAOS are three to four times higher than the standard stress sweep result (τ0 = 92 Pa)". The same cream gave three answers, from 92 to 394 Pa, depending on the test. They also noted that "wall slip is an integral part of the sensorial experience", which changes the reading on smooth plates.

Storage modulus and loss modulus in plain terms

An amplitude sweep reports two moduli. The storage modulus, G', is the part of the response that springs back, like a rubber block. The loss modulus, G'', is the part dissipated as flow, like stirring honey. With G' above G'' a material is predominantly solid; where they cross is the flow point in the table above.

Park and Song measured commercial petroleum jelly at 37 C, body temperature, at a strain of 0.1 percent over angular frequencies from 0.025 to 100 rad/s. They found that "the storage modulus is always greater than the loss modulus over an entire range of angular frequencies studied", with a phase angle "always lower than 45°", and fitted power laws that give G' of about 4,540 Pa and G'' of about 2,520 Pa at 1 rad/s. Their reading is about use: "it is undesirable for the product to flow down from the skin at an initial stage upon contact". A layer of petroleum jelly on warm skin is still a solid until it is rubbed.

A chemically cross-linked gel has a modulus that barely changes with frequency. Petroleum jelly's rose with a slope of 0.314, and the authors conclude that it "does not form a stable (strong) gel but exhibits a weak gel-like behavior": a network of contacts that can be broken, not chemical bonds, as in every wax and oil balm. Liquid mineral oil, its close chemical relative, has no such network, which is the root of the differences on mineral oil vs petroleum jelly.

Shear thinning: why a balm glides once it moves

Past the yield stress, a balm's apparent viscosity falls the faster it is worked. That is shear thinning, and it is why a product stiff in the tin spreads into a thin film. Park and Song's steady shear data for petroleum jelly at 37 C, from 0.025 to 200 per second, fit a power law with a flow index of 0.172, against 1 for a Newtonian liquid. That is very strong thinning.

The shear rate rises eight thousand fold down the table while the stress needed to keep the material moving rises less than fivefold, from 21 to 99 Pa. The effort of spreading barely increases however fast you rub, which is what makes a stiff ointment easy to apply. Summarising their earlier work, the authors attribute it to "the disruption of a crystalline network under the influence of mechanical shear deformation". The gentle oscillation test also gave a complex viscosity "much larger than the steady shear viscosity over a whole range", because steady shear destroys the structure that a 0.1 percent strain leaves intact: the material in the tin and on the skin are two states of one substance.

Shear thinning is also why drag and poor glide is a separate fault from hardness. A balm with a high yield stress can still glide once it yields, given enough liquid oil; drag comes from what happens after yielding.

Is a balm thixotropic? Mostly not

"Thixotropic" is used loosely for anything that thins when worked, but it means a material that thins under shear and then recovers its structure at rest. If a balm were thixotropic, a tin that had been dug into, or a batch stirred while it set, would regain its firmness after a rest.

Wax networks do not. Wang and colleagues sheared their gels gently for three minutes, hard at 10 per second for three, then gently for fifteen to see how much stiffness came back. "The 3BW and 6CRW samples showed very low recoveries of 1.89% and 0.88%", that is, the beeswax and carnauba gels regained under 2 percent and under 1 percent of their storage modulus. They note an earlier study in which a 3 percent candelilla gel in high oleic safflower oil recovered 1.19 percent, "attributed to the disintegration of the junction zones" of its crystal network. The ethylcellulose gel, a polymer network rather than a crystal one, recovered 68.36 percent under the same test.

So a wax and oil balm is shear thinning but, for practical purposes, not thixotropic. The loose usage persists because both properties start with thinning under a finger; the difference only shows afterwards. Broken crystal contacts stay broken at room temperature. The wax is still crystalline, but the connected scaffold is gone, and only melting and recrystallising rebuilds it.

Shear during setting: the factor of 38

Penagos and colleagues made 5 percent carnauba oleogels in sunflower oil two ways: at rest in the lab, and in a pilot-scale scraped surface heat exchanger, which shears the product continuously as it cools. "The CRW oleogel had a yield stress of 16.32 ± 2.08 Pa when prepared statically and a yield stress of 0.43 ± 0.04 when prepared on a pilot scale." Same wax, same oil, same concentration, both measured at 5 C, and the yield stress fell by a factor of about 38. The authors describe the sheared product as having "notably smaller crystal sizes, reduced rigidity".

That is why pour temperatures has you stir the jug only while the melt is still liquid, down to the pour window, and then leave the poured balm alone. Once the network has started to form, each stroke breaks structure that will not come back. Whipping is crystallisation under shear by design, so a whipped butter is held up differently, as the whipped body butter collapse page explains.

What a yield stress has to beat

Column figures below are order of magnitude arithmetic from a balm density of about 900 kilograms per cubic metre, not measurements.

Holding particles needs almost no yield stress once the network exists, which is why the lip scrub formula treats sedimentation as a filling fault: the risk is the minute before the network forms. A free-standing stick needs a structure in the hundreds of pascals, tens of times any gel at its gelling minimum, which is why a tube carries 10 to 20 percent wax against 7 to 12 in a tin on lip balm ratios. Oil escaping a network that is too coarse is a separate failure, covered in balm sweating. Some products are mostly a yield stress: a lip gloss balm has to flow through a wiper and stop two seconds later, and a hair pomade holds a style by resisting the bending stress of hair until a comb pushes it past yield.

Temperature: the yield stress fades long before the balm melts

As temperature rises, part of the wax dissolves back into the oil and the network loses contacts. It does not need to melt to fail.

Wang and colleagues heated their gels at 2 C per minute and recorded where G' and G'' crossed, the point at which the sample became more liquid than solid. The 3 percent beeswax gel crossed at 46.71 C and the 6 percent carnauba gel at 76.96 C. Beeswax itself melts at 62 to 65 C, so the dilute beeswax network gave way roughly 15 to 18 degrees below its wax's own melting range. That agrees with the 46 to 50 C calorimetric peaks for beeswax oleogels on the oleogel page.

Tsupko's commercial lip balms showed the same in finished products. Their calorimetric melting points were 57 C for the beeswax balm and 54 C for the plant wax balm, yet between 25 and 50 C, still below melting, "wax-based lip balms have lost their strength (in terms of G′LVR) by an order of 10". A balm in a car at 50 C looks solid and is about a tenth as stiff. That is the mechanism behind balm melting in the heat and behind tubes leaking in warm weather.

Skin temperature matters too. Gillece and colleagues found that for their buttery cream, "at 32°C, the stress at maximum shear strain rate drops by 302 Pa" compared with 25 C. A balm judged on a workbench is in a different state from the one it is in on a lip.

Brittle or plastic: why some balms crack

Su and colleagues, writing about candelilla wax oleogels, note that most oleogels "can match the hardness of the solid fats but lack plasticity", and they "undergo brittle fracture rather than plastic flow when a stress greater than the yield stress of the oleogels is applied". That is what a maker sees when a candelilla or carnauba heavy balm crumbles under a fingernail instead of smearing.

The narrow gap between carnauba's two amplitude sweep points, 1.82 and 4.36 Pa, against beeswax's 2.29 and 18.55, says the same thing: one structure goes from intact to broken quickly, the other yields progressively. High melting waxes grown slowly push a formula towards brittle, and in the extreme towards the grit described on gritty balm from hard wax.

Judging yield stress at the bench without a rheometer

None of these gives pascals, but each compares batches fairly if temperature and geometry stay fixed.

1. The inverted jar. Set 50 g in a small jar for 24 hours and turn it over. If it holds, it has a yield stress; this is the gelling test used in the oleogel studies.
2. The slump test. Set the balm in a ring on a tile, lift the ring, and measure the sag after an hour at a fixed warm temperature such as 35 C. Keep the height fixed, since a taller cylinder loads its base harder.
3. The spatula trail. When a trail drawn through the cooling melt stops closing, the yield stress has arrived. Add particles before that point and stop stirring after it.
4. The stick test. Extend a stick to a fixed height at a fixed warm temperature and time how long it stands.

Judge on day three rather than day one, for the reasons in cooling rate and crystallisation, and see sensory testing on why durometer readings on a fat network are only an index.

What this means for a formula

Raising yield stress buys heat resistance, a stick that stands and a structure that holds pigment. It costs pick-up and, past a point, glide. More wax raises it steeply, undisturbed cooling keeps it, and shear during setting throws most of it away. The face balm keeps its network weak so light pressure breaks it; a tube wants the opposite. Balm changed after shipping covers a network that partly melted and reformed, and balm too hard or too soft adjusts in single points of wax.

Where these numbers stop applying

The yield stresses here come from wax oleogels in rice bran and sunflower oil near their gelling minimum, at 25 C or, for the carnauba comparison, 5 C, and the moduli from one petroleum jelly at 37 C. None was measured on a balm with shea, cocoa or mango butter, whose fat crystals add a second network, or on castor oil or jojoba. A balm at 8 to 15 percent wax is much stronger than a 3 percent gel, but no study we found gives that figure with its method, so this page does not state one. The cream figures are emulsions, included only to show how much the method moves the answer, and the recovery figures come from one protocol with fifteen minutes of rest.

What the numbers settle is the shape of the answer. A balm holds its shape because a crystal network gives it a yield stress. That stress is small, falls steeply with warmth, depends on how the balm was cooled, and once broken at room temperature does not come back.

Reviewed and updated 19 October 2026. Spotted an error? Tell us and we will fix and log it.