A ripple or a chunk has to be reformulated before it goes in

Everything in the tub reaches the same temperature

It is an obvious point that is nonetheless routinely forgotten. Whatever goes into an ice cream sits at minus eighteen or thereabouts along with everything else, and almost nothing that is pleasant at room temperature is still pleasant there. Caramel sauce becomes a hard seam. Chocolate chips become small stones with a waxy finish. A brownie chunk becomes a cold, dense brick with ice in it. Each of these is a formulation failure rather than an accident.

The commercial trade treats inclusions as separate products with their own specifications, developed to behave correctly at freezer temperature rather than in a bowl. That is the difference between a ripple that stays soft in a shop tub and one made at home from the same sauce that won’t yield to a spoon. The sauce wasn’t wrong. It was simply designed for a different temperature.

A ripple stays soft because its freezing point was moved

A sauce that remains plastic at freezer temperature has to contain enough dissolved material that most of its water cannot freeze. In practice that means a very high sugar content, usually with a substantial proportion of glucose syrup or invert sugar, because what matters is the number of dissolved molecules rather than their mass. A little salt does disproportionate work for the same reason, since salt dissociates into two particles and is very effective per gram.

Fat and alcohol are the other two levers. A fat-based variegate contains little water to freeze in the first place, which is why chocolate-flavoured ribbons are often more fat than sauce. A splash of spirit lowers the freezing point of the water phase directly. What none of these can do is make the ribbon taste the same as it did warm, because cold suppresses aroma and a formulation loaded with glucose reads as less sweet than the sugar in it suggests.

Water moves between the inclusion and the ice cream

Anything with a different water activity from the surrounding ice cream will exchange moisture with it, slowly, for as long as they sit together. A dry biscuit or wafer draws water out of the ice cream at its boundary and softens, which is why a cookie piece that was crisp on the day it went in is leathery a fortnight later. It is the same migration that ruins a wafer cone left too long in a filled state.

It runs the other way too. A wet inclusion such as a fruit piece or an under-formulated brownie gives up water into its surroundings and develops ice crystals of its own, which is the gritty, frosty texture people find in fruit chunks. The commercial answer is to bring the water activity of the inclusion into line with the base, usually by sweetening it heavily or by coating it in fat, so there is no gradient for water to follow.

Fat-based inclusions win, and then have their own problem

Chocolate is the most common inclusion precisely because it contains almost no water, so nothing in it freezes and nothing migrates. The difficulty is that cocoa butter is very hard at minus eighteen, and a chip of it in a cold mouth takes far too long to melt, which is why it reads as waxy and why the flavour arrives late or not at all. The mouth has already been chilled by the ice cream, so the melting is slower than it would otherwise be.

Manufacturers deal with this by using a coating fat that melts lower than cocoa butter, and by making the pieces thin — flakes and shavings rather than chunks. A thin flake has enough surface area to melt in the time a mouthful lasts. It is the same reasoning that makes a chocolate shell on a bar work best when it is thin, and it applies to anything fatty that has to melt in a mouth that is already cold.

Timing, placement and what to expect at home

Inclusions go in after the freezing, not before, for straightforward mechanical reasons. Solid pieces would damage the dasher and the scraper blades, and a ripple beaten into the mix disperses into a uniform colour instead of remaining a ribbon. In a factory both are fed into the product as it leaves the continuous freezer, which is why a shop ripple has clean edges: it was placed into a semi-solid stream that could not mix it in.

A domestic version follows the same principle. Layer the sauce as you fill the container rather than stirring it through, fold pieces in when the base is thick enough to hold them in suspension, and expect a caramel that was soft in the pan to set firm unless it has been formulated with glucose and salt to stay soft. Warming the tub a little before serving disguises a hard ribbon, which works, and it is a workaround rather than a solution.