The cooking and cooling behaviour of commercially available vanilla puddings

05 November 202014 min reading

Markus Löns Business Development Manager Food Brabender GmbH & Co. KG

“Thanks to the performance of the ViscoQuick, starch parameters can now be measured within 10 minutes. The latest innovations built into the device allow heating rates of up to 20 °C/min. and cooling rates of up to 15 °C/min. without the need to connect a heating/cooling thermostat.”

INTRODUCTION AND OBJECTIVES The consumption of vanilla pudding is very popular among the population of many countries and regions. This is shown by the large number of suppliers who offer a broad range of variants, divided into cold-produced products and those that need to be cooked. No wonder that even medium-sized discounters and supermarkets may have listed 2-5 different ready-to-cook vanilla pudding mixes suppliers.

The main ingredient of the mixes is starch, mainly made from maize. Furthermore, they contain table salt, aroma and colourings, partly real bourbon vanilla is added. Customers experience product quality parameters as taste, colour and consistency only with their senses. As this is very subjective, this study concentrates exclusively on one measurable quality parameter: viscosity. In addition to the already mentioned taste and colour, this is one of the main sensory properties and thus responsible for the popularity of vanilla pudding.

This study should examine the extent to which commercially available puddings differ in their consistency.

MATERIALS AND METHODS A total of 10 different commercial pudding mixes were purchased in supermarkets. In addition, a product manufactured for industrial use in the bakery industry was tested but was not included in the evaluation because of its gelatinisation and cooling properties, which deviated greatly from the other samples.

The main focus was on the final viscosity of each stirred and boiled pudding mix at 50 °C (standard temperature/10 min. running time), 20 °C (room temperature/20 min.) and 6 °C (refrigerator temperature/25 min.). The qualitative determination of the gelatinisation properties of starch is mainly carried out up to 50 °C in quality assurance and product development, but the pudding itself is later usually consumed at room or refrigerator temperature. In addition, the influence of milk and the addition of sugar were considered.

Fig. 1: Brabender ViscoQuick®

The Brabender ViscoQuick used for the test series has an integrated heating/cooling system that operates without an additional heating or cooling thermostat. Owing to the high heating/cooling rates, the individual measurements could be performed in a relatively short time (for measuring times see above). Due to the evaluation unit directly built into the device, necessary changes of the test parameters and correlations could be made without an additional PC. The necessary cooling profiles provided reproducible results. As a true stand-alone and compact unit without a cooling thermostat or separate computer, the ViscoQuick adapts perfectly to often meet tight space conditions in many laboratories.

As already mentioned above, cooling rates specially adapted to the respective final temperature were programmed in the software to achieve the desired final temperatures in a reproducible manner. The evaluation was performed in Brabender units (BU), which have become established worldwide for the description of viscosities in starch gelatinization.

During a ViscoQuick analysis, a characteristic curve is created during heating, maintaining the temperature and subsequent cooling. The changing consistency of the suspension during the test duration is recorded and evaluated as a force-time diagram. Special attention is usually paid to the following points, which are determined, for example, during analysis with the Brabender Viscograph or


  • Start of gelatinization (A)
  • Maximum viscosity (gelatinization maximum / B)
  • Minimum viscosity (C)
  • Viscosity at the end of the cooling phase (D)
  • Viscosity changes during maintaining the temperature (B-C)
  • Viscosity changes during cooling time (2-3)
Tab. 1: Overview of methods Tab. 2: Overview of test encoding Test series 1: Analysis at different temperature profiles The existing procedures and standard regulations (ICC/AACCI/ISO1)) describe temperature ramps of 30 °C (start temperature) and 50 °C (end of the cooling phase) as methods. The heating and cooling ramps of these methods are sometimes relatively low, which means that the measurements are very precise, but can take up to >1.5 hours. The consistency at 20°C and 6 °C (assumed room/fridge temperature) was also investigated. Between the heating and cooling phase, different holding times at 93 °C are also specified depending on the laboratory equipment. Distilled water was used as liquid. Fig2. Example: ViscoQuick curve of corn starch Thanks to the performance of the ViscoQuick, these typical starch parameters can now be measured within 10 minutes. The latest innovations built into the device allow heating rates of up to 20 °C/min. and cooling rates of up to 15 °C/min. without the need to connect a heating/cooling thermostat.

Thus, the gelatinization properties can be measured very precisely and reproducibly within this short period of time (e.g. 10 min. for corn starch) and displayed graphically (see Fig.1).

The examination of the different samples should be carried out under conditions nearest to the end user. This means that the manufacturer's instructions and recommendations should be observed, and the samples should be processed as it is typical for households. In the first step, the process parameters listed in Tab. 2 were set on the ViscoQuick. The recommended weight for the ViscoQuick for the measurement of corn starch is 105 g water and 10 g starch. The amount of pudding mix weighed in per test was made according to the manufacturer's specifications and was converted accordingly from 500 ml liquid to 105 ml.

The pudding mix to be examined were each dissolved in the corresponding liquid and placed into the measuring cup of the ViscoQuick. The premixing of the powder with sugar, as recommended by the manufacturer, was omitted to evaluate only the pure mix in the first step.

The following parameters served as the basis for the first test setup:

-Starting temperature: 30 °C -Holding temperature: 93 °C -Final temperature: 50 °C -Speed: 250 min-1 -Measuring range: 150 cmg -Test time: 10 min.

The heating rate of 20 °C per minute was much higher than with the methods known and used so far. This considers the fact that modern kitchen stoves (e.g. induction) but also industrial cooking systems can bring liquids to the boiling temperature in much less than 10 minutes. The subsequent holding time of 3 min. is based on the short cooking time usually recommended by the manufacturers.

In addition to the standard parameters already mentioned, test runs had been primarily oriented towards practical application and the state-of-the-art technical possibilities. This refers to the cooling phase with lower final temperatures (20 °C / 6 °C) for which special cooling ramps were developed. These reflect the temperatures at which pudding is usually consumed (20 °C room temperature, 6 °C fridge temperature). Significant difference in consistency was expected compared to each other and to the temperature of a recently cooked pudding (50 °C).

Test series 2: Analysis under variation of the liquid In the laboratory, experiments are usually performed with distilled water. Besides the easier handling of water this has ethical reasons, as milk is a high-quality foodstuff. However, to be as close as possible to cooking pudding as a final product, milk was used for a selected number of tests and samples in this study. It was initially assumed that all pudding mixes behave similarly in terms of cooking and cooling when water is substituted by milk.

Test series 3: Analysis under variation of sugar addition Since the final product usually contains sugar, this influence was also investigated. Again, it was assumed that the influence of sugar would be similar for all pudding mixes, so this was only performed on some samples.

In addition, the time of addition of pudding mixes and sugar was varied during the measurements. In common laboratory measurements, all materials (liquid and solids) required for testing are added at the beginning. Since in the preparation of pudding, the powder, which is mixed with sugar and then dissolved with little liquid, is only stirred in when the main part of the milk is boiling, this influence on the viscosity was also considered.

Fig. 3: Test scheme

Two test arrangements were considered, whereby the sugar was always mixed with the pudding mix in a dry state (manufacturer's information), then stirred with cold liquid and transferred to the measuring cup. In variant one, all components (powder, sugar, milk) were already contained in the measuring cup when the experiment was started. In contrast, 86 % of the milk was available at the beginning of variant two, which was heated at 20 °C/min. according to the specifications. After exactly 3 minutes, shortly after reaching the 93 °C holding temperature, the solids mixed with the remaining milk (14 %) were added during the running process through an opening in the lid. During dosing, the speed of the paddle was automatically reduced to the programmed speed of 50 rpm, only to return to the original value after 1 minute.

In this study, however, the main focus was only on the different viscosities and the viscosity changes at the end point of the respective cooling to 50/20/6 °C, as only these consistencies - especially those at lower temperatures (20/6 °C) - are of interest from the consumer's point of view. For this purpose, two special profiles were developed in advance, which automatically enabled reproducible cooling to 20 °C and 6 °C, respectively. The programmed cooling rates were between 15 °C/min. and 1.5 °C min., depending on the selected final temperature.

RESULTS OF THE STUDY Test series 1: Analysis at different temperature profiles As already described, in the first test series all samples were measured with the parameters listed under ST-50 (Tab. 1). Due to the high heating and cooling rates possible with the ViscoQuick, the measuring time was reduced to 10 minutes compared to the frequently used methods. The sample mix for industrial use showed viscosities that were in some cases more than twice as high as the "supermarket samples". The cause was identified as the much higher initial weight (approx. 35 %) specified by the manufacturer. As a result, the weight was reduced to the 7.77 g specified for most products, which resulted in a much better comparability. However, the data were not considered in the further course of the series, as all other mixes were used according to manufacturers' recommendations.

Fig. 4: Method ST-50 °C – Comparison of the 10 samplesThe course of all curves is very similar and shows the usual behaviour of corn starch, which was listed in most ingredient lists. In the other samples only starch was listed, so that no exact statement can be made. A total of 8 curves are only +/- 20 BU apart at measuring point B (maximum viscosity), 2 samples (M-03/M-10) deviate more than average, M-10 (the industrial product) very strongly.

When looking at the consumer-relevant end temperatures of 20 °C/6 °C, it is immediately noticeable that the industrial product sample continues to show the greatest deviations from the total, which even increases as the temperature drops. It can be assumed that the manufacturer deliberately relies on a consistency that differs significantly from the average as a sales argument. In the further evaluation of the results the industrial product sample is not included due to the significant differences.

Fig. 5: Method RT-20 °C – Comparison of 10 samples

When considering all values for the final viscosity across temperature ranges, the majority of all cooked and cooled puddings behave very similarly. The differences, some of which are only minor, can be recorded and described by a sensitive and reproducible measuring laboratory device. The only exception in terms of gelatinization and cooling properties is the product manufactured for industrial use. Whether the differences in the other samples are recognized by the consumer and considered in the purchase selection, however, would have to be determined by professionally conducted sensory tastings. Since the sensory perception of the consistency is very subjective, it certainly will pay off to apply an objective measuring method such as the ViscoQuick measurement.

Fig. 6: Method FT-6 °C - Comparison of 10 samples

Test series 2: Analysis under variation of the liquid In contrast to the previous test arrangements, only three randomly selected mixes were used for test series 2 and 3. The solid content of milk was not considered when substituting distilled water with milk. As a result, it could have been assumed that the lower proportion of liquid available would lead to an increase in viscosity.

Fig. 7: Method FT-6°C – Comparison of 3 samples with distilled water and milk

Yet the corresponding curves of all 3 samples show a contrary image. Throughout the entire duration of the experiment, the consistencies of the "milk experiments" are below those of the measurements performed with water. Moreover, it could be observed that gelatinization starts much more abruptly when using milk and the gelatinization curve is also steeper than with water. Before reaching the maximum, the slope flattens out, creating a kind of plateau. With water, however, reaching the maximum is characterized by a relatively clear peak.

The reasons for these different results are probably to be found in the main ingredients (carbohydrates, proteins, fats) of milk, however, their individual effects were not analyzed in the study. Nevertheless, the influencing factor milk should be considered in product development and recipe recommendations for the customer.

Test series 3: Analysis under variation of sugar addition The 3rd research question of the topic "Viscosity measurements of vanilla pudding mixes" deals with the influences which result from the addition of sugar. This takes into account not only the change that occurs when sugar is added. The timing of the addition has also been varied.

While in laboratories all components are often already in the measuring cup of a viscometer at the start of the measurement, in practice and according to the manufacturer's instructions, pudding powder and sugar are only added when the liquid is boiling. For this reason, it was obvious to adapt the test arrangement accordingly to show possible differences.

Fig. 8: Method FT-6°C – Comparison addition of sugar M-03

Fig. 8 shows the curves of the test arrangements performed with milk: "without sugar/blue", "sugar addition directly/green" and "sugar addition after 3. min/red".

The gelatinization of the two experiments "without sugar/sugar addition direct" is almost identical for the first 10 min. Only after this time the consistency of the two samples will differ. When the final temperature of 6 °C is reached the measured value is 160 BU higher than the sugar-free pudding.

Due to the later addition of pudding powder and sugar to the already boiling milk in the third test arrangement, there is a spontaneous gelatinisation of the starch and a time shift of the entire curve. While this apparently has no influence on the gelatinisation maximum (all three tests here have a delta of 20 BU), the gelatinisation minimum is 40 or 50 BU below the other patterns.

As in previous discussions of the results, the final point is consistency at assumed refrigerator temperature. The sample "sugar addition after 3 minutes" was of particular interest because here the pudding powder was also added after 3 minutes and therefore the cooking test was carried out exactly according to the manufacturer's instructions. The final consistency is almost 100 BU below the variant "sugar addition directly" and slightly higher than the variant with milk only. The curve of the two samples with sugar could indicate that the viscosities will equalise during longer storage in the refrigerator and that the time of its addition has no greater influence.

CONCLUSION The study found that one pudding mix was very different from all other samples. One can only speculate about the reasons. All other patterns showed predominantly similar or equal viscosities. However, the change in the ranking (highest - lowest) of the individual pudding samples when the final temperature was changed is remarkable. There were samples that became firmer as they cooled down, others softer. Since for the consumer not only colour and taste but also consistency can be a purchase criterion, measurements up to 20 °C/6 °C as well as the effects of milk and sugar on the consistency of a vanilla pudding mix ought to be included in the focus of future product developments and quality controls.

Literature references: -ICC: International Association for Cereal Science and Technology, Vienna, Austria -AACCI: Cereals & Grains Association, St. Paul, USA -ISO: International Organization for Standardization, Geneva, Switzerland

Articles in Article Category