1-Introduction

Waffles are flat fine bakery products, which are baked between two hot iron plates (waffle iron) or within a specific waffle machine. There exist many different varieties of this baked good, which are different in structure and texture, e.g. flat, soft or cross waffles.

Flat waffles are a dry pastry with a porous structure, which is mainly due to evaporating water. Waffles are used for different purposes (rolled or casted ice-cream cones), in layers incorporated in sweets (flat waffle, waffle sticks) or also as waffle sheathing for a chocolate filling (hollow waffle).

Waffle recipes are quite simple and do not need many ingredients, whereas besides water, flour is the main component. Also milk, sugar, salt, baking improver and/or fat may be added. Most waffle-masses are fluid and pumpable. The flour used should be weak, with low protein and wet gluten content, in order to obtain a complete structure of the waffle sheet. The water absorption rate of waffle flours should be low, in order to obtain already with a small water addition a well pumpable dough, which allows for a wide application of water as baking agent (water steam proving).

Within the industrial production the waffle mass is pumped through a tubing system. During that process mechanical energy is imposed into the mass, which leads to a viscosity change if gluten protein is present. The so called gluten aggregation can influence the dosing, and thus leads to the fact that waffle forms are not filled properly, or that the baked product shows unwanted effects like cracks. It is also possible that clots are formed, which lead to a blockage of the dosing dies.

A quick and reliable method for characterizing the most main flour parameters (wet gluten, protein content, and water absorption) is of great importance for the producers of waffle products. The evaluation of flours with a Brabender GlutoPeak, and the GlutoPeak evaluation ”Rapid Flour Check” provide such a method, which implements a quick characterization of flours by mixing a small quantity of flour and water within a controlled environment, in order to analyse and monitor the gluten development within that dough. As an onward development there exists now the “Low Protein Check” (LPC)-method, which is specifically designed for properties of low protein flours.

Within the here mentioned project these GlutoPeak methods were tested, implementing 20 different flour samples. The aim of this test was to check, if these methods give reliable results and offer the same possibilities to obtain information in regards to mixing and baking of waffle products within a short time range, in comparison to standard routine analysis. The flour samples where provided by mills and food manufacturers. The project partners were the companies Bühler AG from Uzwil, Switzerland, as well as the Brabender GmbH & Co. KG, Duisburg, Germany. The baking tests, the planning, realization, and evaluation where carried out at the Bühler/Haas food technical centre, Leobendorf (Austria). There was also the GlutoPeak instrument stationed, which was provided by the company Brabender.

Pic. 1: Brabender GlutoPeak

2-Material and Methods

For the project work, 20 different wheat flours were available, which stand for typical waffle flours or cookie flours. As a baking test, standard ice cones were produced and tested. All flour samples were tested with the GlutoPeak and standard laboratory analyses (in accordance to ICC) prior to the waffle dough preparation and baked to ice cones. With the exception of one sample, all other samples fulfilled the basic requirements of a waffle dough (pls. see table 1).

2.1 Laboratory Analysis Using the Brabender GlutoPeak

The first analysis was carried out with the GlutoPeak (pic. 1) in accordance with the standard method “Rapid Flour Check” and the available evaluation algorithm. 

After defining the moisture content of the flour sample, approx. 9 g water (depending on the water content of the corresponding sample) was filled into the measuring bowl and inserted into the GlutoPeak. By using a connected thermostat, the measuring opening in the housing was heated up to 36 °C, in order to provide for equal test conditions, i.e. homogenous temperature distribution in the measuring bowl. After 2 minutes, in which an even temperature distribution within the water filled measuring bowl is given, the corresponding flour quantity (depending on the water content of the specific flour) was added.

                                                                    Pic. 2: GlutoPeak diagram of a waffle flour

By pressing down the measuring head, at which at its lower side the measuring paddle is mounted, the GlutoPeak is closed and the test starts automatically. Before the test, the corresponding test parameters were stored into the MetaBridge instrument software. The motor, built-in the measuring head, drives the paddle and thus generates the rotation.

Depending on the flour property, the measurement takes 2 – 5 minutes at 2750 min-1. Due to the high energy input, a gluten network develops depending on time and corresponding to each gluten quality and quantity, showing a peak (gluten peak) as highest value (torque). Afterwards the network is destroyed due to the mechanical energy input, the curve declines and the measurement is ended. A typical curve diagram is depicted in pic. 2.

After the test ends, the highest torque is listed as peak (BEM/max. torque), and the necessary time (PMT/Peak Maximum Time) is evaluated. Additional measuring points are the torque every 15 seconds before and after the maximum (AM/PM). The Rapid Flour Check analysis also provides for correlations towards other measuring procedures. These are protein content, wet protein content, water absorption, and the W-value (Alveograph process), whereby the latter mentioned was not considered.

After a short cleaning, the next measurement was done. Due to the rapid test procedure and easy cleaning, up to 8 measurements per hour were possible.

The “Low Protein Check” (LPC) method, which was developed during the project, was used in a second step. Hereby approximately 12 g water per 11 g flour (depending on the water content of each flour) are weighed. This method is done with reduced speed of 2500 min-1 and a temperature of 35°C. The earlier mentioned evaluation is not changed by using this method.

All tests were carried out as triple measurements and the results were collected and evaluated with the included software (MetaBridge).

2.2 Laboratory Analysis according to ICC-Standard Methods

Most flour samples available were provided with the appropriate specifications. In most cases, these contained an average range within which the corresponding flour parameter should result. However, in some cases it was not detectable if standard analysis methods were applied for testing.

In order to get reliable information about the available samples, all samples were checked by an accredited laboratory in Austria in accordance with the standard ICC methods.

All results of the flour analysis in accordance with ICC-Methods are specified in table 2.

2.3 Producing a Waffle Dough

                                      Pic. 3: Typical carrier for industrial „Waffle-Cone-Baking Oven“ as laboratory instrument

The recipe used for making the waffle mass is depicted in table 3. In fixed order, water and sodium bicarbonate were given into an IKA laboratory stirrer equipped with a R1300 blade mixer, and mixed with moderate speed until a complete homogenization was reached.

Following the next process steps, flour was inserted stepwise and the mixer speed was increased. After adding all flour, a further mixing of the waffle mass is done until again a complete homogenization was reached. Then followed by the next step the pre-prepared 1:1 mixture of sunflower oil and soy-lecithin, was added into the mass und mixed for another 60 seconds.

Immediately after mixing, the viscosity was measured using a gravity feed cup of 100 ml and an 8 mm die (diameter) (according to ASTM D 333). In case that the measured viscosity showed a deviation towards the defined viscosity of 20 seconds (+/- 1 sec.) the water amount of the recipe was adjusted and a new waffle mass was prepared.

By adjusting the viscosity of each dough to 20 seconds (+/- 1 sec.) the mechanical behaviour during baking can be considered similar for all flours, thus assuming that the product properties can only depend on the flour property itself. The necessary water/flour-relation is an important parameter for developing a specific viscosity and is recorded for all test masses.

2.4 Baking Waffle (Casted Cones)

Baking of cones, was carried out with an electrical heated laboratory instrument type LB-SWAK_STAK, equipped with an industrial baking pan and a corresponding core plate (pic. 3).

The pan was heated up to 180°C, whereas the cores were heated up to 190°C. Using a spraying device, 7 ml waffle mass was inserted and the baking time was 80 seconds. Prior to closing the carrier for the final baking process, it was closed and opened twice (so called breathing: 2 sec. open, 2 sec. closed) in order to let steam evaporate out of the system.

After baking, the cones were removed from the open baking pan and left for cooling at room temperature for approximately 10 minutes. The cooled cones were packed in PE wrapping, hot sealed, and stored at 8°C until the texture analysis took place.

2.5 Texture Analysis of the waffles

The texture analysis was carried out with a TA.XT.plus Texture Analyser (Stable Micro System) using an ice cream cone pan (A/ICC). This tool measures the force, which is necessary to create a break, when a ball like item of a certain diameter and weight is pressed into the cone.

The most important texture parameters are the distance the ball moves, as well as the force which is necessary to break the cone. In order to carry out the texture analysis for every flour, 10 visually inspected cones were tested. The obtained results are summarized and evaluated afterwards.

3-Results and Discussion

3.1 Comparison GlutoPeak “Rapid Flour Check” with Standard Analysis

The GlutoPeak method “Rapid Flour Check” was developed for flours destined for baked goods, which require medium till high gluten level quality behaviours. Although flours for waffle production do not require high standards in regards to protein content, the method should be verified for its applicability. 

Comparing the results of the “Rapid Flour Check” with the results obtained with the standard analysis, in principle the correlation of all relevant parameters (protein content, wet gluten, and water absorption) is good (pic. 4 – 6), however showing relatively high deviations.

Brabender states, that the correlated protein content is varying by +/- 0.8 within a range of 10.5 to 15 %. Testing the flour samples, similar results are obtained (apart from W700 having a high protein content). The wet gluten numbers should range within +/- 2.3 in comparison to the standard analysis, which was nearly reached for the flour samples tested.

Pic. 5: Correlation of wet gluten measured with    GlutoPeak „Rapid Flour Check“ and routine analysis

In both cases the GlutoPeak method underestimated the results of the standard measurements, and was somewhat lower in protein content by around 3.5 %, and 11 % by wet gluten content. 

The GlutoPeak method, however, overestimated the water absorption by around 2 % (see pic. 6). Brabender specifies a deviation in water absorption of +/- 2.8, which the Bühler results confirm nearly. It is clearly recognisable that the deviation is increasing with lower WA values (waffle flour of a WA < 55 %). This is a disadvantage of the “Rapid Flour Check” method, as it is designed for flours with higher protein quantities and water absorption.

Abb. 6: Correlation of water absorption measured with GlutoPeak „Rapid Flour Check“ and routine analysis

In order to maximise the potential of the GlutoPeak results and optimise the significance in regards to waffle flours, Brabender included the newly developed “Low-Protein“ method in the course of the test series. The already obtained data were re-evaluated and compared afterwards with the results of the ICC-Method. The results obtained showed a good correlation between both methods, which is commented later in the course of this study.

3.2 Parameter for Mixing and Baking of Waffle Masses Adapted on Basic GlutoPeak Data 

In order to verify a correlation between required water quantities - respectively product weight on one side, and flour parameters on the other side - the data of the standard analysis and GlutoPeak analysis are dealt with separately. Primarily it presents that a correlation exists towards the standard laboratory data. Following it is tested if the correlation with standard laboratory data is also observed with the GlutoPeak data.

Standard Analysis

The basic correlation of the required water amount (for a determined dough viscosity) and the water absorption of the flour, is well known within the waffle production and can also be illustrated for similar flours (see pic. 6).

Flours with a higher water absorption need more water for reaching a defined dough viscosity. This can be a disadvantage during waffle baking, as the baked waffle normally has a water content of 1 – 2 %. Thus the complete water included in the waffle mass, has to evaporate during baking to reach that residual. In the baking process this evaporating water, acts as a leavening agent thus leading to very brittle or even incomplete products (inaccurate distribution of the mass within the baking pan). A water reduction, however, is not always possible when those flours are involved, as the thereof resulting increase in viscosity is influencing the dough- and product properties negatively. A decisive factor is the water demand of the flour.

An easy detectable effect in regards to the water requirement is the product density, and the specific weight. The higher the water amount, respectively the water demand, the lower the density of the waffle product (i.e. product weight). A high water absorption leads to products with a low density, as qualitatively depicted in pic. 8. Besides the flour properties, parameters like baking temperature influence (especially the core temperature can strongly vary during batch baking) the dosing volume etc. and density of waffle products.

Considering during baking these additional influencing factors, it implies that the correlation of the product weight with the water absorption is less significant than the correlation with the water demand in dough preparation.  

Pic. 9: Water demand of the dough correlating to the water absorption GlutoPeak

More important this applies to the batch baking of waffle sheets on laboratory scale, where significant differences in regards to waffle density may occur.

GlutoPeak-Data

The correlation of the water absorption obtained with the GlutoPeak method “Rapid Flour Check” and the water demand results in less significant values in comparison with standard analysis (see pic. 9).

Sample W700 (very high protein- and wet gluten content) is clearly identifiable within the sample points. If only the waffle flours are considered, there is no clear trend visible. The same applies to the product density (Pic. 10). Therefore, the GlutoPeak parameter WA is less significant compared to the water absorption, measured by standard analysis.

Pic. 10: Waffle weight correlating to the water absorption GlutoPeak

A better correlation is achieved, when the parameter G (wet gluten content) is selected instead of the parameter WA (Pic. 11 and 12). 

3.3 Predicting Product Properties  on Basis of GlutoPeak Data 

Predicting waffle properties from flour data is difficult, as not only the dough is influencing the product, but also the temperature/ temperature deviations in the oven, as well as dosing (quantity, speed). The data acquired from the texture analysis of the waffle product is already exhibiting a significant static variance and does not correlate well with the flour properties.

Considering the test set-up (see Pic. 13) the essential properties in regards to the product stability, are the distance the ball is pressed into the cone (until the first crack appears), as well as the corresponding force maximum for that crack.

Both parameters are severely influenced by the waffle structure near the steam vent of the baking pan. Flours of low water absorption show generally also at that point homogenous, dense structures (see pic. 14).

Stronger doughs tend to withstand the steam flow, if they have higher water absorption properties, and present streamline structures close to the steam vents (pic. 15). The stronger the flour, the more intensive these structures are, which lead to areas with less density and corresponding lower stability.

As mentioned earlier, the influence of flour is only a partial aspect, which is influencing the stability of the waffle material. Therefore, a direct correlation, even with the data obtained with standard analysis is difficult (pic. 16). Visibly flours with high water absorption tend to form low stability products. As an example, the sample W700 with its high water absorption of > 60 % results in less stable products, compared to the other flours.

However, also flours with lower water absorption can result in less stable products. This principle process is also exhibited, if correlated between the designated primary peak value Fmax with the GlutoPeak parameter G (see pic. 17).

4-Modification GlutoPeak-Method for Low Protein Flours

As the test results of the Rapid Flour Check (RFC) method did not produce an optimum information, the test evaluation was carried out with the newly developed Low Protein Check Method.

Pic. 15: Steam flow vent flour M04 (left) / Steam flow vent flour W700 (right)

By using this modified method, flour data of weak flours with satisfactory accuracy – compared to ICC methods, can be assessed with a rapid method (see pic. 18 – 20).

Contrary to the RFC method the LPC method is verified to flours with low protein (<11 %) and wet gluten content (< 25 %). 

The basic predictions in regards to product properties (e.g. stability) will lead to similar conclusions (see pic. 23) as this is already possible with data obtained relating to ICC-Methods. This leads to the conclusion that the “Low-Protein”-Method of the GlutoPeak is a valuable alternative to the ICC-Methods.

As already mentioned the prediction of product properties, in regards to weight or even more evident the mechanical properties, obtained only from flour data is rather inconsistent. These properties are not only defined by the water absorption of the flour used, but by far more other values.

During the GlutoPeak tests there were two flour samples which showed the phenomenon, that no peak developed. The necessary aggregation of gluten forming proteins obviously did not take place, and lead to the fact that no relevant change in torque is measured (see pic. 24/25). This test series was carried out as triple tests in order to rule out any application mistakes. Due to the curve progression and the resulting missing values, it was not possible to include these flours into the evaluation.

As a reason for the missing peak, it is assumed that the gluten building flour protein contained, was too low to aggregate for influencing the torque. There could also be the reason of a protein damage due to a denaturisation caused by too high temperatures during the milling process.

5-Conclusion

Every prediction could be reproduced in principle. This is valid for the standard laboratory analysis as well as for the results obtained with the GlutoPeak measurements. However, the results of the Rapid Flour Check method showed only a good correlation to a certain degree with the ICC laboratory data. Especially the water absorption – a crucial factor for characterisation of waffle flours – was insufficiently displayed.

Pic. 20: Correlation of water content measured by GlutoPeak Low Protein Check and routine analysis

From 20 tested flours, 19 samples were defined as suitable raw material for waffle production. Only one flour sample was outside the specification shown in Tab. 1. The selected random sample is considered as homogenous. 

Predicting quality trends of baked waffles is independently of the flour analysis used, very difficult, as a number of further influencing factors are affecting the product properties.

Pic. 21: Comparison of ICC- and Low Protein GP-Method relating to the water demand prognosis.

This is particularly valid, when not only the raw material flour, but also the baking parameter have a significant influence on the measuring data (e.g. texture of waffle cones). Laboratory data of the baking results correlated better, when acquired via ICC-Methods in comparison to the RFC-GlutoPeak evaluation statistical wise. This is due as the RFC-Method was developed rather for gluten strong flours. 

Pic. 22: Comparison of ICC and Low Protein GP-Method relating to the product weight prognosis.

Because of that, the flours were also tested and evaluated with the newly developed Low Protein Check method, which at the beginning of the project, however, was not yet available. The corresponding optimized LPC-GlutoPeak method was modified by including the evaluation algorithm towards the ICC-standard methods, and adapting the necessary parameter. Thereby equivalent results were generated in comparison with the ICC- laboratory data.

Pic. 23:  Comparison between ICC- and Low Protein GP Method for predicting the product stability

With the exception of these two flours, which did not show any peak when applying the LPC method and therefore showing no results, the tests carried out with the GlutoPeak are suitable to describe the flour quality. Due to its flexibility, easy handling, small sample quantities, and short measuring times a quick characterization of waffle flours is provided. The stored correlations, like wet gluten content, showed a good reference value with the obtained results, by applying the usual ICC Methods.

Based on the flour characterization provided with the LPC –Method, a quick test of much smaller flour samples for its standard suitability as waffle flour is as well possible, as a quick comparison of different flour batches. Furthermore, the measuring results may be used to make a quick estimation in regards to the necessary water demand, leading to a possible change of recipe or energy consumption. Predicting the product properties is just going by flour analysis, independently from the measuring method, not possible. 

The company Bühler AG, founded in 1860, is worldwide market leader for mills and milling systems, and equally throughout the food- and feed industry. The Bühler Food Equipment section covers the demand of customers in the waffle industry. Since 2018 the waffle innovation centre in Leobendorf, Austria, is developing specific custom made production equipment and carries out corresponding pilot tests on laboratory scale.

For 95 years Brabender produces measuring- and control instruments for testing the material quality in all sectors of research, development- and production within the food- and feed industry. It is a leading worldwide supplier for laboratory equipment in the milling, baking- and starch industry. The Brabender GlutoPeak is a relatively new quick test procedure and determines the aggregation behaviour and thus the gluten quality within  5 – 10 minutes.

For further questions, relating to this study, please, contact:

zuleyka.rodriguez@buhlergroup.com

markus.loens@brabender.com