WO2023041596A1

WO2023041596A1 - Method for tinting a batch of paint

Info

Publication number: WO2023041596A1
Application number: PCT/EP2022/075551
Authority: WO
Inventors: Eric Cornelis VAN WINDEN; Wilhelmus Johannes MOOIJMAN
Original assignee: Akzo Nobel Coatings International B.V.
Priority date: 2021-09-17
Filing date: 2022-09-14
Publication date: 2023-03-23

Abstract

The invention relates to a method for tinting a batch of paint to match a targetcolor. The method involves taking a sample of paint from the batch of paint,determining its color, then correcting the color of the sample of paint in accordancewith a first color correction recipe, determining the color of the color-corrected5sample of paint, and correcting the color of the batch of paint in accordance with asecond color correction recipe. The second color correction recipe is calculatedpartially based on scaling of the first color correction recipe. The invention furtherprovides a device configured and arranged for executing the method according tothe invention.

Description

METHOD FOR TINTING A BATCH OF PAINT

Field of the invention

The present invention relates to tinting of batches of paint.

Background

Tinting of batches of paint, also known as shading, is a process of adjusting color of a produced batch of paint in case it deviates from the desired standard color. Tinting is often used in the manufacture of batches of paint, particularly of a large volume (e.g. 1000 liters). Paint is normally produced by mixing plurality of toners in accordance with a known recipe of the standard color. However, the result of a first mixing can have a too large color deviation from the desired standard color due to e.g. process fluctuations, quality variations in the raw materials (i.e. pigments in the toners), and other reasons.

A tinting process involves relatively minor additions of toners to the produced batch in order to correct its color. The additions should be small because of economic reasons (too much waste of material if the produced paint volume is much higher than ordered) and because the size of the tank is limited. Also, no pigments can be taken out of the formula so only additions can be made. These constraints make it complicated to correct the color of an already produced batch of paint and not any color correction method can be used.

US20080052023 A1 describes a method for color matching a reference color formulation to a defined color shade standard. The method is based on a comparison of measured reflection spectra of color shades of known pigmentation and the corresponding theoretical expectation values. The more samples are available, the more information can be collected about the coloristic deviations between the materials used for the pigment calibration and the actually employed raw materials for the color matching. The method is said to stabilize after three to five correction steps.

W02006/052561 A2 discloses a method for guiding a refinisher user through adjustments of a candidate refinishing paint color formed from a combination of a predetermined number of tints T1 through Tn toward a predetermined target paint color. Each tint is also displayed in a respective field of the video monitor. In response to the selection of one of the tints, vector analysis is used to calculate an updated candidate color point in the multidimensional color space. Upon selection of a tint a rendering of an updated candidate refinishing paint color corresponding to the updated candidate color point in the multidimensional color space is displayed in the second field. US4917495 A discloses a portable colorimeter and a method for characterizing the optical properties of a colored surface containing metallic or pearlescent particles. The colorimeter employs three multiangular spectrophotometric measurements to derive color constants for the surface. The method includes determining the tristimulus values of the color of the sample surface from low resolution spectral reflectance data preferably using twelve detector elements.

Although methods for tinting are known, they often require many iteration steps, or result in unsatisfactory results, especially when the stored K and S values are inaccurate for the used pigments in the toners, or due to process parameter deviations. In such cases, the same inaccuracy drags on in the iteration steps and cannot provide for a satisfactory color match.

It is desired to provide a tinting method of a batch of paint with a high accuracy of matching to the target color and requiring a minimum number of color correction steps, most preferably, only one step. It is further desired to provide a system that implements the method.

Summary of the invention In order to address the above-mentioned desires, the present invention provides, in a first aspect, a method for tinting a batch of paint having color A1 to match a target color T, wherein tinting is done by addition of one or more toners to the batch of paint, comprising the steps of: a) taking a sample of paint from the batch of paint, b) determining color A1 of the sample of paint, c) calculating color A2 within a predetermined tolerance distance AT from target color T in a color space, said color A2 having the closest match with color T based on the available toners for tinting, d) calculating a first color correction recipe based on the color difference between colors A1 and A2 in the color space, e) adding one or more toners to the sample of paint in accordance with the first color correction recipe to obtain a color-corrected sample of paint, f) determining color A2’ of the color-corrected sample of paint, g) calculating color A3 within the predetermined tolerance distance AT from target color T in the color space, wherein the vector from color A1 to color A3 in the color space is obtained as a sum of two vectors v1 and v2, wherein the first vector v1 starts from color A1 , has the same direction as the vector from color A1 to color A2’ and is scaled with a scaling factor S with respect to that vector, and the second vector v2 corresponds to the addition of one or more toners, wherein the scaling factor S, the length and direction of the second vector v2 are varied within available colors in the color space and within available toners and wherein the distance A3 - T in the color space is minimized, h) calculating a second color correction recipe, which is a sum of partial color correction recipes p1 and p2, wherein partial color correction recipe p1 is calculated by scaling of the first color correction recipe with scaling factor S, and partial color correction recipe p2 corresponds to vector v2 in the color space, i) adding one or more toners to the batch of paint in accordance with the second color correction recipe.

In another aspect, the present invention provides a device configured and arranged to execute the method according to the invention.

Brief description of the figures

Figures 1A-1 D show a schematic representation of an exemplary method for adjusting the color value of a batch of paint in a color space.

Detailed description of the invention

The invention is based on a judicious insight that a quick and highly accurate tinting of a batch of paint can be done by first tinting a small sample of paint and using that information including a step of scaling the color correction recipe, in order to calculate a more precise color correction recipe for the whole batch of paint.

Advantages of the present invention include a much more precise tinting of the batch of paint to target color T in fewer steps compared to existing methods. Existing methods are based on iteration and include a series of color correction steps to approach the target color T, typically 3-5 steps. This is especially the case when the stored K and S values for the pigments are out of date or are inaccurate because a certain colorant in the toners (used to prepare the batch of paint) has a deviating color value. In such cases, the same inaccuracy drags on in the iteration steps and cannot provide for a satisfactory color match in few steps.

To the contrary, in the present method, only one color correction step is performed on the batch of paint, which is sufficient for most applications. This is because the color correction recipe calculated for the whole batch already takes into account any deviations between predicted and determined color values that were observed during a separate color correction performed on a small sample of paint from the same batch. The improved recipe is based for a part on a scaled version of the first color correction recipe, which is a unique feature of the present invention.

A batch of paint is typically prepared in accordance with a standard recipe, i.e. by mixing one or more toners and optionally other components (e.g. base paint, reducer, etc.). The batch typically has a large volume, e.g. 1000 liters of paint. Toners typically contain one or more pigments and a binder resin system. Pigments can refer to both solid color pigments and effect pigments. There is normally a limited number of toners, e.g. 2-50, preferably 2-30, more preferably 10-30. In some embodiments the number of toners can be below 10, such as 2-8 or preferably 3-6.

In the method according to the invention, first, a sample of paint is taken from the batch of paint. The sample has a significantly smaller volume compared to the volume of the whole batch of paint. Preferably, the volume of the sample is rather small, e.g. at least 100 times smaller than the batch of paint, more preferably at least 500 times smaller. For example, the sample can have a volume of 1 liter and the batch 1000 liter.

Next, the color A1 of the paint sample is determined, preferably by obtaining a reflectance curve, e.g. spectrophotometrically. Color A1 also represents the color of the whole batch, since the sample was taken from the batch. The color determination can be done on wet or dry paint. Preferably, a test panel is sprayed with the paint from the sample and allowed to dry, after which the color of the dry panel is determined. It is also possible to do color measurements on wet paint samples. From these measurements, the color values of the sample, e.g. L, a, b values of the CIELab color space, can be derived in a way well-known to a person skilled in the art of colorimetry or can be measured directly with an appropriate measuring device. For the ease of reading, “color values” are also referred to as “color” throughout this application. The color space is typically two or three-dimensional space such as RGB, CIELAB, or multi-dimensional (e.g. for effect pigments). Then, color A2 is calculated that is sufficiently close to target color T, that is, within a predetermined tolerance distance AT from target color T in the color space. T is the target color, which can be retrieved from a database containing previous measurements, or it can be determined by obtaining respective reflectance spectra e.g. of an existing panel painted in the target color T. Since there is only a limited number of toners, it is often not possible to obtain any color in the color space and color A2 needs to be calculated as something that can be achieved with the available toners sufficiently close to T. From the plurality of possible colors A2 based on available toners, the closest match with color T is selected as the color A2.

This is illustrated in Figure 1A, which shows colors A1 , A2 and T in a color space. For the ease of understanding the figures show a two-dimensional color space to illustrate the present invention. In Figure 1 A, point A1 represents the measured color of the paint sample (which is also the color of the batch). In order to correct the color of the sample A1 to approach T as close as possible, color A2 close to T is calculated. Color A2 is located within a predetermined tolerance distance AT from target color T in the color space and is as close to target color T as can be achieved with the available toners.

Further, a first color correction recipe is calculated taking into account the color difference between A1 and A2 in the color space. This can be done with known methods, e.g. using the Kubelka-Munk equation based on wavelength dependent absorption (K) and scattering (S) values for the pigments used in the toners and the Duncan rule. The result of this calculation is a first color correction recipe that specifies which toners and in which amounts need to be added, based on their known K and S values, to obtain color A2.

In a next step, the color of the sample is corrected by adding one or more toners to it according to the first color correction recipe. Preferably, more than 1 toner is added. Preferably, the number and/or the volume of the toners to be added is minimized. Preferably, not more than 10 toners are used, more preferably not more than 5 toners are used, yet more preferably not more than 3 toners, and in some cases not more than 2 toners. Similarly, the volume of the toners to be added is preferably as small as possible - in order not to add too much volume to the prepared batch of paint.

In a following step, the color of the color-corrected sample is determined (e.g. spectrophotometrically), which results in the actual color value A2’, which is illustrated in Figure 1 B. The actual color A2’ is thus often different from the predicted color A2. As mentioned earlier, the reasons for this discrepancy may be quality variations in the raw materials (i.e. pigments in the toners), not-up-to-date K and S values for the used pigments and process fluctuations.

Although the actual color A2’ is not yet a good match with target color T, the inventors of the present invention have realized that it provides valuable information, which can be used to calculate an improved color correction recipe to correct the color of the batch of paint from A1 to T. Particularly, in accordance with the invention, the first color correction step provides the change direction (represented as the direction of the vector from A1 to A2’ in the color space), which can be in part used in calculation of an improved color correction recipe.

In order to calculate the second (improved) color correction recipe, first, color A3 close to T needs to be determined, which is based on the information obtained from the first color correction step and is sufficiently close to T (within the predetermined tolerance distance AT). This is illustrated in Figure 1 C. According to the present invention, color A3 is calculated using two vectors in the color space.

Particularly, the vector from color A1 to color A3 in the color space is obtained as a sum of two vectors v1 and v2. The first vector v1 starts from color A1 , has the same direction as the vector A1 - A2’ and is scaled with a scaling factor S with respect to that vector. This means that the first vector v1 has a variable length but fixed direction and starting point. The first vector v1 corresponds to a scaled version of the first color correction recipe. The second vector v2 corresponds to the addition of one or more of available toners. The second vector v2 may start from the end of the first vector v1 , it has a variable length and direction and ends at color A3. The scaling factor S, the length and direction of the second vector v2 are varied within available colors in the color space and within available toners, wherein the distance A3 - T in the color space is minimized. In other words, the sum of the two vectors is optimized in order to achieve the closest match between colors A3 and T for the available toners.

The sum of two vectors v1 and v2 can be represented in the color space in various ways. Vectors v1 and v2 can have the same starting point such as color A1 and have different directions. Also, it can be represented that vector v2 starts at the end point of vector v1 and ends at color A3. This is not critical to the invention. In any representation, the sum of vectors v1 and v2 is vector A1 -A3 that ends at color A3 close to target color T.

As mentioned above, the first vector v1 is obtained by scaling of vector A1-A2’. In practice, the first color correction step often turns out to be an “overshoot”, meaning that the vector A1 -A2’ is too long and too much paint has been added in the first color correction step. In such case, the length of vector v1 will be a part of the length of vector A1-A2’ (scaling factor S is less than 1). In other cases, it can be that vector A1 -A2’ is too short, meaning that scaling factor S is more than 1 . It is further possible that the scaling factor S is equal to 1 .

Once suitable v1 vector is found, it is converted into a respective partial color correction recipe p1 for toner(s) addition. Hence, the first vector v1 determines the first partial recipe p1 , which together with the second partial recipe p2 (as explained further below) makes up the combined, second color correction recipe.

First partial recipe p1 is completely based on the first color correction recipe scaled with scaling factor S to the length of vector v1 . This means, e.g. that in p1 the same toner or toners are used as in the first color correction recipe, but in different amounts (when S < 1 , this means in lower amounts). The amounts of toners in recipe p1 are adjusted proportionally to the scaling factor S, which is the length of vector v1 compared to the length of vector A1-A2’. E.g. if in the first color correction recipe 5% of a black toner was used, and the scaling factor is 0.5 (the length of vector v1 is 50% of the vector A1-A2’), then in the recipe p1 the amount of black toner will be 2.5%.

The second vector V2 corresponds to the addition of one or more toners. Vector v2 is preferably constructed from the end of the first vector v1 , has a variable length and direction and ends at color value A3 that is possible to achieve for the available toners. The second vector v2 can be constructed based on the known Kand S values of the pigments used in the available toners but can also be constructed based on historical data of previous color correction steps.

In a particularly preferred embodiment, vector v2 corresponds to the addition of only one toner. This limits the possible directions for vector v2 as each toner defines a particular direction in the color space. This simplifies the optimization process of the sum of vectors v1 and v2 considerably. In the embodiments where it is desired to only use the same toners as present in the original standard recipe of the batch of paint, the number of toners for vector v2 is further limited. Since the number of available toners is finite, it is possible in this embodiment to know beforehand which directions vector v2 can have. In case of 5 toners, for example, 5 possible vector directions for vector v2 can be identified upfront. The length of vector v2 remains variable as it represents the amount of the toner to be added and this can vary.

If possible directions of vector v2 are known beforehand, it is possible to make a preselection of those vectors v2, which point in the direction of target color T rather than away from target color T. The vectors v2, which point away from target color T, can then be disregarded in the optimization step of the vector sum v1 and v2. Terms “pointing in the direction of T” and “pointing away from T” can be defined based on the projection of vectors v2 onto the line A1 - T. If the projection of vector v2 onto that line points in the same direction as vector A1-T, then vector v2 is considered to point in the direction of T. If the projection of vector v2 onto that line points in the opposite direction as vector A1-T (and in the same direction as vector T-A1), then vector v2 is considered to point away from T. The advantage of this pre-selection is that it reduces the number of variables used in the optimization step, which simplifies the optimization process and increases the chances of a good fit of color A3 to target color T.

Preferably, vectors v1 and v2 are not determined separately but together in one optimization step since they depend on each other. During the optimization, the scaling factor S, the length and direction of vector v2 are varied within available colors in the color space and within available toners, wherein the distance between colors A3 and T in the color space is minimized. In other words, the sum of the two vectors is optimized within the available constraints in order to achieve the closest match of A3 and T.

Algorithms for optimization of a sum of two vectors are known to the skilled person. Such algorithms calculate various combinations of vectors and find the optimized combination(s) which meet all pre-defined requirements or constraints. Methods of vector sum optimization include for example QR decomposition, which uses matrix calculations, and the Broyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm, which is generic method for minimization of various values. Preferably, QR decomposition is used.

Once partial recipes p1 and p2 are calculated, the second color correction recipe is calculated as a sum of partial color correction recipes p1 and p2. Partial color correction recipe p1 corresponds to the vector v1 in the color space and partial color correction recipe p2 corresponds to vector v2 in the color space. Since vector v1 is a scaled vector obtained by scaling of vector A1-A2’, the partial color correction recipe p1 is a scaled version of the first color correction recipe, as explained above, the scaling factor S being determined during the optimization. The second partial color correction recipe p2 corresponds to the optimized vector v2 in the color space, which corresponds to the addition of one or more toners (and preferably only one toner).

The sum of the two mentioned vectors is optimized in order to find color A3 that gives the closest match with target color T, based on the available toners and restrictions applied on the system. A closest match means that the distance in the color space between color A3 and T is minimized. As mentioned earlier, since the number of toners (and hence, available colors of the pigments therein) is very limited, there are only certain points in the color space that can represent A3. From those possibilities, color A3, closest to color T in the color space, is selected. This is illustrated in Figure 1 D, where several possible colors A3 and different combinations of vectors vi and V2 are shown.

Additionally, the sum of vectors vi and V2 can be optimized (or the distance A3 - T minimized) in order to use the least number of toners in the second color correction recipe, preferably 1-2 toners, most preferably only 1 toner. The sum of the vectors vi and V2 can also be optimized in order to use as few new toners as possible. Preferably no new toners are used. New toners are the toners that are not yet present in the prepared batch of paint, color of which needs to be adjusted. This is beneficial, since the more different pigments (toners) are used, the more chance that other hues are introduced, which may require further color correction.

Also, the sum of vectors vi and V2 can be optimized (or the distance A3 - T minimized) in order to use the least volume of toners in the second color correction recipe. The less the volume of toners in the second color correction recipe, the less total increase in volume of the batch of paint, leading to less paint waste. In the color space, this is represented by the vectors v1 and v2 having the shortest length, particularly, so that the sum of the lengths of vectors v1 and v2 is minimized.

In practice, for each available toner it is possible to calculate all possible A3 color values, which can be achieved in the second color correction step, and then choose the one that is closest to T and/or meets one of other requirements.

As previously mentioned, the second color correction recipe is calculated as a sum of partial recipes p1 and p2 determined earlier. Although these partial recipes are calculated separately, it does not mean that the toner(s) need to be added to the paint batch in two steps. Toner addition can be done at once in accordance with a combined recipe that is a sum of recipes p1 and p2. The sum of partial recipes means a recipe, in which the concentrations of the same toner are added up and if different toners are used, all toners are used in the combined recipe in the respective concentrations as in the partial recipes p1 and p2.

Optionally, a further sample of paint can be taken from the batch of paint and subjected to a color correction according to the second color correction recipe. The color of this second, color-corrected sample of paint can then be determined to check whether it is close enough to target color T. The deviation of the actual (determined) color A3’ from predicted color A3 is expected to be small, i.e. much smaller than the deviation of A2’ from A2. The reason for a much smaller deviation is that the correction method already accounted for deviations in raw materials and process parameters, in contrast to conventional methods which only use K and S values. Therefore, in practice, testing of the second color correction recipe on a second sample from the batch of paint is not necessary.

In the next step, the color of the whole batch of paint is corrected by adding one or more toners in accordance with the second color correction recipe.

This results in a batch of paint, the color of which has an improved match with the desired target color T.

The above-described method can be suitably computer implemented.

In another aspect, the present invention provides a device configured and arranged for executing the method according to the invention. The method as described herein can be executed on a computing device such as a point-of-sale computer system or a mobile computing system, e.g. a smartphone, tablet, laptop, computer or the like. The method can also be provided by a computer program product, such as an app, loaded and executed on a general purpose computer or a mobile computing system. It will be appreciated that all features and options mentioned in view of the methods apply equally to the device, and vice versa. It will also be clear that any one or more of the above aspects, features and options can be combined.

For the purpose of clarity and a concise description features are described herein as part of the same or separate embodiments, however, it will be appreciated that the scope of the invention may include embodiments having combinations of all or some of the features described.

In the claims, any reference sign placed between parentheses shall not be construed as limiting the claim. The word ‘comprising’ does not exclude the presence of other features or steps than those listed in a claim. Furthermore, the words ‘a’ and ‘an’ shall not be construed as limited to ‘only one’, but instead are used to mean ‘at least one’, and do not exclude a plurality. The mere fact that certain measures are recited in mutually different claims does not indicate that a combination of these measures cannot be used to an advantage.

Claims

1 . A method for tinting a batch of paint having color A1 to match a target color T, wherein tinting is done by addition of one or more toners to the batch of paint, said method comprising the steps of: a) taking a sample of paint from the batch of paint, b) determining color A1 of the sample of paint, c) calculating color A2 within a predetermined tolerance distance AT from target color T in a color space, said color A2 having the closest match with color T based on the available toners for tinting, d) calculating a first color correction recipe based on the color difference between colors A1 and A2 in the color space, e) adding one or more toners to the sample of paint in accordance with the first color correction recipe to obtain a color-corrected sample of paint, f) determining color A2’ of the color-corrected sample of paint, g) calculating color A3 within the predetermined tolerance distance AT from target color T in the color space, wherein the vector from color A1 to color A3 in the color space is obtained as a sum of two vectors v1 and v2, wherein the first vector v1 starts from color A1 , has the same direction as the vector from color A1 to color A2’ and is scaled with a scaling factor S with respect to that vector, and the second vector v2 corresponds to the addition of one or more toners, wherein the scaling factor S, the length and direction of the second vector v2 are varied within available colors in the color space and within available toners and wherein the distance A3 - T in the color space is minimized, h) calculating a second color correction recipe, which is a sum of partial color correction recipes p1 and p2, wherein partial color correction recipe p1 is calculated by scaling the first color correction recipe with scaling factor S, and partial color correction recipe p2 corresponds to vector v2 in the color space, i) adding one or more toners to the batch of paint in accordance with the second color correction recipe.

2. The method according to claim 1 , wherein the distance A3 - T in the color space is minimized using QR decomposition or the Broyden-Fletcher-Goldfarb- Shanno (BFGS) algorithm.

3. The method according to claim 1 or 2, wherein the first color correction recipe is calculated based on the absorption K and scattering S values of the pigments used in the toners.

4. The method according to any one of claims 1 -3, wherein the distance A3-T in the color space is additionally minimized in order to use the least number of toners in the second color correction recipe.

5. The method according to any one of claims 1 -4, wherein the distance A3-T in the color space is additionally minimized in order to use the least volumes of toners in the second color correction recipe.

6. The method according to any one of claims 1 -5, wherein the distance A3-T in the color space is additionally minimized in order to only use the toners already present in the batch of paint.

7. The method according to any one of claims 1 -6, wherein colors A1 and A2’ are determined by obtaining respective reflectance spectra of either dry paint after being sprayed on a panel, or wet paint samples.

8. The method according to any one of claims 1-7, which is computer implemented. 16

9. The method according to any one of claims 1 -8, wherein the target color T is either retrieved from a database or is determined by obtaining respective reflectance spectra of a painted panel with color T.

10. Device configured and arranged to execute the method according to any one of claims 1-9.