EP2179257A1

EP2179257A1 - Learning method for producing color formulas

Info

Publication number: EP2179257A1
Application number: EP08785519A
Authority: EP
Inventors: Christian Bornemann; Heiner Cloppenburg; Carlos Vignolo; Jürgen LOHMANN; Stuart Kendall Scott
Original assignee: BASF Coatings GmbH
Current assignee: BASF Coatings GmbH
Priority date: 2007-08-14
Filing date: 2008-08-13
Publication date: 2010-04-28
Also published as: WO2009021720A1; JP5538222B2; JP2010536046A; DE102007038483B4; CN101784871B; CN101784871A; US10996110B2; US20110097691A1; DE102007038483A1

Abstract

The invention relates to a learning method for calculating color formulas of pigmented color shades adjusted to match a target color shade, comprising the steps i) preparation of an effect matrix for the components present in a coloring system by suitable scaling, i) determining the optical material parameters of the target color shade, ii) selecting a suitable starting formulation, iii) determining the color difference between the initial formulation and the target color shade, iv) calculating a first adjusted color formulation considering the effect matrices, v) repeating the steps iv) and v) until an acceptable remaining color difference has been achieved, in which the effect matrices are constantly updated with information relevant to color shades in the running process, achieving an improvement to the prior known method of color shade formulation calculation both with regard to reduction in the number of shade steps required, and with regard to minimizing the remaining color difference.

Description

Learning process for the preparation of color formulas

The invention relates to a learning process for the preparation of color formulations, which can be adapted in a few steps to a fixed color specification.

In the production of paint batches, in particular in the automotive industry, it is an important task to represent the color tone, which is produced by weighing the quantities of constituents specified in a color formulation, with respect to a predetermined target color shade with the smallest possible deviation. The formulation ingredients can be both colored in nature, such as color or effect pigments, as well as non-colored nature, including, for example, binders. Additives and solvents are understood. The aim of the toning of the batches is to adjust the color of the batch to the target color tone with as few tinting steps as possible, in the interest of economic efficiency. This adjustment is made by slight changes in the amounts of colored ingredients contained in the recipe and optionally by adding further colored Tönzusätze in low concentrations. The adaptation step is not complete until an acceptable residual color difference has been reached between the color of the batch and the target color tone.

Whereas the described sounding process used to be primarily visual, instrumental control systems are used today. This includes above all the use of a spectrophotometer with which reflection spectra in the visible range of the electromagnetic spectrum are recorded at different illumination and observation angles. From the convolution of such reflection spectra with one type of light and one of the three standard spectral value functions, color indices result that indicate the color locus, that is to say the position of the color tone in the color space. As standard, the color space of the so-called ClELab Coordinates L ^* , a ^* and b ^* established. Color differences dL *, da * and db * then result from the difference between two color loci with respect to the respectively measured coordinates L *, a * and b * of the two color tones.

From the German patent application DE 197 20 887 A1 a method for color recipe calculation in the field of effect surface coatings is known, which is based on the establishment of a calibration scale for each pigment based on a dyeing system. It is described that subsequently the associated reflection factors are determined experimentally. In this way, for each color pigment and for each effect pigment, the corresponding optical material parameters, such as scattering coefficient, absorption coefficient or phase function, which are used in the computational determination of the formulation for a given Effektfarbtcn by computer simulation. The optical material parameters thus describe the properties of the pigments as they are dispersed in the respective binder system. They are wavelength dependent and must be determined for each desired wavelength. Altogether, the quality of the calculated recipes strongly depends on the standardization of all components of a mixed lacquer system and the constancy of the application parameters. Against this background, it should be noted that, due to the complexity and cost of their determination, the optical material parameters are determined only once for the particular pigment in the binder system.

Another computational method for color matching of paints or inks is known from the European patent application EP 0 828 144 A2. In this method, the reflection factors are first determined for the target shade with a goniospectrophotometer and from this the associated standard color values and / or the vectors derived therefrom in the ClELab color space are calculated. Subsequently, from a recipe database on the basis of the reflection spectra or the derived color technical Measures identify those recipes or shades that are most similar in their reflective properties to those of the target hue. With the aid of an action matrix, which is likewise constructed purely mathematically, the residual color tone difference between the target color tone and the database color tone is calculated by means of a calculation algorithm for functional minimization taking into account

Minimized constraints in the sense of the L ₂ standard and created a corrected recipe. The effect matrix is a decisive factor for the quality of the hue adjustment. It is, as already described above roughly for the process of DE 197 20 887 A1, set up by means of experimentally determined calibration scales of the reflection parameters of the pigments underlying the recipe and also stored in the recipe database. Therefore, the effect matrix describes the coloristic effect of the individual recipes in the Reflexicnsraurn or color space angle-dependent with varying

Pigment concentrations.

Based on the method just described, far-reaching studies have been carried out to further improve and refine the calculation programs for determining the target formulation for a color tone adapted to the original. Representative of this aspect of a process for the preparation of color formulations that must be based on a template, are described in the publications WO 2006/052556 A2 and WO 2006/052561 A2 optical conversions of the adjustment process based on the individual steps by a color reproduction in the RGB color space on a monitor for the user to make comprehensible. In this way, a connection is established between the well-established visual adaptation and the computer and instrument-based adaptation, which enables a simulation of the influences of changes of individual pigments and / or constituents of the formulation.

A disadvantage of all previously known methods, however, is that they offer no possibility of fluctuations in the raw and To detect input materials or color differences by different manufacturing and application methods. Experience has shown that such fluctuations can be so great that a re-adaptation of the color shade produced to the target shade is necessary, although the formulation corresponds to the previously determined calibration scales. In industrial series production, this can lead to considerable costs either through necessary reworking of the actually finished products or through time delays through further adjustments.

The object of the invention is therefore to provide a method for color recipe calculation, which also takes into account the fluctuations of the raw materials and starting materials and the effects of process fluctuations and thus the residual color difference to the target color tone while reducing the number of Tönschritte required.

Surprisingly, it turns out that a method for color recipe calculation of pigmented hues adapted to a target shade, comprising the steps i) setting up each effect matrix by suitable

Calibration scales for the constituents contained in a staining system, ii) determining the optical material parameters of the

(Iii) selection of a suitable starting formula; (iv) determination of the color difference between the

V) calculation of a first adjusted color recipe taking into account the effect matrices, vi) repetition of steps iv) and v) until an acceptable residual color difference is reached at which the effect matrices in the running process are constantly updated with hue relevant information, an improvement the previously known methods both by a reduction of Number of Tönschritte required as well as achieved by minimizing the residual color distance.

In the following, it will be explained in more detail how the first created effect matrix for a dyeing system is determined in detail.

For the creation of the effect matrix is first a prerequisite that the optical material parameters of all components of the dyeing system are known.

By adapting the radiative transfer equation in the sense of an I_ ₂ standard, the optical material parameters are adapted to each one

Pigment experimentally determined reflection factors or

Beam density factors determined.

In the case of uni-pigments, the well-known Schuster / Kubelka / Munk

Approximation of the radiative transfer equation completely sufficient. In the context of this 2-flux approximation, a simple relationship between the reflection of, for example, an opaque lacquer layer and the scattering and absorption properties of the pigments contained in this layer can be derived.

The pigment-specific and wavelength-dependent material parameters of the scattering and absorption coefficients must be determined experimentally by means of a calibration scale in a manner known to the person skilled in the art.

To describe effect shades must instead of the known

Schuster- / Kubelka - / - Munk approximation of the radiative transfer equation another, known in the art form of

Radiation transport equation can be used. In addition to the scatter and

Absorption coefficient is additionally to determine the phase function.

For metrological detection of the reflection surfaces, a stationary or portable goniospectrophotometer with symmetrical or asymmetrical measurement geometry can be used. The area of the observation angle to be covered depends on the particular approximation used for the radiative transfer equation. It can both devices with lighting and with

Observation modulation can be used. To determine the coloristic effect matrix, the amounts of the recipe components (N pigments) for a given recipe are slightly varied by their nominal concentration and the associated coloristic effect in the reflection space or in the ClELab space is calculated as an angle (M angle) using the optical material parameters. In other words, the effects of changes in concentration of constituents whose material parameters are known can be calculated as effects in color space or reflection space.

The information content of the effect matrix can now be used immediately or later at any given time to tint a reference recipe to a target point deviating from the reference point. This target point should not be so far away from the reference point that the scope of the approximation used is exceeded.

Basically in the professional world under the term

"Hue relevant information" hue changes dL *, da * and db * for achromatic shades and dL *, dC * and dH * for colored hues as explained above.Such hue changes expressed by dL *, da * and db * for achromatic hues and dL *, dC and dH * for colorful shades can be used in the manufacturing process

Charge arise in a variety of ways. Some key influencing factors are included

Raw material fluctuations, in particular of the pigments used for the production of pastes, which have an influence on the hue, such as, for example, the particle size distribution and surface texture and the layer thickness of the pigments, fluctuations in the weighting of the raw materials

Intermediate production of, for example, pigment pastes and occurring process fluctuations, Variations in the weighting of the intermediates for batch production and occurring process fluctuations and

Application parameters during the coating of the batch, such as, for example, horn air, atomizing air, outflow rate,

Belt speed, high voltage and bell speed, and other environmental parameters such as temperature, humidity, air sink rate or UV radiation.

An ab initio description of the fluctuations in the aforementioned influencing variables and their effects on the color behavior of the charge can not be relevant to practice due to the manifold influences and the complexity of the relationships. In addition, the often insufficiently precise recording of this data prevents such a purely theoretical approach. For this reason, a phenomenological approach is used in d ■ ■ CJΛIO to determine the hue changes dL *, da * and db * for achromatic hues and dL *, dC * and dH * for colorful shades in the batch process. For this purpose, it is possible in practice to describe the hue changes as a function of the concentration of the intermediates used. The hue changes can then be represented as differentials dL7dc da * / dc "db * / dc, and dL7dc" dC * / dc "dH7dc. In the same way, the color tone changes can be detected as a function of the aforementioned process parameters.

The term "color-relevant information" is understood here and in the following as meaning all information and measured values which comprise color tone changes caused by respective quantity or batch changes in the continuous process, and the entirety of the dependencies of the color tone changes in the form of differentials of the effect matrix to be created therefrom summarized under this term.

This information can either be collected in accordance with this invention in addition to existing processes. You can but preferably obtained from existing testing processes. For example, such information may be obtained in routine, ongoing quality control. The color-relevant information is particularly preferably obtained from laboratory tests, quality tests, incoming inspections or operating samples.

The term "dyeing system" as used herein means any combination of at least two different pigments and / or binders Preferably, a dyeing system comprises a large number of different colored or effect pigments containing compositions, which may be termed either a base or pigment paste. The number and variety or selection or combination of the pigment components are not subject to any restriction and can be adapted to the respective requirements, for example, such a dyeing system can comprise all the pigment components of a standardized mixed-lacquer system underlie.

Effect pigments are to be understood as meaning all pigments which have a platelet-like structure and give a surface coating special decorative effects. The effect pigments are, for example, all effect pigments which can usually be used in vehicle and industrial coating or in ink and dye production. Examples of such pigments are pure metal pigments such as aluminum, iron or copper pigments, interference pigments such as titanium dioxide coated mica, iron oxide coated mica, mixed oxide coated mica, metal oxide coated aluminum, or liquid crystal pigments.

The coloring absorption pigments are, for example, customary organic or inorganic absorption pigments which can be used in coating chemistry. Examples of organic Absorbent pigments are azo pigments, phthalocyanine, quinacridone, and pyrrolopyrrole pigments. Examples of inorganic absorption pigments are iron oxide or lead oxide pigments, titanium dioxide and carbon black. Examples of so-called pseudo pigments are those substances which are influencing the topology in terms of effect pigments, but are otherwise coloristically ineffective. They are preferably selected from the group of known fillers.

With the updating of the effect matrices according to the invention by the current color-relevant information, in contrast to the previously known methods, a dynamic is introduced which makes it possible to also detect fluctuations in the raw materials and starting materials as well as fluctuations through production and application processes and to take them into account accordingly. In this sense, the inventive method can be referred to as learning. In this way, it is also possible for the first time to greatly increase the accuracy of the process. Because the accuracy of the process is crucially dependent on the quality of the effect matrices.

In a further embodiment of the present invention, the effect matrices can be expanded with shade-relevant information from laboratory tests, quality checks, incoming inspections and operating samples. In this refinement of the method, the effect matrices, which were first determined during the preparation and production of a base shade or a color paste, are later extended by further parameters. This extension of the effect matrices is preferably done by shade relevant information obtained in the ongoing process by the measurement of hue changes depending on different amounts, application methods or batches. From this information, it is possible to determine the changes in the optical material parameters, which are ultimately included in the effect matrices. The process according to the invention can be used, for example, for the color shade of paints and printing inks or polymer dispersions.

The invention is explained in more detail below with reference to the drawing, without limiting the invention thereto. The picture shows in

Fig. 1 is a schematic representation of the learning correction according to the invention (right) compared to the classical static stitching.

From the same reference mixture, a corrected hue is achieved by stepwise correction, which achieves an acceptable residual color difference with respect to the associated hue standard. As can be seen from the figure, according to the classical method, considerably more tinting steps are required to reach the target point than in the new learning method. A characteristic of the learning, dynamically updated method can be seen in the fact that even in the first tinting step a very close approach to the target point succeeds due to the high quality of the effect matrices.

Claims

claims

1. A method for color recipe calculation of pigmented hues adapted to a target hue, comprising the steps i) Setting up each effect matrix by suitable

Initial formulation and the target hue, v) calculation of a first adapted color recipe taking into account the effects mstrics, vi) repetition of steps iv) and v) until an acceptable residual color difference is reached, characterized in that the effect matrices in the running process are constantly updated with hue relevant information.

2. The method according to claim 1, characterized in that the shade relevant information from laboratory tests, quality checks, incoming inspections and operating samples are obtained.

3. The method according to claim 1 or 2, characterized in that the effect matrices are expanded with shade-relevant information from laboratory tests, quality checks, incoming inspections and operating samples.

4. Use of a method according to any one of claims 1 to 3 for the color tint of paints and printing inks or polymer dispersions.