1 Introduction Lubricating oil plays an important role in guaranteeing smooth and efficient functioning of machinery. However, these oils have a limited period of operation and need to be exchanged in regular intervals. Currently, in general the industry performs oil changes in regular predetermined intervals, which has the disadvantage that exchanges are possibly performed long before they would be necessary. One of the possible reasons for such behavior is an economical factor. From this perspective the simple exchange of the oil is cheaper than the monitoring of the oil quality in regular instances. The problem is that the most widely accepted methods for quality control (QC) of lubricating oils are time-consuming and require various analytical instruments, additional reagents and qualified staff. This influences dramatically the price of each single analysis and make routine quality control uneconomical. Aus der Praxis für die Praxis 35 Tribologie + Schmierungstechnik · 65. Jahrgang · 4/ 2018 Simultaneous monitoring of multiple quality parameters of lubricating oils by Vis-NIR spectroscopy A. Kadenkin, E. Hagemann, D.R. van Staveren* Im Rahmen dieser Arbeit wurden die Einsatzmöglichkeiten der sichtbaren (Vis) und Nah-Infrarot- (NIR) Spektroskopie in der petrochemischen Industrie am Beispiel der Bestimmung verschiedener physikalischer und chemischer Qualitätsparameter in Schmierölen nach ASTM E1655 gezeigt. Hierfür wurden spezifische NIR-Methoden unter Verwendung eines großen Probensatzes bestehend aus verschiedenen Ölen von mehreren Herstellern entwickelt. Die einzigartige Kombination von Nah-Infrarot-Spektroskopie mit sichtbarer Spektroskopie ermöglichte dabei die Ermittlung zusätzlicher Qualitätsparameter wie die Farbzahl. Die ermittelten Standardfehler der Kreuzvalidierung betrugen 0.21 mg KOH g -1 für Säurezahl, 0.14 mg KOH g -1 für Basenzahl, 0.9 für Farbzahl, 0.0031 g cm -3 für Dichte, 0.11 % für Feuchtigkeit, 0.54 cSt für Viskosität bei 100 °C, 11.7 cSt für Viskosität bei 40 °C und 1.5 für Viskositätsindex. Die ermittelten Fehler waren vergleichbar mit den Fehlern herkömmlicher Prüfverfahren für Schmieröle. In einem zusätzlichen Schritt wurden die entwickelten Methoden mit einem unabhängigen Probensatz erfolgreich validiert. Somit hat sich die Vis-NIR-Spektroskopie als eine schnelle, zeit- und kostensparende Methode zur Qualitätskontrolle von Schmieröl erwiesen. Schlüsselwörter Qualitätskontrolle, Nahinfrarot-Spektroskopie, Säurezahl, Viskosität, Feuchtigkeit, Basenzahl, Dichte, Viskositätsindex The possibilities of visible (Vis) and near-infrared (NIR) spectroscopy in the petrochemical industry were demonstrated on the example of the determination of various physical and chemical quality parameters in lubricating oils according to ASTM E1655. Specific NIR methods were developed using a large sample set with different oils from several manufacturers. The unique combination of near-infrared spectroscopy with visible spectroscopy enabled the determination of additional quality parameters like the color number. The determined standard errors of cross-validation were found to be 0.21 mg KOH g -1 for acid number, 0.14 mg KOH g -1 for base number, 0.9 for color number, 0.0031 g cm -3 for density, 0.11 % for moisture, 0.54 cSt for viscosity at 100 °C, 11.7 for viscosity at 40 °C in cSt and 1.5 for viscosity index. The determined errors were comparable with errors of conventional test methods for lubricating oil. The developed methods were successful validated in an additional step by using an independent set of samples. In summary, Vis-NIR spectroscopy has been proven to be a rapid, timeand cost-saving method for quality control of lubricating oil. Keywords Quality control, near-infrared spectroscopy, total acid number, viscosity, moisture, total base number, density, viscosity index Kurzfassung Abstract * Dr. Alexander Kadenkin Elena Hagemann, M. Sc. Dr. Dave R. van Staveren Metrohm AG, 9100 Herisau, Switzerland T+S_4_2018.qxp_T+S_2018 05.06.18 11: 15 Seite 35 Test Method for Acid Number of Petroleum Products by Potentiometric Titration” [3]. This means that if the reference samples used for the development of the NIR calibration were analyzed by accepted primary methods, then the NIR solution can be used for the same type of analysis as the primary method like e. g. titration or viscosimetry. This approach is exemplified in the present work, which demonstrates the possibilities of Vis-NIR in quality control of lubricating oils. This analytical technique was successfully used for the simultaneous determination of color number, total acid and total base numbers, moisture content, density, viscosities at 40 °C and at 100 °C as well as viscosity index. 2 Experimental Materials A sample group of 260 liquid lubricating oils from various suppliers and in different stages of wear (fresh or old oils) were used in the present study. Furthermore different types of oils were used in order to increase the variance in the calibration set. These samples were provi- Aus der Praxis für die Praxis 36 Tribologie + Schmierungstechnik · 65. Jahrgang · 4/ 2018 However, the operation costs of the quality control laboratory for the analysis of lubricating oils and the capital expenditure investment in hardware can be dramatically reduced by using Vis-near-infrared (Vis-NIR) spectroscopy due to several reasons. Firstly, this analytical technique can be used for the simultaneous analysis of multiple physical and chemical parameters, providing results within one minute. This boosts the throughput of the QC lab and at the same time reduces the investment in analytical instrumentation. Secondly, it requires neither additional chemicals nor sample preparation, which significantly reduces the costs of each single sample analysis. Furthermore, due to the intuitive software and strict user rights, daily routine measurements can be even performed by inexperienced operators, without the risk of them accessing set-up menus or having them unintentionally change instrument settings. The procedure of calibration development using NIR is well described in different norms like ASTM E1655 [1] and ASTM D6122 [2]. This is applicable to different analytical applications especially “if the reference method used to obtain reference values… is an established ASTM method” [1] like e. g. ASTM D664 - 11a (2017) “Standard Table 1: Lubricating oil quality parameters with the concentration range covered by the calibration sample set, the conventionally used lab methods, and the corresponding ASTM norm. Parameter Range Number Reference Reference values of samples Method measured according to ASTM Norm Acid number 0.1 - 4.3 mg KOH g -1 131 Potentiometric titration D664 [3] Kinematic viscosity at 40 °C 5.2 - 252.4 cSt 210 Viscosimetry D445 [4] Kinematic viscosity at 100 °C 11.5 - 22.1 cSt 56 Viscosimetry D445 [4] Viscosity index 96 - 136 56 Calculation D2270 [5] Color number 1.8 - 8.0 153 Colorimetry D1500 [6] Moisture content 0.02 - 0.53 % 115 Coulometric KF titration D6304 [7] Base number 0.6 - 14.3 mg KOH g -1 86 Potentiometric titration D2896 [8] Density 0.881 - 0.956 g cm -3 139 Density meter D4052 [9] Table 2: Lubricating oil quality parameters with the concentration range covered by the validation sample set. Parameter Range Number of samples Acid number 1.7 - 3.5 mg KOH g -1 27 Kinematic viscosity at 40 °C 7.7 - 217 cSt 54 Viscosity index 96 - 111 54 Moisture content 0.02 - 0.5 % 48 Density 0.881 - 0.956 g cm -3 54 ded with reference values already determined for the different mentioned parameters. The range of different quality parameters as well as the number of reference values available for each parameter are summarized in Table 1. In addition, Table 1 shows the reference method and corresponding ASTM norms. A second set of 54 samples was used for external validation of the developed methods. This sample set was a hydraulic oil from a completely different manufacturer provided by a separate QC laboratory. The range of different quality parameters as well as the number of reference values available for each parameter are summarized in Table 2. In contrast to the calibration set on- T+S_4_2018.qxp_T+S_2018 05.06.18 11: 15 Seite 36 ly density, viscosity at 40 °C, viscosity index, acid number and moisture content were available as a reference values for this sample set. Sample analysis The samples were placed in 8 mm disposable glass vials without any further sample preparation and measured in transmission mode on a Metrohm NIRS XDS Rapid- Liquid Analyzer over the full Vis-NIR wavelength range of 400 - 2500 nm (Figure 1). Typical Vis-NIR spectra of lubricating oils are shown in Figure 2 on example of hydraulic oils. During the measurement the temperature was kept constant at 40 °C to provide the same measurement conditions as used for determination of kinematic viscosity at 40 °C. A delay time of 30 s was used in order to reach thermal equilibration of the samples prior to each measurement. The software package Vision Air 2.0 Complete was used for data acquisition, data management, and development of the quantitative methods. With this software package quantitative methods can be developed according to ASTM E1655 [1]. Method development The quantification methods for the four individual parameters were developed in Vision 4.1 (Metrohm chemometric software) using the algorithm of Partial Least Squares Regression (PLS) in accordance to ASTM E1655 [1]. Partial Least Squares Regression is a powerful mathematical algorithm used for the quantitative analysis of complex spectral data and is widely used in analytical chemistry, bioinformatics, statistics, neuroscience and anthropology. This algorithm reduces the complete spectra with 4200 data points as in the present case to a smaller set of uncorrelated components, a so called factors, which cover as much of the covariance of the original data as possible [10]. Afterwards, it performs least squares regression on these components, instead of on the original data. The number of factors is specific for each application and constituents and it should be chosen carefully in order to avoid a so called overfitting of the model. According to ASTM E1655 [1], one of the possibilities to select the application specific number of factors is the use of standard error (SE), which is defined as Where ŷ i is the predicted value, y i is the reference value and n is the number of samples used. There are three frequently used types of standard error for PLS regression: • Standard error of calibration (SEC), automatically calculated during the method development. • Standard error of cross-validation (SECV) frequently calculated during the model development, when using internal validation. • Standard error of prediction (SEP) calculated using an independent set of samples. Usually SEC is the error with the lowest value determined during the calibration model development. SECV typically has higher values and can be used for the determination of the optimal number of factors. Finally, SEP, which is determined using an independent set of samples, is higher. The SEP demonstrates the real performance of the developed model. Although the NIRS RapidLiquid Analyzer allows spectral data collection from 400 nm to 2500 nm, the PLS evaluation was performed using specific spectral regions in order to minimize the error of prediction. E. g. for the quantitative analysis of moisture only the specific water bands (900 - 950 nm, 1350 - 1450 nm, and 1850 - 1950 nm) were used. For further improvement of the analytical figures of merit, disruptive and uninformative spectral characteristics like e. g. spectral background was Aus der Praxis für die Praxis 37 Tribologie + Schmierungstechnik · 65. Jahrgang · 4/ 2018 Figure 1: A NIRS XDS RapidLiquid Analyzer was used to collect the spectral data of 260 samples in transmission mode covering the full Vis-NIR wavelength range of 400-2500 nm. Figure 2: Vis-NIR spectra of different hydraulic oils collected over the full wavelength range of 400 - 2500 nm. ! " #· $%& ' ( % ' ) * +',- . T+S_4_2018.qxp_T+S_2018 05.06.18 11: 15 Seite 37 removed using dedicated mathematical spectra pretreatments like 2 nd derivative. An example of pretreated spectra is shown in Figure 3 for the same set of hydraulic oils as shown in Figure 2. It demonstrates the successful removal of the baseline effect. The number of PLS factors, spectral range and mathematical pre-treatments used for different parameters are summarized in Table 3. 3 Results and Discussion Calibration The correlation coefficient R 2 as well as the standard error of cross-validation (SECV) for different parameters in the calibration set are summarized in Table 4. Furthermore, Table 4 provides information about the repro- Aus der Praxis für die Praxis 38 Tribologie + Schmierungstechnik · 65. Jahrgang · 4/ 2018 Figure 3: Vis-NIR spectra from Figure 2 pretreated with the 2 nd derivative over the full wavelength range of 400 - 2500 nm. Table 3: Wavelength range, mathematical pre-treatments and number of factors used for the development of PLS models for different parameters. Parameter Wavelength range in nm Mathematical pre-treatment Number of factors Acid number 1120 - 1370, 1450 -1860 and 1960 -2100 2 nd derivative 7 Base number 1120 - 1370, 1450 -1860 and 1960 -2100 2 nd derivative 7 Color number 410 - 800 1 st derivative 5 Density 1120 - 1370, 1450 -1860 and 1960 -2100 2 nd derivative 7 Moisture 1120 - 1450 and 1860 - 1960 2 nd derivative 6 Viscosity at 100 °C 1120 - 1370, 1450 -1860 and 1960 -2100 2 nd derivative 6 Viscosity at 40 °C 1120 - 1370, 1450 -1860 and 1960 -2100 2 nd derivative 8 Viscosity index 1120 - 1370, 1450 -1860 and 1960 -2100 2 nd derivative 6 Table 4: Coefficient of determination R 2 and standard error of cross-validation (SECV) for the calibration set as well as repeatability and reproducibility as mentioned in the corresponding ASTM norm [3-9]. Parameter R 2 SECV Repeatability and reproducibility as mentioned in the corresponding ASTM norm [3-9] Acid number in mg KOH g -1 0.968 0.21 Repeatability 0.016 mg KOH g -1 at 0.2 mg KOH g -1 equal to 8 %, reproducibility 0.031 mg KOH g -1 at 0.2 mg KOH g -1 equal to 15.5 % Base number in mg KOH g -1 0.998 0.14 Repeatability 3 - 24 % of mean, reproducibility 7 - 32 % of mean depending on the test method used Color number 0.679 0.90 Repeatability 0.5, reproducibility 1 Density in g cm -3 0.981 0.0031 Reproducibility 0.0019 - 0.034 g cm -3 depending on the sample used Moisture in % 0.849 0.11 Repeatability 0.03813 x (measured value) 0.6 , reproducibility 0.4243 x (measured value) 0.6 Viscosity at 100 °C in cSt 0.933 0.54 Repeatability 0.11 - 1.5 %, reproducibility 0.65 - 7.4 % depending on the sample type and temperature used Viscosity at 40 °C in cSt 0.972 11.7 Repeatability 0.11 - 1.5 %, reproducibility 0.65 - 7.4 % depending on the sample type and temperature used Viscosity index 0.945 1.5 Reproducibility is demonstrated on the example of sample analysis at two different labs with 78 (lab 1) and 80 (lab 2) for sample 1 and 144 (lab 1) and 145 (lab 2) for sample 2 T+S_4_2018.qxp_T+S_2018 05.06.18 11: 15 Seite 38 ducibility and repeatability as mentioned in the corresponding ASTM norms [3-9]. Since the samples in the present work stem from different QC laboratories, reproducibility should be taken into account during comparison of conventional methods with NIR method because this analytical figure of merit is the measure of the difference between two single and independent test results obtained by different operators working in different laboratories on identical test material. The analytical figures of merit for NIR spectroscopy summarized in Table 4 are characterized by a very low error, which is in most cases better than the reproducibility of the conventional reference method. This can be e. g. demonstrated on the SECV for the density determination, which is 0.0031 g cm -3 and therefore within the range of the reproducibility of 0.0019 - 0.034 g cm -3 described in ASTM 4052 [9]. The coefficient of determination R 2 for the plots of predicted values versus reference values is close to 1 and demonstrates excellent correlation between the results of both methods. Such plots are exemplarily shown in Figures 4 - 6 on examples of base and acid number as well as viscosity at 100 °C. Only in case of the color index with R 2 = 0.679 the correlation between reference and predicted values is not excellent. This relatively low correlation can be explained by the relatively worse repeatability and reproducibility of the used reference method (0.5 and 1 units respectively) [6]. Additionally, the ASTM color scale is a color scale with a limited resolution of 0.5 units. Validation Analytical figures of merit (coefficient of determination R 2 and standard error of prediction (SEP)) for the analysis of the validation set are shown in Table 5. As expected the SEP values are higher and R 2 are lower than in case of internal validation. This can be explained by statistical reasons since these samples were not included into model building. Furthermore, this sample set consisted of hydraulic oil from a different manufacturer, provided by a separate QC laboratory, and therefore it is not exactly the same type of oil as used during the calibration model development. Nevertheless, the resulted SEP values are in a repeatability range comparable with Aus der Praxis für die Praxis 39 Tribologie + Schmierungstechnik · 65. Jahrgang · 4/ 2018 Figure 4: Correlation plot of reference values from titration versus predicted values from Vis-NIR for the calibration set. The base number varies between 0.6 - 14.3 mg KOH g -1 . A high correlation is observable (R 2 = 0.998). Figure 5: Correlation plot of reference values from viscosimetry versus predicted values from Vis-NIR for the calibration set. The viscosity at 100 °C varies between 11.5 - 22.1 cSt. A high correlation is observable (R 2 = 0.933). Figure 6: Correlation plot of reference values from titration versus predicted values from Vis-NIR for the calibration set. The acid number varies between 0.1 - 4.3 mg KOH g -1 . A high correlation is observable (R 2 = 0.968). Table 5: Coefficient of determination R 2 and standard error of prediction (SEP) for the validation set. Parameter R 2 SEP Acid number in mg KOH g -1 0.769 0.24 Density in g cm -3 0.963 0.0034 Moisture in % 0.849 0.11 Viscosity at 40 °C in cSt 0.906 17.6 Viscosity index 0.822 1.6 T+S_4_2018.qxp_T+S_2018 05.06.18 11: 15 Seite 39 reference methods. Additionally, the linearity of correlation plots is still good, as shown in Figures 7 - 9 for the examples of density, viscosity index and moisture content. 4 Conclusion and outlook A method based on Vis-NIR spectroscopy for rapid simultaneous determination of 8 different chemical and physical quality parameters in lubricating oils from different suppliers and application fields was successfully developed in the current work. A total of 260 samples of oil from different suppliers and application fields served as a basis for this method. It was validated using an independent set of 54 samples, originating from a different manufacturer. The determined analytical figures of merit were comparable to the reproducibility of the conventional ASTM methods. When taking into account that the measurement of one sample takes 2 minutes (1 minute for the pipetting of the sample, 30 s for the thermal equilibration and 30 s the sample measurement) this approach offers several clear advantages over traditional methods. Most importantly, it saves costs and time and, secondly, it can be flexibly deployed either in the lab or atline/ online in a process environment. For this study we have on purpose selected a wide variety of different oils from different applications and manufacturers for our model development. Undoubtedly, additional improvements of the developed methods can be realized, when lubricating oils with a similar matrix, like e. g. hydraulic oils, are used. Further improvements of the methods can be accomplished by using samples with a narrow constituent range and by developing specific methods for lubricating oils from the same supplier. Acknowledgements The authors thank the cooperation partners for providing the samples and the corresponding reference values. References [1] ASTM E1655-05(2012), Standard Practices for Infrared Multivariate Quantitative Analysis, ASTM International, West Conshohocken, PA, 2012, www.astm.org [2] ASTM D6122-15, Standard Practice for Validation of the Performance of Multivariate Online, At-Line, and Laboratory Infrared Spectrophotometer Based Analyzer Systems, ASTM International, West Conshohocken, PA, 2015, www.astm.org [3] ASTM D664-11a(2017), Standard Test Method for Acid Number of Petroleum Products by Potentiometric Titration, ASTM International, West Conshohocken, PA, 2017, www.astm.org [4] ASTM D445-17a, Standard Test Method for Kinematic Viscosity of Transparent and Opaque Liquids (and Calculation of Dynamic Viscosity), ASTM International, West Conshohocken, PA, 2017, www.astm.org Aus der Praxis für die Praxis 40 Tribologie + Schmierungstechnik · 65. Jahrgang · 4/ 2018 Figure 7: Correlation plot of reference values from the density measurement versus predicted values from Vis-NIR for the validation set. The density varies between 0.881 - 0.956 g cm -3 . A high correlation is observable (R 2 = 0.963). Figure 8: Correlation plot of reference values from calculation of viscosity index versus predicted values from Vis-NIR for the validation set. The viscosity index varies between 96 - 111. A high correlation is observable (R 2 = 0.822). Figure 9: Correlation plot of reference values from Karl Fisher titration versus predicted values from Vis-NIR for the validation set. The moisture content varies between 0.02 - 0.5 %. A high correlation is observable (R 2 = 0.849). T+S_4_2018.qxp_T+S_2018 05.06.18 11: 15 Seite 40 [5] ASTM D2270-10(2016), Standard Practice for Calculating Viscosity Index from Kinematic Viscosity at 40 °C and 100 °C, ASTM International, West Conshohocken, PA, 2016, www.astm.org [6] ASTM D1500-12, Standard Test Method for ASTM Color of Petroleum Products (ASTM Color Scale), ASTM International, West Conshohocken, PA, 2012, www.astm.org [7] ASTM D6304-16e1, Standard Test Method for Determination of Water in Petroleum Products, Lubricating Oils, and Additives by Coulometric Karl Fischer Titration, ASTM International, West Conshohocken, PA, 2016, www.astm.org [8] ASTM D2896-15, Standard Test Method for Base Number of Petroleum Products by Potentiometric Perchloric Acid Titration, ASTM International, West Conshohocken, PA, 2015, www.astm.org [9] ASTM D4052-16, Standard Test Method for Density, Relative Density, and API Gravity of Liquids by Digital Density Meter, ASTM International, West Conshohocken, PA, 2016, www.astm.org [10] H. Wold, Estimation of Principal Components and Related Models by Iterative Least Squares. New York: Academic Press, 1966. Aus der Praxis für die Praxis 41 Tribologie + Schmierungstechnik · 65. Jahrgang · 4/ 2018 expert verlag GmbH: Wankelstr. 13, 71272 Renningen Postfach 20 20, 71268 Renningen Tel. (0 71 59) 92 65 - 0, Fax (0 71 59) 92 65 -20 E-Mail expert@expertverlag.de Vereinigte Volksbank AG, Sindelfingen BIC GENODES1 BBV, IBAN DE51 6039 0000 0032 9460 07 Postbank Stuttgart BIC PBNKDEFF, IBAN DE87 6001 0070 0022 5467 07 USt.-IdNr. DE 145162062 Anzeigen: Sigrid Hackenberg, expert verlag Tel. (0 71 59) 92 65 -13, Fax (0 71 59) 92 65-20 E-Mail anzeigen@expertverlag.de Informationen und Mediendaten senden wir Ihnen gerne zu. Vertrieb: Rainer Paulsen, expert verlag Tel. (0 71 59) 92 65 -16, Fax (0 71 59) 92 65-20 E-Mail paulsen@expertverlag.de Die zweimonatlich erscheinende Zeitschrift kostet bei Vorauszahlung im Jahresvorzugspreis für incl. Versand im Inland 189,- € (incl. 7 % MwSt.), im Ausland 198,- € * , Einzelheft 39,- € ; * (in der EU bei fehlender UID-Nr. zzgl. MwSt.); Studenten und persönliche Mitglieder der GfT erhalten gegen Vorlage eines entsprechenden Nachweises einen Nachlass von 20 % auf das Abo-Netto. Für Mitglieder der ÖTG ist der Abonnementspreis im Mitgliedschaftsbeitrag enthalten. Die Abonnementsgebühren sind jährlich im Voraus bei Rechnungsstellung durch den Verlag ohne Abzug zahlbar; kürzere Rechnungszeiträume bedingen einen Bearbeitungszuschlag von 3,- € pro Rechnungslegung. Abbestellungen müssen spätestens sechs Wochen vor Ende des Bezugsjahres schriftlich vorliegen. Der Bezug der Zeitschriften zum Jahresvorzugspreis verpflichtet den Besteller zur Abnahme eines vollen Jahrgangs. Bei vorzeitiger Beendigung eines Abonnementauftrages wird der Einzelpreis nachbelastet. Bei höherer Gewalt keine Lieferungspflicht. Erfüllungsort und Gerichtsstand: Leonberg expert verlag, 71272 Renningen ISSN 0724-3472 4/ 18 Tribologie und Schmierungstechnik Organ der Gesellschaft für Tribologie | Organ der Österreichischen Tribologischen Gesellschaft | Organ der Swiss Tribology Heft 4 Juli/ August 2018 65. Jahrgang Herausgeber und Schriftleiter: Prof. Dr.-Ing. Dr. h.c. Wilfried J. Bartz Mühlhaldenstr. 91, 73770 Denkendorf Tel./ Fax (07 11) 3 46 48 35 E-Mail wilfried.bartz@tribo-lubri.de www.tribo-lubri.de Redaktion: Dr. rer. nat. Erich Santner, Bonn Tel. (02 28) 9 61 61 36 E-Mail esantner@arcor.de Redaktionssekretariat: expert verlag Tel. (0 71 59) 92 65 - 0, Fax (0 71 59) 92 65 -20 E-Mail: expert@expertverlag.de Beiträge, die mit vollem Namen oder auch mit Kurzzeichen des Autors gezeichnet sind, stellen die Meinung des Autors, nicht unbedingt auch die der Redaktion dar. Unverlangte Zusendungen redaktioneller Beiträge auf eigene Gefahr und ohne Gewähr für die Rücksendung. Die Einholung des Abdruckrechtes für dem Verlag eingesandte Fotos obliegt dem Einsender. Die Rechte an Abbildungen ohne Quellenhinweis liegen beim Autor oder der Redaktion. Ansprüche Dritter gegenüber dem Verlag sind, wenn keine besonderen Vereinbarungen getroffen sind, ausgeschlossen. Überarbeitungen und Kürzungen liegen im Ermessen der Redaktion. Die Wiedergabe von Gebrauchsnamen, Warenbezeichnungen und Handelsnamen in dieser Zeitschrift berechtigt nicht zu der Annahme, dass solche Namen ohne Weiteres von jedermann benutzt werden dürfen. Vielmehr handelt es sich häufig um geschützte, eingetragene Warenzeichen. Die Zeitschrift und alle in ihr enthaltenen Beiträge und Abbildungen sind urheberrechtlich geschützt. Mit Ausnahme der gesetzlich zugelassenen Fälle ist eine Verwertung ohne Einwilligung des Verlags strafbar. Dies gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. Entwurf und Layout: Ludwig-Kirn Layout, 71638 Ludwigsburg Impressum Impressum T+S_4_2018.qxp_T+S_2018 05.06.18 11: 15 Seite 41