In the manufacturing of heat-tempered safety glass, temperature and stress variations influence the optical properties of the glass. These stress differences alter the refractive indices of the glass, appearing as shadow-like anisotropy patterns under polarized light. For instance, anisotropy patterns are not visible in diffuse light. However, when direct sunlight (e.g., during sunset) strikes a façade or when light is reflected from a nearby water surface, the light becomes polarized, making the effect more noticeable. The same phenomenon occurs when polarized sunglasses are used.

Measuring Anisotropy – Definition
In heat-tempered glass, light refracts differently along various axes when passing through affected surfaces or points (a phenomenon known as birefringence). This results in a phase difference between light waves, referred to as “optical retardation.” Using machine vision, optical retardation in the glass can be visualized and measured in nanometers (nm).
- In 2018, Apple (Cupertino, CA) initiated a group to establish norms and standards (US ASTM Group).
- US ASTM Group: Defined accurate and repeatable measurement methods. (Members: Arcon, LiteSentry, Softsolution, and Viprotron.) The result was the publication of ASTM C190-21 in 2021.
- Germany/Europe: Criteria for determining the quality level of tempered, uncoated glass based on measurement results were established. DIN Spec 18198 was published in May 2022.
The specification focuses primarily on monolithic structures. However, the evaluation methods described can also be applied to other glass structures, such as multi-layer laminated glass, laminated safety glass, and insulating glass units. When multiple layers of tempered glass are stacked, anisotropy effects can become more pronounced. Additionally, other factors, not yet fully understood, may also influence these cases.
ASTM C190-21 – Methods for Measuring Anisotropy
The device measures the optical retardation of an area based on the principles of digital photoelasticity. It must produce repeatable and direction-independent measurement results. The optical components should be adapted to the wavelength range of the light source, and imaging can be performed using a camera or a linear sensor


The image below is a false color image captured by the Viprotron anisotropy scanner, showing a glass pane where the intensity of anisotropy distribution varies across different regions. Under normal conditions, without polarized sunglasses, the human eye can only detect effects exceeding a threshold value of 50–75 nm.

DIN SPEC 18198:2022–05– Measurement and Evaluation Methods
DIN SPEC 18198 provides guidelines for distinguishing between “good” and “poor” uncoated glass. It specifically identifies the main pane area as separate from the edges and drilling holes, which are critical areas for the formation of anisotropies (iridescence). According to DIN, quality evaluation is performed only on the main pane area (2). The edges (1) and holes (3) are excluded, as they are typically covered by frames or fasteners.
For the actual quality assessment (A = good, B = acceptable, C = poor), DIN specifies two methods, both of which take glass thickness into account.

Evaluation Using Method A – 95% Quantile Value
The calculation is based on all measured retardation values and the determination of their empirical cumulative distribution function. The 95% quantile value means that 95% of the measured retardation values are smaller than this specified value. The quantile value is expressed in nanometers (nm).

Evaluation Using Method B – Isotropy (%) with a Threshold of 75 nm
This method evaluates the proportion of the glass surface area (as a percentage) that remains below the set isotropy threshold of 75 nm (isotropy being the opposite of anisotropy). For example, if only 55% of the surface area of an 8 mm glass pane falls below the 75 nm threshold, the glass quality can be deemed “poor.”
The results of this method always depend on the selected isotropy threshold—the higher the threshold, the “better” the pane will be rated.

Optimizing the Tempering Process with an Anisotropy Scanner
By optimizing the tempering process, it is possible to reduce stress differences and improve the optical quality of glass. However, this requires precise adjustments to various parameters, such as heating, oscillation, movement paths, nozzle and roller spacing, and the configuration of the furnace and chiller. Within the same batch, glass panes may exhibit different types of anisotropy patterns, such as “panther,” “diamond,” “zebra,” or longitudinal patterns. Without advanced machine vision, optimizing these variables is nearly impossible.
To address this need, Viprotron has developed the 5D-Temper Scanner, which can measure not only anisotropy but also white haze, scratches, waviness, and edge kink on glass.

Heating and cooling processes can be visualized, providing operators with a real-time view of the necessary adjustments. Parameter fine-tuning begins with corrections based on the operator’s expertise, and the effects of these adjustments can be evaluated with the next batch.
Isotropy percentage values depend on the selected threshold, but averages become key metrics. However, measurement results with a tolerance exceeding 10 nm are only indicative and do not leverage the full potential of the scanner. The Viprotron scanner achieves a precision level with a measurement tolerance of less than ±7 nm.
Experiences with Anisotropy Scanner Installations
The quality of anisotropy and white haze often does not meet acceptable standards immediately after the installation of the scanner. In the initial measurements, isotropy values may remain at around 60% (with a 60 nm retardation threshold). However, within a week, isotropy values typically rise to 80–90%, thanks to adjustments made to the tempering process based on the scanner’s findings.
A month after installation, anisotropy patterns can be significantly improved or even eliminated entirely.

Sources for additional information:
- ASTM C190-21 Standard Test Method for Measuring Optical Retardation in Flat Architectural Glass
- DIN SPEC 18198:2022-05 Measurement and evaluation methods for optical anisotropic effects in thermally toughened glass
- Saverio Pasetto: „Anisotropy as a defect in the U.K. architectural float heat-treated glass” (University of Bath, 2014)
- Feldmann, Schuler et al: „Methoden zur Erfassung und Analyse von Anisotropien bei thermisch vorgespannten Glasprodukten“ (Glasbau 2017)
- Dr. Decourcelle, Kaminski & Serruys: „Controlling Anisotropy” (Glass Performance Days, Tampere Finland 2017)