Partially Fritted Glazing Optical & Thermal Property Testing Procedures
Partially fritted glasses are different from conventional glasses in two aspects: a) they are not uniform and b) the fritted parts are not specular.
The following practices are employed in the lab to test partially fritted glasses:
- The clear part and fritted part are tested separately;
- The fritted sample is tested in the same way as a specular sample;
- Area-weighted averaging is performed to get the overall results.
The practices are in line with Annex C of the EN 410 standard, but deviate from the other NFRC/ISO/EN standards.
This article aims to explain the step-by-step procedures for partially fritted glass optical & thermal property testing.
- Testing procedures
- Step 1. Preparation of test samples
- Step 2. Individual part/sample testing
- Step 3. Area weighted averaging
- Step 4. Report writing
- More information
Step 1. Preparation of test samples
There are two sample preparation methods:
- Method A: clear part and fritted part on the same sample
- Method B: one clear sample and one fritted sample
Method A: clear part and fritted part on the same sample
The area sampled by the instruments in a measurement is around 1 inch (25 mm) in diameter.
For a partially fritted glass, if there are continuous patches larger than 1 inch (for both the clear part and fritted parts), the glass can be directly tested. Shown below is an example.
For such samples, the clear part and fritted part are measured separately.
Method B: one clear sample and one fritted sample
In practice, most partially fritted glasses are not with large continuous patches. Shown below is a typical partially fritted glass without large continuous patches.
The sample above cannot be directly tested, because the instrument sampling beam covers both the clear part and fritted part. The result variation is large, depending of the fraction of clear/fritted part covered.
In order to address this issue, two samples, one representing the clear part and one representing the fritted part, need to be prepared. The principle is illustrated below. The two samples shall be identical, except that one is without frit and one is with frit.
For such sample pairs, the clear sample and fritted sample are measured separately.
Step 2. Individual part/sample testing
The individual parts (Method A) or individual samples (Method B) are tested separately, in the same way as conventional single glazing glasses, with the following procedures:
- Cleaning and labelling
- Spectral reflectance measurement with FTIR
- Spectral transmittance/reflectance measurement with UV/VIS/NIR
- IGDB file generation with IGDB@OTM
- Result calculation with LBNL Optics/WINDOW or Glazing@OTM
Refer to the article on single glazing testing procedures for more details.
As explained in the beginning of this article, the fritted parts are not specular. Some additional considerations or exceptions are discussed below.
Specular and non-specular materials
Conventional glasses are specular materials. Shown below are the optical paths through a specular glass. There are multiple inter-reflections in the glass. If the incident light is from one direction, the reflected or transmitted light is in one direct too. The instruments can measure such specular glass transmittance/reflectance accurately.
Glasses with frits are non-specular materials. Shown below are the optical paths through a non-specular glass with frit. Despite that the incident light is from one direction, the reflected or transmitted light is in all directions, rather than in one direction. This is called non-specular or diffuse transmission/reflection.
For the multiple inter-reflections in the glass, some light reflected in oblique angles (for example, the long green arrows shown below) may exit from the glass edge, without being detected by instrument.
Additional uncertainty for non-specular glass with frit
As explained above, for a non-specular glass with frit, some inter-reflected light may exit from the glass edge, without being detected by instrument. Due to this additional effect, the measured transmittance/reflectance would be lower.
The magnitude of this additional uncertainty is dependent on a few factors:
- Instrument design
- Glass thickness
- Position of frit
For instrument design, instruments with large size integrating spheres are with smaller uncertainty. For example, 270 mm integrating spheres are better than 150 mm integrating spheres. At OTM, 150 mm integrating sphere is used and the uncertainty is relatively larger.
For glass thickness, thin glasses are with smaller uncertainty than thick glasses.
For position of frit, the uncertainty is smaller when the frit layer is closer to the detector. For the example discussed above, the uncertainty in the right side reflectance measurement is smaller than that in the left side reflectance measurement.
According to the inter-laboratory comparison (ILC) organized by LBNL in 2011, the result variations for fritted glass samples were large. The practices introduced in this article are a workaround with compromised accuracy.
Emissivity test of exposed frit layer
For the fritted part/sample, the frit layer could be embedded in a laminated glass or exposed as the external layer. Shown below are the two scenarios.
The standard emissivity measurement instrument for glass, FTIR, is for specular materials only. As glass is opaque to far infrared radiation, there is no transmission of far infrared radiation through the glass. Only the top layer of a sample needs to be specular.
For a fritted glass with embedded frit layer, the top surface is still a conventional glass and is specular. The FTIR instrument can be used to measure its emissivity as usual.
For a fritted glass with exposed frit layer, the side without frit can be measured with FTIR as usual, but the side with frit cannot. For this scenario, the side with exposed frit layer is measured with the emissometer in the lab, according to ASTM C1371. Shown below is the emissometer used at OTM.
Step 3. Area weighted averaging
On completion of step 2, there are two sets of results, one for the clear part and one for the fritted part. Area weighed averaging is then performed to get the over performance results of the partially fritted glass.
In the area-weighted averaging, the coverage of frit is declared by the client, but not measured in the lab. It is possible to declare multiple coverages.
MS Excel is used for the area weighted averaging and shown below is an example.
Step 4. Report writing
The last step is to write the test report according to the software calculation results.
Shown below is an example result table based on the results shown above for 25% frit coverage.
In summary, for a partially fritted glass, its clear part and fritted part need to test separately and the overall performance is calculated with area-weighted averaging. The fritted part is tested in the same way as a clear part, with some additional considerations or exceptions.
If you are interested in more in-depth information, please continue reading this article. If your concerned questions are not explained, please feel free to leave a comment at the end of this page. This article will be reviewed and updated regularly.
What are the processing time and lead time?
The typical processing time for testing a partially fritted glass is around 1.5 times of that for testing a single glazing and similar to that for testing a DGU, i.e. 6 – 9 hours.
The typical lead time is 3 working days. We may schedule the activities on different days, due to operational reasons.
Can the partially fritted glass be part of a DGU?
For a DGU glass, its outer pane or inner pane can be partially fritted. In that case, the partially fritted pane need to the sampled and tested in the same way as described in this article. The DGU properties are calculated before area-weighted averaging, i.e. two sets of DGU results, one with clear pane and one with fritted pane, are calculated and then area-weighted averaging is performed on the two sets of DGU results.
The amount of work required to test a DGU glass with a partially fritted pane is equivalent to testing a triple glazing system.
In some special cases, both the outer pane and inner pane are partially fritted. In that case, four parts or samples (clear/fritted and inner/outer) need to be tested.
How accurate are the measurement results for partially fritted glasses?
According to the inter-laboratory comparison (ILC) organized by LBNL in 2011, the result variations for fritted glass samples were large.
The practices described in this article are a compromised workaround. It allows quantitative measurement results with existing commercial instruments and with reasonable cost.
There is no well-developed test standard for partially fritted glasses at the moment of writing. The practices are in line with Annex C of the EN 410 standard, but deviate from the other NFRC/ISO/EN standards.
How to test the daylight reflectance of partially fritted glasses?
For the daylight reflectance requirements by BCA Singapore, the test methods for partially fritted glasses are not clearly stated in the officially published documents.
Below are our recommendations to our clients:
- Test the clear part and the fritted part of a partially fritted glass separately
- As the clear part is specular, we will report its visible light reflectance only
- As the fritted part is non-specular, we will report its total/diffuse/specular daylight reflectances.
For the fritted part, it is necessary to highlight that the specular daylight reflectance result is not affected by the additional uncertainty described above.
The total and diffuse daylight reflectances would be lower, due to some light leakage through the glass edge. As the specular daylight reflectance is calculated as the difference between the total and diffuse daylight reflectances, the errors are canceled out. Therefore, the accuracy of the specular daylight reflectance result is not affected.
Work with the expert in Optical & Thermal Measurement solutions.