Update on Taber Abrasion Test
Mr. Dümmler clarified that the aim of the exercise was to compare the Taber vs. the Daiwa wheels.

All the wheels came from the same lots. The results were analysed by the US experts. But there is some doubt about the analysis performed, hence it was decided to re-check the evaluation. This was done by Mr. Buckel (Bayer).

Mr. Buckel presented IGPG-05-04. 1st part is status of ISO work. 2nd part of the presentation is a study on the windscreens from KRD.

During the presentation, there was a debate about the nature of the round robin performed by ISO and the relevant information it could bring:
- The nature of the wheels was not delivered by Taber and Daiwa
- The procedure was well defined, but the there was still a high dispersion in the results
- The aim of the round robin is not
  o to determine the proper haze value for R43
  o to determine whether the Taber test is appropriate, rather
- The aim is to assess the reproducibility and repeatability of the Taber test.

Conclusion: ISO has finalized a test description for glass, but test description for plastic material is ongoing at ISO.

Concerning the study of the windscreens of police vehicles provided by KRD, it was revealed that, while the mileage has an influence on the haze, some other parameters are of some influence.

The analysis of the images tend to show that the Taber test is not relevant for plastic windscreens, and the car wash test + sand drop test best approach the real world abrasion scheme.

It was mentioned that haze measurement could be a wrong measurement of the discomfort felt by the user: little pits could cause the same haze as long scratches, while the latter are less acceptable to the
normal driver than the small pits.

However, it was considered that keeping haze as the criterion would be reasonable in order to keep the discussions simple. In addition, the round robin tests performed to date showed that the haze measurement is reproducible and repeatable.

It was suggested to find a reference material, then use a correction factor. This reference material should be coated plastic material (to avoid too big correction factors) and easily available.

Abrasion wheel feasibility study report
J Mr. Hara presented the document IGPG-05-05e.

The experts were informed that, as a conclusion of this study, there is in J an attempt to reach the 2% delta haze value.

There was however some concern in increasing the particle size. If such big particle happens to remove from the wheel, there will be no abrasive material in that location at all, i.e. all the results could be altered.

Some experts found the principle of the FGMAJ approach not relevant.

A debate took place about the choice of a proper abrasion test and the reasons in favour or against changing the test. There is a need to demonstrate the abrasive process of each material (i.e. glass vs. plastics), in order to assess whether the Taber test is convenient for each. Another concern was the lack of information about the characteristics of the wheels. An advantage of the taber test is that it is a well-recognized test.

French study on Taber test:
A debate followed the presentation.
- should the conclusions be verified, then the problem of the wheels etc can be dropped, because making the influencing parameters tiny permits of course to achieve narrower deviations
- even with the best test protocol, the human factor remains a parameter
- the study does not provide any indication on whether the value of 2% is the relevant one.

The question raised as to whether not using plastic glazing since it reaches the same results as the glass on the field. The reason was that perhaps the plastic glazing can nowadays achieve the qualities of the glass, but cannot fulfil the Taber test performance requirements.

It was considered necessary to get the data from the field test in order to have an assessment of the plastic glazing behaviour in the field.

A fundamental question raised: whether the group does not miss its final target by gathering and accumulating endless data, and whether an acceptable compromise could be reached within the informal group, with unanimity, for a test method (whether Taber or another).

It was suggested to federate the forces in order to make plastic glazing win the battle of introduction in the market.

A fast tour de table revealed that at least six experts are in favour of dropping the Taber test for plastic glazing for windscreens and side windows in front of the driver’s eyes, i.e. elaborating a new test for them.

Keep Taber	Neutral	Drop Taber
PSA Renault AGC NSG FGMAJ JASIC	Sabic Webasto Altuglass JAMA Freeglass Vinçotte	Evonik John Deere Bayer Daimler KRD Momentive Polyplastic MPA

Some companies were in favour of changing their mind should it be proven that the Taber test is relevant for plastic glazing.

France confirmed that their study is still on-going, but did not expect any change in the conclusions.

MPA was of the opinion that the problem could be formulated as follows:
- even with improved discrepancy, the value of 2% would be too difficult to achieve
- the real question would be whether the wiper test would be relevant, in combination with sand drop and Amtec-Kistler tests.

PSA was keen that the coating itself be discussed.

Bayer was keen that the test best represents the wear that can be seen on the real field; it was pointed out that such approach would delay the introduction of plastic windscreen by at least 10-15 years.

Automotive Industry was reluctant in investing resources in a new test with no guarantee that it would be beneficial.

FGMA-J was of the opinion that 2% delta haze could be put into question again, and that new value of haze or criteria could be found.

The group proceeded with another tour de table:
The question was as follows: Are you ready to accept the wiper test (in combination with sand drop and Amtec-Kistler on three different samples), as an ALTERNATIVE to the Taber test in R43, if the group provides “sufficient evidence” that the wiper, the sand drop and Amtec-Kistler tests are together relevant for plastic windshields (glazing in front of the driver’s eyes)?
Ready: Unanimity but Vinçotte
Neutral: Vinçotte
Not ready: Nobody

Examples of definitions for “sufficient evidence”:
“sufficient evidence” means:
- 200 000 km mileage, wear in real field must be comparable to wear after performing the tests.
- Good reproducibility of the test, good correlation to reality by image analysis and perhaps other means
- Definition to be set up by the wiper TF

Another informal group to GRSG could be established to assess whether the Taber test is representative of reality.