T 0760/05 03-06-2008

1. Admissibility of new experimental report

The new experimental report was submitted by the appellant with its letter dated 30 April 2008. It was thus filed only about 5 weeks before the date of the scheduled oral proceedings and thus at the very end of the period as given in the Board's communication annexed to the summons to oral proceedings dated 28 January 2008. Hence it was filed at a very late stage of the appeal proceedings, with the only excuse that the experiments were performed in reply to the Board's communication.

1.1 Firstly, the Board wishes to note that the absence of any particular embodiment of the invention in the patent in suit as well as the lack of any test results was already noted in the decision under appeal. Thus, appropriate test results should have been filed with the appeal.

1.2 Secondly, if an experimental report is submitted as a reaction to a communication of the Board annexed to the summons to oral proceedings, in principle sufficient time should remain for the adversary to study it and perform counter experiments.

Filing an experimental report shortly before the date indicated by the Board as ultimate date for filing submissions runs counter to this principle (see e.g. T 375/91 of 17 November 1994, unpublished in OJ EPO and T 569/02 of 2 June 2004, unpublished in OJ EPO).

The report is consequently late-filed. The Board therefore has restricted itself to a prima-facie relevance test to determine whether there would be a case for admitting it and postponing the oral proceedings as a consequence.

1.3 The report discloses grinding performance results of abrasive disc products made according to three experiments: the first disc does not comprise any non-abrasive particles besides the abrasive particles which have an average aspect ratio of 2:1 with 56% having an aspect ratio of larger than 2:1, the second and third discs each comprise glass beads as non-abrasive particles to an amount of approx. 20% of the abrasive grain weight, the second disc presenting an average largest dimension of said glass beads of about 90.5% of the average largest dimension of the abrasive grits and the third disc presenting an average largest dimension of said glass beads of about 52.4% of the average largest dimension of the abrasive grits.

1.3.1 The description of said experiments does not reveal all the details of how said three discs are made. For example, the amounts of make coat and size coat or the amount of grinding adjuvant are not specified. It is also nowhere stated that the said amounts were the same for all three discs. To the contrary, from the passage "After that, glass beads were electrostatically applied, when needed, filling the spaces between the grains" (see the new experimental report, page 1, sixth paragraph) it could be concluded that besides said non-abrasive particles there was a further difference in the structure of the second and third discs containing glass beads, of which the performance was then compared.

1.3.2 The Board considers that, due to said lacking details, it is not possible to the appellant to repeat these experiments in order to verify them.

1.3.3 The first coated abrasive disc not containing any non-abrasive grits does not correspond to the closest prior art D1 and is thus not relevant at all as comparative example.

The other two coated abrasive discs containing 20% by weight non-abrasive particles, the first one being in accordance with D1 while the second one falls within the scope of claim 1, however, do not allow to deduce that the features of claim 1, e.g. the amount of abrasive grits "at least 25% of which have an aspect ratio greater than 2:1", the non-abrasive particles in an amount of "from 5 to 50% by weight, based on the abrasive grit weight" or having "an average largest dimension that is less than 75% of the average largest dimension of the abrasive grits", and "an outer layer comprising a grinding adjuvant", are actually determinative for the alleged effects.

Any test report made in order to prove such criticality should have comprised more examples and particularly their parameters should have been selected accordingly to be slightly inside and/or slightly outside the ranges and values as specified in claim 1 of the patent in suit in order to convincingly show an effect due to these specific values.

Consequently, the Board considers that the single example and the two comparative examples according to the new experimental report were also not made in accordance with the principle established for comparative tests according to the established case law, i.e. that the comparison with the closest state of the art must be such that said effect is convincingly shown to have its origin in the distinguishing feature(s) of the invention (see Case Law of the Boards of Appeal of the European Patent Office, 5th edition 2006, I.D.9.8).

1.4 In its communication annexed to the summons (see point IV above) the Board had also advised both parties to take note of the amended Rules of Procedure of the Boards of Appeal (RPBA), in particular Article 13 according to the new RPBA in force as from 13 December 2007 which relates to amendments to a party's case after the first exchange of the grounds of appeal and the reply of the respondent. The grounds of appeal were filed with letter of 23 August 2005, the reply of the respondent questioning inventive step in respect of the effects allegedly achieved dates from 8 March 2006. The new test report has been filed more than 2 years later, with letter of 30 April 2008.

Since the new experimental report is considered not to be relevant for the question of inventive step of the subject-matter of the independent claims 1 and 12, particularly because the parameters of the three examples do not reflect the critical features of said claims (see point 1.3.3 above) the Board has decided not to admit the new experimental report into the proceedings.

There was therefore no need to postpone the date for the oral proceedings.

2. Novelty (Article 54 EPC)

The respondent's arguments that the subject-matter of claim 1 as granted would lack novelty over D1 for not meeting the requirements of a selection invention cannot be accepted for the following reasons:

2.1 D1 discloses a coated abrasive article which comprises a backing and at least one binder. The abrasive coating comprises diluent particles and shaped abrasive particles which are bonded to said backing via said binder and may comprise an overlying size coat (see figures 3 and 7; claims 26 to 29; page 10, lines 22 to 30). The diluent particles can comprise (1) a plurality of individual abrasive particles bonded together by an adhesive to form an agglomerate, (2) a plurality of individual non-abrasive particles bonded together by an adhesive to form an agglomerate, (3) a plurality of individual abrasive particles and a plurality of individual non-abrasive particles bonded together by an adhesive to form an agglomerate, (4) individual non-abrasive particles, (5) individual abrasive particles, or (6) combinations of the foregoing (see claims 1 and 13). The shaped abrasive particles can have shapes which can be characterized as thin bodies having faces of triangular, rectangular, including square, circular or other geometric shape and the ratio of the length of the shortest facial dimension of an abrasive particle to its thickness is at least 1:1, preferably at least 2:1, more preferably at least 5:1, and most preferably 6:1; they can have shapes that can be characterized as rods with the ratio of length to maximum cross-sectional dimension being at least 1:1, preferably 2:1, and most preferably at least 3:1 (see page 4, lines 20 to 29; page 8, lines 48 to 50; page 9, lines 1 to 3 and lines 27 to 30). The thickness of said shaped particles preferably ranges from about 25 µm to 500 µm (see page 9, lines 24 and 25). In general the ratio of the size of the shaped abrasive particles to the size of the diluent particles is in the range from 2.5:1 to 0.5:1 and the size of the diluent particles ranges from about 50 to about 1500 µm, preferably from about 100-1200 µm; and "it is preferred that the diluent particles and the shaped abrasive particles be of approximately the same particle size range" (see page 10, lines 14 to 21). The volume ratio of shaped abrasive particles to diluent particles can vary from 95:5 to 5:95, typically from 30:70 to 70:30, and preferably from 40:60 to 60:40 (see page 11, lines 24 and 25).

2.1.1 Thus the general teaching of D1 is silent with respect to any average largest dimension of the non-abrasive particles being less than 75% of the average largest dimension of the abrasive grits (particles), let alone that at least 25% of the abrasive grits should have an aspect ratio greater than 2:1. Although the average largest dimension is comprised somewhere in the range between the smallest and the largest size of the specified size range, it can be close to the largest size of the size range due to a possible particle distribution. Hence it is not conclusive that the requirement of claim 1 of "an average largest dimension that is less than 75% of the average largest dimension of the abrasive grits" is fulfilled. The same conclusion is valid with respect to the aspect ratio of the abrasive grits which according to D1 is at least 1:1.

Furthermore, the general disclosure of D1 does not teach to incorporate from 5-50% by weight, based on the abrasive grit weight, of non-abrasive particles into the abrasive layer, but only of 95:5 to 5-95% by volume.

Consequently, the general disclosure of D1 cannot be considered to be novelty destroying. The same holds true with respect to the specific embodiments of D1 for the following reasons.

2.1.2 D1 discloses only in the context of the examples a more specific procedure for making shaped abrasive particles (see page 11, lines 40 to 42). By using said procedure triangular-shaped particles were made from a gel-dispersion in a mould having dimensions of the mould cavities of 0.29 cm (= 2900 µm) on each side and 0.05 cm (= 500 µm) in depth wherein said dispersion underwent substantial shrinkage during drying. By extrusion of the gelled dispersion rods were produced which during drying broke into lengths and were screened to size. The dimensions of the dried rods were about 0.6 mm diameter (= 600 µm) by about 0.6 to 2.4 mm length (= about 600 to 2400 µm), with the median length being about 1.6 mm (= 1600 µm)" (see page 12, lines 5 to 15). These (unfired) rods were then fired and collected (see page 12, lines 16 to 24).

The dimensions of the fired triangles or rods are nowhere specified in D1. Experimental Report D8 - which discloses measurements of average lengths and aspect ratios which are stated to be based on the shaped abrasive particles used in the examples of D1 -represents a separate, a second, document which disclosure is neither comprised in nor can it be considered as incorporated into the disclosure of D1 for novelty purposes.

In this context it is remarked that the fired triangular abrasive particles according to said general procedure should have a largest dimension smaller than 290 µm due to said described shrinkage (the examples 1 to 18 of D1 only specify the incorporation of "triangular-shaped particles" without mentioning any grade number thereof) whereas an average length of 1.44 mm (= 1440 µm) is specified in D8. Likewise an average length of 1.14 mm (= 1140 µm) is specified in D8 for the shaped abrasive rods - which are there designated "G-36 Rods" - which value, however, does not fit with an average grain size of 710 µm as specified in D4 for Grade 36 (see column 3, lines 9 and 10) but would fit to the rods described in the context of said general procedure for making shaped abrasives.

2.1.3 According to the procedure for making diluent particles the diluent particles referred to as "DP I" according to the examples comprised [in wt.%] 35.5 cured resole phenolic resin, 61.1 sodium cryolite, 1 wood pulp and 2.4 glass bubbles. These particles "were screened to a size range of about 589 to 1350 µm, such that they passed through a 16 mesh stainless steel screen, but were retained on a 34 mesh stainless steel screen" (see page 12, lines 34 to 38; page 15, lines 57 and 58).

Thus D1 does not specify any average particle size of said "DP I" particles.

2.1.4 The same conclusion holds true for the diluent particles according to example 18 which "were screened to a size range of about 297 to 710 micrometers, such that they passed through a 25 U.S. standard screen, but were retained on a 50 U.S. standard screen" and for example 16, which additionally does not clearly specify the particle size of the diluent particles "190 g/m**(2) of DP I" which thus could be that according to the described general procedure, i.e. 589-1350 µm.

2.1.5 The above "average particle size" conclusion holds also true for the abrasive particles of examples 16 and 18 which "consisted of about 480 g/m**(2) of grade 36 rods" (see page 15, line 57 to page 16, line 1 and lines 5 to 9). There exists no further description in D1 of the size range or the aspect ratio of said "grade 36 rods". Therefore the aspect ratio of these rods is also not known. As already mentioned above, said "grade 36" corresponds - according to other patents assigned to the respondent, such as D4 - to an average grain size of 710 µm (see D4, column 3, lines 9 and 10) so that - taking account of said diluent particle size range of 297 to 710 µm - it is not credible that the requirement of claim 1 of "less than 75% of the average largest dimension of the abrasive grits" is conclusively fulfilled. In this context it is also considered that according to D1 "it is preferred that the diluent particles and the shaped abrasive particles be of approximately the same particle size range" (see page 10, lines 20 and 21) so that, also for this reason, it is not conclusive that the aforementioned requirement is actually met.

2.1.6 Said 190 g/m**(2) non-abrasive DP I in combination with 480 g/m**(2) grade 36 rods of example 16 and said 130 g/m**(2) DP I and 480 g/m**(2) grade 36 rods according to example 18, respectively, result in weight ratios of 38,6% and 27,1%, respectively, thus meeting the weight ratio requirement of "from 5 to 50% by weight" of claim 1.

2.1.7 Moreover, the high likelihood - as argued by the respondent - that 40% of said fired abrasive rods had an aspect ratio of at least 2:1, since such a ratio is preferred according to D1 (see e.g. claim 8) is not sufficient with respect to the - according to the longstanding practice of the Boards of Appeal - required standard of a clear and unambiguous explicit and/or implicit disclosure in a document in order that the document is considered to be novelty destroying. Likewise it is clear that this specific aspect ratio of at least 2:1, taken from another passage in the specification of D1, cannot be combined with the specific embodiments according to the examples 16 and 18.

2.1.8 According to the procedure in D1 for making coated abrasive articles the size coat for making the examples comprises 68 wt.% cryolite, i.e. it comprised a grinding aid (see page 12, lines 51 to 55).

2.1.9 Thus the subject-matter of claim 1 differs from the coated abrasives according to examples 16 and 18 of D1 in that at least 25% of the abrasive grits have an aspect ratio greater than 2:1 and that the non-abrasive particles have an average largest dimension that is less than 75% of the average largest dimension of the abrasive grits.

Consequently, the subject-matter of claim 1 as granted is novel with respect to the disclosure of D1.

2.2 Since there is no document on file which discloses a coated abrasive product having all the features of claim 1 as granted the Board considers that the subject-matter of claim 1 as granted is novel (Article 54 EPC).

3. Inventive step (Article 56 EPC)

3.1 D1 represents the closest prior art for disclosing a coated abrasive article comprising a backing and at least one binder and an abrasive coating comprising diluent particles and shaped abrasive particles which are bonded to said backing via said binder, preferably using electrostatic coating (see figures 3 and 7; claims 26 to 29). The aspect ratio of the shaped abrasive particles is at least 1:1, preferably at least 2:1 and most preferably 6:1 (see page 9, lines 1 to 3) and if these particles are elongated, such as rods, their aspect ratio is more preferably at least 2:1 and most preferably at least 3:1 (see page 9, lines 27 to 30). Preferably the shaped abrasive particles are coated onto the make coat in an electrostatic field so that the majority thereof is oriented, i.e. with the longest dimension being perpendicular to the plane of the backing; and the applied size coat contains sodium cryolite, i.e. contains a grinding aid (see page 10, lines 46 to 55; page 12, lines 40 to 55; claim 29; figures 3 and 7). The ratio of the size of the shaped abrasive particles to the size of the diluent particles is in general in the range from 2.5:1 to 0.5:1 (see page 10, lines 14 to 21). Thus the size range of the diluent particles is 40-200% of the size of the abrasive particles. It is preferred that the diluent particles and the shaped abrasive particles have approximately the same particle size range (see page 10, lines 20 and 21).

D1 aims to provide coated abrasive articles containing both abrasive particles having specified shapes and diluent particles which provide a cut that compares favourably with the cut provided by coated abrasive articles containing only high quality abrasive grits, but can be produced at lower costs than the latter (see page 3, lines 2 and 3 in combination with page 11, lines 35 to 39).

3.2 The coated abrasive product of claim 1 as granted differs from the coated abrasives according to D1 in that

i) at least 25% of the abrasive grits have an aspect ratio greater than 2:1, and

ii) that the non-abrasive particles have an average largest dimension that is less than 75% of the average largest dimension of the abrasive grits.

3.2.1 Feature i) appears to result in that such "weak" shaped grits, i.e. those having an aspect ratio of greater than about 1.5, when being aligned with their longest dimension perpendicular to the plane of the backing, improve on the one hand the effectiveness of the abrading process but on the other require larger amounts of size and supersize coats in order to prevent the premature fracture of the grit or even displacement of the whole grit from the backing (see patent, column 1, line 44 to column 2, line 4 and line 45 to column 3, line 1; column 3, lines 38 to 48). It is, however, not apparent from the patent in suit - which does not disclose any example, let alone a comparison with the closest prior art D1 - that said amount of "at least 25% of abrasive grits having an aspect ratio greater than 2:1" causes any different effect compared to an abrasive product of the prior art comprising abrasive grits having also such an aspect ratio but comprising not more than about 20% thereof (see patent, column 3, lines 2 to 7 and lines 24 to 29). In the patent in suit it is merely stated that the products are particularly useful when the abrasive grits have aspect ratios such that at least 40%, and even more preferably at least 75%, exceed 2:1 (see patent, column 5, lines 28 to 31).

3.2.2 With respect to feature ii) it is only stated in the patent in suit that the non-abrasive particles should be small enough to occupy the spaces between the abrasive grits (see patent, column 6, lines 11 to 26) and that they contribute to the more efficient operation of the abrasive particles with which they are mixed (see column 4, lines 45 to 48). Here again, the patent does not give any example of a technical effect obtained with the claimed feature, nor any evidence of the criticality of this size feature. Hence a specific effect of this feature ii) is likewise not apparent.

3.2.3 Taking account of points 3.2.1 and 3.2.2 above the Board therefore concludes that features i) and ii) need not be considered as distinguishing features for defining the objective technical problem since no effect can be attributed to them.

3.3 According to the appellant the problem underlying the patent in suit relates to the improvement of grinding aid efficiency (see column 3, lines 49 to 54) and thus to improving the grinding performance of abrasive articles.

The Board, however, cannot accept this definition of problem for the following reasons.

3.3.1 Although the aforementioned problem is identically described in the application as originally filed, on which the patent in suit is based, it is based on different prior art documents cited in the description. During the examination procedure D1 was then identified as the closest prior art in the description of the patent.

Furthermore, the patent specification does not comprise a single example, let alone a comparative one with respect to any of the cited prior art and particularly not with respect to the closest prior art D1. During the entire examination and opposition proceedings the appellant has also not submitted any results of appropriate comparative tests which would demonstrate an improvement of grinding performance of the abrasive articles according to claim 1 compared to those of D1. A reduction of the amount of grinding aid has not been shown by the patent in suit. This deficiency had already been noted in the impugned decision (see grounds for the decision, point 3.1, last paragraph) and the Board has reiterated this point in its communication annexed to the summons to the oral proceedings. Additionally the Board gave the reasons as to why the test report D9, filed together with the grounds of appeal, does not meet the requirements for comparative tests according to the established case law of the Boards of Appeal (see point IV above).

The appellant has not replied in substance to the reasoning given in said communication. Since it does not refute or overcome these objections, the Board sees no reason to depart from its preliminary opinion. Consequently, no technical improvement can be acknowledged for the coated abrasive according to claim 1 compared to the coated abrasives of the closest prior art D1.

3.3.2 Therefore the objective technical problem can only be defined as a less ambitious one, being the provision of a further coated abrasive product.

3.4 This problem is solved by the process as defined in claim 1 as granted.

3.5 The Board, however, considers that the subject-matter of claim 1 as granted is rendered obvious for the following reasons:

3.5.1 It is considered that the skilled person starting from the examples 16 and 18 of D1 would select the recommended preferred aspect ratios of the shaped abrasive particles of at least 2:1, particularly when considering the specified dimensions of the moulds for the triangular-shaped, square-shaped abrasive articles or the dimensions of the dried rod-shaped abrasive articles obtained according to the described general procedure (see page 12, lines 5 to 15) which in all cases result in aspect ratios well above 2:1.

3.5.2 Concerning the size of the diluent particles, based on the general teaching of a ratio of the size of the shaped abrasive particles to the size of the diluent particles of 2.5:1 to 0.5:1, D1 thus mentions a general range of from 40-200% of the size of the shaped abrasive particles. D1 further specifies that the diluent particles neither may be too small nor be too large relative to the shaped abrasive particles, in order not to adversely affect the coated abrasive article (see page 10, lines 14 to 20). D1 further states that it is preferred that the diluent and shaped abrasive particles be of approximately the same particle size (see page 10, lines 20 and 21).

Taking account of said general range it is, however, clear to the skilled person, particularly when reading example 18, which is the only example specifying the size range of the diluent particles (namely about 297 to 710 µm) in combination with the size (namely grade 36) of the shaped abrasive particles (see page 16, lines 5 to 9) that the two size ranges have to be matched within said general range of 40-200%. Said grade 36 corresponds to an average grain size of 710 µm (see point 2.1.5 above) so that implicitly the maximum value of the abrasive particle size range must be greater than the maximum value of the diluent particle size range of 710 µm. For example 18 the average grain size of the diluent particles will thus be smaller than that of the shaped abrasive particles.

3.5.3 Consequently, the person skilled in the art has to apply his common general knowledge and select the size range of the diluent particles by trial and error, performing experiments within said general range of 40-200%. Such experiments, however, will inevitably include diluent particle sizes close to the minimum and maximum values of said range. Thereby the person skilled in the art would obtain coated abrasive articles having a diluent particle size of which the average largest dimension is less than 75% of the average largest dimension of the abrasive grits as required by claim 1 as granted. The appellant's arguments regarding the size of the shaped abrasive particles therefore cannot hold.

3.6 The appellant's further arguments cannot be accepted for the following reasons.

3.6.1 First of all, the burden of proof regarding the effect allegedly caused by the distinguishing features of claim 1 has shifted from the respondent (opponent) to the appellant, since the present case is determined by a decision of the Opposition Division, which revoked the patent (see Case Law of the Boards of Appeal, 5th edition 2006, section VI.K.5.2.).

3.6.2 Although the test results of examples 16 and 18 in D1 are less positive than the comparative examples this cannot be interpreted that the skilled person has no incentive to produce the abrasive articles according to D1 comprising such a mixture of shaped abrasive particles and diluent particles since there exist other examples in D1 which clearly outperformed the comparative ones, e.g. examples 4, 6, 8, 10 and 11 (examples 10 and 11 except in test procedure II), 12, 14 and 20 (compare Tables 2 to 5) even though the examples 4 and 10-12 do not meet the weight ratio requirement of claim 1.

3.6.3 The argument concerning an effect caused by the distinguishing features of claim 1 - resulting in an improvement over the coated abrasive articles of D1 - therefore has not been supported by any appropriate evidence.

3.7 Therefore the subject-matter of claim 1 as granted lacks an inventive step, and thus does not meet the requirements of Article 56 EPC.

Consequently, the single request is not allowable.