T 0973/04 01-12-2006

1. The appeal is admissible.

2. Main request

2.1 The claimed subject matter

Claim 1 of the patent at issue relates to hypereutectoid steel rails comprising more than 0.85 to 1.20% carbon and a fully pearlite structure with a lamellar space lambda of not more than 100 nm and a ratio Rc (Rc = t2/t1; t2 = thickness of the cementite plates; t1 = thickness of the ferrite plates) of at least 0.15. This particular microstructure is obtained by the process steps set out in either claims 5 to 8.

2.2 Novelty

Novelty has been amply discussed in the impugned decision and in the appeal proceeding. The opponents argued i.a. that the claimed steel rail and the process for producing it were, in particular, anticipated either by the disclosure of OID1 or OIID4.

Document OID1 relates to hardened steel rails of hypereutectoid steel with carbon contents of 0.92 and 0.95% (cf. OID1, page 191, right hand column, Table). As to improving ductility and toughness, the lamellar pearlite structure of the rail is - in a preliminary treatment - spheroidised into globular pearlite. In a further heat treatment, the rails are induction hardened, i.e. at least the rail head is austenized and subsequently cooled at a particular cooling speed to produce a considerable amount of fine pearlite changing from troostite to sorbite (both phases are obsolete technical terms for fine-lamellar and ultra-fine pearlite, respectively); (cf. OID1, page 191, column 1, paragraphs 5 and 6; right hand column, last paragraph to page 192, first column, paragraph 1 and column 2, third full paragraph.) However, OID1 discloses neither the interlamellar spacing lambda of not more than 100 nm and the ratio Rc nor the precise cooling speed and stop temperature set out in the process claimed in the patent.

Likewise, document OIID4 is concerned with the production of steel rails comprising 0.75 to 1.00% carbon. After hot rolling, the rails are cooled from a temperature higher than 750ºC (i.e. >Ac3) at a cooling rate between 20 to 30ºC/s down to a temperature of at most 550ºC to transform the austenite structure into a fine pearlite structure (cf. OIID4, claim 1). Apart from conventional eutectoid steel rails (0.816%C; samples 1 and 2), this document also refers to a hypereutectoid steel rail (sample B: 0.95%C) having a very fine pearlite structure without the formation of free cementite. Sample B however fails to disclose the precise process steps (stop temperature for the accelerated cooling or stop when a specific degree of pearlite was formed) and remains silent about the interlamellar spacing lambda and the ratio Rc of the pearlite structure.

Document OIID17 is essentially concerned with the production of hypoeutectoid steel rails with carbon in the range of 0.64 to 0.82% which is below the claimed carbon range of >0.85 to 1.20%.

The remaining documents are more remote in that they refer to the metallurgical background knowledge. None of them discloses all the technical features of the claimed hypereutectoid steel rails and the method for producing them.

The subject matter set out in claims 1 to 8 as granted (main request) is, therefore, novel.

2.3 Inventive step

It was common ground to all parties at the oral proceedings that document OIID4 qualifies as the closest prior art. As to the issue of inventive step, the Board cannot, however, agree with the patent proprietor's line of arguments brought forward in particular with respect to this document for the following reasons.

Like the opposed patent, OIID4 addresses the problem of increasing the hardness of the pearlite structure of a steel rail to obtain a higher wear resistance (cf. OIID4, page 2, second paragraph; page 3, last line to page 4, line 3; page 4, 3. full paragraph). Specifically, OIID4 relates to the production of wear resistant high carbon (0.075 to 1.00%C) steel rails obtained by accelerated cooling with forced air-water sprays from Ac3 down to at most 550ºC to promote a homogeneous fine pearlite structure in all parts of the rail and a hardness Hv in the range of 340 to 400 (cf. page 4). An overlap exists between the carbon content and hardness properties of the known steel rail with those claimed in the patent.

The patent proprietor's argument that the basic teaching of OIID4 resides in disclosing two eutectoid steel rails (0.816%C) which are quenched with different cooling rates and stop temperatures cannot be followed. The technical information in document OIID4 goes beyond this, teaching the skilled reader that even with carbon contents as high as 0.95%C (sample B) a homogeneous and fully pearlitic microstructure in a rail can be successfully achieved by adhering to a specific cooling regimen. Based on the Jominy test, a CCT (continuous cooling transformation) diagram (Figure 6) has been developed for steel composition B which is given in the Table on page 9. OIID4 further discloses on page 12 a (hot) rolled steel rail R which was produced from heat B and exhibited after quenching with a cooling rate between lines C1 and C2 (10-20ºC/s; see OIID4, Figure 6) a fully fine pearlitic structure free from cementite and having a Vickers hardness Hv of 380 to 410, contrary to what could have been expected from the statements taught in document OID1.

The patent proprietor correctly pointed out that OIID4 does not disclose explicitly the ratio Rc and the interlamellar spacing lambda.

Following the position of appellant II, the ratio Rc is physico-chemically determined by the carbon content of the steel melt when it crystallises to form alpha iron (ferrite) and Fe3C (cementite). Contrary to the interlamellar spacing lambda, the ratio Rc remains essentially unaffected by other parameters such as the cooling rate. This finding, not convincingly refuted by the patent proprietor at the oral proceedings, is corroborated by paragraph [0030] and by the examples 1 to 16 given in Tables 1 and 2 of the patent at issue, which reveal a strong linear correlation between the carbon content C and the ratio Rc. It is, therefore, justified to conclude that Rc is not an independent technical feature which actually effects a patentable distinction between the claimed steel rails and OIID4, since it can be assumed that the claimed ratio is likewise fulfilled in hypereutectic rail R made from the hypereutectic steel alloy B.

Turning to the lamellar spacing lambda, document OID6 teaches that a very fine interlamellar spacing is indispensable for producing harder, more wear resistant pearlitic steel rails (cf. e.g. OID6, Abstract, first paragraph; 1. Introduction, first paragraph; 6. Conclusions, first sentence). This correlation of an improved hardness with a finer interlamellar spacing, well known to the metallurgical expert and reflected in many other documents, is evident for instance from OID6, Table 2. It shows that a hardness of about Hv 370 or more is obtained in a pearlitic steel rail if the mean true interlamellar spacing is less than 100 nm. The patent proprietor correctly argues that these results are obtained with four pearlitic steels with 0.76, 0.77, 0.79 and 0.81%C. It appears, however, from OID6, Figure 7 that the relationship Hv/(lambda-1/2) remains essentially unaffected by the carbon content of the steel. Despite the fact that the correlation between Hv and lambda has been obtained with pearlitic hypoeutectoid or eutectoid steel rails, this finding can therefore be transferred also to hypereutectoid steel rails with carbon contents in the range of 0.85 to 1.20%. In so doing, rail B (0.95%C) in OIID4 having a hardness Hv between 380 and 410 is expected to exhibit an interlamellar spacing of about 100 nm or less, all the more so since the quenching rate of 10 to 20ºC/s fully complies with the one used in the patent.

Apart from these specific considerations, the general teaching of OID6 is that a small interlamellar spacing lambda achieves a better hardness and, in consequence thereof, an improved rail wear resistance. The same basic metallurgical teaching is given e.g. in the textbook reference OID4, page 418, third full paragraph or in document OIID17, page 1, line 85 to page 2, line 4: a fine interlamellar spacing of 90 to 110 nm of a fully pearlitic microstructure is obtained by the accelerated cooling of a steel rail having 0.64 to 0.82%C at a rate of 1.5 to 5.2ºC/s and stopping the cooling between 538 and 371ºC (cf. OIID17, page 2, left hand column, lines 9 to 52). It, therefore, does not involve an inventive step in view of the technical teaching of document OID6 or, in the alternative, of OID4 or OIID17 to adapt the process parameters for quenching so that the interlamellar spacing lambda is reduced to a value as low as possible if the wear resistance of the rails is to be significantly improved.

With particular reference to document OID1, page 191, first column, paragraphs 2 and 3, the patent proprietor argues that the railway companies dissuade from producing hypereutectoid steel rails having more than 0.82% C due to the cementite network in the steel matrix and the embrittlement associated therewith.

The patent itself notes in paragraph [0051] that the claimed pearlitic steel rail may comprise in the pearlite matrix (traces or even considerable amounts of) pro-eutectic cementite which do not adversely affect strength and toughness of the rail. More importantly however is the technical teaching of document OIID4 that the formation of cementite can be minimised or even avoided in hypereutectic steel rails provided the cooling speed is controlled within specific limits.

The subject matter of claim 1 as granted, therefore, lacks an inventive step with respect to the combined technical teaching given in document OIID4 and OID6 or OIID17.

3. Auxiliary requests 1, 2, 2B, 2C and 3 to 7

3.1 Claim 1 of the auxiliary requests 1, 2, 2B, 2C, and 3 to 5 the terms

(i) "and no martensite structure at a segregation portion" as well as

(ii) "optionally Mn: 0.40 to 1.50%".

For support of the terms (i) and (ii), the patent proprietor referred to the patent specification (unamended in view of the application as filed) page 7, lines 23 to 28 to paragraphs [0031], [0032] and [0033], respectively.

As to amendment (i), the cited passage on page 7 of the specification discloses that sufficient recuperation from inside the rail cannot be expected after accelerated cooling down to a temperature of less than 500ºC, and the martensite structure detrimental to the toughness and the wear resistance of the rail is formed at the segregation portion. It does, however, not disclose that no martensite structure at all is formed in the segregation portion when stopping the accelerated cooling at a temperature of 500ºC or above. In the Board's view, however, the extraction of the reverse from what is stated in the cited passage of the patent is not allowed.

Turning to amendment (ii) it cannot be unambiguously derived from claim 1 as granted or from paragraphs [0031] to [0033] or from any other part of the original application whether or not Mn actually represents an optional component. The only option originally disclosed was Si and Mn (see original claim 3 and page 4, lines 9 to 17) and not Mn without Si which would be now covered. The introduction of the term "optionally" Mn thus cannot be allowed.

3.2 The upper limit for the cooling rate of 5ºC/s featuring in claim 1 of auxiliary request 6 is derived from samples 21 to 23, which are all micro-alloyed with vanadium. Nothing is found anywhere in the original application to show that a cooling rate of 1 to 5ºC/s is a preferred range.

Likewise, the lower limit of 1.02% of the carbon range featuring in claim 1 of auxiliary request 7 is derived from example 41.

It is, however, noted that examples 21 to 23 and 41 exhibit a specific steel composition and cooling rate which upon interaction bring about the final microstructure and mechanical properties of the steel rail. Contrary to the patent proprietor's position, the metallurgists know that C, Si, Mn and the microstructure as a consequence of the cooling rate contribute to the strength and toughness of the steel rail (cf. for instance paragraphs [0030] to [0034]). Moreover, the eutectoid composition itself (0.77%C for Fe-C alloys) varies when further alloying elements such as Mn, Si, Cr, etc are added. For corroboration only, reference is made to OID4, page 405/406, point 9,2,2 and page 417, first column, second paragraph. Hence, the individual amounts of elements and cooling rates making up an example cannot be regarded in strict isolation and, therefore, cannot not be used for the definition of a range. This may be done only in very exceptional cases. Reference is made in this context to decision T 201/83 (lead alloy) where the Board first established that, for a given Pb-alloy, only a loose or no connection existed between the components Ca and Mg with regard to their effect and that the actual amount of Ca was not tied to a specific magnesium content. From the detailed considerations given in T 201/83 the conclusion must be drawn that because of the effects of interaction of the constituents making up the claimed steel rails and their properties, it is not possible to make an arbitrary selection of individual features from the single examples for the definition of a (new) range. To disregard the specific context would result in a new selection from the original range which was neither explicitly nor implicitly disclosed. This means that in accordance with the above decision any arbitrary combination of values, isolated from the original text, is not allowed.

3.3 The amendments to claim 1 of the auxiliary requests 1, 2, 2B, 2C and 3 to 7, therefore, contravene the requirements of Article 123(2) EPC.

4. Auxiliary request 2A:

4.1 The process set out in claim 1 of auxiliary request 2A comprises an accelerated cooling at a rate of 1 to 10ºC/s which is stopped when the rails reaches a temperature between 700 and 500º.

In the patent proprietor's view, the claimed method was different to that described in OIID4 in that a cooling rate between 10 and 20ºC/s has to be adhered to for obtaining a pearlite structure and given that in the examples set out in OIID4 the accelerated cooling was stopped at a temperature far below the range of 700 to 500ºC according the claims of the patent.

4.2 Although document OIID4 recommends a cooling rate between 10 and 20ºC/s to promote a homogeneous pearlite structure in the rails, the lower limit of 10ºC/s complies with that specified in claim 1. However, figure 6 of document OIID4 further shows that also for cooling rates somewhat lower than 10ºC/s a fully pearlitic structure or at least a structure comprising 98% pearlite/2% Fe3C can be successfully obtained for steel B (0.95% C). As noted in the patent, a certain amount of pro-eutectic Fe3C can be tolerated in the pearlitic steel rails produced according to the claimed method (cf. [0051]). The technical findings in document OIID4 are therefore consistent with those given in the patent.

Document OIID4 further notes on page 12, penultimate paragraph that the austenite -> pearlite transformation for steel B is finished at 550ºC. This means to the metallurgical expert that there is no need to continue the accelerated cooling below this temperature or even down to ambient temperature given that no further phase transformation will occur.

4.3 The patent proprietor further argues that reducing or even interrupting the accelerated cooling of the steel rail during the austenite -> pearlite transformation, i.e. within the so called "pearlite nose" was not disclosed in document OIID4.

However, the interrupted cooling represents a conventional practice well known to the metallurgist if the complete transformation of the austenite structure into pearlite is to be guaranteed. For corroboration purposes only, reference is in this context to the textbook OID4, page 418, 9.4.1.4. third paragraph and Figure 9.11.a. There an increased cooling rate is applied to conventional rails to cause refinement of the lamellar spacing. In this example, only the rail head is heated to about 900ºC, rapidly cooled to a temperature of about 550 to 650ºC using forced air cooling and held in this temperature range until the austenite is fully transformed to austenite.

4.4 The process features set out in claim 1 of the auxiliary request 2A therefore amount to nothing more than what is done in document OIID4 in combination with conventional practice that is applied by a person skilled in the art. Claim 1 of auxiliary request 2A therefore lacks an inventive step as required by Article 52(1) EPC in conjunction with Article 56 EPC.

5. Given that none of the requests relates to claims satisfying the requirements of the EPC, in particular those of Articles 56 and 123(2) EPC, the patent is revoked.