T 0547/17 22-04-2021

1. Admissibility of the Appeal

The appeal complies with the formal and substantive requirements for admissibility of Article 108 and Rule 99 EPC and is therefore admissible.

2. Main Request - Inventive Step

2.1 The subject-matter of claim 1 of the main request does not involve an inventive step within the meaning of Article 56 EPC.

2.2 Closest Prior Art

The Board is satisfied that document A5 is a suitable choice as a starting point for the examination of inventive step, which is not contested by the respondent.

2.3 Distinguishing Features

2.3.1 For the discussion of the distinguishing features the Board considers it useful to reproduce figure 7 of the opposed patent in order to illustrate some of the features of claim 1 relevant for the present discussion.

FORMULA/TABLE/GRAPHIC

Reproduction of figure 7 of the opposed patent

Claim 1 requires the ratio A/B to satisfy the relation 0.2 is less that or equal to A/B is less that or equal to 1.0. Here, A denotes the overlap between said stator core (15) and the disk portions (301, 302) when viewed from a radial direction and B denotes the axial length of said disk portions. The radial direction is from top to bottom in figure 7 and the axial direction is from left to right.

Furthermore, claim 1 requires the ratio Lp/Lc to satisfy the relation 0.7 is less that or equal to Lp/Lc is less that or equal to 0.9. Lp denotes the axial length of the overlap of the claw-shaped magnetic poles (22, 23) and the stator core (15) when viewed from a radial direction (i.e. from above in figure 7). Lc denotes the axial (i.e. from left to right) length of the stator core (15).

2.3.2 The following discussion will be based on the following distinguishing features of claim 1:

- the ratio A/B is in the range 0.2 is less that or equal to A/B is less that or equal to 1.0, and

- the ratio Lp/Lc is in the range 0.7 is less that or equal to Lp/Lc is less that or equal to 0.9.

It has not been disputed that the remaining features of the claim are known from A5, in particular that this document discloses in the passage from column 5, line 60 to column 6, line 9 values for the claw dimensions which fall within the claimed range for t1/t2.

The appellant contested that 0.2 is less that or equal to A/B is less that or equal to 1.0 could be considered a distinguishing feature since the claimed range was not narrow in view of the geometrically possible range, which was necessarily confined to be between 0 and 1 due to the fact that A was defined to be an overlap. It was not necessary for the Board to decide this question, because even assuming that the claimed range for the ratio A/B was a distinguishing feature, the Board comes to the conclusion that the subject-matter of claim 1 does not involve an inventive step.

It further appears from the passage of A5 cited above that the axial length of the stator core Lx is 5% longer then the length of the pole claw A. This would mean that the ratio Lp/Lc (which is the same as A/Lx in A5) equals 1/1.025 = 0.98, which is outside the claimed range of 0.7 to 0.9. The parties did not contest this.

2.4 Technical Effect and Technical Problem

2.4.1 Concerning the technical effects associated with the distinguishing features, the respondent argued, adducing figures 8 and 9 of the opposed patent, that the claimed ranges maximised the output current of the electric machine. Furthermore, the ratios were interrelated. Varying one pair of parameters such as A/B also influenced the other parameter Lp/Lc, and vice versa. For example, increasing A in relation to B by making the stator core 15 longer would result in a concomitant change of the the ratio Lp/Lc. The appellant, based on their simulation results, contests the assertion that the choice of parameter ranges leads to an optimum output current and asserted that the two parameters did not have any synergy.

The Board notes that neither the opposed patent nor the respondent nor the appellant supply sufficiently detailed information so as to put the Board in a position to find fault with either the experiment or the simulation. For the following discussion, the Board assumes, to the benefit of the respondent, that their experimental results were correct. Since the appellant's simulations are not decisive for the Board's conclusions, the Board did not have to take a decision (and hence discuss) whether to admit them.

For the sake of the argument, the Board therefore concurs with the respondent that the choice of ratios A/B and Lp/Lc in the claimed range both contributed to maximising the output current (rather than just increasing it or causing it to have some arbitrary value). Concerning the technical effect associated with the first distinguishing feature, according to column 9, lines 30 to 40 of the opposed patent in connection with figure 12, the ratio A/B influences the flux linkage between the root portions of the pole claws and the stator core. If the overlap A between them is too small not all flux enters the stator core. The appellant argued that an "overlap" A larger than 1 was geometrically impossible. The Board would agree that A larger than 1 would mean that the axial length of the stator core must be longer than the axial length of the rotor. In that case, clearly some portion of the stator core extending beyond the rotor would not be linked to the magnetic flux emanating from the rotor. Therefore, the technical effect of the first range for the ratio A/B is to maximise the flux linkage at the root portion of the pole claws.

Similarly, the technical effect associated with the second distinguishing feature, according to figure 7, and column 9, line 47 to column 10, line 4 of the opposed patent is to maximise the magnetic flux linkage at the tip of the pole claws. If the pole claw tips protrude too much (i.e. Lp/Lc > 0.9) too high a portion of the flux emanating from the pole claws cannot enter the stator core and is lost as stray flux. If the pole claws are too short and thus their tips are too far away from the edge of the stator core large parts of the latter are not linked. The Board thus agrees that both distinguishing features contribute to maximising the output current.

However, this does not mean that the resulting output current was more than the sum of the individual contributions of the two ratios. The mere fact that both features contribute to maximising the same physical quantity (here the output current) is not sufficient to conclude that there is synergy between them. The flux linkages at the pole claw tips and roots are largely independent of one another. This last point was not contested by the respondent. This, however, shows that the output current produced by both distinguishing features in combination is not higher than the sum of the individual contributions of each ratio. There is therefore no synergy between them.

The respondent argued that, in addition to solving the same problem, the two ratios according to the distinguishing features were interrelated. The Board agrees that the ratios A/B and Lp/Lc are geometrically interrelated. However, whether there is synergy between two features is a matter of whether their technical effects produce a combined technical effect which is greater than the sum of their individual technical effects, and not simply a matter of geometrical interrelation.

In conclusion, both distinguishing features are aimed at maximising the flux linkage but without influencing each other because the flux linkages at the pole claw roots and at their tips are independent of each other. The Board therefore agrees with the appellant, that there is no synergy.

It follows that the first and second partial technical problems addressed by the two distinguishing features are both to maximise the output current.

2.5 Assessment of the Solution

Since there is no synergy, the solutions to the partial problems can be assessed separately.

2.5.1 If a skilled person starting from A5 wanted to increase the output current, they would consult document A6 because it deals with maximising the output current of electric machines according to claim 1.

FORMULA/TABLE/GRAPHIC

Reproduction of figure 16 of document A6

Document A6 suggests in figure 16 and column 10, lines 13 to 48, that the ratio L2/L1, which corresponds to A/B according to claim 1, should be 0.3 or greater in order to maximise the flux link at the at the disk portion, i.e. at the pole claw roots. According to column 10, lines 25 to 48 and figure 17, this also maximises the output current for a given field current.

2.5.2 The skilled person would also consult A8 because it also deals with the same problem.

FORMULA/TABLE/GRAPHIC

Reproduction of figure 4 of A8

In can be seen in the reproduction of figure 4 above that the ratio of d2/L (also called overlap alpha in A8) is equal to the ratio Lp/Lc according to the opposed patent. According to A8, figure 7 and page 6, line 35 to page 7, line 11, a particularly preferred value is 0.88, which falls in the claimed range of 0.7 is less that or equal to Lp/Lc is less that or equal to 0.9.

2.5.3 In the absence of any synergy, the Board cannot see that combining the teachings of all three documents, i.e. choosing a thickness ratio t1/t2 according to the closest prior art A5, an overlap ratio A/B according to document A6, and a length ratio Lp/Lc according to A8, required inventive activity. It may be true that these ratios are somehow geometrically related and need simultaneous adaption so that each individual ratio remains within the range prescribed by the relevant prior art document. However, in the Board's opinion, such simultaneous adaption, when the desired ratios are already given, does not require any understanding of the mechanism of magnetic flux linkage but merely the skill to geometrically dimension the rotor and stator cores according to the prescribed ratios. The Board would concede that inventive activity may lie in recognising in which places the magnetic flux linkage can be improved to increase the output current. However, the individual influences of the two ratios according to the distinguishing features on the current output is already known and described in a detailed manner in the state of the art as demonstrated by documents A5 to A8.

In order to arrive at the subject-matter of claim 1 and thus solve the objective technical problem of maximising the output current, a skilled person can draw on existing knowledge about the geometric rotor and stator parameters which influence the flux linkage and output current as well as the optimum starting values taught in the closest prior art. Systematically varying more than one parameter (i.e. three in the present case), even though some of them might be geometrically interrelated, when given the understanding about the output current in the prior art, can, in the Board's opinion, be characterised as routine experimentation but not as inventive activity.

2.5.4 In the respondent's opinion, the teachings of A6 and A8 were contradictory. A8 suggested making the pole claws shorter with respect to the stator core in order to avoid vibrations, whereas A6 suggested making the pole claws longer in order to have more overlap at the root portion. The Board is not persuaded by this argument. While it is true that A8 suggests employing shorter pole claws in order to avoid vibrations and centrifugal deformation of the pole claws, it can be seen in figure 4 of A8, reproduced above, that A8 suggests making the pole claws shorter at the tip without thereby also decreasing the overlap of the stator core and disk portions at the root of the pole claw tips. The same is true for A6, which shows in figure 16, also reproduced above, that the lap ratio, i.e. the overlap at the root portion of the pole claw, is independent of the overlap at the tip.

2.5.5 The respondent also argued that A6 showed that copper losses, i.e. increased resistivity of the rotor winding due to heating, had to be taken into account when varying the lap ratio, and in particular that if the lap ratio was too large, one could not achieve a maximum current output as shown in figure 17 of A6. The respondent argued further that, since the cavity made by the pole claws of the rotor according to A5 was particularly densely packed with the rotor winding, the above mentioned copper losses would start to negatively affect the output current at smaller lap ratios than the value 0.3, which was applicable only to the particular rotor according to A6.

The Board is also not persuaded by this argument. According to figure 17 of A6 the output current plateaus in the range from 0.3 to 1 because copper losses compensate for the increased flux linkage. The teaching of A6 is therefore merely that there is no gain in increasing the lap ratio (corresponding to A/B) much beyond 0.3. However, figure 17 also clearly teaches that there is no harm in doing so if the aim is to maximise the output current as the latter remains largely constant over the entire range of 0.3 to 1. It has to be borne in mind that claim 1 of the main request is directed to nearly the entire geometrically possible range, starting at an overlap of 0.2 and going up to a full overlap A/B = 1, still claiming that a "maximum" output current could be verified experimentally. The Board notes that claim 1 does not specify what proportion of the pole claw cavity is occupied by the rotor winding. This means, given the knowledge of A6, that the inventors of the opposed patent merely accepted the copper losses which are correctly predicted by A6 and found that they do not appear to influence the output current in an inacceptable manner. The Board is not persuaded that merely accepting a correctly predicted adverse effect from the prior art could be seen as inventive activity. It has also to be borne in mind that if the respondent's argument were correct and in A5, due to the higher degree of packing of the pole claw cavity with the rotor winding, copper losses did appear earlier, a skilled person had a clear motivation due to the teaching of A6 to choose the lap ratio somewhat smaller than the value of 0.3, which is optimal for the rotor of A6. The exact value of the lower boundary will clearly be dependent on details of the alternator, to which the claim is not restricted. The skilled person is given detailed explanations in A6 of the various effects of dimensioning the overlap. In view of this, the Board is convinced that choosing a range that covers nearly the entire range from no overlap to full overlap, and in particular merely accepting that copper losses occur as correctly predicted in the prior art, does not amount to inventive activity.

The Board notes that the respondent's argument concerning allegedly contradictory teachings of A6 and A8 and concerning A6 allegedly teaching away from the claimed ratio A/B were presented for the first time at the oral proceedings before the Board. Given that the Board comes to the conclusion that they are not convincing, it was not necessary to decide whether the argument contained also new facts and whether or not the Board should admit them.

3. Auxiliary Request - Inventive Step

3.1 The subject-matter of claim 1 of the auxiliary request does not involve an inventive step within the meaning of Article 56 EPC.

3.2 The feature added to claim 1 concerns an inclined surface (22a, 23a) inclined radially inward starting from outside an end surface of said stator core (15) which is formed on a shoulder portion of a claw-shaped magnetic pole (22, 23).

A5 discloses in figures 1 and 3 an inclined surface on the magnetic pole claw rotor as shown in the following partial reproduction of figure 1 of A5.

FORMULA/TABLE/GRAPHIC

Partial reproduction of figure 1 of A5.

The rotor is shown to have chamfers, i.e. surfaces which are inclined radially inwards, which start outside the stator core (4). As can be seen in figure 3 of A5, the surfaces are indeed inclined surfaces (chamfers) although they may appear somewhat rounded in the above figure 1. As can be seen in figure 7 of the opposed patent, the expression "start outside the stator core" is supposed to mean that the shoulders lie axially outside of the stator core, as is the case in A5.

The added feature of claim 1 of the auxiliary request is therefore not a distinguishing feature over the closest prior art A5. Therefore, the conclusions concerning lack of inventive step of the main request also apply to the auxiliary request.

3.3 The respondent argued that figures 1 and 3 of A5 were mere schematic figures. As there was no other disclosure of inclined surfaces, A5 did not contain a direct and unambiguous disclosure of this feature.

This does not persuade the Board. What can be considered directly and unambiguously disclosed by a figure has to be decided in each individual case. A skilled person will not necessarily consider a figure to be either entirely schematic or exact in every detail. Figure 1 contains details typical for a technical drawing, such as hatched areas or grooves on the belt pulley (not shown in the above reproduction). As such this figure appears to the Board to be not merely a schematic representation. The respondent failed to explain why a technical draftsman would consistently draw chamfers in figure 1 and 3 of A5, if the rotors did not have chamfers in reality. The Board also finds the respondent's position inconsistent. The respondent argued in the context of the main request that a skilled person could recognise from figure 1 of A5 that the degree of packing of the pole claw cavity (reproduced above) with the rotor winding was particularly dense. However, a skilled person would not recognise from the same figure that the rotor had inclined surfaces on its shoulders.

4. Conclusions

Since neither the subject-matter of claim 1 of the main request nor that of claim 1 of the auxiliary request involves an inventive step, the Board accedes to the respondent's request.