dissenting in part.
I join the court’s opinion in all but two respects, upon which I respectfully dissent. On the first point, I disagree with the ground upon which the court affirms the judgment of the district court that the claims of the '809 and '906 are not infringed. For the reasons that follow, I would reverse the district court’s interpretation of the “axial peripheral lip” limitation, and thus not rely on that limitation to sustain the judgment of noninfringement. Instead, I would apply the correct interpretation of the “wheel face outer surface” limitation to the accused products, and in so doing would employ the, uncontested facts to sustain the district court’s judgment of noninfringement on this alternate ground.
Second, for the reasons stated below, I disagree with the court’s affirmance of the district court’s finding of infringement of claims 11-13, 20, 24, and 25 of the '213 patent. I think we should have rejected these findings as clearly erroneous, on the ground that the evidence failed to prove that the accused devices met the Ford durability testing standard set by the district court as the measurement device for assessing infringement of the “permanently secure” claim limitation.
I
I would reverse the district court’s interpretation of “axial peripheral lip.” However, as I conclude that the construction of “wheel face outer surface” compels a finding of noninfringement, I concur with the court’s opinion on this issue.
The district court’s claim construction of “axial peripheral lip” is unusual. Defining “axial peripheral lip” as the “uncovered *1355surface of the composite wheel that starts at the edge of the cladding, goes up and over the top of the projecting rim lip, and ends where the outer side of the rim lip merges with the tire mounting surface of the wheel rim” (emphasis in original) literally means that the accused products must meet the claim requirement that the ornamental surface “not cover” the axial peripheral lip, since the axial peripheral lip, by the district court’s definition, is uncovered. This renders the district court’s finding of noninfringement somewhat opaque.
The district court’s claim construction views the “axial peripheral lip” as including both some portion of the wheel wall below the innermost edge of the wheel rim (facing the center of the wheel) as well as the top of the wheel rim. This is evident from both the district court’s definition given above, as well as its characterization of the defendant’s ornamental panel as “terminatpng] just shy of the tip of the lip.” The parties concede that the term “axial peripheral lip” has no particular meaning in the art. The standard dictionary definition of “lip” is given as “an edge, rim, or margin,” or even more specifically as the “edge or margin of a hollow vessel.” Webster’s 3rd New International Dictionary 1318 (1993). Here, the diagrams indicate two separate edges on the wheel rim, an interior one facing the center of the wheel, and an external one facing the outside. In this context, the “axial peripheral lip” logically refers to the external one — the one furthest from the center of the wheel. However, there is an alternative possibility: the “axial peripheral lip” could also refer to the combination of both “edges,” along with the flat portion of the wheel rim in between.
Either of these definitions is inconsistent with the district court’s interpretation of “axial peripheral lip,” specifically the inclusion in the definition of some portion of the wheel wall below the innermost edge of the wheel rim. Both are strongly supported by the prosecution history. The claims at issue were amended in response to an office rejection for unpatentability in light of the Beith reference. Beith discloses a lip 43 “formed to grip the edge of the terminal flange 29 of the multi-flanged tire mounting rim 21 and, thereby, aid in fixing the wheel cover 12 to the wheel 13.” Figures 2 (see below) and 5 depict lip 43 doing precisely this — extending over and around the outer circumferential edge of the rim to “grip” the wheel.
[[Image here]]
In response to this rejection, the patentee added the requirement that the ornamental panel “not cover” the axial peripheral lip. As can be seen from Figure 2, in order to distinguish over Beith, the paten-tee needed to emphasize that his ornamental panel did not overlap the very outer edge of the wheel rim.
It is difficult to square the import of this amendment with a claim construction interpreting “axial peripheral lip” as including some undefined portion of the wheel wall below the interior “edge,” particularly if we presume that the patentee only sur*1356rendered enough structure to get out from under Beith.
As a result, I would reverse the district court’s claim construction as to “axial peripheral lip,” interpreting it instead as either the outer terminal edge or, alternatively, the highest point on the wheel rim.
I now turn to the “wheel face outer surface” claim limitation. The claim language expressly characterizes the wheel face outer surface, providing that “said web portion and said peripheral rim portion defin[e] a wheel face outer surface.” 'S09 patent, col. 11, 11. 4-5. This appears to be a closed definition, the wheel face outer surface consisting of only two elements — the central web portion and the peripheral rim portion. The claim language further defines the axial peripheral lip as “circumscribing said peripheral rim portion,” id. at col. 11,11. 1-2, and requires that the ornamental panel cover “said entire wheel face outer surface,” id. at col. 11,1.10 (emphasis added).
To my mind, this claim language explicitly defines the wheel face outer surface as extending precisely up to the terminal boundary of the peripheral rim portion— the axial peripheral lip. Indeed, despite its differing construction of “axial peripheral lip,” this is what the district court concluded as well. This interpretation would thus allow no portion of the peripheral rim portion to be uncovered by the ornamental panel.
Such a conclusion is problematic, however, given Figures 3 and 5 of the '809 and '906 patents. These clearly depict the ornamental panel as only covering part of the peripheral rim portion. The above interpretation would thus read out the preferred embodiment.
As a general rule, claim interpretations, which operate to exclude the preferred embodiment, are “rarely, if ever, correct and require highly persuasive evidentiary support.” Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1583, 39 USPQ2d 1573, 1578 (Fed.Cir.1996). However, we have found that such a conclusion can be mandated by clear intrinsic evidence, such as “unambiguous” claim language. Elekta Instrument v. O.U.R. Sci. Int’l, 214 F.3d 1302, 1308, 54 USPQ2d 1910, 1914 (Fed.Cir.2000). See also Rheox, Inc. v. Entact, Inc., 276 F.3d 1319, 61 USPQ2d 1368 (Fed.Cir.2002) (reaching a claim interpretation excluding some of the preferred embodiments in light of the prosecution history). Here, the claim language is quite explicit in both defining the “wheel face outer surface” and in requiring it to be completely covered by the ornamental panel, and I would find that it constitutes an exception to the general rule.
Indeed, the linkage of the two claim requirements, that the cladding cover the entire wheel face outer surface, but that it not cover the axial peripheral lip only has meaning in the situation where the axial peripheral lip is actually providing a boundary for the overlay. Any other construction of “wheel face outer surface” would essentially mean that there is a separate interior boundary for the overlay coverage. In such a situation, the requirement that the overlay not cover the axial peripheral lip is surplusage.
Rejecting the above construction leaves open a substantial question as to what the “wheel face outer surface” actually is. Lacks’s suggestion that it merely requires partial coverage of the “peripheral rim portion” is extremely unsatisfying. First, the claim language describes “said web portion and said peripheral rim portion defining a wheel face outer surface.” If this language includes both elements in the definition, then full coverage is required. If it does not, Lacks’s suggestion that “partial” coverage of the rim portion suffices would make the use of “entire” completely meaningless — there would be no *1357way to define how much coverage of the peripheral rim portion is required.
The district court used the Hayes diagram to evaluate the accused device. Lacks’s brief references the same diagram. It discloses an ornamental panel terminating, as the district court found, “just shy of the tip of the lip.” Neither of the parties dispute that this is an accurate representation of the accused product.
Employing the claim constructions given above, the district court’s finding of nonin-fringement can be sustained. Even if the accused product meets the limitation requiring that the ornamental panel not cover the “axial peripheral lip,” the ornamental panel does not cover the entire wheel face outer surface, as there is a gap between the ornamental panel and the axial peripheral lip. The district court’s grant of summary judgment of noninfringement of the claims of the '809 and '906 patents was correct, for the reasons I have stated.
II
Hayes and McKeehnie cross-appeal the finding of infringement of the asserted claims of the '213 patent. For the reasons that follow, I agree with Hayes and McKeehnie, and disagree with .the contrary decision by the court affirming the district court’s resolution of this issue.
Claims 11-13, 20, and 24-25 of the '213 patent require that the adhesive “permanently secure” the overlay to the wheel. '213 patent, col. 15, 1. 31. The district court interpreted this to require that the adhesive be “itself capable of securing the overlay to the wheel during the normal useful life of the wheel” (emphasis in original). The Special Master concluded that this test was met if the adhesive:
(1) ... satisfies the specifications set forth on the Hayes engineering drawing for the accused F 150 wheel (Exh. 42) and (2) retains sufficient retentive strength (without reliance on the mechanical locking system) to secure the overlay throughout Ford’s vehicle durability testing, but with a modified upper temperature limit as discussed below.
This is a very reasonable test. Unfortunately, it is not applied. Having established the Ford vehicle durability testing as a key component of the infringement test, the Special Master’s report indicates that not only is there almost no evidence as to what the Ford testing consists of, there is no evidence that there was ever any testing of the wheel on an actual vehicle.
First, the Special Master notes, “[w]ith regard to the Ford durability testing conditions, the Special Master finds the record to be somewhat incomplete. There is no description of what type of corrosion testing is performed as part of such durability tests on the vehicle, or the other durability testing performed at the Ford’s Arizona Proving Ground.... ” The report also states:
With regard to Ford’s own durability testing on driven vehicles, Mr. Thomas Heck [a Hayes director] testified that these generally consisted of corrosion testing and power train durability testing. The accused Hayes wheels had to pass these Ford tests. There was no testimony or documentation concerning what Ford would consider acceptable performance of the wheel overlays during such tests. [Emphasis added.]
In other words, having set up “Ford’s vehicle durability testing” as a key element in evaluating the claim, the Special Master seems to be explicitly acknowledging that it is uncertain as to what that durability testing actually is.
Second, the Special Master also explicitly states that “[t]here is no evidence, however, of vehicle testing on the accused wheels with the mechanical lock component of the overlay system disabled and only the adhesive component of the overlay *1358system operative” (emphasis in original). This, again, is somewhat problematic in light of the Special Master’s own definition of the test for infringement as requiring just such a demonstration.
Lacks accurately points out that the Smithers rotary, radial fatigue, and impact testing was done (pursuant to standard testing procedure) on individual wheels with the mechanical locking system disabled. Under our deferential standard of review of factual issues, these might suffice to support a finding of infringement, had they actually comprised part of the Special Master’s own test for infringement. However, as mentioned above, the Special Master explicitly required that the adhesive demonstrate retentive strength “throughout Ford’s vehicle durability testing” (emphasis added). This language itself contains the word “vehicle,” strongly suggesting that some of the testing should be on an actual vehicle. In addition, despite the apparent uncertainty discussed above as to what Ford’s durability testing consisted of, the Special Master described at least “one phase of Ford’s durability testing” as consisting of an actual vehicle test drive. As cited above, the report also references “Ford’s own durability testing on driven vehicles.” In contrast, the alternative tests are not Ford durability tests, nor even vehicle tests, and thus are not probative under the standard of evaluation defined by the Special Master.
Third, as the Special Master noted, the only other testing performed with the mechanical lock disabled, the McKechnie “pull-off test,” was “not designed as a durability test or simulation of any particular type of removal force that would be encountered in the field,” much less a Ford durability test. Again, were the Special Master to have simply cited this test as evidence in favor of the fact that the accused device met the “permanently secure” claim limitation, it would likely survive our deferential standard of review. Here, however, having established the Ford durability test as an explicit standard for how to evaluate infringement, the Report’s reliance on evidence, which falls outside of this standard, is both logically and legally incorrect.
We thus should conclude that the Special Master’s decision as to infringement of the '213 patent claims is clearly erroneous, and that the district court accordingly should have rejected these findings and entered judgment of noninfringement of the asserted claims of the '213 patent.