Opinion for the court filed by Circuit Judge NEWMAN. Circuit Judge MAYER dissents without opinion.
PAULINE NEWMAN, Circuit Judge.
Modine Manufacturing Co. appeals the decision of the United States International Trade Commission,1 holding that section 337 of the Tariff Act, 19 U.S.C. § 1337, was not violated by the importation of certain automotive condensers. The respondents before the Commission, who participate as intervenors in this appeal, manufacture in Japan and import, sell, and use the accused condensers in the United States: Showa Aluminum Corporation and Showa Aluminum Corporation of America; Mitsubishi Motors Corporation and Mitsubishi Motor Sales of America; and Mitsubishi Heavy Industries, Ltd. and Mitsubishi Heavy Industries America, Inc.
[1549]*1549The issues on appeal are the validity and enforceability of Modine’s United States Patent No. 4,998,580 (the ’580 patent) and infringement by several models of condensers manufactured by Showa and imported, sold, and used by the intervenors. Modine is the appellant on the infringement issues, and the intervenors appeal the issues of validity and enforceability. We vacate the finding of non-infringement and remand for further proceedings based on the correct claim interpretation. On the other issues the Commission’s decision is affirmed.
I
THE PATENTED INVENTION
The invention of the ’580 patent is described by Modine as a highly efficient and environmentally advanced condenser for use in automotive air conditioning. It is more compact, lighter, uses less refrigerant, outperforms prior condensers, and has the additional advantage of being usable with refrigerants other than chlorofluorocarbons. Modine states that it converted the entire industry to a new standard.
Claims 9 and 10 of the ’580 patent, the only claims in suit, are shown with emphasis added to point out the two terms that are the focus of the infringement issues:
Claim 9. A condenser for a refrigerant in a cooling system comprising:
[1]a pair of spaced, generally parallel, elongated cylindrical tubes defining headers;
[2] a vapor inlet in one of said tubes;
[3] a condensate outlet from one of said tubes;
[4] said header tubes each having a series of elongated generally parallel slots with the slots in the series on one header tube aligned with and facing the slots in the series on the other header tube;
[5] a tube row defined by a plurality of straight, tubes of flat cross-section and with flat side walls and having opposed ends extending in parallel between said header tubes, the ends of said flat cross-section tubes being disposed in corresponding aligned ones of said slots and in fluid communication with the interior of said header tubes, at least some of said tubes being in hydraulic parallel with each other;
[6] web means within said flat cross-section tubes and extending between and joined to the flat side walls at spaced intervals to (a) define a plurality of discrete, hydraulically parallel flow paths within each flat cross-section tube that extend between said header tubes; to (b) absorb forces resulting from internal pressure within said condenser and tending to expand the flat cross-section tubes; and to (c) conduct heat between both said flat sides and fluid in said flow paths;
[7] said flow paths being of relatively small hydraulic diameter which is defined as the cross-sectional area of the corresponding flow path multiplied by four (4) and divided by the wetted perimeter of the corresponding flow path;
[8] serpentine fins incapable of supporting said flat cross-section tubes against substantial internal pressure extending between facing flat side walls of adjacent flat cross-section tubes;
[9] each of said flow paths including at least one elongated crevice extending generally along the length of the associated flow path.
Claim 10. The condenser of claim 9 wherein each flow path has a plurality of said crevices.
It is not disputed that all of the elements of the claimed invention have counterparts in the accused condensers, and that infringement turns on the meaning and scope of the terms “flat side walls” and “relatively small hydraulic diameter.” Modine challenges the correctness of the Commission’s claim interpretation and the ensuing finding of non-infringement.
II
INFRINGEMENT
As we have recently held, “[bjecause claim construction is a matter of law,' the construction given the claims is reviewed de novo on appeal.” Markman v. Westview Instruments, Inc., 52 F.3d 967, 979, 34 USPQ2d 1321, 1329 (Fed.Cir.) (en banc), cert. grant[1550]*1550ed, — U.S.-, 116 S.Ct. 40, 132 L.Ed.2d 921 (1995). Disputes as to the meaning and scope of terms as used in the claims are determined as a matter of law, based on the patent specification and the prosecution history if it is in evidence. Id. at 979-80, 34 USPQ2d at 1329-30. As stated in Mark-man, “[w]hen legal ‘experts’ offer their conflicting views of how the patent should be construed, or where the legal expert’s view of how the patent should be construed conflicts with the patent document itself, such conflict does not create a question of fact nor can the expert opinion bind the court or relieve the court of its obligation to construe the claims according to the tenor of the patent. This opinion testimony also does not change or affect the de novo appellate review standard for ascertaining the meaning of the claim language.” Id. at 983, 34 USPQ2d at 1333.
A THE FLAT SIDE WALLS
The Commission, adopting the ALJ’s Initial Determination, held that the term “flat side walls” means the interior walls of the condenser tubes, and that although the Showa side walls are flat in that they are not rounded, they are not “flat” because most (but not all) of the Showa models have fin-like projections on their interior surfaces. On this term construction the Commission concluded that the Showa tubes do not have “flat side walls” and therefore that the claims are not infringed by any of the Showa models.
Modine states that “flat side walls” describes the shape of the tubes, and refers to the ’580 specification which describes the condenser tubes as “noncircular in cross section” and as a “flattened tube.” Modine states that the side walls are flat whether or not they have fins on their inner surfaces, pointing out that the specification as well as the claims show fins on the outer surfaces as well as webs on the inner surfaces. Claim 9 mentions “flat side walls” in several clauses:
[5] a plurality of straight tubes of flat cross-section and with flat side walls ...
[6] web means within said flat cross-section tubes and extending between and joined to the flat side walls ...
* * * * * *
[8] serpentine fins incapable of supporting said flat cross-section tubes against substantial internal pressure extending between facing flat side walls of adjacent flat cross-section tubes;
Clause [5] uses the word “flat” to describe both the cross-section and the side walls, but neither usage of “flat” requires that the interior or exterior wall surfaces be clear, without web or fin. Clause [6] requires a web “joined to” the interior flat side walls, negating the ALJ’s reading that the wall surfaces must be clear. Clause [8] describes serpentine fins on flat side walls that are necessarily the exterior surfaces of the walls, contravening the ALJ’s ruling that “flat side walls” means the interior walls.
The entirety of the claim’s usage of flat side walls is consistent with the specification’s description of the condenser tubes as “flattened” and “not circular.” This plain reading is not affected by webs or fins on either the interior or the exterior surfaces of the walls, or by the crevices of claim 10. Indeed, a claim interpretation that would exclude the inventor’s device is rarely the correct interpretation; such an interpretation requires highly persuasive evidentiary support, whereas in this case it received none, whether from the specification, the prosecution history, or the prior art.
We conclude that the term “flat side walls” means that the tube structure is flat, as the specification states, and does not prohibit the presence of fins, webs, or other attachments to either the interior or exterior surfaces. Those Showa tubes that bear inner fins (the 3mm models), and those that do not (the 2mm models), all have flat side walls as the term is correctly construed. This claim limitation is not a ground for a finding of noninfringement.
B. RELATIVELY SMALL HYDRAULIC DIAMETER
The Commission held that the claim term “relatively small hydraulic diameter” is a limitation to hydraulic diameters no larger than exactly 0.040 inch. Since the hydraulic diameters of the Showa condensers are all larger than 0.040 inch, the Commission held that [1551]*1551the claims are not infringed, literally or under the doctrine of equivalents.
Hydraulic diameter is an engineering designation that measures the flow path within the condenser. It is defined as the cross-sectional area of the corresponding flow path multiplied by four and divided by the wetted perimeter of the corresponding flow path; this definition is included in claim clause [7]. Claims 9 and 10, the claims in suit, describe the hydraulic diameter as “relatively small.” Claims not in suit describe the hydraulic diameter as “about 0.015 to 0.040 inches” (claims 1-3), “in the range of 0.015 to 0.040 inches” (claims 4-5), and “sufficiently small ... so that surface tension and capillary forces acting upon condensate within said flow paths improve heat transfer efficiency” (claims 6-8). The Commission held that “relatively small” in claims 9 and 10 has an upper limit of 0.040 inch.
Ordinarily a claim element that is claimed in general descriptive words, when a numerical range appears in the specification and in other claims, is not limited to the numbers in the specification or the other claims. See Specialty Composites v. Cabot Corp., 845 F.2d 981, 987, 6 USPQ2d 1601, 1604 (Fed.Cir.1988) (“[P]articular embodiments appearing in the specification will not generally be read into the claims____ What is patented is not restricted to the examples, but is defined by the words in the claims.”) It is usually incorrect to read numerical precision into a claim from which it is absent, particularly when other claims contain the numerical limitation. In D.M.I., Inc. v. Deere & Co., 755 F.2d 1570,1574, 225 USPQ 236,239 (Fed.Cir.1985), the court stated:
Where, as here, the limitation sought to be “read into” a claim already appears in another claim, the rule is far more than “general.” It is fixed. It is long and well established. It enjoys an immutable and universally applicable status comparatively rare among rules of law. Without it, the entire statutory and regulatory structure governing the drafting, submission, examination, allowance, and enforceability of claims would crumble.
When a limitation is included in several claims but is stated in terms of apparently different scope, there is a presumption that a difference in scope is intended and is real. Tandon Corp. v. United States Int’l Trade Comm’n, 831 F.2d 1017, 1023, 4 USPQ2d 1283, 1288 (Fed.Cir.1987). Such a presumption can be overcome, but the evidence must be clear and persuasive. Conversely, it is incorrect to construe a claim as encompassing the scope that was relinquished in order to obtain allowance of another claim, despite a difference in the words used. See Builders Concrete, Inc. v. Bremerton Concrete Prods. Co., 757 F.2d 255, 260, 225 USPQ 240, 243 (Fed.Cir.1985).
All rules of construction must be understood in terms of the factual situations that produced them, and applied in fidelity to their origins. Thus, although Modine stresses the rule that the description of the preferred embodiment in the specification does not limit the claims to that embodiment, when the preferred embodiment is described in the specification as the invention itself, the claims are not necessarily entitled to a scope broader than that embodiment. See Autogiro Co. of America v. United States, 384 F.2d 391, 398, 181 Ct.Cl. 55, 155 USPQ 697, 703 (1967) (“where the patentee describes an embodiment as being the invention itself and not only one way of utilizing it,” this description guides understanding the scope of the claims).
Modine accurately states that the ’580 specification describes “about 0.015-0.040 inch” as the “preferred embodiment.” However, it is the only embodiment remaining for, as we shall discuss, the broader range of 0.015-0.070 inch was removed from the specification on refiling of the patent application. Modine thus limited the invention of the ’580 patent to hydraulic diameters in the range of about 0.015-0.040 inch. Although Modine states that it neither abandoned, nor intended to abandon, the broader range of hydraulic diameters, and is prosecuting in another application claims that include hydraulic diameters up to 0.07 inch, that does not control the construction of the claims of the ’580 patent.
Thus we agree with the Commission that “relatively small” in claims 9 and 10, inter[1552]*1552preted in light of the ’580 specification and the prosecution history, is not entitled to the range of up to 0.070 inch as sought by Modine. However, the Commission erred in literally restricting the hydraulic diameter range to an upper limit of exactly 0.040 inch, and in barring access to the doctrine of equivalents. See Hilton Davis Chem. Co. v. Warner-Jenkinson Co., 62 F.3d 1512, 1522, 35 USPQ2d 1641, 1648 (Fed.Cir.1995) (en banc) (“The trial judge does not have discretion to choose whether to apply the doctrine of equivalents when the record shows no literal infringement.”)
1. The Specifícation and the Prosecution History
The ’580 patent evolved from two continuation-in-part applications. The first-filed application, called the “grandparent,” described condenser tubes with flow paths having hydraulic diameters in the range of “about 0.015-0.070” inch. The specification stated that “heat transfer is increased in the range of hydraulic diameters of about 0.015 inches to about 0.070 inches through the use of the invention with some variance depending upon air flow.” The specification also stated that the preferred range was “about 0.015-0.040” inch. The specification included a graph of heat transfer as a function of hydraulic diameter, and described the graph as showing improved performance at hydraulic diameters up to about 0.070 inch.
In the second-filed application, called the “parent,” the hydraulic diameter upper limit of “about 0.070” inch in the grandparent was replaced, at every occurrence in the text, with “about 0.040” inch. The graph showing improved performance at hydraulic diameters up to 0.070 inch was retained in the application, but the explanatory text now described the graph as showing that
heat transfer is advantageously and substantially increased in the range of hydraulic diameters of about 0.015 inches to about 0.040 inches through the use of the invention with some variance depending upon airflow.
In the ensuing prosecution, when the examiner objected that the specification did not show “criticality” of the 0.015-0.040 inch range, Modine argued that this was the peak range “and it is this peak heat range that is sought to be covered by the applicant.” During prosecution of the parent application Modine told the patent examiner of the “Cat-Folded Front” condenser that was made by Modine for the Caterpillar Company for use in tractors, and sold more than a year before the filing date of the grandparent application. The Cat condenser had several structural differences from the condenser described in these applications: it had an overall hydraulic diameter of 0.0496 inch (or 0.04822, the record shows both figures); it did not have a web joined to the tube walls; and it did not have a plurality of elongated crevices in the flow paths. The Cat condenser was treated as prior art, along with several cited references.
Modine again refiled the patent application (the “child” application), without further change in the description of the hydraulic diameter. Although Modine points out that hydraulic diameters up to 0.070 inch continued to be shown in the graph that appeared in all three applications, the replacement of 0.070 with 0.040 in the text requires the conclusion that the applicant limited the invention described in the refiled applications to hydraulic diameters of up to about 0.040 inch.
Modine argues that the limitation in the parent/child specifications to hydraulic diameters of about 0.015-0.040 inch was not required by the prior art, pointing out that it is appropriate to consider not only the changes made during prosecution but also the reason for the changes. See Insta-Foam Prods., Inc. v. Universal Foam Sys., Inc., 906 F.2d 698, 703, 15 USPQ2d 1295, 1298 (Fed.Cir.1990) (“A close examination must be made as to not only what was surrendered, but also the reason for such a surrender.”) Although Modine may be correct that it was not necessary to reduce 0.070 to 0.040, this change was conspicuous and unambiguous. It was made in the context of the cited references and the Cat condenser, and the interested public is entitled to rely on it in interpreting the claim term “relatively small” as used in the ’580 patent.
[1553]*1553
2. The Incorporation by Reference
The ’580 specification incorporates by reference a patent entitled “Method of Making a Heat Exchanger,” U.S. Patent No. 4,688,311 (the ’311 patent). The ’580 specification states that the ’311 patent describes a “highly preferred means by which the tubes 20 with accompanying spacers 40 may be formed.” Modine places great weight on this incorporation to support its position that “relatively small” is correctly construed to mean hydraulic diameters up to 0.07 inch, for the ’311 patent states:
The invention may be used with particular efficacy where the flow passages are to be of relatively small hydraulic diameter as, for example, 0.07 inches or less. When such dimension is selected, particularly where the hydraulic diameter is 0.040 inches or less, the structure is ideal for utilization in a high efficiency condenser.
Modine stresses that the ’311 patent defines “relatively small” as “0.07 inch or less,” and that incorporation by reference has the same effect as if the host patent had set forth the entire text of the incorporated document. See In re Lund, 376 F.2d 982, 989, 153 USPQ 625, 631 (C.C.P.A.1967).
However, incorporation by reference does not convert the invention of the incorporated patent into the invention of the host patent. The use of “relatively small” to describe the condenser flow paths to which the ’311 manufacturing method is applicable did not reinstate into the parent and the child applications the hydraulic diameter range that was deleted from the grandparent. The words “relatively small” appear only once in the ’580 specification, and do not change the presentation of the invention as defined by the hydraulic diameter numerical range stated throughout the specification:
In addition to the utilization of a relatively small hydraulic diameter for the flow paths as mentioned previously, as another facet of the invention, it is contemplated that each of the flow paths have at least one crevice preferably extending along the entire length of the flow path, ...
“Relatively small” does not appear to have an independent meaning in this art that would distinguish between 0.040 and 0.070 inch. In view of the replacement of the “relatively small” parameter of about 0.070 in the grandparent application with the “relatively small” parameter of about 0.040 in the parent and child applications, it does not appear to be correct to read the scope of “relatively small” in the ’311 patent as overriding that replacement.
3. The Prosecution of Claims 9 and 10 (Application Claims 27 and 28)
The claims in suit were added near the end of the prosecution of the child application, and contain the only usage of “relatively small” in the claims. During prosecution Modine cancelled allowed claim 10 and added application claims 25-28. Modine points out that when application claims 25-28 were added Modine told the examiner that the claims were broader than allowed application claim 10 in some respects and narrower in others:
New Claim 25 is somewhat like original Claim 10 although it is broader in some respects and narrower in others.
As regards new Claims 25-28, independent Claim 25 is basically directed to the improved strength characteristic of the invention, See page 7 of the application as filed. Should the Examiner believe it to be dispositive, or otherwise helpful in determining the patentability of these claims, counsel notes that when the hoop strength containment feature of the Cat folded front type tubes is removed by removing the plate fins at the rounds of the flattened tubes, that is, at the radius of the flattened tubes, deformation and/or bursts occur at about 1,800 PSI; whereas the tubes of the present invention remained intact at pressures at least as high as 2,000 PSI. These sort of results are believed to commend the patentability of independent Claim 25.
Thus in presenting this group of claims Modine focused on the additional strength assertedly provided by the web means, not that these claims broadened the scope of the hydraulic diameter. Independent claim 25 used the words “relatively small hydraulic diameter.” Showa argues, and we agree, that the examiner viewed all the claims as referring to the hydraulic diameter range of 0.015 to [1554]*15540.04, for the examiner stated at one point during the prosecution of this claim:
Therefore, to achieve increased heat transfer and to save on material consumption it would have been obvious to one of ordinary skill in the art of heat exchange to reduce the hydraulic diameter of flow passages to within the claimed range of 0.015 to 0.04 as taught by Asselman et al.
The examiner persisted in the rejection of application claims 25 and 26, but held that claims 27 and 28 were allowable. In view of the prosecution history, we conclude that “relatively small” in application claims 27 and 28 (patent claims 9 and 10) did not enlarge the range of hydraulic diameters beyond that described in the specification as the invention.
The specification and prosecution history of the ’580 patent do not permit a construction of “relatively small” to include the 0.070 inch range that was described in the grandparent application, when that range was reduced in the parent and child to about 0.040 inch. However, neither are the claims correctly construed as limited to exactly 0.040 inch. Although the Commission correctly held that “relatively small” in claims 9 and 10 is limited by the description of the invention in the specification, the Commission incorrectly limited the hydraulic diameter to exactly 0.040 inch, for that is not the description in the specification and is not required by the prosecution history.
The specification uses the qualifier “about,” and also states that the optimum hydraulic diameter varies with the conditions. Such broadening usages as “about” must be given reasonable scope; they must be viewed by the decisionmaker as they would be understood by persons experienced in the field of the invention. Andrew Corp. v. Gabriel Electronics, Inc., 847 F.2d 819, 821-22, 6 USPQ2d 2010, 2013 (Fed.Cir.), cert. denied, 488 U.S. 927, 109 S.Ct. 312, 102 L.Ed.2d 330 (1988). Although it is rarely feasible to attach a precise limit to “about,” the usage can usually be understood in light of the technology embodied in the invention. When the claims are applied to an accused device, it is a question of technologic fact whether the accused device meets a reasonable meaning of “about” in the particular circumstances. Thus we turn to the factual aspects of the infringement determination.
C. LITERAL INFRINGEMENT
The record shows hydraulic diameter ranges of the nine Showa models before the Commission as follows: 0.0484-0.0519 inch; 0.0453-0.0520 inch; 0.0477-0.0577 inch; 0.0577-0.0606 inch; 0.0482-0.0497 inch; 0.061-0.065 inch; 0.0445-0.0682 inch; 0.0424-0.0573 inch; and 0.0513-0.0547 inch. The ALJ also referred to a model having a range of 0.0453-0.0477 inch. The Commission, incorrectly limiting “relatively small” to precisely 0.040 inch, did not consider variability based on the nature of the coolant, as stated in the specification, and did not determine whether any of the Showa models were within a reasonable literal scope of “relatively small” interpreted as meaning “about 0.015-0.040” inch.
Precedent illustrates the fact-dependency of determinations of the technologic scope of “about” and similar terms, depending on their contexts and the precision or significance of the measurements used. See, e.g., Quantum Corp. v. Rodime, PLC, 65 F.3d 1577, 36 USPQ2d 1162 (Fed.Cir.1995) (the addition of “approximately” during reexamination was a significant broadening of the claims); Hybritech, Inc. v. Abbott Labs., 849 F.2d 1446, 1455, 7 USPQ2d 1191, 1199 (Fed. Cir.1988) (“at least about 108 liters/mole” is literally satisfied by 4.8 x 107 liters/mole and 7.1 to 7.5 x 107 liters/mole); W.L. Gore & Assoc., Inc. v. Garlock, Inc., 842 F.2d 1275, 1280, 6 USPQ2d 1277, 1282 (Fed.Cir.1988) (an “imprecise limitation, such as the phrase ‘about 100% per second’ ” is to be considered in determination of infringement); Rosemount, Inc. v. Beckman Instruments, Inc., 727 F.2d 1540, 1546-47, 221 USPQ 1, 7 (Fed. Cir.1984) (“ ‘close proximity’ is as precise as the subject matter permits”).
The Commission’s determination of literal infringement was based on a hydraulic diameter limit of exactly 0.040 inch, with no consideration of the scope of “about” and no determination of the effect of relevant factors such as the nature of the coolant and the [1555]*1555precision of measurement. The finding of non-infringement was based on an incorrect claim construction, leading to an inadequate application of the claims to the accused devices. The finding is vacated. On remand the Commission shall determine whether any of the accused condenser models literally infringes the claims, upon construction of the claim term “relatively small” as meaning a hydraulic diameter in the range of about 0.015-0.040 inch, and upon applying the claims to the various accused Showa models.
D. INFRINGEMENT BY EQUIVALENCY
The Commission held that the doctrine of equivalents did not apply because Showa did not “unscrupulously” copy the Modine condenser. This ruling was based on an incorrect view of the law. As was explained in Hilton Davis, 62 F.3d at 1522, 35 USPQ2d at 1648, there is no equitable threshold for determination of the factual question of infringement by equivalency. Although the ALJ received evidence on the facts relevant to equivalency, the findings were made in the context of an incorrect view of prosecution history estoppel, the ALJ holding that Modine’s claims were limited to hydraulic diameters no larger than exactly 0.040 inch. The ALJ’s determination of estoppel was based on the same factors that led to the incorrect claim interpretation.
Modine argued to the Commission that the teachings in the specification, including the graph showing superior coolant activity up to a hydraulic diameter of 0.070 inch, support a range of equivalents up to 0.070 inch. Modine alternatively argued that even if the Commission found that Modine was not entitled to this range of equivalents, it is entitled to establish equivalency up to the overall hydraulic diameter of the prior art Cat condenser, i.e. about 0.48-0.49 inch.
Discussing the imported accused condensers, the ALJ found as fact that their function and result are the same as those of the claimed invention, but that the imported condensers do not meet the “same way” test because of the presence of internal fins in some of the Showa models. However, Modine’s evidence was substantially unrebutted that the presence of inner fins did not substantially change the way the condensers function, by surface tension and capillary forces. Although the intervenors argue that their condensers with larger hydraulic diameters are less efficient, equal performance is not required to establish equivalency. Laitram Corp. v. Cambridge Wire Cloth Co., 863 F.2d 855, 859, 9 USPQ2d 1289, 1294 (Fed. Cir.1988), cert. denied, 490 U.S. 1068, 109 S.Ct. 2069, 104 L.Ed.2d 634 (1989). The ALJ’s finding is against the heavy weight of the evidence. There was not substantial evidence supporting the finding of non-equivalence.
However, the ALJ correctly recognized that prosecution history estoppel limits the application of the doctrine of equivalents, even when the function/way/result or other test of equivalency is met by the accused devices. Prosecution history estoppel implements the principle that a patentee can not obtain, in an infringement suit, protection of subject matter that was relinquished in order to obtain allowance of other subject matter during prosecution of the patent application. Mannesmann Demag Corp. v. Engineered Metal Prods. Co., 793 F.2d 1279, 1285, 230 USPQ 45, 48 (Fed.Cir.1986) (the relinquished subject matter must be material to the issuance of the patent). The standard for determining whether particular subject matter was relinquished and was material is an objective one which we determine as a matter of law, LaBounty Mfg., Inc. v. United States Int’l Trade Comm’n, 867 F.2d 1572, 1576, 9 USPQ2d 1995, 1998 (Fed.Cir.1989), and is based on the reasonable reading, by a person of skill in the field of the invention, of the entire prosecution history.
We have discussed the prosecution history ante, and concluded that in connection with the patent application that led to the ’580 patent, Modine relinquished the range of hydraulic diameters that extended to 0.070 inch, based in substantial part on the hydraulic diameter of the prior art Cat-Folded Front condenser. Although Modine points out that the ’580 invention differs in several respects from the Cat condenser, the prosecution history shows that the hydraulic [1556]*1556diameter of the Cat condenser was a factor in limitation of the ’580 claims. The change in the description of the hydraulic diameter in the specification from grandparent to parent/child application, and the arguments to the patent examiner, highlighted the applicant’s action in distinguishing the ’580 claims from the Cat condenser.
Thus we conclude that the available range of equivalency is limited, by estoppel, to the hydraulic diameter of the Cat condenser. Within this boundary, however, the prosecution history and the prior art do not eliminate equivalents if substantial identity is shown. The controlling criterion, as reaffirmed in Hilton Davis, 62 F.3d at 1518, 35 USPQ2d at 1645, is whether the accused device is substantially the same as the claimed invention. See Graver Tank & Mfg. Co. v. Linde Air Prods. Co., 339 U.S. 605, 607, 70 S.Ct. 854, 855-56, 94 L.Ed. 1097 (1950) (insubstantial changes do not avoid the application of the doctrine of equivalents).
The ALJ incorrectly held that Modine was estopped to assert equivalency against any condenser with a hydraulic diameter larger than exactly 0.040 inch. The holding is vacated, and the case is remanded to the Commission for findings in accordance with the doctrine of equivalents.
Ill
VALIDITY AND ENFORCEABILITY
The Commission held that the ’580 patent was valid and enforceable. The intervenors appeal these holdings, stating that there is an on-sale bar, that the claims are invalid for obviousness or indefiniteness, and that Modine committed inequitable conduct in its prosecution of the ’580 patent.
A Prior Sale, 35 U.S.C. § 102(b)
If Modine is entitled to the filing date of its grandparent or parent application, 35 U.S.C. § 120, the parties agree that there is no on-sale bar based on Modine’s own sales of the claimed condenser. The Commission found that the two-pass condenser, as it was described in various documents, was described as well as enabled in the parent application. Although Mitsubishi argues that the claims of the parent application did not describe a two-pass condenser, the Commission found, and we agree, that the disclosure was present in the specification. A later application is entitled to the earlier filing date for all common subject matter that is contained in the earlier application, whether the subject matter appears in the body of the specification or in the claims or drawings.
The Commission found that the parent application adequately described, in compliance with 35 U.S.C. § 112 ¶ 1, the subject matter that is now claimed. See Ralston Purina Co. v. Far-Mar-Co. Inc., 772 F.2d 1570, 1574, 227 USPQ 177, 178 (Fed.Cir.1985). The Commission’s findings are supported by substantial evidence. The determination that there is not an on-sale bar is affirmed.
B. Obviousness, 35 U.S.C. § 103
The Commission held that invalidity based on obviousness had not been established, in that there was no teaching or suggestion that prior art condensers should be modified in the several ways that are embodied in the ’580 invention. The Commission appropriately considered the superior results achieved, the long-felt need for an automotive condenser that would overcome the limitations in cooling capacity and coolant nature that had been reached by conventional automotive condensers, and the commercial success of the ’580 invention. The Commission observed that reduction of hydraulic diameter had previously been associated with an undesirable increase in refrigerant-side pressure drop, and that Modine’s result was unexpected as well as superior.
Reversible error has not been shown in the Commission’s determination that the patent is not invalid on the ground of obviousness. The decision is affirmed.2
[1557]*1557
C. Indefiniteness, 35 U.S.C. § 112
The intervenors argue that the claims are invalid for indefiniteness if “relatively small” is construed as larger than exactly 0.040 inch. The ALJ had adopted this position. However, as we have discussed ante, technical terms are not per se indefinite when expressed in qualitative terms without numerical limits.
When claims are amenable to more than one construction, they should when reasonably possible be interpreted so as to preserve their validity. Whittaker Corp. by its Technibilt Div. v. UNR Indus., Inc., 911 F.2d 709, 711, 15 USPQ2d 1742, 1744 (Fed.Cir.1990); ACS Hosp. Sys., Inc. v. Montefiore Hosp., 732 F.2d 1572, 1577, 221 USPQ 929, 932 (Fed.Cir.1984). In this case the specification itself used the terms “relatively small,” and “about 0.015-0.040,” and the construction required to preserve the claims’ validity was simply that “relatively small” and “about 0.015-0.040” not include invalidating prior art. It was evident from the prosecution history that the patentability of claims 9 and 10 did not require an exact numerical limit of the hydraulic diameter. Mathematical precision should not be imposed for its own sake; a patentee has the right to claim the invention in terms that would be understood by persons of skill in the field of the invention. See Shatterproof Glass Corp. v. Libbey-Owens Ford Co., 758 F.2d 613, 624, 225 USPQ 634, 641 (Fed.Cir.), cert. dismissed, 474 U.S. 976, 106 S.Ct. 340, 88 L.Ed.2d 326 (1985) (“if the language is as precise as the subject matter permits, the courts can demand no more”).
Thus although we do not endorse the Commission’s reason, the Commission’s holding that the claims are not invalid for indefiniteness is affirmed.
D. Inequitable Conduct
The Commission held that the ’580 patent was not unenforceable for inequitable conduct, finding that there was no withholding of material information with intent to deceive or mislead the patent examiner into granting the patent. Both materiality and intent are essential factual predicates of inequitable conduct, and each must be proved by clear and convincing evidence. Kingsdown Medical Consultants, Ltd. v. Hollister Inc., 863 F.2d 867, 9 USPQ2d 1384 (Fed.Cir.1988) (en banc), cert. denied, 490 U.S. 1067, 109 S.Ct. 2068, 104 L.Ed.2d 633 (1989).
The ALJ had criticized Modine’s choice of the data that were included in the grandparent application as filed, for Modine had replaced a graph of computer-generated heat transfer data, that appeared in an early draft of the patent application, with later-obtained data and comparison with a different prior art condenser. Modine explained at trial the flaws in the first set of data, and the reasons for the change to data that were believed to be more accurate and to present a more useful comparison. Although there was no challenge at trial to either the correctness or the veracity of this explanation, the ALJ nonetheless found that Modine intended to deceive the patent examiner. The ALJ also criticized Modine’s description of the prior art and the arguments presented to the examiner concerning the prior art.
The Commission found that comparative data with Modine’s most efficient prior condenser were included in graphs in the patent application, and that certain computer-generated early data were replaced with more accurate data. Substantial evidence supports the Commission’s findings that there was neither material withholding nor intent to deceive in Modine’s selection of data and in the prosecution of the patent application. We remark that the rule of Kingsdown evolved in response to the “plague” of collat[1558]*1558eral attacks, of which this is an example, wherein routine patent practice is challenged without substance. See Northern Telecom, Inc. v. Datapoint Corp., 908 F.2d 931, 939, 15 USPQ2d 1321, 1327 (Fed.Cir.1990).
The holding that there was not inequitable conduct is affirmed.
Summary
Having interpreted the claims de novo, we vacate the Commission’s rulings on the issues of infringement, and remand for findings and redetermination with respect to literal infringement and infringement under the doctrine of equivalents. In all other respects the Commission’s decision is affirmed.
AFFIRMED IN PART, VACATED IN PART, AND REMANDED
MAYER, C.J., dissents.