2017 WL 239344, at *2 & n.6 (D. Del. Jan.
19, 2017). Evidence about the stamped AXN products
indicated stamping by a single non-commercial “prototype
shop” named Vehma. J.A. 1061–62. The evidence indi-
cated that the stamped AXN product had a UTS that did
not exceed 1,500 MPa. ArcelorMittal’s own expert testi-
fied that, after hot-stamping, the AXN product had a UTS
of 1,442 MPa. ArcelorMittal,
2017 WL 239344, at *2 &
n.6. ArcelorMittal’s expert further testified that a UTS of
1,442 MPa would be “equivalent to something that’s a
little bit over 1,500 [MPa].”
Id. at *2. Evidence also
indicated that Ford Motor Company sought to order steel
sheets from AK Steel that were to be hot-stamped by a
commercial hot-stamper for production of automobile
parts, but that those orders were never fulfilled.
Id.
The 2010 action proceeded to trial and, in January
2011, resulted in a jury verdict of non-infringement under
the doctrine of equivalents, “in part based on the lack of
evidence of the hot-stamped steel sheet meeting the
district court case, No. 1:13-cv-00685-SLR, filed in 2013
and involving primarily the RE940 patent. A detailed
summary of the 2010 action is set out in the opinions of
this court in the prior appeals. See ArcelorMittal I, 700
F.3d at 1317–19; ArcelorMittal Fr. v. AK Steel Corp.,
786
F.3d 885, 887–88 (Fed. Cir. 2015); ArcelorMittal v. AK
Steel Corp.,
856 F.3d 1365, 1367–68 (Fed. Cir. 2017).
Only the history relevant to this appeal is recited in this
opinion.
6 ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION
limitation requiring a [UTS] of 1500 MPa.”
Id. at *3.
ArcelorMittal appealed, and this court remanded after
reversing in part on a separate issue.
3. The 2013 Litigation
In April 2013, while the 2010 case was on remand,
ArcelorMittal filed the complaint related to this appeal.
ArcelorMittal first asserted the RE153 patent and later
amended the complaint to substitute the RE940 patent.
AK Steel subsequently filed a motion to dismiss the case
on the basis of collateral estoppel, arguing that the action
was estopped by the verdict in the 2010 action. According
to AK Steel, the accused products in this case—steel
sheets marketed and sold under the trade name
ULTRALUME—are the same as the AXN steel sheets in
the 2010 case.
In response, ArcelorMittal argued that new evidence,
obtained after the 2011 verdict, established that AK
Steel’s ULTRALUME products were materially different
from the AXN products because the ULTRALUME sheets
were hot-stamped to achieve a UTS exceeding 1,500 MPa.
In particular, ArcelorMittal relied on a declaration by its
employee, Marc Millius. The declaration provided that
Mr. Millius became aware, as of December 2012, that AK
Steel was supplying the market with steel sheets that
were hot-stamped by a commercial hot-stamper 2 to a UTS
exceeding 1,500 MPa.
On April 19, 2016, the district court denied AK Steel’s
motion to dismiss. The court allowed discovery, but
limited it to a single deposition of an AK Steel representa-
tive and the production of AK Steel’s manufacturing
specifications for the period of 2010 to 2013. J.A. 538–39.
2 The specific hot-stamper is referred to generically
throughout this opinion to protect information designated
by the parties as confidential.
ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION 7
The district court reasoned that limited discovery was
necessary to resolve the question of whether the AK Steel
products had changed since the 2010 case.
Id. The
district court cautioned that it was “reluctant to simply
open the doors to discovery without some modicum of
proof (more substantive than the [Millius] declaration
already submitted) that the accused steel sheet products
have changed.”
Id. (emphasis added).
After limited discovery concluded, AK Steel moved for
summary judgment based on collateral estoppel. J.A. 1–2.
In support, AK Steel argued that there was no evidence
that its pre-stamped steel sheets or its conduct had
changed since 2010. See J.A. 8–9; 564–66; 582. AK Steel
reasoned that because the accused products were the
same products found not to infringe in the 2010 action,
collateral estoppel applied, and it was entitled to sum-
mary judgment as a matter of law.
In opposition, ArcelorMittal argued that the evidence
before the district court now showed that AK Steel com-
mercially marketed and supplied steel sheets to various
auto producers that were hot-stamped to a UTS exceeding
1,500 MPa. ArcelorMittal pointed to an AK Steel market-
ing brochure describing how ULTRALUME products
underwent hot-stamping that converted the steel’s crys-
talline microstructure into martensite, and that hot-
stamped steel sheets are sold as auto parts with a UTS of
“1,400 MPa and higher.” J.A. 1095–1100. ArcelorMittal
also submitted a supplemental declaration from Mr.
Millius that provided detail on his observations that AK
Steel’s sheets were hot-stamped to exceed a UTS of 1,500
MPa (and supplied to a major auto manufacturer 3 in
December 2012). The supplemental Millius declaration
3 As with the hot-stamper referred to above, the
specific auto manufacturer is referred to generically
herein.
8 ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION
included an exhibit consisting of a slide from a webinar
conducted by the auto manufacturer that he personally
attended, which includes a bar graph that shows the auto
manufacturer’s various suppliers of 1,500 MPa UTS steel
sheets, including AK Steel and ArcelorMittal. J.A. 1362–
66. According to the declaration and as shown on the
slide, the bar on the graph representing steel supplied by
AK Steel indicated a tensile strength exceeding 1,500
MPa.
Based on the evidence, ArcelorMittal requested addi-
tional discovery under Federal Rule of Civil Procedure
56(d). J.A. 1354–60. ArcelorMittal argued that addition-
al discovery was necessary to present additional facts
related to the accused products and business conduct
subsequent to the verdict in the 2010 action.
Id. The
district court denied the requested additional discovery
and granted AK Steel’s motion for summary judgment of
non-infringement.
The district court agreed that ArcelorMittal’s in-
fringement claims with respect to the RE940 patent were
barred by collateral estoppel. ArcelorMittal,
2017 WL
239344, at *3–4. The district court found that “the prod-
uct at issue is the same and cannot—prior to hot stamp-
ing—be infringing as the steel sheets have a tensile
strength of about 600 MPa.”
Id. at *3. The district court
noted a lack of evidence in the 2010 case that AK Steel’s
steel sheets were hot-stamped to achieve a UTS exceeding
1,500 MPa. According to the district court, “[t]hat [AK
Steel] may ship the steel sheets to a third party hot
stamper on behalf of its purchaser is of no consequence to
the infringement analysis.”
Id. The district court further
noted that neither the chemistry of the steel sheets nor
the tensile strength of the sheets as shipped by AK Steel
were in dispute.
Id. The district court concluded that
these facts foreclosed the possibility of direct infringe-
ment.
Id. The district court also concluded that with no
direct infringement, there could be no indirect infringe-
ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION 9
ment.
Id. at *4 (citing Limelight Networks, Inc. v. Akamai
Techs., Inc.,
134 S. Ct. 2111, 2115 (2014)).
In denying ArcelorMittal’s request for additional dis-
covery under Rule 56(d), the district court reasoned that
discovery was unnecessary because, in its view, the evi-
dence showed that auto manufacturers bought the steel
sheets from AK Steel and shipped them to the hot-
stampers or had AK Steel ship them on their behalf.
Id.
at *3 & n.9. The district court also rejected ArcelorMit-
tal’s arguments that its divided or indirect infringement
theories could justify discovery. Id. at *4.
DISCUSSION
We review summary judgment rulings under the law
of the relevant regional circuit, in this case the Third
Circuit. See Intellectual Ventures I LLC v. Capital One
Fin. Corp.,
850 F.3d 1332, 1337 (Fed. Cir. 2017). The
Third Circuit reviews the grant of summary judgment de
novo. Gonzalez v. Sec’y of Dep’t of Homeland Sec.,
678
F.3d 254, 257 (3d Cir. 2012). “When reviewing a grant of
summary judgment, the court must view the facts in the
light most favorable to the nonmoving party and draw all
inferences in that party’s favor.”
Id. (internal quotation
marks omitted). Similarly, whether collateral estoppel
applies is a question of law, reviewed de novo. See Jean
Alexander Cosmetics, Inc. v. L’Oreal USA, Inc.,
458 F.3d
244, 248 (3d Cir. 2006).
1. Collateral Estoppel of Non-infringement
A primary issue in a collateral estoppel analysis with
respect to non-infringement is whether the accused prod-
uct is the same—i.e., the issue sought to be precluded is
the same as that involved in the prior action—or whether
the accused products have materially changed. The
question on appeal, then, is whether changes, if any, in
the accused products are “material” for the purpose of
collateral estoppel application. We conclude that the
10 ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION
evidence in this case supports that the products are not
materially the same.
Under Third Circuit law, collateral estoppel applies
when “(1) the issue sought to be precluded [is] the same as
that involved in the prior action; (2) that issue [was]
actually litigated; (3) it [was] determined by a final and
valid judgment; and (4) the determination [was] essential
to the prior judgment.” Anderson v. Comm’r,
698 F.3d
160, 164 (3d Cir. 2012) (quoting In re Graham,
973 F.2d
1089, 1097 (3d Cir. 1992)). Unless circumstances have
materially changed such that they constitute controlling
facts, “collateral estoppel remains applicable.” See Scoop-
er Dooper, Inc. v. Kraftco Corp.,
494 F.2d 840, 846 (3d Cir.
1974). Where aspects of collateral estoppel involve sub-
stantive issues of patent law, we apply Federal Circuit
precedent. Aspex Eyewear, Inc. v. Zenni Optical Inc.,
713
F.3d 1377, 1380 (Fed. Cir. 2013).
In the patent context, the Federal Circuit has followed
Supreme Court precedent and held that defensive collat-
eral estoppel applies with respect to a patent’s validity if
(1) the patent was found invalid in a prior case that
proceeded to final judgment where all procedural oppor-
tunities were available to the patentee; (2) the issues
litigated were identical; and (3) the party against whom
estoppel is applied had a full and fair opportunity to
litigate. See Abbott Labs. v. Andrx Pharm., Inc.,
473 F.3d
1196, 1203 (Fed. Cir. 2007) (summarizing Blonder–
Tongue Labs., Inc. v. Univ. of Ill. Found.,
402 U.S. 313,
330–33 (1971)). The court should determine “whether a
patentee has had a full and fair chance to litigate the
validity of his patent” by considering “whether without
fault of his own the patentee was deprived of crucial
evidence or witnesses in the first litigation.”
Id. at 333.
Collateral estoppel may bar litigation in cases with differ-
ent but related patents when there are common issues.
See Amgen, Inc. v. Genetics Inst., Inc.,
98 F.3d 1328, 1332
(Fed. Cir. 1996).
ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION 11
With respect to non-infringement, where an alleged
infringer prevails, the accused products gain non-
infringing status and the alleged infringer acquires the
status of a non-infringer to the extent that the accused
products remain the same. Young Eng’rs, Inc. v. U.S. Int’l
Trade Comm’n,
721 F.2d 1305, 1316 (Fed. Cir. 1983). If
accused devices in a second suit remain “unchanged with
respect to the corresponding claim limitations at issue in
the first suit,” then the patentee is precluded from pursu-
ing her claims a second time. Nystrom v. Trex Co.,
580
F.3d 1281, 1285 (Fed. Cir. 2009). However, “conduct of a
different nature from that involved in the prior litigation”
will not be given preclusive effect. Young Eng’rs,
721 F.2d
at 1316.
Defensive collateral estoppel of non-infringement,
therefore, may apply in the “limited circumstances[]
where it is shown that a close identity exists between the
relevant features of the accused device and the device
previously determined to be [non-]infringing” such that
they are “essentially the same.” See Yingbin-Nature
(Guangdong) Wood Indus. Co. v. Int’l Trade Comm’n,
535
F.3d 1322, 1333 (Fed. Cir. 2008). “Accused devices are
‘essentially the same’ where the differences between them
are merely ‘colorable’ or ‘unrelated to the limitations in
the claim of the patent.’” Acumed LLC v. Stryker Corp.,
525 F.3d 1319, 1324 (Fed. Cir. 2008) (quoting Foster v.
Hallco Mfg. Co.,
947 F.2d 469, 480 (Fed. Cir. 1991)); see
also Brain Life, LLC v. Elekta Inc.,
746 F.3d 1045, 1055
(Fed. Cir. 2014) (observing that products that were not
“materially different” from products at issue in prior
litigation would bar the subsequent assertion of the same
claims). The proponent of claim or issue preclusion bears
the burden of showing that the accused devices are essen-
tially the same as those in the prior litigation. See Young
Eng’rs,
721 F.2d at 1316 (claim preclusion); In re Katz
Interactive Call Processing Patent Litig.,
639 F.3d 1303,
1311 (Fed. Cir. 2011) (issue preclusion).
12 ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION
2. Evidence of Material Differences
We conclude that AK Steel failed to meet its burden
that the ULTRALUME products in the present action are
materially the same as the AXN products in the 2010
action, especially when the facts are viewed in the light
most favorable to the non-movant, ArcelorMittal. In the
2010 action, there was no evidence that the AK Steel AXN
products met the “in excess of 1,500 MPa” UTS limitation
or that any AXN products were commercially stamped
and supplied to an auto manufacturer at all. In this
action, there is evidence that AK Steel is a supplier of
such products to at least one major auto manufacturer.
The September 2016 supplemental Millius declaration
supports the allegation that AK Steel is a supplier of steel
sheets that have a UTS exceeding 1,500 MPa to the auto-
producing industry. J.A. 1362–66. The declaration was
supported by an exhibit of a presentation slide taken from
a December 2012 auto manufacturer-hosted webinar that
provided tensile test information for hot-stamped steel
from several of the auto manufacturer’s suppliers.
Id.
The auto manufacturer’s webinar slide, entitled “Tensile
Strength Comparison by Supplier,” includes a bar graph
representing tensile strengths of steel sheets that, accord-
ing to the declaration, correspond to each supplier, includ-
ing ArcelorMittal and AK Steel.
Id. For both
ArcelorMittal and AK Steel, the graph indicates a UTS in
excess of 1,500 MPa.
Id. Although the parties dispute
the significance of the auto manufacturer’s tensile test
data, it is significant that the auto manufacturer’s own
2012 test data shows that AK Steel’s products were
thermally treated to achieve a UTS in excess of 1,500
MPa. J.A. 1364 ¶ 5, 1366. This evidence, when viewed in
the light most favorable to ArcelorMittal as the non-
movant, reflects a material difference in the accused
product and AK Steel’s conduct since the 2011 verdict.
See Anderson v. Liberty Lobby, Inc.,
477 U.S. 242, 255
(1986).
ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION 13
We disagree with AK Steel’s argument that this slide
is “identical to the testing introduced in the [2010] [c]ase.”
Appellee Br. 46. AK Steel points to testimony from Arce-
lorMittal’s expert in the 2010 action who heat treated and
tested several steel sheets from various suppliers, using
what he called a “conservative heat treatment” that
resulted in a UTS of 1,442 MPa. J.A. 705. But the auto
manufacturer’s test data shows strengths of steel hot-
stamped by a commercial hot-stamper (as named in the
first Millius declaration) to meet the auto manufacturer’s
commercial specifications. AK Steel in attorney argument
characterizes the auto manufacturer’s test data as “exper-
imental,” but there is no evidence supporting such a
characterization. Compare Appellee Br. 46, with J.A.
1364 ¶ 4. AK Steel’s attorney argument is not sufficient
to rebut ArcelorMittal’s evidence. See Icon Health &
Fitness, Inc. v. Strava, Inc.,
849 F.3d 1034, 1043 (Fed. Cir.
2017). In addition, that the auto manufacturer’s test data
was prior to the RE940 patent’s issue date is not material.
In the 2010 action, the evidence was that the AXN prod-
ucts were of “experimental grade,” i.e., not for sale, and
were thermally treated by the Vehma “prototype shop,”
i.e., non-commercial hot-stamper; there was no evidence
that those products were supplied to an auto manufactur-
er. See ArcelorMittal,
2017 WL 239344, at *2, 1061–63;
see also, e.g., J.A. 1065 (“The Vehma trial was rudimen-
tary. . . . [It] should not, by any means, be used as a mass
production trial.”). The evidence in this case indicates
that the production process is no longer “rudimentary,”
that commercial stampers are utilized, and that the
stamped steel is being commercialized. This crucial
evidence represents a material difference in the accused
products and did not exist during the 2010 action.
Finally, AK Steel’s own product brochure, made after
the verdict in the 2010 action, lends further support to
ArcelorMittal’s contention that AK Steel’s conduct and
the accused products have changed. The brochure ex-
14 ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION
plains that blanks produced from ULTRALUME are hot-
stamped to undergo “a phase transformation” to its micro-
structure that results in “a high strength phase – marten-
site.” J.A. 1089. The brochure advertises that this
process “increases the tensile strength of the steel from
approximately 600 MPa to 1,400 MPa and higher.”
Id.
(emphasis added). AK Steel’s brochure is evidence in this
action that supports an allegation that AK Steel is com-
mercially marketing and supplying steel sheets to auto
manufacturers that may exceed 1,500 MPa UTS after hot-
stamping.
At minimum, this evidence contravenes the undisput-
ed evidence in the 2010 action that AXN steel sheets hot-
stamped by Vehma did not exceed 1,500 MPa UTS, and
that none of the stamped AXN sheets were supplied to, or
used by, auto manufacturers. In aggregate, this evidence
constitutes, as the district court put it, more than a modi-
cum of proof that the accused products materially differ
from those in the 2010 action with respect to claimed
limitations. Differences with respect to the claimed
limitations constitute changes in controlling facts, such
that collateral estoppel does not apply. See Scooper
Dooper,
494 F.2d at 846. For this reason, the district
court erred in finding that the accused products cannot as
a matter of law infringe and granting summary judgment
of non-infringement on the basis of collateral estoppel.
We express no opinion on whether the RE940 patent is
actually infringed or who might infringe it, only that
there is now evidence that the accused products, the
stamped steel sheets originating from AK Steel and
supplied to auto manufacturers, are materially different
from the products in the 2010 action.
3. The Relevance of Hot-Stamping
The parties dispute the extent to which there may be
differences in the chemical composition between the AXN
and ULTRALUME products and the effect of those differ-
ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION 15
ences, but those are not the only facts relevant to the
claim limitations at issue. Consistent with the claim
language of the patent, the mechanical resistance of the
accused products after the claimed thermal treatment,
i.e., hot-stamping, which transforms the crystalline mi-
crostructure of the steel (as opposed to its chemical com-
position), is also relevant to the claim limitations at issue.
The district court erred when it focused solely on the
pre-stamped product and stated “[t]hat [AK Steel] may
ship the steel sheets to a third party hot stamper on
behalf of its purchaser is of no consequence to the in-
fringement analysis.” ArcelorMittal,
2017 WL 239344, at
*3. The asserted product-by-process claims are not di-
rected to unstamped blanks that have not yet been sub-
ject to the claimed thermal treatment process step. It is
well established that “[e]ach element contained in a
patent claim is deemed material to defining the scope of
the patented invention.” Warner-Jenkinson Co. v. Hilton
Davis Chem. Co.,
520 U.S. 17, 29 (1997). As we have
previously affirmed, “process terms in product-by-process
claims serve as limitations in determining infringement.”
Abbott Labs. v. Sandoz, Inc.,
566 F.3d 1282, 1293 (Fed.
Cir. 2009) (en banc) (quoting Atl. Thermoplastics Co. v.
Faytex Corp.,
970 F.2d 834, 846–47 (Fed. Cir. 1992)). An
“inventor is absolutely free to use process steps to define
[a] product” and “will not be denied protection.” Id. at
1294. The court cannot ignore the inventor’s definition of
her product. Id. In view of the claim language, the
claimed thermal treatment process step, i.e., hot-
stamping, and resulting UTS, is of significant conse-
quence to the infringement analysis.
The district court recognized that the jury verdict was
based on a lack of evidence that AK Steel’s 2010 hot-
stamped steel sheet met the claimed thermal treatment
limitation requiring a UTS of 1,500 MPa. Yet in conclud-
ing that the product now at issue cannot infringe, the
district court instead focused on the accused product prior
16 ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION
to hot-stamping, having the same composition as the 2010
products and a tensile strength of about 600 MPa. By
simply focusing on the composition and mechanical re-
sistance of the pre-processed, blank steel sheets, and by
ignoring the claimed thermal treatment process step
conducted by hot-stampers and the resulting UTS, the
district court erred in determining whether a material
claimed element in the patent as defined by the inventor
had changed in the accused products. See ArcelorMittal,
2017 WL 239344, at *3–4. Indeed, it is the thermal
treatment and resulting UTS claim elements that Arce-
lorMittal contends have changed since the 2011 jury
verdict, making the now-accused products materially
different from the products in the 2010 action.
Contrary to the district court’s finding, that the ther-
mal treatment is performed by a third party hot-stamper
does not foreclose the possibility of direct infringement
here. See
id. If the third party hot-stampers produce
steel sheets with a UTS exceeding 1,500 MPa after ther-
mal treatment, as ArcelorMittal alleges, the hot-stampers
might very well directly infringe the RE940 patent.
Therefore, the district court erred in dismissing Arce-
lorMittal’s potential indirect infringement claims due to a
lack of direct infringement. See
id. at *4 (citing Limelight
Networks,
134 S. Ct. at 2115).
4. Additional Discovery Is Warranted
In granting summary judgment, the district court de-
nied ArcelorMittal’s request for additional discovery to
oppose AK Steel’s motion for summary judgment under
Rule 56(d). The district court based its decision on its
finding that “the product at issue [in this case] is the
same [as the product at issue in the 2010 case] and can-
not—prior to hot stamping—be infringing as the steel
sheets have a tensile strength of about 600 MPa.” Arce-
lorMittal,
2017 WL 239344, at *3.
ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION 17
The standard of review for the district court’s deci-
sions related to discovery under Rule 56(d) in the Third
Circuit is abuse of discretion. Renchenski v. Williams,
622 F.3d 315, 339 (3d Cir. 2010). To disturb such a deci-
sion, “[a]ppellants must demonstrate that the district
court’s action made it impossible to obtain crucial evi-
dence.” Wisniewski v. Johns-Manville Corp.,
812 F.2d 81,
90 (3d Cir. 1987) (internal quotation marks omitted).
Additionally, “[a]n abuse of discretion may be established
under Federal Circuit law by showing that the court made
a clear error of judgment in weighing the relevant factors
or exercised its discretion based on an error of law or
clearly erroneous fact finding.” Wind Tower Trade Coal.
v. United States,
741 F.3d 89, 95 (Fed. Cir. 2014) (quoting
Qingdao Taifa Grp. Co. v. United States,
581 F.3d 1375,
1379 (Fed. Cir. 2009).
As explained above, we conclude that new evidence of
material differences in the accused products exists. This
means that the district court erred in limiting discovery
on the basis of the issues presented in the 2010 action.
The denial of ArcelorMittal’s requested discovery kept it
from obtaining crucial evidence. See Wisniewski,
812 F.2d
at 90.
Discovery is necessary to determine whether and to
what extent AK Steel supplied auto manufacturers with
steel sheets with a UTS in excess of 1,500 MPa after
thermal treatment, its knowledge and intent in doing so,
and the nature of the relationships between AK Steel, the
hot-stampers, and the auto manufacturers during the
relevant timeframe. On remand, we direct the district
court to allow discovery as requested by ArcelorMittal
under Rule 56(d) in its opposition to summary judgment.
CONCLUSION
Because evidence shows material differences between
the accused product in this action and the product in
18 ARCELORMITTAL ATLANTIQUE v. AK STEEL CORPORATION
earlier litigation, the district court erred in applying
collateral estoppel. We vacate and remand.
VACATED AND REMANDED
COSTS
No costs.