RT remains the only means available to nondestructively test
EV motors for R&D and defect analysis.
Typically, the features that need to be inspected in both
major components of a motor—the rotor and the stator—are
the copper wire coils, their insulation, and the plastic or epoxy
materials used to hold everything in place. By its very nature,
imaging low-density components buried inside high-density
parts is one of the most difficult challenges in RT, because the
low-density parts experience tremendous scatter artifacts from
the higher-density parts. These artifacts can mask features in
the image and may also lead to false-positive identification of
discontinuities.
Though difficult, it’s not impossible if enough voltage,
power, and resolution are available, along with supplemen-
tary practices that include using filtration, collimation, and
scatter-correction software, as demonstrated by the sharply
rendered 3D image of a multi-material motor assembly shown
in Figure 14. Figure 15 shows a 2D slice taken through the
3D data, in which different materials are visible. Using this
method, it is possible to extract, for example, only the copper
wires to inspect them for short circuits or breaks without ever
needing to disassemble the motor.
The clarity of these images is achieved through mathemat-
ical scatter-reduction calculations. The 2D scanned images,
known as “projections,” are reconstructed into 3D data, and
the different materials are separated by their gray value, with
lighter shades of gray (or “grayscale”) representing denser
materials. Scatter-reduction software calculates what is true
material and what is scatter and subtracts the scatter from the
data. A second reconstruction is carried out to produce the
final dataset (see Figures 16a and 16b).
Conclusion
The challenges of providing NDT and measurement solu-
tions for EVs are different from those faced by the IC vehicle
industry. The nature of EV components is such that their
delicate parts are almost always buried and sealed to make
them robust, making it impossible for them to be inspected
using any method other than RT. The pressure to improve yield
rates for production components, along with the demand for
performance and improved product reliability from EV pow-
ertrains, is enormous—and probably greater than any other
pressure on the RT NDT industry today to provide solutions.
This article provided an overview of current practices, but it
is only a snapshot in time, as developments in RT are being
made as rapidly as breakthroughs and improvements in EV
powertrains themselves.
REFERENCES
1. Fraunhofer Research Institution for Battery Cell Production FFB. 2024.
“Mastering Ramp-up of Battery Production.” White paper. https://www.
ffb.fraunhofer.de/en/publications/White_papers_environment_reports_
studies/Mastering_Ramp-up_of_Battery_Production.html.
2. Excillum AB. 2025. “The metal-jet technology” [online]. Accessed
5 December 2025. https://www.excillum.com/products/metaljet/.
3. Nikon Metrology Inc. 2025. “Unique 225kV Rotating.Target 2.0" [online].
Accessed 5 December 2025. https://industry.nikon.com/en-us/products/x-
ray-ct/x-ray-source-technology/.
4. Varex Imaging Corporation. 2025. “Photon Counting" [online]. Accessed
5 December 2025. https://www.vareximaging.com/photon-counting/.
ME
|
ELECTRICVEHICLES
Figure 14. 3D rendering of the complete MeV scan volume of a multi-
material motor assembly.
Figure 15. 2D slice through the 3D MeV scan volume of the multi-
material motor assembly.
Figure 16. (a) Uncorrected 2D slice of MeV scan (b) corrected 2D slice of
MeV scan, using scatter-correction software.
60
M AT E R I A L S E V A L U AT I O N • J A N U A R Y 2 0 2 6
CREDIT:
VAREX
IMAGING
CREDIT:
VAREX
IMAGING
CREDIT:
VAREX
IMAGING

ABSTRACT
The increasing prevalence of battery systems in
automotive engineering, along with ever-shortening
development cycles, demands advanced and efficient
inspection methods to ensure performance, safety,
and longevity. This paper presents the capabilities
of the XXL computed tomography (CT) technology
tailored for large-scale battery analysis. While state-
of-the-art microfocus CT systems offer excellent
performance for cell-level inspections, their limitations
become apparent when addressing larger and more
complex battery cell assemblies. We therefore explore
use cases beyond the scope of conventional CT,
including large battery modules, complete packs
with multiple modules, cell-to-chassis architectures,
and even fully assembled electric vehicles. In these
scenarios, the XXL-CT—with its high penetration power
enabled by a 9 MeV linear accelerator (linac) X-ray
source—supports geometric evaluation of batteries
and their surroundings, as well as changes induced by
lifecycle and abuse testing or crash events, all without
disassembly. Finally, we provide an outlook on the
next evolutionary step in XXL-CT: a giant gantry-based
CT system and its potential to increase the application
range and throughput even further. This larger system
will begin operation in Fürth, Germany, at the end of
2026.
KEYWORDS: XXL-CT, high-energy CT, battery testing, gantry-CT,
MeV CT, linac, X-ray imaging
1. Introduction
Battery packs for electric vehicles (EVs) contain a few hundred
to several thousand tightly packed cells within complex metal
and polymer structures. Ensuring safety and quality through-
out the battery lifecycle requires nondestructive inspection of
internal features that are hidden once the pack is assembled.
X-ray computed tomography (CT) has become an indispens-
able inspection tool for batteries, supporting quality control
and failure analysis at all stages—from cell development
through production and end-of-life (EOL) assessment. State-
of-the-art microfocus CT systems with up to 450 kV and small
focal spots enable high-resolution imaging of individual cells.
Such systems excel at detecting small discontinuities (anode
overhang, foreign particle contamination, weld defects) in
cylindrical, prismatic, or pouch cells [1, 2, 3].
However, when battery cells are assembled into large
modules and packs, conventional CT systems encounter severe
limitations. Battery modules often span over a meter in size,
with a combination of highly absorbing metal enclosures and
other (often low-density) peripherals attached, which severely
affect image quality. While size limitations can sometimes be
addressed by using offset-scan or helical scan techniques, the
image quality issues from insufficient penetration power and
multimaterial mixes still remain.
These challenges become even more pronounced for
battery packs weighing up to 600 kg and extending up to
2 m in longitudinal direction. In these cases, laminography
has recently been adopted to examine the battery pack from
the top—i.e., the direction with the shortest X-ray penetra-
tion length—in order to detect small particles or winding
changes and errors in the assembled cells. However, due to
the physical limitations of this method, it is not possible to
obtain a precise depth representation that can visualize the
layer structure with features such as the height position and
orientation of each cell or intrusion depth in impact and roll-
over tests.
In practice, battery inspections with CT have therefore
largely been limited to single cells or cut‑down modules to
reduce overall penetrations lengths. But as battery modules
and EV packs grow larger, the necessity to inspect complete
assemblies for cell alignment or dimensional accuracy
becomes increasingly important [3]. This is especially relevant
in the context of cell-to-pack or cell-to-structure designs.
High-energy CT systems in the MeV range have emerged
to fill this gap. By using linear accelerator (linac) X-ray
XXL-CT FOR MULTISCALE BATTERY TESTING:
FROM MODULES TO COMPLETE ELECTRIC
VEHICLES
NILS REIMS*†, MICHAEL SALAMON†, AND MICHAEL BOEHNEL†
† Fraunhofer IIS (Institute for Integrated Circuits) – Development Center X-ray
Technology EZRT, Erlangen, Germany
*Corresponding author: nils.reims@iis.fraunhofer.de
Materials Evaluation 84 (1): 61–67
https://doi.org/10.32548/2026.me-04555
©2026 American Society for Nondestructive Testing
NDTTECHPAPER
|
ME
J A N U A R Y 2 0 2 6 • M AT E R I A L S E V A L U AT I O N 61