MULTIPLE NDT METHODS IN THE AUTOMOTIVE INDUSTRY

P. Buschke, W.Roye, T. Dahmen

GE Inspection Technologies, Huerth, Germany
Abstract: The automotive industry must meet the requirements for increased safety (product liability), stricter
regulations regarding environmental protection, as well as the demands of car drivers for more luxury features.
Costly recall actions, ever increasing demands on the quality management and the changing legal framework with a
simultaneously harsher competition force the industry towards a more economic direction. Improved accessories
for vehicles lead to increased weight. Higher weight also means higher gasoline consumption. This is only one of
the reasons for todays lightweight design of new vehicles. Whereas steel was with more than 80 percent the most
important material for the automotive industry in the past, other materials such as plastics, magnesium and
aluminum will become more important in the future. The use of several materials will inevitably lead to new
joining techniques since the traditional joining techniques, e.g. spot welding for body constructions, can only be
partly used. A combination of methods will replace a single dominant method in the future. Due to the changed
conditions with regard to material and joining technique, quality assurance must likewise adapt itself to the changes
in production conditions. New test techniques are not only meant to determine the current quality of a bonded joint
but also more and more to positively influence the whole process by corresponding measured variables. The
industry will have to invest in safety and quality in the future as well because the conditions related to product
liability will continue to cause recall actions and other cost-intensive measures. Advanced processes will be in
demand, measured by (NDT) variables in the field of quality management.
The lecture sheds light on feasible solutions in the automotive industry (e.g. quality inspection of joining techniques
in car body framework and of casted components) using various methods (ultrasonic, x-ray etc.) and the resulting
benefits for the users.
References: W. Roye Nondestructive Inspection Methods in the automotive industry, 8th ECNDT Conference,
Barcelona (Spain), June 17-21, 2002

Spot welds
The computer aided spot weld inspection, see Figure 1, has achieved a high state of the art and is applied
by almost all car manufacturers. Here, the principle is briefly discussed, see Figure 1:

Figure 1: Spot Weld Inspection

In the case of a good spot weld [1] we obtain an echo

sequence as shown. Due to the coarse grained material and
the related sound attenuation the echo sequence decays
quickly.
If the spot is too small [2], the echo sequence contains
intermediate echoes due to the fact that the sound field
diameter is larger than the spot diameter.

A stick joint [3] is transparent for the sound but the lower
sound attenuation leads to a longer echo sequence.

Finally, if there is no bonding [4], one obtains an echo

sequence from the first plate only.

Figure 2: The principle of spot weld

inspection

Laser seam welds

Overlapped laser seam welds, e.g. in the roof
channel of a car body, can be inspected by
means of the ultrasonic transmission
technique, as shown in figure 3: The emitter
probe E generates a guided wave through the
plate and the laser weld L. In case of a good
weld performance, the ultrasonic signal is
received by the receiver probe R.
First prototypes contained standard anglebeam probes with a flowing water coupling
device. But as any fluid for coupling purposes
is not permitted because of the later painting
process,
GE
Inspection
Technologies
developed a system with roller probes which
enable dry coupling. The system is shown in
Figure 4. It contains

Figure 3: Inspection Scheme

the roller probes,

a probe holder with springs, which enable the geometric adaptation to the changing surface of the object
construction,
a guide system which ensures the correct position of the probes along the whole length of the laser weld seam
and finally a position encoder.

The inspection is computer aided.

The computer receives the sound
amplitude values from the ultrasonic
system and the corresponding
position data from the position
encoder, and a dedicated software
thus allows an inspection procedure
including documentation.
This type of ultrasonic transmission
technique will not be able to detect
extremely small defects in the weld,
however, practical application tests
in several automobile companies
demonstrate, that all relevant bad
through welding defects can be
detected reproducibly.

Figure 4: The roller probe setup for the laser weld inspection

Gas shielded welding

Different components of a car body are joined by MIG (metal inert gas) or MAG (metal active gas) welds. In most
cases the geometry does not allow any type of ultrasonic impulse echo method. Therefore for this kind of weld, GE
Inspection Technologies proposes the ultrasonic transmission technique, as shown in Figure 5:

Also for this application a special probe

system was developed, see Figure 6:
The emitter probe contains three
transducers, which generate sound beams
into three different directions. The sound
waves propagate through the plates and
the weld as a guided waves and can be
received on the other side by means of
the second probe, e.g. K2MNE. In order
to simplify the inspection procedure, the
emitter probe is equipped with four
magnetic feet for self fixation to the
metal object.
Gas shield welding is applied e.g. for the
joints of door hinges, the bows of the
seat fixture and several other
components. In all cases a good weld is
indicated by a high ultrasonic amplitude
and a bad throughwelding by a lower
signal, Figure 7.

Emitter Probe
Receiver Probe

Figure 5: Inspection layout for MIG/MAG welds

Figure 6:Special probe W45/3xB2K

High amplitude
Good weld

Low amplitude
Bad weld

Figure 7: Ultrasonic inspection of MAG welds

Adhesive Joints
More and more adhesive joints are applied in the
automobile and aircraft industries. This concerns the
joints of metal plates as well as combinations of steel,
polymeric materials and glass.
The example in Figure 8, two adhesively bonded steel
plates with a thickness of 1 mm, were inspected.
Principally there are two possibilities:
If a relatively high sound frequency is used, e.g. 20
MHz, then we obtain a good echo sequence out of the
first plate, which is highly attenuated by the adhesive
if the adhesion is good, see Figure 9. In case of a bad
adhesion or missing adhesive, the sound damping is
non-existent, which leads to a long echo sequence.
However, a disadvantage of this method is, that only
the first interface between plate and adhesive can be
inspected, because the high frequency sound does not
interact with the adhesive material.
If we apply a low frequency, e.g. 2 MHz, then an
interaction with the adhesive can be observed,
however, it is not possible to obtain resolved echoes
in the time domain. For this case the spectral analysis
in the frequency domain is proposed.

Wasser

Plate
1
Blech 1
Adhesive
Kleber
Blech 2
Plate2

Figure 8: Inspection of adhesive joints

Figure 9: Echos of a 20MHz probe

As an example Figures 10 and 11 each present the AScans of the interfering signals of the plate and the
adhesive. A gate is set to the interference from the
adhesive. The time range of the gate is presented once
more in the left upper side and the corresponding
frequency spectrum is presented at the right side. The

Figure 10: Good adhesion

printed values mean: fp Peak frequency, fc Center

frequency and BB Bandwidth.
It can be clearly seen, that different gluing qualities lead
to different interference patterns in the time domain which
can easily be evaluated in the frequency spectrum.

Figure11: Bad adhesion

Laser welds of airbag gas generators

An airbag gas generator is shown in figure 12. It contains
several laser welds.
The partial design of the object with weld is presented in
Figure 13. The task is to measure the depth of the weld and
to detect defects, e.g. pores, in the welded zone. For this
purpose it is recommended to apply the ultrasonic
immersion technique in combination with the point focussed
probe H10MP15 (f =10 MHz, focus depth = 15 mm in
water). With a radial insonification and by means of an
object rotation and probe shifting a C-Scan-Image (Helical
Scan) is obtained, as presented in Figure 14.
Here, the echo amplitude is plotted to the rotation angle (0 to
360 degrees) and the probe shift (0 to 17 mm). The
amplitude is colour coded: red represents a high amplitude
of the backwall of the first plate or of defects in the weld
and green presents a low amplitude or missing signal. This
allows the detection and localisation of defects as well as the
evaluation and determination of the weld depth, as
demonstrated in the magnified C-Scan part, Figure 15.

Figure 12: Airbag gas generator

H 10 MP

~15

Figure 13: Partial section with Laser

Weld Seam

Figure 14: Circumferential C-Scan

The accessories used for this application were:
ultrasonic instrument: Krautkramer USD15,
probe: H10MP15,
an immersion tank with a shift axis and a
rotation device and
Software: UltraMAP for motor control, data
acquisition and generation of C-Scan images.

Weld Depths

Defect Indication

Figure 15: Magnified Part of the C-Scan

Bonding test between the cylinder liners and the casing of motor units
The considered motorblock unit is made of cast
aluminum and contains cylinder liners made of gray
cast iron with a wall thickness of 4 mm. The inside
diameter is 80 mm.
The task is to inspect the bonding between the cast
iron liner and the aluminium casting.
As shown in Figure 16, the wall of the liner is scanned by
vertical beaming using the immersion probe H5K and a 450
deflection mirror. The quality of bonding can be read from
the amplitude of the backwall echo: A small echo indicates
good bonding and a high signal the bad bonding.
In order to generate a C-scan, the total liner surface
area is helically scanned from inside. To achieve this,
the probe is shifted vertically, and the deflection
mirror is rotated around its axis.
Figure 17 shows a motor unit and the probe holder
with the deflection mirror. This system permits a
helical scan of the entire liner surface area.

Good Bonding

Bad Bonding

Figure 16: Al-motor block with cast iron liner

Figure 18 presents typical C-scans as a result of the

bonding test. The high amplitudes (bright) to be seen
there indicate areas of bad bonding, i.e. the lack of
fusion.
The accessories for this application were:
Instrument: Krautkramer USD15
Probe: H5K
Scanning system
Software: K-Scan
Figure 17: The motor block and the ultrasonic
probe with deflection mirror

Figure 18: C-scan-plots of 4 cylinders

Wall thickness measurement on cast camshafts

Typical casting defects are core mismatches.
The consequence of such defects in camshafts are failures because of extreme dynamic loads due to reduced wall
thicknesses.

Figure 19: Ultrasonic measurement of thicknesses on a camshaft

The recommended solution is to measure the wall thickness on the bearing points. The simple, annular geometry of
these points enable the shaft to be tested in the casted state. Further processing actions on any defective parts can
thus be excluded.

Figure 20: Lineplots of circumferential thickness measurement

This method makes it possible to recognise the core mismatch by a reduced wall thickness and excessive wall
thickness on the opposite side (180). The waterflow technique, using a so-called squirter or bubbler probe, is
applied according to the arrangement shown in the sketch, Figure 19. This ensures a constant coupling in the area of
the bearings around the entire circumference during camshaft rotation. Figure 20 presents two lineplots (thickness
to rotation angle). The upper plot represents a uniform wall thickness with a d < 0.6 mm and the lower plot shows
a varying thickness with a d > 2.5 mm.

Hardness test on the running surface of a camshaft

In order to attain adequate service lives, that means high resistance capacity to wear, camshaft running faces are
hardened, for example in the laser beam remelting process.
It must be ensured in this process that the martensitic microstructure is developed uniformly over the camshaft
contour by exact thermal processes and time sequences.
This process is checked by means of hardness testing. This calls for a method that keeps the necessary time
expenditure within certain limits even with statistically required multiple measurements.
For this purpose we recommend the hardness test according to the UCI method (Ultrasonic Contact Impedance).
When using the well known Vickers method, the indentation area of the diamond has to be determined using a
microscope. The hardness H then is expressed according to the relation

H=

F
A

where F is the indenting Force and A the indentation area.

In case of the UCI method, the Vickers diamond is mounted on a vibrating rod, which changes the vibration
frequency during the indentation process. The frequency change is proportional to the indentation area. This allows
a fast and mobile test procedure because the microscopic area measurement is no longer necessary.
Using the test load of 0.3 kgf (3 N) and 1 kgf (10 N)
respectively, an approximative nondestructive surface
hardness determination is achieved (experience shows
that the indentation depths of Vickers diamonds are at
around 4 to 7 m).
The measuring process is considerably simplified by
the special test support MIC 225 for hardness testing
on camshafts. This is achieved by mounting the
camshaft in V-blocks and an air bearing slide carriage
with probe support guiding the measuring device
parallel to the shaft axis. The torsion of the shaft
during the measuring process is prevented by
magnetic locking, which ensures an exact positioning
of the probe on the camshaft base circle or camshaft
tip.
The recommended equipment for this application:
Hardness tester:
MIC 10, MIC 10DL
Probes:
MIC 2003-A, MIC 201-A
Accessory:
MIC 225

Figure 21: Camshaft Hardness Testing setup

Besides the UCI-technique further methods for the mobile hardness testing are available: The rebound method and
the TIV-method (Through Indenter Vision). The selection of the right method depends on the type of inspection
task.
Radioscopic inspection of Aluminium Castings
For the inspection of aluminium or magnesium castings, e.g. wheels, steering housings etc., the X-Ray technique is
commonly applied, where the radioscopic method reveals a real time procedure.
The basic setup is shown in Figure 22. The specimen is positioned on a manipulator system between the X-Ray
tube and the imaging device an Image Intensifier or a flat panel detector. The relation between the distances FDD
(Focus Detector Distance) and FOD (Focus Object Distance) define the geometric enlargement of the displayed
object details.

X-ray image
monitor image

FDD
digital signal
FOD
data acquisition and
image evaluation
( SABA 2000 )

specimen
flat panel detector
X-ray tube

image processing
monitor

Figure 22: Radioscopic Setup

The digital image signal is displayed directly on a monitor or can be processed, in order to automatically detect
(computer aided) flaws.
Figure 23 presents a view inside a
radiation protection cabinet, where a
steering housing is mounted on the
manipulator. In the upper left corner a
radiographic image is included.
The available contrast of the radioscopy permits
detection of flaws with minimum extensions of
approximately 1% of the actually penetrated
wall thickness. The lateral resolution depends
of the focal spot size, the geometrical
magnification factor and the resolution of the
imaging device.

Figure 23: View inside a radiation cabinet and a radioscopic

image of a steering housing

3D-Computer Tomography
X-ray transmission techniques such as radiography and radioscopy, deliver shadow images without any
information about the depth of a flaw or any geometrical indication.

attenuation image
FDD

reconstruction
network

digital signal

FOD

test
sample

flat panel
detector

visualisation
Figure 24: Setup for the Computer Tomography
The complete information however can be obtained using tomographic methods. Here, the test sample is rotated in
the X-ray beam, see Figure 24, and the intersections of the object or even three dimensional images are obtained by
means of a computer reconstruction.
Highly sophisticated computer tomography allows detection and localization of flaws such as gas pores or
shrinkage cavities and the determination of geometrical properties such as internal wall thicknesses and core
mismatches. For example the 3D-CT is used for the inspection of motor blocks or hydraulic labyrinth systems or
pump housings, as shown in the Figures 25 and 26.

Figure 25: 3D-Tomogram of a

hydraulic labyrinth block

Figure 26: 3D-Tomogram of a

pump housing