development Materials Testing

Innovative Testing Concept

of Cylinder Heads
The increasing demands for improving the specific power output of modern power-trains and for reducing the component
weight via lightweight design, require detailed expertise about the damage mechanisms occurring under real service load-
ing conditions. Taking into account requirements for accelerated, cost-effective component testing and damage analysis,
IABG developed a new test stand for thermo mechanical fatigue testing of cylinder heads. Thereby, both the test and devel-
opment process of cylinder heads may be accelerated significantly. The implementation of an in-situ crack detection sys-
tem enables the investigation of crack initiation and propagation as well as a 24 / 7 service of the test stand.

34 mtZ 12I2009 volume 70

 1 The Thermo-mechanical The Authors
Fatigue Process of Cylinder Heads

Numerous failures of cylinder heads Dipl.-Ing. (FH)

originate from cracks in the combustion Florian Achatz
chamber’s pent roof – induced by ther- is project manager
mo mechanical fatigue. Critical areas are at cylinder head test
often valve reinforcing ribs, or spark stand, tests and
plug/injector nozzle caps. Analyses division,
Under full-loading conditions, the pent at IABG mbH in otto­
roof is subject to high thermal input. In brunn (Germany).
the formentioned critical areas this caus-
es thermal expansion – which is simulta-
neously restricted by surrounding zones Dr. Jürgen Fröschl
that are cooler and often more inflexible is expert for service
geometrically. Depending on the load strength, test and Analy­
conditions, the resulting compression ses division, at IABG mbH
stresses in these key areas may lead to lo- in ottobrunn (Germany).
cal plastic deformation. Once the thermal
input is then reduced, as a result of load
reduction or even engine braking, where Dr. Steffen Rödling
the injected fuel volume is removed and is head of department
“cold air” ingested, strong tensile stresses service strength, func­
can occur in these zones. This can lead tional testing and test
once again to plastic deformation. As a rigs, tests and Analyses
result of the fact that compression stress- division, at IABG mbH in
es – i.e. negative stresses – occur at high ottobrunn (Germany).
temperatures and tensile stresses under
low-temperature conditions. This kind of
thermo mechanical loading is often re- Dipl.-Ing. (FH)
ferred to as “out-of-phase loading”. Jochen Gärtner
High tensile stresses occurring in cool- is senior engineer for de­
down phases are often the main reasons velopment and construc­
of crack initiation and propagation in tion of customer specific
the above-mentioned areas, as a result of test rigs, test and Analy­
thermo mechanical fatigue. Some of ses division, at IABG mbH
these cracks may have the potential to in ottobrunn (Germany).
extend as far as the coolant channels

Figure 1: TMF load cycle

mtZ 12I2009 volume 70 35

development Materials Testing

cylinder head to be tested is placed in a

hinged assembly and connected to the
test rig with the aid of a computer-opti-
mized adapter – the crankcase dummy.
The crankcase dummy simulates the me-
chanical and thermal interface of the
cylinder head with the crankcase. Each
pent roof is then heated cyclically with a
gas torch up to maximum operating tem-
perature, and cooled via a cooling circuit
that accurately represents a real engine
system. By varying the test rig param­
eters, it is possible to modify the spatial
temperature distribution and gradients
specified by the customer during an it-
Figure 2: Cylinder head test erative calibration process that occurs
stand concept prior to the actual test runs. The test rig
concept is illustrated in Figure 2.

2.1 Test Rig Concept

causing engine failure. Due to this, nu- gine to undergo an engine test run. IABG A load cycle on the cylinder head test rig
merous engine manufacturers put a fo- has therefore implemented a cylinder consists of two successive phases: the
cus of their research work on the investi- head test stand with an in-situ crack de- heating phase and the cooling phase. An
gation of the thermo mechanical fatigue tection. This new developed test rig is a entire loading cycle requires about a
processes and the development of new tool that can be introduced into the de- minute to be completed. Figure 3 illus-
strength hypotheses. Furthermore, a spe- velopment process both at a very early trates a representative temperature pro-
cific research focus is put on the investi- stage for design variant studies as well as file during a load cycle.
gation of time-dependent fatigue behav- for statistical strength assessments of In the heating phase, the cylinder
ior of modern light metal alloys leading cylinder head designs at a very late stage. head’s pent roofs are each heated using
to a significant change of the local com- Due to this, the number of cost intensive an oxy-propane compressed air torch.
ponent strength during service [1]. These endurance tests on fired engine test The maximum output of these gas tor–
so-called aging effects may be caused for benches can be reduced significantly. ches is 75 kW each. Up to six of these ad-
example by diffusion processes. Further- justable-output burner units are avail­
more, the high stresses occurring in the able on the test rig.
high-temperature phases can also initi- 2 Thermal Cylinder Head Test Rig The cylinder head is subjected to a
ate creep as well as deposit build-up, sig- continuous stream of coolant during the
nificantly effecting the material behav- In order to meet the above-mentioned re- entire cycle; two coolant circuits are pro-
ior. The primary stress conditions present quirements for significantly accelerated vided for this purpose. During the test
and the simultaneous, locally-occurring component testing, coupled with a real- cycle’s heating phase, the heated water
mechanisms in the material itself are istic depiction of damage formation, the circuit is utilized to support the heating
presented in Figure 1. In order to obtain
reliable damage calculation of cylinder
heads, it is therefore essential to map
these material-specific and largely super-
imposed processes to special material
models. The modeling approaches neces-
sary serve not only to describe the peri-
odic and fluctuating cyclic deformation
behavior, but also the damage behavior
caused by the local stresses, [2].
Improved computational techniques
and a wide variety of design variants re-
quire new time- and cost-efficient testing
methods – especially for cylinder heads
experiencing high thermal stresses. Such
methods must enable the depiction of
the real thermo mechanical damage
process, without requiring the whole en- Figure 3: Temperature protocols at reference points

36 MTZ 12I2009 Volume 70

behavior of the cylinder head. Optionally,
a coolant circuit with vehicle-matched
coolant quantities can also be used in
this phase. This circuit permits volume
flow rates and coolant temperatures as
found in real engine service conditions.
Within 30 s, coolant temperatures of up
to 120 °C can be achieved. By regulating
the volume flow and throttling back the
return line, the pressure in the cylinder
head coolant inflow can be increased to a
maximum of 4 bar during the heating
phase. By generating flow rates and tem-
peratures similar to those found in a fired
engine, it becomes possible to simulate
flow patterns in the critical area between
cylinder head and coolant. At the same
time, it is also possible to induce lo­calized
boiling – or to eliminate it entirely.
In the cooling phase, the gas burners
are switched off and the coolant is
switched to the cold water circuit. The
maximum rate of flow for this circuit is
200 l/min, with the average coolant tem-
perature held constant at 26 °C. Convec-
tion cooling is also applied using an air
stream, which flows through the burner
nozzles during the cooling phase.
The test rig provides connectors for
up to 80 thermocouples for recording
measurement data. Measurement equip-
ment is also available for other indica-
tors, such as pressure and flow rates.
To ensure that pent roof inspections
can be carried out quickly and simply,
the test rig is equipped with a pneumatic
boom, which enables the cylinder head
to be rotated 90° out of the test state and
into an inspection position. In this posi-
tion, incipient cracks can be detected
and the crack propagation process can
be identified with increasing numbers of
cycles. In the inspection position, a cam-
era is introduced and the individual pent
roofs are photographed.
2.2 Crankcase Dummy
Testing a cylinder head under realistic
conditions requires the integration of the
thermal and mechanical interface to the
crankcase. This is achieved by means of a
specially-designed crankcase dummy.
Structurally, the dummy is first adapt-
ed to suit the contours of the interface
from cylinder head to crankcase. Then,
the external geometry and coolant in-
flow/outflow points are adjusted to suit
the test rig conditions. The cylinder head
screw fixings are adjusted to suit the con-
ditions in the crankcase, and sealed us-
ing the original cylinder head gasket.
Computational Fluid Dynamic (CFD)
computation is used to design the fluid
flow capabilities of the crankcase dummy.
In this process, models with varying con-
structional details are computed together
with the cylinder head and iteratively op-
timized until the conditions correspond
to those found in the fired engine, Figure 4.
This process is essential for achieving an
accurate temperature distribution for
heat transfer from the cylinder head to
coolant, comparable to that of the real en-
gine. By establishing the correct velocity
(and thus pressure) distribution, coolant
boiling can be avoided in the cylinder
head – and thus a distortion of the heat
input when compared to the real engine.
3 Test Program
The standard test program on the test rig
itself consists of start-up, the actual en-
durance test and the evaluation followed
subsequently.
3.1 Start-up
Before starting the actual test run, the
transient temperature distribution in
the pent roof on the cylinder head test
rig is optimized in order to meet the cus-
tomers demands. For this, a cylinder
head equipped with thermocouples is
fixed into the test rig. An iterative pro­
cess is used to continually adjust the test
rig parameters until they match the fired
engine’s temperature distribution. Test
rig parameters include: gas burner out-
put, flame profile and the geometrical
orientation of burners to cylinder head,
plus control times and coolant streams
in both phases of the cycle.
3.2 Endurance Test
Following start-up, the endurance test is
started with the configured parameters.
For this stage, a test cylinder head is
mounted, together with cylinder head
gasket and screw fittings.
Test rig indicators such as gas and
coolant flow rates are used to monitor
the stability of the endurance test. In ad-
dition, a reference control point is used
to measure the temperature stability of
each test subject.
Figure 4: Comparison of CFD-simulation:
Entire engine vs. cylinder head test stand
In endurance test mode, the cylinder
head test rig can be operated unattended
for a 24-hour service with automated
crack reporting. The test sample is sub-
jected to visual crack monitoring at regu­
lar intervals, to ensure that incipient
cracks are detected as soon as the crack
starts to form. Crack propagation is
tracked photographically once cracks
are initiated. For this purpose, the test
rig is equipped with a specialized cam-
era, taking automatic snapshots at pre-
set cycle intervals.
3.3 Evaluation
The endurance test ends when the re-
quired load cycle count is reached or
when the specimen fails. The crack sec-
tions detected are further verified using
a dye penetration process, a test report is
created and the test sample is handed to
the client for further analysis. IABG can
also offer damage assessment, associated
computation and ancillary material
characterization services, conducted in a
fatigue and materials laboratory that is
accredited according DIN 17025.
4 Comparison of Cylinder Head
Test Rig and Engine Test Bench
Until now, the standard technical proce-
dure has been to test cylinder heads using
an engine endurance test and a complete
engine setup. This kind of testing is ex-
pensive and time-consuming, since a com-
plete engine needs to be set up and oper-
ated. A single temperature cycle requires
start-up, warm-up and cool-down phases
for the entire engine and thus takes sev-
eral minutes. These laborious setup re-
MTZ 12I2009 Volume 70 37
development
Figure 5: Typical failure mode of a specimen tested in the cylinder head test stand
quirements mean that damage assess-
ment of the specimen can occur only at
the end of the test run, which prevents
the acquisition of data on damage forma-
tion and the damage process.
The cylinder head test rig described
here thus constitutes an innovative and
superior approach to testing cylinder
heads. The depiction of damage-causing
temperature cycles as thermal shocks
within about a minute enables significant-
ly accelerated testing. As a result of the
shorter test duration times compared to
traditional engine test benches, and the
lower costs thus achieved, the cylinder
head test rig makes it economically feasi-
ble to assess material, manufacturing and
38 MTZ 12I2009 Volume 70
Materials Testing
design variants, even where test data has a
sound statistical basis. As many as 1,000
temperature cycles per day can be attained
during an endurance test. Furthermore,
the test process does not require the entire
engine, eliminating the effort required for
its setup and operation. Various cylinder
head models may thus even be tested while
still in the prototype stage. Another advan-
tage when compared to an engine test
bench is the in-situ detection and tracking
of cracks in the pent roof area. This ap-
proach improves comparisons made with
computational models and increases the
relevance of damage estimations.
To verify the new testing model, a
comparison was made between both
test techniques, in relation to damage
­incidence and the temperature cycles
achieved until damage occurred. The
damage patterns attainable using the cyl-
inder head test rig match those occurring
during engine endurance tests. Figure 5 il-
lustrates a typical result about crack loca-
tion and orientation of a cylinder head
tested in the presented test stand.
5 Summary
The thermal cylinder head test rig presents
an innovative method for assessing and op-
timizing cylinder heads in relation to their
service strength vs. thermo mechanical fa-
tigue at a very early stage in their develop-
ment process. The test rig makes it possible
to reduce the testing times and costs sig-
nificantly in comparison with standard
engine test benches, thus shortening the
development time for new cylinder head
designs. The test rig’s novel, open design
also makes it possible to detect crack for-
mation and investigation of crack propaga-
tion in the pent roof across thermo shock
cycles for the very first time. Inspection
procedures do not require the assembly of
the entire engine for fueled operation.
References
[1] Ehart, R.; Nefischer, P.; Riedler, M.: Material­
modellierung und Lebensdauerabschätzung bei
thermomechanisch beanspruchten Motorbauteilen
unter Berücksichtigung von Alterungsvorgängen,
S. 109-122, DVM-Bericht 133, Berlin 2006
[2] Riedler, M.: Methodikfindung zur Simulation von
thermomechanisch beanspruchten Motorbauteilen
aus Aluminiumlegierungen, Fortschritt-Berichte
VDI, Reihe 5, Nr. 718, Düsseldorf 2005