See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/321634027

A failure analysis of the exhaust valve from a heavy duty natural gas engine

Article  in  Engineering Failure Analysis · December 2017

DOI: 10.1016/j.engfailanal.2017.12.001

CITATIONS READS

3 6,838

3 authors:

Muhammad Imran Khan Muhammad Arsalan Khan

39 PUBLICATIONS   752 CITATIONS   
University of Engineering and Technology, Peshawar
4 PUBLICATIONS   5 CITATIONS   
SEE PROFILE
SEE PROFILE

Dr. Abdul Shakoor

University of Engineering and Technology, Peshawar
56 PUBLICATIONS   209 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Composite materials View project

Packaging Products for the Industry View project

All content following this page was uploaded by Muhammad Imran Khan on 21 December 2017.

The user has requested enhancement of the downloaded file.

 Engineering Failure Analysis 85 (2018) 77–88

Contents lists available at ScienceDirect

Engineering Failure Analysis

journal homepage: www.elsevier.com/locate/engfailanal

A failure analysis of the exhaust valve from a heavy duty natural

T
gas engine
⁎
Muhammad Imran Khana, , Muhammad Arsalan Khanb, Abdul Shakoorb
a
Institute of Petroleum Engineering, Heriot Watt University, UK
b
Department of Mechanical Engineering, University of Engineering & Technology, Peshawar, Pakistan

AR TI CLE I NF O AB S T R A CT

Keywords: Increasingly stringent emission standards are changing the conditions that valve systems in heavy
Exhaust valve duty engines are exposed to. Increased pressures and temperatures are challenging system en-
Intergranular corrosion durance. A consequence of these changing conditions is a reduction in the levels of soot pro-
Inconel-751 duction that had formerly generated protective films. In order to help combat this, nickel-based
Failure analysis
super alloys have been widely used in applications requiring strength at high temperature. This
Nickel super alloy
study presents a premature failure case of a set of exhaust valves belonging to a heavy duty
natural gas engine; where the valves were manufactured from one of these alloys, the pre-
cipitation hardened Inconel-751. The failure occurred at approximately 5000 operating hours
after its first commissioning, whilst the standard expected service life of the valves is 20,000 h.
Several examinations employing multiple techniques were carried out in order to identify the
root cause of failure, whilst comparing results against those of a new valve. It was found that
there was some mechanical lapse in proper sitting of the valve, which had been responsible for
unwarranted overheating especially at thinner sections. Microstructure examination revealed
that overheating had been responsible for a creep-rupture failure accentuated by precipitation of
undesirable constituents at grain boundaries.

1. Introduction

The Global demands for inexpensive and clean power is increasing day by day and in this regard, the market of stationary
generators have proven to be one of the major sources of power providers globally. In Asia, generators powered by stationary gas
engines are becoming increasingly popular in areas where natural gas is readily available. Currently Caterpillar Inc., GE Waukesha
and Cummins Power Generation are the key players in Global and Asian gas engine market [1].
The inlet and exhaust valves are one of the vital components of an IC engine. The major function of an Inlet valve is to control the
flow of air to the combustion chamber while the exhaust valves are used to monitor the outward passage of flue gases from the
combustion chamber [2–4]. Their operation has a direct effect on performance parameters (power, torque, fuel consumption and etc.)
and also the engine emissions. Wear and failure of the exhaust valves, is an unavoidable problem in Internal combustion engines
which ultimately leads to under performance, large down time, and high maintenance costs. Numerous numbers of alterations in the
design, material and production techniques have greatly enhanced the running life and performance of the exhaust valves, but these
up gradations cannot keep their pace with the continuous rise in the requirements of enhanced engine performance in our global
competitive environment [4,5].

⁎
Corresponding author.
E-mail address: mk42@hw.ac.uk (M.I. Khan).

https://doi.org/10.1016/j.engfailanal.2017.12.001
Received 8 October 2017; Received in revised form 28 November 2017; Accepted 1 December 2017
Available online 05 December 2017
1350-6307/ © 2017 Elsevier Ltd. All rights reserved.
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

Nomenclature EDS Energy Dispersive Spectroscopy

IC Internal Combustion
CeC Carbon-Carbon OEM Original equipment Manufacturer
CeH Carbon-Hydrogen SEM Scanning Electron Microscope
CH4 Methane TGA Thermo Gravimetric Analysis
DTA Differential Thermal Analysis XRD X-ray Diffraction

There is very limited number of peer reviewed studies available about the failure analysis of exhaust valves especially that of Inconel-
751. A failure analysis of the exhaust valve stem in a Waukesha P9390 gas engine was performed by Kwon OG et al. [6]. They
observed a significant loss of hardness in valve material and concluded overheating as failure cause. The significant hardness loss, the
extensive surface oxidation and fretting/galling on the valve stem were indicative of the overheating. Yu ZW et al. [7] performed a
failure analysis and metallurgical investigations of diesel engine exhaust valves. Their fractographic studies indicated that formation
of the lamellar structure in the material of the valve head was the dominant mechanism for the fatigue failures in the exhaust valve.
Vardar N. et al., [8] investigated the failure of exhaust valve failure in heavy duty diesel Engine. This was carried out by using several
experimental tests methods like optical emission spectroscopy, optical microscopy, scanning electron microscopy SEM and EDX. They
concluded that the valve was failed and broken down prior to its desired service life. Very recently Witek L et al., [9] investigated the
fracture problem of the exhaust valve of a passenger car diesel engine in order to explain the reason of premature fracture of the
valve, author used finite element model consisting of the valve, the seat and the guide. The results shows that irregular depositions of
the carbon on the seat face of the valve caused large amplitude of bending stress in the valve stem which subsequently caused the
premature fatigue failure. Scott CG et al., [10] studied the effect of valve deposit morphology and composition on the erosion-
corrosion of valve seat surfaces. The study provides some initial evidence that although valve seat deposits may have played a role in
valve failure but the erosion-corrosion of exhaust valves was not exclusively related to the thickness of valve seat deposits. Forsberg P
et al. [4] investigated the wear mechanism of three pairs of exhaust valves and valve seat inserts with the same material and design
properties but with different service condition. The study revealed that oxidation and formation of a tribofilm was the dominant wear
mechanism.
Although the alloy Incolnel-751 is being used extensively in modern heavy duty natural engines by various manufactures around
the world but to the best of our knowledge, there is no systematic empirical research exists addressing the thermal performance of
Inconel-751 in IC engines. The aim of this study is to analyze the failure mechanism of exhaust valves made of Inconel-751 failed due

Fig. 1. Showing Cylinder head from a Caterpillar G3406 natural gas engine with guttered exhaust valves and showing high deposits after 10,000 h.

78
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

to guttering in heavy duty natural gas engine.

2. Problem statement

In the present work, failure analysis of exhaust valves form Caterpillar G3406 gas engine coupled with electric generator, were
carried out. This engine was a 4 stroke, 6 cylinder 14.59 l natural gas fueled engine. The Genset was installed at an oil field located in
Pakistan for continues power generation. It was commissioned on August 2013 and had been taken into operation since that date. The
average total power consumption of the oil field was 40% of the generator capacity i.e. 200 Kva. It was reported that the Genset has
trouble of tripping under low frequency just after 5000 h. in service. The problem was investigated by the OEM service team and it
was wrongly concluded that the problem was due to stucking of exhaust bypass valve of turbocharger.
The exhaust was repaired and problem was fixed. It was also found that the pressure of cylinder # 5 and 6 were 180 psi and
160 psi respectively against OEM standard value of 220 psi. As the required power load was very low against the Genset rated
capacity, hence it was decided to run the engine up to 10,000 h. Service life, the OEM recommended schedule for top-end over-
hauling. After completing 10,000 h. of service operation the top-end overhauling of the engine was performed. All parameters
recorded before going for overhaul execution phase. The problems observed with the Genset were: Genset shut down under fre-
quency, white smoke from exhaust and abnormal jacket water temperature. The recorded pressure values for cylinder # 1, 2, 3, & 4
were complied with OEM normal pressure range of 220 psi, while pressure value for cylinder 5 & 6 were 80 psi and 50 psi respec-
tively which is considered very poor. After disintegrated the engine head guttering was found in the exhaust valve of the cylinder # 5
& 6 which was the main cause for poor cylinder pressure of these cylinders as shown in Fig. 1. The initial cause of guttering could not
be determined, however, records of engine temperatures support the existence of guttering as they showed elevated exhaust tem-
peratures at the affected cylinder beginning roughly 5000 h into operation of the generator. After 5000 h of operation, the tem-
perature of jacket water cooling systems was increased from normal average temperature of 185 °F to 205 °F.

3. Factors affecting the life of exhaust valve

3.1. Effects of temperature

Exhaust valves mostly fail due to overheating which manifests itself in the form of decreased hardness, erosion-corrosion and
extensive oxidation of the valve surface. Because of their installation in the path of flue gases, exhaust valves are constantly exposed
to hot corrosive gases and the temperature these valves are subjected to, ranges from 750 °C to 950 °C [3,11,12]. In addition, Exhaust
valves are exposed to a much higher temperature than the intake valves because the incoming fresh charge/air at atmospheric
temperature cools the intake valves [8]. Due to abnormally high temperature in the region of exhaust valve and its seats, the material
prematurely loses its mechanical properties like hardness, elasticity, strength, etc. And in addition the continuous valve closing and
opening leads to the valve and valve seat wear.
This should also be noted that in case of inlet valve, the valve head is kept cool owing to the flow of cool inlet charge over it, and
very little heat is rejected to anywhere other than the valve seat. But in the case of an exhaust valve, a high heat flux is not only
absorbed by the valve head, but also by the area behind the head. For instance about 76% of thermal input to an exhaust valve exits
when it sits in its seat and the remaining 24% leaves by means of the stem [13,14]. Effective heat transfer to the seat insert and into
the cylinder head is, therefore, essential. Heat transfer can be significantly affected by seating deposits. If deposits are allowed to
build up, they may not only lead to an increase in valve temperature, but may also break away locally and create a leakage path,
leading to valve guttering.

3.2. Effects of loading and seating

Studies have shown that the exhaust valve sear wear mainly involves two distinct mechanisms. The first one is the impact of valve
on the seat as it closes and second mechanism is Micro-sliding, caused by the elastic deformation at the valve seat interface as the

Fig. 2. Relation of valve recession with (i) valve closing velocity (ii) combustion load and (iii) valve misalignment Source: Prepared by the authors based on [16].

79
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

valve head is pressed into its seat by the high combustion pressure [15]. Experimental work performed also show that combustion
load, valve misalignment and most importantly, the valve closing velocity have a significant effect on valve face recession [16],
which is shown in the Fig. 2.

3.3. Effects of combustion deposits

Depositions building up on valve face and valve seats leads to poor sealing and seat can have an insulating effect which causes
poor cooling and makes the valve run hot. If deposition is built up on valve face in the form of one spot, then poor sealing appears.
This poor sealing leads to leaking of the combustion and creates hot spot on valve sealing face which in turn results in channeling
effect. The deposition on valve seat, valve face, valve stem, is from combined reaction of impure contaminated fuel and lubricating
oil. They are composed almost entirely of inorganic compounds. The oxides e.g. sulfur dioxide act with calcium contents in lu-
bricating oil to form low melting point salts. These salt particles get deposited on valve surfaces in their molten state. These salt
particles get cooled sufficiently to adhere to valve stem and valve spindle and do not get carried away by the exhaust gases. As
operating temperature is more than 700 °C the deposited salts change into molten state. This molten state salts flow along the grain
boundaries depending on valve material and dissolves the protective oxide on the grain boundaries. This phenomena is called
intergranular corrosion, which leads to burning of the valve sealing face.
In a modern engine with low oil consumption running on a typical full boiling range fuel, combustion chamber deposits are
formed primarily by the fuel though the lubricant also contributes to their formation. Deposit growth is a dynamic and reversible
process, which at a given time, reflects the balance between the formation and removal processes. The single most important
parameter that controls deposit formation is the surface temperature, which itself changes as deposits grow because of their in-
sulating properties. Changes in engine operating parameters that tend to increase surface temperatures reduce deposit formation
[12]. Thus increasing the speed, load and coolant temperature or making the fuel/air mixture slightly richer than stoichiometric all
reduce deposit growth. Boiling point is the most important fuel or lubricant property in combustion chamber deposit formation - the
higher the boiling point, the greater the deposit forming tendency.
Studies into the characteristics of combustion or ash deposits on the sealing faces of valves and seats have shown that these play
an important role in the guttering process [17]. Such scales are ubiquitous, consisting of, for example, sulphates, phosphates and
oxides of inorganic fuel and oil constituents. Not only do these scales interfere with the tight geometric fit of valves and seats, but
they also restrict heat flow out of the valves. Around 75–80% of the heat absorbed by exhaust valves exits through the contact
between valve and valve seat [15]. Valves can therefore be insulated by excessive scale, raising their temperature and making them
more susceptible to distortion or damage. Finally, uneven flaking or spalling of the scale itself under the closing action of valves can
create sufficient localized leakage paths to initiate valve guttering.

3.4. Effects of the fuel and lubricant

Natural gas which is primary fuel used in gas engine mainly consist of methane (CH4). The four CeH bonds of methane give it a
higher specific heat content than liquid fuels like gasoline or diesel that contain some lower energy CeC bonds. Consequently, it
burns hotter than other fuels and can cause severe oxidation and nitration of the engine oil [21]. In addition, since it is already a gas,
methane does not cool the intake air by evaporation as liquid fuel droplets do. This has a significant effect on the intake and exhaust
valves because there is no fuel-derived lubricant for the valves like liquid droplets or soot. Consequently, gas engines are solely
dependent on the lubricant ash to provide a lubricant between the hot valve face and its mating seat. The ash is the portion of the
lubricant that is left behind as a deposit after complete burning of the oil. It is whitish-gray and comes from the metallic detergents
(calcium and barium) and antiwear (zinc) additives. In addition, ash concentration and composition, thermal expansion, melting and
boiling points of the ash deposits, which are majorly influenced by the additives and detergents employed in the engine oil, play a
vital role in exhaust valve failure mechanism [8]. Too little ash or the wrong type can accelerate valve and seat wear, while too much
ash may lead to interference in the geometrical fit of the valve and its seat, leading to distortion, leakage, guttering and hot spots
subsequently resulting in valve torching [10,22]. A Well formulated gas engine oil can not only provide excellent wear resistance,
cooling and deposit control, but misapplication of such an oil not designed for gas engines can degrade life expectancy, and per-
formance of the engine and also negatively impact operating costs [15,16,21]. Too much ash deposit may be caused by using an oil
with too high of an ash content, over lubrication or many other mechanical factors. It is noteworthy to mention that exhaust valve
failure can result from both lubricant (too much ash) and non-lubricant factors, as well as from valve recession itself. This is a
complicated process as these factors are intertwined with each other, which makes the process of finding the root cause of a valve
failure difficult. The ash content of engine oil used in this case study was 0.14% which comes under category of low ash oil.

3.5. Valve materials

As the inlet and exhaust valves are exposed to varying operating conditions in heavy duty engines, so the materials used in their
manufacturing are also generally different. Specifically, the material used for exhaust valve must have high resistance to corrosion at
the elevated operating temperatures, sufficient strength and hardness to resist tensile forces and wear, adequate fatigue and creep
strength. In addition, coefficient of thermal expansion should be low to avoid excessive thermal stresses and most importantly they
should have a high thermal conductivity for good heat dissipation [11,18,19]. Exhaust valves of heavy duty engines are generally
made of austenitic steel. Typical austenitic alloys for exhaust valve are 21-2 N, 21-4 N and 23-8 N, Inconel-751, Pyromet 31 and

80
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

Nimonic 80A. In gas engines the exhaust valves have high levels of cobalt, nickel, iron, chromium, titanium and other elements to
increase hardness [20].

3.6. Failure's mode of exhaust valves

The exhaust valves are exposed to thermal and mechanical overstress which can be sources of valve failures. During each
combustion event, high stresses are imposed on the combustion chamber side of the valve head. These generate cyclic stresses
peaking above 200 MN/m2 on the port side of the valve head. The most common type of failure that can be caused by these thermal
stresses is the valve and seat wear. The understanding of wear mechanisms is complicated by inconsistent patterns of valve failure.
For example, failure may occur in only a single valve operating in a multi-valve cylinder. Furthermore, the apparent mode of failure
may vary from one valve to another in the same cylinder or between cylinders in the same engine. It has been noted that in practice,
every engine manufacturer has had share of their exhaust valve failure problems but no hard and fast rules can be used to predict or
achieve a satisfactory exhaust valve life. The literature study shows that valve wear generally occurs by the three main modes namely
the abrasive wear, adhesive wear, and corrosive wear [4,20,23–25].
Abrasive wear can be observed in exhaust valve and its seat, by the presence of gouging and scaring on the mating surfaces.
Relative sliding, wear particles of the mating surfaces and solid combustion products from the fuel, lubricating oil and air con-
tainments are influencing factors in this mode of wear [4,15]. Micro welding or bonding of the mating surfaces has been char-
acterized as Adhesive wear mechanism. At high temperatures, the plastically deformed asperities or projections present on the mating
surfaces are eventually welded together when the valve is pressed into its seat [25]. Incompatible valve and seat materials, critical
temperature, minimal solid film lubrication, and High contact pressure result in adhesive wear [15].
Presence of High operating temperatures, high ash level and Corrosive constituents from combustion products, might lead to
Corrosive wear in Exhaust valves [10,26]. On micro level, presence of high ash level, might lead to insulation of the exhaust valve,
which will not only increase the temperature of the valve, but will also make it prone to the chemical attack by the fuel and deposits
compounds, present in the form of salts and oxides of Sulfur, Magnesium, Calcium, Phosphate, Zinc, Silicon [18,27]. Corrosion
manifests itself in the form of Intergranular attack on the grain boundaries leading to guttering of the exhaust valves [28,29]. The
above mentioned wear mechanisms lead to three main types of valve failure that have been known to occur on the valve face, seating
and sealing face of the valves, which are: Valve recession, Guttering and Torching [4,23,25,26,30].

3.6.1. Valve recession

Valve recession is said to have occurred if wear of the valve and seat insert contact faces has caused the valve to ‘sink’ or recede
into the seat insert, thereby altering the closed position of the valve relative to the cylinder head (as shown in Fig. 3). Valve recession
is the most common form of valve wear in gas-fired engines; valve recession occurs gradually over thousands or tens of thousands of
hours. The recession occurs by metal abrasion, high temperature corrosion, frictional sliding, and adhesion mechanisms [31]. Ex-
cessive amount of valve recession leads to incorrect or incomplete seating which leads to losses in the cylinder pressure [32] Valve
and Seat recession mostly occur due to impact of valves on the seat, and by the systematic gouging, dulling, deformation and eventual
wear out, of the exhaust valve or seat insert material [30,33,34].

Fig. 3. Showing valve recession phenomenon [32].

81
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

3.6.2. Guttering
In case of the valve not seating properly, which might be due to valve deformation or due to the interference of the solid deposits,
the pressure inside cylinder will be lost in the form of a leakage path for the exhaust gases [32]. This leakage path will lead to
material losses in the valve and its seat by means of an erosive-corrosive mechanism [10]. Over repeated engine cycles, this leakage
path will widen and will eventually lead to valve guttering [35] which can be easily identified in leaked engine cylinders by a distinct
hissing sound [32]. Literature Study about the ash deposits or scales on the valve sealing face have clearly confirmed their influence
in the valve guttering process [15,20]. Guttering of exhaust valves will eventually result in an alarming leakage of the cylinder
compression, and will also manifest in the form of power loss due to misfiring [36,37,10].

3.6.3. Torching
Valve torching is the phenomena which may happen where temperature of the valve-face or valve-seat surface is elevated to the
melting point of valve material by pre-ignition. This is caused by the hot surface of valve or deposit particles which act as a “glow
plug” and ignite the fumigated fuel-air mixture during the compression cycle, earlier than the spark plug. The burning gases are
driven through valve and seat interface leakage passages by the high differential pressure of abnormal combustion, heating the
surface to the melting point. Like a cutting torch, the molten metal is carried away from the valve surface into the exhaust passage by
the jet of burning fuel gas and air. Torching can occur in natural gas engines that operate with excessively advanced spark plug
ignition timing for a given fuel, which produces abnormal combustion. With abnormal combustion there is a high pressure rise before
or near piston top-dead-center. The high pressure rise causes compression heating of the burning gases and a more rapid heat release
rate, raising the valve temperature. Deformation of the valve head and cylinder head is increased. The higher heat release rate
increases radiation and convection heat transfer to the combustion chamber and valve surfaces, causing a large surface temperature
rise.

4. Experimental procedures

The failed exhaust valves under investigation were subjected to the following procedures to assess failure causes:

(i) Visual examination of inner and outer surfaces

(ii) Energy dispersive spectroscopy (EDS) analysis for elemental composition
(iii) Hardness measurement
(iv) Scanning electron microscope (SEM) examination of the valve head
(v) Optical microscopy of the valve head
(vi) X-ray diffraction (XRD) analysis of valve-face deposits
(vii) Thermo gravimetric analysis (TGA) for thermal analysis of valve material

First, the inner and outer surfaces of the failed exhaust vales were visually examined, the amount of the obvious guttering was
listed and the typical macro-morphology was taken using the digital camera (NIKON L100). Then the sample pieces of failed valves
were cut from the corresponding parts based on the requirements of different tests. Some of the samples were cross-sectioned and
mounted in epoxy. They were then grounded by different grit size silicon carbide (Sic) emery papers in a sequence of 320, 600, 800,
1000, and 1200 and subsequently mechanically polished with diamond paste (6, 3, and 1 μm) and 0.05 μm alumina suspension.
Pressurized water coolant was applied during grinding process to prevent sample heating. Ultrasonic cleaning for 2 min was carried
out after each step of diamond paste polishing to remove any diamond abrasive on the sample surface. The samples were then etched
for 60 s using etching solution of Modified Glyceregia (30 ml HCl, 25 ml Glycerol, 10 ml HNO3). After etching microstructure analysis
of the polished samples were performed using SEM (JSM-5910, JEOL Japan). After thermal aging, microstructure of the polished

Fig. 4. Failed valves.

82
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

samples was performed using optical microscope (Olympus GX51, *Olympus Co. Japan) and Scanning Electron Microscope - SEM
(JSM-5910, JEOL Japan). The chemical composition of the exhaust valves was analysed using EDS (Oxford Ins Inca-200) coupled
with SEM on the basis of ASTM A751-2008. Hardness of the valve materials was found with micro-hardness tester (Model HMV-2T,
Made by Shimadzu). X-ray diffraction (XRD) measurements were carried out in order to determine the constitution of the phases
present combustion product of valve surface deposits.

5. Results and discussion

5.1. Visual inspection

As shown in Fig. 4 the two out total twelve exhaust valves have suffered from severe erosion–corrosion with the axial distributing
wide channel. It can be observed from the morphology of the guttering channel on the surface of failed valves have macro-features
consistent with erosion induced by turbulence of high velocity of hot gas stream.

5.2. Chemical analysis

Chemical Composition of the failed valves and new un-used valve (from warehouse) specimen was studied using EDS and the
result is shown in Table 1. The composition is in closed matching with Nickel Alloy Inconel-751which is a precipitation hardened Ni-
Cr-Fe alloy, having traces of Al and Ti and is widely used in the diesel engine exhaust valves.

5.3. Hardness test

Polished samples were taken from the new and failed exhaust valves, in accordance with ASTM E92 standards. A Vickers hardness
test with 30 g force was performed for 15 s and it was repeated for 4 times for each sample according to ASTM-E92 standard. The
average hardness of the unused new valve was determined to be 348 HV which comply with the standard hardness value of Inconel-
751 super alloy. The hardness value of the failed valve was 255 HV, which is at lower side of unused valve. Normal heat treatment of
the Inconel-751 would produce a hardness of about 350 HV. The measured hardness of about 255 HV indicates that the valve had
been heavily overaged. This hardness is effectively the same as in the solution annealed condition of the said alloy which is normally
the softest condition available. Although the specific aging characteristics for the alloy and the normal valve operating temperatures
are not known it strongly suggests that the valve has been running hot.
The measured hardness value of about 240 HV for the Failed valve shows an alarming percentage decrease of 27%, this clearly
points to the fact that the hardness of the failed valve had alarmingly fallen, which may be the effects of intergranular corrosion IGC.
The possible reasons for decrease in the hardness of failed valve is the increased in the amount of γ’ precipitates during the service
(see Section 5.4 for detail).

5.4. Microstructure analysis

SEM Analysis of the polished samples revealed the microstructure of failed and new-unused exhaust valves. It was found that the
failure had taken place primarily due to microstructural reasons. The γ matrix of the new un-used valve (Fig. 5) contained precipitates
γ’ presumably of nickel based intermetallics like Ni3 (Al, Ti, and Fe) at the grain boundaries, these precipitates have assumed a
cylindrical morphology and their presence along the grain boundaries in a heterogeneous manner, greatly diminished the grain
boundary separation strength and presence of such precipitates at the triple points might have led to initiation of stress rupture.
Essentially this is a case of combined failure characterized by creep fatigue interaction.
It seems that the solutionizing heat treatment was not properly performed. Most likely the solutionizing temperature and time for
homogenization were not adequate. The quenching practice might have had slower cooling rate too.
During service, the precipitates have increased in amount; this means that areas adjacent to grain boundaries have been depleted
in strengthening precipitates and initiated void formation at grain boundary triple points. Classically we may say that the failure has
been creep rupture accentuated by the precipitation of undesirable constituents at the grain boundary with unusual morphology.
The microstructure of failed valve (Fig. 6) contains some unusual phases at the grain boundary; the material is subjected to harsh
condition of high temperature fatigue where flow stress of the material becomes lower. Therefore continued use has opened voids at
the triple points; because the environment is corrosive, it is quite possible that at high temperature hydrogen aided failure is possible.
It is for certain that corrosion aided thermal fatigue have played big role in accelerating creep damage. In Nickel super alloys, where
chromium is added to the matrix for its role in optimum corrosion resistance, continuous over heating led to over-aging of the

Table 1
Chemical composition of failed valve material (mass %).

Super alloy Elements

Inconel-751 Ni Cr Fe Ti Al
72.30 16.52 7.85 2.25 1.10

83
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

Fig. 5. SEM image of the new exhaust valve specimen.

Fig. 6. SEM images of the failed valve specimens.

material, which ultimately resulted in formation of chromium-depleted zones adjacent to the grain boundaries which is a well-known
failure mechanism in super alloys and steels called “sensitization”.
This precipitation depletes the Cr content near the grain boundaries, thereby decreasing its corrosion resistance, as shown in the
EDS analysis (Fig. 7) of the failed valve, where chromium content in the near boundary region has depleted.
The mechanism involved is known as the Intergranular corrosion (IGC), and is such a form of corrosion whence the boundaries of
crystal lattice are more prone to corrosion than their insides. Micro hardness tests in accordance with ASTM E92 standards indicated a
percent decrease of 27% confirming our speculation.
In summary it seems there was some mechanical lapse in proper sitting of the valve which has been responsible for unwarranted
overheating especially at thinner sections. This has led to particle coarsening which is found in the microstructure of failed valve. The
grain boundary corrosion is also there and at the same time intergranular precipitates of γ’ have become coarse and non-uniformly
distributed especially near the grain boundaries. The significant loss in hardness of failed valve is also due by this particle coarsening
phenomena. One of the possible reasons for the particle coarsening is rise in temperature which has softened the matrix which made
it vulnerable to failure at the working pressure, especially in a situation where grain boundary has weakened and section thickness
has locally diminished due to erosion-corrosion. These have led to premature failure.

84
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

Fig. 7. EDS results for a) EDS at the Grain Boundary b) EDS at a random location in Grain.

One of the possible reasons of the overheating is the lack of tappet clearance, resulting in light seating and carbon build up on the
seating face. Both factors lead to destruction of the thermal heat path outlet from the valve face through the valve seat to the coolant.
This causes a considerable rise in valve head temperature, particularly in the valve face area. Eventually, the conditions exceed the

Fig. 8. XRD spectrums for the exhaust valve deposits.

85
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

material's resistance to hot corrosion or burning. As the localized gas leak increases, so does the torching effect through the gap,
eventually producing the characteristic gutter.

5.5. Combustion deposits analysis

The compounds formed by combustion deposits on the valve surface were obtained by XRD spectrum as shown in Fig. 8.
Using XRD software, the deposits were identified as Anhydrite, Hydroxyl apatite, Zincate and Calcite (Table 2). Ca5(PO4)3OH is
calcium phosphate hydroxide (hydroxyl apatite), a form of enamel that is difficult to remove. It is possible that this deposit protects or
seals the valve from attack of high temperature exhaust gases and prevents high temperature corrosion.
No traces of free carbon were detected in the solid deposits, this fact was also verified by TG/DTA of the deposits, as it would have
reacted with anhydrite at 800–900 °C to form calcium Sulphates. When analyzing the behavior of solid combustion deposits, the most
direct method is to refer to Thermogravitmetric TG and differential thermal analysis DTA. In the Fig. 9, weight loss can be clearly seen
(Blue Line), with the corresponding heat flow (Red Line), demonstrating the effectiveness of TGD/TA in studying the characteristics
of our sample deposits.
For the TG/DTA analysis of Exhaust valve deposits, the weight was kept as 9.510 mg, which was held for 1.0 min at 40.00 °C after
which it was heated from 40 °C to 1000 °C at 10.00 °C/min, while Nitrogen gas was supplied at 20.0 ml/min during the whole
process. The positive and negative heat flows in the DT curve represent exothermic and endothermic processes, respectively. From
ambient temperature to 200 °C, the evaporation of moisture present in the form of free water or absorbed water occurred which lead
to the initial weight decrement and the first endothermic peak, and this subsequently means that the fuel natural gas was not well dry
and there was free water in the gas, This behavior is similar to the peak as a result of H2O, in the TGA of injector deposits [38].
In the TG curve there is significant weight loss between 200 and 400 °C, which is due to loss of lattice water in Hydroxyl apatite
Ca5(PO4)3(OH). This results in an exothermic reaction which is represented by the 2nd peak at 280 °C in DT curve. Beyond the
temperature of 490 °C till 1000 °C, the mass of the deposits remained negligibly unchanged at 8.5 mg with a total of about 10% loss in
its weight.
If free carbon was present in the exhaust valve deposits, then further decrease in the sample weight would have occurred. These
solid deposits cling to the surface of the valve face, and act as a thermally insulating film, and as we know that about 76% of the heat
absorbed by the exhaust valve during combustion and exhaust gases flow, is given off when the exhaust valve sits in its seat, this will
restrict the cooling process [15] therefore Exhaust Valve will be insulated by the excessive scale, and will become susceptible to
intergranular attack IGA, ultimately leading to guttering of the said valve.

6. Conclusions

The exhaust valves were analysed to be typical of Nickel based supperlloy Inconel-751. This alloy is commonly used for high
performance valves. The valves probably failed as a result of overheating. The possible cause for overheating is lack of tappet
clearance, which results in light seating and carbon build up on the seating face. Both factors lead to destruction of the thermal heat
path outlet from the valve face through the valve seat to the coolant. This causes a considerable rise in valve head temperature,
particularly in the valve face area. Eventually, the conditions exceed the material's resistance to hot corrosion or burning. As the
localized gas leak increases, so does the torching effect through the gap, eventually producing the characteristic gutter.
The significant overaging of the alloy caused the particle coarsening which subsequently resulted in the decrease of hardness.
There are some traces of low melting point compounds in the valve surface deposits which can cause hot salt corrosion that attacks
the grain boundaries. The grain boundary corrosion is there and at the same time intraganular precipitates of γ’ have become coarse
and non-uniformly distributed especially near the grain boundaries. The grain size in the failed valve indicates that performed heat
treatment had not achieve the aim of optimal creep resistance. Probably the microstructure is set to achieve optimal fatigue re-
sistance. The microstructure of the failed valve indicates without doubt on the creep type of fracture (coalescence of cavities along
grain boundaries).

7. Recommendations

Based on the results of this investigation, the major causes of valve failure are distortion of the valve seat, deposits on the valve
and small tappet clearance. Therefore the following items may be considered for future operational and design concerns regarding
prevention of exhaust valve's failure of a natural gas engine. These recommendations are not identified as a comprehensive list, but

Table 2
Phases present in the combustion deposits.

Name of compound Chemical formula Crystal shape Weight %

Anhydrite CaSO4 Orthorhombic 82

Hydroxyl apatite Ca5 (PO4)3(OH) Hexagonal 6
Zincate ZnO Hexagonal 7
Calcite CaCO3 Hexagonal 5

86
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

Fig. 9. TG/DTA for the exhaust valve deposits. (For interpretation of the references to color in this figure, the reader is referred to the web version of this article.)

may provide partial guidance for safe operation of natural gas engines:

▪ Check the clearances at regular intervals as specified in the OEM service manual or when there is abnormal increase in the
temperature of jacket water.
▪ To prevent the intergranular corrosion in exhaust valve system made of a Nickel based super alloy, engine oil with Mg-based
additives is recommended.

References

[1] H.O. Hardenberg, The Middle Ages of the Internal-combustion Engine: 1794–1886, SAE, 1999.
[2] Dongare, S. and P. Narwade, Optimization of IC Engine Valve for Stationary Engine on Fillet Basis.
[3] Y. Wang, et al., The effect of operating conditions on heavy duty engine valve seat wear, Wear 201 (1–2) (1996) 15–25.
[4] P. Forsberg, P. Hollman, S. Jacobson, Wear mechanism study of exhaust valve system in modern heavy duty combustion engines, Wear 271 (9) (2011)
2477–2484.
[5] P. Forsberg, Combustion Valve Wear: A Tribological Study of Combustion Valve Sealing Interfaces, PhD Thesis Uppsala University, 2013.
[6] O.G. Kwon, M.S. Han, Failure analysis of the exhaust valve stem from a Waukesha P9390 GSI gas engine, Eng. Fail. Anal. 11 (3) (2004) 439–447.
[7] Z. Yu, X. Xu, Failure analysis and metallurgical investigation of diesel engine exhaust valves, Eng. Fail. Anal. 13 (4) (2006) 673–682.
[8] N. Vardar, A. Ekerim, Investigation of exhaust valve failure in heavy â?? duty diesel engine, Gazi Univ. J. Sci. 23 (4) (2010) 493–499.
[9] L. Witek, Failure and thermo-mechanical stress analysis of the exhaust valve of diesel engine, Eng. Fail. Anal. 66 (2016) 154–165.
[10] C. Scott, A. Riga, H. Hong, The erosion-corrosion of nickel-base diesel engine exhaust valves, Wear 181 (1995) 485–494.
[11] G.U. Kumar, V.R. Mamilla, Failure analysis of internal combustion engine valves by using ANSYS, AIJRSTEM 14 (183) (2014) 169–173.
[12] M. Karamangil, A. Avci, H. Bilal, Investigation of the effect of different carbon film thickness on the exhaust valve, Heat Mass Transf. 44 (5) (2008) 587–598.
[13] W. Giles, Valve problems with lead free gasoline, SAE Technical Paper, 1971.
[14] M. Lane, P. Smith, Developments in sintered valve seat inserts, SAE Technical Paper, 1982.
[15] R. Lewis, R.S. Dwyer-Joyce, Automotive engine valve recession, Professional Engineering, 2002.
[16] R. Lewis, R.S. Dwyer-Joyce, Combating automative engine valve recession, Tribol. Lubr. Technol. 59 (10) (2003) 48–51.
[17] R. Scott, Stationary natural gas engine lubrication, Machinery Lubrication, Noria, US, 5/22/2015.
[18] K.D. Ramkumar, et al., Investigations on the microstructure, tensile strength and high temperature corrosion behaviour of Inconel 625 and Inconel 718 dissimilar
joints, J. Manuf. Process. 25 (2017) 306–322.
[19] Y. Zhu, Y. Guo, L. Yang, Microstructure-based strength distribution across the welds of nickel-based superalloy inconel 751 and austenite steel 21-4N joined by
inertia friction welding, Metall. Mater. Trans. B 44 (2) (2013) 396–405.
[20] V.A. Johnson, C.W. Galen, Diesel exhaust valves, SAE Technical Paper, 1966.
[21] M. Logan, K. Carabell, N. Smrcka, Gas Engines Need Gas Engine Oils, Citeseer, 1999.
[22] C. Munro, Analysis of a failed Detroit diesel series 149 generator, Eng. Fail. Anal. 35 (2013) 499–507.
[23] D. Godfrey, R.L. Courtney, Investigation of the mechanism of exhaust valve seat wear in engines run on unleaded gasoline, SAE Technical Paper, 1971.
[24] Y.K. Lavhale, J. Salunke, Overview of failure trend of inlet & exhaust valve, Int. J. Mech. Eng. Technol. 5 (3) (2014) 104–113.
[25] R. Zhao, et al., Wear mechanism analysis of engine exhaust valve seats with a laboratory simulator, Tribol. Trans. 40 (2) (1997) 209–218.
[26] M. Tauschek, J. Newton, Valve seat distortion, SAE National Passenger Car, Body and Materials Meeting, Detroit, Mich., Society of Automotive Engineers, 1953.
[27] N.K. Raghuwanshi, A. Pandey, R. Mandloi, Failure analysis of internal combustion engine valves: a review, Int. J. Innov. Res. Sci. Eng. Technol. 1 (2) (2012)
173–181.
[28] G. Mahobia, N. Paulose, V. Singh, Hot corrosion behavior of superalloy IN718 at 550 and 650 C, J. Mater. Eng. Perform. 22 (8) (2013) 2418–2435.
[29] T. Hejwowski, Investigations of corrosion resistance of Fe-, Ni- and Co-based hardfacings, Vacuum 80 (11 − 12) (2006) 1386–1390.

87
M.I. Khan et al. Engineering Failure Analysis 85 (2018) 77–88

[30] R. Van Diesel, et al., Engine valve seat and insert wear, SAE Technical Paper, 1989.
[31] A. Pandey, R. Mandloi, Surface morphology and wear investigation of valves used in LPG-fueled passenger car engines, Asian J. Appl. Sci. Eng. 4 (1) (2015)
37–46.
[32] W.R. Pyle, N.R. Smrcka, The effect of lubricating oil additives on valve recession in stationary gaseous-fueled four-cycle engines, SAE Technical Paper, 1993.
[33] T. Ootani, et al., Impact wear characteristics of engine valve and valve seat insert materials at high temperature (impact wear tests of austenitic heat-resistant
steel SUH36 against Fe-base sintered alloy using plane specimens), Wear 188 (1) (1995) 175–184.
[34] A. Pandey, R. Mandloi, Effects of high temperature on the microstructure of automotive engine valves, Int. J. Eng. Res. Appl. 4 (3) (2014) 122–126.
[35] E.B. Arnold, et al., Development and application of a cycle for evaluating factors contributing to diesel engine valve guttering, SAE Technical Paper, 1988.
[36] J. Heywood, Internal Combustion Engine Fundamentals, McGraw-Hill Education, 1988.
[37] J. Tantet, P. Brown, Series 3 Oils and Their Suitability for Wider Applications, NPRA, Tech, 1965.
[38] R. Venkataraman, S. Eser, Characterization of deposits formed on diesel injectors in field test and from thermal oxidative degradation of n-hexadecane in a
laboratory reactor, Chem. Cent. J. 2 (1) (2008) 25.

88

View publication stats