U.S. patent application number 09/835797 was filed with the patent office on 2001-12-13 for system for reduction of harmful exhaust emissions from diesel engines.
Invention is credited to Kenneth, Voss E., Wildman, Timothy D..
Application Number | 20010049936 09/835797 |
Document ID | / |
Family ID | 27409933 |
Filed Date | 2001-12-13 |
United States Patent
Application |
20010049936 |
Kind Code |
A1 |
Kenneth, Voss E. ; et
al. |
December 13, 2001 |
System for reduction of harmful exhaust emissions from diesel
engines
Abstract
Methods and apparatus for reducing the TPM level of a diesel
engine exhaust stream by providing a suitable oxidation catalyst
into the exhaust train. The oxidation catalyst may be incorporated
into a thermal insulative coating on the inner surface of the
exhaust train, particularly the exhaust manifold and exhaust pipes
prior to the turbocharger. Alternatively, when the exhaust train
includes a turbocharger, the catalyst can be in a separate
monolithic unit between the engine and the turbocharger. The system
may also include an improved diesel oxidation catalyst unit having
a metal monolithic substrate. The oxidation catalyst can also be
incorporated into a thermal insulative coating inside the
cylinders, particularly on non-rubbing surfaces such as The
invention also includes the use of a protective mullite top coat on
the thermal coating. A further embodiment is the use of a stainless
steel bond coat to bind the thermal coating to a metallic
substrate, particularly an aluminum substrate.
Inventors: |
Kenneth, Voss E.;
(Somerville, NJ) ; Wildman, Timothy D.; (Monmouth
Junction, NJ) |
Correspondence
Address: |
Chief Patent Counsel
Engelhard Corporation
101 Wood Avenue
Iselin
NJ
08830-0770
US
|
Family ID: |
27409933 |
Appl. No.: |
09/835797 |
Filed: |
April 16, 2001 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
09835797 |
Apr 16, 2001 |
|
|
|
09390192 |
Sep 7, 1999 |
|
|
|
6256984 |
|
|
|
|
09390192 |
Sep 7, 1999 |
|
|
|
08838907 |
Apr 11, 1997 |
|
|
|
6006516 |
|
|
|
|
08838907 |
Apr 11, 1997 |
|
|
|
08635345 |
Apr 19, 1996 |
|
|
|
5987882 |
|
|
|
|
Current U.S.
Class: |
60/299 ; 422/180;
427/250 |
Current CPC
Class: |
B01D 53/944 20130101;
F02B 51/02 20130101; Y02T 10/126 20130101; C23C 30/00 20130101;
F01N 13/14 20130101; F01N 13/102 20130101; F01N 3/2807 20130101;
Y10T 428/12056 20150115; F02B 3/06 20130101; Y02T 10/12
20130101 |
Class at
Publication: |
60/299 ; 422/180;
427/250 |
International
Class: |
F01N 003/10; C23C
016/00; B01D 053/34; F01N 003/20 |
Claims
We claim:
1. In a diesel power system which includes a diesel engine and an
exhaust train through which the exhaust from the diesel engine
passes, a method of reducing the total particulate matter emissions
in said exhaust from the diesel engine comprising: a) thermally
insulating at least a portion of the surface of said exhaust train
which comes into contact with said exhaust with a thermal barrier
coating; and b) incorporating an oxidation catalyst into at least a
portion of the thermal barrier coating in operative contact with
the exhaust.
2. The method of claim 1 wherein said exhaust train includes an
exhaust manifold mounted on said engine for receiving the exhaust
from said engine, wherein the step of thermally insulating
comprises insulating at least a portion of the surface of the
manifold which comes into contact with the exhaust, and the
oxidation catalyst is incorporated into at least a portion of the
surface of the manifold which is thermally insulated.
3. The method of claim 2 comprising insulating substantially all of
the surface of the manifold which comes into contact with the
exhaust, and the oxidation catalyst is incorporated into
substantially all of the surface of the manifold which is thermally
insulated.
4. The method of claim 2 wherein the exhaust train further
comprises a turbocharger mounted downstream of the manifold and
operationally connected to said manifold by a connecting pipe, and
wherein the step of thermally insulating comprises insulating at
least a portion of the surfaces of both the manifold and the
connecting pipe which come into contact with the exhaust.
5. The method of claim 4 comprising insulating substantially all of
the surfaces of the manifold and the connecting pipe which come
into contact with the exhaust.
6. The method of claim 5 wherein the oxidation catalyst is
incorporated into substantially all of the surfaces of the manifold
and connecting pipe which are thermally insulated.
7. The method of claim 1 wherein the step of thermally insulating
comprises insulating substantially all of the surface of the
exhaust train which comes into contact with the exhaust from where
the exhaust exits the diesel engine to a preselected point
downstream on the exhaust train.
8. The method of claim 7 further wherein the oxidation catalyst is
incorporated into substantially all of the surface which is
thermally insulated.
9. The method of claim 1 wherein the diesel engine comprises one or
more cylinders having combustion chambers wherein the method
further comprising thermally insulating the inner surfaces of the
combustion chamber with a thermal barrier coating.
10. The method of claim 8 further comprising incorporating an
oxidation catalyst into the thermal barrier coating of the
combustion chamber in operative contact with the gases therein.
11. The method of claim 1 wherein said oxidation catalyst comprises
a base metal oxide.
12. The method of claim 11 wherein said oxidation catalyst
comprises a rare-earth metal oxide.
13. The method of claim 12 wherein said oxidation catalyst
comprises praseodymium oxide, cerium oxide or combinations thereof,
or a mixed oxide containing praseodymium, cerium or combinations
thereof.
14. In a diesel power system which includes a diesel engine and an
exhaust train through which the exhaust from the diesel engine
passes, and wherein the exhaust train comprises a turbocharger, a
method of reducing the total particulate matter emissions in said
exhaust from the diesel engine comprising providing an oxidation
catalyst in said exhaust train between the engine and the
turbocharger, wherein the oxidation catalyst is in operative
contact with the exhaust.
15. The method of claim 14 wherein the oxidation catalyst is
deposited on at least a portion of the surface of said exhaust
train between the engine and the turbocharger which comes into
contact with said exhaust.
16. The method of claim 15 wherein the oxidation catalyst is
deposited on substantially all of the surface of said exhaust train
between the engine and the turbocharger which comes into contact
with said exhaust.
17. The method of claim 14 wherein the oxidation catalyst is
mounted on the surface of a monolithic support.
18. The method of claim 17 further comprising thermally insulating
at least a portion of the surface of said exhaust train between the
engine and the turbocharger which comes into contact with said
exhaust with a thermal barrier coating.
19. The method of claim 18 comprising thermally insulating
substantially all of the surface of said exhaust train between the
engine and the turbocharger which comes into contact with said
exhaust with a thermal barrier coating.
20. The method of claim 18 further comprising incorporating an
oxidation catalyst into at least a portion of the thermal barrier
coating in operative contact with the exhaust.
21. The method of claim 14 wherein said oxidation catalyst
comprises a rare-earth metal oxide.
22. The method of claim 21 wherein said oxidation catalyst
comprises praseodymium oxide, cerium oxide or combinations thereof,
or a mixed oxide containing praseodymium, cerium or combinations
thereof.
23. In a diesel power system which includes a diesel engine and an
exhaust train through which the exhaust from the diesel engine
passes, a system for reducing the total particulate matter
emissions in said exhaust from the diesel engine comprising: a) a
thermal barrier coating on at least a portion of the surface of
said exhaust train which comes into contact with said exhaust; and
b) an oxidation catalyst incorporated into at least a portion of
the thermal barrier coating in operative contact with the
exhaust.
24. A diesel engine exhaust manifold comprising a thermal barrier
coating on at least a portion of the inner surface of said manifold
and an oxidation catalyst incorporated into at least a portion of
the thermal barrier coating, said catalyst located to be in
operative contact with an exhaust stream passing through the
manifold.
25. In a diesel engine having one or more cylinders which have
combustion chambers, a catalyzed thermal barrier coating for the
surfaces of components of the combustion chambers comprising: a) a
thermal barrier coating deposited on the surfaces of said
components; and b) an oxidation catalyst which is provided at the
surface of the thermal barrier coating.
26. The catalyzed thermal barrier coating of claim 25 wherein the
oxidation catalyst is incorporated into the thermal barrier
coating.
27. The catalyzed thermal barrier coating of claim 25 wherein the
oxidation catalyst is coated onto the thermal barrier coating.
28. The catalyzed thermal barrier coating of claim 25 wherein said
components are selected from the group consisting of the piston
crowns, cylinder heads, and valves.
29. The catalyzed thermal barrier coating of claim 25 wherein the
oxidation catalyst comprises a base metal oxide.
30. The catalyzed thermal barrier coating of claim 29 wherein the
oxidation catalyst comprises a rare-earth metal oxide.
31. The catalyzed thermal barrier coating of claim 30 wherein the
oxidation catalyst comprises praseodymium oxide, cerium oxide or
combinations thereof, or a mixed oxide containing praseodymium,
cerium or combinations thereof.
32. In a diesel engine having one or more cylinders which have
combustion chambers, a method of reducing the total particulate
emissions in the exhaust from the diesel engine comprising: a)
depositing a thermal barrier coating on the surface of components
in the combustion chambers; and b) providing an oxidation catalyst
at the surface of said thermal barrier coating.
33. The method of claim 32 wherein the oxidation catalyst is
incorporated into the thermal barrier coating.
34. The method of claim 32 wherein the oxidation catalyst is coated
onto the thermal barrier coating.
35. The method of claim 32 wherein the coating is deposited on
components of the combustion chamber selected from the group
consisting of the piston crowns, cylinder heads, and valves.
36. The method of claim 32 wherein the oxidation catalyst comprises
a base metal oxide.
37. The method of claim 36 wherein the oxidation catalyst comprises
a rare-earth metal oxide.
38. The method of claim 37 wherein the oxidation catalyst comprises
praseodymium oxide, cerium oxide or combinations thereof, or a
mixed oxide containing praseodymium, cerium or combinations
thereof.
39. An improved ceramic thermal barrier coating for a metallic
substrate in which the improvement comprises a mullite top
coat.
40. The ceramic thermal barrier coating of claim 35 comprising a
bond coat on the metallic substrate, an yttria stabilized zirconia
intermediate coat, and the mullite top coat.
41. The ceramic thermal barrier coating of claim 36 in which the
bond coat comprises an MCrAlY alloy.
42. The ceramic thermal barrier coating of claim 36 in which the
bond coat comprises a martensitic stainless steel.
43. The ceramic thermal barrier coating of claim 30 further
comprising an oxidation catalyst provided at the surface of the
mullite top coat.
44. A method of protecting a ceramic thermal barrier coating
comprising depositing a mullite top coat onto the ceramic thermal
barrier coating.
45. The method of claim 44 wherein the thermal barrier coating
comprises a bond coat on the metallic substrate and an yttria
stabilized zirconia intermediate coat onto which the mullite top
coat is deposited.
46. The method of claim 44 further comprising providing an
oxidation catalyst at the surface of the mullite top coat.
47. A thermal barrier coating for an aluminum substrate comprising
a stainless steel bond coat deposited on the aluminum substrate and
a ceramic thermal barrier coating deposited on the stainless steel
bond coat.
48. The coating of claim 47 wherein the stainless steel is
martensitic.
49. The coating of claim 48 wherein the stainless steel is a type
431 martensitic stainless steel.
50. The coating of claim 47 wherein the ceramic thermal barrier
coating deposited on the stainless steel is an yttria stabilized
zirconia.
51. A method of bonding a ceramic coating to an aluminum substrate
comprising depositing a bond coat of stainless steel on the
aluminum substrate, and then depositing a top coat of the ceramic
on the bond coat.
52. The method of claim 51 wherein the stainless steel is
martensitic.
53. The method of claim 52 wherein the stainless steel is a type
431 martensitic stainless steel.
54. The method of claim 51 wherein the bond coat is deposited by
thermal spraying.
55. The method of claim 51 wherein the ceramic is an yttria
stabilized zirconia.
Description
RELATED APPLICATIONS
[0001] This application is a continuation-in-part of U.S. Ser. No.
08/635,345, filed Apr. 19, 1996.
BACKGROUND OF THE INVENTION
[0002] 1. Field of the Invention
[0003] This invention relates to a system for the reduction of
harmful exhaust emissions from diesel engines, and more
particularly to a system for increasing the effectiveness of the
oxidation of the oxidizable components in the exhaust
emissions.
[0004] 2. Description of Related Art
[0005] Diesel engine exhaust is a heterogeneous mixture which
contains not only gaseous emissions such as carbon monoxide ("CO"),
unburned hydrocarbons ("HC") and nitrogen oxides ("NO.sub.x"), but
also condensed phase materials (liquids and solids) which
constitute the so-called particulates or particulate matter ("PM").
The total particulate matter ("TPM") emissions are comprised of
three main components. One component is the solid, dry, solid
carbonaceous fraction or soot. This dry carbonaceous matter
contributes to the visible soot emissions commonly associated with
diesel exhaust. A second component of the TPM is the soluble
organic fraction ("SOF"). The soluble organic fraction is sometimes
referred to as the volatile organic fraction ("VOF"), which
terminology will be used herein. The VOF may exist in diesel
exhaust either as a vapor or as an aerosol (fine droplets of liquid
condensate) depending on the temperature of the diesel exhaust, and
are generally present as condensed liquids at the standard
particulate collection temperature of 52.degree. C. in diluted
exhaust, as prescribed by a standard measurement test, such as the
U.S. Heavy Duty Transient Federal Test Procedure, discussed further
below. These liquids arise from two sources: (1) lubricating oil
swept from the cylinder walls of the engine each time the pistons
go up and down; and (2) unburned or partially burned diesel
fuel.
[0006] The third component of the particulates is the so-called
sulfate fraction. Diesel fuel contains sulfur, and even the low
sulfur fuel available in the U.S. may contain 0.05% sulfur. Upon
combustion of the fuel in the engine, nearly all of the sulfur is
oxidized to sulfur dioxide which exits with the exhaust in the gas
phase. However, a small portion of the sulfur, perhaps 2-5%, is
oxidized further to SO.sub.3, which in turn combines rapidly with
water in the exhaust to form sulfuric acid which collects as a
condensed phase with the particulates as an aerosol, or is adsorbed
onto the other particulate components, and thereby adds to the mass
of TPM.
[0007] Emissions from diesel engines have been under increasing
scrutiny in recent years and standards, especially for particulate
emissions, have become stricter. In 1994 the particulate emission
standards in the U.S. for new engines allowed no more than a total
of 0.1 grams per brake horse power hour (g/BHP-h). For diesel
engines in buses operating in congested urban areas the particulate
emissions standard was even stricter, 0.07 g/BHP-h TPM. Both of
these standards were seen as significant reductions relative to the
prior particulate emission standard of 0.25 g/BHP-h which had been
in effect since 1991. Starting in 1994, for the first time, engine
technology developments alone were found to be incapable of meeting
the new standards, and for some engines aftertreatment technology,
for example, diesel oxidation catalyst (DOC) units, as discussed
further below, were necessary.
[0008] Current engines are generally capable of meeting the 1994
NO.sub.x emissions standards of 5.0 g/BHP-h, but by only a slim
margin. Diesel engines, because they operate with a great excess of
combustion air (lean exhaust) typically have emissions of CO and
gas phase HC's which are well below the 1994 emissions standards of
15.5 g/BHP-h and 1.3 g/BHP-h, respectively. Therefore, the key
emission control concerns for diesel engines now and for the
immediate future are the reduction in particulates (TPM) and
NO.sub.x emissions.
[0009] Emissions of NO.sub.x from diesel engines can be reduced by
retarding injection timing. However, this is accompanied by a
corresponding increase in particulate emissions, particularly of
the dry carbon or soot portion. Emissions of NO.sub.x can also be
reduced by applying exhaust gas recirculation (EGR) technology.
However, this is also accompanied by a corresponding increase in
particulate emissions. Thus, both of these engine technologies are
constrained by a trade-off or balance between TPM and NO.sub.x
emissions.
[0010] Additional EPA requirements went into effect at the
beginning of 1995 which apply to urban buses equipped with engines
manufactured prior to 1994. These requirements apply to engines in
service when they come due for rebuilding. Following engine rebuild
the requirements must be met. One portion of these requirements
specifies that if technology can be demonstrated for particulate
reduction for these pre-1994 bus engines, that technology would be
mandated for use on those engines for which it is certified. Two
tiers of such technology/emissions reduction attainment were
promulgated including:
[0011] 1. Meet the 1994 Emissions Standard of 0.1 g/BHP-h TPM, with
a technology cost cap of about $8,000.
[0012] 2. Reduce Engine-Out TPM Emissions by at least 25%, with a
technology cost cap of about $3,000.
[0013] The first of the above attainment levels, which is
considered the stricter of the two requirements, if demonstrated
and certified, takes precedence. Thus, the 25% TPM reduction tier
was considered a "fall-back" position, if the 0.1 g/BHP-h TPM tier
could not be met. It is clear from the strict emissions
requirements for new diesel engines used in urban buses and the
attainment requirement for pre-1994 bus engines that a major
challenge exists for this type of application.
[0014] Diesel engines used in urban bus applications in the U.S.
are of many types, both two-cycle and four-cycle, supplied by a
range of engine original equipment manufacturers (OEM's). However,
a large percentage of urban transit buses have two-cycle engines
from one manufacturer (Detroit Diesel Corp.). The emissions
reduction system of this invention is considered to be applicable
to any diesel engine for lowering emissions and the level of
emissions reduction attained is expected to be dependent on the
specific engine, its operating parameters and baseline engine-out
emissions. However, this invention has been found to be especially
useful for two-stroke diesel engines, and as demonstrated herein,
can be used with such engines manufactured prior to 1994 to bring
them into compliance with the 1994 particulate emissions standard
of 0.1 g/BHP-h TPM, as discussed above.
[0015] Oxidation catalysts comprising a platinum group metal
dispersed on a refractory metal oxide support are known for use in
treating the exhaust of diesel engines in order to convert both HC
and CO gaseous pollutants and particulates, i.e., soot particles,
by catalyzing the oxidation of these pollutants to carbon dioxide
and water. Such catalysts have generally been contained in units
called diesel oxidation catalysts (DOC's), or more simply catalytic
converters or catalyzers, which are placed in the exhaust train of
diesel power systems to treat the exhaust before it vents to the
atmosphere. However, by the time the exhaust gas reaches the
catalyzer, it has generally lost a considerable amount of heat,
both by radiation through the engine and exhaust system walls, and
by intentional power transfer at the turbocharger. Because the
efficiency of such catalytic oxidation processes is generally a
direct function of the gas temperature, such temperature losses can
have a significant negative impact on the effectiveness of the
catalyzer.
[0016] One approach to improving the effectiveness of the catalyzer
is to maintain the exhaust temperature at as high a level as
possible, from the combustion chamber and through the connecting
exhaust train to the catalyzer. Heat-insulating structures and
heat-insulating coatings, i.e., thermal barrier coatings have been
employed by those skilled in the art to enhance the thermal
efficiency of internal combustion engines by permitting more
complete fuel burning at higher temperatures. Typically, such
heat-insulating coatings have been applied to all of the chamber
surfaces, including the cylinder walls and head and piston
combustion faces to prevent heat loss. Heat-insulating structures
and heat-insulating coatings have also been used in automobile
exhaust systems to maintain high exhaust temperatures required by
thermal reactors and catalytic converters, thus reducing the
emission of unburned hydrocarbons emitted into the atmosphere as an
undesirable component of exhaust gas.
[0017] U.S. Pat. No. 5,384,200 is directed to particular thermal
barrier coatings and methods of depositing such coatings on the
surfaces of combustion chamber components. As discussed in that
patent, insulating the combustion chamber components reduces the
amount of heat loss in the engine. The higher temperature in the
combustion chamber results in a more complete combustion of the
fuel in the chamber, and also results in a hotter exhaust being
delivered to any downstream catalytic converters to promote more
effective oxidation of the oxidizable components of the exhaust
stream.
[0018] The use of thermal barrier coatings has also been suggested
for engine components other than in-cylinder surfaces. In a paper
entitled "High Performance Coatings for Diesels and Other Heat
Engines", by Roy Kamo, presented at the Thermal Spray Coatings
Conference, Gorham Advanced Materials Institute, Orlando, Fla., on
Sep. 12-14, 1993, it is suggested that engine performance can be
improved by applying thermal barrier coatings to various engine
components. In addition to in-cylinder surfaces such as the piston
crown and cylinder head, the article also suggests the exhaust
port, exhaust manifold and turbocharger housing.
SUMMARY OF THE INVENTION
[0019] A typical diesel power system includes a diesel engine and
an exhaust train through which the exhaust from the diesel engine
passes. The present invention is directed to methods and apparatus
for reducing the total particulate matter emissions in said exhaust
from the diesel engine. One embodiment of the method of the present
invention comprises thermally insulating at least a portion of the
surface of the exhaust train which comes into contact with the
exhaust with a thermal barrier coating, and incorporating an
oxidation catalyst into at least a portion of the thermal barrier
coating in operative contact with the exhaust. This is accomplished
by thermally insulating at least a portion, preferably the hottest
portion of the exhaust train which carries the hot exhaust gas
stream from the diesel engine to the atmospheric vent. The
insulation is applied to surfaces of the exhaust train which are in
direct contact with exhaust gas, that is, the inside surfaces of
the exhaust train components. The oxidation catalyst is
incorporated into at least a portion of the thermal coating, and
optionally into substantially all of the thermal coating.
[0020] Preferred oxidation catalysts for use in the present
invention are base metal oxides, particularly the rare-earth metal
oxides, or mixtures of materials containing the base metal oxides
or rare-earth metal oxides. Preferred rare-earth oxide catalysts
for use in this invention are praseodymium oxide and ceria. Good
results are also obtained with other rare-earth oxides, as
discussed further below. The base metal oxides can be used alone,
or in combination with catalytic platinum group metals, such as
platinum, palladium and rhodium.
[0021] By insulating the exhaust train in accordance with the
present invention, the effectiveness of the oxidation of the
oxidizable components of the exhaust is increased in a downstream
diesel oxidation catalyst (DOC) unit, and this decreases the level
of undesirable emissions in the exhaust. Incorporating an oxidation
catalyst into the thermal coating further reduces the emissions,
particularly the TPM emissions. The catalyzed thermal barrier and
the downstream DOC unit combine for a significant reduction in the
overall pollution level in the exhaust.
[0022] Typically, the exhaust train of a diesel power system
includes a manifold to collect the exhaust from the engine and
channel it into one or more exhaust pipes. Being closest to the
engine, the manifold is generally the hottest section of the
exhaust train. Therefore, in a preferred embodiment of the present
invention, at least a portion of the inner surface of the manifold
is insulated to reduce the amount of heat lost through the manifold
walls and thus maintain the exhaust at high temperature.
Preferably, substantially the entire inner surface of the manifold
is coated, that is at least about 90% of the area exposed to the
hot exhaust gases.
[0023] From the manifold, pipes carry the exhaust through various
apparatus which may be present in the exhaust train. Typically, a
turbocharger is provided downstream of the manifold. Such devices
are well known to those skilled in the art. A turbocharger
mechanically extracts power from the exhaust stream, such as by a
compressor driven by the exhaust, and transfers it to the inlet air
stream to improve the overall efficiency of the diesel power
system. As a result of such power extraction, the temperature of
the exhaust gas generally drops significantly, such as about
100.degree. F. or more, as it passes through the turbocharger. It
is therefore a further preferred embodiment of the present
invention not only to insulate the manifold, but also to insulate
the pipe or pipes connecting the manifold to the turbocharger, when
a turbocharger is present.
[0024] After the exhaust exits the turbocharger, it is at a lower
temperature. Downstream of the turbocharger, many commercial diesel
power systems include a diesel exhaust oxidizer (DOC), as discussed
above, for oxidizing the oxidizable components of the exhaust
stream. Generally, the hotter the exhaust is when it enters the
catalytic oxidizer, the more effective the oxidizer is in oxidizing
the harmful oxidizable components. It is therefore a further
embodiment of the present invention to insulate the pipes
connecting the turbocharger to the downstream catalytic
oxidizer.
[0025] A further embodiment of the present invention is to combine
the insulative coating of the exhaust train, as discussed above,
with insulative coating of the surfaces of the combustion chamber
components, in order to maximize the combustion of the fuel in the
combustion chamber and to further impede heat loss from the exhaust
stream. The surfaces to be coated can include the piston crown, the
cylinder head and the valve faces, as well as any other surfaces
which are exposed to the combustion.
[0026] As discussed above, catalytic converters in diesel power
systems are located in the exhaust train, and the effectiveness of
the catalysts in such converters is reduced by temperature loss in
the exhaust train. In accordance with one aspect of the present
invention the catalytic oxidation of oxidizable components in the
exhaust stream is improved by providing catalysts in the thermal
barrier coatings which are applied to the exhaust train. By
providing the catalysts in the high temperature end of the exhaust
train, the catalysts are able to act on the exhaust gas when it is
at its highest temperature. Furthermore, because such oxidation is
an exothermic reaction, it is possible that this catalytic
oxidation may increase the temperature of the exhaust gas, thus
promoting more effective oxidation downstream at the catalytic
converter.
[0027] In another embodiment of the present invention, in which the
diesel power system exhaust train includes a turbocharger, the
method of reducing the total particulate matter emissions in the
exhaust simply comprises providing an oxidation catalyst in the
exhaust train between the engine and the turbocharger. In this
case, the oxidation catalyst can be mounted on the operating
surfaces of a monolithic support, of the type well known in the
art. As discussed further below, the turbocharger can significantly
reduce the temperature of the exhaust gases. By providing catalyst
in the exhaust train prior to the turbocharger, the catalytic
oxidation can be conducted at the elevated exhaust temperatures
before the turbocharger is reached. Optionally, the inner surfaces
of the exhaust train can also be insulated, as discussed above.
Also, if the surfaces are insulated, additional catalyst can be
incorporated into a portion or substantially all of the
insulation.
[0028] In a further embodiment of the present invention, when the
combustion chamber is provided with a thermal barrier coating, an
oxidation catalyst is provided on or in such coating in the
combustion chamber. Oxidation catalysts in the combustion chamber
can promote more complete oxidation of the fuel, thereby decreasing
the amount of undesirable emissions sent to the exhaust train.
[0029] In a particular embodiment of the present invention,
catalytic ceramic coatings are applied to the in-cylinder surfaces
of the combustion chambers of an internal combustion engine,
especially a compression ignition (diesel) engine. These coatings
are of low thermal conductivity compared with standard metal parts,
and provide a thermal barrier at the combustion chamber walls which
keeps heat in the cylinder and promotes more complete combustion of
the fuel, thereby giving greater fuel efficiency and lower
emissions of particulates (SOF and dry-carbon/soot). In addition
the coatings are provided with catalytic surfaces which further
promote combustion of unburned fuel and particulates for increased
efficiency and lower particulate emissions.
[0030] Another aspect of the present invention is that it was found
that providing smooth and non-porous surface properties to the
coatings, also contributes to improved combustion of unburned fuel
and soot. This is believed to be the result of low drag at the
surface which allows the swirl and mixing of the fuel. Such smooth
and non-porous surfaces also reduce adsorption of the fuel which
impacts the coated combustion chamber walls. Such adsorption can
cause slow and incomplete combustion. The smoothness of thermal
sprayed coated surfaces can be further improved by sanding or
polishing. The increased smoothness can further reduce drag at the
surface and thereby improve the mixing and swirling action of the
fuel-air mixture, which in turn leads to better combustion.
[0031] In a particular embodiment of the present invention a top
coat of mullite is used to protect the ceramic coating. The ceramic
coatings are essentially the same as otherwise described herein for
insulative coatings, except that a top layer of an alumino-silicate
(mullite) ceramic layer is provided. The layer should be at least
about 2 mils thick, and preferably about 3 to 5 mils thick. Under
this typically is the yttria-stabilized zirconia layer and beneath
that the bond coat. These coatings are therefore three layer
coatings with an overall thickness approximately the same as the
above coatings without mullite. The reason for the mullite top
layer is to provide a very chemically inert ceramic surface with
high resistance to the corrosive and aggressive materials
encountered in the combustion and exhaust (sulfur and sulfates,
calcium, zinc, nitrogen oxides and nitrates, chlorides, phosphorus,
etc.). The "as sprayed" mullite top layer is less porous (about 3%
porosity) than the zirconia layer, which is about 10% porosity.
[0032] The combustion chamber surfaces which are coated with the
catalytic materials, such as catalytic oxides, will be at least the
crowns (including the bowls of pistons which have bowls for
receiving injected fuel, as is well-known in the art). The exhaust
valves and cylinder head fire decks can also be coated with the
catalytic oxides. The overall coating thickness should be
relatively thin, for example, less than about 20 mils. and
preferably less than about 15 mils. These thicknesses are the total
of all coatings, such as a bond coat, ceramic thermal barrier
layer(s) and catalytic oxide layer(s). This thickness criteria is
to produce a coating with good thermo-mechanical properties and
which will exhibit good long-term durability in-cylinder.
[0033] Another aspect of this invention is the use of stainless
steel bond coats, particularly martensitic steels, to give good
durability when the coatings are applied to aluminum alloy
surfaces. Aluminum alloy pistons are used in many internal
combustion engines, thus the need for the special bond coat. The
typical alloy bond coats used with prior thermal spray coatings
have been comprised of MCrAlY alloys where M=Ni, Co, Fe, etc., as
is well-known in the art. In accordance with the present invention,
stainless steel, particularly martensitic stainless steel, such as
type 431 SS, has been found to be particularly useful as a bond
coat to bond ceramics, such as yttria stabilized zirconia, to
aluminum substrates.
[0034] It is believed that the reason such stainless steels are
effective as bond coats is related to the relative coefficients of
thermal expansion (CTE) of the components. Aluminum has a CTE of
about 23.0, and aluminum alloys have CTE's in the general range of
about 21 to 25. The ceramic thermal barrier coatings have CTE's of
roughly about 6 to 8, with standard 7% yttria stabilized zirconia
being about 7.6. The stainless steels used in the present
invention, particularly martensitic stainless steels, have
relatively low CTE's, which approximate those of the ceramic
thermal barrier coatings. Such steels are commonly referred to as
being "low-shrink", because of their relatively low CTE. At the
same time, such stainless steels also bond well to the aluminum
substrate. It is believed that the use of a stainless steel which
has a coefficient of expansion relatively close to that of the
ceramic coatings acts to keep the ceramic firmly adhered to the
aluminum. The CTE of type 431 SS martensitic stainless steel is
about 6.6, which is even less than that of the yttria stabilized
zirconia used as the thermal barrier coating in some of the
examples below. The stainless steel bond coat does not move
differentially to the zirconia ceramic, and remains firmly anchored
to the aluminum substrate. As a result, the tendency for the
coating to crack and spall off such aluminum substrates is greatly
reduced
[0035] A further aspect of the present invention is providing an
improved diesel oxidation catalyst (DOC) unit in the exhaust train.
The improved DOC unit more effectively oxidizes the oxidizable
components of the exhaust stream, thus reducing the TPM level of
the exhaust. The improved DOC unit comprises a metal monolithic
catalyst support rather than a ceramic support as used in other
units.
BRIEF DESCRIPTION OF THE DRAWINGS
[0036] FIG. 1 is a schematic representation of a diesel power
system.
[0037] FIG. 2 is an illustrative graphical printout of a combined
thermogravimetric/differential thermal analysis (TGA/DTA) for an
active catalyst, in this case containing ceria.
[0038] FIG. 3 is a printout of a TGA/DTA for relatively inert
cordierite.
[0039] FIG. 4 is a comparative bar graph of different
configurations of the present invention and comparative baseline
data.
[0040] FIG. 5 is a graph showing the comparative effects of the
injection timing setting on TPM and NO.sub.x emission levels.
DETAILS OF PREFERRED EMBODIMENTS
[0041] A simple, generalized schematic of a diesel power system 10
such as is used in commercial buses is depicted in FIG. 1. Diesel
fuel is combusted in one or more cylinders 13 of diesel engine 12.
The exhaust from diesel engine 12 is collected in exhaust manifold
14 and channeled downstream by one or more pipes 16. The exhaust
train of such diesel systems generally includes a turbocharger 18,
which extracts power from the exhaust and transfers it to the air
intake of the engine 12 by means which are well known in the art.
After turbocharger 18, pipe 20 carries the exhaust downstream to an
optional diesel oxidation catalyst (DOC) unit 22, or more simply a
catalyzer, in which a catalyst is provided to promote the oxidation
of harmful emissions contained in the exhaust. Downstream of
catalyzer 22 pipe 24 carries the exhaust to a muffler 26 from which
the exhaust exits through pipe 28 to the atmosphere. Optionally and
preferably, catalyzer 22 and muffler 26 are combined into a single
catalyzer/muffler unit, thus eliminating the need for two separate
units and intermediate connecting pipes.
[0042] The exhaust gas exiting engine 12 is at elevated
temperature, and cools as it passes through the exhaust train. In
particular, when a turbocharger is present, the temperature of the
gas is significantly reduced by the process which transfers power
to the inlet air. The temperature drop across the turbocharger can
be in excess of about 100.degree. F. In any case, as the exhaust
gas moves through the manifold and down the exhaust train to the
atmosphere, the gas loses heat through the walls of the exhaust
train by radiation or other heat transfer means.
[0043] In accordance with a preferred embodiment of the present
invention, an insulative layer 30 is provided on the inner surface
of at least a portion of the exhaust train. Although insulative
layer 30 is shown on the inside of manifold 14, turbocharger 18 and
connecting pipes 16, 20, 24 and 28, it is not necessary for all of
these surfaces to be coated to achieve the desired improvement in
the oxidation of the oxidizable exhaust components. The exhaust gas
is generally at its hottest temperature when it leaves the engine,
and it is more effective to insulate the hottest components of the
exhaust train to maintain the exhaust at the highest possible
temperature. Therefore, it is preferred to at least insulate part
of the manifold, more preferably substantially the entire manifold,
that is, at least about 90% of the surface exposed to the exhaust
gas. When a turbocharger is in the system, it is also preferable to
insulate the pipe 16 connecting the manifold to the turbocharger.
After the exhaust gas passes through the turbocharger, it is at a
significantly lower temperature, therefore insulating the piping
downstream of the turbocharger is not as significant to the overall
improvement of oxidation efficiency. Because one of the purposes of
insulating the exhaust train is to increase the effectiveness of
the catalyzer unit 22, there is little to be gained in regard to
emissions control by insulating the exhaust train downstream of the
catalyzer.
[0044] In one embodiment of the invention, an oxidation catalyst is
incorporated into at least a portion of the thermal insulative
coating 30. Preferably, the catalyst is provided in the highest
temperature regions of the exhaust stream, that is in the manifold
14 and the pipe 16 connecting the manifold to the turbocharger 18.
The catalyst can be disposed on selected portions of the insulated
surface 30, or on substantially all (that is, at least about 90%)
of the insulated surface.
[0045] Optionally, a separate oxidation catalyst unit 32 may be
provided upstream of the turbocharger. This unit can include a
monolithic catalyst support, as is well known in the art, on which
the oxidation catalyst is disposed. Such a catalyst unit can be
provided in place of incorporating catalyst into coating 30, or can
be in addition to catalyzed coating 30. When a catalyst unit 32 is
provided, it is not necessary to insulate the exhaust train with
coating 30. However, preferably at least the exhaust manifold 14
and the connecting pipe 16 are insulated, as discussed above.
[0046] Optionally, the inner surface of the cylinders 13 can also
be coated with either catalyst, thermal barrier insulation, or
catalyst incorporated into thermal barrier insulation, as discussed
above. The preferred sites for coating in-cylinder are the piston
crown 34 or the cylinder head 36, also known as the firedeck, and
valve faces. When coating in-cylinder, appropriate adjustments
would also have to be made to maintain the desired compression, as
would be understood by one skilled in the art.
[0047] The preferred catalysts are oxidation catalysts which can
combust the hydrocarbons (referred to as SOF or VOF) and
carbonaceous materials (referred to as insolubles or soot) which
make up nearly all of the total particulate mass (TPM) emitted in
the exhaust of diesel engines. The catalysts used for this
invention include base metal oxides with activity for burning SOF
and/or insolubles. These oxides can be used in combination with one
another, or also optionally in combination with precious metals
(e.g. Pt, Pd, Rh, etc.), to get enhancement in overall activity
some cases. Test results demonstrating relative performance of
different base oxides and their combinations are given below.
[0048] The base metal oxide combustion catalysts can be used in and
with a diesel engine system in a number of ways for destroying SOF
and insolubles, thereby reducing TPM emissions.
[0049] The catalysts can be used within the combustion chambers of
the diesel engine. They can be applied to the combustion chamber
surfaces, including the piston crowns, valve faces and firedeck
area of the cylinder heads. These are surfaces which can come into
contact with SOF and insolubles which can be burned there. This is
especially true of the piston crowns which are known to encounter
impingement of at least a portion of the fuel or partially burned
fuel directly on their surfaces following fuel injection. In
addition, optical observations have shown high concentrations of
soot at the outside of the flame and near the piston crown
surfaces. The catalysts can be applied directly to the combustion
chamber surfaces as a coating or they can be applied onto the
surface of another coating which as already been put onto the
combustion surfaces, such as a thermal barrier coating. This can
have an added benefit in that the thermal barrier coating can keep
heat in the cylinder and provide a surface to keep the catalyst hot
and thereby more active, plus it can provide a stable, compatible
surface for applying the oxide catalyst. The catalysts could be
applied to the above mentioned surfaces by a variety of means known
to the art including: impregnation with a solution of soluble
precursor followed by thermal or chemical decomposition to obtain
the active oxide, thermal spraying processes such as flame spraying
or plasma spraying of the oxide catalysts or their precursors, or
by application of a slurry of the catalyst material together with
appropriate binders, if needed, followed by thermal treatment to
dry, cure and set the catalyst coating.
[0050] The catalysts can also be applied as a coating, in manners
such as those described above, to the internal surfaces of the
exhaust system of the engine, including the ports in the cylinder
heads, exhaust manifolds, exhaust pipes and turbocharger housing.
As discussed above, the exhaust is significantly cooled as it
passes through a turbocharger. Therefore, the catalysts are
especially useful in the exhaust system sections prior to the
exhaust reaching the turbocharger, where the exhaust temperature is
at its highest, thereby providing the most sensible heat for
enhanced catalyst activity. As with the in-cylinder coatings above,
the catalyst can be applied directly to the inner surfaces of the
exhaust system components or applied to the surface of another
coating, e.g. a thermal barrier coating, which has already been put
onto the surfaces. The benefits of incorporating the catalyst into
the surface of the thermal barrier coating would be the same as
those described in regard to in-cylinder usage above.
[0051] In some engines, such as Detroit Diesel's 6V92TA, there is a
relatively long run of exhaust manifolds and piping connecting the
exhaust ports with the turbocharger. This particular engine has
seven sections of connecting pipes, including two exhaust manifolds
(it's a V-configuration engine), two connector pipes, one from each
manifold, which feed into a Y-connector, which in turn is connected
to a turbo-extension pipe, which finally connects to an elbow
connector, which is connected to and feeds the combined engine
exhaust to the turbocharger. This pre-turbo exhaust system not only
provides a relatively large interior geometric surface area onto
which to put a catalyst coating, as described above, but also has
several straight runs of piping with sufficient inner diameter
which could be fitted or configured with a catalyst coated
flow-thru honeycomb substrate, such as that depicted as component
32 in FIG. 1 and discussed above.
[0052] Such a separate catalyzed honeycomb unit 32 can provide a
much greater geometric surface area for contacting the catalyst
with the exhaust, thereby enhancing overall effectiveness and
activity for TPM reduction. For example, the turbo connector pipe
noted above has an inner diameter of 3.31" with a length of 11.38"
which could accommodate a flow-thru catalytic element. Such a
catalyst would be near the engine exhaust port, and before the
turbocharger. This would provide for higher temperatures for the
catalyst and thereby better activity and more rapid "light-off"
after initial engine "start up", these terms being well known in
the art, for burning the VOF and insoluble components of the TPM's.
A metallic monolithic flow-thru honeycomb would be an especially
good candidate as a substrate to place in a such a high temperature
and pressure position, because of good mechanical strength and
relatively low pressure drop. Another advantage of a metallic
substrate is that they are typically made by wrapping corrugated
metal foils in layers from the inside out, thereby allowing the
fabrication of practically any diameter substrate to fit any inner
diameter of exhaust piping. The oxide catalysts, possibly in
combination with one or more precious metals, can be coated onto
the metallic flow-thru substrate as a washcoat by conventional
techniques, including dip-coating or flow coating. The preferred
catalytic materials for use in the configurations described above
have been defined, at least initially, by laboratory testing which
is described below.
Laboratory Testing
[0053] The identification of candidate catalysts and ceramic
coating materials was first determined by laboratory testing. A
test was developed to screen various materials for their
capabilities to catalyze the combustion of diesel particulates. For
the test, the diesel particulates were simulated by a mixture of
lube oil (model SOF) and carbon black (model soot). The lube oil
and carbon black were mixed in the ratio of 30:70 to approximate
their proportions in representative diesel particulate emissions.
The lube/carbon black mixture was further combined with a powdered
sample of the candidate catalyst material in the ratio of 20 parts
model particulate and 80 parts catalyst to form a uniform, intimate
mixture. The performance of the candidate catalyst for burning the
admixed model particulates was evaluated using a
thermogravimetric/differential thermal analysis unit (TGA/DTA; STA
1500, Polymer Labs). In a given test run a small sample (e.g., 30
mg) of the catalyst/model particulate mixture was placed in the
sample pan of the TGA/DTA and heated in a flow of air from ambient
temperature to about 1000.degree. C. at a ramp rate of 20.degree.
C./min. The TGA/DTA unit was used to simultaneously measure the
weight loss and heat evolution from the test sample as a function
of temperature. The test was conducted with an atmosphere of
flowing air to simulate the lean environment found in diesel
exhaust.
[0054] The results from a representative test run of a candidate
catalyst are shown in FIG. 2. The sample weight loss (TGA, lower
trace) exhibits two breaks at different temperatures. The lower
temperature break (about 240-343.degree. C.) is due to the loss of
the lube oil (volatilization +combustion), and the higher
temperature break (about 617-713.degree. C.) is due to the
combustion of the carbon black. The heat evolution (DTA, upper
trace) due to combustion of lube and carbon black are seen as
separate peaks (exotherms) in the curve at about 358.degree. C. and
687.degree. C., respectively. All catalyst candidates exhibited the
two breaks for weight loss in the TGA trace as the lube and carbon
black were burned, respectively. However, some catalyst candidates
exhibited little or no activity for burning the lube oil and as a
result exhibited no DTA peak in the lube oil temperature range and
all the lube oil loss was due to volatilization alone. This can be
seen in FIG. 3 for cordierite powder, an inert material, which was
mixed with the model particulate mixture and run in air. As can be
seen, cordierite is completely inactive for burning lube oil and
the weight loss in the temperature range of 236.degree. C. to
334.degree. C. is due entirely to volatilization of the lube.
[0055] The carbon black is a nonvolatile solid and thus all samples
exhibited a DTA peak in the carbon black temperature range due to
combustion. As can be seen, even with an inert material such as
cordierite a carbon black DTA peak is obtained, as shown in FIG.
3.
[0056] Catalyst candidates differed in performance for burning the
lube and carbon black, as evidenced by the temperatures at which
the lube oil and carbon black combustion occurred. For the
screening tests the figure of merit was taken as the temperature of
the inflection in the weight loss curve (maximum rate of weight
loss) and the temperature of the peak in the DTA curve (maximum
rate of heat evolution from combustion). The best candidate
catalyst materials were those which exhibited the lowest
temperatures of TGA inflection and DTA peak for the combustion of
lube and carbon black. That is, the best catalyst materials shift
the burning of lube or carbon black to the lowest possible
temperature. This can be seen for carbon black burning by comparing
the DTA Peak temperature for the catalyst candidate run in FIG. 2
(687.degree. C.) to that for cordierite in FIG. 3 (708.degree.
C.)
[0057] The catalyst candidates screened in this test procedure were
mainly metal oxides, although non-oxides are within the scope of
this invention and similar screening tests can be performed on
representative samples. In addition, oxides or non-oxides in
combination with one or more precious metals is considered within
the scope of the invention.
[0058] Pure metal oxides were tested and mixtures of metal oxides
were tested. For the latter some were based on candidate catalytic
metal oxides impregnated onto low surface area, non-active oxides
via solution impregnation from a water soluble precursor followed
by calcination in air. The metal oxides chosen have been of both
low and high surface area and were tested in powder form as
described above. The low surface area oxides were of special
interest because for this invention the form envisioned is of a
ceramic coating applied to metal substrates such as piston crowns,
valves, cylinder heads, exhaust manifolds and exhaust pipes by
methods such as thermal spray techniques (plasma or HVOF spraying),
and as a result lower or very low surface area coatings are
expected to result. These coating techniques are known to give
thermo-mechanically stable and highly adherent coatings which would
be needed for the in-cylinder parts. Higher surface area oxides are
also considered because of the potential to apply them to the
substrate by slurry or wet coating processes. This could be done
for the exhaust pipes which are exposed to less extreme
thermo-mechanical conditions than are in-cylinder parts.
Furthermore, both high surface area and low surface area oxides or
ceramic materials would be candidates for catalysts to be used in
enhancing TPM insolubles or dry carbon burning in diesel oxidation
catalyst units or catalyzed diesel soot filters.
[0059] Test results for a series of candidate oxides and mixed
oxide systems are given in Table I. In Table I, the TGA inflection
temperatures and DTA peak temperatures are given for each
candidate, and the results showed significant variations for
different candidates. For purposes of the present invention, the
lowest possible temperature is desired for DTA peak and TGA
inflection. As will be shown by the results presented below, it is
of particular importance for the catalyst to affect the combustion
of the carbon component of the TPM, because the downstream diesel
oxidation catalyst units have been found to be much more effective
on the VOF component than on the dry soot. Therefore, the best
candidate catalysts for use in the present invention are those with
the lowest DTA inflection and TGA peak temperatures for carbon
black combustion. These results indicated that praseodymium oxide
(Pr.sub.6O.sub.11) showed a significant ability to decrease the TGA
and DTA values for dry carbon burning, especially Samples 21 and
23-27.
[0060] High surface area cerium oxide (Sample 20) also showed very
good combustion characteristics (low TGA and DTA values) for carbon
black burning. In addition, this cerium oxide (Sample 20) also was
the best for burning the lube oil component of the model
particulates (i.e., lowest DTA peak temperature). As can be seen
from Table I, samples of cerium oxide with lower surface areas
(e.g., Sample 16 with SA=0.9 m.sup.2/g and Sample 17 with SA=8.9
m.sup.2/g) did not exhibit as low TGA and DTA values for carbon
black combustion as did Sample 20. This shows that with ceria,
surface area, and specifically high surface area, is important for
obtaining the best activity for carbon black combustion, as well as
lube oil combustion. Thus, high surface area cerium (or cerium
containing) oxides are included as preferred catalysts.
[0061] Praseodymium oxide, also a preferred catalyst, can be seem
form Table I to be highly active for carbon black combustion.
Furthermore, praseodymium oxide is highly active with a low surface
area (3.7-23 m.sup.2/g, Samples 23-25). Thus, surface area of
praseodymium oxide is not as important a factor for high carbon
black combustion activity as is for ceria. However, praseodymium
oxide does not exhibit as high an activity for lube oil
combustion.
1TABLE I Laboratory TGA/DTA Test Results with Various Oxides for
Lube Oil & Carbon Black Combustion in Air TGA Inflection DTA
Peak Temp. (.degree. C.) Temp. (.degree. C.) SA Lube Lube Sample
Catalyst (m.sup.2/g) Oil Carbon Oil Carbon 1 alpha-Alumina 8.9 287
697 388 687 (Al.sub.2O.sub.3) 2 Mullite <1 278 688 None 698
(alumina-silica) 3 Cordierite Powder <1 284 694 None 708 4
Chromium Oxide 2.9 294 649 387 651 (C.sub.2O.sub.3) 279 637 407 625
5 Yttrium Oxide 10.3 290 685 371 707 (Y.sub.2O.sub.3) 6 Zirconia
(ZrO.sub.2, 0.4 286 641 None 632 Yttria Stabilized) 7 Lithium
Niobate 0.9 291 687 None 712 8 Niobium(V) Oxide 1.7 290 712 399 699
(Nb.sub.2O.sub.10) 9 Terbium Oxide 1.2 303 660 300 693
(Tb.sub.6O.sub.11) 10 Europium Oxide 5.6 306 665 353 667
(Eu.sub.2O.sub.3) 11 Samarium Oxide 1 301 669 337/ 703
(Sm.sub.2O.sub.3) 394 12 Tantalum(V) Oxide 0.3 288 679 331 680
(Ta.sub.2O.sub.10) 13 Ceria (CeO.sub.2), 280 687 382 718 7 wt % on
Mullite 14 Ceria (CeO.sub.2), 292 664 358 687 7 wt % & Pt 0.1
wt % on Mullite 15 Ceria (CeO.sub.2), 271 637 344 604 10 wt %
Citrate Process & Pt 0.1 wt % on Cordierite 16 Ceria
(CeO.sub.2) 0.9 289 703 403 728 99.9% 17 Ceria (CeO.sub.2) 90% 8.9
307 646 369 638 18 Ceria (CeO.sub.2) 96% 282 677 406 671 19 Ceria
(CeO.sub.2) 96% 290 703 None 709 After Plasma Spraying 20 Ceria
(CeO.sub.2) 99% 165 252 584 236 541 21 Praseodymium 275 653 410 662
Oxide (Pr.sub.6O.sub.11) 7 wt % on Mullite 22 Pr.sub.6O.sub.11 (10
wt % 284 692 390 708 Citrate Process & Pt 0.1 wt % on
Cordierite) 23 Pr.sub.6O.sub.11 4 297 593 285/ 572 399 24
Pr.sub.6O.sub.11 (96%) 3.7 303 612 445 581 25 Pr.sub.6O.sub.11
(HSA) 23 299 565 317/ 540 414 26 Pr.sub.6O.sub.11 (HSA) with 277
591 328 564 1 wt % Pt 27 Pr.sub.6O.sub.11 (96%) after 306 607 386
591 HVOF spraying 28 CeO.sub.2 (96%) & 293 693 336 693
Pr.sub.6O.sub.11 (96%) mixed in 70:30 ratio then plasma sprayed
Note: HSA = High Surface Area.
EXAMPLES
[0062] The emissions reduction system which constitutes the present
invention was tested and demonstrated by engine tests conducted
using a two-cycle diesel urban bus engine. The tests were conducted
in accordance with the U.S. Heavy Duty Transient Federal Test
Procedure (FTP), defined in 40 CFR, Part 86, Subpart N (1995),
incorporated herein by reference. The FTP outlines the specific
requirements for setting up, mapping and running a test engine for
the performance and emissions evaluation. A standardized 20-minute
transient test cycle was used, including rapid changes of speed and
torque which the engine must produce. The tests were run using 1994
emissions grade #2 diesel fuel (0.05 wt % S). Both cold start
(after an overnight equilibrium soak period) and hot start
transient tests were run and the data collected. Composite
results,of both cold and hot start transients were reported for
certification purposes. The composite results were calculated as a
weighted average of the cold start (1/7) and hot start (6/7)
transient results as required by the FTP. For some engine
configurations only hot start transient data were collected to
assess the relative contribution of the components comprising that
configuration. The key results included determination of brake
specific emissions (TPM, NO.sub.x, gas phase HC's and CO) and TPM
breakdowns (insolubles, VOF) expressed in g/BHP-h. Engine fuel
consumption was also determined (lb/BHP-h).
The Test Engine and Test Strategy
[0063] The test engine was a MY 1987 DDC 6V-92TA MUI (Serial #
6VF120195), which was a 6 cylinder, two-cycle diesel engine with a
V-configuration and having a displacement of 9.05 liters (552 cu.
in.). The engine was turbocharged, aftercooled and had mechanical
governing and mechanical unit fuel injectors. This engine is
typical of those used in a large proportion (about 40%) of urban
transit buses. The engine developed about 294 HP at 2100 RPM.
[0064] The engine was rebuilt and configured to give various test
cases for evaluation using the U.S. HD Transient Test (FTP). These
configurations (examples) included the components of the current
invention, as well as, other configurations (examples), including a
baseline, for comparative purposes. The exhaust system used for
testing and included the fittings for incorporation of the exhaust
catalyst (DOC) into the exhaust train. The DOC element was located,
for these tests, 6 feet downstream of the exhaust manifold. The
engine was run on a break-in cycle for at least 75 hours prior to
the tests.
[0065] Fuel injection timing for the engine was controlled by the
injector "height" setting and was varied in the tests from 1.460"
to 1.520" with the greater injector height being a more retarded
injection configuration. Most of the test runs were done at an
injection setting of 1.466". The throttle delay for the tests was
set at 0.594".
[0066] Ceramic Coatings: The catalyzed and non-catalyzed ceramic
coatings used in the examples below to demonstrate the invention
are described below:
[0067] 1. Non-catalyzed Insulative Coatings: These coatings,
referred to as "Coating A", are basic ceramic coatings applied to a
metal surface via the plasma spray process, as are well-known in
the art. The plasma sprayed coatings were applied by Engelhard
Surface Technologies in East Windsor, Conn. Such coatings consist
of:
[0068] a. A metal alloy bond coat (e.g. super alloy or 431 SS)
which acts as an interlayer between the metal surface of the part
and the ceramic layer. Particularly good results are obtained using
a well-known class of metal-chromium-aluminum-yttrium alloys,
commonly referred to as MCrAlY alloys, wherein the metal is
preferably nickel, cobalt, iron or combinations thereof. Such
alloys are well known in the art, as for example, in U.S. Pat. No.
5,384,200, cited above and incorporated herein by reference.
[0069] b. An yttria-stabilized zirconia ceramic layer (e.g. 7-20 wt
% yttria) deposited on top of the bond coat.
[0070] The coatings used for the present invention are relatively
thin (<15 mils thick) compared with those typically used in the
art (>60 mils thick). This gives them better thermo-mechanical
stability and thereby good durability to survive the temperature
gradients and stresses encountered in the combustion chamber of the
engine. The bond coat layer for these coatings is typically 3-5
mils thick with the remainder comprising the yttria stabilized
zirconia. The porosity of the yttria-stabilized zirconia can be
controlled by adjusting the plasma spray parameters, but typical
for this application is about 10% porosity.
[0071] 2. Non-catalyzed Mullite Insulative Coatings: These coatings
have the same basic structure as the above non-catalyzed insulative
coatings, except that the top-most 3-5 mils thickness is comprised
of an alumino-silicate (mullite) ceramic layer. Under this is the
yttria-stabilized zirconia layer and beneath this is the bond coat.
These coatings are therefore three layer coatings with an overall
thickness approximately the same as the above coatings without
mullite. The reason for the mullite top layer is to provide a very
chemically inert ceramic surface with high resistance to the
corrosive and aggressive materials encountered in the combustion
and exhaust (sulfur and sulfates, calcium, zinc, nitrogen oxides
and nitrates, chlorides, phosphorus, etc.). The "as sprayed"
mullite top layer is less porous (about 3% porosity) than the
zirconia layer, which is about 10% porosity. In the following
examples, the mullite non-catalyzed insulative coating is
designated "Coating B", and comprises the same first two layers as
Coating A, with a mullite top coat.
[0072] The non-catalyzed mullite insulative coating used in the
examples below and identified as "Coating B" comprised a nominal 4
mil thick bond coat of NiCrAlY alloy and a nominal 7 mil thick
ceramic layer of 8% yttria stabilized zirconia. On top of the
ceramic layer, a nominal 2 mil thick layer of mullite was
deposited.
[0073] 3. Catalyzed Insulative Coating: These coatings are
catalyzed versions of the above non-catalyzed or mullite
non-catalyzed insulative coatings. The specific catalyzed coating
used in the examples below consisted of the above Coating B applied
by a plasma spray process which had been impregnated with a
candidate metal oxide catalyst, and is designated "Coating C". In
this case the catalytic metal oxide was praseodymium oxide
(Pr.sub.6O.sub.11) which was applied to the Coating B surface (and
pores) via an aqueous solution of praseodymium nitrate precursor,
followed by drying and thermal decomposition of the nitrate to the
corresponding oxide. This was done in three applications to achieve
a praseodymium oxide loading equivalent, for example, to about 0.25
g of Pr.sub.6O.sub.11 on the surface of a coated piston crown. The
praseodymium oxide catalyst was chosen as a candidate because of
promising performance for burning dry carbon and to an extent SOF
in model TGA/DTA lab tests. However, for the purposes of the
present invention a variety of catalysts can be used, including
other base metal and rare earth oxides, as well as, precious
metals.
[0074] For all of the coatings, the surface of the metal part to be
coated was first pretreated by grit blasting. This resulted in a
clean, corrosion-free and roughened surface to insure good adhesion
and mechanical interlocking of the plasma sprayed bond coat to the
metal part.
[0075] Ceramic (catalyst) Coated Parts: The engine parts coated
with ceramic (catalyzed) coatings for these tests included the
following:
[0076] 1. Piston Crowns: The pistons for these tests were standard
commercial parts. The specific pistons initially had a compression
ratio of 15:1. The top rim was relieved to avoid interference with
the cylinder head coatings. The pistons were of a open bowl, low
swirl profile. The piston crowns were plasma spray coated with
Coatings A or B. Application of the plasma spray coating changed
the dimensions and bowl volume of the piston crown slightly, thus
the coated pistons were calculated to have a compression ratio of
about 17:1. One set of Coating B coated pistons had the catalyst
(praseodymium oxide) applied to make a set of Coating C coated
pistons. The catalyst constituted a very thin coating and thus did
not change the relative compression ratio of the pistons.
[0077] 2. Engine Heads & Valves: The coatings were applied by
the plasma spray process to the area of the engine cylinder heads
enclosed by the combustion chambers ("fire deck"). During the
coating process the exhaust valves were installed in the heads and
seated. In this way both the "fire deck" region and the exhaust
valve faces could be plasma spray coated simultaneously with
Coatings A or B. One set of heads and valves which had been coated
with Coating B had the catalyst (praseodymium oxide) coating
applied to make a set of Coating C coated heads and valves. The
exhaust ports on these heads had also been plasma spray coated with
Coating B but did not have the praseodymium coating applied in that
area of the head.
[0078] 3. Exhaust Manifolds and Pre-Turbo Pipes: The plasma spray
process is a line-of-sight process and as such it can be used to
coat a part as long as the area of the part to be coated is
accessible to the spray stream of the plasma spray gun. It is thus
possible to spray the interior of parts, such at tubes and pipes,
as long as the inner surface is accessible. Larger tubes (e.g.
>about 2" id) can be coated through each end for lengths of up
to 1-2 ft. Smaller tubes can be coated on the inside using special
mini-guns or mini-torches. The plasma spray process can not;
however, be used to coat around acute corners or in areas which can
not be made accessible. The exhaust manifolds and pre-turbo
connecting pipes for the DDC 6V92TA engine were large enough in
diameter and straight enough that it was possible to coat the inner
surface to a significant extent (about 75%) with the plasma spray
coating of Coating B. One set of exhaust manifolds and pre-turbo
pipes coated with Coating B had the catalyst (praseodymium oxide)
also applied to make a set coated on the inner surface with Coating
C.
[0079] The Exhaust Catalyst (DOC): A diesel oxidation catalytic
converter is comprised of a catalyst coating, typically composed of
a mixture of catalytically active base metal oxides and optional
precious metals applied to the flow channel surfaces of a suitable
monolithic honeycomb substrate. This substrate with the catalyst
washcoat applied is contained within a metal canister (the can)
designed to hold the substrate in place and allow for the flow of
the exhaust gases from the engine to pass over the catalyst.
[0080] Current Production DOC (DOC-A): A current production
catalyst for this application, designated "DOC-A" was obtained for
comparative testing. This catalyst was coated onto an extruded
cordierite ceramic honeycomb catalyst substrate measuring 9.5"
dia..times.6" long. The cell spacing was 200 cpsi, with a wall
thickness of 12 mil. The geometric surface area was 13 m.sup.2. The
total catalyst washcoat loading was 2.95 g/in.sup.3, for a total of
1254 g. The catalyst washcoat consisted of a bottom coat of 1.00
g/in.sup.3 highly milled high surface area (150 m.sup.2/g) alumina,
with a pore volume of 0.5 cc/g. The topcoat included the same
alumina (1.05 g/in.sup.3) plus high surface area (150 m.sup.2/g)
aluminum-stabilized ceria (0.90 g/in.sup.3), having a pore volume
of 0.1 cc/g, with platinum at 10 g/ft.sup.3 distributed equally on
the alumina and ceria in the topcoat. This catalyst was broken-in
(de-greened) on the engine overnight before testing.
[0081] Improved DOC (DOC-B): The improved catalyst used for this
invention, designated "DOC-B" is comprised of high surface area,
high pore volume bulk metal oxides of aluminum and cerium which in
turn contain platinum impregnated on their surfaces. This catalyst
was coated onto a metallic wrapped foil honeycomb substrate with
the dimensions of 10.5" dia..times.6" long. The cell spacing was
310 cpsi, with a wall thickness of 4 mil. The geometric surface
area was 27 m.sup.2. Such substrates are manufactured in Sweden by
EcoCat, a subsidiary of Sandvik Steel, and are made in accordance
with one or both of U.S. Pat. Nos. 4,633,936 and 5,085,268, both of
which are incorporated herein by reference. As with DOC-A, the
total catalyst loading was 2.95 g/in.sup.3, but in this case for a
total of 1534 g. The catalyst coating consisted of: [1] A thin
"etch coat" of highly milled high surface area (150 m.sup.2/g)
alumina to insure good adherence of the catalyst washcoat to the
metallic honeycomb surface; [2] An undercoat consisting of high
surface area alumina mixed with high surface area
aluminum-stabilized ceria (150 m.sup.2/g) in a ratio of 34:66 by
weight; and [3] A topcoat consisting of very high surface area and
pore volume alumina (250-300 m.sup.2/g) mixed with high surface
area aluminum-stabilized ceria in a ratio of 54:46 by weight, and
having platinum impregnated onto the alumina and ceria at about
0.3% by weight of the topcoat loading with the platinum distributed
equally on the alumina and ceria in the top coat. This catalyst
DOC-B was prepared and coated onto the metallic honeycomb
substrate. The loading levels for this catalyst were as
follows:
[0082] 1. Etch coat of 0.20 g/in.sup.3 highly milled high surface
area alumina
[0083] 2. Bottom coat of 0.60 g/in.sup.3 of the high surface area
alumina plus 0.20 g/in.sup.3 of the high surface area ceria.
[0084] 3. Topcoat of 1.05 g/in.sup.3 of the very high surface area
alumina plus 0.90 g/in.sup.3 of the high surface area ceria and
containing 10 g/ft.sup.3 platinum distributed about 90% on the
alumina and about 10% on the ceria in the topcoat.
[0085] The catalyst coated substrate was then put into a canister
and broken in (de-greening) on a diesel engine for 40 hours prior
to the tests.
[0086] Test Configurations: The DDC 6V92TA MUI engine was set up
with various configurations of ceramic (catalyzed) coatings and
with and without aftertreatment catalysts (DOC's) to demonstrate
the current invention and assess the relative contributions of the
various components of the invention. The various configurations
included: [A] Coated or uncoated piston crowns; [B] Coated or
uncoated heads & valves; [C] Coated and uncoated exhaust
manifolds and pre-turbo exhaust pipes; [D] Insulated exhaust pipes
between the turbo-outlet and the exhaust catalyst (DOC); [E]
Presence or absence of the exhaust catalyst (DOC) in the exhaust
stream and [F] Type of exhaust catalyst (DOC), improved or
production. The configurations (examples) tested are given in TABLE
II which shows 9 Examples plus baselines. The examples are
summarized as follows:
Example 1
[0087] The engine configuration which includes catalyzed Coating C
on the piston crowns, heads and valves and in the exhaust manifolds
and pre-turbo pipes, plus insulated exhaust pipes between the
turbo-outlet and the DOC and with the improved diesel oxidation
catalyst DOC-B.
Example 2
[0088] Same as Example 1 but using the production exhaust catalyst
DOC-A instead of the improved DOC-B.
Example 3
[0089] Same as Example 1 but with stock, uncoated heads and
valves.
Example 4
[0090] Same as Example 1 but without the insulation on the exhaust
pipes between the turbo-outlet and the DOC.
Example 5
[0091] Same as Example 1 but with stock, uncoated exhaust manifolds
and pre-turbo exhaust pipes.
Example 6
[0092] Same as Example 5 but without a diesel oxidation catalyst
(DOC) in the exhaust stream.
Example 7
[0093] The engine configuration which includes non-catalyzed
Coating B on the piston crowns, heads and valves and in the exhaust
manifolds and pre-turbo pipes, plus the improved DOC-B.
Example 8
[0094] Same as Example 7 but with stock, uncoated exhaust manifolds
and pre-turbo exhaust pipes.
Example 9
[0095] Same as Example 8 but without a diesel oxidation catalyst
(DOC) in the exhaust stream.
Baseline
[0096] The current engine rebuilt to its baseline condition for a
1979 DDC 6V92TA MUI and without coatings or a diesel oxidation
catalyst (DOC) in the exhaust stream.
[0097] Baseline ('94)
[0098] This is included for comparison purposes and refers to DDC
6V92TA MUI baseline engines tested in 1994. The engine was rebuilt
to a 350 HP rating and again to a 277 HP rating with rebuild kits
representative of a MY 1985 engine.
[0099] Test Results: The results of transient emissions tests for
the configurations corresponding to the Examples listed in TABLE II
above are given in TABLE III. The first column in TABLE III refers
to the engine test configuration (Examples). The second column
gives the throttle delay setting in inches (0.594" in each case
here). The third column gives the injection timing used for the
individual test run or set of test runs. The next column designates
the type of HD transient test (cold start, hot start or composite).
The next four columns give the brake specific emissions for gas
phase HC's, CO, NO.sub.x and TPM, respectively. The next two
columns give the breakdown of the TPM into insolubles and VOF
components. The final column gives the brake specific fuel
consumption for the test run.
[0100] In addition to the data in TABLE III the results for hot
start transient tests for the examples are shown graphically in the
bar chart in FIG. 2. These hot start results are all for the 1.466"
fuel injection timing setting. FIG. 2 gives the average particulate
emissions for each example, either as TPM or particulate breakdowns
into insoluble and VOF portions, where available. As will be
discussed further below in regard to the specific examples,
catalyzed coatings in the exhaust train, particularly in the
exhaust manifold and pre-turbo pipes, are particularly effective
for reducing the insoluble soot component of the total particulate
matter (TPM) emissions.
[0101] Results for Example 1, an example of the present invention,
are given in TABLE III for injection timing settings from 1.460" to
1.520" (most retarded injection). As can be seen the 0.1 g/BHP-h
TPM standard was readily met for injection settings of 1.460",
1.466" and 1.475". As the injection was progressively retarded the
TPM emission level increased, but only slightly, and TPM emissions
from 0.076 to 0.084 g/BHP-h were achieved. The technology used for
the Example 1 configuration achieved dramatically lower TPM
emissions compared with the baseline engine configurations with
timing setting of 1.466" from 1994 tests. The baseline engines gave
TPM emissions of at least 0.200 g/BHP-h. In the current tests
retarding the injection timing was accompanied by reduction in
NO.sub.x emissions. For Example 1 with injection timing settings of
1.466" and 1.475" lower NO.sub.x emissions were achieved (10.11 and
8.97 g/BHP-h, respectively) than for the 194 baseline engine
configurations (10.3 to 11.7 g/BHP-h) while maintaining TPM
emissions well below the 0.1 g/BHP-h standard. The level of
NO.sub.x emission could be reduced even further to 5.07 g/BHP-h
(essentially the 1994 NO.sub.x emission standard) by additional
retarding of injection timing to 1.520". This was accompanied by a
level of TPM emissions (0.141 g/BHP-h) which exceeded the 0.1
g/BHP-h standard. However, it was still lower than the TPM
emissions level from the baseline engine (>0.2 g/BHP-h). The
relationship of both TPM and NO.sub.x emissions levels for Example
1 as a function of injection timing setting is shown graphically in
FIG. 4. Example 1 with injection timing setting in the range of
1.466" to 1.475" appears to best fulfill the goals of meeting the
0.1 g/BHP-h TPM emissions with good NO.sub.x emissions levels. The
TPM breakdown for Example 1 (1.466") showed 0.049 g/BHP-h
insolubles and 0.031 g/BHP-h VOF showing that this configuration
gave substantial reduction in both fractions compared with the
baseline results. The brake specific fuel consumption for Example 1
was found to be comparable or lower than that of the '94 baseline
engine configurations showing that implementation of the technology
of this invention is not accompanied by a fuel efficiency
penalty.
[0102] The results for Example 2 show that when the larger volume
(8.5 liter), improved DOC-B of this invention (on 10.5"
dia..times.6" long 310 cpsi metallic substrate) was replaced with
the smaller volume (7.0 liter) production DOC-A (on 9.5"
dia..times.6" long 200 cpsi ceramic substrate), the TPM emissions
increased to 0.116 g/BHP-h and the 0.1 g/BHP-h standard could not
be achieved. It, therefore, appears that the improved catalyst
exhibited a TPM emissions performance advantage of 0.037 g/BHP-h
compared with the production catalyst. Although the TPM breakdown
for Example 2 has not yet been determined it is expected that the
improved catalyst gives higher VOF conversion than the production
catalyst because of its improved washcoat properties and larger
catalyst volume. The NO.sub.x emissions levels with the two
catalysts were comparable. Gas phase HC emissions were about 20%
lower with the improved catalyst. However, CO emissions were
slightly higher, possibly due to partial oxidation of the greater
relative amount of VOF and gas phase HC's converted by the improved
catalyst. The brake specific fuel consumption for Example 2 was
comparable with Example 1. These results show that the improved
DOC-B and larger catalyst volume are key components for the best
performance of the invention.
[0103] The results for Example 3 show that the engine configuration
with the uncoated heads and valves run with an injection timing
setting of 1.460" exhibited slightly higher levels of emissions
than for Example 1 with coated heads and valves run at the same
injection timing setting. The TPM emissions for Example 3 were ca
0.014 g/BHP-h higher than for Example 1. The composite NO.sub.x
emission levels of Examples 3 and 1 were comparable for the 1.46"
timing setting. However, gas phase HC's and CO were slightly higher
for Example 3 than for Example 1. This indicated that the
configuration with coated heads and valves was better for emissions
performance, but the 0.1 g/BHP-h TPM emission goal could still be
met with uncoated heads and valves and the configuration of Example
3 would be simpler and lower in cost from a manufacturing point of
view. Thus Example 3 might be considered useful under some
circumstances. The composite brake specific fuel consumption for
Example 3 appears to be slightly lower than for Example 1, but this
could be within the experimental error of the measurement.
[0104] For Example 4 the insulation was removed from the exhaust
pipes between the turbo-outlet and the diesel oxidation
catalyst(DOC). The cold start TPM emission result for Example 4 was
about 15% higher than the cold start results for Example 1 showing
the positive effect of the insulation on reducing emissions for the
cold start test condition. The average hot start TPM emissions for
Example 4 were also found to be slightly higher showing that the
insulation has a positive overall effect on particulate emissions.
TPM breakdowns show slightly higher insolubles and slightly lower
VOF for Example 4 compared with Example 1. The reason for this is
not fully understood, but could be related to inhibition of the
development of insoluble related components in the hotter insulated
exhaust pipes (Examples 1 & 3) compared with the un-insulated
exhaust pipes (Example 4). The results for Example 4 show that it
was not necessary to have the insulation in place to meet the 0.1
g/BHP-h TPM emission standard, but its beneficial effect can give a
slightly greater delta TPM to allow for greater DF and better
operation under cold ambient conditions. Composite NO.sub.x
emission levels were comparable for Example 4 and Example 1.
Surprisingly, gas phase HC and CO emissions were slightly lower
without the insulation. The fuel consumption was comparable both
with and without the insulation.
[0105] For Example 5 the Coating C coated exhaust manifolds and
pre-turbo exhaust pipes of Example 1 were replaced with stock,
uncoated manifolds and pre-turbo exhaust pipes. The test results
showed that the TPM emission level for this configuration was
substantially higher (0.121 g/BHP-h) than when the coated manifolds
and pipes were used (0.080 g/BHP-h). The TPM performance advantage
with the coated manifolds and pipes appears to be 0.041 g/BHP-h.
Comparison of the TPM breakdowns for Examples 1 & 5 shows that
the key contribution of the use of Coating C coated exhaust
manifolds and pre-turbo pipes is in reduction in the insoluble
portion of the particulates. The level of NO.sub.x emissions for
Example 5 was comparable with that of Example 1 as was the brake
specific fuel consumption levels. For Example 5 gas phase HC
emissions were slightly lower and CO emissions were slightly higher
than for Example 1. These results show that the coated exhaust
manifolds and pre-turbo pipes were an important component
contributing to the performance of the system which constitutes
this invention. This was a surprising result in that intuitively
one would not consider the inner surface of the exhaust manifolds
and pipes as potentially having much effect on emissions
considering the relatively low surface area. However, there are
considerable changes in flow direction as the exhaust passes
through the manifolds and pipes and the momentum of components of
the particulates (VOF aerosols, dry carbon particles and carbon
particles with adsorbed VOF) might cause them to impact on the
inner walls at pipe bends and angle changes, thus allowing for
interaction with the ceramic (catalyzed) coatings and burning.
[0106] For Example 6 the configuration consisted of Coating C
coatings in the combustion chamber only. The coated exhaust
manifolds and pre-turbo exhaust pipes were replaced with stock,
uncoated manifolds and pipes as with Example 5, plus the diesel
oxidation catalyst (DOC) was removed. The test results showed
relatively high TPM emissions level for this configuration (0.188
g/BHP-h), compared with Example 5 and especially compared with
Example 1. Compared with Example 5, it appeared that the TPM
emissions performance advantage with the diesel oxidation catalyst
(DOC) was 0.067 g/BHP-h. Compared with Example 1, it appears that
the TPM emissions performance advantage for the coated exhaust
manifolds and pre-turbo pipes plus the diesel oxidation catalyst
was a total of ca 0.108 g/BHP-h. The TPM breakdown for Example 6
shows that nearly all of the increased TPM emission compared with
Example 5 was due to greater VOF contribution. This is consistent
with the high VOF removal performance observed with the DOC
present. Comparing Example 6 with the baseline engine
configurations indicates that the TPM emissions performance
advantage for the combustion chamber coatings alone was about
0.012-0.013 g/BHP-h. The emission levels of NO.sub.x and brake
specific fuel consumption for Example 6 were comparable with those
for Examples 1-5 at the equivalent injection timing setting.
[0107] The results of these tests show that substantial reduction
in TPM emissions for a 2-cycle DDC 6V92TA MUI bus engine can be
dramatically reduced and the 0.1 g/BHP-h TPM standard can be met
with the technology described in this invention record. The key
components of the invention appear to be the improved diesel
oxidation catalyst (DOC), the ceramic (catalyzed) coatings on the
inner surfaces of the exhaust manifolds and pre-turbo exhaust pipes
and the ceramic (catalyzed) coatings on the combustion chamber
surfaces (piston crowns, heads and valves). Each of these
components has its own incremental effect on reducing the dry
carbon and VOF portions of the TPM emissions. We believe that the
combination of components used to achieve these results provides a
system capable of significant reduction in TPM emissions from
diesel engines. This coupled with the possibilities of reducing
NO.sub.x via injection retard without great increases in TPM
emissions gives a system to address the two key emissions
challenges for diesel engines, namely TPM and NO.sub.x.
NO.sub.x Benefits
[0108] The NO.sub.x and TPM trade-off is well known in industry.
Each diesel engine model will produce a set of NO.sub.x and TPM
emissions depending on the engine parameters. This set of NO.sub.x
and TPM values can be plotted as shown in FIG. 5, which is a
representation of the values reported for Example 1, in Table III,
compared to the Baseline (350 HP) values. Engineers calibrate
engines to achieve the desired balance of NO.sub.x-TPM emissions by
adjusting various engine operating parameters. The primary engine
parameter is the timing of fuel injection, which is the time during
the cycle at which diesel fuel begins to enter the combustion
chamber. Retarding the injection (i.e., starting injection later in
the cycle) has the well-known effect of reducing NO.sub.x at the
expense of increasing TPM. The results depicted in FIG. 1 show that
the system of the present invention can significantly reduce
NO.sub.x levels with only a relatively small increase in TPM
emissions.
2TABLE II Description of Emissions Control Systems Tested Pre-Turbo
Insulated Head & Manifold & Post-Turbo Piston Valve Pipes
Exhaust Coating Coating Coating Pipes Catalyst Example 1 Coating C
Coating C Coating C Yes DOC-B Example 2 Coating C Coating C Coating
C Yes DOC-A Example 3 Coating C None Coating C Yes DOC-B Example 4
Coating C Coating C Coating C No DOC-B Example 5 Coating C Coating
C None Yes DOC-B Example 6 Coating C Coating C None Yes None
Example 7 Coating B Coating B Coating B No DOC-B Example 8 Coating
B Coating B None No DOC-B Example 9 Coating B Coating B None No
None Baseline None None None None None Baseline None None None None
None ('94) (277 & 350 HP)
[0109]
3TABLE III Summary of HD Transient Engine Emissions Tests Throttle
Injection Transient Measured Emissions (g/BHP-h) TPM Breakdown BSFC
Delay (") Timing (") Test Type HC CO NOx TPM Insolubles VOF
(lb/BHP-h) Ex. 1 0.594 1.46 Cold-3 0.172 0.401 10.65 0.076 0.458
Hot-4 0.161 0.254 10.93 0.077 0.058 0.019 0.44 Hot-5 0.162 0.275
11.06 0.079 0.438 Comp. 0.162 0.274 10.89 0.077 0.442 0.594 1.466
Cold-4 0.219 0.524 9.94 0.076 0.043 0.033 0.453 Hot-8 0.154 0.422
10.14 0.08 0.049 0.031 0.436 Hot-9 0.173 0.431 10.09 0.081 0.049
0.032 0.431 Comp. 0.163 0.436 10.11 0.08 0.049 0.031 0.438 0.594
1.475 Hot-6 0.17 0.309 8.97 0.089 0.054 0.03 0.442 0.594 1.52
Hot-19 0.1 0.31 5.07 0.141 0.122 0.019 0.449 Ex. 2 0.594 1.466
Hot-15 0.201 0.41 10.32 0.116 0.081 0.035 0.439 Ex. 3 0.594 1.46
Cold-2 0.215 0.481 10.3 0.089 0.042 0.047 0.457 Hot-2 0.203 0.422
10.91 0.092 0.046 0.046 0.425 Hot-3 0.208 0.473 10.89 0.092 0.045
0.047 0.43 Comp. 0.205 0.43 10.83 0.091 0.046 0.046 0.429 Ex. 4
0.594 1.466 Cold-5 0.17 0.48 9.8 0.083 0.457 Hot-16 0.154 0.359
10.17 0.08 0.054 0.026 0.433 Hot-20 0.15 0.51 10.22 0.09 0.065
0.025 0.441 Hot-21 0.15 0.49 10.21 0.089 0.439 Comp. 0.15 0.51
10.16 0.09 0.443 Ex. 5 0.594 1.466 Hot-17 0.135 0.509 10.1 0.121
0.094 0.027 0.437 Ex. 6 0.594 1.466 Hot-18 0.358 1.298 10.2 0.188
0.108 0.08 0.439 Ex. 7 0.594 1.46 Hot-31 10.7 0.097 0.079 0.018
1.47 Hot-29 0.16 0.59 9.87 0.102 0.083 0.019 0.436 1.485 Hot-32 7.8
0.14 0.128 0.015 1.5 Hot-33 6.5 0.18 0.157 0.019 Ex. 8 0.594 1.47
Hot-40 0.185 0.81 9.72 0.116 0.0974 0.0186 0.441 Ex. 9 0.594 1.47
Hot-39 0.413 1.481 9.84 0.191 0.441 Baseline 0.594 1.466 Cold Hot
Hot Comp. Baseline 0.594 1.466 Cold 0.5 0.9 10.3 0.2 0.07 0.13
0.441 ('94) Hot 0.5 1.6 10.3 0.21 0.08 0.13 0.422 350 HP Comp. 0.5
1.5 10.3 0.21 0.08 0.13 0.424 Baseline 0.636" 1.466" Cold 0.5 1.7
11.6 0.24 0.494 ('94) Hot 0.5 0.8 11.7 0.2 0.451 277 HP Comp. 0.5
0.9 11.7 0.2 0.457
Example 10
[0110] This example illustrates the incremental improvement of
emissions performance (particularly particulate emissions) obtained
with in-cylinder thermal barrier coating and more so with catalyzed
thermal barrier coatings. This effect was shown by engine emissions
testing using the U.S. Heavy Duty Transient Test Procedure, as
previously discussed.
[0111] A 1986 DDC 6V92-TA MUI engine was rebuilt using standard
rebuild kit parts. This engine was built to a 294 HP configuration,
and fitted with 9F-80 fuel injectors and 17:1 compression ratio
pistons. The engine was run with a fuel injection timing setting of
1.460" (injector height in inches) and a throttle delay of 0.636"
(throttle actuator adjustment in inches). The engine was equipped
with a standard exhaust system and no after treatment catalyst was
used. In this way the baseline engine-out emissions of a standard
engine rebuild were evaluated. The engine was broken in for the
required 100 hrs and then tested for emissions in hot start
transient test runs.
[0112] Next the engine was rebuilt again, but was equipped with
uncatalyzed thermal barrier coatings (the above "Coating B") on all
piston crowns, the head firedecks and exhaust valve faces. (The
inlet valves, which are located in the cylinder side walls of this
engine rather than in the head, were not coated.) This engine, also
294 HP, was fitted with 9F-80 fuel injectors and the compression
ratio with the coatings was 17:1 as with the standard rebuild,
above. This coated engine was also run with a fuel injection timing
setting of 1.460" and a throttle delay of 0.636". This engine was
run with the same standard exhaust system as used for the standard
rebuild, above. the engine was broken in for 100 hrs and tested for
emissions performance in hot start transients.
[0113] Next the engine was rebuilt again, but this time it was
equipped with catalyzed thermal barrier coatings (the above
"Coating C") on all piston crowns, the firedecks and valves. The
catalyst was Pr.sub.6O.sub.11 as described previously for "Coating
C". This engine was built to a 294 HP configuration with 9F-80 fuel
injectors and compression ratio of 17:1 as with the standard
rebuild and uncatalyzed thermal barrier coating rebuild, discussed
above. The engine was equipped with the same standard exhaust train
used for the standard build and uncatalyzed thermal barrier coated
engines, above. This engine was broken in and tested for emissions
performance in hot start transient tests.
[0114] As is apparent for the three engine builds described above,
as much care was taken as possible to have the builds be identical
and the engines run in the same way with the only key variable
being the in-cylinder, combustion chamber surface coatings.
[0115] The emissions results of the hot start transient tests for
the three engine builds described above are shown in TABLE IV
below:
4TABLE IV Engine Configuration Emissions (g/BHP-h) and Run No. HC
CO NOx PM Standard Rebuild 1 0.50 1.45 11.44 0.193 2 0.50 1.47
11.50 0.194 3 0.51 1.57 11.28 0.200 4 0.51 1.56 11.15 0.195 Avg.
0.51 1.51 11.34 0.196 .+-. 0.003 Uncatalyzed Coating ("Coating B")
1 0.58 1.21 11.05 0.185 2 0.58 1.23 10.90 0.188 3 0.58 1.17 11.10
0.185 Avg. 0.58 1.21 11.05 0.186 .+-. 0.001 Catalyzed Coating
("Coating C") 1 0.57 1.04 10.70 0.170 2 0.58 1.13 10.70 0.176 Avg.
0.58 1.09 10.70 0.173 .+-. 0.003
[0116] As can be seen the average particulate emissions level of
the engine coated with the uncatalyzed coating ("Coating B") was
lower than the average particulate emissions level of the standard
rebuild engine by 0.010 g/BHP-h. Furthermore, the average
particulate emissions for the engine coated with the catalyzed
coating ("Coating C") was lower than average particulate emissions
for the engine coated with the uncatalyzed coating ("Coating B") by
0.013 g/BHP-h. The emissions ranges (average .+-.2 sigma) for each
of these engines do not overlap and thus are significantly
different. These results show that the catalyzed coatings give a
real and distinct effect in reducing the particulate emissions for
the diesel engine.
Example 11
Stainless Steel Bond Coats
[0117] Experiments were conducted to test thermal barrier coatings
for resistance to rapid thermal cycling using various bond coats
for bonding a standard 7% yttria stabilized zirconia ceramic top
coat to an aluminum substrate, simulating the conditions to which
an aluminum piston would be exposed in operation in a diesel
engine. Based on manufactures literature, the ceramic is considered
to have a coefficient of thermal expansion (CTE) of about 7.6,
while the aluminum substrate has a CTE of about 23.0. A "thermal
cycle machine" was developed to perform the tests. The device was
made to move a test sample between a heat source and a cooling flow
for a programmable period of time in order to simulate typical
engine run conditions. In the case of the evaluation of aluminum
piston bond coats, a three-inch diameter disk sample is used. This
is coated on one face with the bond coat and ceramic top coat
system for the purpose of seeing if the top coat will remain
adherent to the bond layer. The heat source is a large propane air
burner whose flame impinges directly on the coated disk. Cooling is
done by a high volume compressed air flow. In operation the test
disk is shuttled back and forth between the heat and cool locations
for timed durations that allow a range of 900.degree. F. (high) to
200.degree. F. (low) to be reached. A cycle was developed in which
a 2.5 minute dwell in each station achieved the target
temperatures.
[0118] Four bond materials were chosen for this experiment. In each
case, the bond coat was applied by plasma spraying the metal in
powder form onto the aluminum disk. The materials tested were
Nichrome.RTM. (80%Ni/20%Cr) alloy (Metco 43C), having a CTE of
about 7.3; 95% nickel/5% aluminum alloy (metco 480), having a CTE
of about 10; NiCrAlY (Praxair 211), having a CTE of about 13.0; and
type 431 martensitic stainless steel (Metco 42C), having a CTE of
about 6.6. For each sample, the number of thermal cycles to failure
was measured. For purposes of this test, coating failure is
considered when the top coat begins chipping and separating from
the bond coat. The Nichrome sample failed after 47 cycles; the
nickel/aluminum sample after 83 cycles; the NiCrAlY sample after 23
cycles; and the stainless steel sample still had not failed after
123 cycles, at which time the test was discontinued. This showed
the distinct improvement obtained by using a stainless steel bond
coat under simulated engine conditions to bond a ceramic coat to
aluminum, particularly a "low-shrink" martensitic steel such as
type 431 SS.
* * * * *