Geometric Parameters

Influencing Flow in an
Andreas Maier
Research Engineer
Axisymmetric IC Engine Inlet Port
Terry H. Sheldrake
Research and Development Manager
Assembly: Part II—Parametric
Dennis Wilcock Variation of Valve Geometry
Emeritus Professor of Engineering
The influence of inlet valve geometry on IC engine port flow characteristics was investi-
School of Computing, gated systematically by varying the key valve geometric parameters of fillet radius, cone
Engineering and Technology, angle, and seat angle. The analysis of flow through the port was measured for 36 valves
University of Sunderland, having geometries determined by R/D ⫽ 0.20– 0.30, cone angle 0–20 deg and seat angle
Sunderland, United Kingdom 30–45 deg. Discharge coefficients were measured to gain an understanding of the general
flow performance. These were used in conjunction with heat transfer data to gain a clear
understanding of the state of the boundary layer on the valve and port seats for varying
lifts and pressure drops. This established the presence or otherwise of the four flow
regimes within the valve passage. The results demonstrate that the cone and seat angle
have major effects on valve flow performance and enable engine designers to choose
more readily the valve characteristics they desire. 关S0098-2202共00兲00404-1兴

Introduction revealed that a wide overlap gave better flow performance. Seat
angle variations were examined using four seat geometries be-
In an internal combustion engine, volumetric efficiency, and
therefore power output, could be increased with improvements to tween 0 and 60 deg to quantify its effect on valve flow character-
the design of the inlet port and particularly the inlet valve. Part I istics. The 45 deg seat angle valve showed a distinct transition
of this paper clearly identified the sensitivity of valve flow char- between the four flow regimes observed for various valve lifts.
acteristics to small changes in valve geometry by presenting data The valve with a 30 deg seat angle showed the transition between
for two valve geometries varying in only their valve seat angle. the different flow pattern less clearly.
The present work conducted a parametric investigation into the In the subsequent investigations of Wood et al. 关4兴 and Kastner
influence of valve geometry on valve flow characteristics by test- et al. 关5兴, attempts were made to reduce the tendency of the flow
ing a range of inlet valve geometries. to separate from the sealing faces of the valve passage.
Previous studies paid much attention to testing specific valve Wood et al. 关4兴 described various attempts to reduce the flow
geometries with respect to their flow performance. Luke 关1兴 was resistance by rounding the sharp corners of the valve and seat
the first to indicate that the geometry of valve and port plays a faces, by varying the radius of the port elbow and by varying the
significant role in determining the performance characteristics of fillet radius of the valve head in order to approach the perfor-
the induction system. Although no geometrical details were given, mance of a perfect venturi nozzle. Kastner et al. 关5兴 also reported
the so-called ‘‘conical-seated’’ valve showed a performance supe- on a range of modifications to the valve and port to improve its
rior to that of the ‘‘flat-seated’’ valve. Lewis and Nutting 关2兴 flow performance. The modifications included a tapered port to
investigated the arrangement of valves in pairs as compared with avoid abrupt area changes reducing wall separation and shaping
single ones and valves of different sizes. Aspects of valve geom- the passage to a converging-diverging nozzle. Although these at-
etry and its influence on the performance of the valve/port assem- tempts produced considerable improvements in discharge coeffi-
bly were extensively investigated by Tanaka 关3兴. The study fo- cient C D , which were however mostly in the low lift range
cused on the influence of fillet radius, valve head angle, seat (L/D⫽0⫺0.10) with values exceeding unity, the modifications
width, and seat angle. The valve models were tested under steady resulted in complex geometrical configurations, which might
flow conditions in an axisymmetric port/cylinder arrangement. prove impractical for production engine designs. Annand 关6兴 fur-
The effects of varying parameters defining the valve geometry ther investigated the effects of seat width, fillet radius, and seat
were examined by scrutinizing the variation in mass flow rate angle on discharge coefficients C D with an isolated valve and
through the valve passage for different lift conditions at a fixed results mainly agreed with the findings of Tanaka 关3兴. Valves with
pressure differential across the port. Generally, the fillet radius did a 45 deg seat angle clearly showed the transition between the
not have a significant effect on the flow quantity, but a large fillet different flow regimes whereas the 30 deg seat indicated the tran-
radius was found to have a rather bad influence on flow charac- sition points less distinctly. However, Annand recommended tak-
teristics. The effect of valve cone angle was investigated on the ing into account mechanical considerations proposed by Pope 关7兴
basis of four valve cone angle geometries ranging from 0 to 45 for the final choice of the appropriate seat angle.
deg. The study showed a pronounced effect of the upper valve An extensive study of the flow structure at the valve exit plane
face corner on flow separation, which was delayed to higher lifts and of the flow field in the cylinder was carried out by Bicen and
for a larger valve cone angle. The investigation of the seat width Whitelaw 关8兴 and Vafidis and Whitelaw 关9兴 who measured dis-
charge coefficients and the detailed velocity distribution using
Contributed by the Fluids Engineering Division for publication in the JOURNAL LDA for various valve geometries.
OF FLUIDS ENGINEERING. Manuscript received by the Fluids Engineering Division
January 11, 2000; revised manuscript received June 6, 2000. Associate Technical Vafidis and Whitelaw 关9兴 investigated a 45 deg seat angle valve
Editor: D. Williams. with and without rounded inner corners of the valve and seat

658 Õ Vol. 122, DECEMBER 2000 Copyright © 2000 by ASME Transactions of the ASME

Downloaded 10 Jul 2010 to 129.173.72.87. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
 sealing faces upon their effect on flow transition. The results sug-
gested that only the valve with sharp corners exhibited the four
flow regimes.
Bicen et al. 关10兴 presented a summary of the work reported by
Bicen and Whitelaw 关8兴 and Vafidis and Whitelaw 关9兴 in compar-
ing discharge coefficients for the 45 and 60 deg seat angle valves
and additionally the detailed flow characteristic at the exit of the
valve passage. The results showed that the performance of the 45
deg seat was generally superior to the 60 deg seat over the range
of lifts tested (L/D⬍0.26).
Gosman and Ahmed 关11兴 subsequently reported on the mea-
surements of discharge coefficients, mean velocity and turbulent
stress fields obtained by Hot-Wire Anemometry 共HWA兲 at various
valve lifts in an axisymmetric valve/port assembly with a valve
geometry 共45 deg seat angle and rounded corners兲 similar to the
one investigated by Bicen et al. 关10兴.
Weclas et al. 关12兴 presented a comprehensive investigation into Fig. 1 Valve geometry
the flow separation in the inlet valve passage. Their study concen-
trated on the flow characteristics of two intake port geometries: an
Those criteria clearly identify that practical valve design may dif-
idealized directed port of a research engine and a helical port of a
fer from the recommendations for optimum flow performance. A
direct injection Diesel engine. Furthermore the investigation con-
parametric study was therefore designed to investigate the behav-
sidered the effect of a variation in valve geometry. Two valves
ior of valve flow with the limits of the geometry variation taken
with a 45 deg valve seat angle that differed in the fillet radius and
from general design criteria for the inlet valve within the automo-
the valve head angle were examined for their discharge character-
tive industry 共TRW Automotive Systems Ltd 关13兴兲. The three
istics in the directed port.
most influential valve parameters, which are illustrated in Fig. 1,
Those studies identified that the inlet valve flow is clearly de-
determining valve flow were varied as follows:
pendent on valve geometry with the general flow characterized by
the four flow regimes 共Tanaka 关3兴兲. The occurrence of the four
Geometrical Range Incremental
flow regimes and the transition points were found to be influenced Parameter Change
by the valve and port geometry. The investigations showed that
with certain geometrical modifications the transition points can be Fillet Radius R/D 0.20–0.30 0.05
altered leading to improvements in the valve flow characteristic Valve Cone Angle 0–20 deg 10 deg
and therefore the performance of the inlet port. However, the in- Valve Seat Angle 30–45 deg 5 deg
vestigations described were mostly limited to specific valve ge- This resulted in a total number of 36 valve geometries to form the
ometries and did not provide a complete analysis of the flow present parametric study.
through the intake valve, the flow separation phenomenon and its
dependence on geometrical parameters. Results and Discussion
No comprehensive and systematic data exist about the influence
of valve and port geometries on discharge coefficients. Neither The magnitude of discharge coefficients, transition in flow re-
has there been any previous detailed investigation into the various gimes, and the various flow patterns occurring in the valve pas-
flow regimes that influence flow performance. The present work sage are dependent on the valve geometry as Part I of this paper
therefore investigated in detail the influence of valve geometry on clearly identified. The combination of the geometrical parameters
the passage flow characteristics by varying incrementally the key valve fillet radius, cone angle, and seat angle and their influence
parameters on the valve flow characteristics in terms of discharge coefficients
was investigated for the full range of 36 valve geometries de-
• fillet radius signed for the present investigation. Measurements were carried
• cone angle out for lift conditions ranging from L/D⫽0.02 to 0.24 at pressure
• seat angle differentials across the valves between 20 mm H2O and
100 mm H2O in incremental steps of 20 mm H2O.
of the valve geometry using discharge coefficient measurements The detailed behavior of the boundary layer in the passage was
and heat transfer data from the transient liquid crystal technique to resolved using heat transfer data from the transient liquid crystal
resolve the boundary layer flow in the valve passage. technique. The measurement techniques and the data interpreta-
tion applied for the present parametric study were established in
Part I of this paper. The results in this paper are presented and
The Valve Models discussed on the basis of the largest pressure differential investi-
gated since it was found that the pressure drop across the passage
To carry out the parametric investigation, a relevant range for a had no influence on the discharge coefficient characteristic. The
variation of the valve geometry had to be developed to cover measurements showed good repeatability with the differences
practical inlet valve designs. Apart from having a very good flow well within the estimated uncertainty of ⫾3.6 percent for the dis-
performance, the overall design requirements of an engine inlet charge coefficient results.
valve can be summarized as follows:
• Adequate seating area 共contact area between valve head and Influence of Valve Cone Angle
seat兲 to improve cooling of the valve during the valve closing The effect of the valve cone angle on discharge coefficient
period. characteristic was examined for the various seat geometries inves-
• Shape of the valve to keep thermal stresses to a minimum. tigated. Results of the mass flow rate and discharge coefficient
• Valve head as light as possible since it is an accelerated mass characteristics for the three valve cone angles for a 45 deg seat
and therefore exerts high forces on the valve train; it is possible geometry are presented in Fig. 2共a,b兲, respectively, and reflect the
that for high engine speeds the resonant frequency of the valve pronounced effect of valve cone angle on valve flow characteris-
train is approached. tics which were consistent for all the other seat geometries
• Valve shapes that are not subject to engine deposits and wear. investigated.

Journal of Fluids Engineering DECEMBER 2000, Vol. 122 Õ 659

Downloaded 10 Jul 2010 to 129.173.72.87. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
 bination with a 45 deg seat would not show flow regime II and III
and transition was assumed to be directly from flow regime I to IV
in the medium lift range. The assumed steady increase in the flow
separation would therefore be restricted to the seat face which
may be a result of the increased axial flow momentum induced by
a large cone angle.
The cone angle of the valve therefore affected the discharge
characteristic over the full range of lifts. A large cone angle was
thought to direct the flow in a more axial direction resulting in
high discharge coefficients in the low lift range. This influence
had the opposite effect in the medium to high lift range with a
small cone angle resulting in earlier transition to flow regime II
and latter transition to flow regime IV. A small valve cone angle
was assumed to direct the flow toward the plane of the cylinder
head, which suppressed flow separation from the seat face and
resulted in a superior flow performance in the medium to high lift
range.
The influence of the valve cone angle on the passage flow char-
acteristics was examined by means of heat transfer coefficient
data, which was used to resolve the detailed behavior of the
boundary layer inside the valve passage. Distribution of heat
transfer coefficients on the valve seat and seat sealing face is
presented in Fig. 3共a,b兲 and Fig. 4共a,b兲 for the lift conditions
L/D⫽0.06 and L/D⫽0.18, respectively. For the low lift condi-
tion the heat transfer data indicated that the flow separated and
reattached to the seat sealing faces for all three cone angle con-
figurations, reflected by a local peak in heat transfer coefficient in
Fig. 3共a,b兲. The passage flow was therefore in regime I for all
three cone angle geometries at this lift condition. The heat transfer
results on the valve seat and seat sealing face clearly illustrated
the effect of the valve cone angle on the flow direction into the
valve passage. A 0 deg valve cone angle directed the flow in a
more radial direction and produced the largest separation area
from the valve seat face. This influence of valve cone angle on
flow direction into the passage resulted in a reversed characteristic
on the seat sealing face with a small valve cone angle suppressing
the flow separation area.
At L/D⫽0.18, those influences of the valve cone angle were
continued. The results on both seat sealing faces reflected the
significance of the valve cone angle on flow direction into the
passage with the flow being fully separated from the valve seat
face for the 0 deg valve cone angle configuration and separating
and reattaching for the 10 and 20 deg cone angle geometries. The
Fig. 2 „a… Mass flow rate versus lift; variation of valve cone
angle; „b… discharge coefficients versus lift; variation of valve liquid crystal data indicated that the separation area on the valve
cone angle seat face was decreased with an increase in valve cone angle. This
effect was reversed on the seat sealing face with the separation
area being increased with a larger valve cone angle. The 10 and
20 deg cone angle geometries showed fully separated flow from
A flat valve shoulder 共0 deg cone angle in Fig. 2兲 showed a first the seat sealing face with the passage flow being in regime IV for
change in slope at about L/D⫽0.11 in the mass flow rate versus these cone angle configurations. For the 0 deg cone angle geom-
valve lift plot 共Fig. 2共a兲兲 with a corresponding reduction in the C D etry, the flow reattached at the trailing edge of the seat sealing
value, indicating a transition in flow regime, whereas the 10 and face which reflects transition to flow regime III for this
20 deg cone angle continued the effective mode of flow regime I valve geometry as concluded from the discharge coefficient
up to L/D⫽0.14. However, the 10 deg cone angle showed an
measurements.
immediate change in slope again at L/D⫽0.16. This would sug-
The influence of valve cone angle on passage flow characteris-
gest that the flow separated from the valve seat face at L/D
⫽0.14, with the effective flow area remaining constant over a tics can therefore be summarized as follows. A large cone angle
small range of valve lifts, resulting in a small recovery in dis- suppresses flow separation from the valve seat face resulting in an
charge coefficient, which was followed immediately by the flow improved flow performance in the low lift range. This influence of
also detaching from the seat face 共transition from flow regime II cone angle is reversed in the medium to high lift range since a
to III兲. The latter was assumed to have produced the continuous small cone angle directs the flow toward the plane of the cylinder
drop in discharge coefficients at L/D⫽0.16 as shown in Fig. 2共b兲. head, suppressing flow separation from the seat face. The flow
At L/D⫽0.19, the flow was thought to reattach to the valve face, performance of a small valve cone angle was therefore superior in
which was consistent with flow regime IV. the medium to high lift range as compared with a large valve cone
The results for a 20 deg valve cone angle suggested that the angle. This behavior was consistent for all seat angles investi-
separation process from valve and seat face was a continuous one gated. However, the effect became more significant as the valve
in the lift range between L/D⫽0.125 and 0.16. Tanaka 关3兴 and seat angle was increased in that the flow separation from the valve
Annand 关6兴, however, concluded that a 20 deg cone angle in com- seat face was most pronounced for large seat angle configurations.

660 Õ Vol. 122, DECEMBER 2000 Transactions of the ASME

Downloaded 10 Jul 2010 to 129.173.72.87. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
 Fig. 3 „a… Heat-transfer coefficient versus lift; valve seat; L Õ D
Ä0.06; „b… heat transfer coefficient versus lift; port seat; L Õ D Fig. 4 „a… Heat transfer coefficient versus lift; valve seat; L Õ D
Ä0.06 Ä0.18; „b… heat transfer coefficient versus lift; port seat; L Õ D
Ä0.18

Influence of Fillet Radius seat face, the effect of fillet radius for the 35 deg seat valves was
Three fillet radius ratios (R/D⫽0.20, 0.25, and 0.30兲 were in- found to be insignificant and discharge coefficient curves almost
vestigated and their influence on the flow behavior was examined matched for the various fillet radii investigated.
for the various valve seat and valve head configurations. Results As was observed for the valve cone angle, the effect of fillet
for the discharge coefficients are presented in Fig. 5 and Fig. 6, radius was most pronounced for the 45 deg valve seat geometries
respectively, for a 35 and 45 deg seat angle geometry, which presented in Fig. 6 for a 0 deg valve cone angle geometry. For this
reflect the general behavior of the influence of fillet radius on geometry, a large fillet radius resulted in an improved flow per-
valve flow. formance in the lift range smaller than 0.05 with the influence
The results for the 35 deg seat angle, shown in Fig. 5, suggested almost diminishing in the medium to high lift range 0.12⬍L/D
that there was almost no influence of the fillet radius on the gen- ⬍0.25. It was reasoned that a large fillet radius would not induce
eral discharge characteristics over the full range of lifts investi- as much radial momentum in the flow as a small fillet radius,
gated. However, a small influence of the fillet radius was shown in resulting in a decreased flow separation area from the upper valve
the low lift range (L/D⬍0.08) and therefore in the region where face corner and consequently producing higher discharge coeffi-
flow transitions were assumed for this geometry. A larger fillet cients in this range of lifts. The investigation of Vafidis and
radius resulted in slightly higher discharge coefficients since it Whitelaw 关9兴 showed that for a 45 deg seat geometry a small
was assumed to direct the flow in a more downward direction separation area was formed at the upper valve face corner at low
suppressing flow separation from the valve seat face at low lifts lifts. The present results would suggest that the fillet radius had
which results in slightly higher discharge coefficients. In the me- some influence on the flow separation from the upper valve seat
dium to high lift range, where the flow was attached to the valve face corner, which would therefore explain its pronounced effect

Journal of Fluids Engineering DECEMBER 2000, Vol. 122 Õ 661

Downloaded 10 Jul 2010 to 129.173.72.87. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
 Fig. 5 Discharge coefficients versus lift; variation of fillet Fig. 7 Discharge coefficients versus lift; variation of seat
radius angle

in the low lift range for only the 45 deg seat geometry. The small nozzle shape, which resulted in high discharge coefficients at low
separation area formed at the upper corner of the valve seat lifts as compared with the other seat geometries. Furthermore, it
face was therefore suppressed by the axially directed flow for a was assumed that any flow separation area formed at the upper
larger fillet radius resulting in an improved flow performance valve face corner was decreased or suppressed for the 30 deg seat
in the low to medium lift range for that valve configuration. In geometry as the cone angle was increased. This characteristic of
the high lift range, however, the differences for the various fillet the 30 deg seat geometry resulted in an improved flow perfor-
radii were shown to be insignificant within the experimental error mance at low lifts. A sudden drop in discharge coefficients oc-
band. curred at a relatively low lift for a 30 deg seat (L/D⫽0.05 in Fig.
8兲, which was assumed to be caused by early flow separation from
the shallow seat face. This transition point from flow regime I to
Influence of Valve Seat Angle IV was the only discontinuity observed in the discharge coeffi-
The seat angle was varied between 30 and 45 deg in incremen- cient versus lift plot for a 30 deg seat as shown in Fig. 7 and Fig.
tal steps of 5 deg in the present parametric study and results for 8 and hence the only change in flow pattern in the valve passage
discharge coefficients are shown in Fig. 7 and Fig. 8 for a 0 and for this seat geometry. This characteristic resulted in a poor flow
20 deg valve cone angle respectively. performance in the medium to high lift range for this valve con-
A 30 deg seat angle generally showed very high discharge co- figuration. The present findings would therefore support the rec-
efficients at low lifts 共i.e., Fig. 8兲. In this lift range, the valve ommendations of Tanaka 关3兴 and Annand 关6兴 to use a 30 deg
shoulder and the 30 deg seat formed the passage similar to a valve seat angle for valves that are predominantly operated at low

Fig. 6 Discharge coefficients versus lift; variation of fillet Fig. 8 Discharge coefficients versus lift; variation of seat
radius angle

662 Õ Vol. 122, DECEMBER 2000 Transactions of the ASME

Downloaded 10 Jul 2010 to 129.173.72.87. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
 lifts due to their superior performance in this region, even though
the range of effective operation for a 30 deg seat configuration
was very small according to the current results.
For a 35 deg seat angle, the sudden drop in C D observed for the
30 deg valve seat geometry at around L/D⫽0.06 was diminished
and resulted in a sharp but continuous decrease in discharge co-
efficients for increasing valve lifts, as shown in Fig. 8 for a 20 deg
valve cone angle. This characteristic was further relaxed as the
seat angle was increased to 40 and 45 deg, which may be ex-
plained by the fact that the early separation from the seat face for
the 30 deg seat angle was a sudden one and resulted in a sharp
decrease in discharge coefficients, whereas for a larger seat angle
this separation process was delayed and produced a continuous
decrease in flow performance. The latter characteristic was indi-
cated in Fig. 8 in that the transition from flow regime I to presum-
ably IV for a 20 deg valve cone angle was considerably delayed as
the valve seat angle was increased.
The recovery in discharge coefficient, which occurred at ap-
proximately L/D⫽0.04 for the 30 deg seat geometry having a flat
valve shoulder, was also reflected for the 35 deg seat geometries
with a 0 and 10 deg valve cone angle at L/D⫽0.06 as shown i.e.,
in Fig. 7 for a 0 deg valve cone angle. However, this temporary
improvement in discharge coefficients at relatively low lifts was
decreased for a 40 deg valve seat angle. The recovery in flow
performance was assumed to be produced by an increased radially
diverted flow in case of a small valve cone angle suppressing flow
separation from the seat face which resulted in the effective flow
area remaining constant over a small range of valve lifts. The
effect was consequently most pronounced for a shallow seat and
continuously decreased as the seat angle was increased.
The large increase in discharge coefficients in the medium lift
range was observed for only the 35 deg and 40 deg seat angle
geometries having a flat valve shoulder 共0 deg valve cone angle兲,
shown in Fig. 7, with valves for C D exceeding 0.9 in case of a 40
deg seat angle, which could be of significant interest for practical
inlet valve designs. This peculiar characteristic was therefore a
combined effect of valve shoulder and valve seat angle and it was
explained in Part I of this paper as a result of a delayed flow
separation from the seat face leading to a significant pressure re-
covery in the valve passage and consequently high flow perfor-
mance. However, the steep rise was followed by a sharp drop in
flow performance with discharge coefficients being similar to the
other seat geometries investigated at high lifts (L/D⬎0.16).
The 45 deg seat angle valve had a generally inferior perfor-
mance in the low lift range and showed the transition to the four Fig. 9 „a… Heat transfer coefficient versus lift; valve seat; L Õ D
flow regimes as outlined in Part I of this paper for a 0 and 10 deg Ä0.12; „b… heat transfer coefficient Fig. versus lift; port seat;
valve cone angle 共see Fig. 7 for a 0 deg valve cone angle兲. The L Õ D Ä0.12
transition from flow regime I to II was delayed for the 10 deg
valve cone angle as a result of suppressed flow separation from
the valve seat face caused by the larger valve cone angle. The
change to flow regime II was immediately followed by a further
transition presumably to flow regime III at L/D⫽0.14 and con- the flow being fully separated for the 40 and 45 deg seat angle
tinued to L/D⫽0.18 where the final state of flow was reached for configuration, respectively. This effect of the seat geometry on
this valve geometry. flow separation was the opposite on the seat sealing face with the
The 20 deg cone angle showed only one clear transition for the flow for the 30 and 35 deg being fully separated and therefore in
45 deg seat geometry 共shown in Fig. 8 at L/D⫽0.15) in agree- flow regime IV for these seat configurations, which was in agree-
ment with the findings of Tanaka 关3兴. The transition was therefore ment with the flow interpretation from the discharge coefficient
assumed to be from flow regime I to IV with suppression of flow measurements. For the 40 and 45 deg seat geometry, the passage
regime II and III, which resulted in a smooth change in the slope flow was in regime II and attached to the sealing face at passage
of the discharge coefficient graph for increasing valve lifts. exit at this lift condition. The 40 and 45 deg seat geometries
The occurrence of the different flow regimes was verified from represented the only seat configuration indicating flow regime II
heat transfer data in the valve passage which are presented for the in the present investigation.
lift condition L/D⫽0.12 and L/D⫽0.18 in Fig. 9共a,b兲 and Fig. The data presented for the lift configuration L/D⫽0.18 in Fig.
10共a,b兲, respectively, for a 0 deg valve cone angle. The heat trans- 10共a,b兲 continued this trend of the influence of valve seat angle on
fer results clearly reflected the pronounced effect of the valve seat flow separation. The reattachment location on the valve seat face
angle on flow separation and the location of flow reattachment at moved in downstream direction as the seat angle was increased
low lifts. The results at the lift condition L/D⫽0.12 clearly illus- with the flow being fully separated from the valve seat face for the
trated that the reattachment location on the valve seat face moved 45 deg seat configuration at this lift condition. The heat transfer
in downstream direction for an increase in valve seat angle with coefficient distribution on the seat sealing face indicated fully

Journal of Fluids Engineering DECEMBER 2000, Vol. 122 Õ 663

Downloaded 10 Jul 2010 to 129.173.72.87. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
 ometries indicating the transition points to the four modes of flow.
For the other valve geometries investigated in the current study,
flow regime II and III were suppressed.

Conclusions
The present parametric study of the influence of valve geometry
on inlet port flow showed that a variation of valve cone angle
defined the flow into the passage over the full range of valve lifts
with its influence decreasing towards high valve lifts. A small
valve cone angle directed the flow radially, which suppressed flow
separation from the seat sealing face in the medium to high lift
range but enhanced the tendency of the flow to separate from the
valve seat face. This influence resulted in a reversed effect at low
lifts. A large valve cone angle induced an increased axial momen-
tum in the flow and guided it into the passage, which resulted in
high discharge coefficients in this lift region.
The fillet radius was found to have a minor influence on the
flow characteristics for 30, 35, and 40 deg seat geometries. Some
influence of the fillet radius was observed at low valve lifts and
was most pronounced for a 45 deg seat geometry.
The influence of the valve seat angle on the valve performance
characteristics was most significant of all the geometrical param-
eters varied. In the low lift range, discharge coefficients continu-
ously decreased as the valve seat angle was increased. In the
medium to high lift range this effect was shown to be reversed
with a 45 deg seat geometry becoming the best performing
geometry.
An interesting characteristic was observed for the 35 and 40
deg valve seat geometries having a flat valve shoulder with the
discharge coefficients rising to relatively high values in the me-
dium lift conditions making these geometries of significant inter-
est for actual engine valve designs.
The current study presented the first large scale investigation of
the valve geometry effects on port flow characteristics by combin-
ing experimental data from flow and heat transfer measurements
to obtain highly detailed information about the influence of the
geometrical parameters on the flow characteristics in and around
the valve passage.

Nomenclature
CD ⫽ discharge coefficient
d, ⌬ ⫽ differential
D ⫽ valve head diameter
Dp ⫽ inlet port diameter
h, HTC ⫽ heat transfer coefficient
Fig. 10 „a… Heat transfer coefficient versus lift; valve seat; L ⫽ valve lift
L Õ D Ä0.18; „b… heat transfer coefficient versus lift; port seat; ␮ ⫽ kinematic viscosity
L Õ D Ä0.18 P ⫽ pressure
R ⫽ valve fillet radius
Re ⫽ Reynolds number 兵 ␳ D p v a / ␮ 其
va ⫽ velocity of inlet flow
separated flow for the 30, 35, and 40 deg seat geometries with
the flow reattaching to the trailing edge for the 45 deg seat References
configuration.
关1兴 Luke, C. E., 1906, ‘‘The Pressure Drop through Poppet Valves,’’ Trans.
The liquid crystal data therefore confirmed an early transition to ASME, 27, pp. 232–301.
flow regime IV for the 30 and 35 deg seat geometry with the flow 关2兴 Lewis, G. W., and Nutting, E. N., 1918, ‘‘Airflow through Poppet Valves,’’
being fully separated from the seat sealing face, which resulted in NACA Report, No. 24.
the inferior performance in the medium to high lift range for these 关3兴 Tanaka, K., 1929 ‘‘Airflow Through Suction Valve of Conical Seat,’’ Aero-
nautical Research Institute Report, Tokyo Imperial University, Part 1, pp. 262;
seat configurations. Part 2, p. 361.
The variation in seat angle had the most pronounced effect on 关4兴 Wood, G. B., Hunter, D. U., Taylor, E. S., and Taylor, C. F., 1942, ‘‘Air Flow
the discharge characteristic of all three valve parameters varied in through Intake Valves,’’ Trans. of SAE, 50, p. 212.
the present investigation. In the low lift range discharge coeffi- 关5兴 Kastner, L. J., Williams, T. J., and White, J. B., 1963–1964, ‘‘Poppet Inlet
Valve Characteristics and their Influence on the Induction Process,’’ Proc. Inst.
cients steadily decreased as the valve seat angle was increased Mech. Eng., 178, Part 1, No. 36, pp. 955–975.
from 30 to 45 deg, which was reversed in the high lift range 共flow 关6兴 Annand, W. J. D., 1969, ‘‘Engine Breathing,’’ Auto. Eng., 7, pp. 50–55.
regime IV兲 when the flow started to reattach to the valve seat but 关7兴 Pope, J. A., 1967, ‘‘Techniques Used in Achieving a High Specific Airflow for
was fully separated from the seat face. The four flow regimes High-Output, Medium-Speed Diesel Engines,’’ ASME J. Eng. Power, 89, pp.
265–275.
were thought to be exhibited only by the 45 deg seat geometry 关8兴 Bicen, A. F., and Whitelaw, J. H., 1983, ‘‘Steady and Unsteady Flow through
having a 0 and 10 deg cone angle. Three discontinuities in the an Intake Valve,’’ Imperial College, Mech. Eng. Dept. Report, No. FS/83/11.
mass flow characteristic were solely observed for these seat ge- 关9兴 Vafidis, C., and Whitelaw, J. H., 1984, ‘‘Steady and Pulsating Air Flow

664 Õ Vol. 122, DECEMBER 2000 Transactions of the ASME

Downloaded 10 Jul 2010 to 129.173.72.87. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm
 Through a Stationary Intake Valve of a Reciprocating Engine,’’ Imperial Col- sional Prediction of Flow in an Axisymmetric Port/Valve Assembly,’’ SAE
lege, Mech. Eng. Dept. Report, No. FS/84/04. Technical Paper 870592.
关10兴 Bicen A. F., Vafidis C., and Whitelaw J. H., 1984, ‘‘Steady and Unsteady Air 关12兴 Weclas, M., Melling, A., and Durst, F., 1998, ‘‘Flow Separation in the Inlet
Flow through an Intake Valve of a Reciprocating Engine,’’ Proc. Symp. of Valve Gap of Piston Engines,’’ Prog. Energy Combust. Sci., 24, No. 3, pp.
Flows in IC Engines., pp. 47–55. 165–195.
关11兴 Gosman A. D., and Ahmed A. M. Y., 1987, ‘‘Measurement and Multidimen- 关13兴 TRW Automotive Systems Ltd, 1997, private correspondence.

Journal of Fluids Engineering DECEMBER 2000, Vol. 122 Õ 665

Downloaded 10 Jul 2010 to 129.173.72.87. Redistribution subject to ASME license or copyright; see http://www.asme.org/terms/Terms_Use.cfm