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

net/publication/242727542

Design and Development of a 2Step Rocker Arm

Article · April 2007

DOI: 10.4271/2007-01-1285

CITATION READS
1 973

3 authors, including:

Timothy Kunz
Delphi
3 PUBLICATIONS   27 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Timothy Kunz on 10 February 2015.

The user has requested enhancement of the downloaded file.

SAE TECHNICAL
PAPER SERIES 2007-01-1285

Design and Development of a 2-Step Rocker Arm

N. Hendriksma, T. Kunz and C. Greene
Delphi Corporation

Reprinted From: Variable Valve Actuation, 2007

(SP-2135)

2007 World Congress

Detroit, Michigan
April 16-19, 2007

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-0790 Web: www.sae.org
By mandate of the Engineering Meetings Board, this paper has been approved for SAE publication upon
completion of a peer review process by a minimum of three (3) industry experts under the supervision of
the session organizer.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise,
without the prior written permission of SAE.

For permission and licensing requests contact:

SAE Permissions
400 Commonwealth Drive
Warrendale, PA 15096-0001-USA
Email: [email protected]
Fax: 724-776-3036
Tel: 724-772-4028

For multiple print copies contact:

SAE Customer Service

Tel: 877-606-7323 (inside USA and Canada)
Tel: 724-776-4970 (outside USA)
Fax: 724-776-0790
Email: [email protected]

ISSN 0148-7191
Copyright © 2007 SAE International
Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE.
The author is solely responsible for the content of the paper. A process is available by which discussions
will be printed with the paper if it is published in SAE Transactions.

Persons wishing to submit papers to be considered for presentation or publication by SAE should send the
manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.

Printed in USA
 2007-01-1285

Design and Development of a 2-Step Rocker Arm

N. Hendriksma, T. Kunz and C. Greene
Delphi Corporation

Copyright © 2007 SAE International

ABSTRACT fewer cylinders exacerbates noise and vibration issues,

further limiting deactivation at low engine speeds. Also,
2-Step variable-valve lift and timing is a high-value many Type II engines are already equipped with dual
technology for the further development of automotive independent cam phasing (DICP) which provides a fuel
internal combustion engines. 2-Step valve train systems economy benefit compared to a fixed cam timing
provide improved engine efficiency, emissions, and baseline [3, 4]. Since this is achieved through reduced
performance using components that are relatively low- pumping losses, the net benefit of adding cylinder
cost and compatible with new and existing cylinder deactivation will be correspondingly reduced compared
heads. This paper describes the design and to cam in block pushrod applications.
development of a 2-Step rocker arm using a combination
of analytical tools and physical testing. Prototype For this class of DICP applications, recent research has
hardware was built to confirm the design. Performance identified 2-Step valve lift with early intake valve closing
and durability test results are presented. as a promising alternative to cylinder deactivation [5, 6].
This strategy provides excellent fuel economy benefit
INTRODUCTION without the negative side effects on acoustics, vibration,
and calibration complexity. An impediment to
Variable valve actuation (VVA) is an essential technology implementing this approach has been the lack of low
for improving the fuel efficiency of the internal cost, durable hardware, particularly the rocker arm
combustion engine. Until recently, economic factors component. This paper describes the design and
have prevented mainstream implementation of many development of a productive 2-Step rocker arm using a
VVA technologies. However, increasing fuel prices and combination of analytical tools and physical testing.
customer demand for higher fuel economy have made
more VVA strategies economically feasible for high
volume production. For example, DaimlerChrysler and
General Motors have successfully implemented cylinder
deactivation in high volume pushrod engine applications
[1, 2].

Similar focus is now on overhead cam applications,

specifically the Type II architecture shown in Figure 1.
This valve train configuration has been widely adopted
for its advantages of low friction, high stiffness, low
reciprocating mass, compact packaging, and low cost.
However, the lack of a fixed support to anchor the rocker
arm combined with the limited packaging space
significantly complicates the design of VVA components
for this valve train.

Given the proven fuel economy benefit of cylinder

deactivation on pushrod engines, many manufacturers Figure 1. Type II Valve train
are considering this technology for overhead cam type
engines as well. For engines with fewer than 8 cylinders, DESIGN CONSIDERATIONS
however, the fuel economy benefit actually realized may
be less than anticipated, particularly for applications with Numerous product goals, requirements, and constraints
relatively large vehicle weight to displacement ratios. had to be considered throughout the design process.
These vehicles can not operate in deactivated mode to Since the 2-Step rocker arm was a new product, not all
the same extent as more lightly loaded systems and requirements were well defined and some had to be
1
estimated or assumed based on experience. Tradeoffs Switching mechanisms are also susceptible to a
between conflicting design objectives had to be identified behavior, hereafter referred to as an ejection, which
and resolved. Some of the key design considerations occurs when the lock mechanism is only partially
are discussed below. engaged at the start of the high mode lift event. In some
cases, the lock mechanism disengages prior to fully
PACKAGING completing the high mode lift event and significant
Packaging was a significant constraint in the design the impact forces are generated when the lift differential
2-Step rocker arm. To minimize friction, the initial goal between low and high modes is large.
was to have rolling cam follower contact in both low and
high mode operation. However, such concepts were too 2-STEP ROCKER ARM CONCEPTS
wide to fit most applications. Using a combination of
roller and slider reduced the arm width to approximately A concept generation activity produced several potential
24 mm which can be packaged in many applications. design alternatives. A selection process was then used
For fuel economy reasons, the low friction rolling follower to identified two concepts, denoted dual slider and center
contact was selected for low mode operation. The slider slider, for continued development and hardware
follower is better suited to high mode operation because evaluation. Both concepts required a camshaft with
the slider friction decreases with increasing engine three lobes per arm.
speed and the higher slider stiffness is beneficial for high
speed dynamics. DUAL SLIDER DESIGN
The dual slider design, Figure 2, utilizes a single center-
MASS AND STIFFNESS roller to transmit low lift cam motion, and dual outer
It was anticipated the 2-Step Rocker Arm would have sliders to transmit high lift cam motion. The roller and
greater mass and rotating inertia than a conventional bearing are similar to those in conventional Type II
finger follower rocker arm because of the additional rocker arms. The sliders pivot about a shaft positioned
components needed. The added mass lowers both the near the HLA socket and external torsion springs control
system natural frequency and increases the tendency for the lost motion of the sliders during low mode operation.
the follower to separate from the cam at elevated engine
speeds. Using simulation tools, the effect of increased
Connector
mass can be managed with valve spring design and
optimization of the valve lift event. For stiffness, the goal
was to be similar to conventional Type II rocker arms.

ROCKER ARM RATIO

A rocker arm ratio of 1.7:1 was targeted for both high
and low modes, similar to conventional rocker arm
systems. Analysis showed that even relatively small
increases in rocker arm ratio significantly reduced the Torsion Lost
Motion Spring Lock Block
effective stiffness and increased the valve lift loss due to
hydraulic lash adjuster (HLA) leakdown.
Figure 2. Dual Slider 2-Step Rocker Arm
ACTUATION SYSTEM COMPATIBLITY
Another objective was that the 2-Step rocker arm be For high mode, additional oil pressure is applied to a
compatible with actuation systems that are low cost, piston above the HLA socket. This moves the connector
compact and reliable. Hydraulic actuation has been and pulls the lock block under the sliders, preventing
used successfully in numerous production switching slider rotation with respect to the rocker arm body.
systems and was the primary direction for the 2-Step When the oil pressure is reduced, a compression spring
rocker arm. This led to a requirement that the locking pushes the block back to the low mode position.
mechanism be located in the vicinity of the HLA socket.
CENTER SLIDER DESIGN
RELIABILITY AND DURABILITY The center slider design, Figure 3, utilizes two outer
The 2-Step rocker arm must be as reliable and durable rollers to transmit the low lift cam motion and a single
as the conventional rocker arm it replaces. high lift follower (HLF) to transmit the high lift cam
Requirements for conventional Type II rocker arms were motion. The rollers are crowned to avoid edge loading
well understood from prior production experience. So, with the cam lobes. Both rollers are fixed to the shaft so
efforts were focused on identifying and understanding the three components rotate as an assembly. The roller
new failure modes early in the design stage. The rocker shaft is supported by the rocker arm body using a
arm must be robust to failures in other components in patented bearing arrangement. In low mode, the HLF
the system as well. For example, malfunctions of the may not be aligned to the cam because the rollers
actuation system can result in low mode operation at dictate the relationship between the rocker arm and
higher than expected engine speeds. camshaft. However, the slider loads generated in low
mode are relatively small so any edge loading that
occurs does not cause distress. An internal

2
compression spring controls the motion of the slider DESIGN & ANALYSIS
during low mode operation.
The use of simulations and analytical tools early in the
design process was critical to successfully develop and
optimize the 2-Step rocker arm. Existing analytical tools
and methods were used to evaluate the conventional
aspects of the 2-Step rocker arm but new analysis tools
and procedures were developed for the unique features
of the rocker arm. The following sections detail the
analysis and simulation tools used to optimize the center
slider design.

MASS AND STIFFNESS

Finite element analysis (FEA) was used to optimize the
mass, rotating inertia, and stiffness of the rocker arm.
Valve train simulation routines were used to predict the
Figure 3. Center-Slider 2-Step Rocker Arm forces applied to the rocker arm in both low and high
mode operation. The stiffness of the rocker arm was
The rocker arm is switched to high mode using a calculated for both low and high mode operation and
hydraulically actuated lock pin located above the HLA compared to a baseline conventional roller finger follower
socket. The pin prevents the HLF from reciprocating in (RFF). The computed stiffness of the 2-Step rocker arm
the outer arm so the high lift cam actuates the valve. in low mode was about 5% lower than the baseline. The
Once the rollers separate from the cam, the center slider high mode stiffness was approximately 10% higher than
aligns with the cam lobe face equally distributing the load the baseline.
across the surface.
For the development valve train, the added mass of the
CONCEPT SELECTION 2-Step rocker arm increased the total effective
reciprocating mass by 20 grams (at the valve) compared
The two concepts were compared analytically and to the conventional system, an increase of about 21%.
experimentally. Testing uncovered a number of issues The added mass was managed with increased valve
for the dual slider. The exposed components (lock spring force and valve lift profile design. Dynamics
block, connector and torsion spring) were prone to testing of the center slider 2-Step rocker arm confirmed
damage during handling and assembly operations. Also, high speed dynamic performance using productive valve
the location of the locking mechanism at the valve end of springs and lift profiles.
rocker arm produced unacceptable centrifugal forces at
high speeds and was not robust to ejection related SOCKET STABILITY
impact forces. Finally, cam/slider friction was higher with The socket was designed so the rocker arm does not
the dual slider design. Figure 4 compares cam drive separate from the HLA. As shown in Figure 5, the lines
torque measurements made with similar lift events for from the center of the offset radii passing through the
the same test conditions. Based on the analysis and test ball center define two boundary limits for the reaction
results, the center slider 2-Step rocker arm was selected force vector produced at the ball/socket interface [8].
as the primary design direction.

Reaction Force
Vector
Cam Drive Torque

Dual Slider
2-Step Rocker
Arm Socket

Socket/HLA Contact Limits

Center Slider

HLA

0 2000 4000 6000 Figure 5. Pictorial Description of Socket Stability

Engine Speed (RPM)

Figure 4 Motored Valve Train Cam Drive Torque

3
The socket stability ratio indicates the relative location of ROCKER ARM RATIO
the reaction force vector with respect to the two The rocker arm ratio (RAR) here is defined as the ratio of
boundary limits. The ratio is defined to be zero when the valve lift to cam lift. Given the sensitivity of system
reaction force vector bisects the two limits and +/-1 at the stiffness and leakdown lift loss to this parameter, a high
respective boundary limits. Stability ratios greater than 1 priority was placed on achieving the target RAR of 1.7:1
or less than -1 indicate less than full circumferential for both rocker arm modes. The average RAR was 1.66
contact between the ball and socket. As the absolute in low mode and 1.73 in high mode.
value of the stability ratio grows, there is an increasing
risk of the rocker arm dislodging. The 2-Step rocker arm PACKAGING
had acceptable socket stability in both modes of The packaging evaluation was accomplished with two
operation (Figure 6). Physical testing confirmed the activities. First, a tolerance variation analysis along the
analysis; no rocker arms separated from the HLA during axis of the camshaft was performed using rocker arm
normal operation or during over speed and ejection and typical cylinder head tolerances. Second, three
testing. dimensional motion analysis was done to confirm
sufficient clearances to other surfaces and components
throughout the full rotation of the camshaft for both low
1.0 and high mode operation.
Socket Stability Ratio

The tolerance analysis along the axis of the camshaft is

0.5 essentially one-dimensional and was completed using a
Low Mode
spreadsheet application. Arithmetic and statistical
0.0 values were calculated; thermal effects were also
considered to account for the different expansion rates of
cylinder head and cam materials. The analysis was used
-0.5
to identify significant sources of variation, establish
High Mode
various dimensions, and to develop suitable tolerances.
-1.0 It was also used to determine the low and high mode
-100 -50 0 50 100 cam/follower contact widths for subsequent contact
stress calculations.
Cam Rotation Angle - deg
Figure 6. Socket Stability Analysis CAM/SLIDER INTERFACE
Typically, cam and follower material combinations for
VALVE TIP INTERFACE rollers are different than for sliders. Other variable lift
The oscillating motion of the rocker arm produces a systems with both roller and slider followers have used
combination of rolling and sliding (scrub) contact composite camshafts to address conflicting material
between the rocker arm pallet and valve tip. To ensure requirements [11]. For the 2-Step rocker arm, the goal
durability of this interface, an analysis was performed, was to utilize a single cam material to minimize cost and
evaluating scrub power, scrub work, contact velocity, and risk due to the lobe widths and the number of cam lobes
contact area. The rocker arm geometry was adjusted to required.
achieve an acceptable balance between low and high
modes, as illustrated in Figure 7. Durability testing of Initial durability testing with steel followers and steel
both low and high modes confirmed acceptable wear at cams showed poor results. Cam and follower scuffing
the valve tip interface. occurred rapidly. Several follower materials were tested
as well as different levels of surface finish and camshaft
phosphate coating. These did not make satisfactory
Valve Tip High Mode improvements in durability. Consequently, various
follower coatings were investigated, and a diamond-like
carbon (DLC) coating was selected based on durability
test results with steel and nodular iron camshafts. A
feature was incorporated in the outer arm to provide oil
spray directed to the cam and slider interface when in
high mode.

Contact stress was minimized by optimizing follower

radius and valve lift profiles. Figure 8 shows the
Low Mode relationship between follower radius and contact stress
for a given valve lift event. Figure 9 shows that a profile
with similar lift area and duration but designed
specifically for the slider radius reduced the peak contact
Figure 7. Valve Tip Contact Area stress compared to simply converting the lift event used
with a production RFF.

4
 In some cases, the width of the theoretical Hertzian
Peak stress contact ellipse exceeds the actual physical width
increases available, a condition known as unsupported contact
width. Durability tests were conducted to evaluate the
increasing effect of unsupported contact width. Components were
Contact Stress

follower radius built to produce the maximum stress conditions as well

(same lift profile) as the worst case unsupported contact width conditions.
After 700 hours with a nodular iron camshaft, the wear
Stress at results were acceptable.
maximum lift
decreases LOST MOTION SPRING

Design Considerations
The design of the lost motion spring was challenging
because of packaging and load requirements. If the
Cam Rotation Angle spring pre-load reaction force at the HLA is too large, the
HLA can not expand properly to recover the fluid
Figure 8 Static Cam Contact Stress versus Follower displaced by leakdown during the lift event. Therefore,
Radius the preload must be limited to an acceptably low value or
a travel limit feature must be incorporated in the rocker
Another consideration for contact stress was edge arm. The lost motion spring is exposed to high cycle
loading of the cam and slider. Unlike the dual slider fatigue conditions so material, processing and stress
rocker arm, the center slider rocker arm aligns to the limits for the lost motion spring were similar to those
cam lobe face in high mode, which equally distributes the typically used for valve springs.
load across the surfaces. Successful durability tests
were completed with both flat and crowned cam lobes Lost Motion Spring Concept Selection
and showed that the high lift follower did not require a The initial version of the 2-Step rocker arm employed a
crown to avoid edge distress. torsion lost motion spring as shown in Figure 10.
However, prototype testing indicated significant friction
related hysteresis of the spring force and manufacturing
variation of the spring preload. With additional
with converted roller
development, it was possible to package a compression
profile
lost motion spring, also shown in Figure 10. Spring force
hysteresis was significantly reduced and the spring
Contact Stress

preload variation was low, eliminating the requirement for

a travel limit based on the HLA reaction force. The
spring is well protected from potential surface damage
detrimental to high cycle fatigue applications.

Kinematic Analysis
with profile Because of the large number of variables and nonlinear
designed for slider relationships, a spreadsheet application was developed
to facilitate lost motion spring design. The desired low
and high mode valve lift events are specified as inputs
Cam Rotation Angle along with all other geometry data necessary to define
the system. At the start of the simulation, the program
Figure 9 Cam Contact Stress for Two Different Valve Lift automatically derives the low and high mode cam lobe
Profiles shapes required to produce the specified valve lift
events.
CAM/ROLLER INTERFACE
To meet the 24 mm width target, the rollers and low lift The spreadsheet displays a graphical representation of
cam lobes are relatively narrow compared to the mechanism using markers to identify key points.
conventional valve trains. An analysis was done using During the simulation, these coordinates are
contact widths obtained from the cylinder head axial continuously updated, thereby animating the motion of
tolerance analysis. Existing analysis programs were the mechanism. The visual feedback helps confirm
used to calculate contact characteristics and stress but correct data input and illustrates relationships between
new spreadsheet models were developed to evaluate components. Parameters of interest, such as cam force,
conditions unique to the dual roller interface. For relative rotation of the high lift follower, and lost motion
example, roller diameter variation and roller shaft spring force are calculated at each cam rotation angle,
bending produce an axial shift of the cam/roller contact output to the spreadsheet, and automatically charted.
area. For a typical computer, an analysis takes a few seconds
so many iterations can be completed in a relatively short
period of time.
5
 50

Force Between HLF & Cam - N

40
Torsion Lost Spring
Motion Spring
30

20

10 Net
0

-10
Compression Lost Inertial
Motion Spring -20
-100 -50 0 50 100
Cam Rotation Angle - deg

Figure 12 Force components between HLF and cam

at an engine speed of 3000 RPM.

Low Mode Overspeed

Figure 10 Torsion and compression lost motion The potential for low mode operation at elevated engine
spring versions of the center slider rocker arm. speeds was also considered. Designing the spring to
maintain contact under these conditions was not
Analysis Results practical for packaging and inertia reasons. Figure 13
The valve lift events shown in Figure 11 were analyzed shows the spring and inertial cam forces for an over
with respect to lost motion spring requirements. The low speed condition of 4000 engine RPM. The net cam force
lift event is advanced relative to the high event in keeping first becomes negative after the point of maximum HLF
with the early intake valve closing strategy [5,6]. relative rotation. Consequently, the separation occurs on
the closing flank of the cam lobe and results from the
inability of the HLF to follow the receding cam lobe as
12
opposed to tossing the HLF over the nose of the lobe
High Lift Profile with over travel of the follower. The subsequent impact
forces are less severe compared to an over travel
condition.
Lift - mm

Low Lift 80
Profile
Force Between HLF & Cam - N

60 Spring

40

0 20
-100 -50 0 50 100
0
Cam Rotation Angle - deg
Net
Figure 11 Valve Lift Profiles -20
Inertial
The lost motion spring and HLF inertial forces are shown
in Figure 12 in terms of the force each generates at the -40
interface between the cam and HLF at an engine speed -100 -50 0 50 100
of 3000 RPM. Positive values indicate compressive Cam Rotation Angle - deg
force while negative values represent forces acting to
separate the components. The net force, which is sum Figure 13 Force components between HLF and cam
of the spring and inertial components, must be positive in at an over speed condition of 4000 engine RPM
order to maintain contact between the two components

6
The compression spring design was evaluated using a
motored valve train. Rocker arms were tested in low HLF
mode over a range of speeds for 360 hours,
representing over 25 million spring cycles. The average
load loss was approximately 10%. Over speed tests Lock Pin
were conducted and the HLF motion was observed using
high speed video. As predicted by the model, over travel
did not occur, even when significant separation was
observed between the HLF and cam lobe on the closing
flank of the lobe. Lock Pin Support

STRUCTURAL (STRESS, MASS, STIFFNESS)

The 2-Step rocker arm was optimized for stress, mass,
and stiffness using FEA. The outer arm and HLF were
optimized individually and then the assembly was Figure 15 Improved Lock Pin Support Provided by the
evaluated in both low and high modes using three times Angled Lock Pin Orientation.
the maximum kinetostatic loads. For example, the
highest tensile stresses in the outer arm were at the top MECHANICAL LASH AND VARIATION
of the pallet as shown in Figure 14. The maximum A clearance between the lock pin and the HLF saddle,
allowable stress depends on the manufacturing process, called mechanical lash, is necessary to ensure pin
material, and heat treatment and was determined in engagement for all cases of variability. The clearance
concert with fatigue testing. Investment cast and Metal tolerance is important because it contributes to valve lift
Injection Molding (MIM) designs were evaluated. variation in high mode operation. The lash variation is a
complicated nonlinear function of numerous tolerances
so a module of UGS NX solid modeling software, Vis
VSA™ [12], was used to calculate the mechanical lash
variation as a function of the constituent component
tolerances.

The mechanical lash variation in a 2-Step rocker arm

only produces valve lift variation when the rocker arm
operates in high mode, typically at engine speeds above
3000 RPM. At idle conditions, the 2-Step rocker arm
normally operates in low mode where the valve lift
variation is similar to a conventional hydraulically lashed
valve train. With cylinder deactivation, the valve lift
Figure 14 Example of Finite Element Analysis of the variation occurs at idle so the lash variation must be
Outer Arm. tightly controlled for idle quality. The acceptable lash
variation for 2-Step systems will be application specific.
NEEDLE BEARINGS
The center slider rocker arm employs an innovative LOCK PIN EJECTIONS
bearing design consisting of needle bearings packaged Lock pin ejections occur when the lock pin is only
in the sidewalls of the outer body. Shaft diameter, minimally engaged at the start of the high mode lift
needle diameter and length, and the number of needles event. The rocker arm initially follows the motion of the
were selected based on bearing calculations. The high mode cam but jumps to the low mode cam when
bearing capacity, applied loads and speeds were used to the lock pin loses engagement. During the transition, the
calculate bearing life. Testing was performed to confirm system masses are accelerated by the valve spring,
bearing durability. resulting in significant impact forces.

LOCKING MECHANISM Most systems utilizing switchable components are

The locking mechanism for the 2-Step rocker arm designed to avoid ejections by using multiple control
consists of a round pin which mates with a valves and by limiting the maximum switching speed.
corresponding saddle in the HLF. Figure 15 depicts the Even so, it is typically a requirement that the components
location of the lock pin relative to the HLF and HLA be robust to some finite number of ejections in case of
socket. This orientation reduced the cantilever of the pin system malfunctions. However, several applications
in the engaged position which minimized pin edge have been produced that switch at conditions where
loading and improved the stiffness of the locking ejections can not be avoided [9, 11]. In those cases, the
mechanism. system and components were developed to be robust to
the effects of the ejections.

7
Theoretical and experimental means were used to
evaluate the impact loads and to identify potential design 14
features minimizing ejection occurrence. Type 1 Type 2 Type 3

Theoretical Evaluation and Results

An explicit nonlinear dynamic FEA was performed to
simultaneously calculate mechanism motions, forces and High Lift

Lift - mm
stresses following an ejection. Figure 16 depicts the
simulation results of the force acting to separate the
roller and shaft. Given the complexity of the analysis,
several rocker arms were strain gauged to correlate the
dynamic FEA. Ejection testing was done to confirm Low Lift
durability of the selected design.

0
-75 -50 -25 0 25 50 75
Cam Rotation Angle - deg

Figure 17 Examples of Valve Motion Following a Lock

Pin Ejection.

Ejection Characterization
The ejections were categorized into one of three regimes
based on the timing of the impact following the pin
ejection. Type 1 ejections occur during low lift valve
opening; Type 2 occur during low lift valve closing, and
3000 Type 3 occur after low lift valve closing. Examples of
2500 measured valve motion for each of the three ejection
2000
types are shown in Figure 17.
Pullout Force - N

1500
70%
1000

500 60%
0
Relative Percentage

-500
50%
-1000
40%
-1500
0 0.00005 0.0001 0.00015 0.0002 0.00025 30%
Time - sec
Figure 16 Dynamic FEA Example 20%

Experimental Evaluation 10%

A test procedure was developed to duplicate ejections as
they would occur in an engine. An accelerometer was 0%
attached to a motored valve train and used to trigger an Type 1 Type 2 Type 3
oscilloscope set to capture output from valve lift and Type of Ejection
gallery pressure transducers. The control valve was
switched at a constant frequency of 4 cycles per second Figure 18 Relative Percentage of Ejection Types for
with no synchronization of the switch timing relative to Five Different Rocker Arms
the camshaft rotation angle. The camshaft was rotated
at 1500 RPM while maintaining the oil temperature at 40 Test Results
degrees C. The rocker arms were tested individually to Five 2-Step rocker arms were evaluated in the ejection
simplify data acquisition and post-processing. test for 50 hours. The relative frequency of each ejection
type is displayed in Figure 18 and is similar for all five
A spreadsheet application was developed to post- arms. Wear of the lock pin and HLF saddle was
process the ejection event data. The application detects monitored and found to be acceptable.
ejections by comparing the acquired valve lift and
velocity to those of non-ejecting lift events. Because the rocker arm is not retained by the HLA or
valve, there was concern that the rocker arm would be

8
dislodged. However, this behavior was not observed High Mode
during considerable ejection testing. High mode durability tests were completed with no
significant wear of cam or follower. Some pitting was
PERFORMANCE AND DURABILITY TESTING observed early in the test schedule but did not progress
over the remainder of the test. Figure 20 shows a
Numerous tests were completed with 2-Step rocker photograph of a high lift follower after the high speed
arms. Table 1 summarizes the testing, and further detail durability test. Testing with degraded and contaminated
follows. oil is continuing.

Table 1 Testing Summary

o Durability
o Structural Fatigue – Low and High Mode
o 25 Million Low Mode Cycles
o 1000 Hour High Mode Idle
o 1000 Hour Low Mode Idle
o 900 Hour High Mode High Speed
o 3 Million Switching
o 5000 Ejection Durability
o Performance/Function
Figure 20 High Lift Slider After 900 Hour High Speed
o Lost Motion Spring Durability Test
o Cam/Slider Friction
o Valve Train Dynamics – Low and High Mode Switching
o Low and High Mode Over Speed A 3 million cycle switching test was performed in a
o Stiffness and Load Test to Failure motored valve train running at 1500 cam RPM. The
control valve was set for synchronized switching at a rate
o Development Testing
of 4 cycles per second. Oil temperature was maintained
o Casting and MIM Fatigue at 120 degrees C and pressure at 3 bar. The test ran for
o Cam/Slider and Cam/Roller Interfaces approximately 250 hours and no failures were observed.
o Bearing Testing
o Roller Shaft Retention Testing Ejections
o Lost Motion and Lock Pin Springs Individual rocker arms were subjected to repeated
o Switching and Ejection Testing ejections in a motored valve train. To maximize the
ejection rate, the gallery pressure was adjusted to
DURABILITY position the lock pin with minimal engagement. The
rocker arms showed no loss of function or performance
Low Mode after accumulating 5000 ejections.
Low Mode durability testing was conducted to evaluate
the cam/roller interface, the bearings, and the lost motion DYNAMICS
spring assembly. The test of limit rollers resulted in no Valve train dynamics were tested in low and high modes.
significant wear even with the most extreme parts. The limiting speed for low mode was 4400 engine rpm
Figure 19, shows a picture of a roller that had worst case with the most aggressive profile tested (see Table 5 in
unsupported contact width. Reference 6). The limiting speed for high mode was
7400 engine rpm with nominal mechanical lash and 6600
engine rpm with maximum mechanical lash. A valve
seating velocity limit of 1 meter per second was used as
the main indicator of limiting speed.

The valve train was also subjected to momentary speeds

to 6000 engine rpm in low mode and 7500 engine rpm in
high mode with no failures.

STRUCTURAL

Static Testing
Stiffness was measured by applying a load to the valve
pallet and measuring deflection with the rocker arm
constrained as it is in the valve train. This process was
Figure 19 Worst Case Roller After 700 Hour repeated for both low and high modes.
Development Test
9
The static failure load for each mode was found by 3. Jung, Hosuk H., et al., Comparison of Dual Retard
increasing the load until the part failed. The failure VCT to Continuously Variable Event Valvetrain”, SAE
locations matched those predicted by the FEA. Paper 2004-01-1268, 2004.
4. Kramer, Ulrich, and Phlips, Patrick, “Phasing
Fatigue Testing Strategy for an Engine with Twin Variable Cam
The assembly was loaded sinusoidally between 50 N and Timing”, SAE Paper 2002-01-1101.
the peak load and tested to failure or suspended at ten 5. Sellnau, M., and Rask, E., “Two-Step Variable Valve
million cycles. The results were used to define stress Actuation for Fuel Economy, Emissions, and
limits for use with ongoing FEA activity. Additional tests Performance”, SAE Paper 2003-10-0029, 2003.
continue to optimize the design and evaluate variation in
6. Sellnau, M., et al. “2-Step Variable Valve Actuation:
the casting and MIM processes.
System Optimization and Integration on an SI
Engine”, SAE Paper 2006-01-0040, 2006
CONCLUSIONS 7. Leone, T.G., and Pozar, M., “Fuel Economy Benefit
of Cylinder Deactivation – Sensitivity to Vehicle
1. A 2-Step rocker arm was developed for Type II valve
Application and Operating Constraints”, SAE Paper
trains using a combination of theoretical and
2001-01-3591.
experimental methods.
8. Meeusen, H. J., “Overhead Cam Valve Train Design
2. Of the packaging constraints, arm width was the Analysis with a Digital Computer”, SAE Paper
most significant. The current design has a width of 660350, 1966.
24 mm. 9. Inoue, K, et al., “A High Power, Wide Torque Range
3. Two concepts were developed to the prototype Efficient Engine with a Newly Developed Variable-
hardware stage. Based on testing and analysis the Valve-Lift and Timing Mechanism”, SAE Paper
center slider design was selected. 890675, 1989.
10. Brustle, C., and Schwarzenthal, D., “VarioCam Plus
4. The stiffness of the 2-Step rocker arm was similar to – A Highlight of the Porsche 911 Turbo Engine”, SAE
conventional rocker arms. The inertia was higher
Paper 2001-01-0245, 2001.
but it was possible to achieve comparable valve train
11. Shikida, T., et al., “Development of the High Speed
performance with productive valve springs and lift
2ZZ-GE Engine”, SAE Paper 2000-01-0671, 2000.
events.
12. NX Vis VSA Software, UGS Corporation, Plano,
5. Performance and durability tests were successfully Texas, 2006.
conducted to confirm the design.
DEFINITIONS, ACRONYMS, ABBREVIATIONS
ACKNOWLEDGMENTS
CAD: Computer Aided Design
The authors acknowledge the following individuals for
their contributions to the success of this project: Hermes DICP: Dual Independent Cam Phasers
Fernandez, Ahmet Becene, Don Cole, Andy Lipinski,
Ryan Fogarty, Greg Naber, George Postlethwait, Mark DOHC: Double Over Head Camshafts
Spath, Bill Robison, Al Stone, Richard Roe, Mark
Sellnau, and Paul Van Heyningen
FE: Fuel Economy
CONTACT INFORMATION FEA: Finite Element Analysis
Nick Hendriksma: [email protected] HLF: High Lift Follower
Timothy Kunz: [email protected]
Cynthia Greene: [email protected]
HLA: Hydraulic Lash Adjuster
REFERENCES MIM Metal Injection Molding
1. Falkowski, A., et al., “Design and Development of RAR Rocker Arm Ratio
the DaimlerChrysler 5.7L HEMI Multi-Displacement
Cylinder Deactivation System”, SAE Paper 2004-01- RFF Roller Finger Follower
2106, 2004.
2. Albertson, W., et al., “Displacement on Demand for RPM Revolutions Per Minute
Improved Fuel Economy without Compromising
Performance in GM’s High Value Engines”,
Powertrain International, Volume 7, Number 1, pg
25-40, 2004.

10

View publication stats