Вы находитесь на странице: 1из 36

Fatigue Analysis of Electric Actuator

Torque Train Components -


Avoid Modifications and Maintain Margin

By
Neal Estep
Kalsi Engineering, Inc.

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing


INTRODUCTION

Why is this important?


– There may be instances where an actuator is required
to operate at a torque level that is in excess of the
manufacturer’s current rating due to:
Margin improvement initiatives
JOG MOV PV program implementation
Etc.
– Analysis can often times justify operation of the
existing actuator in an over-torque situation
Avoids costly actuator replacement modifications while
maintaining safety
(NOTE: Consider motor-gearing capability first)

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 2


INTRODUCTION

This presentation will describe the actuator


torque-train model, including:
– Testing,
– Model development,
– Model validation and
– Implementation.
The main focus will be on implementation.

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 3


BACKGROUND - TESTING
Generic thrust rating increase achieved in early 1990's
under a 4-year comprehensive program supported by
industry-wide participation.

Qualification Test Fixture

One of the most cost-effective programs; Eliminated need for actuator


replacements; Saved millions of dollars for each utility
Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 4
BACKGROUND - TESTING

Test Scope:
– Phase 1: SMB-000, -00, –0, –1
– Phase 2: SMB-2, SB-00, –0, –1 & -2, & H0BC
Test Objectives:
– 200% thrust
– 2,000 strokes, 5 stalls, seismic testing (SMB-000 through 1),
2,000 more strokes
Actuator Motor Motor Gear Worm Set Overall Spring Pack
Model Size Set (Ratio) Ratio Ratio Assembly,
(ft-lb) Part No.
SMB-000-5-1800 5 19/26 (1.368) 50 to 1 68.4 Heavy 0101-092
SMB-00-25-1800 25 33/32 (.970) 45 to 1 43.6 Heavy 0301-113
SMB-0-40-1800 40 27/45 (1.667) 37 to 1 61.7 Heavy 0501-184
SMB-1-60-1800 60 22/50 (2.273) 34 to 1 77.3 Heavy 0701-212
SMB-2-80-1800 80 20/50(2.5) 33 to 1 82.5 Heavy 0901-212

SMB Test Specimen Description

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 5


BACKGROUND - TESTING
Test Results:
– SMB-000 through SMB-1 and SB-00 through SB-1 qualified to
162% of thrust rating for 2,000 cycles
– SMB/SB-2 qualified to 120% of thrust rating for 2,000 cycles
– H0BC qualified to 150% of torque for 37 cycles
– Phase 1 results endorsed by manufacturer and met NRC
acceptance
– Limitorque issued Technical Update 92-01 allowing a 140% of
thrust rating for SMB-000, -00,- 0, and -1 for non-sponsoring
utilities.
Torque rating was not to be exceeded
– For sponsoring utilities thrust rating increases up to 200%
permissible for allowable number of valve strokes based on:
Actuator size, MOV specific load history, and Implementation
guidelines

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 6


BACKGROUND - TESTING

Some torque train component fatigue/wear related


failures occurred during the generic test program
– Parts were replaced, testing continued to achieve thrust goal

Worm shaft &


Worm bushing
Worm Shaft

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 7


BACKGROUND - TESTING

Worm

Worm Shaft

Typical torque-
related failures

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 8


BACKGROUND - TESTING

Torsional Component Average Test Torque (% of No. of Cycles to Failure In


Failure Rating) Test
SMB-000
8620 worm 117 755
4320 worm 117 2,458
4320 worm 117 1,648
Worm shaft 106 4,870
SMB-00
Worm 96 3,774
SMB-1
Worm shaft 141 1167
Worm shaft 141 714

Torque Train Component Failure Summary

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 9


BACKGROUND – METHODOLOGY
Difference Between Thrust and Torque Evaluations
Thrust components are subjected to only one stress
cycle for each MOV stroke (O-C-O or C-O-C)
In contrast, torque train components are subject to
multiple stress cycles of variable magnitude for each
MOV stroke.

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 10


BACKGROUND - METHODOLOGY
Variables
Torque component load magnitudes depend on:
– Valve type
– Stem load profile
– Stem screw thread characteristics
– Worm gear ratio
– Worm/worm gear friction coefficient
– Limit switch helix drive gear forces
Number cycles depends on:
– Number of valve operations
– Motor speed
– Stem loads
– Valve stiffness
– Stem screw lead
– Worm gear ratio
– Torque spring pack stiffness

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 11


BACKGROUND - METHODOLOGY

Event Description
A–B Thread clearance
B–C Backseating
C–D Actuator stopped
D–E Actuator reverses direction – tensile
force relieved
E–F Thread clearance
F–G Seating
G–H Actuator Stopped
H–A Actuator reverses direction –
compressive force relieved

Typical Stem Thrust and Torque


Profiles Measured
in Phase I Test Program

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 12


BACKGROUND - METHODOLOGY

Critical torsional components:


– Worm
– Worm shaft at two locations -
worm-shaft interior contact point (SMB 000 – 2)
limit switch helical drive gear (SMB 0 – 2)
Two different mechanical models are required:
– SMB-000/00 actuators have a cantilevered shaft
design at the motor pinion/drive gear
– SMB-0/1/2 actuators are supported by an oilite
bushing mounted in the housing

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 13


BACKGROUND - METHODOLOGY

l sh

Mechanical Model and Load Diagram for Mechanical Model and Load Diagram for the
the Worm/ Worm Shaft Assembly of Worm/ Worm Shaft Assembly of SMB-0/1/2
SMB-000/00 Actuators Actuators

Fw = Worm/worm gear force


Fd = Motor pinion/drive gear force
Fb = Worm shaft bushing contact point
Bs = Shaft bearing reaction force
Bw = Worm bearing reaction force
Y axis aligned with worm to worm gear contact point
θd = pinion contact point offset angle

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 14


BACKGROUND - METHODOLOGY

The computer computational algorithm performs


the following calculations:
– Static stresses for worm and worm shaft
– Mean dynamic stresses for worm and worm shaft
– Alternating dynamic stresses for worm and worm
shaft
– Fatigue life for worm and worm shaft for each load
condition
– Total cumulative fatigue damage due to multiple
strokes and multiple load cases using Miner’s Rule
– Remaining life based on torque

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 15


BACKGROUND - METHODOLOGY
Example Analysis Results Summary

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 16


BACKGROUND - METHODOLOGY
Example Static Stress Results

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 17


BACKGROUND - METHODOLOGY
Example Alternating Stress Results

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 18


BACKGROUND - VALIDATION

Torque Methodology Validation Results

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 19


IMPLEMENTATION

Evaluations of actuator thrust and torque capability at


the new requirements can be performed and will result
in one of the following conclusions:
1. Actuator can operate for 2000 strokes (qualified life).
OR
2. Actuator can operate with valve stroke restriction. NOTE:
Actuator operation can continue indefinitely IF recommended
component replacement and inspection intervals are followed.
Evaluation capabilities include SMB/SB/SBD (size 000
through 2) and HBC

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 20


IMPLEMENTATION

For implementation, the


individual valve load profile
must be modeled as a
series of linearly increasing
ramps or constant load
“dwells” using the
convention shown to the
left.

Definition of Load Profiles used in Torque Fatigue Model

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 21


IMPLEMENTATION

Ramp Profiles used to Model Torque Fatigue for a Static Valve Stroke

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 22


IMPLEMENTATION

Gate Valve Opening Dynamic Stroke Ramp


and Dwell Definitions
Gate Valve Closing Dynamic Stroke
Ramp and Dwell Definitions
Backseat

Ramp and Dwell Load


Profiles used to Model
Torque Fatigue for a
Dynamic Valve Stroke

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 23


IMPLEMENTATION

Future load profiles can be derived from existing


data using conservative assumptions

Ramp1b
0

Stem Thrust
or
Torque
Extrapolated Conditions

Existing Test Conditions

New Dwell1
Max Torque or
Thrust Goal
Ramp1a Dwell1

Extrapolation of Load Profiles to New Thrust or


Torque Goals

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 24


IMPLEMENTATION

Typically, there will be four analyses performed


for each actuator:
– Analysis 1 – Past Static Strokes
– Analysis 2 – Future Static Stokes (at the new thrust
and torque goals)
– Analysis 3 – Past Dynamic Strokes
– Analysis 4 – Future Dynamic Strokes (at the new
thrust and torque goals)

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 25


IMPLEMENTATION

A fatigue usage factor can be determined for


each analysis and combined as follows:
n1 n 2 n
+ + ...... + n = Cumulative fatigue damage ≤ 1.0
N1 N 2 Nn

Where
nn = applied cycles under load case #n
Nn = allowable design cycles under load case #n

The remaining design life is calculated in terms


of a single load case as shown below.:
n2 = N2 (1- cumulative fatigue damage)

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 26


IMPLEMENTATION

Example 1
General Input Data
Valve Thrust Torque
Size Rating Rating
Worm Shaft Valve mm Motor KN [lbf] N-m
Mtl Mtl Type [in] Class Actuator OAR RPM [ft-lb]
150 62.28 339
4320 4140 Gate 1500 SB-00-15 59.4 1800
[3] [14000] [250]

Analysis Goals
Future Future
Torque Goal Future Thrust Past Static Future Static Past Dynamic Dynamic
N-m [ft-lbs] Goal KN [lbs] Strokes Strokes Strokes Strokes
373 [275] 87.2 [19600] 123 298 100 100

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 27


IMPLEMENTATION

Past static strokes were all below the current torque rating, therefore
2,000 valve stroke cycles are permitted by the manufacturer:
Past Max Torque Design Life Strokes Fatigue Usage Factor
N-m [ft-lbs] Past Static Strokes (n) (N) (n/N)
295.6 [218] 123 2000 0.0615

Also, the past DP strokes were also below the current torque rating,
therefore, 2,000 valve stroke cycles are permitted by the manufacturer:
Past Max Torque Past Dynamic Strokes Design Life Strokes
N-m [ft-lbs] (n) (N) Usage Factor (n/N)
288.8 [213] 100 2000 0.0500

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 28


IMPLEMENTATION

The usage factor for future dynamic strokes is shown below:


Design Valve Strokes Design Valve Strokes Estimated Future Usage Factor (n/N)
Based on Worm Life Based on Shaft Life Required Valve
(N) Strokes (n)
104 2000 100 0.96

The remaining allowed future static strokes can be calculated using


Miner’s Rule:
Future static strokes = Nfuture_static x (1 – Σ(UsageFactors))
However, for this case the past static and dynamic, and future dynamic
strokes would consume all of the fatigue life.
Options:
1. Re-examine input requirements for the future dynamic valve stroke
conditions to see if the number of required future valve strokes, the
torque magnitudes and/or durations can be reduced.
2. Replace torque train components after a certain number of dynamic
valve strokes.
3. Use a larger actuator.

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 29


IMPLEMENTATION

Example 2
General Input Data
Valve Thrust Torque
Size Rating Rating
Worm Shaft Valve mm Motor KN [lbf] N-m [ft-
Mtl Mtl Type [in] Class Actuator OAR RPM lbs]
254 200 1152
4620 4140 Gate 1500 SB-1-25 171.6 1800
[10] [45,000] [850]

Analysis Goals
Future Torque
Goal N-m [ft- Future Thrust Past Static Future Static Past DP Future DP
lbs] Goal KN [lbf] Strokes Strokes Strokes Strokes
1220 [900] 280 [63,000] 110 220 100 100

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 30


IMPLEMENTATION

For this case, the future thrust and torque goals


are modest.
The analyses show that 2000 cycle life is
available for all load cases:
– Past static and dynamic
– Future static and dynamic
Therefore, any combination of load cases is
permitted up to the original 2000 cycle qualified
life as follows:
n1 n2 n3 n4 n1 + n 2 + n 3 + n 4
+ + + = ≤ 1 .0
2000 2000 2000 2000 2000

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 31


IMPLEMENTATION

Example 3
Charging system inlet isolation valves (10 valves)
– 3" Anchor Darling Double Disc Gate Valve, Class 1500
– Limitorque SMB-00 Actuator
Thrust Rating: 14,000 lbs
Torque Rating: 250 ft-lb
Worm Set Ratio: 45:1
Overall Gear Ratio: 49:1
Stem: 1.25", 1/3 pitch 2/3 lead
JOG PV implementation required an increased valve factor,
which resulted in the following thrust and torque capability
changes:

Torque (ft-lb) / % Rating Thrust (lb) / % Rating


Pre-JOG 250 / 100% 15,060 / 108%

Post-JOG 300 / 120% 18,000 / 129%

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 32


IMPLEMENTATION

Post-JOG Implementation Options


– Replace with SMB-0 (seismic evaluation, snubber upgrade,
etc.)
– Evaluate acceptability to operate under increased torque
levels using LiFE software
Evaluation Results
– Static Stroke
Allowable thrust cycles = 2,000
Allowable torque cycles = 2,000
– Design Basis Dynamic Stroke
Allowable thrust cycles = 2,000
Allowable torque cycles = 59 (only one is required)

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 33


IMPLEMENTATION

Evaluation Conclusion
– MOV specific evaluation using LiFE shows that the
existing SMB-00 actuator can satisfy cumulative
static and dynamic design basis requirements for
this MOV at torque and thrust levels exceeding the
rating without compromising safety.

Cost Savings exceeded $400,000

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 34


IMPLEMENTATION

Example 4: HBC Evaluation


BIF Butterfly Valve
Limitorque H2BC with SMB-00 Actuator
– Torque Rating: 2,200 ft-lb
– Worm Ratio: 70:1
Desired torque level: 2,800 ft-lb (127% of rating)
for static strokes
Result
– 1,385 allowable valve cycles under over-torque
conditions based on the MOV specific evaluation

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 35


Conclusions
– Actuator torque fatigue analysis is very sensitivity to
the valve load profile history and the analysis is
much more involved than that for thrust fatigue
– A conservative estimate of the past and future static
and dynamic valve stroke load profiles can be used
to develop input data for the fatigue life model
– Depending on the resulting stress levels and number
of fatigue cycles, restricted or unrestricted actuator
life is possible
– Evaluation of actuator/HBC thrust and torque
capability in excess of the ratings can eliminate the
need for expensive actuator modifications while
safely increasing or maintaining margin

Tenth NRC/ASME Symposium on Valves, Pumps, and Inservice Testing 36