OUTLINE • A WEB WEB DEFLECT DEFLECTION ION SURVEY SURVEY IS.. IS.. • WHY WHY RUN RUN WEB WEB DEFLE DEFLECT CTIO IONS NS • TOOLS TOOLS AND TECHN TECHNIQ IQUES UES • WEBMAP WEBMAP DEVELO DEVELOPME PMENT NT • CASE CASE HIST HISTOR ORIE IES S • DEMON DEMONSTR STRATI ATION ONS: S: SOFTWAR SOFTWARE E and HARDWARE 1/10/01
2
OUTLINE • A WEB WEB DEFLECT DEFLECTION ION SURVEY SURVEY IS.. IS.. • WHY WHY RUN RUN WEB WEB DEFLE DEFLECT CTIO IONS NS • TOOLS TOOLS AND TECHN TECHNIQ IQUES UES • WEBMAP WEBMAP DEVELO DEVELOPME PMENT NT • CASE CASE HIST HISTOR ORIE IES S • DEMON DEMONSTR STRATI ATION ONS: S: SOFTWAR SOFTWARE E and HARDWARE 1/10/01
2
A WEB DEFLEC DEFLECTION TION SURVEY IS: • MEASUR MEASURIN ING G CRANK CRANKSHA SHAFT FT WEB WEB FLEXURE AS A FUNCTION OF CRANK ROTATION
1/10/01
3
1/10/01
4
WHY CONDUCT A SURVEY • REALI REALIGNM GNMENT ENT / MAJOR MAJOR REP REPAI AIR R • ROUTI ROUTINE NE MECHAN MECHANIC ICAL AL INSPECTION
1/10/01
5
Repairing Block Separation
Courtesy: Adhesive Services
1/10/01
6
CRANKSHAFT BEHAVIOR
• Large crankshafts are flexible • Can sag 60+ mils from gravity • Crankshaft assumes frame shape
1/10/01
7
FACTORY ASSEMBLY: TCV-16
Courtesy: Clark Bro’s Co.
1/10/01
8
FINAL GRIND
Courtesy: Fusion, Inc. 1/10/01
9
CRANKSHAFT BEHAVIOR Location of max stress
• Bending takes place primarily in the webs! • Web acts like a hinge
Apparent hinge points
1/10/01
• Max stresses occur at the fillets 10
WEB DEFLECTION MECHANICS • If the bearings in a machine are not precisely aligned, the crankshaft webs will move farther apart or closer together as the crankshaft is rotated.
Bearings aligned web deflection zero
1/10/01
Bearings misaligned webs closer together with pin up
Bearings misaligned webs farther apart with pin down
11
GAS PLANT COMPRESSOR
DON’T I WISH!
Courtesy: Philadelphia Resins 1/10/01
12
KING RANCH GAS PLANT TLA-8
1/10/01
13
Clark Cross-section 1/10/01
14
DISTORTION GAGE BETWEEN WEBS
1/10/01
15
THE WEB DEFLECTION IS OVER THE LIMIT !! The Superintendent is going to kill me! Easy, Now, Don’t Panic.
1/10/01
16
GET AN ACTION PLAN ! • Check tools/ Rerun survey. • Suspect bad bearings before suspecting distorted frame. • Check bearing clearances. • Replace bearings as necessary. • A bent crankshaft is rarely the cause of excessive deflections. 1/10/01
17
The Web Deflection Survey • Tools • Techniques • Recording the data (Forms to use) • Units (Inches, Mils, MM, Whatever) • Maximum Limits
1/10/01
18
TOOLS FOR THE SURVEY • INSIDE MICROMETER • STARRETT DISTORTION GAUGE • INDIKON ELECTRONIC GAUGE • LASER ALIGNMENT KIT
1/10/01
19
INSIDE MICROMETER • USED BY DRESSER-RAND • CAN REACH BETWEEN “VEE” RODS • SLOWER AND MORE DIFFICULT THAN USING A DISTORTION GAUGE
1/10/01
20
STARRETT DISTORTION GAUGE • TRADITIONALLY RECOMMENDED • SUBJECT TO SOME UNCERTAINTY • FULLY ADEQUATE
1/10/01
21
DISTORTION GAGE AND EXTENSIONS IN BOX
1/10/01
22
STARRETT TECHNIQUES • USE FACTORY POINTS • AVOID BENT EXTENSIONS • REVIEW ACCURACY PERIODICALLY • USE CARE REGARDING PUNCH MARK INDENTIONS
1/10/01
23
INDIKON ELECTRONIC • VERY HIGH RESOLUTION • SOMEWHAT COMPLICATED • PROVIDES ANGULAR DATA • RECOMMENDED, BUT IS NOT NECESSARY WHEN USING WEBMAP
1/10/01
24
LASER ALIGNMENT KIT • THEORETICALLY EXCELLENT • NOT YET DEVELOPED AS AN APPLICATION • DETECTS OFFSET AS WELL AS ANGULAR MISALIGNMENT
1/10/01
25
FACTORS AFFECTING DATA • Condition of the Tools • Positioning of the Tool • Temperature of the Engine • Temperature of the Instrument • Condition of the Punch Marks • Extensions: Bolts or Counterweights 1/10/01
26
EXTENSIONS • RECORD EXTENSION LENGTHS • DO NOT RECORD “CORRECTED” READINGS AS DATA • CORRECT FOR EXTENSION LATER • WEBMAP “CORRECTS” READINGS
1/10/01
27
Maximum Deflection Limits • Three mils is about the number you will hear for most large engines (10 to 20 inches of stroke) • It is best considered a function of the stroke length • Traditional formula 0.18 mils / inch of stroke (Courtesy: Cooper Energy Services)
1/10/01
28
Recording the Data • Do not use inches • Make written marks in Mils • Use a form for recording data • Employ “circles” for the data, not a table • Show approximate angle for data points
1/10/01
29
1/10/01
30
This Many Inches Deflection? This Many WHAT? • 0.00025
“Inches,” You say?
• 0.0025 Or was it this many? • Uh...Well. It was about a quarter of a mil. I was looking at it!
1/10/01
31
INCIDENTAL DATA • DATE AND TIME • ENGINE SUMP TEMPERATURE • MECHANICAL CONSIDERATIONS • COMPLETE IDENTIFICATION
1/10/01
32
Typical Deflection Readings
As Viewed From Oil Pump End
1/10/01
33
FILE The DATA • Keep a good file on location or at the nearest field office. • Send a copy to the area engineer.
1/10/01
34
WEBMAP DEVELOPMENT • EXXON’S KING RANCH GAS PLANT • ALIGNMENT PROBLEMS • GROUT CREEP • ASSISTED BY RAY JAMES • CONVERTED TO PC BY GMRC 1/10/01
35
THE QUESTION WEBMAP ANSWERS
Courtesy: Cooper B essemer
1/10/01
36
MATHEMATICAL BASIS • ASSUME A HARMONIC FUNCTION • DEVELOP A SIMPLE ANALOGY • INCLUDE ENGINE GEOMETRY • THE RESULTS APPLY DIRECTLY TO THE SHAFT, BUT INDIRECTLY TO THE FRAME
1/10/01
37
1/10/01
38
1/10/01
39
1/10/01
40
1/10/01
41
APPLYING THE THEORY • CHOOSE A ZERO REFERENCE • FIT WEB DEFLECTION DATA TO A SINE FUNCTION • FORMULA FOR CORRECTIONS FOR EXTENSIONS • RELATES ENTIRE SURVEY FOR VERTICAL AND HORZ. PROFILES 1/10/01
42
1/10/01
43
WEBMAP ANALYSIS • SCREENS FIELD DATA • CORRECTS FOR EXTENSIONS • EXTRAPOLATES FOR 360 DEGREES • GENERATES A LONGITUDINAL PROFILE OF THE I N F E R R E D FRAME POSTURE
1/10/01
44
EXAMPLES OF WEBMAP • KVR CASE CONFIRMS CRUCIAL DATA • HRA CASE ILLUSTRATES HEAT EFFECTS • OFFSHORE GMVA DATA CONFIRMS FRAME DISTORTION MEASUREMENTS 1/10/01
45
INGERSOLL-RAND 3400 HP KVR
1/10/01
46
1/10/01
47
1/10/01
48
1/10/01
49
1/10/01
50
1/10/01
51
CRANKSHAFT ALIGNMENT AND WEB DEFLECTIONS • WHY RUN A WEB SURVEY • DOING SURVEYS RIGHT • WHAT WEB SURVEYS REVEAL • THEY WILL ALWAYS BE NECESSARY
BUT... 1/10/01
52
BUT...THE WEB SURVEY • IS ONLY AN I N D I R E C T INDICATOR OF FRAME ALIGNMENT • BEST PRACTICES CALL FOR D I R E C T MEASUREMENTS OF FRAME SETTING POSTURE AND POSITION?
