Tuesday, February 21, 2012
Sunday, February 19, 2012
Nature of Efficient Design
"designs will fail if subjected to overload...that's just the nature of efficient design"
Broken crankshaft at the fillet
Designs may not fail immediately; this is the trap that entices in the consumer--the desire for more power, greater performance than designed. A popular performance modification.
It's delayed failure due to metal fatigue that kills. A seemingly successful patch, a popular engine modification, an antique airplane used for aerobatics, antique air tankers:
- Aloha Airlines Flight 243
- Japan Flight 123 520 people killed
- N2969 Turbo Mallard Wing Breaking Off
- China Airlines Flight CI-611 225 people kilIed.
A flight-proven repair, a proven design, long years of successful service were all present in the accidents listed above. Recently I received an inquiry from someone who was surprised that his crankshaft had broken. In talking to him he proudly listed all of his "performance" upgrades that had been done on his engine. Why was he surprised that his crankshaft broke? The only way it would NOT break is if the engineer who designed the crankshaft was inefficient in his design.
Friday, February 17, 2012
Torque Wrench - Accuracy and Precision
Accuracy and Precision are terms that have very different meanings in the measurement industry whereas in common language there are often used interchangeably with the same meaning. This leads to confusion. Accuracy describes the average whereas precision describes the standard deviation of the average. Precision is often contained by the tolerance. Tolerance is used to define the allowable deviation from average (the amount of acceptable variance, or precision). Confused? An example might help.
To be precise is to hit the same spot every time. Notice on the target the shooter is precise but not accurate.
Now look at the next target. The shooter is accurate (when you average the numbers) but not precise.
The next target is both precise and accurate. Precise is repeatability and accurate hits the spot.
These examples show that we need to know both: accuracy and precision. One or the other gives us little useful information.
This has special meaning for torque wrenches. When the gears wear in a snap style torque wrench the precision gets awful. Lets say we are calibrating your torque wrench and set it to 40 lb-ft. We will click it 10 times and average the 10 readings to arrive at the number we put on your calibration report we send back with the wrench. The first pull and your wrench clicks at 50, next click at 30 and so on until we reach our 10 pulls. We average the values and it comes out to 40. Perfect accuracy!
Your wrench may never have clicked at 40. When you use this wrench you think --aww it's perfectly accurate at 40 which means that if it clicks at 40 it is at 40. But you are wrong because your wrench is not precise. When you use the wrench you are not averaging over 10 pulls -- you get what you get-- so you want your wrench to click at the same value each time (precise) and you want it to click at the correct value(accurate).
So now you know that your torque wrench that advertises itself as accurate to within 5% really means: "On the average I click to within 5% of the set point; however, any one discrete click may occur at any value." Rather worthless information isn't it.
But if the wrench manufacturer also states that are statistically certain that 95% of the time the wrench clicks to within 1 lb-ft. of the set point, then we are more confident that when we use the wrench we will establish the amount of tightening to within the manufacturer's torque tolerance.
To be a little more precise in my writing, calibration shops now use the term uncertainty rather than certainty to define the degree of statistic confidence they have in their reported value. Any modern calibration statement must include an uncertainty value.
To be precise is to hit the same spot every time. Notice on the target the shooter is precise but not accurate.
Now look at the next target. The shooter is accurate (when you average the numbers) but not precise.
The next target is both precise and accurate. Precise is repeatability and accurate hits the spot.
These examples show that we need to know both: accuracy and precision. One or the other gives us little useful information.
This has special meaning for torque wrenches. When the gears wear in a snap style torque wrench the precision gets awful. Lets say we are calibrating your torque wrench and set it to 40 lb-ft. We will click it 10 times and average the 10 readings to arrive at the number we put on your calibration report we send back with the wrench. The first pull and your wrench clicks at 50, next click at 30 and so on until we reach our 10 pulls. We average the values and it comes out to 40. Perfect accuracy!
Your wrench may never have clicked at 40. When you use this wrench you think --aww it's perfectly accurate at 40 which means that if it clicks at 40 it is at 40. But you are wrong because your wrench is not precise. When you use the wrench you are not averaging over 10 pulls -- you get what you get-- so you want your wrench to click at the same value each time (precise) and you want it to click at the correct value(accurate).
So now you know that your torque wrench that advertises itself as accurate to within 5% really means: "On the average I click to within 5% of the set point; however, any one discrete click may occur at any value." Rather worthless information isn't it.
But if the wrench manufacturer also states that are statistically certain that 95% of the time the wrench clicks to within 1 lb-ft. of the set point, then we are more confident that when we use the wrench we will establish the amount of tightening to within the manufacturer's torque tolerance.
To be a little more precise in my writing, calibration shops now use the term uncertainty rather than certainty to define the degree of statistic confidence they have in their reported value. Any modern calibration statement must include an uncertainty value.
Monday, February 13, 2012
Friday, February 03, 2012
Pilot Description of Broken Impulse Coupling Spring
"Aircraft has a 600 rpm drop on left magneto during run-up. Otherwise engine operates smoothly."
Diagnosis: Retarded engine timing caused by broken impulse coupling spring.
Underlying Condition leading to Failure: Corrosion pitting
Prevention: Replace impulse coupling spring on engines that show signs of exterior corrosion and at a maximum of 500 hour intervals or in accordance with manufacturer's recommendations.
Additional Warnings: D-2000 and D-3000 magnetos, broken impulse coupling can lead to severe loss of engine power and forced landing. Replace spring at each annual inspection might be something to consider.
Additional Instructions:
Here are two videos I did that shows you how to wind impulse coupling springs
 |
| 10-52949 Impulse Coupling Spring |
Diagnosis: Retarded engine timing caused by broken impulse coupling spring.
Underlying Condition leading to Failure: Corrosion pitting
Prevention: Replace impulse coupling spring on engines that show signs of exterior corrosion and at a maximum of 500 hour intervals or in accordance with manufacturer's recommendations.
 |
| Red rust caused by condensation collecting in magneto. |
Additional Instructions:
Here are two videos I did that shows you how to wind impulse coupling springs
Bendix S20/S200 Magnetos
Slick Magnetos
Thursday, February 02, 2012
Rapidly and Ruthlessly eliminating "customers"
Something a little different today--business advise.
I always keep a $5.00 bill in my wallet. A $5 has the picture of Lincoln and as we know Lincoln freed the slaves. My $5 serves to remind me to always charged for my professional services -- that's what puts food on my table. Only slaves work without getting paid. If I make a mistake, then of course I take care of it to the best of my ability (not God's ability). I have always tried to provide excellent value and not charge for stuff that I didn't earn; but at the end of the day I'm in it for the money as that is what pays the bills.
In response to an email about trying to please a difficult customer, doing free labor as a "favor" to satisfy that difficult customer and being burned in the process...
I've been in your position before and it is uncomfortable and costly. In this industry one has to exercise care in deciding what airplane and which customers you can adequately service and which ones you cannot. I rapidly and ruthlessly eliminate "customers" that are not profitable to my business. Such a policy has not only proved profitable but also lowers my stress levels. I highly recommend it.
30+ years of running a successful aviation business -- paying every bill on time and never late with a payroll
I always keep a $5.00 bill in my wallet. A $5 has the picture of Lincoln and as we know Lincoln freed the slaves. My $5 serves to remind me to always charged for my professional services -- that's what puts food on my table. Only slaves work without getting paid. If I make a mistake, then of course I take care of it to the best of my ability (not God's ability). I have always tried to provide excellent value and not charge for stuff that I didn't earn; but at the end of the day I'm in it for the money as that is what pays the bills.
In response to an email about trying to please a difficult customer, doing free labor as a "favor" to satisfy that difficult customer and being burned in the process...
I've been in your position before and it is uncomfortable and costly. In this industry one has to exercise care in deciding what airplane and which customers you can adequately service and which ones you cannot. I rapidly and ruthlessly eliminate "customers" that are not profitable to my business. Such a policy has not only proved profitable but also lowers my stress levels. I highly recommend it.
30+ years of running a successful aviation business -- paying every bill on time and never late with a payroll
Subscribe to:
Posts (Atom)